Martin Shaw

New wars of the city: 'urbicide' and 'genocide'

This is a paper in two halves: in the first part, I explore the role of cities in warfare historically as a means of framing the contemporary targeting of cities in war and political violence and the emergence of 'urbicide'. In the second part, however, I try to show how the targeting of urban populations is not something separate from other kinds of violence. I point out that targeting urbanity has gone hand in hand, in key examples, with campaigns against ethnic groups and against rural, peasant populations. I use this insight as a way into a general argument about the classification and understanding of political violence and its relationship to war. I argue that 'urbicide' (like 'ethnocide', 'politicide' and other 'cides' that have been identified) is not a separate phenomenon from genocide, but one of its forms. From this I make a final move in which I argue that genocide itself is not separate from war, but a specific form of war that must be understood together with it. 

War and the city: historical perspective

War is commonly understood as a phenomenon of one form of spatial organisation, the territorial nation-state. And yet in the history of human society, the state was not always nation-based, and in the history of states, control of urban space has often been pivotal to their survival. Even in the era of nation-states, warfare frequently revolved around the holding or capture of cities. Sieges were defining moments.

At the end of the twentieth century, as nation-states fragment and the system of nation-states is partially superseded by global politics, the city has a new centrality in the ‘new wars’ of genocide. This centrality has not been understood, however, because the wars that have followed the Cold War have been misapprehended as ‘ethnic’ conflicts. In this paper I attempt to outline, schematically, the changing historical significance of cities in warfare, and to develop some theses on the new wars as wars of the city.

In the origins of the modern polis, city and state were one. War in Greek civilisation was a struggle of city-states, and although warfare often took place outside the city, the urban centre itself always the ultimate prize. The Roman empire, based on the greatest city-state of all, built fortified cities as centres of its far-flung power. Cities were bulwarks of Roman civilisation, defended against barbarian hordes. The sack of Rome itself symbolised the defeat of its civilisation and the descent into fractured political authority and social precariousness.

For more than a thousand years afterwards, state power remained fragmentary and its borders uncertain, and the city retained a special role in the state. The city in medieval Europe was a fortified space, the only relatively certain territory wherein the writ of rulers ran. The city was the redoubt which the ruler could be reasonably sure of defending when more remote territories fell to an invader, were under the control of rebellious lords, or were plagued by robbers and bandits. The siege remained the ultimate moment of war, when the centre fell.


War and the modern city

Modernity transformed the relationship of cities and states. States became, as Giddens (1985) has put it, ‘bordered power containers’. Within the territories of states, the writ of rulers was consolidated by extended ‘surveillance’. Between states, borders became demarcations of violence: now the whole territory claimed by the state would be defended. The state was defined by nation, not city, and the whole population of the nation, even in border regions - even sometimes beyond the borders - became part of the national ‘defence’ of the state.

So although cities grew enormously as wealth and population expanded, their special military significance changed. The extended cities of modernity surpassed the historic fortified boundaries; the fortifications fell into irrelevance or decay rather than being re-built with each phase of expansion. Many new industrial cities grew from what had previously been insignificant villages. While the gap between city and country remained, and indeed was in some ways intensified in the early phases of industrialism, it was no longer a military border.

This is not to say that cities lost all military significance. Capitals still remained the political and administrative if not military centres of state power, and their capture remained the ultimate symbol of conquest and national survival. The successive falls of Paris, for example, in the Franco-Prussian and the two world wars, epitomised France’s repeated humiliations by German power. The fall of Madrid was a decisive defeat for Spain’s Republic in the Civil War. The defence of Stalingrad, in contrast, was a powerful symbol of Soviet defiance and a sign of Hitler’s ultimate failure.

Industrial cities were also of great strategic significance, as they were the engines of the new industrialised war. The difference was largely that cities were no longer built for military defence, and increasingly they were militarily indefensible. Fortifications were no longer generally fixed structures, and where they were - as in the Maginot Line - they were not around cities. Cities were therefore increasingly open to conquest - and not only to occupation on the ground. The new techniques of total warfare brought special dangers to cities. The tank, developed in the First World War for use in the open battlefields of the Western front, was an instrument of indiscriminate destruction in urban settings, although its mobility could also be hampered in narrow streets. The warplane, which first flew over the trenches in 1914-18, was recognised as an instrument of urban mass terror long before its emblematic use at Guernica in 1937.

The full potential of modern war for urban killing was demonstrated in the dreadfully misnamed ‘strategic bombing’ of Dresden, Hamburg and Tokyo in the latter stages of the Second World War. The atomic bomb completed this new vulnerability of the city: whole conurbations and their populations were destroyed instantaneously at Hiroshima and Nagosaki. At the same time, new and vile forms of urban life were developed in the Nazi genocide. Hitler’s war machine, having first turned Jewish ghettos into grotesque caricatures of city life, then created special cities of death, the extermination camps, for their victims.

The relationship between the ‘deliberate’ Nazi genocide and the ‘strategic’ mass murder of Allied bombing campaigns is interesting. With a single atomic bomb, any city could become an instant Auschwitz. In the nuclear age, the city was no longer so important as the industrial engine or political motor of total war, but urban areas remained strategic targets as NATO and the Warsaw Pact developed computerised maps of doom. In the age of the inter-continental ballistic missile, the capacity to simultaneously destroy all the major centres of urban life became a symbol of the degeneration of war. Nuclear war was no longer simply genocidal, but produced ‘mutually assured destruction’, an ‘exterminism’ (as E. P. Thompson, 1982, called it) which threatened human life as such. In their attempt to prevent this catastrophe, peace movements persuaded municipal councils to declare cities across the world ‘nuclear-free zones’. Fortunately, the threat to cities from nuclear war has remained hypothetical in the period after 1945, although cities have remained highly volunerable to aerial attack, as the 1991 destruction of Baghdad’s infrastructure showed.


The guerrilla threat to the city

An actual threat to many cities has been posed by guerrilla war. Although guerrilla war has been romanticised as a progressive movement for social change - the ‘revolution in the revolution’ as the French theorist Régis Debray (1967) famously called it - it involved a reaction to the classic urban-democratic model of revolution. In nineteenth and early twentieth century revolutions, middle- and working-class city-dwellers challenged authoritarian and aristocratic rulers in order to create democratic forms of power. The socialist and communist traditions initially represented extensions of this model. St Petersburg in 1917 and Barcelona in 1936 both followed the pattern that originated in Paris in 1789, in which democratic urban revolt provided leadership to peasant uprisings. Following the degeneration of soviet democracy into party dictatorship in Russia, the Stalinised Communist parties renounced the democratic urban revolution - in some cases in favour of reformism, in others of guerrilla struggle. Guerrilla war represented an authoritarian form of revolutionary change that has been hostile to the plural, creative dynamics of modern city life.

The guerrilla-based Communist strategy, first fully developed in China, involved a militarised party mobilising peasant support to surround the cities, before entering them as conqueror. Communist parties that achieved power in this way created centralised dictatorships that repressed democratic urban life. The degree of repression varied greatly, from the relative openness and tolerance of Tito’s Yugoslavia to the extreme closure and cultural monolithism of Hoxha’s Albania.

The anti-urban bias of Communist regimes originating in rural wars was evident, however, in some of their most destructive phases. In Mao’s ‘Cultural Revolution’ in the late 1960s, urban intellectuals - artists, teachers, officials - were targeted in officially orchestrated mass violence, and often punished by being deported to the countryside where their bodies and in some cases minds were broken by crude physical labour. Mao’s clients, the Cambodian Khmer Rouge, carried this anti-urbanism to even viler extremes, deporting the entire population of cities and exterminating them in an anti-urban genocide.

In both these cases, the city was seen as a source of moral pollution, to be ‘cleansed’ by contact with the hardworking peasantry. The critique of urban ‘decadence’ had much in common with Nazi hatred of the plural urban culture of Weimar cities. Although genocide is often thought of as carried out by one ethnic or national group against another, it is important to remember that the Cambodian genocide was partially an urbicide - one of its prime motives was the destruction of urban culture and the killing of all those contaminated by it.

In the North, however, with the end of the Cold War and the lifting of the nuclear threat, the city has ceased to be a symbolic place of death. It has started to be seen once more - despite the other dangers of crime and pollution which still threaten it - as a place of life, a centre of the pluralism, diversity and creativity of modern civilisation. Military bases are typically located in rural areas, leaving the city as a demilitarised zone. Although militaries consume the scientific and technical knowledge which are produced in urban universities, these functions are almost hidden in institutions dominated by a civilian ethos.

The contemporary West is therefore a ‘post-military’ society (Shaw, 1991) in which military institutions no longer have a central place. Conscription has been abolished or minimised in most advanced societies. Where non-military service is an easy option, educated urban youth are quick to escape - in Germany, for example, over one-third of young men opt for civilian rather than military service. In professional armed forces, it is rare that recruits come from educated urban elites, although urban workers - blighted by the unemployment produced by industrial decline - have provided disproportionate numbers of soldiers.

Post-military urban culture is still saturated with military symbols, but the cult of heroic armed forces has given way to the fascination of the high-tech weapon in a commercialised ‘armament culture’ (Luckham, 1984). Even real wars are fought, so far as Western urban dwellers are concerned, in remote landscapes encapsulated on their television screens. Michael Mann (1988) called it ‘spectator sport’ militarism: mediated violence in which the modern urban dweller participates, not as a foot-soldier as his or her grandfather did, but as a consumer of images. Viewers know that real violence occurs in a war like the Gulf, as a study I carried out showed (Shaw, 1996, pp.127-38) but they expect that their own sons and brothers will not be killed. Governments calculate their strategic options on the basis of avoiding death to modern urban-dwellers.

Military institutions themselves are becoming ‘civilianised’, adapting to the exigencies of high-tech, low-casualty war. Technical, political and managerial knowledge have become more important. Women and even (in some cases) homosexuals have been accepted into many branches of military power. The prime function of armed force has begun to shift from warfighting to war management (more commonly called peacekeeping and peace enforcement). In this sense, the more advanced Western militaries exhibit a more modern, urban, even semi-democratic ethos. The military sociologist Charles Moscos (2000) identifies the officer of the future as an urbanised ‘soldier-scholar’ - a far cry from the not-so-distant days when many officer corps were often recruited from aristocratic landowners and reflected their social manners and prejudices.


Anti-urbanism in the new wars

While the city becomes demilitarised and conventional military forces urbanised, new forms of violence within and against the city have made it the focus of new wars. In the West, inner-urban areas of large cities and conurbations are often centres of violent crime, but typical violence is more individual than collective. Nevertheless, the era of mass unemployment, often concentrated in some inner- and outer-urban areas, has given rise to more pervasive, often drug-related violence that increasingly takes a gang form. In the United States especially, many areas of cities are battlegrounds for gang violence. This low-grade urban warfare has been both cause and consequence of the middle class flight to the suburbs. In poorer countries, these problems are often even more acute. As urban areas have burgeoned, shanty towns have arisen on the edges of older cities, where the poor and unemployed are concentrated in slum conditions. Both individual and collective violence are often rife.

Gang violence is often dubbed ‘warfare’ and has many similarities with more fully-fledged ‘new wars’ of the 1990s (Kaldor, 1999). Violence between supporters of the African National Congress and Inkatha in the Zulu areas of Natal, which reached its peak during South Africa’s transition to democracy, mobilised young, unemployed men in the townships. Although called ‘township wars’, much of the worst violence was in rural areas, where supporters of the ethnically-based Inkatha were often stronger than the urban, multi-ethnic, civic-nationalist ANC. These smouldering mini-wars of the new South Africa were not specifically urban, and indeed incorporated a definite anti-urban element.

An anti-urban element is a common characteristic of ‘new wars’ even where they mobilise urban discontents. Ethnic-nationalist political movements often depend on a specific rural and small-town hatred of the city. They are often led by urban intellectuals from rural backgrounds, by those who have shared but remain uncomfortable with cosmopolitan urban culture - like the leader of the Serbian genocide in Bosnia, Radovan Karadzic, who came from Montenegro to practise as a pyschiatrist in plural Sarajevo. Their support is often strongest in relatively isolated rural areas - like the ‘Krajina’ strongholds of Serbian nationalism in Croatia and the Herzegovinan fiefdom of Croatian nationalism in Bosnia-Herzegovina.

Pluralist, non-ethnic democratic politics, in contrast, is nearly always strongest in larger urban areas. It is rooted in the character of these areas: education, intermarriage, cultural and media diversity led, for example, to far more people in Sarajevo than in other areas identifying themselves non-ethnically, as Bosnians or Yugoslavs rather than Muslims, Serbs or Croats, in the last Yugoslav census. As the power of nationalist authoritarianism has waned, it has done so first in the large urban areas - both Croatia’s Franjo Tudjman and Serbia’s Slobodan Milosevic annulled democratic local elections in their capital cities, won by the opposition, in their attempts to hang on to power.

I pointed out above a common misunderstanding of genocide as the destruction of one ethnic group by another - a misconception which is clearly disproved by the case of the anti-urban genocide in Cambodia. This mistake has been repeated in many analyses of the wars in Yugoslavia. These, like the Nazi war against the Jews, were explicitly genocidal wars, in which civilian groups were as much, or more, the enemy than opposing states or armies. They were initiated by urban political elites and orchestrated using modern urban mass media. Their political support was rooted, however, in the narrow ethnic traditions of the rural and small-town Serb- and Croat-populated areas, and directed against the cosmopolitan way of life in the large urban areas.

There were in fact two sides to the genocides carried out by Serbian and Croatian forces. On the one hand, there was the ‘ethnic cleansing’ (a term which originates with Serbian nationalists and should not be used as a neutral description). In this, non-Serb (chiefly Muslim and Croat) and non-Croat (Muslim and Serb) populations were forcibly expelled from their homes, land, villages and towns, on grounds of their ethnicity, and ethnically ‘pure’ mini-states were established. On the other hand, the genocidists aimed to destroy plural, multi-ethnic urban communities, which just as much as the other ethnic communities offended against their ethnic-national ideals and were obstacles to their victory. Sarajevo - historically a centre of all the major religions, of co-habitation and tolerance, of education, intellectual and artistic life, of high intermarriage and Yugoslav identity - was anathema to the Serbian and Croatian nationalists.

The two sides to the genocide came together as the ‘ethnic cleansers’ saw the educated urban elite within the enemy communities as their first target in each town and village they conquered. Teachers, officials and other ‘intellectuals’ were the first to be selected for deportation and killing. On the other side, Bosnian (Muslim) nationalism was the only nationalist movement that did not have a genocidal project, mostly tolerated other ethnicities and allowed political pluralism within its territory. The struggle for an independent, plural, democratic Bosnia and the campaign of Bosnian nationalists were, for the most part, the same fight, although Muslim nationalism at times showed the same intolerance and tendency towards persecution as the other ethnic political projects.

While the Bosnian war can be considered a war of the city, in the sense that the viability of plural, democratic urban life was a key point of contention, it is important to emphasise that it was also a war within the city. One of the striking features of the new wars is the economic degeneration that accompanies them. Rooted often in economic failure - the failure of market reform in Yugoslavia, the more comprehensive failures of many African economies - new wars are also wars of privatisation in the basest sense. Even where as in Bosnia the central protagonists were state forces (the ex-Yugoslav army still under Serbian state command, the incipient Croatian and Bosnian armies), a key role was played by private gangs. Raiding parties of extreme Serbian nationalists, led by notorious criminals like Arkan, were unleashed by President Milosevic as the front-runners of genocide. In both Croatia and Bosnia, these ‘unofficial’ warriors often initiated the indiscriminate killing, burning and rape of conquered communities.

Everywhere in the Yugoslav wars, indeed, private gain was in the forefront of genocide. The nationalist killing-gangs looted on a big scale. ‘Weekend warriors’ travelled to Bosnia from Serbia-Montenegro and Croatia, to take pot-shots at helpless Sarajevo residents or to loot abandoned villages, often returning home with someone else’s TV or video recorder. Individuals and families annexed the farms, houses and other property of dispossessed neighbours, just as overall, the nationalist movements annexed whole towns and villages. On the borders between the parties' territories, refugees and aid convoys alike were robbed; it was very common for people to be charged extortionate fees to flee their homes to safer areas.

Within even the most cosmopolitian urban areas, the shortages of war gave rise to huge black markets. At times almost the only viable economic activity was that of criminal gangs. Given the lack of Bosnian military forces at the outset of the war, militia led by notorious underworld figures played a prominent part in Sarajevo’s defence (only later in the war did the state assume real control of a national Bosnian army).

The industrial city, particularly, fell into decay as a functioning economic unit. Most industries ceased to operate. Only tiny fractions of pre-war economic output and employment were maintained. Urban economies were even more damaged than those of at least the safer rural areas, where subsistence and local market farming remained possible. Even cities that were saved from destruction or genocidal occupation fell apart, therefore, under the economic pressures of the new wars. There was large-scale emigration, especially of younger, more educated people, who were replaced by refugees ‘cleansed’ from small towns and villages.

These experiences of new urban warfare close to the heart of Europe have been repeated, under even worse conditions, throughout the Caucasian region of the former Soviet area, and in large parts of Africa. Sarajevo and Tuzla, for all their human and physical degradations, have continued to function as urban communities. The same cannot so easily be said of Mogadishu, where modern urban life more or less disintegrated and the gangs and warlords reigned supreme, or of many other towns in the new zones of war.


Violence against the peasantry

This discussion has shown that throughout history, the city has been at the centre of warfare. The concentration of human populations in larger settlements has provided resources for rulers and would-be rulers to wage war. At the same time, however, it has provided highly vulnerable targets for warfare. In the modern city, which is no longer physically organised for defence, the balance has shifted overwhelmingly in favour of vulnerability. Historically huge populations, whose lives are more dependent than ever before on sophisticated technical systems, are easy targets for the killing-power and physical destructiveness of advanced military technology. Moreover, it is what the city represents that is at stake, as much as the existence of its inhabitants and their physical surroundings.

In this sense, contemporary wars are wars of the city, threatened from within as well as without by social forces which deny the vital impulses of modern urban civilisation. It is tempting, therefore, to emphasise this element of modern warfare as a distinctive phenomenon, 'urbicide'. I shall return in the following section to the definition of this concept and its relationship to others. Here, however, I want to underline the important historical counterpoints to such a line of argument. Very simply, attacks on urban centres and urban values have only ever been one dimension of each phase of warfare. For every city besieged in pre-modern wars, many fields and farms were also devastated. If plural urban centres constitute a particular provocation to contemporary nationalists, so too do populations of the 'enemy' ethnic groups. Anti-urbanism has never been a unique or predominant goal of warfare, but has always been an element in a larger enterprise - in which other aims have usually been in the forefront.

To concretise this argument, we have only to match anti-peasant and anti-rural campaigns, often within the same episodes of violence or by the same perpetrators, to anti-urban cases. For urban rulers, the peasantry has always been an object of mistrust. Among the vast, dispersed peasant populations of historic empires, rulers frequently saw the danger of rebellion. In modern invasions, peasant populations have often not only been raped and pillaged but brutally targeted with concentrated violence. The Japanese armies that perpetrated the Rape of Nanking, China's then capital, also systematically razed the countryside, destroyed villages and slaughtered peasants. (Their response to Mao's idea of the peasantry as the 'sea' in which the Communists swum was the chilling order, 'drain the sea.') The Nazi forces that invaded the Soviet Union in 1941 and flattened its cities also torched village after village, to deny the basis of Soviet resistance. American forces bombed Hanoi, but they napalmed the countryside, obliterating villages and forcing their inhabitants into 'strategic hamlets'.

Alongside these examples of anti-peasant intertwined with anti-urban warfare, even more revealing is the incidence of targeted anti-peasant violence by the totalitarian regimes that epitomise anti-urbanism. Soviet power began, of course, in the time of Lenin and Trotsky, with a struggle to prevail against the mass of the Russian peasantry, during the Civil War. It turned simultaneously against the pluralism and diversity of the urban working class parties and unions, and as centralised bureaucratic power was consolidated, against the intelligentsia too. Stalin then turned back on the peasantry in his infamous 'liquidation of the kulaks' of 1929-30, the 'terror famine' and the forced collectivisation that killed millions and wrecked Soviet agriculture later in the decade. (Conquest 19??)

In China, where the anti-urban bias of Communism has been widely recognised, there were equally destructive anti-peasant policies. Mao Zedong's own terror famine, the 'Great Leap Forward' of 1959-61, is only now becoming as well known as the anti-urban terror of the later Cultural Revolution. Although this was possibly the largest single episode of mass death in the twentieth century, with as many as 30 or even 40 million victims, it was less well known precisely because of the greater ease, for a totalitarian regime, of hiding rural than urban suffering. (Becker, 19??) In Cambodia, too, the Khmer Rouge did not just empty the cities and destroy the intelligentsia: in their attempt to remake the people they also systematically relocated the rural population and wiped out the kin networks of the peasants. Peasant ways of life and beliefs were just as much a target as urban ways and beliefs. The party aimed to destroy all pre-existing social organisation and values. (Kiernan, 19??)

Further complicating these stories is the fact that in all cases, national and ethnic identities were also targeted. It is obvious that the Japanese and German invasions were directed, respectively, against the Chinese and Slav peoples as well as state institutions. But for Stalin, too, the terror famine was also about destroying Ukranian national identity; Mao's campaign was also aimed against Tibetan national consciousness; Pol Pot specifically targeted the Vietnamese population and ethnic minorities.


Urbicide in context

How then do we understand the significance of anti-urbanism in the context of such complex, multiple targeting in war and other state-organised political violence? This question takes us directly into the conceptual minefield that is the current state of classification and understanding of political violence. Campbell and Coward (in this volume) attempt to deal with this question by proposing that 'there is a certain kinship between urbicide and genocide.' This idea of affinity suggests two overlapping or related, but ultimately discrete phenomena. Indeed they suggest a definition of urbicide that emphasises its distinctiveness from genocide, when they propose that it 'entails the destruction of buildings and urban fabric as elements of urbanity. Buildings are destroyed because they are the condition of possibility of urbanity. Since urbanity is constituted by heterogeneity, urbicide comprises the destruction of the conditions of possibility of heterogeneity.'

This definition, centred on the destruction of buildings and 'urban fabric', de-emphasises the central feature of anti-urbanism in sieges like that of Sarajevo: the violence directed at the population as such, in order to destroy their resistance and to undermine the multicultural centres of the independent Bosnian state. It is proposed here, in contrast, that we understand both the destruction of buildings and the targeting of urbanity as elements of genocidal war, alongside the attack on Bosnian statehood and on hostile ethnic national identities (Bosnian Muslim, etc.). All of these are explained by the aim of Serbian nationalists to create an ethnically defined state in much of Bosnia, and their determination to destroy the power and ability to resist of the enemy state and enemy population. Thus urbicide was part of the war which also involved genocide.

How then do we understand the relationships between these? Campbell and Coward suggest that 'the meaning of ‘genocide’ is played out in each and every death, each and every time. Since genocide is enacted in each and every death it expresses a relation between what is destroyed and the meaning of destruction that is other than the simple death of the individual. It is integral to our understanding of ‘genocide’ that we recognise what ‘it’ is that is destroyed, and the meaning of the destruction. In genocide ‘it’ is a member of a national or ethnic group and the destruction has the meaning of the eradication of this group.'

So far, so good. But if we define anti-urbanism in terms of its targeting of urban communities, rather than artificially separating the 'destruction of urban fabric' from the destruction of people and social relations, then the ‘it’ is that is destroyed in urbicide is not so distinct from that destroyed in genocide. On the contrary, to the extent that the Serbian forces aimed simultaneously to destroy both the plural, cosmopolitan and the 'ethnic' Bosnian-Muslim populations and characters of Sarajevo, in their attacks on the city, it is very difficult to separate urbicide and genocide (in the conventional understanding of the latter). It is not so much a question of affinity as of intimate interconnection between these aims, both in Serbian policy and ideology, on the one hand, and in Sarajevan experience, on the other. The perpetrators combine the two ideas in a single policy and course of action; the victims experience the two more or less simultaneously. Who was to say, when someone was shot by a sniper or a Mosque was destroyed by bombardment, whether this represented urbicide or genocide, or both together?

This argument draws our attention to two central features of war and political violence: first, that perpetrators often have multiple targets, and secondly, that victims often experience violence and killing as relatively indiscriminate rather than as the heavily categorised violence of perpetrator ideologies (and legal or academic classification). Military campaigns target organised state and military enemies, but they also often target various civilian groups as enemies. When campaigns are translated into action, they often go beyond the pseudo-rationality of political-ideological targeting. So in any given war or campaign a number of different things may be going on simultaneously. People who are tortured, wounded or killed by armed violence, or who see their homes, towns and symbolic buildings destroyed, do not necessarily know precisely, still less care, which of the goals of political and military leaders is being worked out in their suffering. For them, the violence is often 'senseless'.

The aim of social theory must be, therefore, to grasp the unities of the relationships and processes of violence and destruction, rather than to counterpose different dimensions of these actions/experiences as categorical opposites. This task requires both a general theoretical re-unification of the theory of war and genocide, and specific historical understandings of the relationships in particular episodes. In the remainder of this paper I shall make some proposals, based on the argument in my forthcoming War and Genocide (Shaw, 2003).


War, genocide and their many 'cides'

The major theoretical problem of the burgeoning academic study of genocide is its common categorical separation of genocide from war. This separation goes back to the origins of the international legal definition of genocide, in the 1948 Convention. The victors of the Second World War wished to define the heinous crimes of the principal losers (Nazi Germany and Imperial Japan), in targeting civilian populations as such, but in a way that clearly demarcated them from their own crimes. Genocide was thus defined as certain kinds of 'acts committed with intent to destroy, in whole or in part, a national, ethnical, racial or religious group, as such'. The acts included not only 'Killing members of the group', or 'Causing serious bodily or mental harm' but also 'Deliberately inflicting on the group conditions of life calculated to bring a out its physical destruction', 'Imposing measures intended to prevent births within the group' and 'Forcibly transferring children of the group to another group.' (For the full text, see Roberts and Guelff, 2000, pp.180-84)

Genocide as a legal concept was an extension of the previous prohibitions on the killing of civilians in war, and the recognition that these had been breached in ever more systematic ways. Genocide was one of the three main categories of crime in war that were prosecuted at Nuremberg, alongside 'crimes against humanity' and 'war crimes' as such, and this has remained the pattern in the cases brought before the International Criminal Tribunal for Former Yugoslavia in the last few years. However the concept of genocide as the destruction of a civilian social group as such opened the way to separating genocide from war. This suited the Second World War Allies, who wished to emphasise the difference between the killing of civilians through bombing (Dresden, Hiroshima) and the genocide of peoples (Auschwitz). However as the world war has receded into memory, at the beginning of the twenty-first century there is an even more powerful impetus to separate even its paradigmatic genocide, the Nazi Holocaust, from the war itself. (Although the 'good war' of the Allies is recognised as having ended it.) 'Holocaust' and 'genocide' studies are mushrooming as fields distinct from, and largely unconnected to, war studies.


These tendencies involve considerable distortion of the historical context in which the idea of genocide emerged. The Nazis' 'war against the Jews' (Davidowicz, 19??) was waged alongside, and as part of, their wars against the Slav peoples of Poland and the Soviet Union, as well as their more conventional wars against their state enemies. These were from the start not separate events, but fully intertwined campaigns: the entrapment of Polish Jewry in the ghettos was part of the larger forced dispersal of the Polish population following the 1939 invasion. The organised, large-scale murder of Jews began in the invasion of the USSR alongside the killings of Communists, prisoners of war and other civilians. Only in the second half of the war was the industrial extermination of Jews, Gypsies and others carried out in the camp system, fed by railway from all over militarily occupied Europe.

The international legal concept was sophisticated in that it saw the intentional destruction of a target group as such as the aim of genocide, and recognised that variable extents of actual killing or other physical destruction would accompany this. However the phrase 'in whole or in part' allowed great ambiguity and did not clarify the relationship between killing and genocide. The international concept was also flawed from its inception in specifying 'national, ethnical, racial or religious groups'. As critics pointed out during the drafting, this excluded political groups (and social classes), prime targets of Soviet mass murder (the USSR was of course a major contributor to the drafting process) (Kuper, 198?). Certainly, from a theoretical point of view, there is no principled reason to define the destruction of one kind of social group as a supreme crime and of another as not.

This is a central flaw in the legal idea of genocide and means that for social-scientific use, the concept must be substantially adapted. Two major strategies appear to be available. The first is to develop a more elaborate scheme of classification, in which the notion of genocide as a specific type of mass killing is complemented by many other categories of violence. The second is to expand the meaning of genocide to include violence against all types of social groups, in a larger theorisation of war.

A major example of the first approach is the table developed by Michael Mann (2001) to classify 'The Extent of Cleansing and Violence in Inter-Group Relations'. He cross-tabulates extents of 'cleansing' and violence, ranging from the absence of both (multiculturalism and toleration) to genocide as the most extreme form of both, combining 'premeditated mass killing' with 'total cleansing'. In his schema, certain types of action are classified as types of murderous cleansing: e.g. ethnocide is 'total cleansing' accompanied by 'unpremeditated mass deaths', while politicide and classicide are forms of 'partial cleansing' accompanied by 'premeditated mass killing'.

Mann's table, which altogether has 18 boxes, is an interesting way of separating out different levels and types of violence and expulsion. However it has three major problems. First, the central role it gives to 'cleansing' as a social-scientific category is troubling because it legitimates the ideal of racial or ethnic purification. Second, it presents violence as a question of the relations between social groups, rather than as something perpetrated by organised, armed bodies against largely unarmed civilian populations. (Mann wishes to emphasise the wider social participation in genocides: but this definition obscures the central role of organised political power.) Third, Mann's restriction of genocide to total expulsion combined with premeditated killing legitimates the idea of genocide as a maximum case, separated both from 'unintentional' mass death among ethnic groups and the destruction of other types of groups such as social classes.

Mann's definitions are in accord with the international legal concept of genocide in separating the destruction of ethnic from political and class groups, as separate categories. (They would clearly permit the addition of urbicide as a distinctive type, too.) However although his categories clearly incorporate, ad hoc, the links between war and 'murderous cleansing' - '"Callous" war, civil war & class war & revolutionary projects' and 'Exemplary & civil war repression, systematic reprisals' are two of his categories - this classification obscures some crucial relationships. It presents the 'cides' - genocide, ethnocide, politicide, classicide - all as forms of 'cleansing' rather than of war.

Mann's classification is thus based on several theoretical errors. It presents the expulsion rather than the destruction of a social group as the common characteristic of these various types of political violence. It makes it difficult to see processes of simultaneous destruction of different kinds of social group (ethnic, class, urban, etc.) in a common frame. It makes it difficult to see various 'levels' of destruction as part of the same kind of process, And it makes it more difficult to see how closely the destruction of social groups is related to the destruction of armed enemies in violent conflicts. That is to say it separates types of political violence against civilians from war, although the forms of violence under discussion are generally perpetrated in the context of war or by heavily militarised regimes. (For examples, see Shaw, 2003, Chapter 2, Table 2.2.)

The theoretical alternative that is proposed here is to go back to the understanding of genocide as an extension of the wider illegitimate violence against civilians in war. My approach adopts the core meaning of genocide in the definition, viz. the destruction of social groups by organised, armed actors, centrally involving killing and other physical and mental harm, but rejects the inappropriate secondary specification of the types of social groups (as 'national, ethnical, racial or religious'). In these terms, then, the destruction of classes, political groups and plural urban populations are as much genocide as is the destruction of ethnic groups. Indeed, since these groups are often targeted together, as we have seen in the cases of the Serbians, the Nazis and the Khmer Rouge, it is essential for social understanding to explore the processes that lead to this common victimisation. It is also crucial to explore the relations between these different dimensions in the experience of victims.

Genocide differs from war in general in that in that war is, ideal-typically, the conflict of two (or more) organised, armed groups; genocide is an illegitimate form of war in which an organised, armed group defines a largely unarmed civilian population as its enemy, to be destroyed by force. The link between the two is the tendency for war to degenerate from the contest of armed groups, by the extension of violence to the civilian population linked to the armed enemy. Degenerate war is the use of armed force against a civilian population as the extension of military struggle, e.g. the Allied bombing of civilians in order to pressurise the Axis powers into surrender. It will be evident how narrow is the dividing line between this kind of war on civilians and genocide, where a civilian population (e.g. the Jews) is the enemy as such. (For the definition of war, degenerate war and genocide, see Shaw, 2003, Chapter 2, Table 2.3.)

The matter is confused by the fact that genocide usually occurs in the context of war: so that overall, the Nazis' campaign in 1939-45 and the Serbians' in 1991-95 could accurately be described as genocidal war, combining war against armed enemies with genocide of civilian groups. Civilian groups (Jews, Muslims, urban populations) were enemies in their own rights, to be murderously 'cleansed'; but they were also linked to organised armed enemies. To grasp either the war or the genocide, we need to understand the two as parts of the same process.



This chapter has argued that, while historically cities have always been implicated in wars and specific anti-urban violence has been a feature of recent warfare, it does not make sense to separate urbicide (or other 'cides') from genocide, or indeed genocide from war, in genocidal war. The argument about urbicide has thus led into a general argument about the nature of genocide and its relationship to war. We can only understand any of these types as the way in which intentional violent actions by the perpetrators combined with the experiences of victims in a single process. Urbicide is a form of genocide, the fundamentally illegitimate form of modern war in which a civilian population as such is targeted for destruction by armed force.



Becker, J. Hungry ghosts: China's secret famine. London, John Murray, 1996.

Conquest, R. The harvest of sorrow: Soviet collectivization and the terror famine, London, Weidenfeld and Nicolson,  1986

Davidowicz, L. The War Against the Jews. Harmondsworth, Penguin, 1985

Debray, R. Revolution in the revolution, London, Monthly Review Press, 1967

Giddens, A. The Nation-State and Violence. Cambridge, Polity, 1985.

Kaldor, M. New and Old Wars, Cambridge, Polity, 1999

Kiernan, B. The Pol Pot Regime: Race, Power and Genocide in Cambodia Under the Khmer Rouge, 1975-79. Cambridge, Mass.: Yale University Press, 1996

Kuper, A. Genocide. Harmondsworth, Penguin, 1981

Luckham, Robin, ‘Of Arms and Culture’, Current Research on Peace and Violence, VII, 1, 1-64, 1984

Mann, ‘The Roots and Contradictions of Modern Militarism’, States, War and Capitalism, Blackwell 1988, 166-87

Mann, M., 'The colonial darkside of democracy', the global site, 2001

Moskos, C., Williams, J. and Segal, D., eds. (2000) The Postmodern Military: Armed Forces after the Cold War, New York: Oxford University Press

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

Shaw, M. Post-Military Society. Cambridge, Polity, 1991

Shaw, M. Civil Society and Media in Global Crises: Representing Distant Violence, London, Pinter 1996

Shaw, M. War and Genocide. Cambridge, Polity, 2003.

Thompson, E.P., ‘Notes on Exterminism, the Last Stage of Civilization’, in New Left Review, ed., Exterminism and Cold War, London, Verso 1982


Bibliographic note

For fuller discussions of urbicide, albeit adopting a rather narrower definition than this chapter, see David Campbell's and Martin Coward's chapter in this volume, and Coward's paper, 'Community as Heterogenous Ensemble', Alternatives, 27, 2002, pp.29-66. On genocide, Adam Kuper's Genocide, Harmondsworth, Penguin, 1981, is still a very useful starting point, and three volumes that offer varying conceptual and comparative frameworks are G. Andreopoulos, Genocide: Conceptual and Historical Dimensions, Philadelphia: University of Pennsylvania Press, 1994, F. Chalk and K. Jonasson, The History and Sociology of Genocide: Analyses and Case Studies, Cambridge, Mass., Yale University Press, 1990, and Helen Fein's Genocide: A Sociological Perspective, London: Sage, 1993. The argument for looking at the Nazi genocide in the context of the war is best made in Christopher Browning's The Path to Genocide, Cambridge, Cambridge University Press, 1992. An excellent case study of a multi-targeted 'genocide' is Ben Kiernan's The Pol Pot Regime: Race, Power and Genocide in Cambodia Under the Khmer Rouge, 1975-79. Cambridge, Mass.: Yale University Press, 1996. A good introduction to the study of contemporary wars is Mary Kaldor's New and Old Wars, Cambridge, Polity, 1999. The case for seeing genocide as a form of war is presented in Martin Shaw, War and Genocide, contemporary wars is Mary Kaldor's New and Old Wars, Cambridge, Polity, 2003.



The first part of this paper was originally published in Spanish as 'Nueva guerras urbanas', Dos, dos: revista sobre las cuidades, 2, Vallidolid, Spain, pp 67-75, 1997, and I am grateful to the editors of that journal for first stimulating me to write about the issue. This work then appeared in English as 'New wars of the city' on my personal website, at, 2000. The text has been fully revised, as well as substantially extended, however, in the current chapter.



Martin Shaw is a sociologist of war and Professor of International Relations and Politics, University of Sussex, Brighton, England, and the editor of He is the author of many books on war, the state and global politics, most recently War and Genocide (Polity, Spring 2003).