The Spanish philosopher Álvaro d'Ors, addressing a conference just a few days after the defeat of the Third Reich, began: ʻI belong to a generation that was born under the sign of Mars: a generation that first saw the light when war was laying waste the fields of Europe; which reached adulthood at the splendid moment of religious crusade that was our War of Liberation; and that now confronts the grandiose and tragic spectacle of a universal, total war such as was never seen in any previous century’.Footnote 1 The Spain of 1945 knew nothing of the atrocities that had taken place in Poland. Even so, d'Ors must have been aware that what he called a ʻgrandiose spectacle’ was really a terrible scenario involving the murder, rape, exile and orphaning of millions of people deprived of physical, sexual and material security. It was 1945 and d'Ors – the philosopher of the Spanish Crusade, of legitimate violence, the glorification of just and necessary war – was drawing on his own experience to normalise the extreme violence that had taken over Europe. He was proclaiming the identity of a generation that saw death as productive, destruction as constructive: a generation of war. The generation of fascism, the generation whose used terror as a political weapon, the generation who advocated the expulsion and elimination of the adversary – such was the generation born under the sign of Mars.
The history of interwar Europe can be recounted in terms of homogenisation, confrontation, elimination and expulsion. It was a time of crisis, when attempts to grasp and retain power were accompanied by violence on almost every part of the continent – violence which, from the erasure of the distinction between the military and civil spheres during the First World War to the attempted exterminations and racial hierarchisations during the Second, turned the first half of the twentieth century into the most brutal, bloody – and in consequence, foundational – period in a millennium of European history. The era that began in 1917 and ended in 1949 was a time of warfare, extermination and mass deportation all over Europe. It was also the era in which civil war became a major agent of transformation among European societies. While the first-mentioned aspect is covered by a vast theoretical corpus and a multitude of comparative studies, the second is not. In spite of their great historical importance, civil wars have received less attention than international wars as generators of collective violence.
In this article, therefore, I shall traverse half a century of European history, starting with the rather obvious assumption that, in spite of the general lack of comparative and theoretical analyses of its nature, the historical process that supplied the context for collective violence was, if not predominantly, certainly recurrently, that of internal – civil – war. Civil wars have been universally perceived as the epitome of suffering, cruelty and pain. But in most cases this fact has not been supported by any analysis of why the violence occurred. Initially, scholarly attention was focused on war itself, rather than violence, which was not included among the central elements that define internal conflicts in Europe.Footnote 2 Partly in order to escape this constricting theoretical framework, Stanley G. Payne, in the only major comparative study of civil wars in Europe, has treated them as a multifactorial process that is, to a great extent, reducible to a half a century of continuous conflict between revolution and counter-revolution, from which other elements derive.Footnote 3 This may be accurate, but it is still the case that analyses of the logic of violence in civil wars do not always produce such clear-cut results.
As Stathis Kalyvas has convincingly demonstrated, war can generate violence that was completely unintended by the main actors.Footnote 4 In other words, violence can have its own logic, but it is not suspended in space and/or time. Rather it is determined by the context of the war: civil war reduces the cost of violence because it destroys institutional sanctions. This works both ways: violence may have its own dynamics, which influence the context (rather than the other way around); yet, historically, it is war that generates a framework that favours and multiplies violence.Footnote 5 The two processes are not necessarily mutually exclusive when it comes to the analysis of historical contingencies. The European civil wars of our time, being both national and international, regular and irregular, were waged against armies and against civilians and so dissolved the distinction between combatants and non-combatants. Being mostly total wars, they superimposed political, national and symbolic conflicts upon one another. This multiplication of conflicts helped to deepen and intensify the politics of violence.
I therefore suggest that Europe's internal wars were exceptionally violent for a number of different reasons, including the seizure and control of power, symbolic conflicts, the break-up of communities and the juxtaposition of separate conflicts under the civil war umbrella. Some of these reasons are, of course, military: for example, civil wars are wars of intersecting belligerence. As wars became total wars, this inevitably meant that non-combatants became progressively more involved as part of the state or quasi-state apparatus of war – or as military targets, particularly from the First World War onwards.Footnote 6 There are also political elements, which as generalisations require some nuancing: both revolution and counter-revolution and fascism and anti-fascism are part of the macro-narratives that have fed into analysis of Europe's 1917–1949 internal conflicts in terms of civil war.
Not all internal conflicts, including the civil wars of Russia and Finland, are universally accepted as such. The Spanish Civil War (1936–1939) was obviously just that; but internal conflicts also affected other territories such as Italy, the Balkans, France and Greece up to 1949, sometimes but not always within the context of the world war. There is little previous work on which to base a comparative analysis of these civil wars and the logic of their violence. Nevertheless, this article will attempt to evaluate the usefulness of the concept of ʻcivil war’ in analysing the violence of internal conflict in interwar Europe. War, particularly civil war, is a highly codified form of violence.Footnote 7 Nonetheless the ubiquity of the term means that its application is not merely a historiographical but also a political and cultural act, as well as a statement of identity. Here civil wars are defined as open wars, preceded by a reciprocal declaration of hostilities by parties that previously belonged to the same political unit. From a comparative viewpoint, however, there are certain common elements that define civil war and explain the degree and intensity of its violence.Footnote 8
Revolution and Counter-Revolution
The twentieth century was colossally violent owing to the global agglomeration of multifactorial historical processes, each attended by its own circumstances and within its own context. Whereas collective violence in wars of occupation, ethnic cleansing and genocide has become the domain of ʻgenocide studies’, with its sometimes rather sweeping theories, violence in civil wars has not received the same theoretical or methodological attention. These theories point to nation states as the primary and major location of collective violence in twentieth-century Europe,Footnote 9 analysing their politics from the starting premise that they have been almost entirely murderous, responding to pre-decided motives and plans directed against homogeneous groups of victims identified by some common, usually metaphorical, characteristic.Footnote 10 Nonetheless, as suggested by Kalyvas and (in passing) Christian Gerlach, civil war entails a multiplicity and fragmentation of factors, levels and perpetrators, at both the micro and macro level, which make its violence particularly difficult to analyse through a homogenising lens.Footnote 11 To examine these processes, as Mark Mazower has pointed out, it is essential to start not just with theory but also with historical contingency.Footnote 12
In perspective, the worst practices of collective violence appear to depend on – without being the inevitable outcome of – specific contexts, including open war, civil war and the infiltration of peacetime politics by the logic of war. Europe's great collective massacres all took place under the aegis of war or as the result of warfare as waged against non-combatants.Footnote 13 Indeed, one of the characteristics of contemporary history is the ever-increasing proportion of civilian deaths in war – which is perfectly logical because the dynamics behind these wars aimed to transform the societies in which they took place. In fact, the same dynamics apply to both civil and international wars: conceptualisation of civilians as prime military targets, the proliferating dynamic of revolution versus counter-revolution and the spread of eliminationist ideologies, such as fascism, which glorify violence and death. All these must be ranked among the factors that encouraged the convergence and consolidation of power in twentieth-century Europe in the form of mass violence.Footnote 14 With hindsight it can be seen that the process was also powerfully affected by the modernisation and accumulation arising from industrialisation. Nevertheless, as Gerlach pointed out, all this preparation and accumulation of factors did not inevitably have to lead to collective violence.
More precisely, collective violence takes place in situations where there is conflict and a perceived crisisFootnote 15 over fairly short periodsFootnote 16 that include some decisive moments and are part of contexts such as a coup d’état or an open war. However, it also depends heavily on the reaction that it generates in the states where it takes place – on the nature of their institutions, power relations and economic structures. From a historical perspective, the first of Europe's great twentieth-century civil wars launched the confrontation between revolution and counter-revolution, owing to the expansion of both processes in Europe. From this viewpoint it is difficult to avoid the sort of global and transtemporal macro-interpretation that attributes the intersecting violence of civil war to two grand concepts – revolution and counter-revolution – constantly at war with each other throughout the twentieth century, and which used two kinds of terror, the red and the white, as tools in their struggle for power over, and violent repression of, another grand concept, the people. It may be that all too often such definitions gloss over internal, local or community dynamics; or that the terminology, particularly ʻcounter-revolution’, underestimates the fact that, as stated by the philosopher Joseph de Maistre, ʻcounter-revolution is not the opposite of a revolution, but . . . an opposing revolution’.Footnote 17 We might even agree with Arno Mayer that the dynamics of revolution and counter-revolution can explain the sudden proliferation of European civil wars. This does not necessarily mean, however, that the same dynamics can explain violence in the context of such wars.
In any case, it is highly likely that revolution has received much more attention than war from researchers intent on tracing the entire process back to the events of 1917.Footnote 18 It is true, however, that the historical importance of the Russian civil war, and the way it unleashed political tendencies and energies that were to influence other comparable processes, makes it a turning point in the history of violence in European civil wars. The war between White Russians and Bolsheviks (counter-revolutionaries and revolutionaries according to 1917 criteria), which dragged on until 1923, left no alternative but to locate the seizure and retention of revolutionary power in an armed context, testing the extent to which both sides created a space for political cleansing, repression and the exploitation and/or elimination of the adversary.Footnote 19
The Russian civil war was the first in Europe in which we may (at least if we apply the customary interpretive standards) clearly identify these two grand projects, along with two further categories of central importance to Europe: Red Terror and White Terror. The former was ʻrevolutionary’ and the latter ʻcounter-revolutionary’, and each had its own path to tread, its own complex modes of operation and its own narrative deployments. According to Figes, the Cheka (the Emergency Committee directing the struggle against counter-revolution, sabotage and speculation) ordered about 250,000 executions of ʻenemies of the people’ under the Decree passed on 5 September 1918 for the protection of the Soviet Republic against its class enemies – which also spawned the infamous ʻde-cossackisation’, the disappearance or deportation of some half-million members of a Cossack minority that numbered some three million people, all of them identified as military and class enemies. It is impossible to comprehend the nature of Soviet polities of violence after the 1918 Decree unless they are put in a civil war context. In August 1918 the order to the Vecheka (the All-Russian Extraordinary Commission for Combating Counter-Revolution and Sabotage) to intern all suspect elements, White guards and kulaks opened the door to isolating the enemies of the revolution, resulting in a vast extension of the General Directorate of Internment Camps, more commonly known as the Gulag system (Glavnoe Upravlenie Lagerei), under the aegis of the Cheka, which set up the Chief Directorate of Forced Labour, the GUPR. The latter reported in around 1921 that there were 41,000 forced labourers in concentration camps and 73,000 in eighty-four internment camps. Only two years later the number of internment camps had risen to 355. This inexorable increase must be attributed to internal warfare.
Nevertheless, counter-revolutionary violence was not long in coming, and the Bolsheviks were not its sole target. Jews, too, fell victim to White Russian violence. Arno Mayer estimates the number of killings at between 100,000 and 150,000 in the Ukraine and southern Russia. In the Don province about 45,000 people were executed or hanged, and reprisals against combatants and non-combatants took place in all zones controlled by the Kolchak government, which is thought to have ordered 25,000 executions in the Ekaterinburg province alone.Footnote 20 This was still less than the violence perpetrated by the revolutionariesFootnote 21 and less than the number of combatants who died in battle: 1.2 million Bolsheviks and about 400,000 Whites, not to mention the tens of thousands of peasants who died in revolts and battles against the Red Army and the million or more civilians who died in the east of the former empire. In any case, the figures for the victims of violence, cleansing and purges show how difficult it is to distinguish between death and murder, between civilians who were executed and civilians who died as a consequence of war.
Factors contributing to the high index of violence against non-combatants in the Russian civil war include power, identity, national and international politics and the wider context of an international war – the violence of which was unprecedented in contemporary history. But civil wars are not international conflicts – even if they subsume such conflicts – or dual wars, but rather ʻcomplex and ambiguous processes that favour united action by local and supra-local actors, some civilian and some in the armed forces, whose alliance gives rise to very diverse kinds of violence’. Hence these processes are defined in two dimensions – fragmentation and sovereignty – with the main bones of contention being control, popular support, collaboration and the discouraging of collaboration with the enemy. In such wars violence is greater where sovereignty is fragmented and disputed. The Russian war is a good example; so is the Spanish. In the Finnish civil war of winter–spring 1918, the internal fragmentation of power may have been the main factor. There was no clear casus belli – although there was a recognisable enemy – but the coup d’état, coupled with division within the army and in politics, large-scale mobilisation and the reciprocal call to arms in January 1918 triggered a territorial division and a revolutionary process in areas under Social Democrat control. Needless to say this also triggered violence against internal enemies.Footnote 22
The revolutionaries were defeated but not crushed. Some 6,500 people died in battle, according to statistics quoted by Risto Alapuro; 1,650 people were executed in the Red Terror and 8,400 in the White – this out of a population of three million. After the end of the war and a chaotic retreat by the Reds during the abandonment of Tampere (on which occasion they murdered about 600 people) came the White Terror. Alapuro calculates that no fewer than 5,600 executions were ordered by ad hoc tribunals – about 200 a day. After the war's end another 12,500 died in White prisoner-of-war camps that housed about 82,000 people. In other words, many more people died as a result of policies of violence than died in battle. More died after the war than during it, and the counter-revolution killed considerably more than the revolution. Thus it is not so easy to identify the Finnish civil war as a war of elimination in a context of total civil war.Footnote 23
The wars in Russia and Finland are somewhat fuzzy in outline, but at least the two had an identifiable declaration, opening and ending of hostilities. The case of Hungary is more complex because the boundary between what can and what cannot be defined as civil war becomes blurred as we analyse the anti-communist coup d’état, the White and Red Terrors and the wresting of power from Béla Kun in 1919. Clearly the existence of intersecting violence, while it may imply belligerence on both sides, cannot be the sole explanation for an internal war. Similarly, coups d’état cannot per se be identified as civil wars, insofar as attempts at self-defence by a persecuted opposition cannot be identified as war. A struggle for independence may turn into an internal war, but it does not do so inevitably.Footnote 24 The case of Ireland in 1922 is symptomatic, insofar as if the term ʻcivil war’ is applied to the armed struggle in Ireland it assumes a debatable compromise between two opposing identities – the very two involved in the conflict – while some authors argue that it took about a century to change from covert to open warfare, by way of rebellions, internal and inter-communal conflicts and reciprocal terrorism.Footnote 25
Of course these three concepts – civil war, overt war, covert war – are merely conventional. Covert war, in particular, is problematic from a comparative viewpoint. A civil war cannot be covert, except as a narrative metaphor; therefore, the Irish conflict cannot, strictly speaking, be a civil war. But is there such a thing as a civil war in the strictest sense? By any criteria – the reciprocal violence exercised in superimposed conflicts that were both multifactorialFootnote 26 and multidirectional;Footnote 27 the involvement of non-combatants; the attempts to secure civil backing and the instrumentalisation of historiography based on hermetic and totalising categories such as people, nation or community – the Irish conflict was not a European civil war.Footnote 28 At least, it was not if such war is seen, continental-style, as a dynamic of revolution versus counter-revolution,Footnote 29 or if we subject it to the proviso that there must be open hostility between two claimants to national legitimacy. But, again, it all depends on the definition of civil war used.Footnote 30
After the First World War, the history of Western Europe acquired a complexity that cannot be reduced to the binary schematic of revolution versus counter-revolution. That dualism was complicated by the emergence of fascism as a vehicle for conservative revolution, on the one hand, and the anti-socialist counter-revolution, on the other. In these cases, and in that of Spain, the counter-revolution was not reactive but preventive. Italy in 1922 and Germany in 1933 have often been described in terms of civil war, insofar as the ascent of fascism led to savage repression of revolutionary parties, as in Hungary.Footnote 31 If these conflicts are to be identified as civil wars, the assumption must be that civil war violence of this sort does not require a state of open warfare or a declaration of war. In point of fact, the universally acknowledged ʻmodel’ for civil war – the Spanish civil war of 1936–39 – is the very opposite of a model example. The preventive counter-revolution in Spain is assumed to have generated a reactive revolution and subsequently a civil war.Footnote 32 Moreover, an enormous proportion of its violence occurred in the first few months of the conflict, long before it can be identified as a civil war in the military sense of the term.
The Fascist Era
The Spanish Civil War has been linked to European civil wars and similar conflicts because it was in part a struggle between revolution and counter-revolution, between imagined and collective entities embodied, respectively, in republicans and rebels. However, the Second Spanish Republic before July 1936 was not exactly revolutionary.Footnote 33 What initially took the brakes off the use of violence was not, in the first instance, a revolution but rather a coup d’état.Footnote 34 Thus, although the Spanish counter-revolutionaries almost attained their objective of crushing communism as in Germany or Hungary, they committed the ʻsin’ of ushering in a period that was revolutionary, whereas the regime they were shooting at was not. So even in Spain, the difficulties of applying this model are obvious.
We must, therefore, try another approach, one based on the profound impact that policies of violence had on the non-combatant population of Spain. Displacement and mass violence led to an enormous loss of population, including deaths in battle, murder behind the lines and the exile of republican soldiers and civilians. It was a war of forced displacements and the homogenisation and persecution of minorities whose identity depended on their political stance.Footnote 35 It was the longest of the civil wars that comprise the conventional frame of reference; it was also proportionately the bloodiest. In appropriately assessing its violence, it is crucial not to note how many victims there were in absolute terms but rather how many there were relative to the size of the population. None of Europe's internal wars throughout the twentieth century approached the murderous levels reached in the Spain of 1936: nearly 3 per cent of the population were killed in the ʻred’ zone and over 5 per cent in the ʻblue’ zone.Footnote 36
Rebel violence caused at least 100,000–130,000 deaths by direct violence (political cleansing, occupation of territory), judicial violence, attacks on the civilian population (including the bombing of cities) and extra-judicial murder in prisons or quasi-prisons, including concentration camps and forced-labour camps. About 52,800 of these deaths occurred in the first few months after the coup d’état, many even before Franco had emerged in October 1936 as Head of State and Generalísimo – which throws some doubt on the label ʻfranquista’ which is so often attached to this violence. However, the fragmentary rearguard revolution in places where the coup d’état was unsuccessful – a revolution that took violence as a concomitant of the seizure and exercise of power – killed about 38,000 people during the first few months of the war, out of a total of about 55,000 during the war as a whole. Thus, out of the figure of approximately 185,000 deaths accepted by historians for the period from 1936 to 1948 (the year in which the victors formally announced the cessation of hostilities), about 90,000 were killed in 1936. To put it another way: of the killings that took place through the twelve years of war, the great majority happened in the first six months.
Thus, although the violence was unleashed by the state of war, the ways in which it was practised had their own dynamics. Throughout the Spanish Civil War of 1936–39, the overall proportion of non-combatant victims to combatant victims is something more than half. However, in 1936 the former outnumbered the latter to a hugely disproportionate extent. Wherever the coup d’état was successful, the maintenance of public order was automatically equated with elimination of the opposition, followed immediately by a huge campaign of cleansing or purging (of which the details varied from place to place), ushered in by the proclamation of martial law by local and supra-regional authorities. This was carried out by armed civilians – most notably the Falange, the fascist party which later became the single official party that sustained Franco's regime – and by cleansing squads, or sometimes by the rebel army itself. Although ʻthe enemy’ was not always identified beforehand, everybody knew who he was and who had to be murdered in any particular place. Rebel violence was – quantitatively – mass violence, but it was also selective.
In the Republican zone, the first manifestations of revolution were symbolic rather than relational – once the coup d’état had been foiled, its leaders killed and control of public order transferred to armed parties or trade unionists. As Mary Vincent has pointed out, these early manifestations constituted a war of religion. Violence inspired by hatred of the Catholic Church was, like all the revolutionary violence, concentrated in the first few weeks of the war.Footnote 37 In Tarragona, twenty-eight of the fifty recorded murders in the first fortnight (23 July to 4 August) were of priests or other religious individuals.Footnote 38 Clergy, and anyone else that could be linked with the Church, were one of the primary targets of revolutionary violence – if not the primary target. Historians accept a figure of about 6,800 regular and secular clergy. It was a war of religion, a class war, a national war, a revolutionary war and a struggle for power. Certain areas were harder hit by violence: while the death rate across Spain approached 3 per cent of the population, it doubled in Madrid to 6.8 per cent, which means that about one in every 147 people living in Madrid was killed. The rate in Catalonia was right on the average, at about 2.9 per cent; but in certain places, such as Cervera, the killing rate was above 20 per cent. In Sant Vicenç de Montalt it reached 45 per cent.Footnote 39
Overall, the revolutionary violence of 1936 represented about 80 per cent of the total for the entire war. From August 1936 onwards – from the day after at least thirty political prisoners were murdered in Madrid's ʻmodel’ prison – revolutionary justice was dispensed by people's tribunals. This did not put a stop to extrajudicial violence, including the massacres of Paracuellos del Jarama and Torrejón de Ardoz in November 1936. As José Luis Ledesma has pointed out, ʻfive months into the war, with twenty-seven to come, one in five of the total victims had already been killed’.Footnote 40 In the rearguard zone, the occupation of territory, as part of a total war (which is what this Spanish war was), always involved some degree of direct violence throughout the three remaining years of the war, along with a new logic of repression, recovery, re-education and re-use.Footnote 41 A leading characteristic of civil war violence is its use against ‘fifth-columnists’ – a use which was more intensive in Spain than in any other of Europe's internal wars. Internal enemies are more likely to feature in a civil war – in a single country where the boundaries between the sides are less clearly marked – than in an international war.Footnote 42 This enemy becomes an obsession: he is persecuted, tracked down and eliminated, and his real potential for action is overestimated, turning his elimination into a primary objective.
The efficacy of this violence, both in 1936 and subsequently, is demonstrated by the fact that there was practically no guerrilla warfare by partisans in the rearguard zone. People's tribunals, military tribunals, classification committees, audit offices, concentration camps and forced-labour camps for prisoners, including political prisoners, all contributed to the purging of the political opposition and established model forms of violence that would outlast the Francoist victory and continue unchanged until at least the mid-1940s. Formally, they continued until 1948. This formalisation did not put an end to direct violence, which continued primarily in the form of killings and the cleansing of conquered territory. But it was not the only possible model.Footnote 43 The violence of the Spanish Civil War stands out from that of other European internal wars for many reasons, particularly its nature, percentage and tempo. The violence, killing and murder for identifiable causes, such as those which took place in Finland or Ireland, accelerated towards the end of these conflicts. Over and above the numerous complex realities seen on the ground, a macro-interpretive approach focuses on military-style judgments: violence is visited by the victors on the vanquished as a punishment. In Spain, on the other hand, the killing accelerated in 1936, before the civil war had really started and as the result of a preventive counter-revolutionary coup that unleashed a reactive revolution and a total war. Levels of violence during that year were high owing to the fragmentation of power and the generalised use of violence as a mechanism for appropriating, maintaining and controlling that power. At the time, however, it was also the result of a decision to forego safeguarding present success in favour of developing the society of the future. The main characteristic of violence in 1936 Spain was that it encompassed every aspect of society and could be used to purge it. Moreover, it served to close up cracks in the political and symbolic order that had not been resolved during the years of Republican political reform.
From this point of view, the Spanish Civil War was not exceptional. The notion of civil war has become an analytical tool, used to explain complex conflicts such as those during the Second World War which, although internal, were encouraged by an external invasion. Even where no open warfare took place, some historians have postulated the existence of latent civil wars in order to explain the background to the rise of phenomena such as European fascism. This is still problematic because, as Claudio Pavone has pointed out, when a state fragments under external pressure, the very concept of civil war loses precision and merges with the concepts of national liberation and collaboration.Footnote 44 Nevertheless, as we have seen, exactly the same lack of precision applies to the other processes of European civil war, particularly when there is a merging of complex situations, such as collaboration with a foreign occupying power and guerrilla warfare.
This analytical model fits the internal conflicts in the Balkans and the ʻfirst stage’ of the Greek civil war. The occupation of Greece by the Axis powers (1941–1944) changed the country profoundly, sowing the seeds of civil war and prompting the rapid growth of a communist party that successfully organised one of the strongest resistance movements in occupied Europe, as well as instigating an internal conflict against the right-wing and collaborationist armed factions which began in 1943.Footnote 45 In 1944 the National Liberation Front and its military arm, the Greek People's Liberation Army (EAM-ELAS), came up against a government backed by Britain. After the failure of the Athens uprising the government started a counter-revolutionary campaign, one aim of which was to disarm the paramilitary parties, with mass arrests affecting up to 50,000 members of the communist militias. EAM sources put the number killed at 1,192.Footnote 46
Such was the prologue to the armed confrontation of 1947–49: Europe's last civil war before Yugoslavia fell apart in the 1990s. The Yugoslav war arose from the ashes of a multidirectional internal war encouraged by the Axis occupation, which either set up fascist states, as in Ante Pavelić’s Independent State of Croatia (Nezavisna Država Hrvatska; NDH), or led to occupation and collaboration, as in Milan Nedić’s Serbia. The combination of factors – military, political, national, ethnic, linguistic and religious – and the number of contending factions (Serbia, the NDH, Germany, Italy, Tito's partisans, Mihailović’s Chetniks) delivered up the former Kingdom of Yugoslavia to killings, deportations and cleansings – state against state, state against guerrillas and guerrillas against guerrillas – raising the number of victims (in Biondich's estimation) to nearly a million, the worst perpetrators being the Croatian Ustaše.Footnote 47 The killing of nearly 600,000 Serbs, Muslims and Jews by Pavelić’s Ustaše in Croatia is a paradigm for the analysis of homogenising, eliminationist violence. However, if we treat the Balkan conflict of 1941–45 as a civil war we are likely to underestimate the main factor behind these policies of violence: the fascist occupation.Footnote 48
The Spanish Civil War may be the best known such conflict of the fascist era, but it was by no means the only one. Indeed, we can extend the term to cover the internal conflicts that, over the course of the Second World War, swept first through occupied Europe and then the Axis countries themselves. The Italian war was both an internal conflict and a border war on the southern European frontier of the Third Reich. The conflict of 1943–45, previously seen as a war of resistance to occupation and a partisan struggle, is now being reinterpreted as another civil war, although there has been strong conceptual and political resistance to this shift. Three or four factions (fascists, anti-fascists, Germans, Allies) participated in the fighting, killing and vengeance that characterised the internal war after the armistice of 1943. Once again there was no clear casus belli apart from the fact that the partitioning of the country into two zones (both under foreign occupation, with two self-proclaimed governments and, above all, a powerful partisan army) led to a violent armed confrontation between two claims to legitimacy. As the Italian Social Republic (Repubblica Sociale Italiana; RSI) reverted to so-called sansepolcrismo, a pure and virginal image of pre-regime revolutionary fascism (referring to the square in Milan where fascism was born, San Sepolcro) and multi-directional violence, Italy experienced a kind of internal fascist palingenesis in a context of simultaneous world and civil war.Footnote 49
The Italian war was one against civilians. Claudio Pavone accepts a figure of over 187,000 victims between 1943 and 1945, of whom 120,000 were non-combatants. Violent fascist reprisals – torture, execution and deportation – killed between 10,000 and 15,000 partisans and civilians.Footnote 50 As Toni Rovatti has pointed out, the fascists’ own (obviously distorted) estimates for executions in Italy between October 1943 and April 1945 were something over 1,400 on the Italian side and about 800 on the German side. Although these are the fascists’ own figures, everything indicates that judicial executions were vastly outnumbered by extrajudicial ones, as in every European civil war. If the numbers of extrajudicial executions were proportionate to the judicial ones, the main perpetrators must have been the fascist authorities in the RSI rather than the Germans. It is no coincidence that it was during this period of internal war, occupation and fascist radicalisation that Jews and partisans were deported to labour camps and extermination camps in Eastern Europe.Footnote 51 Nowhere, either in Italy or elsewhere in occupied Europe, were the deportations a unidirectional phenomenon involving only two parties: the process was in part internal, and still more, inter-communal as regards both victims and perpetrators. The de facto civil war and the phenomenon of superimposed wars (national war, civil war, class war) which continued right through the Second World War in Europe and persisted after it – in the split between the National Front (which was the political wing of the armed force the Francs-Tireurset Partisans; FTP) and the Gaullists in France, for example, or between communist partisans and moderate (Catholic) resistance in Italy – is crucial to an understanding of the complexity and multiplicity of the causes of violence in the most practical sense.Footnote 52
The combination of total war, national war, political or class war and war of religion that characterised internal conflicts between 1939 and 1945 explains, or helps to explain, the fact that all these wars were ‘dirty wars’, waged largely against non-combatants, using the techniques and tactics of violence to the detriment of individual rights and for reasons that were substantially supra-individual. The armistices that put an end to the main source of violent eliminationist policies – fascism – did not put a stop to revenge attacks, destruction of property, expulsions or killings. In fact, collective violence in the aftermath of the Second World War was openly vengeful, on the part of communities decimated during the war, Soviet soldiers against their beaten enemies and partisans and guerrillas. The world war provided a context for a series of national wars, each with its own rhythm and its own logic, and each in turn harboured its own logic of violence under the umbrella of the world war.
Other processes that can be analysed as internal conflicts (though they are not the only ones) are the measures taken to purge Europe of fascism – a quintessentially violent political cleansing which not only killed some 10,000 people in Italy and 9,000 in France but also produced the highest rates of generalised political arrest ever seen in Europe.Footnote 53 In Norway 55,000 members of the fascist Nasjonal Samling; NS) were tried and imprisoned; in Holland 200,000 people were put under investigation; most of the 29,000 people incarcerated in France in 1946 were political prisoners, i.e. they were found guilty of collaborating with the fascists. The figures for arrests in Italy were even higher. In Spain it is thought there were at least 180,000. The violence of war, continuing into the aftermath, promoted expulsions and homogenisation. As Mazower argues, the de facto disappearance of minorities as a ʻproblem’ took place in a context that favoured extreme internal violence.Footnote 54 The same could be said of the so-called civil wars on the western borders of the Soviet Union from 1941 to 1947 – the last gasps of the ultra-violent clashes between collaborationist fascism and communism.Footnote 55
The last of the great European civil wars of the first half-century took place in Greece. With hindsight, it can be seen as a turning point from the earlier dynamic and the new dynamic of communism versus anti-communism, ushering in a new logic – the ʻCold War’. Between 1947 and 1949, after the disarming of EAM-ELAS and the repression of communism,Footnote 56 the third phase of the internal conflict would, once again, include national and international factors; the Truman Doctrine and the confrontation between Tito and Stalin were crucial to defeat of the communist guerrillas.Footnote 57 The statistics of its violence, however, seem to have little to do with international politics. As previously said, grassroots logic may have little to do with the higher logic of diplomacy and economics.Footnote 58 The Axis occupation killed some 40,000 civilians, to which the Greek resistance added another 15,000. The counter-revolutionary terror claimed another 3,000 lives and the civil war proper killed 35,000 soldiers, plus about 4,000 civilians killed by the insurgents and 5,000 by government forces.Footnote 59
Conclusion
The Greek war over, the hurricane of civil war did not revisit Europe until many years later. A historian who sets out to analyse forms of violence in European civil wars over the first half of the twentieth century will be forced to consider, among many other factors, the multi-directional nature of the processes that trigger them. The logic of these forms of violence combines local and regional dynamics with general, supranational contexts such as revolution versus counter-revolution, or fascism versus anti-fascism. It links motives, desires, fears and aspirations, from individual experience all the way to government policies.Footnote 60 And, as previously mentioned, it leads to the conclusion that the main factor in any explanation of civil wars is likely to be the superposition of wars on top of wars.
Closer examination of the reasons for such forms of violence requires a detailed exploration of the ideological, cultural, political, economic and identity-related factors at the local, regional, supra-regional, national and supranational level. For this reason, and because of the social and historiographical importance of the topic, any history of civil wars must be a comparative history. It must go beyond mere juxtaposition, beyond generalisations such as slow modernisation, beyond commonplaces such as structural poverty or atemporal idealisations such as inherited domination or ancestral imbalances and beyond interpretations of violence that rely on pathological (madness, sickness) or moral archetypes. It is difficult indeed to think of a kind of war that is so resistant to identification: a war whose very name or denomination (and this is common to all civil wars) is repudiated by all the parties involved, and which combines brief, violent processes such as coups d’état with long-term dynamics such as total war.
The alternative requires an analysis of civil war violence starting with praxis and context: the language it uses, the interpretations (mostly positive) that it attracts and, fundamentally, its logic. Wars of revolution against counter-revolution, as in Russia or Finland, internal wars between fascism and anti-fascism, as in Spain or Italy, and battles between partisans and/or against occupiers and collaborators, as in France, Yugoslavia or Greece, were marked by extreme multiplicity and multi-directionality which affected loyalties, individual actions and attitudes to the enemy. Not all these wars were equally violent, however. The percentage of killings and mechanisms of repression were particularly significant in Finland and even more so (both relatively and absolutely) in the Spanish Civil War, which is the easiest to identify as a result of internal logic and processes. Nevertheless, in spite of internal differences some preliminary conclusions can be drawn.
In a global perspective, genocide and mass murder are not always associated with a state of war, nor is there any reason why they should be. The Ukrainian man-made famine (Holodomor) of 1933 and the killings in Maoist China – including the hundreds of thousands of Tibetans killed in 1950 – were not directly associated with a state of war.Footnote 61 However, in Europe, levels of violence are always considerably lower where there is no war. In Spain the number of political killings, along with other indicators of collective violence such as concentration camps and forced labour, dropped sharply after 1948, at the end of the war that began in 1936. During the long fascist period in Italy, most of the violence (quantitatively speaking) occurred during the Second World War, the civil war and the liberation. The same can be said of Nazi Germany or Pavelic's Croatia. To widen the analysis, although it was eliminationist fascism that stained interwar Europe with the thick ochre of violence, the project for the radical racist hierarchisation of Europe would have been unthinkable without a state of war. Moreover, this reaffirms the porosity of processes and policies of mass violence: after the Shoah, the killing of Soviet political prisoners was the biggest ever mass murder of a particular category of victims by a particular group of perpetrators – the Nazi authorities and the army. In figures, out of 5,700,000-plus prisoners, only about 930,000 survived.Footnote 62 The phenomenon of (international) wars superimposed on (internal) wars accounts for the fact that policies of violence did not end in Europe in 1945, along with the Second World War. To draw too firm a line between two periods on either side of this date is a distortion which prevents us from seeing that the end of violence after the Second World War was not a fact but a process, involving both continuities and discontinuities, with one logic of violence succeeding another some – sometimes abruptly, sometimes progressively – between 1945 and the end of the decade.
The logic of civil war and the logic of civil war violence may not be the same thing, but they are surely interconnected. In many cases, civil war violence is not aimed solely at elimination. Internment camps did not always aim to kill their prisoners but rather to re-educate and exploit them; if public rape and humiliation were visited on left-wing women in the Spanish Civil War, and on female collaborators in France and Italy after the Liberation, the main aims were re-education and expiation. If people were exiled, or forcibly resettled, during or after internal wars, the idea was to get rid of them – physically and symbolically – rather than kill them. The dispute over legitimacy in a civil war means that perpetrators and victims are also participants in symbolic combats. Such wars use violence as a mechanism for the assumption and retention of power at every level; they are also performative elements for transforming society and building the future. In most cases the principal actor is not the state but several para-states competing for power and for control of the administration, armed forces and symbolic capital of the nation. Recent research has shown that while policies of violence need the dynamic of war, they can become independent of it when they are put into practice.Footnote 63 From the perspective of geopolitics, identity or culture, the assumption and retention of power may turn out to be less important to an understanding of civil war violence because they may not both be subject to the same logic. They may well be interrelated, yet potentially interdependent. When we are analysing historical contingencies in all their complexity, what the historian may or may not consider logical is irrelevant.
Following on from this, civil wars become more violent as they become more complex. The Spanish Civil War, the Second World War and the internal conflicts within the latter were civil wars, justified on a national basis, fought over issues of class and religion. They were national wars of independence against an enemy from outside, wars against class enemies, wars against the ghosts of a recent revolutionary past, wars of religion, political wars, international wars, military wars, total wars, wars of territorial occupation. It is this superposition, together with the fact that (as Victor Serge points out) civil war does not recognise non-combatants, that determines the dimension and degree of internal violence. Civil wars decide the hierarchy, and even the appropriation, of a nation's or community's symbolic capital and sense of identity. They are invariably struggles over the future shape of society, which means that they always involve some sort of purification. This is obvious in the case of the English Civil War of the 1640s, the French wars of religion, the civil war that followed the French Revolution and the wars in Russia, Yugoslavia and Italy in the twentieth century. The Spanish Civil War of 1936–39 was the one that combined all the possible wars and fractures: that is the reason why it is commonly considered as the paradigm of all European Civil wars.
At a time when military technology was far ahead of information technology, it was easier to destroy than to understand the enemy, easier to wage war than to use politics as a way of transforming society. This is why war was so popular with political regimes that desired transformation, a new start and a foundation for a new nation. As d'Ors remarked, violence in civil wars was to varying degrees a mechanism for the cleansing and transformation of society. This partly explains why the civil wars – although they inflicted much less death and suffering on both civilians and combatants than world wars – are still seen as the supreme epitome of cruelty and barbarism.
