The security dilemma and ethnicconflict

Barry R. Posen

 

The end of the Cold War has been accompanied by the emergence of nationalist, ethnic and religious conflict in Eurasia. However, the risks and intensity of these conflicts have varied from region to region: Ukrainians and Russians are still getting along relatively well; Serbs and Slovenians had a short, sharp clash; Serbs, Croats and Bosnian Muslims have waged open warfare; and Armenians and Azeris seem destined to fight a slow-motion attrition war. The claim that newly released, age-old antipathies account for this violence fails to explain the considerable variance in observable intergroup relations.

 

The purpose of this article is to apply a basic concept from the realist tradition of international relations theory, 'the security dilemma', to the special conditions that arise when proximate groups of people suddenly find themselves newly responsible for their own security. A group suddenly compelled to provide its own protection must ask the following questions about any neighbouring group: is it a threat? How much of a threat? Will the threat grow or diminish over time? Is there anything that must be done immediately? The answers to these questions strongly influence the chances for war.

 

This article assesses the factors that could produce an intense security dilemma when imperial order breaks down, thus producing an early resort to violence. The security dilemma is then employed to analyse two cases - the break-up of Yugoslavia and relations between Russia and Ukraine - to illustrate its utility. Finally, some actions are suggested to ameliorate the tendency towards violence.

 

THE SECURITY DILEMMA

 

The collapse of imperial regimes can be profitably viewed as a problem of 'emerging anarchy'. The longest standing and most useful school of international relations theory - realism - explicitly addresses the consequences of anarchy - the absence of a sovereign - for political relations among states.1 In areas such as the former Soviet Union and Yugoslavia, 'sovereigns' have disappeared. They leave in their wake a host of groups - ethnic, religious, cultural - of greater or lesser cohesion. These groups must pay attention to the first thing that states have historically addressed - the problem of security - even though many of these groups still lack many of the attributes of statehood.

 

Realist theory contends that the condition of anarchy makes security the first concern of states. It can be otherwise only if these political organizations do not care about their survival as independent entities. As long as some do care, there will be competition for the key to security - power. The competition will often continue to a point at which the competing entities have amassed more power than needed for security and, thus, consequently begin to threaten others. Those threatened will respond in turn.

 

Relative power is difficult to measure and is often subjectively appraised; what seems sufficient to one state's defence will seem, and will often be, offensive to its neighbours. Because neighbours wish to remain autonomous and secure, they will react by trying to strengthen their own positions. States can trigger these reactions even if they have no expansionist inclinations. This is the security dilemma: what one does to enhance one's own security causes reactions that, in the end, can make one less secure. Cooperation among states to mute these competitions can be difficult because someone else's 'cheating' may leave one in a militarily weakened position. All fear betrayal.

 

Often statesmen do not recognize that this problem exists: they do not empathize with their neighbours; they are unaware that their own actions can seem threatening. Often it does not matter if they know of this problem. The nature of their situation compels them to take the steps they do.

 

The security dilemma is particularly intense when two conditions hold. First, when offensive and defensive military forces are more or less identical, states cannot signal their defensive intent - that is, their limited objectives - by the kinds of military forces they choose to deploy. Any forces on hand are suitable for offensive campaigns. For example, many believe that armoured forces are the best means of defence against an attack by armoured forces. However, because armour has a great deal of offensive potential, states so outfitted cannot distinguish one another's intentions. They must assume the worst because the worst is possible.

 

A second condition arises from the effectiveness of the offence versus the defence. If offensive operations are more effective than defensive operations, states will choose the offensive if they wish to survive. This may encourage pre-emptive war in the event of a political crisis because the perceived superiority of the offensive creates incentives to strike first whenever war appears likely. In addition, in the situation in which offensive capability is strong, a modest superiority in numbers will appear to provide greatly increased prospects for military success. Thus, the offensive advantage can cause preventive war if a state achieves a military advantage, however fleeting.

 

The barriers to cooperation inherent in international politics provide clues to the problems that arise as central authority collapses in multiethnic empires. The security dilemma affects relations among these groups, just as it affects relations among states. Indeed, because these groups have the added problem of building new state structures from the wreckage of old empires, they are doubly vulnerable.

 

Here it is argued that the process of imperial collapse produces conditions that make offensive and defensive capabilities indistinguishable and make the offence superior to the defence. In addition, uneven progress in the formation of state structures will create windows of opportunity and vulnerability. These factors have a powerful influence on the prospects for conflict, regardless of the internal politics of the groups emerging from old empires. Analysts inclined to the view that most of the trouble lies elsewhere, either in the specific nature of group identities or in the short-term incentives for new leaders to 'play the nationalist card' to secure their power, need to understand the security dilemma and its consequences. Across the board, these strategic problems show that very little nationalist rabble-rousing or nationalistic combativeness is required to generate very dangerous situations.

 

The Indistinguishability of Offence and Defence

 

Newly independent groups must first determine whether neighbouring groups are a threat. They will examine one another's military capabilities to do so. Because the weaponry available to these groups will often be quite rudimentary, their offensive military capabilities will be as much a function of the quantity and commitment of the soldiers they can mobilize as the particular characteristics of the weapons they control. Thus, each group will have to assess the other's offensive military potential in terms of its cohesion and its past military record.

 

The nature of military technology and organization is usually taken to be the main factor affecting the distinguishability of offence and defence. Yet, clear distinctions between offensive and defensive capabilities are historically rare, and they are particularly difficult to make in the realm of land warfare. For example, the force structures of armed neutrals such as Finland, Sweden and Switzerland are often categorized as defensive. These countries rely more heavily on infantry, which is thought to have weak offensive potential, than on tanks and other mechanized weaponry, which are thought to have strong offensive potential. However, their weak offensive capabilities have also been a function of the massive military power of what used to be their most plausible adversary, the former Soviet Union. Against states of similar size, similarly armed, all three countries would have considerable offensive capabilities - particularly if their infantries were extraordinarily motivated - as German and French infantry were at the outset of World War I, as Chinese and North Vietnamese infantry were against the Americans and as Iran's infantry was against the Iraqis.

 

Ever since the French Revolution put the first politically motivated mass armies into the field, strong national identity has been understood by both scholars and practitioners to be a key ingredient of the combat power of armies.2 A group identity helps the individual members cooperate to achieve their purposes. When humans can readily cooperate, the whole exceeds the sum of the parts, creating a unit stronger relative to those groups with a weaker identity. Thus, the 'groupness' of the ethnic, religious, cultural and linguistic collectivities that emerge from collapsed empires gives each of them an inherent offensive military power.

 

The military capabilities available to newly independent groups will often be less sophisticated; infantry-based armies will be easy to organize, augmented by whatever heavier equipment is inherited or seized from the old regime. Their offensive potential will be stronger the more cohesive their sponsoring group appears to be. Particularly in the close quarters in which these groups often find themselves, the combination of infantry-based, or quasi-mechanized, ground forces with strong group solidarity is likely to encourage groups to fear each other. Their capabilities will appear offensive.

 

The solidarity of the opposing group will strongly influence how each group assesses the magnitude of the military threat of the others. In general, however, it is quite difficult to perform such assessments. One expects these groups to be 'exclusive' and, hence, defensive. Frenchmen generally do not want to turn Germans into Frenchmen, or the reverse. Nevertheless, the drive for security in one group can be so great that it produces near-genocidal behaviour towards neighbouring groups. Because so much conflict has been identified with 'group' identity throughout history, those who emerge as the leaders of any group and who confront the task of self-defence for the first time will be sceptical that the strong group identity of others is benign.

 

What methods are available to a newly independent group to assess the offensive implications of another's sense of identity?3 The main mechanism that they will use is history: how did other groups behave the last time they were unconstrained? Is there a record of offensive military activity by the other? Unfortunately, the conditions under which this assessment occurs suggest that these groups are more likely to assume that their neighbours are dangerous than not.

 

The reason is that the historical reviews that new groups undertake rarely meet the scholarly standards that modern history and social science hold as norms (or at least as ideals) in the West. First, the recently departed multi-ethnic empires probably suppressed or manipulated the facts of previous rivalries to reinforce their own rule; the previous regimes in the Soviet Union and Yugoslavia lacked any systemic commitment to truth in historical scholarship. Second, the members of these various groups no doubt did not forget the record of their old rivalries; it was preserved in oral history. This history was undoubtedly magnified in the telling and was seldom subjected to critical appraisal. Third, because their history is mostly oral, each group has a difficult time divining another's view of the past. Fourth, as central authority begins to collapse and local politicians begin to struggle for power, they will begin to write down their versions of history in political speeches. Yet, because the purpose of speeches is domestic political mobilization, these stories are likely to be emotionally charged.

 

The result is a worst-case analysis. Unless proven otherwise, one group is likely to assume that another group's sense of identity, and the cohesion that it produces, is a danger. Proving it to be otherwise is likely to be very difficult. Because the cohesion of one's own group is an essential means of defence against the possible depredations of neighbours, efforts to reinforce cohesion are likely to be undertaken. Propagandists are put to work writing a politicized history of the group, and the mass media are directed to disseminate that history. The media may either willingly, or under compulsion, report unfolding events in terms that magnify the threat to the group. As neighbouring groups observe this, they do the same.

 

In sum, the military capability of groups will often be dependent on their cohesion, rather than their meagre military assets. This cohesion is a threat in its own right because it can provide the emotional power for infantry armies to take the offensive. An historical record of largescale armed clashes, much less wholesale mistreatment of unarmed civilians, however subjective, will further the tendency for groups to see other groups as threats. They will all simultaneously 'arm' - militarily and ideologically - against each other.

 

The Superiority of Offensive over Defensive Action

 

Two factors have generally been seen as affecting the superiority of offensive over defensive action - technology and geography. Technology is usually treated as a universal variable, which affects the military capabilities of all the states in a given competition. Geography is a situational variable, which makes offence particularly appealing to specific states for specific reasons. This is what matters most when empires collapse.

 

In the rare historical cases in which technology has clearly determined the offence-defence balance, such as World War I, soldiers and statesmen have often failed to appreciate its- impact. Thus, technology need not be examined further, with one exception: nuclear weapons. If a group inherits a nuclear deterrent, and its neighbours do as well, 'groupness' is not likely to affect the security dilemma with as much intensity as would be the case in non-nuclear cases. Because group solidarity would not contribute to the ability of either side to mount a counterforce nuclear attack, nationalism is less important from a military standpoint in a nuclear relationship.

 

Political geography will frequently create an 'offence-dominant world' when empires collapse. Some groups will have greater offensive capabilities because they, will effectively surround some or all of the other groups. These other groups may be forced to adopt offensive strategies to break the ring of encirclement. Islands of one group's population are often stranded in a sea of another. Where one territorially concentrated group has 'islands' of settlement of its members distributed across the nominal territory of another group (irredenta), the protection of these islands in the event of hostile action can seem extremely difficult. These islands may not be able to help one another; they may be subject to blockade and siege, and by virtue of their numbers relative to the surrounding population and because of topography, they may be militarily indefensible. Thus, the brethren of the stranded group may come to believe that only rapid offensive military action can save their irredenta from a horrible fate.4

 

The geographic factor is a variable, not a constant. Islands of population can be quite large, economically autonomous and militarily defensible. Alternatively, they can have large numbers of nearby brethren who form a powerful state, which could rescue them in the event of trouble. Potentially, hostile groups could have islands of another group's people within their states; these islands could serve as hostages. Alternatively, the brethren of the 'island' group could deploy nuclear weapons and thus punish the surrounding group if they misbehave. In short, it might be possible to defend irredenta without attacking or to deter would-be aggressors by threatening to retaliate in one way or another.

 

Isolated ethnic groups - ethnic islands - can produce incentives for preventive war. Theorists argue that perceived offensive advantages make preventive war more attractive: if one side has an advantage that will not be present later and if security can best be achieved by offensive military action in any case, then leaders will be inclined to attack during this 'window of opportunity'.5 For example, if a surrounding population will ultimately be able to fend off relief attacks from the home territory of an island group's brethren, but is currently weak, then the brethren will be inclined to attack sooner rather than later.

 

In disputes among groups interspersed in the same territory, another kind of offensive advantage exists - a tactical offensive advantage. Often the goal of the disputants is to create ever-growing areas of homogeneous population for their brethren. Therefore, the other group's population must be induced to leave. The Serbs have introduced the term 'ethnic cleansing' to describe this objective, a term redolent with the horrors of 50 years earlier. The offence has tremendous tactical military advantages in operations such as these. Small military forces directed against unarmed or poorly armed civilians can generate tremendous terror. This has always been true, of course, but even simple modern weapons, such as machine guns and mortars, increase the havoc that small bands of fanatics can wreak against the defenceless: Consequently, small bands of each group have an incentive to attack the towns of the other in the hopes of driving the people away. This is often quite successful, as the vast populations of war refugees in the world today attest.

 

The vulnerability of civilians makes it possible for small bands of fanatics to initiate conflict. Because they are small and fanatical, these bands are hard to control. (This allows the political leadership of the group to deny responsibility for the actions those bands take.) These activities produce disproportionate political results among the opposing group magnifying initial fears by confirming them. The presence or absence of small gangs of fanatics is thus itself a key determinant of the ability of groups to avoid war as central political authority erodes. Although almost every society produces small numbers of people willing to engage in violence at any given moment, the rapid emergence of organized bands of particularly violent individuals is a sure sign of trouble.

 

The characteristic behaviour of international organizations, especially the United Nations (UN), reinforces the incentives for offensive action. Thus far, the UN has proven itself unable to anticipate conflict and provide the credible security guarantees that would mitigate the security dilemma. Once there is politically salient trouble in an area, the UN may tr£ to intervene to 'keep the peace'. However, the conditions under which peacekeeping is attempted are favourable to the party that has had the most military success. As a general rule, the UN does not make peace: it negotiates cease-fires. Two parties in dispute generally agree to a cease-fire only because one is successful and happy with its gains, while the other has lost, but fears even worse to come. Alternatively, the two sides have fought to a bloody stalemate and would like to rest. The UN thus protects, and to some extent legitimates, the military gains of the winning side, or gives both a respite to recover. This approach by the international community to intervention in ethnic conflict, helps create an incentive for offensive military operations.

 

Windows of Vulnerability and Opportunity

 

Where central authority has recently collapsed, the groups emerging from an old empire must calculate their power relative to each other at the time of collapse and make a guess about their relative power in the future. Such calculations must account for a variety of factors. Objectively, only one side can be better off. However, the complexity of these situations makes it possible for many competing groups to believe that their prospects in a war would be better earlier, rather than later. In addition, if the geographic situation creates incentives of the kind discussed earlier, the temptation to capitalize on these windows of opportunity may be great. These windows may also prove tempting to those who wish to expand for other reasons.

 

The relative rate of state formation strongly influences the incentives for preventive war. When central authority has collapsed or is collapsing, the groups emerging from the political rubble will try to form their own states. These groups must choose leaders, set up bureaucracies to collect taxes and provide services, organize police forces for internal security and organize military forces for external security. The material remnants of the old state (especially weaponry, foreign currency reserves, raw material stocks and industrial capabilities) will be unevenly distributed across the territories of the old empire. Some groups may have had a privileged position in the old system. Others will be less well placed.

 

The states formed by these groups will thus vary greatly in their strength. This will provide immediate military advantages to those who are farther along in the process of state formation. If those with greater advantages expect to remain in that position by virtue of their superior numbers, then they may see no window of opportunity. However, if they expect their advantage to wane or disappear, then they will have an incentive to solve outstanding issues while they are much stronger than the opposition.

 

This power differential may create incentives for preventive expropriation, which can generate a spiral of action and reaction. With military resources unevenly distributed and perhaps artificially scarce for some due to arms embargoes, cash shortages or constrained access to the outside world, small caches of armaments assume large importance. Any military depot will be a tempting target, especially for the poorly armed. Better armed groups also have a strong incentive to seize these weapons because this would increase their margin of superiority.

 

In addition, it matters whether or not the old regime imposed military conscription on all groups in society. Conscription makes arms theft quite easy because hijackers know what to look for and how to move it. Gains are highly cumulative because each side can quickly integrate whatever it steals into its existing forces. High cumulativity of conquered resources has often motivated states in the past to initiate preventive military actions.

 

Expectations about outside intervention will also affect preventive war calculations. Historically, this usually meant expectations about the intervention of allies on one side or the other, and the value of such allies. Allies may be explicit or tacit. A group may expect itself or another to find friends abroad. It may calculate that the other group's natural allies are temporarily preoccupied, or a group may calculate that it or its adversary has many other adversaries who will attack in the event of conflict. The greater the number of potential allies for all groups, the more complex this calculation will be and the greater the chance for error. Thus, two opposing groups could both think that the expected behaviour of others makes them stronger in the short term.

 

A broader window-of-opportunity problem has been created by the large number of crises and conflicts that have been precipitated by the end of the Cold War. The electronic media provide free global strategic intelligence about these problems to anyone for the price of a shortwave radio, much less a satellite dish. Middle and great powers, and international organizations, are able to deal with only a small number of crises simultaneously. States that wish to initiate, offensive military actions, but fear outside opposition, may move quickly if they learn that international organizations and great powers are preoccupied momentarily with other problems.

 

CROATS AND SERBS

 

Viewed through the lens of the security dilemma, the early stages of Yugoslavia's disintegration were strongly influenced by the following factors. First, the parties identified the re-emerging identities of the others as offensive threats. The last time these groups were free of constraint, during World War II, they slaughtered one another with abandon. In addition, the Yugoslav military system trained most men for war and distributed infantry armament widely across the country. Second, the offensive appeared to have the advantage, particularly against Serbs 'marooned' in Croatian and Muslim territory. Third, the new republics were not equally powerful. Their power assets varied in terms of people and economic resources; access to the wealth and military assets of the previous regime; access to external allies; and possible outside enemies. Preventive war incentives were consequently high. Fourth, small bands of fanatics soon appeared on the scene. Indeed, the political and military history of the region stressed the role of small, violent, committed groups; the resistance to the Turks; the Ustashe in the 1930s; and the Ustashe state and Serbian Chetniks during World War II.

 

Serbs and Croats both have a terrifying oral history of each other's behaviour. This history goes back hundreds of years, although the intense Croat-Serb conflict is only about 125 years old. The history of the region is quite warlike: the area was the frontier of the Hapsburg and Turkish empires, and Croatia had been an integral part of the military apparatus of the Hapsburg empire. The imposition of harsh Hungarian rule in Croatia in 1868; the Hungarian divide-and-conquer strategy that pitted Croats and Serbs in Croatia against each other; the rise of the independent Serbian nation-state out of the Ottoman empire, formally recognized in Europe in 1878; and Serbian pretensions to speak for all south Slavs were the main origins of the CroatSerb conflict. When Yugoslavia was formed after World War I, the Croats had a very different vision of the new state than the Serbs. They hoped for a confederal system, while the Serbs planned to develop a centralized nation-state.7 The Croats did not perceive themselves to be treated fairly under this arrangement, and this helped stimulate the development of a violent resistance movement, the Ustashe, which collaborated with the Fascist powers during the 1930s.

 

The Serbs had some reasons for assuming the worst about the existence of an independent Croatian state, given Croatian behaviour during World War II. Ustashe leadership was established in Croatia by Nazi Germany. The Serbs, both communist and non-communist, fought the Axis forces, including the Croats, and each other. (Some Croats also fought in Josef Tito's communist partisan movement against the Nazis.) Roughly a million people died in the fighting - some 5.9% of Yugoslavia's pre-war population.8 The Croats behaved with extraordinary brutality towards the Serbs, who suffered nearly 500,000 dead, more than twice as many dead as the Croats.9 (Obviously, the Germans were responsible for many Serbian deaths as well.) Most of these were not killed in battle; they were civilians murdered in largescale terrorist raids.

 

The Croats themselves suffered some 200,000 dead in World War II, which suggests that depredations were inflicted on many sides. (The non-communist, 'nationalist' Chetniks were among the most aggressive killers of Croats, which helps explain why the new Croatian republic is worried by the nationalist rhetoric of the new Serbian republic.) Having lived in a pre- and post-war Yugoslavia largely dominated by Serbs, the Croats had reason to suspect that the demise of the Yugoslavian Communist Party would be followed by a Serbian bid for hegemony. In 1971, the Croatian Communist Party had been purged of leaders who had favoured greater autonomy. In addition, the historical record of the Serbs during the past 200 years is one of regular efforts to establish an ever larger centralized Serbian national state on the Balkan Peninsula. Thus, Croats had sufficient reason to fear the Serbs.

 

Serbs in Croatia were scattered in a number of vulnerable islands; they could only be 'rescued' by offensive action from Serbia. Such a rescue, of course, would have been enormously complicated by an independent Bosnia, which in part explains the Serbian war there. In addition, Serbia could not count on maintaining absolute military superiority over the Croats forever: almost twice as many Serbs as Croats inhabit the territory of what was once Yugoslavia, but Croatia is slightly wealthier than Serbia.10 Croatia also has some natural allies within former Yugoslavia, especially Bosnian Muslims, and seemed somewhat more adept at winning allies abroad. As Croatia adopted the trappings of statehood and achieved international recognition, its military power was expected to grow. From the Serbian point of view, Serbs in Croatia were insecure and expected to become more so as time went by.

 

From a military point of view, the Croats probably would have been better off postponing their secession until after they had made additional military preparations. However, their experience in 1971, more recent political developments and the military preparations of the Yugoslav army probably convinced them that the Serbs were about to strike and that the Croatian leadership would be rounded up and imprisoned or killed if they did not act quickly.

 

Each side not only had to assess the other's capabilities, but also its intentions, and there were plenty of signals of malign intent. Between 1987 and 1990, Slobodan Milosevic ended the administrative autonomy within Serbia that had been granted to Kosovo and Vojvodina in the 1974 constitution." In August 1990, Serbs in the Dalmatia region of Croatia held a cultural autonomy referendum, which they defended with armed roadblocks against expected Croatian interference.12 By October, the Yugoslav army began to impound all of the heavy weapons stored in Croatia for the use of the territorial defence forces, thus securing a vast military advantage over the nascent armed forces of the republic.13 The Serbian window of opportunity, already large, grew larger. The Croats accelerated their own military preparations.

 

It is difficult to tell just how much interference the Croats planned, if any, in the referendum in Dalmatia. However, Croatia had stoked the fires of Serbian secessionism with a series < t ominous rulings. In the spring of 1990, Serbs in Croatia were redefined as a minority, rather than a constituent nation, and were asked to take a loyalty oath. Serbian police were to be replaced with Croats, as were some local Serbian officials. No offer of cultural autonomy was made at the time. These Croatian policies undoubtedly intensified Serbian fears about the future and further tempted them to exploit their military superiority.

 

It appears that the Croats overestimated the reliability and influence of the Federal Republic of Germany as an ally due to some combination of World War II history, the widespread misperception created by the European media and by Western political leaders of Germany's nearsuperpower status, the presumed influence of the large Croatian emigre community in Germany and Germany's own diplomacy, which was quite favourable to Croatia even before its June 1991 declaration of independence.14 These considerations may have encouraged Croatia to secede. Conversely, Serbian propaganda was quick to stress the German-Croatian connection and to speculate on future German ambitions in the Balkans.15 Fair or not, this prospect would have had an impact on Serbia's preventive war calculus.

 

RUSSIA AND UKRAINE

 

Through the lens of the security dilemma, several important factors in Russian-Ukrainian relations can be identified that suggest that the potential for conflict is not as great as for Yugoslavia. First, the propensity of Russians and Ukrainians to view one another's cohesion as an offensive military threat is slight. A principal stabilizing factor here is the presence of former Soviet nuclear forces in both Russia and Ukraine, which provides each republic with a powerful deterrent. Second, each side's perception of the other's 'identity' is comparatively benign. Third, settlement patterns create comparatively less pressure for offensive action. These three factors reduce the pressure for preventive war.

 

The nuclear forces of the former Soviet Union - both those clearly under Commonwealth (effectively Russian) control and those with a more ambiguous status in Ukraine - have probably helped stabilize Russian-Ukrainian relations. This is because nuclear weapons make it dangerous for either to launch a campaign of violence against the other. Mutual deterrence prevails. In a clash of wills between two nucleararmed states about attacks on minority populations, the state representing the interests of the victims would have more credibility; it would be the defender of the status quo. The potential military consequences of each side's 'groupness' is thus muted.

 

Most of the Soviet nuclear forces came under the control of the Russian Republic, thereby rendering large-scale anti-Russian violence in Ukraine very risky. The presence of large numbers of nuclear weapons on Ukrainian soil gives Ukraine a nuclear 'threat that leaves something to chance'. Although these weapons are believed to remain under the technical control of the Commonwealth (Russian) command structure, military action by Russians against Ukraine could precipitate a Ukrainian attempt to seize these weapons. Given the significant representation of Ukrainians in the Soviet officer and non-commissioned officer corps, it is quite likely that there are many Ukrainians who know a lot about nuclear weapons, making their seizure quite plausible. This would be a novel kind of nuclear crisis, but it would probably be enough of a crisis to produce the prudent behaviour among nuclear powers that existed during the Cold War. An overt nationalist political campaign in Russia for action against Ukraine could also provoke Ukrainian seizure of these weapons.

 

Russian and Ukrainian histories of each other, as well as their past relations, are less terrifying than those found among groups within the former Yugoslavia. There is no record of large-scale Russian-Ukrainian military rivalry and no clear, salient incident of nationalist bloodletting. However, one dangerous historical episode could play a significant role in the development of an anti-Russian, Ukrainian history: the communist war on independent farmers and its concomitant famine in 193032 killed millions.17 If Ukrainians begin to blame the famine on Russians, this would be quite dangerous politically. If, instead, the famine continues to be blamed on a Communist Party headed by a renegade Georgian psychopath, then this experience will cause less trouble. Ukrainian President Leonid Kravchuk, in his public utterances, tends to portray the Bolsheviks, not the Russians, as the culprit.18

 

That the famine has not played a large role in Ukrainian nationalist rhetoric is a good sign, but this event provides potential tinder. Russian nationalists should therefore be very careful how they portray future Russian-Ukrainian relations. If they project a subordinate status for Ukraine, then Ukrainian nationalists will have a strong incentive to portray the famine as a Russian crime in their effort to build cohesion to resist Russian domination. Izvestia reports that Sergei Baburin, leader of the Russian Unity bloc in the Russian parliament, informed the Ukrainian ambassador that 'either Ukraine reunites again with Russia or there will be war'. Such statements will be heard and acted upon in Ukraine.

 

It is difficult for Ukrainian nationalists to argue convincingly that they were exploited by Russia. Ukrainians seem to have achieved at least proportional representation in the Soviet governing and military apparatus. They produced a share of Soviet gross national product (GNP) more than proportional to their share of population, and the kinds of goods they produced suggest that Ukraine enjoyed a fair share of industrial investment. Ukrainian nationalists assert, however, that the Soviet Union extracted substantial economic resources from Ukraine - perhaps as much as half of Ukrainian GNP.

 

Of greater importance, Ukrainian nationalists believe and many scholars agree that both the Russian empire and the later Soviet Union did everything possible to retard the growth of an independent Ukrainian identity and to Russify Ukraine. This experience led to the reassertion of Ukraine's cultural and political identity. Alarmingly, Rukh, the main pro-independence party in Ukraine, has apparently drifted towards a more virulent nationalism, one that portrays Russia and Russians as the enemy.

 

These worrisome signs must be put in context, however. In general, ethnic hatred has not played a great role in Ukrainian efforts to define their state. Initially, both of the large political parties in Ukraine tried to accommodate all groups in the country. There is no record of Ukrainian persecution of resident Russians. The Ukrainians and the Russians living in the eastern part of the country have had amicable relations for a great many years. A majority of Russians voted for Ukrainian independence. There are no reports of Ukrainian nationalist gangs operating against Russians.

 

The history of relations between Russians and Ukrainians is thus conducive to peace. Neither has strong reasons to assume that the other's 'groupness' constitutes a strong offensive threat to its survival. That said, Russian-Ukrainian political history is conducive to Ukrainian mistrust, and the famine is a singular historical episode that could prove problematic.

 

The security situation between the two republics is favourable from a stability standpoint. The 12 million Russians in the Ukraine (who constitute 21% of the population) are not settled in small vulnerable islands; many of the areas of settlement are proximate to each other and to the Russian border. Others are proximate to the Black Sea coast, which may help explain the intensity of the dispute about the ultimate disposition of the Black Sea Fleet. Large numbers of Russians are still to be found in the armed forces of the newly independent Ukraine, complicating any Ukrainian state action against resident Russians. The expulsion of Russians from eastern Ukraine would thus be a tough job for the Ukrainians. Russia is also a nuclear power and thus in a position to make credible threats to protect the safety of its own. In addition, the proximity of many Ukrainian Russians to the border of the Russian Republic would facilitate a conventional rescue operation, should that prove necessary. The fact that Russia has at least three times the population, wealth and probable conventional combat power of Ukraine would favour such a rescue. In sum, Russia is not being forced to take offensive conventional action to protect its nationals in Ukraine. Because Russians can probably protect their brethren in the Ukraine later, they have only limited incentives to solve the problem now.

 

To say that the Russians can protect their brethren, however, is not to say that military intervention in Ukraine would be cheap or safe. The Ukrainians inherited ample stocks of armaments from the Soviet Union; the Ukrainian presence in the Soviet military made fatuous any Russian thoughts of spiriting away this vast quantity of military equipment and guarantees that the Ukrainian military will know how to use the weaponry in its possession.27 Efforts to coerce Ukraine would likely precipitate Ukrainian efforts to seize the nuclear weapons now within its territory. Thus, although Russia clearly has the power to protect Ukrainian Russians in the event of oppression, lacking such a provocation, Russian nationalists would have great difficulty convincing their compatriots that Ukraine is ripe for the picking.

 

Finally, unlike Yugoslavia, external factors reinforce restraint in Russian-Ukrainian relations. Because they are quite close to Western Europe and heavily armed, it is reasonable for Russians and Ukrainians to assume that conflict between the two republics would be condemned by outside powers. Each side has reason to fear being branded the aggressor in such a conflict because the United States and the Europeans lack any deep organic ties to either Russia or Ukraine. Thus, Western diplomacy should encourage even-handedness towards the two parties. Thus far, the West has shown a tendency to patronize the Ukrainians and dote on the Russians; this is a mistake. It would be better for both to believe that whoever was labelled the aggressor in a Russian-Ukrainian conflict could end up earning the enmity of the wealthiest and most powerful coalition of powers in the history of the world.

 

In sum, although there are some danger signs in Russian-Ukrainian relations, the security dilemma is not particularly intense in this case. To the extent that Western powers have an interest in peace between these two powers, efforts should be made to preserve this favourable state of affairs.

 

COMPARISON SUMMARY

 

A brief review of these two cases highlights the factors that favoured war in Yugoslavia and that still favour peace in Russian-Ukrainian relations. This comparison also identifies some early warning indicators that should be monitored regarding Russia and Ukraine.

 

In Yugoslavia, Croats and Serbs found each other's identity a threat because of the primitive military capabilities they could field and the terrible record of their historical relationship. In the Russia-Ukraine case, nuclear weapons mute the conventional competition, making group cohesion less of a military asset. If Ukraine eliminates its nuclear arsenal, as it has pledged to do, it will increasingly come to rely on nationalism to strengthen an army that will only be able to stand against Russia through superior motivation. Eliminating Ukraine's nuclear arsenal will therefore make Russia stronger and Ukraine more nationalistic. This could prove dangerous.

 

In Yugoslavia, Serbs in Croatia were militarily vulnerable, and Serbs in Serbia had only one way to defend them - a speedy, powerful offensive. Russians in Ukraine are less geographically isolated and can be protected in several ways: Russians in Ukraine may be able to defend themselves by virtue of their numbers and their presence in the Ukrainian army; Russia itself could make nuclear threats; and the Russian army will probably maintain a marked quantitative superiority over Ukraine, which would facilitate a counter-offensive rescue operation, should one be needed. Systematic de-Russification of the Ukrainian armed forces, accompanied by a precipitate decline in Russia's military capabilities, would therefore be a sign of trouble in Russian-Ukrainian relations.

 

Although Ukrainians and Russians in the eastern Ukraine do live together, no violent bands have emerged and begun to engage in intercommunal terror. In Yugoslavia, such bands emerged early in the dissolution process. It may be that the Russian presence in the Ukrainian army has helped discourage such developments, or it may be that there are enough lawless places in the former Soviet Union to absorb those prone to violence. Aspiring Croatian and Serbian thugs had no other outlet for their violent inclinations. The appearance of small Russian or Ukrainian terrorist groups could have a powerful incendiary effect on relations between the two republics and would thus indicate trouble.

 

In Yugoslavia, the Serbs had many incentives for preventive war. They outnumbered the Croats by only two to one and enjoyed no economic advantage. The Croats were likely to find allies within the former Yugoslavia. They were also likely to find allies abroad. Serbia was less well placed. Serbia enjoyed privileged access to the spoils of Yugoslavia, so it was initially much more powerful militarily than Croatia. The combination of dependence on an offensive to protect brethren in Croatia, and a temporary but wide military advantage, proved to be too large a temptation to resist.

 

The Russians have few incentives for preventive war. With three times the human and material resources of Ukraine, it is unlikely that the balance of military power will soon shift against them, nor does it seem likely that Ukraine will be better than Russia at finding allies abroad. Ukrainian pledges to become a non-nuclear state make it attractive even for nationalist Russians to postpone aggression until later; making war now would be a risky proposition. If Ukraine's economy recovers much more quickly than Russia's, or if Ukraine finds powerful allies abroad while Russia finds itself isolated, or if Russia begins to fear that endless border wars will tie down many of its forces in the future, Russians might begin to think more about preventive action against Ukraine.

 

Even if many of the factors that currently favour peace change, Russia's possession of nuclear weapons should continue to mute its incentives for defensively motivated, preventive conventional war. It should be noted, however, that nuclear powers had a tendency to solve security problems conventionally - when they could - during the Cold War.

 

CONCLUSION Three main conclusions follow from the preceding analysis. First, the

 

security dilemma and realist international relations theory more generally have considerable ability to explain and predict the probability and intensity of military conflict among groups emerging from the wreckage of empires.

 

Second, the security dilemma suggests that the risks associated with these conflicts are quite high. Several of the causes of conflict and war highlighted by the security dilemma operate with considerable intensity among the groups emerging from empires. The kind of military power that these groups can initially develop and their competing versions of history will often produce mutual fear and competition. Settlement patterns, in conjunction with unequal and shifting power, will often produce incentives for preventive war. The cumulative effect of conquered resources will encourage preventive grabs of military equipment and other assets.

 

Finally, if outsiders wish to understand and perhaps reduce the odds of conflict, they must assess the local groups' strategic view of their situation. Which groups fear for their physical security and why? What military options are open to them? By making these groups feel less threatened and by reducing the salience of windows of opportunity, the odds of conflict may be reduced.

 

Because the international political system as a whole remains a selfhelp system, it will be difficult to act on such calculations. Outsiders rarely have major material or security interests at stake in regional dis-

 

putes. It is difficult for international institutions to threaten credibly in advance to intervene, on humanitarian grounds, to protect groups that fear for the future. Vague humanitarian commitments will not make vulnerable groups feel safe and will probably not deter those who wish to repress them. In some cases, however, such commitments may be credible because the conflict has real security implications for powerful outside actors.

 

Groups drifting into conflict should be encouraged to discuss their individual histories of mutual relations. Competing versions of history should be reconciled if possible. Domestic policies that raise bitter memories of perceived past injustices or depredations should be examined. This exercise need not be managed by an international political institution; non-governmental organizations could play a role. Discussions about regional history would be an intelligent use of the resources of many foundations. A few conferences will not, of course, easily undo generations of hateful, politicized history, bolstered by reams of more recent propaganda. The exercise would cost little and, therefore, should be tried.28

 

In some cases, outside powers could threaten not to act; this would discourage some kinds of aggressive behaviour. For example, outside powers could make clear that if a new state abuses a minority and then gets itself into a war with that minority and its allies, the abuser will find little sympathy abroad if it begins to lose. To accomplish this, however, outside powers must have a way of detecting mistreatment of minorities.

 

In other cases, it may be reasonable for outside powers to provide material resources, including armaments, to help groups protect themselves. However, this kind of hard-bitten policy is politically difficult for liberal democratic governments now dominating world politics to pursue, even on humanitarian grounds. In addition, it is an admittedly complicated game in its own right because it is difficult to determine the amount and type of military assistance needed to produce effective defensive forces, but not offensive capabilities. Nevertheless, considerable diplomatic leverage may be attained by the threat to supply armaments to one side or the other.

 

Non-proliferation policy also has a role to play. In some cases, nuclear weaponry may be an effective way of protecting the weak from the strong. Russia may behave with considerable restraint towards Ukraine as long as some nuclear weapons remain on Ukrainian territory, vulnerable to Ukrainian seizure. However, once the last weapon is gone, Russian nationalists may become much more assertive.

 

The future balance of power between Ukraine and Russia is less conducive to good relations than the current one, which is the reason Ukrainians have sought Western security guarantees as a quid pro quo

 

for ratifying the Strategic Arms Reduction Talks (START) Treaty, for adhering to the Nuclear Non-Proliferation Treaty and for ridding themselves of nuclear weapons. Absent such guarantees and the measures needed to render them credible, Ukrainians can be expected to prolong the 'transition' phase to the non-nuclear status that they have promised.29 It would be politically difficult for the United States to reverse the arms control initiatives already launched, but it is reasonable to stretch out their implementation. Recent suggestions to accelerate the denuclearization of Ukraine (and Belarus and Kazakhstan), therefore, have it exactly backward.30 The West should hold Ukraine to a steady, proportional withdrawal schedule over the longest period consistent with the prescribed outline of the START I agreement. Some of the benefits of nuclear deterrence could thus be secured during the coming difficult political and economic transition in Russia and Ukraine.

 

It will frequently prove impossible, however, to arrange military assets, external political commitments and political expectations so that all neighbouring groups are relatively secure and perceive themselves as such. War is then likely. These wars will confirm and intensify all the fears that led to their initiation. Their brutality will tempt outsiders to intervene, but peace efforts originating from the outside will be unsuccessful if they do not realistically address the fears that triggered the conflicts initially. In most cases, this will require a willingness to commit large numbers of troops and substantial amounts of military equipment to troubled areas for a very long time.