ISJ 2 Index | Main Newspaper Index

Encyclopedia of Trotskyism | Marxists’ Internet Archive

International Socialism, Autumn 1999


Neil Davidson

The trouble with ‘ethnicity’


From International Socialism 2:83, Summer 1999.
Copyright © International Socialism.
Copied with thanks from the International Socialism Archive.
Marked up by Einde O’Callaghan for ETOL.


The concept of exploitation is central to the Marxist understanding of history and contemporary society. But not all social conflicts can be immediately reduced to the struggle between exploiters and exploited, and to explain these conflicts we require other concepts. The most important is that of oppression. This refers to systematic discrimination by one social group against another on the grounds of characteristics either inherited (skin colour, gender) or socially acquired (religious belief, sexual orientation). The experience of oppression cuts across class lines, although that experience is more or less severe depending on where its victims are placed within the class structure. Some forms, like the oppression of women, have persisted throughout the existence of class society, while others, like racism, are specific to capitalism alone. Sometimes the reasons, or pretexts, for the oppression of a group may change over time. During the feudal era, for example, Jewish people were persecuted for their religious beliefs, but as capitalism developed persecution increasingly took place on the grounds of their supposed race. Whatever the reason or pretext, however, ruling classes throughout history have instigated or endorsed the oppression of different groups in order to maintain or create divisions amongst those over whom they rule. Recently, groups have increasingly been subjected to oppression on the grounds of their ethnicity. The most extreme form of such oppression has become known as ‘ethnic cleansing’.

The term ‘ethnic cleansing’ is an English translation of the Serbo-Croat phrase etnicko ciscenje. It was first used in Yugoslavia, not in the conflicts which erupted after the end of the Cold War, but by the Croatian Ustashe during the Second World War to describe its policy of killing or expelling Serbs, Jews, Gypsies and Muslims from the fascist state the Ustashe briefly set up with Nazi support. The first use during the current events was by the Croatian Supreme Council in 1991, after the Croatian declaration of independence from Yugoslavia, to describe the actions of Serb guerrillas who were attempting to drive Croats out of areas where Serbs were in the majority: ‘The aim of this expulsion is obviously the ‘ethnic cleansing’ of the critical areas [to] be annexed to Serbia.’ The phrase only began to appear in the British press – and thereafter in popular usage – during the war which began in Bosnia-Hercegovina the following year, when Bosnian Serb forces, initially backed by the Milosevic regime in Belgrade, started expelling Muslims and Croats from those parts of the state territory that the Bosnian Serbs considered to be Serbian. [1]

Since then the term has been used not just to describe events in former Yugoslavia (where all sides became involved in the practice to some extent), but in similar – and in some cases even worse – occurrences distant in space and time. On the one hand, the term was being extended spatially to events, such as the massacres in Rwanda during 1994, which took place in societies geographically distant from Yugoslavia and were quite different in terms of their historical development. On the other hand, the term was also extended chronologically back to events, such as the expulsion and killing of Armenians by Turks at the end of the First World War, that were historically distant and had not previously been discussed in these terms. [2]

’Ethnic cleansing’ presupposes the existence of different ethnic groups. The majority of people who opposed the bombing of Yugoslavia also opposed the ‘ethnic cleansing’ of Kosovan Albanians which NATO used to justify it, arguing that it both intensified the hatreds which made ethnic cleansing possible and made it easier to carry out by forcing the removal of the international monitors who had provided some check on the Serb paramilitaries. However, the opponents of the war tended to share with supporters of the war – and indeed with the people carrying out the ‘ethnic cleansing’ – the view that there were genuine ethnic differences between groups in former Yugoslavia. From the anti-war perspective, ethnic differences such as those between the Serbs and the Kosovan Albanians should be mutually respected rather than made the occasion for oppression, but the differences themselves could and should not be denied. This position is inadequate, and I want to argue instead that we need to go beyond opposition to ‘ethnic cleansing’ – which of course means all ‘ethnic cleansing’, not only that of the Kosovan Albanians – and question the validity of the term ethnicity itself.

Since the argument that follows may be liable to misrepresentation, I should perhaps make one central point clear from the start. Ethnicity is often equated with culture, most frequently with that of minority populations in Western Europe and North America, or with non-Western cultures more generally. Older readers may remember a time in the late 1960s and early 1970s when Guatemalan pottery or Afghan textiles were regularly described as ‘ethnic’ when being marketed in Britain, as if ‘ethnicity’ was some special property which they possessed. I am not arguing against cultural diversity, still less suggesting that socialists should abandon their duty to defend people whose culture is under threat, or who are suffering from any of the other forms of oppression outlined above. Nearly 100 years ago, Lenin pointed out the necessity for socialists to be ‘tribunes of the people’ in words which still retain their relevance: ‘Working class consciousness cannot be genuine political consciousness unless the workers are trained to respond to all cases of tyranny, oppression, violence and abuse, no matter what class is affected – unless they are trained, moreover, to respond from a social democratic [i.e. revolutionary socialist] point of view and no other’. [3] For socialists, therefore, it makes no difference whether particular groups of people are oppressed because of their language, religion, nationhood, or ethnicity. In each case our duty is to defend the oppressed and show solidarity with them, particularly where socialists themselves belong to the dominant linguistic, religious, national or – assuming for the moment that such a thing exists – ethnic group.

My point is rather that the way in which the notion of ‘ethnicity’ is currently and increasingly being used contains a number of problems for the left. Two stand out in particular. On the on hand, those who approve of ethnicity as the affirmation of a cultural identity, in so far as they emphasise supposedly innate differences between human social groups, are in danger of opening the door to the current form of racist ideology. On the other hand, those who disapprove of ethnicity as a manifestation of (real or imagined) exclusionist tribalism are in danger, in so far as they suggest that ‘ethnic’ nationalisms are particularly prone to oppressive behaviour, of obscuring those characteristics which all nationalisms have in common, whether they are oppressor, oppressed, or fall into neither of these categories. Our first task is therefore to distinguish between the various ways in which the term ‘ethnicity’ has been used, and assess their respective validity.

Kinship, occupation and identity

‘Ethnicity’ has been defined in three ways: first, where members of a group have a common line of descent, and consequently a shared kinship; second, where they have a common position within the international division of labour and consequently a shared occupation; third, where they have one or more cultural attributes in common and consequently a shared identity. The first and second reasons assume that ethnicity can be defined objectively, the third that it can be defined subjectively. As we shall see, it is this subjective definition which is currently dominant.

Kinship: Social groups which share a common line of descent are usually referred to in anthropology as endogamous groups, or groups whose members interbreed exclusively with each other, thus maintaining the same genetic inheritance. Such groups would have been universal at the origins of human evolution but are, however, virtually impossible to find today. Indeed, recent archaeological and anthropological work suggests that mass human migration – often across entire continents – occurred much earlier in history than was previously believed, and resulted in the erosion of endogamy within the original tribal societies. One writer notes that as a result of these factors, within tribal society ‘the common ancestry of “the people“ was always partially fictive’. [4] But once we move onto the terrain of recorded history the multiple genetic inheritance of the global population is an indisputable fact that also makes the existence of different ‘races’ impossible to sustain. Susan Reynolds has rightly criticised the tendency among medieval historians to describe the barbarian invaders of the Roman Empire as biologically distinct ‘tribal’ entities, simply because of the continued use of their original group name. ‘This must be wrong,’ she points out. ‘Once barbarians had been converted to orthodox Christianity and prohibitions on intermarriage had been lifted, it must have been hard to distinguish them from “Romans“ who were already mixed genetically and were increasingly barbarised culturally’. [5]

The main constituent nations of Britain are a case in point. Early in the 18th century Daniel Defoe mocked the pretensions of his countrymen to ethnic purity in his satirical poem, ‘The True-Born Englishman’:

In eager rapes, and furious lust begot,
Between a painted Briton and a Scot:
Whose gen’ring offspring quickly learnt to bow,
And yoke their heifers to the Roman plough:
From whence a mongrel half-bred race there came,
With neither name nor nation, speech or fame:
In whose hot veins now mixtures quickly ran,
Infus’d betwixt a Saxon and a Dane.
While their rank daughters, to their parents just,
Receiv’d all nations with promiscuous lust.
This nauseous brood directly did contain,
The well-extracted blood of Englishmen.

As Linda Colley comments, ‘Defoe’s uncompromising insistence on the ethnic diversity of England, its early exposure to successive invasions from continental Europe, and the constant intermingling of its people with the Welsh and Scots, was fully justified in historical terms’. [6] Similar intermingling took place in Scotland during the ‘Dark Ages’ between 400 and 1057. ‘The period has, with justice, been called “an age of migrations“,’ writes Michael Lynch, ‘when the different tribal peoples – Picts, Scots, Angles, Britons and Scandinavians – who inhabited the mainland of modern day Scotland moved, fought, displaced and intermarried with each other’. [7] And to these, of course, could be added the Norman English who were invited to settle in Scotland during the reign of David I (1124–1153), and who were themselves descended from Viking settlers in part of what is now France.

In an extreme case like that of the native Australians it might be supposed that endogamy was maintained until the arrival of the European colonists, but in fact they too had interbred with Papuan and Polynesian immigrants many centuries before the Dutch or the British set foot on their continent. [8] As the late Eric Wolf wrote of the ethnic composition of the world in 1400, ‘If there were any isolated societies these were but temporary phenomena – a group pushed to the edge of a zone of interaction and left to itself for a brief moment in time’. [9] In short, even before capitalism had penetrated all corners of the world in the search for markets and raw materials, the growth of trade, conquest and migration had already made the existence of endogamous gene pools increasingly rare. Of course, this does not mean that various groups have not claimed, and in some cases perhaps even believed, that they were descended from the pure stock of some ancestral group, but it is important to understand that these claims and beliefs are based on a myth of kinship, not a reality.

Occupation: Like the modern notion of ‘race’, the origins of this second definition lie in the colonial expansion of capitalism outside of its European heartlands. From the origins of systemic racial slavery in the 16th century, ‘race’ has been a general term to override differences between different peoples by categorising them on the basis of physical characteristics, of which skin colour was the most important – although, as we shall see below, the instances of racism directed against the Catholic Irish and by extension the Highland Scots were exceptions in that they were based on religion and language rather than physical appearance. [10] There were massive differences in terms of social development between the Shona speaking peoples of southern Africa who built and lived in the stone city of Great Zimbabwe during the 15th and 16th centuries and the hunter-gatherers who inhabited Australia at the same time. Yet to ideologists of ‘race’ they were both indistinguishably ‘black’. At first racism was used to justify the allocation of specific roles within the system during the process of primitive accumulation (i.e. as slaves), but latterly racism was used to consign members of ‘races’ who had migrated to the metropolitan centres to become either part of the reserve army of labour or workers with the worst pay and conditions in the labour force.

‘Ethnicity’, on the other hand, was a term designed to distinguish between groups within overall ‘racial’ categories in those sections of the labour market in which they had established themselves. The capitalist mode of production requires the subordination of labour to capital, but in the European colonies it also required that the labour force was internally divided. As Eric Wolf notes, the allocation of workers to invented ethnic categories is doubly effective in this respect, first ‘by ordering the groups and categories of labourers hierarchically with respect to one another’, and second ‘by continuously producing and recreating symbolically marked “cultural“ distinctions among them’. On the one hand, groups were allocated specific roles both within the production process and within social life more generally. On the other hand, they were encouraged to identify with these roles and to defend them against other groups. Wolf is therefore right to say that these ethnic identities are not ‘“primordial“ social relationships’, but ‘historical products of labour market segmentation under the capitalist mode [of production]’. [11] Sometimes these built on existing division of labour within pre-colonial society; sometimes they were wholly new and based on the division of labour within the new industries that the colonists established.

In Rwanda and Burundi before colonisation there were three distinct groups – the Twas, the Hutus and the Tutsis. The most numerically significant were the Hutus and the Tutsis, membership of which passed down through the male side of the family. [12] Is this an example of ‘kinship’ ethnicity which I earlier consigned to prehistory? In fact, although group membership at birth was based on that of the male parent, it was possible to move from the Tutsis to the Hutus in the course of life. All three groups spoke the same language, and the distinctions between them were principally based on the fact that they performed different social roles: the Hutu in farming, the Tutsi in cattle rearing and the Twa in hunting. A cattle owner was a Tutsi by definition, which meant that Hutus could ‘become’ Tutsis if they were able to accumulate sufficient wealth to become cattle owners themselves, a transition that was marked ceremonially. Since longhorn cattle were the main form of disposable property, people who owned cattle were therefore a significant part of the ruling class, but there were also Tutsis who owned few cattle and whose social position was proportionally less important.

The situation is further complicated by the fact that Hutus were members of that section of the ruling class who owned large farms, without becoming Tutsis. As Charlie Kimber writes, ‘The Hutu-Tutsi distinction in pre-colonial Rwanda and Burundi was not a simple class distinction (because you could be a poor Tutsi or a rich Hutu), nor was it an ethnic distinction (because you could be born into one group and die as another)’. [13] It did, however, become an ethnic distinction with the arrival first of the German and then the Belgian colonial administrations. Under these regimes, real occupational stratification which designated people as being Hutu, Tutsi and Twa was transformed into imaginary ethnic distinctions between separate ‘tribal’ groups, one of whom (the Tutsis) was privileged over the others in the colonial hierarchy and in the immediately post-colonial state.

Rwanda shows how the existing occupational roles of existing populations can become the basis of new ethnicities imposed by colonialism. More commonly, ethnicities have developed among migrant groups responding to dislocation and industrialisation, as the emergence of ethnicity in the Gezira region of the Sudan after the beginning of cotton production in 1925 demonstrates. Some of the workers recruited by the British migrated for this purpose from various West African groups, all of whom were Muslims and most of whom spoke Hausa. Unlike the local Sudanese, these immigrants were already accustomed to wage labour in their own homelands, which had been industrialised earlier, and were consequently more likely to meet their quotas. The British tended to replace local workers with the immigrants. In response to their displacement, the Sudanese began to refer disparagingly to the West Africans as ‘Fellata’, a term which has overtones of slavish obedience. The West Africans in turn began to distinguish themselves from the Sudanese precisely on the basis of their supposedly greater capacity for hard work, a distinction linked to the adoption of a fundamentalist Islam far stricter than that practised by the Sudanese, and which was enshrined in their self description as ‘Takari’, a respectful term to describe pilgrims to Mecca from West Africa. As Eriksen concludes from this episode, ‘Contemporary ethnicity or “tribalism“ is not, in other words, a relic of the past but a product of modernisation processes leading up to the present’. [14]

The sugar industry in the French colony of Mauritius provides an extreme example of how a wide range of characteristics can be fitted to occupational roles, becoming ethnicities as a result. Indians – members of a national group – were recruited as labourers in the canefields. Brahmins from among these Indians – adherents of a religious sect – were made foremen. Creoles – descendants of slaves and consequently identifiable by their skin colour – tended to be skilled workers. Chinese or ‘Mulattos’ (’half-castes’) – a national group and one defined by skin colour – held the middle managerial positions. The estate managers were invariably French settlers, who were both a national group and identifiable by their white skin but, needless to say, they were not considered to have an ‘ethnicity’. [15]

We might say, therefore, that the term ‘ethnicity’ is valid in this sense where it is used to describe the way in which existing occupational patterns in pre-capitalist societies were used by European colonists to classify the population as supposedly endogamous groups, or where the migrations set in train by colonialism had led groups to define themselves as either endogamous, or in possession of some quality or characteristic which distinguished them from the native populations around them. What has confused the issue is that the word ‘ethnicity’ was not in general use at the time these developments were taking place (roughly between 1875 and 1945), but this would not be the first time that something has existed in the world before the language has been developed to describe it. [16] Nevertheless, it could be usefully employed now in relation, for example, to the situation of Chinese traders in Indonesia or Korean traders in Los Angeles. But the term is generally not used with this degree of specificity. On the contrary, it is the third and final notion of ‘ethnicity’ – that of identity – which is currently sweeping all before it.

Identity: In answer to the central question of why groups come to identify themselves as having a particular ‘ethnicity’, Anthony Smith has argued that an ethnic community – that is, a community whose members have not had their ‘ethnicity’ imposed on them from outside, but distinguish themselves in this way – has six main attributes: ‘a collective proper name, a myth of common ancestry, shared historical memories, one or more differentiating elements of common culture, an association with a specific “homeland“, and a sense of solidarity for significant sectors of the population’. [17] This attempts to incorporate a belief in kinship relations as part of the definition. It is not clear why these are ‘ethnic’ attributes rather than simply ‘national’ ones. Indeed, the definition of a nation given a few pages earlier by Smith could be substituted without affecting his argument: ‘A nation can therefore be defined as a named human population sharing an historic territory, common myths and historical territories, a mass public culture, a common economy, and common legal rights and duties for all members’. [18] (The similarity is perhaps unsurprising, given that Smith is attempting to argue for the importance of ethnicity in the formation of national identity.) It is perfectly possible for a particular social group to identify themselves as having an ‘ethnic’ identity without possessing all or any of the attributes listed by Smith, as the example of the Bosnian Muslims makes clear. Like all classificatory lists, the elements are completely arbitrary. Perhaps in realisation of this, some writers have abandoned any attempts at precise definition.

In 1953 David Reisman became the first person to use the term ‘ethnicity’ to mean identity and he was quickly followed in this by other North American sociologists. [19] In their hands the term was used to describe those groups who did not belong to the white Anglo-Saxon Protestant (WASP) population – that is to say, everyone who was not descended from the original English, Scottish, and ‘Scots’ Irish (Protestant Irish) settlers. Exceptionally, German immigrants were allowed to merge with the WASPs, at least where they too were Protestants. [20] Ethnicity was therefore reserved for ‘minorities’ identified by attributes as diverse as skin colour (blacks), religion (Jews) or country of origin (Italians). This lack of specificity brings to mind the famous conversation between Humpty Dumpty and Alice during her adventures through the looking glass:

‘When I use a word,’ Humpty Dumpty said, in a rather scornful tone, ‘it means just what I choose it to mean – neither more nor less.’ ‘The question is,’ said Alice, ‘whether you can make words mean so many different things.’ ‘The question is,’ said Humpty Dumpty, ‘which is to be the master – that’s all’. [21]

Stuart Hall, here playing the role of Humpty Dumpty, has offered the following definition of ‘ethnicity’: ‘By “ethnicity“ we mean the commitment to those points of attachment which give the individual some sense of “place“ and position in the world, whether these be in relation to particular communities, localities, territories, languages, religions or cultures’. [22] If the term encompasses ‘communities, localities, territories, languages, religions or cultures’, then it is difficult to see what could not be defined as ‘ethnic’. A US sociologist, Abner Cohen, once proposed that City of London stockbrokers should be considered an ‘ethnic’ group by virtue of their group identity. [23] He was not being entirely serious, but the proposal takes the logic of ‘ethnic identity’ to its conclusion in Bedlam. More seriously, the census which British citizens will be required to complete in 2001 asks respondents to define their own ‘ethnicity’ from a core list which consists of four nations (Bangladesh, China, India and Pakistan), one continent (Asia) and two skin colours (black and white) – although these are subdivided, the first into ‘Black: African’, ‘Black: Caribbean’ and ‘Black: Other’, and the latter into ‘White: British’ and ‘White: Other’. In fact, most uses of the word ‘ethnic’ are in place of some other word (like ‘communities, localities, territories, languages, religions or cultures’), the use of which would give far greater precision of meaning. The result of not doing so, as the South African Marxist Neville Alexander rightly says, is ‘to reduce the diverse reasons for the emergence of group solidarities to a single quality called “ethnicity“, thereby obscuring precisely what has to be explained – the basis of such solidarity’. [24] Ethnicity, in short, becomes a way of labelling people through the use of an ideological super-category that includes virtually any characteristic they might conceivably possess.

There is a further problem. Hall assumes that ‘ethnicity’ can be divided into ‘bad’ and ‘good’ forms that more or less correspond to that of the majority populations of metropolitan imperialist states on the one hand and of their minority immigrant communities on the other. Of the former, Hall writes, ‘In the face of the proliferation of cultural differences, and the multi-ethnic character of the new Britain, and threatened on the other side by the encroaching trauma of an emerging European identity, we have seen over the past decade a particularly defensive, closed and exclusive definition of “Englishness“ being advanced as a way of warding off or refusing to live with difference – a retreat from modernity no exercise in managerial newspeak or the “new entrepreneurialism“ can disguise or deflect’. [25] Of the latter, however, we learn that it is ‘not an ethnicity which is doomed to survive, as Englishness was, only by marginalising, dispossessing, displacing and forgetting other ethnicities’. On the contrary, these immigrant communities have a ‘politics of ethnicity predicated on difference and diversity’. [26] It is difficult to see how Hall could explain conflict between youth of Afro-Caribbean and south Asian descent on this basis (or indeed, that between those of Afro-Caribbean and Korean descent in Los Angeles). And while a model of ‘ethnicity’ derived from the British (or rather ‘English’) experience can certainly be generalised to other Western European imperialist nations like France, in a region like the Balkans, where historically there has never been a dominant ‘ethnic’ group, it has no explanatory power whatsoever.

The internationalisation of capital, crisis and identity politics

The editors of a recent reader on ethnicity begin by reflecting on the sudden upsurge of interest in their subject:

For at least 150 years liberals and socialists confidently expected the demise of ethnic, racial and national ties and the unification of the world through international trade and mass communications. These expectations have not been realised. Instead we are witnessing a series of explosive ethnic revivals across the globe. In Europe and the Americas ethnic movements unexpectedly surfaced from the 1960s and 1970s, in Africa and Asia they have been gaining force since the 1950s, and the demise of the former Soviet Union has encouraged ethnic conflicts and national movements to flourish throughout its territory. Since 1990, 20 new states based largely upon dominant ethnic communities have been recognised. Clearly ethnicity, far from fading away, has now become a central issue in the social and political life of every continent. The ‘end of history’, it seems, turns out to have ushered in the era of ethnicity. [27]

Why has the upsurge of ‘ethnic’ identification taken place now? For this sense of ‘ethnicity’ to become established required two general conditions. The first condition is the need to distinguish one group from another. As Thomas Eriksen has stressed, ‘Ethnicity is essentially an aspect of a relationship, not a property of a group.’ In other words, cultural distinctiveness in itself does not confer ‘ethnicity’ on a group, but only when it is contrasted with the culture of another group: ‘For ethnicity to come about, the groups must have a minimum of contact with each other, and they must entertain ideas of each other as being culturally different from themselves’. [28] But for this to happen the differences must themselves be considered important, and there are only certain circumstances in which this is the case. The most important of these circumstances, and the second condition, is rapid social change. As Malcolm Cross notes, ‘A man living in a world where change is largely absent does not need to be reminded of his culture in order to affirm his identity’. [29] Where that change is destructive of established ways of life – and in some cases whole societies – and class politics does not offer an alternative, then distinguishing oneself as part of a specific group in order to struggle over the resources, or scavenge what you can from the rubble left by the onward march of international capital, may appear to be the only available option even where group membership may previously have meant little or nothing to the people concerned.

Across the developing world in particular, the state is increasingly failing to deliver any form of social redistribution to the most disadvantaged. And some areas, most of which are in Africa, have seen not just increasing poverty but actual social collapse, brought on by economic crisis, which the state has been unable to prevent. In these circumstances an ‘ethnic’ community, often constructed by colonial powers that have long since departed, can provide what the state cannot. As David Brown notes, ‘If the state claiming to be the cultural nation cannot offer the necessary protection, then it is the cultural nation claiming to be the potential state which offers the next best bet’. [30] Rwanda provides a particularly tragic example of what can result from the residue of Western invented ‘ethnicities’ in a situation of acute social crisis. The genocide of 1994, far from being the expression of age old ‘ethnic’ animosities, was prepared by the destructive impact of colonialism on Central Africa.

The left should be at the forefront of opposition to this, reasserting the realities of class against the myths of ethnicity, but all too often it is handicapped by its refusal to accept that ‘identity’ can ever be irrelevant, or mystified, or simply a cover for sectional interests. As Adam Kuper writes, ‘So although the popular American notion of cultural identity has been stretched beyond ethnic groups to other kinds of minorities, it remains doubly essentialist: one has an essential identity, and this derives from the essential characteristic of the collectivity to which one belongs’. [31] In most cases, however, there are no ‘essential’ characteristics – nor, indeed, have many of these collectivities existed for any length of time: ‘In actuality, a sense of ethnic community can develop among individuals who neither share significant cultural attributes nor who are particularly distinctive from their neighbours; and it can refer to commonalties of circumstance which developed within living memory, and to attributes which clearly do not objectively derive from common ancestry’. [32] The more developed world – in this case the Balkans – provides us with the best example of how, unlike Central Africa, ‘ethnicities’ can arise with virtually no prior basis.

Unlike their parents, or even their grandparents, many of the people who came to be described as ‘ethnic Muslims’ in Bosnia-Hercegovina had never been inside a mosque in their lives – at least until they began to be identified in this way for the purposes of persecution, when religion took on a new significance for them. As this example suggests, the distinction between ‘imposed’ and ‘chosen’ ethnic identities is not one which can be sustained, since there are many cases where groups which have been identified as possessing a particular attribute and discriminated against on that basis have subsequently chosen to militantly assert that identity in response to their oppressors. But this process is not inevitable. As Misha Glenny writes of the Bosnian Muslims, ‘Although largely secular, the explicit religious origins of the Muslims’ identity (they have no specific ethnic or linguistic criteria to differentiate themselves from Serbs or Croats, neither do they have a Belgrade or Zagreb to turn to for material, political or spiritual aid) have made the process of defining their nationhood exceptionally difficult.’ It is interesting that Glenny, who is otherwise one of the most insightful commentators on the Balkan situation, sees this as a problem, rather than a hopeful basis for overcoming the divisions within Bosnian society, noting that ‘many Muslims incline towards aspects of either Serbian or Croatian culture’. [33] But before sides became fixed in the Bosnian war it was by no means certain that residual religious belief would be inflated until it became an imaginary essence by which people were defined: ‘Before the war ... when the Serbs still hoped to keep Bosnia in Yugoslavia, the media frequently highlighted similarities with the Muslims, while Croats often stressed that Bosnia had been part of historical Croatia and that most Bosnian Muslims were originally of Croatian descent’. [34] In other words, these Muslims could have been absorbed into either Serb or Croat ‘ethnicities’, in which case the supposedly essential nature of their Islamic identity would never have arisen.

The developed world has not remained untouched by the rise of – or perhaps one should say the retreat to – ‘ethnicity’. The crisis in Western Europe and North America is clearly not of the same order as that in the Balkans, still less Central Africa, but similar pressures are at work. Where reforms are increasingly hard to come by, two collective solutions remain for improving conditions. One is the road of class struggle, of forcing redistribution either directly from the bourgeoisie in the form of higher wages and better conditions, or indirectly by forcing the state to intervene through legislation or increased taxation. The other road, the road more frequently travelled, is not to struggle for redistribution from the capitalist class to the working class, but to struggle – or more precisely, to lobby – for resources to be redistributed from one section of the working class to another, or from one region to another, or from one ‘ethnic’ group to another. If groups can become politically organised, and consequently put electoral pressure on local or national politicians, they, or more usually their representatives, can campaign for ‘affirmative action’ on their behalf. [35]

The latter strategy has a long history in post-war Britain stretching back to the 1960s. It was only after the onset of economic crisis in the 1970s that it came to full maturity. Ambalavaner Sivanandan notes acidly the ‘scramble for government favours and government grants ... on the basis of specific ethnic needs and problems’ by ‘minority’ groups following the Brixton riots of 1981 and the recommendations of the Scarman Report. The problem is not simply the compromises and downplaying of radical demands which are required to receive state funding, but the fact of competition between communities that increasingly divides them into rival ‘ethnic’ groups. [36] It is not even the case that such funding as is available invariably goes to the working class areas, since the middle class can play the lobbying game to far greater effect, and will generally reap whatever benefits are to be had.

It would be bad enough if accepting the existence of ethnicity merely meant condoning an endless splintering into rival groups to divide up the crumbs left by global capital, but there are even worse implications. The most serious of these is the relationship between ethnicity and racism.

‘Ethnicity’ and the new racism

For many on the left (as well as academics and officials in government agencies) it is perfectly acceptable to talk about ethnicity (without quotation marks), whereas it is no longer acceptable to talk about ‘race’. There are, in other words, no such things as ‘races’, but there are such things as ‘ethnicities’. As Steve Fenton writes:

The term race is associated with mistaken science. It connotes physical difference and, frequently, colour. It is typically seen as malign, and racial ideologies have been associated with compulsion and regimes of oppression. By contrast, ethnic can be taken as an analytic term in social science, is often seen as the voluntary identification of peoples, and as (at least potentially) benign. [37]

The problem is that the notion of ethnicity is all too often used to invoke precisely the qualities that used to be invoked under the now discredited notion of ‘race’. To understand why, it is necessary to trace the previous major shifts in racist ideology.

Marxist accounts of the origins and development of racist ideology tend to see three moments in the history of capitalism as decisive in determining its precise form. The first is slavery, and the need to justify enslaving millions of fellow human beings at the very moment when men were being declared equal and in possession of certain unalienable rights. The second is colonialism, and the need to justify the conquest and subsequent domination of foreign peoples. The third is immigration, and the need to justify discrimination against peoples who were usually encouraged to come to the metropolitan centres in the process of reconstruction after the Second World War. The respective justifications for the treatment of non-white populations differed in each case, moving from their less than human nature (making it permissible to enslave them) to their backwardness (requiring the guidance of the more advanced white ‘races’), to the competition they posed to the white populations for jobs and housing (requiring an end to immigration and in extreme versions the repatriation of existing immigrants). [38] In an important book published in 1981, however, the Marxist philosopher Martin Barker argued that we were now seeing the rise of a ‘new racism’ which:

… can refuse insults: it need never talk of ‘niggers’, ‘wogs’ or ‘coons’. It does not need to see Jews as morally degenerate, or blacks as ‘jungle bunnies’. Nonetheless, in subtle but effective ways it authorises the very emotions of hostility that then get expressed in these terms. [39]

The ‘newness’ of this racism is not in its reliance on the pseudo-sciences of sociobiology and ethnology for justification – pseudo-science has been a feature of racism from the invention of phrenology in the mid-19th century onwards – but in the claim that it demonstrates the social incompatibility of groups with different cultures. There are two historical precedents for this shift in meaning, in which an entire people were defined by virtue of what had previously been seen as an acquired characteristic – religion in the first, culture more generally in the second – rather than biology.

The first was in Ireland. In his work on the origins of racism, Theodore Allan defines racial oppression as the reduction of ‘all members of an oppressed group to an undifferentiated social status, a status beneath that of any member of any social class within the colonising population’. Allan argues that racism originated not from innate propensities on the part of different groups to distinguish themselves from and discriminate against other groups (the ‘psycho-cultural’ argument), but as a conscious ruling class strategy to justify slavery as an economic system in the epoch where formal male equality was increasingly the norm (the ‘socio-economic’ argument). Although his argument is mainly concerned with the racial oppression in the Americas, Allen sees a precursor of white colonial attitudes to the Native Americans and African slaves in the British (i.e. Lowland Scots and English) treatment of the Irish from the Anglo-Norman period onwards. With the Reformation, however, the religious difference between the Protestantism of the British settlers and the Catholicism of the Irish natives provided an additional element to the prejudices of the former: ‘What had fed primarily on simple xenophobia now, as religio-racism, drank at eternal springs of private feelings about “man and god“.’ There were also more material reasons. As Allan strongly argues, the construction of ‘religio-racism’ against the entire Irish population was a conscious choice on the part of the English ruling class and its Scottish allies. Ireland was a crucial strategic territory in the struggle between Catholic and Protestant Europe, hence the impossibility of co-opting sections of the Catholic Irish ruling class for the purposes of social control: they could not be trusted to take the British side in the conflicts with Catholic Spain and France. The alternative was to attempt to convert the Catholic population to Protestantism, but this was unthinkable for most of the 18th century for two reasons. First, the Ascendancy comprised a relatively small minority of the population whose wealth and power would have been threatened if a majority had been allowed to share its legal privileges. Second, the majority of Protestants below the ruling class proper were Dissenters, most of them Presbyterians, and consequently excluded from the privileges available to communicants with the Anglican Church of Ireland.

Mass conversion of the Catholic population was likely to lead to the converts joining the Dissenting branch of Protestantism rather than that of the great landowners, raising the prospect of the majority of the population uniting against the Ascendancy. After this came near to happening anyway in 1798, the British ruling class and its Irish extension responded by incorporating the Dissenting element through the Orange Order, but, more importantly, by shifting the nature of Catholic oppression from a racial to a national basis ‘by the incorporation of the Irish bourgeoisie into the intermediate buffer social control system’. In short, once Catholics were allowed to participate in ruling Ireland, the system of ‘religio-racial’ oppression had to be abandoned. [40] There are problems with this analysis, not least in the explanatory framework where changes are seen as the result of intentional manoeuvres by the ruling class. It is also the case that Irish people in Britain continue to experience racism as the dominant form of oppression. Nevertheless, Allan is clearly right to note that the use of religion – an attribute that we would now regard as ‘ethnic’ – as the basis of racial identification was rare at the time. In a situation where the oppressed population was the same skin colour as the oppressors, this shift was probably inevitable.

The second precedent was in South Africa. One of the intellectual founders of apartheid (which means ‘separate development’) in South Africa was W.W.M. Eislen. As Adam Kuper points out, Eislen rejected the notions of black inferiority dominant among his countrymen: ‘Not race but culture was the true basis of difference, the sign of destiny.’ But, he said, although different cultures should be valued in their own right, their individual integrity should also be preserved: ‘If the integrity of traditional cultures were undermined, social disintegration would follow.’ Segregation of the races was necessary not to preserve unequal relations between white and black, but the cultural differences between them. [41] This was the theoretical basis on which apartheid was built.

What is disturbing, given these precedents, is that the notion of ‘ethnicity’, particularly when it is used in its cultural sense, has increasingly become a substitute for ‘race’, a coded way of reinventing racial categories without making skin colour the key issue, in similar ways to those pioneered in Ireland and South Africa. And it is not simply racists who are responsible for this. The Race Relations Act of 1976 defines a ‘racial group ... by reference to one or more of the following: colour, race, nationality (including citizenship) or ethnic or national origins’. [42] A Commission for Racial Equality publication setting out the Racial Equality Standard for local government in Scotland asks that ‘the Standard should be adopted and used both by authorities that have relatively substantial ethnic minority populations in their areas and those whose ethnic populations are smaller and more scattered’. [43] The assumption that ethnicity represented a minority deviation from a majority norm (is there an ‘ethnic majority’?) should in itself make us deeply suspicious, but it is only since the 1970s that the racist undercurrents of the term have become completely obvious. As Neville Alexander points out, quoting one of the US sociologists responsible for popularising the term during the 1940s, ‘ethnicity’ is useful ‘as a means of avoiding the word, yet retaining its meaning’. [44]

Alex Callinicos rightly argues that the ‘new’ racism has arisen as a result of the discrediting of the notion of biologically distinct races, partly as a result of advances in knowledge which have undermined any scientific basis for such beliefs, and partly (and one suspects far more) as a result of the use to which such beliefs were put during the Holocaust. (Hence the modern convention, which I have followed here, of placing the word ‘race’ in quotes, indicating that the concept is wholly ideological and has no referent in the world.) Callinicos also argues, however, that the ‘newness’ of this racism is more apparent than real, since biological racism and, related to this, ideas of black inferiority, are still very much alive and, in any case, the ‘new’ cultural racism often involves the same type of stereotyping as the ‘old’ biological racism. [45] There is some force in these criticisms. After reading the attempt by Charles Murray in The Bell Curve to explain black underachievement on the basis of genetic inheritance, or an attempt in The Independent to distinguish between Serbs and Albanians on the basis of their hair colour, it would be very foolish to predict the disappearance of biological racism. Nevertheless, there are reasons to believe that the dominant form of racist ideology is taking a new form in which questions of ‘ethnicity’ are central.

First, the emphasis on culture is not related to biology in the sense of indicating that some human beings are genetically superior or inferior to others, but in the sense that human beings are naturally hostile to those with different cultures: ‘… we may all share a common human nature, but part of that very shared nature is the natural tendency to form bounded social units and to differentiate ourselves from outsiders’. [46] This has become part of the discourse even of the extreme right. During the campaign which preceded the elections to the European Parliament on 10 June 1999, a British National Party leaflet intended for distribution across Scotland called for opposition to the ‘uprooting of our culture and to mass immigration’, and support for ‘the preservation of our unique Scottish identity within a free Britain’. Voters were invited to find out more about the fascist election campaign to save sterling and ‘to preserve the cultural and ethnic identity of Scotland and the British people’. [47] The object here is to bait the hook that catches the unwary with references to culture and ethnicity that make a point of not referring to racial stereotypes.

Second, and more importantly, we are seeing the naturalisation of ‘ethnic’ characteristics. Attributes or properties like religion or language that were once regarded as socially acquired and consequently amenable to change are increasingly being treated as if they were naturally occurring and permanent. Indeed, in the case of nationalism, Tom Nairn has argued that they are naturally occurring and permanent:

… differential cultural development (including language) may have had a function unsuspected by previous historians and theorists ... If internal species-diversity through cultural means has always been ‘human nature’, presumably it will go on being so – in a way that has nothing to do with blood or race. [48]

The idea supported by NATO and large sections of the liberal press – that different ‘ethnic’ groups in the Balkans ‘naturally’ want to separate themselves from other groups, if necessary by terror and expulsion – is clearly one practical application of this theory. As long as ethnicity is assumed to have a real existence, then the pressure is always there to accept the logic of these supposed differences – that all the states in the Balkans must have a single dominant ‘ethnic’ group, no matter what the cost to the other groups who might have lived there for as long or longer.

The same thinking lies behind the Northern Ireland peace agreement. Increasingly the language in which the conflict in the North is described abandons the notions of ‘religious sectarianism’ in which it was conducted for so long, and adopts that of ‘ethnic division’ instead. Religious belief may decline with secularisation, as it has across most of the British Isles, but if religion has become part of your very nature, then all you can do is keep the two sides (’communities’) hermetically sealed from each other. In other words, Protestant and Catholic religious beliefs have become the basis of ‘ethnic identities’, and these in turn are assumed to function in the same way as ‘racial characteristics’ once did. In addition to passing over the role that the British state has had in creating and maintaining the conflict, this ideological transformation also has another advantage for the ruling class in that it also absolves it from finding any permanent solution other than ‘peaceful coexistence’. The question of language illustrates both how the state pretends to be ‘even handed’ while supporting the Unionist position, and the way in which the social divisions in the North are treated as ‘cultural’.

Since 1968 there has been a revival of interest in Irish culture, and of the Gaelic language in particular, among Catholics in the North. Since this interest has rightly been associated with political Republicanism, or at the very least with the desire to assert a political identity in the face of a state which denies or marginalises it, the educational and cultural activities involved have tended to be organised and financed by the communities themselves rather than by the state. In 1994, however, the government sponsored Cultural Traditions Group expressed its concern that Gaelic was associated with Republicanism and arranged to fund a trust to enable Protestants to learn the language in settings where they would not be troubled by these associations. As Bill Rolston points out, this is not ‘symmetry’ but ‘an exercise in depoliticising a cultural movement. It is multiculturalism as counter-insurgency.’ The same Cultural Traditions Group also provides funding for the ‘Ulster Scots’ language, whose supporters set up the Ulster Scots society in, by a curious coincidence, 1994. [49] There is no such language (even ‘Scottish Scots’ is a dialect – in fact several dialects – of English) but such claims help to establish the myth that there are two parallel communities, with equivalent traditions, not a divided society in which one community is oppressed.

The difficulty is that if we lack a word to describe the victims of racism, since we reject the concept of ‘race’, then the concept of ethnicity seems to offer an alternative. If what I have argued here is correct, however, ethnicity is rapidly turning into the thing it was originally introduced to oppose. What Adam Kuper has written of contemporary US anthropology seems applicable to much of the left: ‘It repudiates the popular ideas that differences are natural, and that cultural identity must be grounded in a primordial, biological identity, but a rhetoric that places great emphasis on difference and identity is not best placed to counter these views. On the contrary, the insistence that radical difference can be observed between peoples best serves to sustain them’. [50]

‘Ethnic’ versus ‘civic’ nationalism

There is, however, another danger with the use of ‘ethnicity’. It is sometimes argued that ‘ethnic’ nationalisms that supposedly lead to the purging of entire populations as in Yugoslavia can be combated by an alternative ‘civic’ nationalism based on politics, not tribe. James Kellas describes this as ‘“inclusive“ in the sense that anyone can adopt that culture and join the nation, even if that person is not considered to be part of the “ethnic nation“.’ [51] ‘Civic’ nationalism is frequently presented as the only true form of nationalism. Certain nationalisms – like that of Serbia – are said to be inherently oppressive precisely because they are based on an ‘ethnic’ identity. The contrast is often made between this kind of nationalism and one described as ‘civic’ or ‘social’ – Scottish and Catalan nationalism, for example, are frequently described in this way, not least by Scottish and Catalan nationalists themselves.

George Kerevan, former Trotskyist and currently the SNP spokesperson for the environment, used his column in The Scotsman newspaper recently to distinguish the nationalism of his party from that of the Milosevic regime:

There is nationalism in the sense it applies to Hitler or Milosevic. Call it ethnic or tribal nationalism. In fact, don’t call it nationalism at all, because it’s not about building modern nations. This is a reactionary, tribal, exclusive ideology espoused in times of economic and political change by those social orders who are being usurped or threatened by the process of modernisation ... But there is another, totally different meaning of the word nationalism – nation building. Building the common institutions of an inclusive civil society that alone mobilises the talents, energies and co-operation of the population to create a modern industrial society. [52]

Note that nationalisms of which Kerevan disapproves (not least because they threaten to discredit his own nationalism by association) are dismissed as mere ‘tribalism’. Conversely, ‘civic’ nationalism is with equal frequency presented as not really being a form of nationalism at all. Only the ‘tribalism’ of Milosevic, which, as Michael Ignatieff puts it, ‘legitimises an appeal to blood loyalty’, is designated as such. In either event, the desired effect is to protect ‘civic’ nationalism from any suggestion that it appeals to blood and soil. [53] Marxists do distinguish between different forms of nationalism, in particular between those of oppressors and oppressed, but this is not what is being argued here. [54] What is interesting about the argument about ‘civic’ nationalism is that it is precisely the one that has historically been used to defend multinational oppressor nationalisms like those of Britain and the US.

During the Scottish parliamentary elections of May 1999 The Daily Record issued a warning to its readers that the nationalism of the Scottish National Party could lead to the type of brutality exercised by the Serbs against the Kosovan Albanians. Here the British nationalism supported by both The Daily Record and its party of choice, the British Labour Party, simply disappears from view, despite the fact that it has been used to mobilise support for actual, as opposed to hypothetical, bloodletting for nearly 300 years and has recently been doing so again in the Balkans. The notion that British nationalism is not ‘really’ nationalism at all is of course a venerable theme of ruling class ideologues. It was first systematically expressed by the historian Lord Acton in an article of 1862 where he argued that the multinational character of the British nation ensured that ‘freedom’ (in the economic sense understood by mid-Victorian liberals) was secure: ‘The combination of different nations in one state is as necessary a condition of civilised life as the combination of men in society.’ One of the benefits conferred by this arrangement was that the ‘intellectually superior’ would elevate ‘inferior races’ hitherto corrupted by despotism or democracy. How different this beneficent fusion was to the situation elsewhere in Europe: ‘Where political and national boundaries coincide, society ceases to advance, and nations relapse into a condition corresponding to that of men who renounce intercourse with their fellow men’. [55]

This analysis, or second or third hand versions of it, clearly informs the attitude of contemporary supporters of the British state, such as Gordon Brown, who are happy to dwell on their abhorrence of Scottish nationalism while simultaneously offering their support for British bombers whose sides are decorated with the Union Jack. Social psychologist Michael Billig has characterised the everyday nationalism of the established imperial states as ‘banal nationalism’: ‘“Our“ nationalism is not presented as nationalism, which is dangerously irrational, surplus and alien.’ Other people have nationalism; at best, ‘we’ have patriotism. Billig aims his comments specifically at the situation in the US, but they have a broader applicability: ‘The wars waged by US troops; the bombing in Vietnam and Iraq; the bombast of successive US presidents; and the endless display of the revered flag; all of these are removed from the problems of overheated nationalism’. [56] As these comments suggest, there are significant difficulties for socialists in attempting to use ‘civic’ nationalism as an alternative to ‘ethnic’ nationalism. Two in particular stand out.

The first is that the category of the ‘civic’ avoids any engagement with the fact that there are certain activities which nation states must undertake, regardless of how non-ethnic they may be. As Billig complains, Ignatieff ‘does not describe how “civic nationalists“ create a nation-state with its own myths; how the civic nations recruit their citizenry in wartime; how they draw their boundaries; how they demarcate “others“ behind those boundaries; how they resist, violently if necessary, those movements which seek to rearrange the boundaries; and so on.’

The second is that, as we have seen, ethnicities can be invented to categorise groups by their enemies, or as self identification by those groups themselves, without any reference to real or imaginary kinship relations: culture can just as easily be made the basis of ethnicity as blood and soil tribalism. Precisely because ethnicity is a socially constructed category, however, ethnic categorisations can be produced anywhere with the same disastrous results that we have seen for the last ten years in the Balkans. Consequently there is no reason why ‘civic’ nationalism cannot be transformed into ‘ethnic’ nationalism in its turn under certain determinate conditions, just as it was in Germany – a modern, developed and highly cultured capitalist society – during the 1930s. [57] This is a conclusion that adherents of ‘civic’ nationalism are, of course, most anxious to avoid.

The example of Scotland is worth considering in this context for two reasons: first, because the historical record demonstrates how even this most civil of societies first rose on a sea of ethnic blood; second, because the contemporary situation contains all the elements for an ‘ethnic’ nationalism to arise – and in this Scotland is no different from most other Western European nations, although it tends to evade the scrutiny to which English nationalism is rightly subject.

The modern Scottish nation was created through two processes: first, the destruction of the Highland society and the incorporation of its imagery into the national self image; second, the consolidation of that image through participation in the conquest and colonisation of North America and India. Both processes included ferocious episodes of what we would now call ‘ethnic cleansing’.

The Highlanders were considered to be no better than the Catholic Irish – indeed, their language and persons were often described in this way in both the Lowlands and England. One self proclaimed ‘gentleman’ of Derby, who had Highlanders quartered on him during the Jacobite occupation of that town in 1745, expressed every existing prejudice possible about the Highlanders in the space of one brief letter. First was their appearance: ‘Most of the men, after their entrance into my house, looked like so many fiends turned out of hell to ravage the kingdom and cut throats; and under their plaids nothing but various sorts of butchering weapons were to be seen.’ Even though these fiends in human form proceeded to eat and drink this gentleman out of house and home (although unaccountably failing to cut either his throat or those of his family), he could still find amusement in their religious observance: ‘What did afford me some matter for an unavoidable laughter (though my family was in a miserable condition) was to see these desperadoes, from officers to the common men, at their several meals, first pull off their bonnets, and then lift up their eyes in a most solemn manner, and mutter something to themselves, by way of saying grace – as if they had been so many primitive Christians.’ As if, indeed. His greatest abuse, however, is reserved for their language: ‘Their dialect (from the idea I had of it) seemed to me as if a herd of Hottentots, wild monkeys in a desert or vagrant Gypsies had been jabbering, screaming, and howling together; and really this jargon of speech was very suited to such a set of banditti’. [58] The conflation of ‘Hottentot’, ‘monkey’ and ‘Gypsy’ is suggestive and horrifying, but no different from what was commonly said about the Irish. And this is how they were treated in the aftermath of the Battle of Culloden. Discussing the brutality shown to the defeated Scottish Highlanders by the British army, the historian Alan Macinnes has written that the actions of the victorious Hanoverian troops involved ‘systematic state terrorism, characterised by a genocidal intent that verged on “ethnic cleansing“.’ [59] At the forefront of these atrocities were the Lowland Scots.

As the warrior vanguard of British imperialism, however, the Highlanders behaved no better than the Lowlanders or the English. The Native Americans, to whom the Highlanders have been so frequently and inaccurately compared, might have expected different treatment at their hands than was generally dispensed by settlers from elsewhere in the British Isles. Alas, this was not the case. There were individual examples of intermarriage, or even of Highlanders adopting Native American lifestyles, but as James Hunter writes, ‘Most North American Indian native peoples ... would have been hard pressed to distinguish between the behaviour of Scottish Highlanders or any other of the various types of European with whom they came into contact.’ In some cases this behaviour contained some particularly bitter ironies: ‘Emigrants to Cape Breton Island, many of them refugees from clearances ... showed not the slightest scruple about displacing the area’s traditional inhabitants, the Micmac, from territories the latter had occupied for much longer than there had been Gaelic speaking Scots in Scotland’. [60] Scotland was of course itself an imperial power, or, as an integral part of the British state, at least a major component of one. We are fortunate to have an excellent description of imperial rule in Asia by James Callender, a Scottish radical active during the 1780s and 1790s:

In Bengal only, we destroyed or expelled within the short period of six years no less than five million industrious and harmless people; and as we have been sovereigns in that country for about 35 years, it may be reasonably computed that we have strewn the plains of Indostan with 15 or 20 million carcasses ... The persons positively destroyed must, in whole, have exceeded 20 million, or 2,000 ... acts of homicide per annum. These victims have been sacrificed to the balance of power, and the balance of trade, the honour of the British flag. [61]

Nor is the type of racism associated with empire something of the distant past. As late as 1923 a committee of the Church of Scotland, asked to consider the effects of Irish immigration to Scotland, produced a report to the General Assembly in which the Catholic Irish were described as ‘a people by themselves, segregated by reason of their race, their customs, their traditions, and, above all, by their loyalty to their church’. The Scottish and Irish ‘races’ could never mix, or even live together, because:

The Irish are the most obedient children of the Church of Rome; the Scots stubbornly adhere to the principles of the reformed faith [Protestantism]. The Irish have separate schools for their children; they have their own clubs for recreation and social intercourse; they tend to segregate in communities, and even to monopolise certain departments of labour to the exclusion of Scots. [62]

It should be obvious, therefore, that it is historically inaccurate to claim that the Scottish nation has had a purely ‘civic’ national identity, and it is politically myopic to imagine that a full blown ‘ethnic’ nationalism could not re-emerge here under certain conditions. The materials are there in the traditions of Protestant sectarianism, militarism, or even simply ‘whiteness’. On the latter point it is worth noting that the Commission for Racial Equality reported in May 1999 that Scotland had 1,087 recorded racial incidents during 1997–1998, compared to 441 in Wales and 13,437 in England. Although Scotland is home to only 2.1 per cent of ‘ethnic minorities’ in Britain, it recorded 7.3 per cent of all racially motivated incidents. In central Scotland, where the majority of incidents were reported, the percentage was 15 times higher than in central London. None of these remarks are intended to contribute to The Daily Record style hysteria about Scottish nationalism. The chances of an ethnic national movement arising in the near future strike me as unlikely, and the Scottish National Party is equally unlikely to be a vehicle for such a nationalism should it arise, but it is necessary to remind ourselves that there are no nations on earth, be their nationalisms ever so ‘civic’, where ‘ethnic’ divisions could not be invented and ‘cleansing’ imposed if the material conditions were right.


It could be argued that I am displaying too great a concern with mere terminology and, given the way in which the academic left is currently obsessed with language, this would be an understandable response. Nevertheless, the dire political consequences that have previously followed the widespread adoption of certain terms (’patriarchy’, for example) tend to suggest that terminological shifts not only register changed ways of thinking, but also encourage such changes. As the Russian Marxist Valentin Voloshinov wrote, ‘The word is the most sensitive index of social changes’, and if, as Voloshinov also suggests, the word is ‘an arena of class struggle’, then it is high time that we began to wage it over the word ‘ethnicity’. [63]

In his recent book on culture, Adam Kuper concludes with sentiments that are equally relevant to the discussion of ethnicity:

… unless we separate out the various processes that are lumped together under the heading of culture, and look beyond the field of culture to other processes, then we will not get very far in understanding any of it. For the same sort of reason, cultural identity can never provide an adequate guide for living. We all have multiple identities, and even if I accept that I have a primary cultural identity, I may not want to conform to it. Besides, it may not be very practical. I operate in the market, live through my body, struggle in the grip of others. If I am to regard myself only as a cultural being, I allow myself very little room to manoeuvre, or to question the world in which I find myself.

Kuper notes that there is a final objection to defining ourselves in this way, which he describes as ‘moral’, but which is actually political: ‘It tends to draw attention away from what we have in common instead of encouraging us to communicate across national, ethnic and religious boundaries, and to venture between them’. [64] Although rendered in liberal individualistic terms, this is well said. For socialists, the aim is to overcome the divisions which are increasingly described as ‘ethnic’ by removing the oppressions that give them significance, not to perpetuate or add to them. This may mean supporting oppressed nations or peoples, but the notion of ‘ethnicity’ is ultimately a means of dividing people up into ever more arbitrary classifications. At best, under the guise of celebrating ‘cultural difference’, it only obscures what most working people, which is most people, have in common by emphasising relatively superficial aspects of our social world. At worst, in a struggle for scarce resources such as that currently being played out in the Balkans, it can be used as a means of marking down certain people for persecution. As I have tried to suggest, there is no reason why we in Britain should feel complacent about the implications of ‘ethnic cleansing’ for ourselves. The necessary elements of ‘ethnicity’ can always be assembled from whatever historical relics are lying around, if economic crisis and social collapse are sufficiently severe. The anthropologist Marcus Banks wrote recently of ethnicity: ‘Unfortunately ... it is too late to kill it off or pronounce ethnicity dead; the discourse on ethnicity has escaped from the academy and into the field’. [65] This is too pessimistic. To dispense with the concept, we must first dispense with the social conditions that require the thing to which it refers, but it is possible to make a start. To paraphrase Alasdair MacIntyre in another context, understanding the uses to which ‘ethnicity’ has been put leads comprehensively to the conclusion that it is a term which no honest person should continue to use. [66]


This article is based on a talk given at the Why Is NATO Bombing Yugoslavia? conference organised by the Edinburgh Campaign Against War In Europe and held at the University of Edinburgh on 23 May 1999. Thanks to Margaret MacDonald, Laura Mitchell and Gerry Mooney for helpful comments.

1. M. Banks, Ethnicity: Anthropological Constructions (London and New York 1996), pp. 167–168; Notes and Queries, The Guardian, 17 June 1999. The first area to be ‘cleansed’ in this way seems to have been the Croat village of Kijevo in the otherwise Serb dominated province of Krajina during August 1991. See L. Silber and A. Little, The Death Of Yugoslavia (Harmondsworth 1995), pp. 188–190.

2. See, for example, M. Mann, The Dark Side of Democracy: the Modern Tradition of Ethnic and Political Cleansing, New Left Review 235 (May–June 1999), pp. 31–32, 42–43.

3. V.I. Lenin, What Is To Be Done?, Selected Works, vol. 1 (Moscow 1975), p. 145.

4. P.L. van den Berghe, Race and Ethnicity: a Sociobiological Perspective, Ethnic and Racial Studies, vol. 1, no. 4 (October 1978), p. 404.

5. S. Reynolds, Medieval Origines Gentium and the Community of the Realm, History, vol. 68, no. 224 (October 1983), p. 379.

6. L. Colley, Britons (London 1992), p. 15.

7. M. Lynch, Scotland: A New History (Revised Edition, London 1992), p. 12. The origins of this process were famously mocked in the deathless prose of Sellar and Yeatman: ‘The Scots (originally Irish, but by now Scotch) were this time inhabiting Ireland, having driven the Irish (Picts) out of Scotland; while the Picts (originally Scots) were now Irish (living in brackets) and vice versa. It is essential to keep these distinctions clearly in mind (and vice versa).’ W.C. Sellar and R.J. Yeatman, 1066 And All That (Harmondsworth 1960), p. 13.

8. S. Rose, L.J. Kamin and R.C. Lewontin, Not In Our Genes (Harmondsworth 1984), p. 126.

9. E.R. Wolf, Europe and the Peoples Without History (Berkeley and Los Angeles 1982), p. 71.

10. The Catholic Irish are discussed below. For a brief discussion of how increasing social differences between Lowland and Highland Scotland led the inhabitants of the former region to regard those of the latter as a distinct ‘race’ or ‘ethnicity’, see N. Davidson, Scotland’s Bourgeois Revolution, in C. Bambery (ed.), Scotland, Class and Nation (London 1999), pp. 58–61, 120–121.

11. E.R. Wolf, op. cit., pp. 380, 381.

12. The third group, the Twas, are pygmies and consequently physically distinct from the other two, but in any event they comprised only 1 percent of the population at the time of colonisation.

13. C. Kimber, Coming to Terms With Barbarism in Rwanda and Burundi, International Socialism 73 (Winter 1996), pp. 128–129.

14. T.H. Eriksen, Ethnicity and Nationalism (London and Chicago, 1993), pp. 87–88.

15. Ibid., p. 82.

16. ‘The surest sign that a society has entered into the secure possession of a new concept’, writes Quentin Skinner, ‘is that a new vocabulary will be developed in terms of which the concept can then be publicly articulated and discussed.’ He takes the example of how the term ‘state’ emerged during the Reformation to describe ‘a form of public power separate from both the ruler and the ruled, and constituting a supreme political authority within a certain defined territory’. Yet the state existed for thousands of years before the word came into use. Q. Skinner, The Foundations Of Modern Political Thought, vol. 2 (Cambridge 1978), pp. 351–354.

17. A.D. Smith, National Identity (Harmondsworth, 1991), p. 21.

18. Ibid., p. 14.

19. T.H. Erickson, op. cit., p. 3. Their usage has clear affinities with the notion of a ‘status group’ introduced by Max Weber to describe differentiation at the social level, as opposed to occupational ‘class’ at the economic and ‘party’ interest at the political. See M. Weber, Class, Status, Party, in H.H. Gerth and C. Wright Mills (eds.), From Max Weber: Essays In Sociology (London 1948), pp. 186–194. Weber was sceptical about the usefulness of ‘ethnicity’ as a means of describing identity and tended to use the term more to describe the character of endogamous groups. For an example of this see, for example, Structures of Power, ibid., p. 173. It should be noted that under Stalinism the notion of ‘ethnicity’ was also put to work in similarly ideological ways. During the Brezhnev period in particular a group of Soviet anthropologists, the most important of whom was Yulian Bromley, developed a theory of ethnos which claimed that the constituent peoples of the republic retained an ethnic identity which had continuity throughout the feudal, capitalist and socialist epochs. The function of this theory, whose anti-Marxist content went some way to ensuring it a relatively friendly reception in the West, was to explain why various national characteristics of, say, the Armenians, had failed to wither away under ‘socialism’, by reference to an trans-historical ‘ethnicity’. At the same time the celebration of various ‘ethnicities’ could be used as a covert way of allowing the subject peoples of the Stalinist Empire a means of cultural expression without threatening the integrity of the state. For relatively sympathetic accounts of these theories, see M. Banks, op. cit., pp. 17–24; T. Shanin, Soviet Theories of Ethnicity: the Case of a Missing Term, New Left Review 158 (July–August 1986), pp. 118–122. John Foster, one of the few Stalinist in Britain to retain a respect for the outpourings of the Soviet Academy right till the end, attempted to put these theories of ethnos into effect with respect to Scotland in several articles where he claimed that Scottish nationhood was first formed during the feudal epoch. Such positions are of course standard among the nationalists who march to Bannockburn every year to celebrate the victory of the Scottish ‘nation’ over Edward II in 1314, although they have managed to reach this position without the assistance of Bromley et al. See, in particular, J. Foster, Nationality, Social Change and Class: Transformations of National Identity in Scotland, in D. McCrone, S. Kendrick and P. Straw (eds.), The Making of Scotland: Nation, Culture and Social Change (Edinburgh 1989), p. 40.

20. M. Banks, op. cit., p. 72. Banks suggests that the rise of Nazi Germany was a reason why Germans made less of their origins than, for example, Poles, but the ‘assimilationist’ nature of German immigrants was well established long before 1933.

21. L. Carroll, Through the Looking Glass and What Alice Found There, in The Penguin Complete Lewis Carroll (Harmondsworth 1982), p. 196.

22. S. Hall, Brave New World, Marxism Today, vol. 32, no. 10 (October 1988), p. 29.

23. A. Cohen, Two-Dimensional Man (London 1974), p. 15.

24. No Sizwe (N. Alexander), One Azania, One Nation (London 1978), p. 137. Max Weber made the same point, in more academic language, much earlier in the century: ‘All in all, the notion of “ethnically“ determined social action subsumes phenomena that a rigorous sociological analysis ... would have to distinguish carefully ... It is certain that in this process the collective term “ethnic“ would be abandoned, for it is unsuitable for a really rigorous analysis.’ And in the next paragraph Weber describes how the ‘concept of the ethnic group ... dissolves if we define our terms exactly’. See M. Weber, Economy And Society, vol. 1 (Berkeley and Los Angeles, 1968), p. 395.

25. S. Hall, Our Mongrel Selves, New Statesman and Society, 19 June 1992, p. 6.

26. S. Hall, The New Ethnicities, J. Hutchinson and A.D. Smith (eds.), op. cit., p. 163.

27. J. Hutchinson and A.D. Smith (eds), Preface, op. cit., pv.

28. T.H. Eriksen, op. cit., pp. 13–14.

29. M. Cross, Colonialism and Ethnicity: a Theory and Comparative Case Study, Ethnic And Racial Studies, vol. 1, no. 1 (January 1978), p. 40.

30. D. Brown, Why is the Nation-State so Vulnerable to Ethnic Nationalism?, Nations And Nationalism, vol. 4, no. 1 (January 1998), p. 13.

31. A. Kuper, Culture: The Anthropologist’s Account (London and Cambridge, Massachusetts 1999), pp. 238–239. Kuper is South African and correctly describes himself as a liberal. Like that other exiled son of the British dominions, the Australian Robert Hughes in The Culture Of Complaint, he treads a fine line between acute commentary on follies of much ‘radical’ thought and a dismissal of the problems it seeks to address.

32. D. Brown, op. cit., p. 13.

33. M. Glenny, The Fall Of Yugoslavia (Harmondsworth 1992), pp. 141–142.

34. A. Bell-Fialkoff, A Brief History of Ethnic Cleansing, Foreign Affairs, vol. 72, no. 3 (July 1993), p. 121.

35. E.J. Hobsbawm, Ethnicity and Nationalism in Europe Today, in G. Balakrishnan (ed.), Mapping the Nation (London and New York 1996), p. 260.

36. A. Sivanandan, R[ace] A[wareness] T[raining] and the Degradation of Black Struggle, Communities of Resistance (London and New York 1990), p. 94.

37. S. Fenton, Ethnicity: Racism, Class and Culture (Houndmills 1999), p. 69.

38. See P. Alexander, Racism, Resistance and Revolution (London 1987), ch 1.

39. M. Barker, The New Racism (London, 1981), pp. 4–5.

40. T.W. Allan, The Invention Of The White Race, vol. 1 (London and New York 1994), pp. 14–21, 32, 48, 65–66, 77–79, 92. The second volume of this important work has yet to appear.

41. A. Kuper, op. cit., pp. xii–xiii.

42. The Race Relations Act 1976, S.3 (1), (2).

43. H. Ousley, Preface, Commission for Racial Equality, Racial Equality Means Quality (London 1995), p. 5.

44. No Sizwe, op. cit., p. 137.

45. A. Callinicos, Race And Class (London, 1993), pp. 32–33.

46. M. Barker, op. cit., p. 5.

47. In fact, many members of the Communication Workers Union refused to deliver Nazi election material, but not, alas, those delivering to Edinburgh postal district EH6.

48. T. Nairn, Does Tomorrow Belong to the Bullets or the Bouquets?, New Statesman and Society, 19 June 1992, p. 31. For similar arguments, couched within an explicit appeal to sociobiology, see P.L. van den Berghe, op. cit..

49. B. Rolston, Culture as a Battlefield: Political Identity and the State in the North of Ireland, Race And Class, vol. 39, no. 4 (April–June 1998), pp. 28–30.

50. A. Kuper, op. cit., pp. 238–239.

51. J.G. Kellas, The Politics Of Nationalism And Ethnicity (Houndmills, 1998), p. 65.

52. G. Kerevan, Milosevic’s “Nationalism“ Is Not Ours, The Scotsman , 5 April 1999. Compare Lord Acton in 1862: ‘These two views of nationality, corresponding to the French and English systems, are connected in name only, and are in reality the opposite extremes of political thought.’ Lord Acton, Nationality, in G. Balakrishnan (ed.), op. cit., p. 29.

53. M. Ignatieff, Blood And Belonging (London 1993), p. 6.

54. See N. Davidson, In Perspective: Tom Nairn, International Socialism 82 (Spring 1999), pp. 124–125, and A. Callinicos, Marxism and the National Question, in C. Bambery (ed.), op. cit..

55. Lord Acton, op. cit., p. 31.

56. M. Billig, Banal Nationalism (London 1995), pp. 55, 57.

57. See N. Davidson, In Perspective: Tom Nairn, op. cit., p. 113; D. Gluckstein, The Nazis, Capitalism And The Working Class (London 1999), ch 1.

58. Extract of a Letter from a Gentleman at Derby, The Gentleman’s Magazine and Historical Chronicle, vol. 16 (January 1746), p. 16.

59. A.I. Macinnes, Clanship, Commerce and the House of Stuart, 1603–1788 (East Linton 1996), p. 211. See also p. 32. On p. 215 the Hanoverian High Command is no longer described as having a policy ‘verging’ on ‘ethnic cleansing’, but of ‘opting’ for it without qualification.

60. J. Hunter, A Dance Called America (Edinburgh 1994), p. 237. Hunter recounts one of the more positive encounters between the Highland Scots and the Native Americans, in this case the MacDonalds of Glencoe and the Nez Perce, in his Glencoe and the Indians (Edinburgh 1996).

61. J.T. Callender, The Political Progress of Great Britain; or an Impartial Account of the Principal Abuses in the Government of this Country from the Revolution in 1688 (Edinburgh 1792), pp. 1–2.

62. Report of Committee to Consider Overtures from the Presbytery of Glasgow and from the Synod of Glasgow and Ayr on Irish Immigrants and the Education (Scotland) Act 1918, 29 May 1923 (Edinburgh 1923); Report on the Schemes of the Church of Scotland with Legislative Acts Passed by the General Assembly (Edinburgh 1923), pp. 750, 760–761. Some academics such as Steven Bruce have argued that this was an unrepresentative committee ‘whose views and enthusiasm were not shared by the majority’. The key, according to Bruce, is that the church lacked the willingness to act (see S. Bruce, Sectarianism in Scotland: a Contemporary Assessment and Explanation, D. McCrone and A. Brown (eds.), The Scottish Government Yearbook 1988 (Edinburgh 1988), p. 165, n25). In fact the membership of the committee included both the Moderator and the Procurator, the leading officials of the Church of Scotland.

63. V.N. Voloshinov, Marxism And The Philosophy Of Language (London and Cambridge, Massachusetts 1986), pp. 19, 23.

64. A. Kuper, op. cit., p. 247.

65. M. Banks, op. cit., p. 189.

66. A. MacIntyre, “Ought“, Against the Self-Images of the Age (London 1971), p. 156.

Top of page

ISJ 2 Index | Main Newspaper Index

Encyclopedia of Trotskyism | Marxists’ Internet Archive

Last updated on 6.5.2012