2

Genocide as a form of war

The conventional definition of genocide can be summarised as the deliberate destruction of a people, principally but not only by means of killing some of its members. Uniquely, this definition of genocide depends heavily on an international legal document. The Genocide Convention (see box 2.1) was agreed in 1948 by the United Nations, which had been established three years earlier by the victors of the Second World War. The term genocide had indeed only recently been coined, in response to the Nazis' extermination of the Jews. Accordingly, it took this as its basic standard, although it defined genocide in terms that would also qualify many less extreme episodes as genocide.

The Convention's definition clearly limited genocide to the deliberate destruction of groups as such. It implicitly drew a particularly sharp line between genocide and other forms of mass killing. It separated genocide from even the most degenerate war, such as the annihilation of civilian populations for strategic reasons. As we have seen, this was what the Allies themselves had practised, and it was about to become the norm of the new nuclear age. It was clearly true that Germany (see Episode IV) and Japan (Episode V) had slaughtered those whom they regarded as inferior races; the Nazis in particular had regarded these groups as distinct enemies whose destruction was desirable in itself. For the Allied governments on the other hand, the destruction of Germans and Japanese as such was not an end in itself. But mass killing of civilians was an intended and desired consequence of the Allied bombing, carried out in order to defeat the Axis powers (Episode 6). This fine line, which separated Allied mass slaughter from Auschwitz, remained sacrosanct in the convention.

The Convention said that genocide was about the destruction of national, ethnic, racial and religious groups. It excluded the annihilation of groups defined by other characteristics such as class or political affiliation - so that Stalin's liquidation of the kulaks (or 'rich' peasants: Episode III) and eastern European political elites could not be counted. But in the same year that the United Nations adopted the genocide convention, it also adopted its Universal Declaration of Human Rights. From a universal human standpoint, it is clearly untenable to lay down that the destruction and mass killing of some sorts of human group (races, nations or religions) should be regarded as a particularly heinous crime, while that of others (classes, professions or political groups) should not. Yet the restrictive international definition gives special status to the former groups. Given that perpetrators of mass killing often target both kinds of group, in some combination, the restriction often makes little sense in practice. Are we to say, for example, that when the Khmer Rouge targeted people because of their Vietnamese origins, they were practising genocide, but when they killed people because of their education or social class, they were not? Such a legalistic distinction would fragment the significance of the Cambodian 'killing fields' and stop us understanding the criminal destruction of the Cambodian people as a single, if multi-targeted, process (as I explain in Episode VII).

The Convention said, too, that genocide involved the destruction of groups in whole or in part, through a variety of means (five were named, of which killing was the first and foremost). This term has been instrumental in most of the ambiguity surrounding the meaning of genocide. On the one hand, it reflects the valid and very important understanding that the process of destroying a group is the same, whether the perpetrators are wholly or partially successful - whether they destroy the entirety, the majority or only a minority of the group. On the other hand it reflects a central uncertainty, in all the literature about genocide, concerning the meaning of the destruction of a group. Moreover the Convention's specification of genocidal means points, apparently, in contradictory directions. The primary criterion of killing points towards an understanding of genocide as the attempt to eliminate members of the group, and indeed another criterion refers to 'physical destruction'. However the secondary criteria, especially the prevention of births and forcible transfer of children to another group, suggest that the destruction of group identity is genocidal even where its individual members survive physically.

Clearly, while the framers specified that 'any' of these acts was genocidal, they understood all the various means as driving towards the destruction of a group as such. So destruction cannot be understood simply in terms of killing people. Not even Hitler succeeded in killing all the Jews, even within occupied Europe, and it is not clear that he concretely intended this before 1941. There was manifest continuity between earlier phases of violence against Jews and the final stage of mass extermination. The intention to destroy the Jews as an economic, political and cultural force in German and European society was manifested much earlier, and there was a sense that systematic killing completed (as 'final solution') a process begun in many other measures, less final and comprehensive in their violence. Moreover the Nazi violence against other groups, such as Poles, was also considered genocidal by the Nuremburg Tribunal even though there was never any intention of mass extermination. Subsequent debate has tended to emphasise the uniqueness of the Holocaust's horror, in comparison to other mass slaughters - even those committed by the Nazis. This has helped to enshrine a maximum concept of genocide as the complete extermination of a group, involving killing of almost all its members, in legal, popular and even academic understanding. This is clearly mistaken.

To make more adequate sense of the idea of destruction in genocide, we need to grasp that it involves three stages:

1. The identification of a social group as an enemy in an essentially military (rather than merely political, economic or cultural) sense, i.e. as a group against which it is justified to use physical violence in a systematic way.
2. The intention to destroy the real or imputed power of the enemy group, including its economic, political, cultural and ideological power, together with its ability to resist this destruction.
3. The actual deployment of violence to destroy the power of the enemy group through killing and physically harming a significant number of its members, as well as other measures.

The Genocide Convention clearly specifies that genocidal acts must be committed with intent to destroy the group. This is also a valid and essential element of any definition. Clearly, in the standard cases of genocide (also I argue in others that don't fit the Convention's criteria) it is central to understanding mass killing that we grasp the intention of powerful, armed forces to destroy the power of an 'enemy' group through violence. Intention does not mean, of course, a clearly stated policy: on the contrary, since genocide is always and everywhere illegitimate, the destruction of civilian groups is almost invariably cloaked in subterfuge and euphemism. Orders are hidden and indirect, and the chain of responsibility needs careful elaboration before criminal liability on the part of leaders can be established. Intention may be inferred, moreover, from patterns of actions that would hardly have occurred were it absent. Where terror and killing are directly inflicted by organized groups, intention is usually relatively easy to establish. Where people are dying from the indirect effects of policies, as in famines (see box 2.2), the question of intentionality is more difficult.

However the criterion of intention also leads, if interpreted one-sidedly, to the danger of taking the standards of the genocidists for reality. Certainly, we cannot explain what happens in genocide without grasping the pepretrator mentalities that define certain social groups as enemies, to be defeated by mass killing and other victimization. And yet genocide, like war, is not just a course of action pursued by one more powerful party, but a clash of social power and experience between two social forces. We should define genocide, therefore, by the experience of the victims as well as the mentality of the genocidists.

The common experience of genocide is of cumulative discrimination of enemies and targets leading to indiscriminate slaughter. For victims, the threat to their social, personal and physical existence is very similar regardless of the particular reasons for it in the minds of their killers. In practice killing has a deeply arbitrary aspect, often appearing 'senseless'. If Nazism is our model of genocide, we need to put alongside the images of clinical selection for the gas chambers the roadside massacres and mass burnings of villagers - often Jews and non-Jews alike - by special commando groups in occupied Soviet territory, which preceded extermination in the camps. From the standpoint of the victims, genocide can be deeply arbitrary. Whether the Nazis were killing you because you were a Jew, a Slav, a Soviet citizen or a Communist, or all or none of these, was not so important as the fact that you were going to be killed. Moreover, all genocidists kill for a range of reasons and at the same time, seemingly, for none in particular. As we shall see later in this book, they target young men because they are potential fighters or producers; old people for the opposite reason, their uselessness; women because of the significance of sexual power. But in the end, killing is killing and tends to lead to more killing. Slaughter can appear utterly casual, arbitrary and senseless. There is (il)logic in terror and murder that surpasses the pseudo-rational discrimination dictated by strategy and political ideology, even as it fulfils them.

Our understanding of genocide needs to grasp how categorical discrimination in the minds of genocidal killers combines with relatively indiscriminate practice and results. We need to recognise how, in the way in which mass killing partially transcends political logic, genocide turns out to be very similar to war. Just as 'strategic' bombing has led to relatively indiscriminate mass slaughter so has 'racial' targeting. It is rarely the case, moreover, that killers focus only on one kind of social target. The logic of terror tends towards the multiplication of enemies. Political enemies and social classes (commonly, educated elites) are often targets of genocidal regimes at the same time as ethnic or national groups. Thus Jews were not only a racial enemy for the Nazis, but represented social privilege and political opposition. At all stages they were only one among the social enemies of Nazism, and not always the most threatened. Since genocide often takes place in the context of interstate war, social enemies are often identified with international state enemies. The Nazis linked the Jews to international finance and Soviet Bolshevism.

These examples draw our attention to the issue of genocidal ideology. This is only implicit in the Convention, in the sense that racism is regarded as an inhuman and anti-scientific doctrine. However it is important to recognise that genocide is generally characterized by pseudo-scientific, irrational and fantastical beliefs. Groups identified as enemies in the minds of perpetrators are often comprised of people who do not recognise themselves as a community, and whose imputed power bears little relationship to reality. Thus the Nazis defined 'the Jews' as a 'powerful' enemy, although they included among them people like people of only part Jewish ancestry who were assimilated in German society and whose wealth and power, such as it was, was hardly connected to Jewishness. Stalin defined kulaks as an enemy of the state, designating certain peasants as 'rich peasants' although they often lived amidst other people described as 'poor peasants' and were often as close to the breadline as the latter. These and other genocides have involved schemes of social classification embedded in fantastical belief-systems, which themselves mutated rapidly according to the exigencies of the political struggles in which the perpetrators were involved.

The many similarities between war and genocide are hardly coincidences. Most genocides take place during or around interstate and/or civil wars. Table 2.3 shows selected major episodes of genocide in the twentieth century. It includes a number of cases that would not qualify as genocide on the conventional international definition, because they were not directed against 'a national, ethnical, racial or religious group, as such'. These are Stalin's 'liquidation of the kulaks', the Nazis' euthanasia programme against the mentally handicapped, Mao's 'Great Leap Forward' and the Indonesian army's massacre of the Communists. It also includes the Cambodian genocide, which would only partially qualify (because it was only partially directed against such groups).

The table shows the social groups against which genocide was committed, and that in most cases there are three central connections to the state and war:

1. The genocidal episode was organized by a state, or a power centre within a state, that can be regarded as the primary perpetrator;
2. The principal organ which carried out the genocide was the army in conjunction with other state organizations such as police, as well as party organizations and paramilitary groups;
3. The genocide took place in the context of a war between the perpetrator-state and organized, armed enemies (often but not always other states).

The table suggests that in some cases, however, genocide was perpetrated outside a direct context of a war with organized armed enemies. In all these cases, however, the perpetrator-states were militarized states with militaristic ideologies, which had recently experienced wars and had strong conceptions of external 'enemies'. They were often also totalitarian regimes, in which there was a strong general tendency to designate internal 'enemies'. They believed, often in irrational ways, that they were threatened by world conspiracies comprising both internal and external enemies. Moreover, military and paramilitary organizations were still prime organs of killing in these episodes.

This contextual evidence is strongly suggestive of connections between war and genocide. It is not sufficient, however, to represent the links between war and genocide as external, causal relations. In no case does war simply cause genocide. If that was the case, there would be far more major genocidal episodes. Rather, it is the case that when armed military force is being extensively used against organized armed enemies, then it is an easier for leaders to take the extraordinary, generally illegitimate steps towards also using armed force against social groups as such. Militaristic and totalitarian ideologies that designate groups as 'enemies' are particularly likely to facilitate plans to destroy social groups as such through armed force. This is best shown by examining particular cases of genocide - as I have done in the sections on different episodes that interleave the main chapters of this book.

This argument suggests that the links between war and genocide are not simply external or causal but are internal to the character of genocide. The simplest way to express this is to say that genocide can best be understood as a form of war in which social groups are the enemies. Table 2.4 shows how genocide can be seen on a continuum from war in general, through degenerate war (which was defined in Chapter 1). The core linkages (and differences) between genocide and other forms of war, shown in this table, can be described as follows:

1. Destroying the power of the enemy. In war in general, the point of organized armed force is to destroy the power of another organized armed enemy together with its ability to resist (Clausewitz's definition). In degenerate total war, the destruction of the organized enemy is extended to include the destruction of the civilian population, but still as a means towards the defeat of the organized enemy. In genocide, organized armed force is used to destroy (social) groups as such (the Convention definition). This means to destroy the social power of a particular group - in economic and cultural as well as political senses - and usually to eliminate or drastically reduce its presence in a particular territory. Destroying the target group's power also involves destroying its ability to resist.
2. Killing. In war, the destruction of an enemy's power involves killing some among its forces (but does not generally involve killing all its forces). In genocide, the deep destruction of the enemy group is defined by the mass killing of some among its members (but does not generally involve killing all of them). In both cases, the extent of killing depends on the precise aims of the parties, the means available and the course of the struggle.
3. Relationship between killing and other means of coercion. War is the extension of politics by other means and typically includes economic, ideological and political coercion alongside military action. In genocide, killing is supplemented by other coercive measures, e.g. discrimination, robbery, expulsion and terror.
4. Nature of the conflict. War in general, including degenerate war, is primarily conflict between two or more organized armed forces, although in degenerate war civilians are targeted as well. Genocide is conflict between organized, armed forces, on the one hand, and civilian populations that are largely unarmed, on the other - although some among the latter may offer armed resistance and the group may be more or less linked to an armed force. Many secondary differences of genocide from other forms of war flow from this difference.
5. Legitimacy. A crucial difference is that war is generally legitimate, and degenerate war - although breaching accepted standards of warfare - masks itself in the general legitimacy of war. But because the enemy in genocide is defined as a civilian group, it is by definition illegitimate. This has all sorts of ramifications: not least that genocide will also tend to be presented as war whenever possible. Although perpetrating genocide usually requires special organizations, it is mostly carried out by or through the general state machine, especially the army, and other established institutions. Thus it generally utilizes the machinery of war, and other organs of state power, and takes place under their cover. It is in the interests of genocidists and genocide deniers to claim the legitimacy of war. The close links between genocide and degenerate war mean that, to identify genocide, it is essential to clarify the difference between legitimacy and illegitimacy within war.
6. Preparation. Like war in general, genocide needs to be understood as a process. Because war is legitimate, long-term war-preparations (e.g. organizing armies, making weapons, advocating military ideas) are generally open and explicit. Genocide is also prepared in complex ways within society, and the institutions and ideas that are mobilized by genocidists are also widely accepted in pre-genocidal periods - indeed they are very much the same institutions and ideas as those mobilized for war. But genocide-preparation, even more than revolution, exists on the margins of social acceptance. The genocidal potentials of armies, ideologies, laws, racism, chauvinism, religious and class hostilities remain, for the most part, abstract. They are not seen as means of genocide. The direct preparation of genocide, in contrast to war, is generally semi-hidden. Therefore the specific machinery of genocide, like the machinery of insurrection in classic revolutions, is usually a novel construction, begun only in the period prior to its use. Indeed, this machinery generally develops in contexts of pervasive violence, linked to war and revolution.
7. Scope. War is a pervasive and complex social institution, so that we refer to a wide range of practices as military, because of their connections with war, without assuming that they are homogenous or that they are simply connected to a maximum case of war like nuclear extermination. Similarly, the broad base of genocide, and the range of its expressions, means that we can also use genocidal as an adjective referring to genocide, without assuming that these are homogenous or involved in group extermination. Genocidal practices may be defined, therefore, as those that treat social groups as enemies whose power and lives may have to be destroyed. Clearly not all hostility towards other social groups, such as racism and chauvinism, is actively genocidal. However such forms of hostility may be seen in certain contexts as pre- or proto-genocidal. These are the kinds of beliefs, ideologies and institutions which are likely to facilitate actively genocidal threats, when states or political movements begin to organize the conditions for widespread violence. In this context, any hostility against enemy groups may threaten genocide, and so pre-genocidal ideas may take on directly genocidal significance. For example, we know with hindsight that the previously abstract genocidal potential of German anti-semitism started to actively threaten genocide, when it was mobilized in the propaganda and street violence of the Nazis.
8. Small-scale episodes. Armed violence may be used against another armed force in a limited context, which does not lead immediately to a fully-fledged war. Likewise such violence may be used against some members of a group, without leading immediately to an overall campaign against the group. Moreover we can see now that such episodes, involving relatively small numbers of killings, often threaten the all-out destruction of a group's power and a much larger-scale slaughter. Kristallnacht, 11 November 1938, the 'night of broken glass' when 91 Jews were killed, is a good example of such a massacre. As a result of such experience we can now read genocidal dangers in other movements, even if large-scale genocide may still be some way off. Genocide may be discerned, therefore, in relatively limited mass killing, short of 'all-out' slaughter. The concept of genocidal massacre has been proposed to cover smaller incidents, which are often a prelude to larger-scale genocide. These are particularly indicative where they are linked to other forms of targeting of given social groups. Understanding their dangers can help to identify the early stages of genocide, and halt the escalation in policies designed to destroy social groups - which may lead to mass killing. Thus in the sense of genocide proposed here, Nazi policy was already becoming genocidal in the 1930s, in relation to Jews - as well as to homosexuals and mentally defective people who were physically attacked in this period. It was clearly so during 1939-41, when Jews were transported across the territory of conquered Poland and forced into ghettos. Not only did maybe half a million Jews and many others die of ill treatment, starvation and disease during this period. These were already policies designed to comprehensively destroy the social power, and existing conditions of social life, of the Jews, well before any comprehensive plans for slaughter were evolved. Moreover, it was not only Jews, but Poles in general, who were targets of genocide in this period. (This multiply genocidal character of Nazi policies was accepted in the Nuremberg trials after the Second World War.)
9. Extent of killing. War can often achieve its ends without wholesale killing of the enemy. Likewise, genocide does not always involve the slaughter of the majority, let alone the entirety, of the 'enemy' group. This is because perpetrators can often achieve the destruction of a group through relatively limited killings, accompanied by other measures. For example, in March 1999 the Serbian regime of Slobodan Milosevic aimed to destroy the Albanian community within Kosovo, primarily by expelling large numbers of its members from the province. In this context, the killing of around 10,000 people, and the forcible eviction and deportation of many others, was more than sufficient to terrorise over a million people out of Kosovo. Many who were not personally forced to leave nevertheless fled on hearing of the murder and terror that was being practised elsewhere in the province.
10. Conflict. War is, by definition, conflict two armed powers. Genocide, likewise, is conflict, and although typically one side is initially unarmed, this may change. Genocide, however successfully perpetrated in the short term, is never uncontested. However unequal the initial struggle between the organized power that wishes to destroy and the group whose power and lives are attacked, in the end that struggle is likely to become more balanced. Groups targeted for destruction cannot but resist - by whatever means they can find. However helpless most of the victims may be, much of the time, there are always passive resisters. There is also usually forcible opposition, sooner or later, either from the population that is directly attacked, or from other forces that see their interests implicated in what is going on. Since genocide usually occurs in the context of international political conflict and wider warfare, victim groups will look for and often find more powerful allies. All-out slaughter may be avoided because genocidal processes are halted. Indeed, genocide is generally ended in war, by more conventional military forces, as Soviet and Anglo-American armies ended the Holocaust. This is another reason why we should understand genocide as a form of war.

What I have argued about the relationships between war and genocide was summarized in Table 2.3, above. This shows the characteristics of genocide as a type of war, and shows how modern degenerate war combines characteristics of war and genocide and serves as a bridge between the two. Of course, as with any typology, characteristics are given in abstract forms: any real war will almost certainly combine the features of more than one type.

I have argued that war, revolution and genocide are closely related modes of organized political violence. In the end, revolution and genocide can be seen as distinct forms of war. The distinctions are often narrow, and they should not be regarded as absolute categorical differences. Nevertheless they have considerable continuing significance.

Between strategic bombing and the Holocaust, for example, there is not a simple difference of war and genocide. But there is a difference: between degenerate total war and genocide as a distinctive form of war in the context of degenerate war. The difference boils down to who was the enemy, and whose power the perpetrators of killing intended to destroy. For the Allies, the prime enemy was still other states; enemy civilians were derivative enemies and their mass killing was (however terribly) incidental to the major goals of the war. There are strong echoes, of course, in the destruction of the 'derivative' enemy, of the genocidal attack on a civilian population as such. But there remained this slim difference: for the Nazis, Jews - and many other social groups - were themselves enemies as such, whose destruction ranked as an aim alongside the destruction of the enemy states that the Nazis also fought.

This distinction is not one between right and wrong. Degenerate war, as well as genocide, clearly involved breaches of all the norms of war on an enormous scale. The importance of the distinction is, first, that it mattered in practice, and especially to potential victims - although we should note that it was less important to actual victims. When the Allies defeated the German and Japanese states, they stopped killing German and Japanese civilians. The Nazis would probably not have stopped killing Jews until they had eliminated virtually all Jews from Europe. Moreover the Allies, through their degenerate war, stopped the open genocide of the Nazis. As I have suggested, genocide has often been halted by the military action of other states or armed groups. These have often perpetrated their own atrocities against civilians during their campaigns. However, the distinction between degenerate war and genocide clearly matters in such situations.

The other importance of the distinction is that it continues to reverberate. Directly genocidal policies have come to dominate the war strategies of many local states and state-like movements in the non-Western world. From Yugoslavia (Episode IX) to Rwanda (Episode X), genocide has become a large part of the practice of 'new wars'. The aim of state elites has been to destroy the power of certain groups, plural urban as well as ethnic communities, and with this the lives of large numbers of their members. In Rwanda, this aim was unprecedentedly open, literally broadcast over the airwaves to simultaneously mobilize perpetrators and threaten victims. In other cases it has been only slightly veiled by more familiar practices of denial.

In contrast, the genocidal implications of degenerate war have stimulated attempts to escape from them, especially in Western military practice. Mainstream war-preparation became exterminist in the preparation of all-out nuclear exchanges, and counter-insurgency war became brutally murderous in Vietnam. They led to powerful disarmament and anti-war movements that posed serious challenges to the legitimacy of warfighting and war-preparation in Western states. But these issues stimulated attempts by the West to renew military power in more limited forms of warfare, with more discriminating technology and strategies that avoid genocidal killing.

The contemporary argument about war is framed, therefore, by a new version of the mid-twentieth century dilemma. In its simplest form, the question appears to be: can new forms of legitimate warfare be means to defeat illegitimate, degenerate war and genocide? We could reframe this categorical opposition as: can one kind of war end another? It leads to further questions: can mainstream military forces genuinely create forms of war, or of military force short of war, in which its degenerative tendency is controlled? How else can genocide, the war of states against social groups, be halted? Underlying these specific questions is the deeper issue of whether the degenerative tendency of war can be overcome within the military tradition, or only by far-reaching political and social change that will gradually render war superfluous.

There are two main answers to these questions from within military thought and practice. One is the model of 'good' local war in non-Western regions, of legitimate defence against genocidal violence, that could re-legitimate warfare there. Some Western analysts, like Michael Ignatieff (1999) argue that reviving the 'warrior's honour' is an important component of restoring order. For them, recreating legitimate warfare is a priority. Support for these arguments is provided by examples of military campaigns, by local states and armies, which have ended some of the most severe genocidal violence. Thus invasions by Vietnam and Tanzania ended genocidal episodes in Cambodia and Uganda, respectively, both in 1979. The military victory of the Rwandan Patriotic Front ended the 1994 genocide. The campaigns of the Croatian and Bosnian armies played a part in ending genocidal Serbian rule in Bosnia in 1995. Likewise the Kosovo Liberation Army played a role, albeit subordinate, in overthrowing Serbian rule in Kosovo in 1999. However many of these forces practiced repressive violence of their own, committing in some cases massacres that may be described as genocidal - if not on the scale of those they defeated. These examples do not, therefore, provide a strong general counter to the perception of the illegitimacy of war.

The second answer comes from developments in the Western way of war. In Kosovo, indeed, the major role in defeating Serbian terror was played by NATO, and this campaign was claimed as 'humanitarian'. But this is only a strong version of the general case that the West has re-invented war in a form that escapes the degeneracy of total war and nuclear weapons. After its disgrace and failure in Vietnam, America faced the implication that war could no longer be successfully used, without unacceptable levels of casualties both among its own soldiers and enemy civilians. These would be reported by television and so de-legitimize the use of force. This led to the idea of the 'Vietnam syndrome', with its stark consequence that the West could no longer afford real wars. The only uses of force that would survive the rigorous new tests of instant media surveillance would be very limited, quick-fix operations, peacekeeping or at most 'peace-enforcement', which would be more like policing than war. This view increasingly became the common sense of Western governments and societies in the late twentieth century. These were 'post-military' societies in which mass military mobilization had been ended (Shaw, 1991).

By early in the new century, however, the West was fighting its third war in little more than a decade. The new global era which opened with the end of the Cold War (1989) had produced the Gulf War (1991), the Kosovo War (1999) and the 'war against terrorism' which opened in Afghanistan (2001-2) and threatened to spread to other regions. In these conflicts, American-led, Western-centred coalitions fought a new kind of war, which represented a serious attempt to escape the Vietnam syndrome. New, more precise bombing appeared to bring war back within the limits of the 'just war' tradition, with 'collateral damage' to civilian lives that was merely 'accidental' and 'proportional' to the advantages of ending or punishing aggression. What I shall call risk-transfer militarism (see Episode XI) appeared to overcome the visible degeneracy of earlier Western warfare. However, since much of the risk was transferred to civilians, there were still strong echoes of degenerate war.

The questions just posed will reverberate throughout this book. In showing how war degenerated and became linked to genocide, I have underlined the huge

question mark that hangs over the legitimacy of war in human affairs. In the twentieth century, war was not just a fact of life. It became a huge problem. Mass killing as a key means of resolving political issues became ever more widespread, and simultaneously increasingly unacceptable. 'War is over', sang John Lennon: this was a statement of historic redundancy, despite the continuing empirical reality of violence. Is this situation changing in the twenty-first century? The attempts to reinvent legitimate forms of war, especially the new Western way of war, must be taken seriously. Do they represent a new, controlled use of force that will play a constructive role in resolving the most serious problems of our time? Or are they dangerous developments, which fail to overcome the degeneracy of war, and could even contribute to disastrous new escalations of violence? These are the challenges that we - worldwide humanity, including the readers of this book - will have to answer in the twenty-first century.

Andreopoulos, G. J., ed. (1994) Genocide: conceptual and historical dimensions. Pittsburgh: University of Pennsylvania Press

Bauman, Z. (1991) Modernity and the Holocaust. Cambridge: Polity

Becker, J. (1996) Hungry Ghosts: China's secret famine. London: John Murray

Chalk, F. and K. Jonassohn (1990) The history and sociology of genocide: analyses and case studies. New Haven: Yale University Press

Charny, I.W., ed. (1999) Encyclopaedia of Genocide. Oxford: ABC-CLIO

Cohen, S. (2001) States of Denial: Knowing about Atrocities and Suffering Cambridge: Polity Conquest, R. (1986) The harvest of sorrow: Soviet collectivization and the terror famine. London: Weidenfeld & Nicolson

Davis, M. (2001) Late Victorian Holocausts: El Nino famines and the making of the Third World. London: Verso

Fein, H. (1993) Genocide: a sociological perspective. Sage 1993

Horowitz, I. L. (1997) Taking lives: genocide and state power. Brunswick, N.J.: Transaction

Ignatieff, M. (1999) The Warrior's Honour. London: Chatto and Windus

Kaldor, M. (1999) New and Old Wars: Organized Violence in a Global Era. Cambridge: Polity

Kuper, L. (1981) Genocide. Harmondsworth: Penguin.

Markusen, E. and D. Kopf (1995) The Holocaust and Strategic Bombing: Genocide and Total War in the Twentieth Century. Boulder: Westview

Roberts, A. and Guelff, R. (2000) Documents on the Laws of War. Oxford: Oxford University Press

Shaw, M. (1991) Post-Military Society. Cambridge: Polity.

Strozier, C.B. and Flynn, M. Genocide, War and Human Survival, Lantham, Md.: Rowman & Littlefield 1996

Michael Ignatieff, The Warrior's Honor: Ethnic War and the Modern Conscience, Chatto and Windus 1998