The term ‘ethnic cleansing’ has acquired general currency with the violent conflicts between communities in the contemporary Balkans. It is often seen as a hateful yet characteristically modern process, the outcome of fanatical chauvinism or warped attempts to create a uniform nation-state. In fact, it is in no way a new phenomenon, and did not require the emergence of nationalism to spring to life. How else did the Anglo-Saxons empty most of England of Celts, banishing them to the western extremities of the island? Or Latins move north into once German lands? Invasion is rarely a matter simply of imposing a new elite; the fate it reserves for earlier inhabitants is usually more drastic. Such was the pattern through Antiquity and the Middle Ages. From the sixteenth century onwards, European expansion involved the constant transfer, confinement or destruction of ‘primitive’ peoples throughout the Americas, in Australia, South Africa, or the Antilles. Again and again, autochthonous populations were reduced—in today’s sanitized vocabulary—to mere ‘ethnic minorities’. Since the Second World War, three spectacular operations of ethnic cleansing have marked the Mediterranean, Middle East and Subcontinent: the partition of India, the creation of Israel and the division of Cyprus.

This trio may have more to teach us than meets the eye. To see why, it is instructive to look at the Balkans themselves. There the conflict between Serbs and Albanians, which led first to massive expulsions from Kosovo of the latter by the former, followed—after NATO intervention—by no less thorough extrusion of the former by the latter, is only to be understood in its regional context. Historically, Kosovo was the core of the Serbian kingdom carved out from the Byzantine Empire in the twelfth century by the Nemanjid dynasty, whose borders extended to what are now Montenegro and Albania. In the thirteenth century, Pec became the seat of the Archbishop of the autocephalous Orthodox Church of Serbia, while the silver and other mines of Kosovo provided much of the wealth of the mediaeval Serbian kingdom. It was there that a—now enlarged—Serbian Empire was destroyed in June 1389 by the Turks at Kosovo Polje, ‘the Field of the Blackbirds’, near Priština, leaving the Ottomans masters of most of Southeast Europe. The Turkish advance into Europe was only finally halted with the relief of Vienna in 1683. Five years later an Austrian counterattack took Belgrade, and the Serbs in Kosovo rose up against their rulers. But when the Habsburg armies were forced to withdraw, they took with them the Patriarch of Pec and 37,000 Serb families, who were resettled in present-day Vojvodina. In 1737 a further Austrian advance, followed by another retreat, led to a Second Migration of Serbs from the Priština region.

In these struggles, religion was at the centre of mutual warfare. As Noel Malcolm writes, in the two decades after 1690 the Turkish authorities—now under threat of eviction—tightened confessional screws on their subject populations: ‘a new wave of Islamicization seems to have taken place’, he notes, ‘using conversion as a pacification measure’, as well as heavier taxation to compel abandonment of Orthodoxy, and slaughter of the local priesthood.footnote¹ Although towns did not recover in size for another two hundred years, the Muslim proportion of the urban population grew steadily, as conversion to Islam meant automatic tax reduction. In these conditions, Kosovo—once the centre of the Serbian empire—became a magnet for Albanian immigration, and in due course even of some colon­ization by Turks, as well as Circassians from the Caucasus. Dervish orders expanded their activities in the region. By the nineteenth century Kosovo was a largely Muslim province of the Ottoman Empire, with some Catholic clans in the mountains. When much of the rest of the Balkans was aflame with revolt against Turkish rule in the epoch after the Napoleonic Wars, its Albanian population remained loyal to the Ottomans, and when the Young Turks took power in Istanbul in the early twentieth century, the provincial elites resisted the replacement of the old Arabic script by the ‘alien’ Roman newcomer, and stood firm for the maintenance of the shari‘a. Serbia finally regained the province only through the Balkan wars of 1912–13.

Reincorporated into Yugoslavia after the First World War, Kosovo became the object of determined efforts by the Serbian monarchy to reverse the demographic tide. With the help of Anglo-American charities, Serbian farmers and soldiers were resettled in the ‘Holy Land’, while many Albanians left for Turkey. Significantly, their interests were represented by the ‘Islamic Association for the Defence of Justice’, which sought not only to preserve the shari‘a but also the waqf, the beys’ feudal estates as well as the maternal language. So successful, however, was recolonization from Belgrade that by 1929 Serbs and Montenegrins constituted 61 per cent of the population of Kosovo. When the Second World War broke out the tables were turned again, as the Albanian population welcomed the Italian invasion of April 1939 and Mussolini integrated Kosovo into a Greater Albania under his rule. Some 100,000 Serbs fled northwards to Serbia, where a vigorous struggle against the Axis forces continued. After 1945, the province was granted autonomy within a Federal Yugoslavia by Tito—though not republican status—but the demographic balance could not be shifted. By 1971 Albanians comprised 73.7 per cent of the population; by 1991 Serbs amounted to only 11 per cent. In 1989 Serbia suppressed the autonomy of the province, ultimately goading the Albanian population into guerrilla warfare against Belgrade. The Kosovan Liberation Army, once inspired by Maoism and listed as a ‘terrorist’ organization by Washington, was soon transmuted into a NATO ally, and victor on the ground after a massive aerial bombardment of Serbia by the Western powers.

‘Ethnic cleansing’?

This is a conflict widely described as triggered by, and issuing into, ‘ethnic cleansing’. The term is unsatisfactory for at least two reasons. The first is the implicit notion that such ‘cleansing’ is an outrage peculiar to nationalist dementia or totalitarian power, which it is the mission of Western democracies to prevent or reverse. In fact, there is virtually no modern state whose emergence has not involved similar processes of—putting it euphemistically—‘national consolidation’, offering scant chance of reparation, let alone reversal. Who imagines that the fate of the Australian aborigines or native Americans is going to be cancelled by retrospective justice? What democracy is more toasted in Western capitals than Israel, founded on mass expulsion of Palestinians from their land, and long engaged in yet further expropriations of territory in the West Bank? ‘Cleansing’ has always been with us; its reception has less to do with ethical sensibilities than with the power of the respective parties at issue in the conflict.

But there is a second, and for our purposes more significant, short­coming to the notion of ethnic cleansing. It concerns the adjective rather than the noun. The term ‘ethnic’ has become a cant word in the social sciences and often in everyday speech, where it is frequently used in a blanket fashion to refer to any collective grouping with a semblance of homogeneity, in situations of conflict or positions of subordination. The concept of ethnicity has been so widely taken up because it gets around the problem of defining what it is that makes a people—that is, an ethnos—distinctive. Is the unity it possesses based on language, faith, descent, or culture in some vague sense? Ethnicity covers all as well as covering up all. It conveys, moreover, the suggestion of primordial differences that would be difficult in any circumstances to shake off. In Kosovo, Serbs and Albanians were divided by language and descent, but the depth of the conflict between them had its roots in a history in which religion loomed much larger than either of these markers. Ottoman rule in the Balkans rested on the claims of Islam, to which any linguistic or genetic group could accede, not on criteria of speech or blood. Serb resistance to it was inseparable from Orthodoxy. The religious antagonism between the two communities has persisted down to recent years, which have seen the wholesale destruction of mosques on one side and of churches on the other, in a climate of mutual fear and detestation. Serbian Orthodox bishops spoke of the danger of ‘genocide’ in Kosovo, warning that Albanians were trying to create an ‘ethnically pure’ state in the province. In Belgrade, Serb intellectuals did not hesitate to talk of the imminent ‘crucifixion’ of the Serb nation.

Bosnia and Ireland

An even clearer example of these tensions can be seen in neighbouring Bosnia. There, Michael Sells remarks, ‘the word “ethnic” in “ethnic cleansing” is a euphemism. Bosnians, Serbs, Croats and Muslims all speak the same language—they are divided only by religious criteria’. Describing the activities of Serb paramilitaries, he writes: ‘Those organizing the persecution identified themselves through explicit religious symbols, such as the three-fingered hand gestures representing the Christian Trinity, the images of sacred figures of Serbian religious mythology on their uniform insignia, the songs they memorized and forced their victims to sing, the priest’s ring they kissed before and after their acts of persecution, and the formal religious ceremonies’.footnote² Attacks on Muslims in Bosnia were not, of course, confined to Orthodox Serbs. Catholic Croats were often equally brutal, in pursuit of what was at one time a common aim of splitting Bosnia between them and eliminating the Muslim population—a programme Tudjman described as ‘Europeanization’. For their part, some Muslims sought to establish an Islamic Republic, and the later recruitment of mujaheddin fighters from Afghanistan to aid the Bosniak cause gave some colour to this notion.

In conflicts like these, the ethnic elements have been minimal. Often, they resemble more closely an internecine struggle between groups that are otherwise linguistically and in some respects even culturally indistinguishable. Attempts have been made to represent the Troubles in Northern Ireland as a battle between ethnic communities.footnote³ In reality, the ‘Scots’ who were settled in Northern Ireland in the 17th century were descendants of the ‘Irish’ who migrated to Scotland in the Dalriadic invasions of the Dark Ages; they could be said to have returned to their homelands. The point is that the difference between their identity and that of the rest of the Irish is not an ‘ethnic’ but a religious one—they are Protestant, and their opponents are Catholic.

Another line of thinking, not only Marxist, interprets such conflicts as struggles against ‘class oppression’. But although the political dominance of one group over another inevitably involves elements of stratification, neither in Yugoslavia nor in Ireland was class the primary line of division between the rival groups. In the pre-modern Balkans, Muslims often dominated the towns and formed the merchant class. But the division between communities was along religious lines, determined by the westward expansion of Islam into Europe, spearheaded by the Turks in the East and by the ‘Moors’ in the South. Of course, people convert from one religion to another for economic or political reasons, as they did in Kosovo, and have done in successive waves in India, where lower castes have variously converted to Islam, to Christianity and more recently to Buddhism. But once the conversion has taken place, the religion is no longer an expression of a class; its world view, its practices and beliefs take over as autonomous forces which may in turn provide or support what is often spoken of as ethnic identity.

Cyprus and Andalusia

There is no more vivid evidence of the power of such alternative ideologies than the fate of visual representations in the Near East. Cyprus offers some striking examples. In the ancient monastery of Chrysoroyiatissa, in the hills above Paphos, the church has been reconstructed after damage from earthquakes as well as bombs—dropped by (what Greeks term) the Turkish ‘hordes’, when the island was invaded and partitioned in 1974. Figurative representations of saints and the sacred decorate the iconostasis that separates the congregation from the Lord’s table. Local tourists or pilgrims lower their heads before one icon after another, bestowing a holy kiss upon each, as if they were living beings rather than simulacra or effigies. For Muslims, such conduct would be inconceivable: it is kaffir, pagan and primitive. Turkish attitudes to it are on display in the mosque of Selim the Conqueror in Nicosia. Once a fine Gothic cathedral, built shortly after the Crusaders’ capture of Cyprus in 1191, it was converted for the use of the faithful after the Turkish invasion of 1570. Walls and columns are painted completely white, save for capitals decorated in the reds, greens and yellows beloved of Islam. No representations of any kind—paintings or sculptures—are to be seen; even abstract or geometrical design is virtually absent. Windows that once contained stained glass now display an oriental grill. On the exterior, gargoyles have been covered by aluminium spouts. Any sculptured features around the building have been hammered with heavy blows, to destroy the imitation of life they once offered.footnote⁴ Here mimesis is aberrant; representation not only worthless, but blasphemous.

The two groups in contemporary Cyprus each have their own sacred and secular scripts, Greek and Arabic, which ensure that they cannot read one another. But the ideological contrast between them is not confined to scripture or liturgy, it extends to all imagery of living things. The world around each community constantly reinforces an opposition that is much more omnipresent than the notion of ethnicity or even formal religion would suggest, leaving the opportunities for reconciliation thin and ineffective. We can see this tension at its most acute and paradoxical at the other end of the Mediterranean. The sweeping autonomy of religion is epitomized here, too, in architecture. The ‘Moorish style’, excluding figuration and excelling in geometric and abstract forms, produced a series of worldwide and world-shattering designs—not ‘expressions’ of ethnicity or politics, but manifestations of a religious system—that spread from the magnificent mosque of Lotfollah in Isfahan eastwards to the Taj Mahal and beyond, and westwards to the Alhambra in Andalusia. There, after the expulsion of the Moors, the great mosque of Córdoba, in a mirror image of the cathedral of Nicosia, was eventually converted into a Christian church.

Yet during the Middle Ages, as Norman Daniel has shown, there was a great deal of commonality between the two civilizations. Indeed, from the end of the seventh century, Spain provided a bridgehead between East and West, or more accurately perhaps between North and South, Islam and Christianity. Under Franco, official Spanish historiography tried to play down the Muslim contribution to Spanish life, let alone European culture. But the love poetry of the troubadours which De Rougement thought a distinctively European invention may well have derived from the small courts of the muluk al-tawa’if, the petty Arab kings of Andalusia in the eleventh century. Daniel writes that ‘courtly poetry in Arabic, often trivial, yet ranged much more widely in theme and treatment than troubadour verse. If the latter had not a special position in European literary history, it might well be regarded as no more than a provincial and decadent offshoot of the court poets of Spain’. Suggesting that ‘the whole romantic tradition in European literature owes an almost disproportionate debt to eleventh-century Spain’, he also argues for ‘evidence of Platonic ideas in Provence at this period, conjectured to derive from Ibn Hazm’.footnote⁵ Born in Córdoba in 994, Ibn Hazm was the author of The Ring of the Dove, a poem about the art of love.footnote⁶ But while it was certainly composed in Spain, and therefore European in a geographical sense, historians from the North have drawn a boundary across the Pyrenees, excluding the Arabic culture of the South from their consideration. Daniel’s account makes clear that this is the wrong frontier from the standpoint of literature. What is European has become confused with what is (specifically) Christian.

In the scientific sphere, there has been more recognition of the extent to which the Renaissance was indebted to Arabic translators of classical sources, though the degree of Arab influence on, say, the schools of medicine at Salerno and Montpellier is still disputed. What is clear is that from the twelfth century onwards, key technological developments—the windmill, the mariner’s compass, the trebuchet, guns and gunpowder, the mechanical clock—were shared, in a profitable interchange of experience that made each side at once creditor and debtor to the other. ‘Only later did Europe draw ahead.’footnote⁷ Europeans, of course, have typically viewed these innovations as their own achievements, ignoring the contribution of Muslims to them.footnote⁸ At the time, however, there was much communication between Europeans and Arabs. It was mainly when they were acting principally as Christians and Muslims that tensions rose. Thus during the Reconquista in Spain, we hear of the killing of Arabs who were learned in religious law, while poets were spared. In secular fields there were fewer barriers.

Members of different religious groups have lived together peacefully in the Mediterranean over long periods, but the potentiality for discord if not outright conflict was nearly always present. Confessional divisions had social consequences, as religious norms shaped different patterns of politics, kinship, family, and other ties. Pierre Guichard contrasts ‘western structures’ and ‘eastern structures’ in Andalusia before 1492, the year that saw the end of the kingdom of Granada.footnote⁹ Both faiths were doctrinally rigid, and passage from one to the other meant an apostasy that did not even guarantee security on the other side, since fear that converts might relapse to earlier ways often made them suspect, as happened to Christians in Moorish Spain, or Jews in Europe. Those who rejected conversion, on the other hand, could take their faith to the point of fanaticism. Daniel avers that no hatred has ever surpassed that of the Christians who supported the Martyrs’ movement of ninth-century Córdoba, who went out of their way to provoke their own destruction by denouncing Islam as absolute evil. This was a strand in mediaeval Christianity that eventually led to the expulsion of the Moriscos and the Jews, and the activities of the Inquisition. For all the benefits of earlier periods of coexistence, the struggle between Christians and Muslims in Spain and Italy ended in a logic of extermination.

Daniel concludes that ‘the moral identity of Europe was preserved by a fiercely determined orthodoxy which wanted nothing to do with any least deviation in the whole field of religion’, and that ‘religion itself became the expression of that same sense of identity’.footnote¹⁰ But it is doubtful whether one can regard religion as an expression of identity. The formula smacks of sociological mysticism. Noel Malcolm too comes close to it, in declaring that ‘when modern concepts of nationhood began to be propagated in the nineteenth century, membership of this church supplied a ready-made “category of Serbianism”’.footnote¹¹ Religion is not an expression or a handle so much as a major constituent of such identity. In East Timor it is the Catholic Church in an otherwise Muslim setting that has helped to make the local population what they are, as daily images of the Bishop’s palace, churches destroyed, people crossing themselves before a statue of the Virgin Mary, remind us.

Death and conversion

Why should religion be such a potent factor in contemporary struggles that are so often misinterpreted as ethnic conflicts? Two reasons suggest themselves. The first is the sheer penetrative depth, and divisiveness, of religion as an ideological system. To believe in one God is to exclude the many. In some religions, in particular those originating in the Middle East, belief links the individual directly to the Creator of the world and hence to the sources of its pleasures and its problems. Here lie the sacred truths of our own past origins, of our present existence and of our future after death. Such beliefs find daily outward form in different ways of worship—crossing of oneself, wearing of hats, gesture of hands, posture of the body—and of attitudes to representation, which can polarize communities radically against each other in times of stress or perception of danger. Should such underlying oppositions explode into open conflict, then intense religious beliefs in the afterlife can lend an absolute value to death in the struggle against rival faiths which few secular ideo­logies can match. In life itself, moreover, there probably lies a second reason for the potential violence of religious conflicts between juxtaposed communities. ‘Cleansing’ can eliminate enemies defined by race or language, as well as by religion. But religious differences can also be erased by conversion, leaving other ‘primordial’ features of an individual or community in place, yet replacing the ideology and institutions that otherwise define them. Might not this very possibility render the boundaries between religions particularly fraught, making them an existential danger zone that calls for supreme defensive measures against any threat of corrosion or contamination—what we now often refer to as a fundamentalist reaction?

Considerations like these have been neglected in our political science. Much of the interesting literature on the rise of nations and nationalism, for example, has tended to look for transcultural factors in their spread across the globe. It is certainly correct to see the universalization of conceptions of the nation-state—units into which the whole world is now divided—as a strictly modern phenomenon, engineered from above. But the bases on which these structures have emerged are not at all unitary, but highly diverse. They will often relate to language or territory, but they may also be associated with religious beliefs. These have been very much underplayed in our theories of modernization. In late Victorian times Frazer argued in The Golden Bough that, historically, magic gave way to religion and religion to science, in evolutionary fashion. Human development could thus be seen as layered into so many geological strata. Since Frazer’s time, secularization of intellectual culture has proceeded apace. If we look at recent works on nations and nationalism, we can see how rarely they pay attention to religion. In their works on the subject—Nations and Nationalism since 1780 (1990); Nations and Nationalism (1983); The Nation-State and Violence (1985)—neither the socialist Eric Hobsbawm, the agnostic Ernest Gellner nor the modernist Anthony Giddens devote any space to it.

Thus a glance at the maps at the end of Hobsbawm’s admirable work shows that they include ‘nationalities’, peoples, languages and political divisions—but not religions. Gellner remarks that ‘among cultures, it is the ones linked to a high (literate) faith which seem most likely to fill the role of crystallizer of discontent’,footnote¹² tacitly demoting religion to a mere vehicle of profounder forces (‘discontent’), rather than taking it as a primary element of identities in its own right. Fred Halliday, writing on the politics of the Middle East, distinguishes between a religious Islam and a secular ‘Islam’, the better to set aside the former; as he puts it, Islam as a religion ‘exists as a system of belief about the supernatural and related questions of morality, destiny and meaning. This is a matter for believers and theologians and is not the subject of what follows’footnote¹³—as if a political analysis of Muslim societies were possible in abstraction from religious doctrines, in contradiction of the dedication of his book to ‘Iranian friends and democrats, opponents of the religiously sanctioned dictatorship’. The limitation of all these approaches is that they make it difficult to understand how others could invest so significantly in beliefs we do not share. The temptation is then to downgrade religious practices or images to objects of art or ‘material culture’ rather than critical constituents of faith.

These are myopias of a secular social science that has forgotten the past of its homelands. If we are to understand political conflicts round the world today, we cannot afford such blind spots: ‘ethnic cleansing’ is too often inspired, or intensified, by confessional animosities. It is no accident that the two defining genocides of the twentieth century, though each was committed by a supposedly secular political force, both engaged ancient religious hatreds: Muslims against Christians in the Young Turk slaughter of the Armenians, Christians against Jews in the Nazi extermination of the Jews. Nor that the longest military occupation of modern times, after massive expulsions, sets Jews over Muslims and Christians in Palestine. In Europe today, church attendance has fallen, but it would be an illusion to think that religion does not remain a defining feature of society at a very general level, capable of coming to the surface in crisis conditions, should ‘our Christian civilization’ be threatened. Advanced capitalism does provide an alternative set of goals in life, as did Communism. But the collapse of the USSR has seen a significant reassertion of religious ideologies both in Russia and in Central Asia; while in the prosperous, consumer-oriented environment of the USA, religious values continue to permeate public life, taking violent as well as peaceful forms. In Turkey or India, the secular legacy of Atatürk and Nehru is under severe challenge from Muslim and Hindu revivalism. In Indonesia, in Egypt, in Nigeria, a confessional cauldron is simmering. European historians, sociologists and political scientists reared in sceptical traditions dismiss the power of religion in the modern world at their peril.