CLIFFORD GEERT2
M Primordial and Civic Ties
The stultifying aura of conceptual ambiguity that surrounds the terms 'nation,' 'nationality,' and 'nationalism' has been extensively discussed and thoroughly deplored in almost every work chat has been concerned to attack the relationship between communal and political loyalties.1 But as the preferred remedy has been to adopt a theoretical eclecticism that, in its attempt to do justice to the muitifaceced nature of the problems involved, tends to confuse political, psychological, cultural, and demographic factors, actual reduction of that ambiguity has not proceeded very far. Thus a recent symposium on the Middle East refers indiscriminately to the efforts of the Arab League to destroy existing nation-state boundaries, those of the Sudan Government to unify a somewhat arbitrary and accidentally demarcated sovereign it3te, and those of the Azerin Turks to separate from Iran and join the Soviet Republic of Azerbaijan as 'nationalism. '* Operating with a similarly omnibus concept, Cofeman' sees Nigerians (or some of them) as displaying five different sorts of nationalism at once—'African/'Nigerian,' 'Regional,''Group,' and 'Cultural.' And Emerson* defines a nation as a 'terminal community—the largest community that, when the chips are down, effectively commands men's loyalty, overriding the claims both of the lesser communities within it and those that cut across it or potentially enfold it within a still greater society ,..,' which simply shifts the ambiguity from the term 'nation' to the
30   THE QUESTION OF DEFINITION
term 'loyalty,' as well as seeming to leave such questions as whether India, Indonesia, or Nigeria are nations to the determination of some future, unspecified historical crisis.
Some of this conceptual haze is burned away, however, if it is realized that the peoples of the new states are simultaneously animated by two powerful, thoroughly interdependent, yet distinct and often actually opposed motives— the desire to be recognized as responsible agents whose wishes, acts, hopes, and opinions 'matter,' and the desire to build an efficient, dynamic modem state. The one aim is to be noticed: it is a search for an identicy, and a demand that that identity be publicly acknowledged as having import, a social assertion of the self as 'being somebody in the world.'' The other aim is practical: it is a demand for progress, for a rising standard of living, more effective political order, greater social justice, and beyond that of 'playing a part in the larger arena of world politics,' of 'exercising influence among the nations.'" The two motives are, again, most intimately related, because citizenship in a truly modem state has more and more become the most broadly negotiable claim to personal significance, and because what Mazzini called the demand to exist and have a name is to such a great extent fired by a humiliating sense of exclusion from the important centers of power in world society. But they are not the same thing. They seem from different sources and respond to different pressures. It is, in fact, the tension between them that is one of the central driving forces in the national evolution of the new states; as it is, at the same time, one of the greatest obstacles to such evolution.
This tension takes a peculiarly severe and chronic form in the new states, both because of the great extent to which their peoples' sense of self remains bound up in the gross actualities of blood, race, language, locality, religion, or tradition, and because of the steadily accelerating importance in this century of the sovereign state as a positive instrument for the realization of collective aims. Multiethnic, usually multilinguistic, and sometimes multiracial, the populations of the new states tend to regard the immediate, concrete, and to them inherently meaningful sorting implicit in such 'natural' diversity as the substantial content of their individuality. To subordinate these specific and familiar identifications in favor of a generalized commitment to an overarching and somewhat alien civil order is to risk a loss of definition as an autonomous person, either through absorption into a culturally undifferentiated mass or, what is even worse, through domination by some other rival ethnic, racial, or linguistic community that is able to imbue that order with the temper of its own personality. But at the same time, all but the most unenlightened members of such societies are at least dimly aware—and their leaders are acutely aware—that the possibilities for social reform and material progress they so intensely desire and are so determined to achieve rest with increasing weight on their being enclosed in a reasonably large, independent, powerful, well-ordered polity. The insistence on recognition as someone who
is visible and matters and the will to be modem and dynamic thus tend to diverge, and much of the political process in the new states pivots around an heroic effort to keep them aligned,
A more exact phrasing of the nature of the problem involved here is that, considered as societies, the new states are abnormally susceptible to serious disaffection based on primordial attachments.' By a primordial attachment is meant one that stems from the 'givens'—or, more precisely, as culture is inevitably involved in such matters, the assumed givens'—of social existence: immediate contiguity and kin connection mainly, but beyond them the given-ness that stems from being bom into a particular religious community, speaking a particular language, or ev^n a dialect of a language, and following particular social practices. These congruities ofblood, speech, custom, and so on, are seen to have an ineffable, and at rimes overpowering, coerciveness in and of themselves. One is bound to one's kinsman, one's neighbor, ones fellow believer, ipso facto: as the result not merely of personal affection, practical necessity, common interest, or incurred obligation, bur at least in great part by virtue of some unaccountable absolute import attributed to the very tie itself. The general strength of such primordial bonds, and the rypes of them that are important, differ from person to person, from society to society, and from time to time. But for virtually every person, in every society, at almost all times, some attachments seem to flow more from a sense of natural—some would say spiritual—affinity than from social interaction.
In modern societies the lifting of such ties to the level of political supremacy—though it has, of course, occurred and may again occur—has more and more come to be deplored as pathological. To an increasing degree national unity is maintained not by calls to blood and land but by a vague, intermittent, and routine allegiance to a civil state, supplemented to a greater or lesser extent by governmental use of police powers and ideological exhortation. The havoc wreaked, both upon themselves and others, by those modem (or semi-modern) states that did passionately seek to become primordial rather than civil political communities, as well as a growing realization of the practical advantages of a wider-ranging pattern of social integration than primordial ties cm usually produce or even permit, have only strengthened the reluctance publicly to advance race, language, religion, and the like as bases for the ikfinition of a terminal community. But in modernizing societies, where the tradition of civil politics is weak and where the technical requirements for an ctTective welfare government are poorly understood, primordial attachments tend, as Nehru discovered, to be repeatedly, in some cases almost continually, proposed and widely acclaimed as preferred bases for the demarcation of autonomous political units. L - ■ ■ i
It is this crystallization of a direct conflict between primordial and civil mmriments—this 'longing not to belong to any other group'—that gives to the
n   THE QUESTION OF DEFINITION
CLIFFORD GEERTZ ii
problem variously called tribalism, parochialism, communalism, and so on, a more ominous and deeply threatening quality than most of the other, also very serious and intractable problems the new states face. Here we have not just competing luyalties, but competing loyalties of the same general order, on the same level of integration. There are many other competing loyalties in the new states, as in any state—ties to class, party, business, union, profession, or whatever. But groups formed of such ties are virtually never considered as possible self-standing, maximal social units, as candidates for nationhood. Conflicts among them occur only within a more or less fully accepted terminal community whose political integrity they do not, as a rule, put into question. No matter how severe they become they do not threaten, at least not intentionally, its existence as such. They threaten governments, or even forms of government, but they rarely at best—and then usually when they have become infused with primordial sentiments—threaten to undermine the nation itself, because they do not involve alternative definitions of what the nation is, of what its scope of reference is. Economic or class or intellectual disaffection threatens revolution, but disaffection based on race, language, or culture threatens partition, inedemism, or merget, a redrawing of the very limits of the state, a new definition of its domain. Civil discontent finds its natural outlet in the seizing, legally or illegally, of the state apparatus. Primordial discontent strives more deeply and is satisfied less easily. If severe enough, it wants not just Sukarno's or Nehru's or Moulay Hasan's head it wants Indonesia's or India's or Morocco's.
The actual foci around which such discontent tends to crystallize are various, and in any given case several arc usually involved concurrently, sometimes at cross-purposes with one another, On a merely descriptive level they are, nevertheless, fairly readily enumerable:6
1) Assumed Bfood Ties, Here the denning element in quasi-kinship, 'Quasi' because kin units formed around known biological relationship (extended families, lineages, and so on) are too small for even the most tradition-bound to regard them as having more than limited significance, and the referent is, consequently, to a notion of untraceable but yet sociologically real kinship, as in a tribe. Nigeria, the Congo, and the greater part of sub-Saharan Africa are characterized by a prominence of this son ofprimordialism. But so also are the nomads or seminomads of the Middle East—the Kurds, Baluchis, Pathans, and so on; the Nagas, Mundas, Santals, and so on, of India; and most of the so-called 'hill tribes' of Southeast Asia.
2) Race. Clearly, race is similar to assumed kinship, in that it involves an ethnobiological theory. But it is not quite the same thing. Here, the reference is to phenorypical physical features—especially, of course, skin color, but also facial form, stature, hair type, and so on—rather than any very definite sense of common descent as such. The communal problems of Malaya in large part focus around these sorts of differences, between, in fact, two phenotypically
very similar Mongoloid peoples, 'Negritu.de* clearly draws much, though perhaps not all, of its force from the notion of race as a significant primordial property, and the pariah commercial minorities—like the Chinese in Southeast Asia or the Indians and Lebanese in Africa—are similarly demarcated.
3) Language. Linguism—for some yer to be adequately explained reasons—is particularly intense in the Indian subcontinent, has been something of an issue in Malaya, and has appeared sporadically elsewhere. But as language has sometimes been held to be the altogether essential axis of nationality conflicts, it is worth stressing that linguism is not an inevitable outcome of linguistic diversity. As indeed kinship, race, and the other factors to be listed below, language differences need not in themselves be particularly divisive: they have not been so for the most part in Tanganyika, Iran (not a new state in the strict sense, perhaps), the Philippines, or even in Indonesia, where despite a great confusion of tongues linguistic conflict seems to be the one social problem the country has somehow omitted to demonstrate in extreme form. Furthermore, primordial conflicts can occur where no marked linguistic differences are involved, as in Lebanon, among the various sorts of Batak-speakers in Indonesia, and to a lesser extent perhaps between the Fulani and Hausa in northern Nigeria.
A) Region. Although a factor nearly everywhere, regionalism naturally tends 10 be especially troublesome in geographically heterogeneous areas, Tonkin, Annam, and Cochin in preparcitioned Vietnam, the two baskets on the long pole, were opposed almost purely in regional terms, sharing language, culture, race, etc. The tension between East and West Pakistan involves differences in language and culture too. but the geographic clement is of great prominence owing co the territorial discontinuity of the country. Java versus the Outer Islands in archipelagic Indonesia; the Northeast versus the West Coast in mountain-bisected Malaya, are perhaps other examples in which regionalism has been an important primordial factor in national politics.
5) Religion. Indian partition is the outstanding case of the operation of this type of attachment. But Lebanon, the Karens and the Moslem Arakenese in Burma, the Toba Bataks, Ambonese, and Minahassans in Indonesia, the Moros in the Philippines, the Sikhs in Indian Punjab and the Ahmadiyas in Pakistani, and the Hausa in Nigeria are other well-known examples of its force in undermining or inhibiting a comprehensive civil sense.
6) Custom. Again, differences in custom form a basis for a certain amount of national disunity almost everywhere, and are of especial prominence in those cases in which an intellectually and/or artistically rather sophisticated group sees itself as the bearer of a 'civilization' amid a largely barbarian population that would be we!! advised to model itself upon it: the Bengalis in India, the Javanese in Indonesia, the Arabs (as against the Berbers) in Morocco, the Amhara in—another 'old' new state—Ethiopia, etc. But it is important also to point out that even vitally opposed groups may differ rather little in their
general style of life: Hindu Gujeratis and Maharashtrians in India; Baganda and Bunyoro in Uganda; Javanese and Sundanese in Indonesia. And the reverse holds also: the Balinese have far and away the most divergent pattern of customs in indonesia, but they have been, so far, notable for the absence of any sense of primordial discontent at all.
[The Integrative Revolution: Primordial Sentiments and Civil Politics in the New States', in Clifford Geenz (ed.), Old Societies and New States: TheQveslfor Modernity in Aim and Africa (Free Press: New York, 1963). 107-13.]
Bfjj   ANTHONY GIDDENS
ES  The Nation as Power-Container
A 'nation', as I use the term here, only exists when a state has a unified administrative reach over the territory over which its sovereignty is claimed. The development of a plurality of nations is basic to the centralization and administrative expansion ofstate domination internally, since the fixing of borders depends upon the reflexive ordering of a state system. We can follow Jones in recognizing four aspects of the transformation of frontiers into borders.1 These he calls allocation, delimitation, demarcation and administration.
The first refers to a collaborative political decision taken among states about the distribution of territory between them. Delimitation concerns the identification of specific border sites/ Demarcation in Jones's scheme—written as a guide for policy-makers and not just an academic study—refers to how borders are actually marked on the physical environment. Many borders, even within the heart of Europe today, are not demarcated. That apparent modern equivalent of the walls built by traditional states, the Berlin Wall, is an anomaly because it symbolizes the failure of a modem state to exert the level of administrative control over its population which its governing authorities deem proper and necessary. The border between East and West Germany must be one of the most highly 'administered', in Jones's terms, in the world. That is to say, a high degree of direct surveillance is maintained along it. Traditional states sometimes constructed frontier posts, demanding payment, and occasionally documentation, of those who travelled through. But where these existed they were usually, in fact, at divisions between provinces rather than between states as such. The coupling of direct and indirect surveillance (customs officials and frontier guards, plus the central co-ordination of passport information) is one of the distinctive features of the nation-state.
A nation-state is, therefore, a bordered power-container—as I shall argue, the pre-eminent power-container of the modem era, [.. .} [Ajmong other
things it involves processes of urban transformation and the internal pacification of states. These are phenomena that go together with the creation of generalized 'deviance' as a category and with processes of sequestration. All traditional states have laid claim to the formalized monopoly over the means of violence within their territories. But it is only within nation-states that this claim characteristically becomes more or less successful. The progress of internal pacification is closely connected with such success—they are, as it were, different sides of the same process,
The objection may be raised that there are very many instances, even in current times, of states whose monopoly of the means of violence is chronically threatened from within by armed groups; that insurgent movements, often poorly armed and organized compared with state authorities, have sometimes challenged and overthrown those authorities; and that there are diffuse levels of violence in minor contexts of even the most politically quiescent societies (crimes of violence, domestic violence and so on). None of these, however, compromise the point at issue, which concerns a comparison between nation-states and traditional states. There are circumstances in which civil war, involving chronic confrontations between armed movements or coalitions of more or less equal strength, have been quite protracted. However, not only are such circumstances highly unusual, the very existence of civil war' presumes a norm of a monopolistic state authority. By contrast, conditions which in a modem state would be defined as examples of 'civil war', that is, divisive 'internal' armed struggles, have been typical of all class-divided societies for very long periods. Again, armed groups or movements today are almost always oriented to the assumption of state power, either by taking over an existing state's territory or by dividing up a territory and establishing a separate state. Such organizations do not and cannot 'opt out' from involvement in state power one way or another as frequently happened in traditional states. Finally, 1 have no wish to underplay the importance or extent of violence that takes place in small-scale contexts in modem societies. But I am principally concerned with the means of violence associated with the activities ul'organized armed forces, not with violence as a more blanket category of the doing of physical harm to others.
Collecting together the implications of the foregoing observations, we can .irrive at the following concept of the nation-state, which holds for all variants .ind is not intrinsically bound to any particular characterization of nationalism, i . . | The nation-state, which exists in a complex of other nation-states, is a set of institutional forms of governance maintaining an administrative monopoly over a territory with demarcated boundaries (borders), its rule being sanctioned by law and direct control of the means of internal and external violence.'
\A Contemporary Critique of ttiitorictxt Materialism, li. The Nanon-Siaie and Violence (Polity
Press: Cambridge. 1585), 119-21.J
Wm A Soclc-Blological Perspective
The most basic question asked by socrobiology as well as sociology is: why are animals social, that is, why do they cooperate? Why are some species more soda! than others? The answer was long intuitively known: animals are social to the extent that cooperation is mutually beneficial. What saciobiology does is supply the main genetic mechanism for animal sociality, namely kin selection to maximize inclusive fitness. Natural selection operates through differential reproduction. Different alleles of the same gene compete with each other, and the ones that are carried by the more reproductively successful individuals have a greater probability of being replicated in the population's next generation. The successful alleles are the ones which, in a given environment, favor the reproductive success or 'fitness' of their carriers.
The great theoretical contribution of sociobiology has been ro escend the concept of fitness to that of'inclusive fitness'.1 Indeed, an animal can duplicate its genes directly through its own reproduction, qi indirectly through the reproduction of relatives with which it shares specific proportions of genes. Animals, therefore, can be expected to behave cooperatively, and thereby enhance each other's fitness to the extent that they are genetically related. This is what is meant by kin selection.1 Animals, in short, are nepotistic. i.e. they prefer kin over non-kin, and close kin over distant kin. This may happen consciously, as in humans, or, more commonly, unconsciously. Kin selection does not presuppose consciousness in order to be operative,
The propensity to be 'altruistic' i.e. to contribute to alter's fitness at the expense of ego's fitness, is directly proportional not only ra the coefficient of relatedness between ego and alter, but also to the benefit/cost ratio of the altruistic act. To use a human example, a post-menopausal mother could be expected to sacrifice her life more readily for a young adult child about to reproduce than a young mother to forego her life for the benefit of her first foetus. The genetic relationship is the same in both cases (namely, one half), but the fitness cost is low in the first case, high in the second Altruism, then, is directed mostly at kin, especially dose kin, and is, in fact, a misnomer. It represents the ultimate form of genetic selfishness. It is but the blind expression of inclusive fitness maximization. In fact, a simple formula leads one to predict
that 'altruism' can be expected if the cost/benefit ratio of the transaction is smaller than the coefficient of relatedness between alter and ego.
There is no reason to doubt that kin selection is a powerful cement af sociality in humans as it is in other animals. Vet, it is also clear that kin selection does not explain all of human sociality. There are, in my view, two additional bases of human sociality: reciprocity and coercion. Rudimentary forms of these are also present in many animals, but human forms of reciprocity and coercion greatly over-shadow in complexity and importance anything we know in other species. Not surprisingly, therefore, even the simplest and smallest human societies, though far less 'perfect' than those of the social insects (termites, ants, bees, wasps), are much more complex than those of any other known species. Reciprocity is cooperation for mutual benefit, and with expectation of return, and it can operate between kin or between non-kin. Coercion is the use of force for onesided benefit, that is, for purposes of intra-specific parasitism or predanon. All human societies continue to be organized on the basis of all three principles of sociality: kin selection, reciprocity, and coercion. However, the larger and the more complex a society becomes, the greater the importance of reciprocity, and. wirh the emergence of the state, coercion becomes in relation ro kin selection.
This is the baresr sketch of an argument which [ . . ] seeks to reduce individual behavior, social structure and cultural superstructure to the competition for scarce resources between individual organisms, each one acting, consciously or unconsciously, to maximize its gains or minimize its losses. This view of human affairs is sufficiently at variance with much of contemporary social science to arouse passionate rejection as a return to simplistic instinct theory, biological reduaionism, speculative evolutionism, social Darwinism, racism, hereditarianism, and so on, [... 1 Suffice it to say that sociobiology is indeed reductionist (as all modern science), evolutionist (as all modem biology), and materialist (as much good social science), but that it is emphatically not a return to social Darwinism, instinct theories or racism, and that it does not belittle the importance of environmental factors, the unique characteristics of Homo sapiens, and the significance of human culture. It merely asserts in the mosrundagmatic fashion that human behavior is the product of a long process of adaptive evolution that involved the complex interplay of genorypical, ecological and cultural factors.
How do these prolegomena relate to race and ethnicity? My centra] thesis is that both ethnicity and'race' (in the social sense) are. in fact, extensions of the idiom of kinship, and that, therefore, ethnic and race sentiments are to be understood as an extended and attenuated form of kin selection. Class relations, on the other hand, are in the realm of reciprocity, and are therefore of a fundamentally different nature. In more genera) form, 1 am suggesting thar there are two broad types of human collectivities: the ones that I shall call Type I tend to be ascriprive, defined by common descent, generally hereditary, and
98   THEORIES OF NATIONALISM
often endogamous, and those of Type II that are joined in the defense of common interests Type I includes racial, caste and ethnic groups, while Type II encompasses such varied associations as trade unions, political parties, professional bodies, sports clubs, neighborhood groups, parent-teacher associations, and so on. Empirically, of" course, a group may have mixed characteristics, as an ethnic ally-based political party, or a hereditary occupational guild, Nevertheless, in rheir ideal-typical form, each kind of group has a clearly distinct basis of solidarity: kinship and interest respectively.
Type 1 groups are generally preferentially or prescript!vely endogamous, but internally subdivided into exogamous kin groups: nuclear families, lineages, clans, kindreds. Indeed, until the last few thousand years of human history. Type i groups were synonymous with human societies. They were small in-bred populations of a few hundred individuals, prototypical 'tribes' that regarded themselves as 'the people', sharing cammon descent, real or putative, and as children of the mythical founder couple or creator god. Members of the tribe, though subdivided into smaller kin groups, saw themselves as a single people, solidary against the outside worid, and interlinked by a web of kinship and marriage making the tribe in fact a superfamily. A high rate ofinbreeding insured tfiat most spouses were also kinsmen. The cultural inventions of unilineal descent and lineage exogamy permitted the extension of that primordial model of social organization ro much larger societies running into the tens of thousands of people, and yet where Type 11 organizations were almost totally absent (with the exception of age sets).
Ethnic groups, for nearly al! of human history, were what geneticists call breeding populations, in-breeding superfamilies, in fact, which not only were much more closely related to each other than to even their closest neighbors, but which, almost without exception, explicitly recognized that fact, and maintained clear territorial and social boundaries with other such ethnic groups. This is, of course, not to deny that migration, conquesr, and interbreeding took place with some regularity, and thus that the common ancestry of'the people' was a J way? partially fictjve. But this was also true of smaller kin groups: the pater is not necessarily the progenitor. That the extended kinship of the ethnic group was sometimes putative rather than real was not the important point. Just as in the smaller km units, the kinship was real often enough to become the basis of these powerful sentiments we call nationalism, tribalism, racism, and etnnocenrrism, The ease and speed with which these sentiments can be mobilized even in modern industrial societies where they have to compere with many Type II groups, rhe blind ferocity of the conflicts to which these sentiments can lead, the imperviousness of such sentiments to rational arguments are but a few indications of their continued vitality and rheir primordialtty.
What 1 am suggesting is chat ethnocemrism evolved during millions, or at least hundreds of thousands of years as an extension of kin selection. Redpro-
city was also involved, especially in the exchange of women in marriage, but as spouses were typically also kinsmen there was no sharp distinction between kin selection and reciprocity. As hominids became increasingly formidable competitors and predators to their own and closely related species, there was a strong selective pressure for rhe formation of larger and more powerful groups. Croup size in hunting and gathering societies was, of course, severely constrained by ecological factors, but, still, there was an obvious selective advantage far kin groups to form those solidary superfamilies we call tribes; this, in turn, as Bigelow1 so clearly argues, necessarily meant organising against other competing groups, and therefore maintaining and defending ethnic boundaries.
Of Type II groups, little needs to be jaidhere With the exception of age sets, they tend to be characteristic of larger, more complex, state-organized societies, and therefore to have arisen much later in human evolution, and to be more exclusively cultural. They are. of course, also important, especially in industrial societies, but they are not primordial, they can be more readily formed and disbanded, they arc more amenable to cool, rational calculations of interest, and they do not as readily unleash orgies of passion. Nor, of course, have they stamped out Type I groups. Another fundamental difference between Type I and Type II groups u that the former tend to be mutually exclusive in membership and thus to form the basis of most primary relationships, while the latter are segmental, and non-murually exclusive. Millions of people in individual societies belong to a multiplicity of Type II groups, few of which involve them very deeply ot permanently. Some people are ethnically alienated, marginal or mobile or they are the product of mixed marriages, but most people belong to a single ethnic group or sub-group, and remain there far life. Even allowing for all the complications of the real world, and the existence of mixed-type groups, the categorical distinction remains nevertheless quite striking
Let us return to Type I groups, our special concern here. I have suggested that they evolved as an extension of kin selection, and thus probably have a partial biological basis, in the same sense as human kinship systems are rooted in biology, This contention is, of course, hotly contested by anthropologists such as Sablins,* who counter that human kinship is cultural, not biological. Almost every aspect oFhuman behavior takes a cultural form, from sneezing and defecating to writing poetry and riding a motorcycle. But this is not to say that some of these things do not also have a biological basis. I am definitely not arguing that we have a gene for ethnocentrism, or for recognizing kin; rather 1 am arguing that those societies thai institutionalized norms of nepotism and ethnocemrism had a strong selective advanrage over those that did not (assuming that any such ever existed), because kin selection has been the basic blueprint for animal sociality. To explain the universality of ethnocentrism and kinship otganization in human societies by invoking culture is completely
100   THEOlllCS 01' NATIONALISM
question begging. Culture is merely a proximate explanation of why people behave ethnoceniricaDv and neporisrically. As every ethnographer knows, when natives are asked why they behave a certain way, they answer: because it is the custom. The anthropologist then translates: because of his culture; the sociologist says: because he has been socialized into the norms of his society: and the psychologist counters because of his learning experiences All of them are right as far as they go, bur none of them has explained why all human societies practice kin selection and are ethnocentric.
So far. 1 hive stressed ethnicity rather than race or caste in my treatment of Type I groups, Caste is a very special case, limited, even if one adopts a wide definition of the teim, to highly differentiated, stratified societies, and may be considered an extreme case of the grafting of the principle of occupational specialization into what is basically a Type 1 group. Castes are not unique in being ocrupaiionally specialized Type I groups. Ethnic and racial groups also tend to become so.1 Castes are merely extreme «ses of occupational specialization linked with rigid endogamy and hierarchization
Race is a different matter. First. [ should make it clear that, even though 1 have presented a partially biological argument, 1 am most emphatically not using the word 'race' in the sense of a sub-species of Homo sapiens. Instead, I mean by 'race' the social definition which it is variously ascribed in different societies. Social race typically seizes on biologically trivial phenorypes, and, equally typically, corresponds only very imperfectly with genetically isolated populations. It thus has no intrinsic biological significance, as indicated by the face that only a few of the world's societies use primarily morphological phenorypes to define themselves, and to differentiate outsiders.
Ai first blush, this would seem to invalidate my argument thai ethnic and racial sentiments represent an extension of kin selection. If chat is the case, why should most human societies seize primarily on such obviously culturally transmitted traits such as language and dialect, religious beliefs, dress, hair styles, manners, scarifications, and the like as badges of group recognition and membership? If the name of the game is to identify kinsmen in order to enhance one's inclusive fitness, then why are not inherited physical characteristics chosen as recognition signals, rather than acquired cultural traits? Sometimes, of course, morphological phenorypes such as skin color, facial features, stature, hair texture, eye color, and so on are used, not only to define group membership, but also, within the group, as tests of ever-questionable paternity. Generally, however, cultural criteria of membership are far more salient than physical ones, if the latter are used at all. Societies that stress physical phenorypes more than cultural traits ire exceptional. Why*
The answer must again be sought in our evolutionary history. Until the last few millennia, thar is. uniil the rise of conquest states, sudden, large-scale, human migration was rare, and human breeding populations were small. There was migration and interbreeding, but on an individual scale, and mostly
berween neighboring groups. The result was chat neighboring populations were typically not sharply discontinuous in their genetic composition. The relatwe proportions of alleles of the same gene often constituted a gradient as one travelled through several breeding populations. Eye color in Europe would be a good example The further north one goes, from, say, Sicily to Sweden, the higher the proportion of lightly pigmented eyes. Vet, at no point in the journey is thete a noticeable discontinuity Eye color, therefore, is a poor criterion of national membership in Europe. Indeed, it varies much more within national groups, and indeed even within families, than between groups.
Now, Eurgpeans do use some morphological phenorypes to distinguish various ethnic groups. They speak loosely of 'Nordic', 'Mediterranean', 'Jewish', and so on. types. In the absence of any other clue, probabilistic guesses are often made on the basis of physical appearance as to a stranger's ethnic origin. Most groups probably have what Hoetink termed a 'somatic norm image.'4 that is. a mental picture of what a model group member looks like. The point, however, is that morphological phenorypes tend to be used either in the absence of more reliable cultural clues [such as language), or when physical appearance 15 widely discrepant from the somatic norm image (as, for instance, in Europe with Asians or Africans)
A good test of group membership for the purpose of assessing kin related-ness must meet the basic requirement of discriminating more reliably between groups than within groups. That is, the criterion chosen must show more ititrrgroup than itura-group variance. Until recently, cultural criteria met that condition far more reliably than physical ones. The problem was for smaU groups to distinguish themselves from their immediate neighbors, not with unknown populations thousands of kilometers away. Even the most trivial differences of accent, dialecr, vocabulary, body adornment, and soon, could be used far more reliably to assess biological relatedness or unrelatedness than any physical phenocype.' Therefore, whatever test was easiest to apply and correlated best with kin relatedness was used. That the correlation was spurious did not matter. What mattered was that it discriminated accurately.
This theory accounts not only for the general prevalence of cultural diacrid-ca in assessing group membership. It also accounts for the appearance of racism when and where it does occur better than any competing theory The kin selection argument predicts that physical criteria will be salient to the extent that they do a good and easy job of orscrimmating km and non-lcin. This obviously occurs in the aftermath ot large-scale, long-distance migration., whether through conquest, incursions, slavery, indenture, or voluntary immigration. The colonial expansion ofEurope beginning some five centuries ago, and all of the massive population transfers it brought in its wake are, of course, the overwhelmingly important genetic event of our species Predictably, 11 brought ahout a great surge in racism, because all of a sudden, it became possible to make a fairly accurate kin selection judgment from a distance of
102   THfcUKlea ur
several hundred meters. The Dutchman at the Cape, the Portuguese in Brazil, the Englishman in Kenya did not have to askqucstions and pickup subtle clues of accent to detect kin relatedness. By using a simple test of skin pigmentation he could literally shoot and askqucstions later at little risk of killing a kinsman.
[...}
We suggested at the outset that there were three main mechanisms of human sociality: kin selection, reciprocity and coercion. Ethnic and racial groups command our unreasoned loyalty because they arc in fact, or at least in theory, stipeifamibes Rut ethnic and race relations arc not only relations of cooperation and amity with the in-group; they ate equally importantly rela< tions of competition and conflict between groups. While intra-group relations are primarily dictated by kin selection, real or putative, intergroup relations are typically antagonistic. Occasionally, ethnic groups may enter a symbiotic, mutually beneficial relationship based, for instance, on the exploitation of two specialized and noncompetitive mches in the same habitat. Relations berween some pastoralist and sedentary groups are of this type. More commonly, there is open competition for, and conflict over scarce resources, and not infrequently the establishment of multi ethnic states dominated by one ethnic group at the expense of others Coercion then becomes the basis of mterethnic (or inrer-racial) relations.
Unlike kin selection and reciprocity which tequire no justification because they contribute to the fitness of all actors in the system, coercion, which leads to asymmetrical parasitism, often does attempt to legitimate itself. Interestingly, there are but two basic ideologies in support of coercion. One seeks to disguise coercion is kin selection, and here we have the many brands of paternalism and familism that have been used cojustify nearly allpre-industrial forms of despotism. The other attempts to present coercion as reciprocity and exchange, it is characteristic of the various 'democratic' ideologies of industrial societies in the last two centuries, from liberalism to socialism. Why this ideological shift from paternalism to liberie, igahte, fraumite in justifying tyranny during the last two centuries?
Perhaps this ideological shift reflects in part the increasing incorporation of small nation-states into multi-national states. Paternalism is a peculiarly well suited ideology for the small, ethnically homogeneous nation-state. Not surprisingly, it was independently reinvented in societies as far distant as China, japan, Inca Peru, Tzarist Russia, Ancient Egypt. Ottoman Turkey. Renaissance Europe and countless African kingdoms. Paternalism works in monoethnic states because the very concept of the nation is an extension of kin selection. For the same reason, it breaks down in multi-ethnic states. It was one thing for the Japanese peasant to look on his emperor as a divine super-father, the living incarnation of Nippon, quite another for the Hindu peasant to regard that polluted beef eater, Queen Victoria, as the living symbol of Mother India. An ideology based on reciprocity, on the other hand, can transcend ethnic bound-
aries, h is therefore a suitable cine for the 90 per cent of the world's states which are multi-ethnic conglomerates, and, furthermore, being ethnically neutral, it exports remarkably well as revolutionary ideology. It is no accident that France launched into the most imperialistic phase of its history immediately after the Revolution
The ideas sketched here are still tentative. They do nor so much supplant other theories of ethnicity and race a? supplement them by putting Them in the broader context of evolutionary thinking They do not purport to explain everything about these phenomena; they do not predict detailed historical occurrences, nor account for subtle cultural differences. They do, however, suggest parsimonious hypotheses to account for features of race and ethnicity which had hitherto remained elusive and problematic. Their plausibility to the reader hinges on whether he accepts the most fundamental paradigm for the evolution of different life forms and societal organization on our planet, Darwinian evolutionary theory, and on whether he is willing to apply that enormously successful model to our own species, or prefers to invoke an act of Special creation for mankind.
mm   ERNEST GELLNER
fUl  Nationalism and High Cultures
[But] nationalism is not the awakening of an old, latent, dormant force, though that is how it does indeed present itself. It is in reality me consequence of a new form of social organization, based on deeply internalized, education-dependent high cultures, each protected by its own state, it uses some of the p re-exisrenr cultures, generally transforming them in the process, but it cannot possibly use them all. There are too many of them. A viable higher culture-sustaining modern state cannot fall below a certain minimal size (unless in effect parasitic on its neighbours); and there is only room for a limited number of such states on this earth.
The high ratio of determined slumberers, who will not rise and shine and who refuse to be woken, enables us to turn the tables on nationalism-as-seen-by-itself. Nationalism sees itself as a natural and universal ordering of the political life of mankind, only obscured by that long, persistent and mysterious somnolence. As Hegel expressed this vision: 'Nations may have had a long history before they finally reach their destination—that of forming themselves into states'.1 Hegel immediately goes on to suggest that this pre-state period is really 'pre-historical' (sic): so it would seem that on this view the real history of i nation only begins when it acquires its own state, If we invoke the sleeping-beauty nations, neither possessing a state nor feeling the lack of it, against the nationalist doctrine, we tacitly accept its social metaphysic, which sees nations as the bricks of which mankind is made up. Critics of nationalism who denounce the political movement but tacitly accept the existence of nations, do nut go far enough. Nations as a natural, Cod-given way of classifying men, as an inherent though long-delayed political destiny, are a myth; nationalism.
64
THEORIES OF NATIONALISM
which sometimes takes pre-existing cultures and turns them uito nations, sometimes invents them, and often obliterates pre-existing cultures (hat is a reality, for better or worse, and in general an inescapable one. Those who are us historic agents know not what they do. but that is another matter. (. .. J
The great, but valid, paradox is this: nations can be defined only in terms of (he age of nationalism, rather than, as you might expect, the other way round. It is not the case that the 'age of nationalism' is a mere summation of the awakening and political self-assertion of this, thai, or the other nation Rather, when general social conditions make for standardized, homogeneous, centrally sustained high cultures, pervading entire populations and not just elite minorities, a situation arises in which well-defined educationally sanctioned and unified cultures constimte very nearly the only kind of unit with which men willingly and often ardently identify. The cultures now seem to be the natural repositories of political legitimacy. Only their does it come to appear that any defiance of their boundaries by political units constitutes a scandal.
Under these conditions, though under these conditions only, nations can indeed be denned in terms buih of will and of culture, and indeed in terms of the convergence of them both with political units, [n these conditions, men will to be politically united with all those, and only those, who share their culture, Polities then will to extend their boundanes to the limits of their cultures, and to protect and impose their culture with the boundaries of their power. The fusion of will, culrure and polity becomes the norm, and one not easily or frequently defied. (Once, it had been almost universally defied, with impunity, and had indeed pawed unnoticed and undiscussed > These conditions do not define the human situation as such, but merely its industrial variant.
It is nationalism which engenders nations, and not the other way round. Admittedly, nationalism uses the pre-existing, historically inherited prolifera-Hon of cultures Of cultural wealth, chough it uses them very selectively, and it most often transforms them radically Dead languages can be revived, traditions invented, quite fictitious pristine purities restated. But this culturally creative, fanciful, positively inventive aspect of nationalist ardour ought not to allow anyone to conclude, erroneously, that nationalism is a contingent, artificial, ideological invention, which might not have happened, if only those damned busy-body interfering European thinkers, not content to leave we!) alone, had not concocted it and fatefuUy injected it into the bloodstteam of otherwise viable political communities. The cultural shreds and patches used by nationalism are often arbittary historical inventions. Any old shred and patch would have served as well. Sunn no way does it follow that the principle of nationalism itself, as opposed to the avatars it happens to pick up for in incarnations, is itself in the least contingent and accidental.
Nothing could be further from the truth than such a supposition. Nationalism is not what it seems, and above all it is not what it seems to itself. The
ernest cellnep 65
cultures it claims to defend and revive are often its own inventions, or are modified out of all recognition Nonetheless the nationalist principle as such, as distinct from each of its specific forms, and from the individually distinctive nonsense which it may preach, has very very deep roots in our shared current condition, is not at all contingent, and will not easily be denied.
Dürkheim taught that in religious worship sociery adores its own camouflaged image. In a nationalist age, societies worship themselves brazenly and openly, spuming the camouflage. At Nuremberg, Nazi Germany did not wotship itself by pretending to worship God or even Wotan; it overtly worshipped itself. In milder but just as significant form, enlightened modernist theologians do not believe, or even take much inteiestin, the doctrines of their faith which had meant so much to their-predecessors. They treat them with a kind of comic auto functiortalism. as valid simply and only as the conceptual and ritual iooIs by means of which a social tradition affirms its values, continuity and solidarity, and they systematically obscure and play down the difference between such a tacitly reductionist 'faith', and the real thing which had preceded it and had played such a crucial pan in earlier European history. 3 part which could never have been plaved by the unrecognizably diluted, watered-down curtent versions.
But the fact that social self-worship, whether virulent and violent or gentle and evasive, is now an openly avowed collective self-wotship. tather than a means of covertly revering society through the image of God, as Dürkheim insisted, does not mean that the current style is any more veridical than that of a Dürkheim ian age The community may no longer be seen through the prism of the divine, but nationalism has its own amnesias and selections which, even when they may be severely secular, can be profoundly distorting and deceptive.
The basic deception and self-deception practised by nationalism is this: nationalism is. essentially, the general imposition of a high culture on society, where previously low cultures had taken up the lives of the majonry, and in some cases of the totality, of the population. It means that generalized diffu. sion of a school-mediated, academy-supervised idiom, codified for the requirements of reasonably precise bureaucratic and technological communication. It is the establishment of an anonymous, impersonal society, with mutually substitucable atomized individuals, held together above all by a shared culture oFthis kind, in place of a previous complex structure of local groups, sustained by folk cultures reproduced locally and idiosyncratic ally by the micro-groups themselves. That is what rtdtty happens.
But this is the very opposite of what nationalism affirms and what nationalists fervently believe. Nationalism usually conquers in the name of putative folk culture. Its symbolism is drawn from the healthy, pristine, vigorous life of the peasants, of the Volk, the narod. There is a certain element of truth in the nationalist self-presentation when the narod or Volk is ruled by officials of
66   THEORIES OF NATIONALISM
another, an alien high culture, whose oppression must be resisted first by a cultural revival and reaffirmation, and eventually by a war of national liberation. If the nationalism prospers it eliminates the alien high culture, but it does noi then replace it by the old local low culture; it revives, or invents, a local high (literate, specialist-transmitted) culture of its own, though admittedly one which will have some links with the earlier local folk styles and dialects. But it was the great ladies at the Budapest Opera who reaJly went to town in peasant dresses, or dresses claimed to be such. At the present time in the Soviet Union the consumers of "ethnic" gramophone records are not the remaining ethnic rural population, but the newly urbanized, appanment-dwelling, educated and multi-lingual population,2 who like to express their real or imagined sentiments and roots, and who will no doubt indulge in as much nationalist behaviour as the political situation may allow.
So a sociological self-deception, a vision ofrealiry through a prism of illusion, still persists, but it is not the same as that which was analysed by Durkheim. Society no longer worships itself through religious symbols; a modern, streamlined, on-wheels high culture celebrates itself in song and dance, which it borrows (stylizing it in the process) from a folk culture which it fondiy believes itseřf co be: perpetuating, defending, and reaffirming.
The Course of True Nationalism Never Did Run Smcath
A characteristic scenario of the evolution of a nationalism [. ,. ] tan something like this. The Rumanians were a peasant population speaking a group of related and more or less mutually intelligible dialects, and inhabiting a series of discontinuous but not very much separated pocltds within the lands of the Empire of Megalomania. The Ruritanian language, or rather the dialects which could be held to compose it. was not really spoken by anyone other than these peasants. The aristocracy and officialdom spoke the language of the Megalomanian court, which happened to belong to a language group different from the one of which the Ruritanian dialects were an offshoot.
Most, but not all. Ruritanian peasants belonged to a church whose liturgy was taken from another linguistic group again, and many of the priests, especially higher up in the hierarchy, spoke a language which was a modem vernacular version of the liturgical language of this creed, and which was also very far removed from Ruritanian The petty traders of the small towns serving the Ruritanian countryside were drawn from a different ethnic group and reíigicn stůl, and one heartily detested by the Ruritanian peasantry.
In the past the Ruritanian peasants had had many griefs, movingly and beautifully recorded in their lament-songs (painsnking))' collected by village schoolmasters late in the nineteenth century, and made well known to the international musical public by the compositions of the gieat Ruritanian na tional composer L). The pitiful oppression of the Ruritanian peasantry
ERNEST CELLNER 67
provoked, in the eighteenth century, the guerrilla resistance led by the famous Ruritanian social bandit K., whose deeds are said still to persist in the local/oik memory, not to mention several novels and two films, one of them produced by the national artist Z., under highest auspices, soon after the promulgation of the Popular Socialist Republic of Ruritania
Honesty compels one to admit that the social bandit was captured by his Dwn compatnots, and that the tribunal which condemned him 10 a painful death had as its president another compatriot. Furthermore, shortly after Ruritania first attained independence, a circular passed berween its Ministries of the Interior. Justice and Education, considering whether it might nor now be more politic to celebrate the village defence units which had opposed the social bandit and his gangs, rather than the said social bandit himself, in the interest of not encouraging opposition to the police.
A careful analysis of the folk songs so painstakingly collected in the nineteenth century, and now incorporated in the repertoire of the Ruritanian youth, camping and sports movement, doesnoi disclose much evidence of any serious discontent on the part of the peasantry with their linguistic and cultural situation, however gneved they were by other, more earthy matters. On the contrary, such awareness as there is of linguistic pluralism within the lyrics of the songs is ironic, jocular and good-humoured, and consists in part of bilingual puns, sometimes in questionable taste It must also be admitted that one of the most moving of these songs—I often sang it by the camp fire at rhe holiday camp to which [ was sent during the summer vacations—celebrates the fate of a shepherd boy, grazing three bullocks on the seigneurial clover (sir) near the woods, who was surprised by a group of social bandits, requiring him to surrender his overcoat. Combining reckless folly with lack of political awareness, the shepherd boy refused and was killed, I do not know whether this song has been suitably rewritten since Ruritania went socialist. Anyway, to return to my main theme: though the songs do often contain complaints about the condition of the peasantry, they do not raise the issue of cultural nationalism.
That was yet to come, and presumably post-dates the composition of the said songs. In the nineteenth century o population explosion occurred at the same time as certain other areas of the Empire of Megalomania—but not Ruritania—rapidly industrialized. The Ruritanian peasants were drawn to seek work in the industrially more developed areas, and some secured it, on the dreadful terms prevailing at the time. As backward rustics speaking an obscure and seldom written or taught language, they had a particularly rough deal m rhe towns ro whose slums they had moved. At the same time, some Ruritanian ladí destined for the church, and educated in both the court and the liturgical languages, became influenced by the new liberal ideas in rhe course of their secondary schooling, and shifted to a secular training at the university, ending nut aspnests but as journalists, teachers and professors. They received encouragement from a few foreign. non-Ruritaman ethnographers, musicologists and
68   THEOKI fci uc i.nnu,______
historians who had come to explore Ruritania. The continuing labour migration, increasingly widespread elementary education and conscription provided these Rumanian awakenen with a growing audience.
Of course, it was perfectly possible for the Ruritanians, if they wished to do so (and many did), to assimilare into the dominant language of Megalomania. No genetically transmitted trait, no deep religious custom, differentiated an educated Ruritanian from a similar Megalomanian. In fact, many did assimilate, often without bothering to change their names, and rhe telephone directory of the old capital of Megalomania (now the Federal Republic of Megalomania) is quite full of Ruritanian names, though often rather comically spelt in the Megalomanian manner, and adapted to Megalomanian phonetic expectations. The point is that after a rather harsh and painful start in the first generation, the life chances of the offspring of the Rumanian labour migrant were not unduly bad, and pTobably at least as good (given his willingness to work hard) as those of his non-Ruritaman Megalomanian fellow-citizens. So these offspring shared in the eventual growing prosperity and general embour-geoisement of rhe region. Hence, as far as individual life chances went, there was perhaps no need fot a virulent Rumanian nationalism.
Nonetheless something of the kind did occur, It would, 1 think, be quite wrong to attribute conscious calculation to the participants jn the movement. Subjectively, one must suppose that they had the motives and feelings which are so vigorously expressed in the literature of the national revival. They deplored the squalor and neglect of their home valleys, while yet also seeing the rustic virtues still to be found in therm they deplored the discrimination to which their co-nacionalj were subject, and the alienation from their native culture to which they were doomed in the proletarian suburbs of the industrial lowns. They preached against these ills, and had the hearing of at least many of their fellows. The manner in which, when the international political situation came to favour ir, Ruritariia eventually attained independence, is now part of the hisroncal record and need not be repeated here
There is, one must repeat, no need to assume any conscious long-term calculation of interest on anyone's part. The nationalist intellectuals were full of warm and generous ardour on behalf of the co-nationals. When they donned folk costume and trekked over the hills, composing poems in the forest clearings, they did not also dream of one day becoming powerful bureaucrats, ambassadors and ministers. Likewise, the peasants and workers whom rhey succeeded in reaching felt resentment at their condition, but had no reveries about plans of industrial development which one day would bring a steel mill (quite useless, as it then turned out) to the very heart of the Ruritanian valleys, thus totally ruining quite a sueable area of surrounding arable land and pasture. It would be genuinely wrong to try to reduce these sentiments to calculations of material advantage or of social mobility. The present theory » sometimes travestied as a reduction of national sentiment to calculation of
prospects of social promotion Bur tins is a misrepresemauvu. UJ M«. it made no sense to ask whether the peasants loved their own culture: they took it for granted, like the air they breathed, and were not conscious of either But when labour migration and bureaucratic employment became prominent features within their social horizon, rhey soon learned the difference between dealing-with a co-national, one understanding and sympathizing with their culture, and someone hostile to it. This very concrete experience taught rhem to be aware of their culture, and to love it (or, indeed, to wish to be rid of it) without any conscious calculation of advantages and prospects of social mobility. In stable self-contained communities culture is often quite invisible, but when mobility and context-free communication come to be of the essence of social life, the culture in which one has been laught to communicate becomes the core of one's identity.
So had there been such ealculariorc (which there was not) it would, in quite a number of cases (though by no means in all), have been a very sound one In fact, given the at least relative paucity of Ruritanian intellectuals, those Rurita-nians who did hav« higher qualifications secured much better posts in independent Rumania than most or them could even have hoped for in Greater Megalomania, where they had to compete with scholastically more developed ethnic groups. As for the peasants and workers, they did not benefit immediately; but the drawing of a political boundary around the newly denned ethnic Ruritania did mean the eventual fostering and protection of industries in the area, and in the end drastically diminished the need for labour migration from it.
What all this amounts to is this: during the early period of industrialization, entrants into the new order who are drawn from cultural and linguistic groups thai are distant from those of the more advanced centre, suffer considerable disadvantages which are even greater than those of other economically weak new proletarians who have the advantage of sharing the culture of the political and economic rulers. But the cultural /linguistic distance and capacity to differentiate themselves from others, which is such a handicap for individuals, can be and often is eventually a positive advantage for entire collectivities, or potential collectivities, of these victims of the newly emergent world. It enables them to conceive and express their resentments and discontents in intelligible terms. Ruritanians had previously thought jnd felt m terms of family unit and village, at mosi in terms of a valley, and perhaps on occasion in terms of religion But now, swept into the melting pot of an early industrial development, they had no valley and no village: and sometimes no family. But ihert mere other impovenshed and exploited individuals, and a jot of them spoke dialects recognizably similar, while most of the betrer-off spoke something quite alien: and so the new concept of the Ruritanian nation was bom of this contrast, with some encouragement from those journalists and teachers. And it was not an illusion1 the attainment of some of the objects of the nascent
70   THEORIES OF NATIONALISM
Rumanian national movement did indeed bring relief of the ills which had helped to engender it. The relief would perhaps have come any way; but in this national form, it also brought forth a new high culture and its guardian state.
[Nations anä NMiandum {Sladcwdl: Oxford, 1983), AB-9. 55-62.]
n
!■ j]
t!
:1
ii
>r
rc :i h bi