Both "integration" and "marginalization" are words that are currently widely used in public discussion of contemporary social structures. They are concepts central to the social science enterprise as well insofar as both refer implicitly to the concept of "society." The problem with the discussion within social science is that, although the concept of society is basic to our analyses, it is at the same time an extraordinarily vague term, and this confuses the discussion about integration and marginalization.

Resumé

The modern world-system has created considerable confusion about what we can mean by integration and marginalization into our societies/states. One of the principles of most sovereign states in the last two centuries is that they are composed of "citizens." Once there were citizens, there were non-citizens as well. Citizenship became something very valuable, and consequently not something one was very willing to share with others. Despite the fact that citizenship is a cherished good, which gives rise to "protectionist" sentiment, migration is a constantly recurring phenomenon in the modern world, which leads to the issue of national integration.

The world revolution of 1968 put into question, for the first time since the French Revolution, the concept of citizenship. What was different about the world revolution of 1968 was that it was an expression of disillusionment in the possibilities of state-level reformism. The post-1968 movements added something new. They insisted that racism and sexism were not merely matters of individual prejudice and discrimination but that they took on "institutional" forms as well. What these movements seemed to be talking about was not overt juridical discrimination but the covert forms that were hidden within the concept of "citizen." The concept of citizenship is, in its essence, always simultaneously inclusionary and exclusionary. We should begin to conceive whether we can go beyond or dispense with the concept of citizen, and if so, to replace it with what?

---

* Immanuel Wallerstein, Fernand Braudel Center, Binghamton University, Binghamton, New York 13902-6000, USA.

The concept of society is, I suppose, millennial, in the sense that it has probably been true for at least ten thousand years, if not longer, that humans have been aware of two things about the world in which they live. They interact on a regular basis with others, usually persons located in propinquity. And this "group" has rules of which they all take account, and which in fact fashion in many ways their consciousness of the world. The membership of such groups, however, is always less than the totality of human beings on the earth, and hence the members always distinguish between "we" and "the others."

The classic myths that humans have tended to create about their own "societies" is that the gods somehow created their particular one, usually created it especially, in some remote era, and that the current members are descended from this favored original group. Aside from the self-serving character of such myths, they also imply consanguineal continuity.

Of course, we know that consanguineal continuity is quite literally a myth, in the sense that no group has ever operated this way perfectly. And we know that this is particularly true of the modern world. Hence, since persons from outside the groups are constantly seeking to enter them or are being pulled into them in one way or another, we talk of integration. And since other people are constantly seeking to withdraw from the groups or are being pushed out of them, we speak of marginalization.

The basic intellectual problem is that the modern world-system has created considerable confusion about what we can identify as our "society," and therefore what we can mean by integration and marginalization into such societies. It is quite clear that, in practice, we have been using the word "society" for at least two centuries now to mean the group that is located within the boundaries of a sovereign state, or sometimes what we think ought to be the boundaries of some sovereign state, existing or to be created. Now whatever is the ancestry of such state-bounded groups, they bear little resemblance to continuing consanguineal groups.

Indeed, one of the principles of most sovereign states in the last two centuries is that they are composed of "citizens," of demos and not of ethnos, and therefore represent a category that is more juridical than cultural in character. Furthermore, the category, citizens, is not at all self-evident in its geographical contours; that is, it is not perfectly congruent with persons resident at any particular point of time in a given sovereign state. Some inside the state are not citizens, and some outside the state are. In addition, while states have quite varying rules about the acquisition (and loss) of citizenship, they all have some rules, as well as rules governing the entry of non-citizens into their territory (immigration) and the legal rights of resident non-citizens. Furthermore, migration (inward and outward) is not an exceptional phenomenon in the modern world-system, but rather a continuing (and relatively massive) phenomenon.

Let us begin at the beginning. The modern world-system was constructed during the long sixteenth century, and its original geographical bounds included a large part of the European continent and parts of the Americas. Within this geographic zone, an axial division of labor grew up that took the form of a capitalist world-economy. An institutional framework to sustain this kind of historical system grew up alongside. One such institutional element, a quite essential one, was the creation of so-called sovereign states that were located within an interstate system. Of course, this was a process and not an event. Historians describe this process when they discuss statebuilding within Europe beginning with the New Monarchies of the late fifteenth century, the rise of diplomacy and its rules beginning with the Italian city-states in the Renaissance, the establishment of colonial regimes in the Americas and elsewhere, the collapse of the Habsburg world-empire in 1557, and the Thirty Years' War culminating in the Treaty of Westphalia with its new foundations for state integration and interstate order.

This process of state-construction was not, however, a process separate from the development of historical capitalism, but rather an integral part of the story. Capitalists were well served by the establishment of such sovereign states, obtaining from them a multiplicity of services: to guarantee their property rights, to provide them with protection rent (see Lane 1979), to create the quasi-monopolies they needed to make significant profits, to advance their interests over those of rival entrepreneurs located in other countries, and to provide sufficient order to guarantee their security.¹ Of course, these states were not equal in strength, and it was precisely this inequality which enabled the stronger states to serve well their entrepreneurs. But there was no land area within the division of labor that was not under the jurisdiction of some state, and therefore there were no individuals who were not subject to some primary state authority.

The period going from the sixteenth to the eighteenth centuries marked the institutionalization of this system. During this period, the original claim to the exercise of sovereignty was put forward in the name of a so-called absolute monarch, although subsequently in some states the ruler was under pressure to share the exercise of these sovereign powers with a legislature or a magistrature. We are still however before the era of passports and visas, or of migration controls, or of significant voting privileges for more than a very small minority of the populations. The mass of the population were "subjects," and a distinction between subjects who had some kind of descent rights and those who did not was seldom invoked and not very meaningful. In the seventeenth century, the juridical and social difference in day-to-day life between say a Breton migrant to Paris and a Rhineland migrant to Leyden (one crossing a not very visible international frontier and the other not) was hard to discern.

The French Revolution transformed this situation, by transforming subjects into citizens. There would be no turning back, either for France or for the capitalist world-system as a whole. The states had become theoretically, and to some degree in practice, responsible to a large group of persons with constituted political claims. During the nineteenth and twentieth centuries, implementation of these political claims may have been slow and quite uneven in reality, but there was a clear triumph of the rhetoric. And rhetoric matters. But once there were citizens, there were noncitizens as well.

The transformation of subjects into citizens was the consequence of pressures both from above and below. Popular demands for participation in governance, what might be called the demand for democratization, expressed itself constantly and in whatever ways it could. It served as an underlying force that found expression in populism and in revolutionary upsurges. The claimants were regularly suppressed but the concept survived in a larval form, always there as potential dynamite even if often weak as an immediate pressure.

The long-run response to these demands of the so-called dangerous classes was the political program of liberalism, the triumphant ideology of the capitalist world-system in the nineteenth century. The liberals proposed a program of rational reform, of measured concessions, of gradual institutional change. The nineteenth-century program of liberalism had three main components: suffrage, redistribution, and nationalism.² Suffrage involved giving the vote to larger and larger segments of those resident in the state. By the twentieth century, universal suffrage of adult males and females (with exceptions for specified categories like felons and the insane) came to be the norm. Redistribution involved state-decreed and state-enforced minimal levels of wages and state-administered social security and welfare benefits, the so-called welfare-state, a program that also became the norm, at least in the wealthier countries, by the mid-twentieth century. The third element in the program, nationalism, involved the creation of a sense of patriotic attachment to one's own state, systematically transmitted primarily by two institutions: primary schools (once again virtually universal by the midtwentieth century) and the armed services (participation in which came to be the norm in most countries, even in peacetime, at least for men). Collective nationalist rituals also became quite frequent everywhere.

If we look at each of these three major political institutions - the suffrage, the welfare state, and nationalist rituals/sentiments - we see immediately the relevance of the distinction citizen/non-citizen, at least as it operated up to twenty years ago or so. Only citizens had votes. It was unthinkable that noncitizens would be allowed to vote, however long they might have been resident in a country. State-administered welfare benefits usually, although not in every case, made distinctions between citizens and non-citizens. And

of course, nationalist rituals/sentiments were the domain of citizens, from
which non-citizens were socially excluded, as a consequence of which the
latter became morally suspect, especially in times of interstate tension.

It is not only that these three institutions were developed as institutions of the separate states, albeit in parallel manner, but that citizens were thereby privileged to be central to the process of constructing and strengthening their own states. Since the states were involved in an interstate competition for the "wealth of nations," and since the privileges of the citizens seemed to depend on the achievement of the states, citizenship was considered to be an exceptional privilege, certainly at least in all those states which were in the upper quartile of the hierarchy of GNP. Furthermore, these states all presented themselves to their citizens as somehow quite special, and this seemed plausible to those who benefited from citizenship.

Citizenship thus became something very valuable, and consequently not something one was very willing to share with others. Citizenship in one's state might be doled out to a few eager applicants, but in general it was an advantage to be hoarded. This was all the more true insofar as the citizens believed that they had struggled internally (and externally) to acquire this privilege, and that it had not been a mere gift to them. They felt they merited the citizenship morally. Thus the fact that citizenship as a concept constituted a demand from the bottom up made it all the more efficacious as a mechanism by which the dangerous classes were tamed from the top down. All the state rituals combined to reinforce the belief that the "nation" was the only society to which one belonged or, if not the only one, the most important one by far.

Citizenship effaced, or at least obscured, all other sorts of conflicts - class conflicts; conflicts between groups or strata defined in terms of race, ethnicity, gender, religion, language, or any other social criterion other than "nation/society." Citizenship brought national conflict to the forefront. Citizenship was intended to be unifying within the state, and it did in practice serve this purpose well, all the more so since citizenship conferred privilege, or at least seemed to do so. The concept of citizen has been in general a quite stabilizing element in the modern world-system. It did reduce intra-state disorder, while at the same time it cannot be argued that it increased significantly inter-state disorder above the level that would probably have existed in its absence. It has not only been a stabilizing concept; it has been a central one. One has but to look at the juridical scaffolding of modern states to realize how much of the legislation and administration of states depends on the category of citizen.

Nonetheless, the concept of citizen has created difficulties, for one of the socio-economic underpinnings of the capitalist world-economy is the imperative of continuing physical flows of the labor force, or migration. Migration is first of all an economic necessity. The constant shifts in the

location of economic activities, combined with the uneven distribution of demographic norms, means that, inevitably, there are disparities in local supplies and demands for specific kinds of workers. Whenever this happens, the interests of some workers and some employers are clearly served by some kind of labor migration, and it therefore tends to occur, with greater or lesser promptness depending on the legal constraints (as well as the practical possibilities of evading these constraints). The disparity of local supplies and demands of the labor force cannot be calculated simply in absolute totals of the labor force. Different groups of workers tend to price themselves at different levels for similar kinds of work. This is what we mean by "historical wages." Hence, it is perfectly possible that, in a given local area, there are persons seeking wage work who will refuse to accept certain types of low-paid wage work, and employers will turn to potential or actual immigrants to fill the needs.

So, despite the fact that citizenship is a cherished good, which gives rise to "protectionist" sentiment, migration is a constantly recurring phenomenon in the modern world. This has been true since the beginnings of the modern world-system. I am not sure that migration, however defined, is really quantitatively greater today than in previous centuries as a percentage of the total population, despite the improvement in transport facilities, but it is certainly a more politically noticed and politically controversial phenomenon.

It is the concept of citizen that has changed the meaning of the term migrant. A person who leaves a rural area or a small town and moves to a large city 50 kilometers away may be going through a social transformation as great as one who moves to a large city 5000 kilometers away. Or, if this is no longer true in many countries in the late twentieth century, it was probably more or less true everywhere until at least 1950. The difference is that the 5000-kilometer migrant is quite likely to traverse a state frontier, whereas the 50-kilometer migrant is unlikely to do so. Hence, the latter is legally defined as a migrant, ergo not a citizen, whereas the former is not.

A significant proportion of migrants tend to stay in the locale (or at least the state) into which they have migrated. They tend to have children who are born in the new locale and who, quite often, are culturally the products of their birthplace and not that of their parents' birthplace. When we discuss the issue of integration, it is the integration of such long-term migrants, and their children, of which we are usually speaking. Receiving countries have different rules about the citizenship of persons born in the country, from the jus soli of the United States and Canada to the jus sanguinis of Japan and in a modified form Germany, with a continuum of possibilities in-between.

Integration is a cultural concept, not a legal one. The concept of integration
assumes that there is some cultural norm into the acceptance of
which one has to be integrated. For some states, which are largely monolingual

and mono-religious, such a norm may seem relatively obvious and not too intrusive, although even in such states one can always find "minorities" who deviate from these normative patterns. For other states, which have more "variegated" populations, dominant norms exist nonetheless, but they seem more overbearing and pernicious. Take the United States. At the time of the founding of the republic, the cultural norm of citizenship was to be an English-speaking Protestant of one of four varieties (Episcopalians, Presbyterians, Methodists, and Congregationalists). Of course, this definition corresponded to the upper stratum but included parts of the middle and lower strata as well. This definition was slowly extended to include other varieties of Protestants. Roman Catholics and Jews were fully included in the cultural definition only as recently as the 19505, at which point politicians began to speak of the "Judeo-Christian heritage." African-Americans have never really been included, whereas Latinos and Asian-Americans seem to be in a holding pattern, awaiting their future admission. Moslems, now for the first time a significant minority, are still excluded.

The U.S. example shows the flexibility that is possible in defining the cultural normative pattern of any particular state. The quasi-official ideological interpretation of this flexibility within the United States is that it shows the capacity of the U.S. political system to incorporate outsiders into the category of citizen, and thereby to "integrate" them into the nation. No doubt it does show this. But it also shows that at no point have all migrants been integrated. One might wonder whether there is not something inherent in the process such that at no point would it ever be true that all outsiders will be incorporated. Emile Durkheim once suggested that, whenever deviance disappears de facto, the social system redefines its norms so as to recreate statistical deviance. Perhaps the same things is true in relation to the concept of citizen. When all residents are de facto integrated, does the "nation" redefine itself so as to recreate "marginals"?

Such an outrageous idea assumes that there is social utility to the creation of marginals, and social scientists have in fact often suggested this in one way or another: the value of a scapegoat on whom to thrust our collective sins; the existence of an understratum to create permanent fear among the dangerous classes that they might be made even worse off than they are, and that they therefore should restrain their level of demands; the strengthening of in-group loyalty by providing visible, and undesirable, contrasting strata. These are all plausible suggestions; they are however also quite general and generic.

I noted earlier that this pattern remained more or less the same from about 1800 to the 19705, intimating that matters have changed somewhat since then. I believe this is true. The world revolution of 1968 marked a turningpoint in the history of our modern world-system in many ways. What has not

been noticed is that one of its consequences was to put into question, for the first time since the French Revolution, the concept of citizenship. It was not merely the fact that 1968 was "internationalist" in spirit. After all, we had internationalist movements already throughout the nineteenth and twentieth centuries: on the one hand, the various workers' internationals, and on the other hand, all kinds of peace movements. As we know, such internationalist movements were not very efficacious in constraining the outburst of nationalist sentiment among its members or audience when tension in the interstate system rose sharply. The most notable instance, regularly noted, was the response of the socialist parties to the outbreak of the First World War. (Haupt 1965). The reason is well explained by Kriegel and Becker, in their book on the French socialist debates in the weeks preceding the outbreak of war in 1914:

It appears thus that a certain socialism is nothing but a modern form of Jacobinism and, faced with one's country in danger, the voice of the "great ancestors" outweighed that of socialist theories whose relevance to the immediate situation was difficult to perceive. In the immense patriotic whirlwind in which the country was enveloped, war was once again seen as capable of achieving old aspirations: instead of human fraternity through peace, it was human fraternity through war, through victory. (Kriegel & Becker 1964, 123).

The internationalist orientation of workers' and peace movements were deeply constrained by the fact that each had created their organizations at the national level. But even more important, they had created their organizations at the national level because they considered that their objectives could best, perhaps only, be realized at the national level. That is, they acted primarily as citizens, joined together in a political effort to influence, even transform, their states. It was their presumption that, by changing their states, they would contribute to creating the international solidarity of which they were partisan. Nonetheless, the political activity was first of all, and most often exclusively, national.

What was different about the world revolution of 1968 was that it was just the opposite, an expression of disillusionment in the possibilities of statelevel reformism. Indeed, the participants went further. They argued in effect that the orientation to national reformism was itself a prime means of sustaining a world-system they wanted to reject. The revolutionaries were not against popular action but against citizen action, even when it claimed to be "revolutionary." It was this stance that perhaps aroused the greatest dismay among those distressed by the 1968 uprisings, especially among the Old Left.

This attitude of the 1968 revolutionaries arose out of two analyses they made of the history of the modern world-system. The first was that the historic two-step strategy of the world's antisystemic movements - first achieve state power, then transform the world - was in their view a historic failure. The revolutionaries of 1968 said in effect that the antisystemic

movements born in the nineteenth and twentieth centuries - the Social the Communists, and the national liberation movements - all had in fact already come to state power, more or less, in the period following the Second World War. But, having done so, they hadn't changed the world.

This first observation was rendered still more critical by the second element of the analysis. Insofar as the antisystemic movements had achieved power, it was indeed true that they had effectuated certain reforms that seemed to be progressive, if not revolutionary. But, but ... these reforms were said to have systematically favored a particular and small segment of the lower strata - those of the dominant ethnic group in each country, males primarily, those that were more educated (shall we say more "integrated"?) into the national culture. There were many others left out, forgotten, "marginalized," who hadn't benefited really even from the limited reforms that had been instituted: the women, the "minorities," and all sorts of nonmainstream groups.

What happened after 1968 is that the "forgotten peoples" began to organize both as social movements and as intellectual movements, and set forth their claims not merely against the dominant strata but against the concept of the citizen. One of the most important themes of the post-1968 movements was that they were not merely opposed to racism and sexism. After all, there had long been movements who fought against racism and sexism. But the post-1968 movements added something new. They insisted that racism and sexism were not merely matters of individual prejudice and discrimination but that they took on "institutional" forms as well. What these movements seemed to be talking about was not overt juridical discrimination but the covert forms that were hidden within the concept of "citizen," insofar as citizen was meant to indicate the combination of competence and inherited rights.

Of course, any struggle against covert denial of rights is plagued by the problem of plausibility, evidence, and ultimately proof. What the movements pointed to was outcome. They argued that, in point of fact, there continued to exist gross differentials in the hierarchical position of multiple groups, and this outcome could only be, it was argued, the outcome of institutional marginalization. As an argument in social science, the assertion that institutional marginalization was systematic and fundamental to the contemporary world-system has basically only two possible responses.

One is the conservative response: to deny the premises. The differential in outcomes in group hierarchization may be patently observable, but it does not follow that the cause is institutional marginalization. It can be argued that other factors explain the differential outcome, factors having to do with cultural differentials among the groups. This line of reasoning faces a simple logical problem. Even if we discover so-called cultural differentials among the groups being measured, how in turn do we explain these differentials -

by other cultural differentials? Ultimately, we must return either to a social structural explanation, which was the case of those who put forward the hypothesis of institutional racism/sexism, or alternatively to a sociobiological one, which quickly glides into classical racism-sexism.

If we wish to reject the conservative stance, and accept the social structural explanation, then the issue shifts from accounting for the differentials to reducing them, presuming that this is seen as a moral good. And indeed, this has been one of the central, if not the central, political debate of the last twenty years. Let us review the various positions put forth in this debate. The simplest position - simplest because it accords best with the traditional arguments of liberal ideology - has been that institutional racism and sexism can be overcome by making overt what was covert. And, many added, since it takes time for the process to work, it can be speeded up by temporary systematic assistance to those against whom institutional marginalization had historically operated. This was the essential case for the original program of this kind, the United States program called "affirmative action."

In effect, affirmative action programs are programs to "integrate" those who in theory should long since have been integrated. They are programs to carry out the original intent of the concept of citizenship which, it was being argued, had somehow been subverted by forces antithetical to the full realization of democracy, or citizenship. Affirmative action programs tended to assume the good faith of the "system," but the bad faith of individual participants. They therefore seldom, if ever, posed the prior question of whether there was anything systemic in the fact that theoretical citizenship had never been fully realized, even for the categories of persons to whom it supposedly applied.

Affirmative action programs had three drawbacks. Even great efforts (political and financial) accomplished limited results. In the first place, there was considerable covert resistance to them, and this resistance found many outlets. For example, trans-group school integration was extremely difficult, as long as de facto housing segregation existed. But to challenge de facto housing segregation meant both to intrude in an area generally considered to be part of individual choice and to tackle the issue of class-based de facto housing segregation (since class and race/ethnicity categories were highly correlated).

In the second place, affirmative action only took into account in some sense those who theoretically had citizen rights. But the definition of these categories was in itself part of the issue. Should the children of migrants (Turks in Germany, Koreans in Japan, etc.) be excluded from the rights enjoyed by the children of non-migrants? Should migrants themselves be excluded? This led to many demands for the extension of citizenship rights to juridical non-citizens - both by the easing of the mechanisms of acquiring citizenship, and even by the formal extension of some rights historically

accorded only to citizens to non-citizens (for example, the right to vote, at
least in so-called local elections).

In the third place, the logic of affirmative action led to the expansion of the kinds of groups making claims, as well as to the subdivision of groups making claims. And inevitably, this led to a de facto quota system that seemed to have no end. Nor was it clear when this temporary adjustment could or would make place for a so-called reformed or fully-implemented citizenship to operate without reference to sub-groups of citizens. This led inevitably to the charge of "reverse racism" - that is, to the charge that the previously marginalized groups were now in fact being juridically favored, and particularly at the expense of other low-ranking groups who had been historically more integrated (say, members of the working classes who were male and of the dominant ethnic group). Affirmative action thereupon became not merely difficult to administer and of uncertain benefits but politically very difficult to sustain. This was true not merely within the states as political structures but within the universities as structures of knowledge as well.

There was of course another path to pursue if one wished to overcome the limitations of traditional concepts of citizenship, limitations in terms of the unequal outcomes. Instead of pursuing further "integration" into the structures of marginalized groups, one could pursue the path of the equality of groups. Whereas affirmative action found legitimation in the liberal concept of the perfect equality of all citizens, the concept of group equality found legitimacy in the liberal concept of the self-determination of nations. To be sure, the latter concept had been intended to apply only to the relations of states to each other, and hence the rights of "colonies" to become sovereign states, but it was only a slight stretch of the concept to apply it to groups within states.

This was the path of group "identity" which, as we know, has found strong support within women's groups, within groups based on race or ethnicity, within groups based on sexuality, and indeed within an expanding number of other groups. The path of group identity has involved the rejection of the concept of integration entirely. Why, said its proponents, should marginalized groups want to integrate into dominant groups? The very concept of integration involves, they argued, the assumption of biological or at least bio-cultural hierarchy. It assumes that the group into which one is being called to integrate is, in some way, superior to the group that has been marginalized. On the contrary, said the proponents of group identity, our historical identity is at least as valid as, if not outrightly superior to, the identity into which we are being called to integrate.

The path of groups proclaiming the validity of their identity, and hence the
need to reinforce group consciousness of their identity, is the path
generically of "cultural nationalism." This is essentially a segregationist

path, but (it turns out) not one necessarily antithetical to state integration.
One can argue for it in the name of a state integration based not on individual
citizens but on collective citizens, so to speak.

The difficulties with this path lie in the definition of the groups which could be the collective citizens. This is not necessarily insoluble. Switzerland historically has acknowledged, in certain ways, collective linguistic citizens. Some persons in Quebec have argued for the recognition of two historic "nations" within the Canadian state. Belgium has gone down this path. Without arguing the specific political situations of each of these cases, it is apparent that, whenever one puts forth the idea of collective citizens, one political dilemma is that there are always unresolved, and perhaps unresolvable, nodal points of non-inclusion (so-called allophones in Canada) or overlap (Brussels in Belgium).

But this is the not the greatest difficulty of cultural nationalism. After all, one can in many cases arrive at political compromises. The greatest problem, as in the case of affirmative action, is the definition of the groups themselves, and for themselves. For, as we know, however we define cultural groups, they contain sub-groups or cross-cutting groups. The discussion within women's movements about the neglect of the interests of women of color (at a national level) or of Third World women (at a world level) by White women has led to divisions parallel to those provoked by the discussion within states of the neglect of the interests of women by men.

Once again, there are ways to handle this politically. They all take the form, more or less, of proposing a "rainbow" coalition, that is, a coalition of all marginalized groups within the state to pursue transformations of common interests to them. But rainbow coalitions too run into two problems: debates about comparative victimship, and decisions about which groups are to be considered marginalized for the purposes of inclusion in the coalition. And they run into the same reaction as affirmative action: the charge of exclusion. If there can be separate schools for Blacks or for women, in order to foster consciousness, may there also be separate schools for Whites or for men? Essentialism is a double-edged sword.

It is no wonder, given the fact that each proposed solution has run into difficulties, that marginalized groups have been deeply divided about their strategy and have been oscillating in their tactics. One might ask the question whether the difficulties do not lie in the fact that, at the bottom, the entire debate about integration and marginalization, even for the post-1968 groups despite their skeptical rhetoric, has been based on the assumptions of the concept of citizenship, and that the concept of citizenship is, in its essence, always simultaneously inclusionary and exclusionary.

The concept of citizen makes no sense unless some are excluded from it.
And the some that are to be excluded must be, in the last analysis, an
arbitrarily-selected group. There is no perfect rationale for the boundaries of

the categories of exclusion. Furthermore, the concept of citizen is bound up with the fundamental structure of the capitalist world-economy. It derives from the construction of a states-system that is hierarchical and polarizing, which means that citizenship (at least in the wealthier and more powerful states) is inevitably defined as a privilege that it is not in its members' interest to share. It is bound up with the need to hold in check the dangerous classes, and they can best be held in check both by including some and excluding others.

In short, I am arguing that the entire discussion about integration and marginalization has led us into a cul-de-sac, out of which there is no exit. Better not to enter it and instead to begin to conceive how we can go beyond the concept of citizen. Of course, this means going beyond the structures of our modern world-system. But, since I believe that our modern world-system is in a terminal crisis (a case I do not have the time to develop now, but see Hopkins & Wallerstein 1996), we should perhaps at least consider the kind of historical system we wish to construct, and whether it would be possible to dispense with the concept of citizen; and if so, to replace it with what?

---

NOTES

1. I have spelled out the historical relationship of the states to the entrepreneurs in "States? Sovereignty? The Dilemmas of Capitalists in an Age of Transition," forthcoming in D. Solinger, D. A. Smith, & S. Topik, eds., The States Still Matter: States and Sovereignty in the Contemporary Global Economy.

2. The historical evolution of this program and its social underpinnings are analyzed in detail in my After Liberalism, esp. Part n, "The Construction and Triumph of Liberal Ideology," 71-122.

REFERENCES

Haupt, G. 1965. Le Congrés manqué. L'lnternationale å la veille de la premiere guerre
mondiale. Paris: Francois Maspéro.

Hopkins, T. K. & Wallerstein, I. (coord.) 1996. The Age of Transition: Trajectory of the World
1945-2025. London: Zed Press.

Kriegel, A. & Becker, J.-J. 1964.1914. La guerre et le mouvement ouvrierfranfais. Paris: Lib.
Armand Colin.

Lane, F. 1979. Profits and Power. Albany, NY: SUNY Press.