Nationalism and Liberalism

Mark Haugaard. The Sage Handbook of Nations and Nationalism. Editor: Gerard Delanty & Krishan Kumar. 2006. Sage Publishing.


Arguably nationalism and liberalism are two of the most successful ideologies of the nineteenth and twentieth centuries (the others include socialism and conservatism) and both are associated with the rise of modernity in the eighteenth century. While they share their origins in modernity, only liberalism’s success was widely predicted and perceived to be the logical outcome of modernization. In contrast, nationalism was seen as a temporary aberration. Modernity presupposes atomized individuals, fitted for industrial production, who reject any forms of essentialism, which is seen as a hangover from feudalism. Modernity is also premised upon the march of reason, which should dispel sources of the self based upon metaphysical notions of community, especially those that lay claim to primordial roots.

By and large, liberals do not find the recent emergence of liberalism in the least perturbing but interpret this fact as a confirmation of the sophistication of their ideology—a manifestation of the growth of reason. In contrast, nationalists frequently dispute the view that nationalism is a phenomenon of historically recent date, tending to see it as a ‘natural condition’ of humankind. Consequently, nationalists rarely develop complex philosophical arguments to justify their position. In Imagined Communities, Anderson argues that nationalism is not an ideology but more a form of sentiment and, in support of this he observes that liberals frequently refer to political theorists, whereas nationalists rarely do so (Anderson 1983: 5).

While liberals expend greater intellectual energy theorizing, nationalists are more willing to sacrifice their lives for the sake of their ideological cause. Or, at the very least, nationalists are more likely than liberals to make reference to, and celebrate, those who have died in their cause. The celebrated tomb of the ‘Unknown Soldier’ commemorates a soldier who died for the cause of nationalism (Anderson 1983: 9-12). There is no liberal conceptual equivalent to this, nor do liberals appear to feel the need for it—there are no annual pilgrimages to the tombs of dead liberals.

These contrasts tell us that, while both are consequences of modernity, nationalism and liberalism are sociologically different phenomena. While I would not go as far as Anderson, arguing that nationalism is not an ideology, it is a mistake to compare them as though they were normative political theories where disagreement is over some philosophical point, as is the case, for instance, in contemporary debates between liberals and communitarians. Hence, I would argue that attempts by thinkers such as Tamir (1993), Miller (1995) and Moore (2001) to bring nationalism and liberalism together purely within the traditions of contemporary political philosophy miss an important aspect of what distinguishes nationalism from liberalism. In this article I shall use a sociological understanding of how these very different ideologies were the outcome of modernity and, from these premises, analyse the normative compatibility of the two ideologies. However, it is essential to begin with a few words concerning nationalism and liberalism as concepts.

The Concepts

Nationalism and liberalism are concepts that do not have stable uncontested definitions. Part of the reason is that both are what Wittgenstein (1967) has termed ‘family resemblance’ concepts. These are concepts which do not have a single core essence that defines them but are rather like the members of a family, who resemble one another by a criss-crossing of characteristics—eyes, complexion, stature and so on.

The differences in form taken by liberalism are largely a reflection of theoretical traditions of thought and premises. The two dominant traditions are utilitarianism and contractarianism, while the most frequently used premises are either toleration or autonomy. Within the liberal perspective the basic unit of analysis is the individual. The latter maximizes autonomy and considers the pursuit of toleration essential for legitimacy. In contrast, any definition of nationalism is largely a reflection of historical circumstance. It used to be taken for granted that all nationalists desired full sovereignty in the form of a state of their own (for instance, Gellner 1983: 1), but, as many commentators have pointed out (for instance, Keating 1996), only some nationalists wish for full sovereignty, while others are content with local autonomy. Nationalism varies according to what is considered central to constituting the nation, which can be language, religion, perceived racial difference or identity. The central unit of analysis of nationalism is not ‘the individual’ but the ‘nation,’ which is a communally constituted entity. Of course, it is individuals who are nationalists but these social agents are perceived as ‘incomplete’ without national membership. The autonomous individual of liberalism is preoccupied with community as a constraint upon freedom, while the nationalist considers community a condition of self-realization. Nineteenth-century nationalist rhetoric used the concept of autonomy, which sounds superficially like liberalism, but for nationalists the autonomy of the individual derives from the autonomy of the nations and is consequently subservient to it.

Like all political ideologies, both liberalism and nationalism are essentially theories stipulating the conditions under which state power structures are legitimate. Because both are family resemblance concepts, any definition will not cover all instances. Most liberals hold that the state is legitimate insofar as it is based upon one of the following: toleration, neutrality, rights, justice as fairness, freedom or autonomy. In contrast, nationalists believe that the state is legitimate insofar as there is congruence between sovereign state and nation, or sufficient state power delegated to the nation for it to flourish. Within this the nation can be defined linguistically, religiously, culturally or in terms of identity.

Nationalism and liberalism should be considered as scalar, rather than absolute concepts. Absolute concepts are like apples and oranges: something is either an apple or an orange, not more or less so. In contrast, as scalar concepts nationalism and liberalism are both held with differing levels of intensity. At one end of the scale there are nationalists and liberals who, for instance, are willing to give their lives either for the nation or freedom of speech while at the other end are those who, in everyday life, express some mild pleasure in seeing ‘their’ national team win a sporting event or express mild dismay at the passing of some law which violates ‘their liberties.’ While individuals occur at the extremes of both liberal and nationalist scales, it is arguably the case that the intense end of the scale is more significant to nationalists than to liberals.

The scalar nature of nationalism and liberalism points to another aspect of these ideologies. As characterized by Giddens (1984), an actor’s knowledge of social life can be divided into two parts: discursive consciousness knowledge, which is knowledge that actors can readily put into words, and practical consciousness knowledge—what Bourdieu called habitus. Practical consciousness knowledge is much greater in its extent than discursive consciousness knowledge and is essential to the everyday social competence of social actors. It is an interpretative horizon, constituted of conceptual categories, upon which we routinely draw in making sense of the world—we see tables and chairs as ‘natural occurrences.’ While there is a continual flow between practical and discursive consciousness, there are some concepts that tend, by their nature, to be more discursive than practical. Knowledge of tables and chairs tends to be practical consciousness, while obscure empirical facts, such as ‘black holes,’ tend to be discursive in nature. By and large, liberalism tends to be more discursive consciousness than nationalism. Liberalism is more akin to black holes, while nationalism tends to be like tables and chairs. The average social actor does not believe that they need to resort to theoretical physics in order to understand tables and chairs, while they do for black holes. They believe that political theory is necessary for the justification of liberalism, not for nationalism. Hence, liberalism tends to be associated with political philosophy while nationalism is seen primarily as a social phenomenon and academic writing on liberalism tends to be from the perspective of political theory, whereas nationalism is treated sociologically. There are no ‘Andersons’ or ‘Gellners’ of liberalism because, like theoretical physics, liberalism is not considered to be a social phenomenon—it is the march of ‘socially unencumbered reason.’ Of course, both these claims are false: liberalism is a social product in the same way as nationalism is and a proper understanding of nationalism presupposes some fairly complex discursive normative theory.

The Origins of Nationalism and Liberalism

Nationalism and liberalism are inextricably bound up with a change of practical consciousness associated with modernity. This fact is not altered by the existence of great liberal and nationalist thinkers—Hobbes, Locke, Herder and Fichte. To be successful even the most highly discursively ideology ‘floats’ in a sea of practical consciousness. While there may have been instances of liberalism or nationalism in the feudal or Classical world, these ideas could not constitute a socially influential ideology without a shift in general practical consciousness knowledge. In the medieval world the language of liberalism and nationalism could not have been generalized. Social actors were neither the autonomous individuals of liberalism nor primarily members of nations, as in nationalism. People were members of local communities, families, status groups and religious faiths, all of which created obligations which made liberal individual autonomy a conceptual impossibility and membership of a nation, if existent at all, just one among many obligations.

The decline of teleology was central to the change. In a teleological system objects move or change because of some essence within them. As argued by Aristotle, what makes an acorn become an oak is the essence of ‘oakness’ within the acorn. The ultimate end or telos is the prime cause of change. The world of teleology was one of essences realizing themselves. Planets did not move according to laws of physics, which are indifferent to purpose, but because of telos. The revolution which Galileo and Newton brought about in physics was the replacement of a purposeful teleological cosmos with an essentially purposeless universe driven by laws that are expressed mathematically.

In politics the teleological world manifested itself in a complex hierarchy of political institutions that reflected particular essences. The ‘Great Chain of Being’ manifested divine purpose whereby aristocrats embodied a different essence from feudal serfs. In social practice this entailed a great cultural difference, which reinforced and legitimized the hierarchy of the feudal world. As described by Elias (2000), the local aristocracy deliberately lived and behaved differently from peasants. They dressed differently (as expressed in sartorial laws), ate differently (using knives and forks while peasants used their hands) and, frequently, even spoke a different language. There is a medieval Danish saying which encapsulates this: ‘An aristocrat speaks Latin to scholars, French to his peers, German to the peasants and Danish to his dogs.’

The adoption of particular forms of behaviour made classes different in their predispositions in much the same way as peoples from disparate cultures are different in today’s world. The legitimacy of the system was rooted in this cultural difference, which created the illusion of deeper essential difference between classes. The manners of the elites were not perceived as a manifestation of ‘arbitrary’ social practices but were considered to reflect essential differences ranging on a continuous scale from the humblest plant, through human society, to God. Any attempt at violating this hierarchy was ‘unnatural’ because it represented violation of essence and was likely to bring divine retribution—witness Shakespeare’s Macbeth.

The vision of the Enlightenment was a continuous one, in which reason was expected to expand indefinitely and everything was reducible to rational laws. Kant defined enlightenment as the courage to think purely according to reason, while authority and tradition are sources of error (Kant 1970 [1784]). This continuity entailed that there was a conceptual move from the physical world to that of politics. Beginning from the premise of a ‘state of nature,’ which precludes tradition and authority, Hobbes (1914 [1651]) argued that he was constructing politics that ‘mirrored’ the natural world and likened it to building a clock—a machine that moves according to the laws of mechanical physics.

Since people along the Great Chain of Being had different essences, it was considered natural that the law to which they were subject should reflect this—different laws for different people. In physics the laws of gravity apply equally and, so too, in liberalism all individuals are subject to the same law, regardless of life plan. Consequently, the different cultures of the hierarchy of the feudal world become quaint arbitrary irrelevancies.

The liberal view of the self is paradigmatically represented by Rawls’s description of the ‘original position,’ which he considers the premise for all considerations of justice. In the original position social actors are behind a ‘veil of ignorance’ whereby ‘no-one knows his place in society, his class position or social status, nor does anyone know his fortune in the distribution of natural assets and abilities, his intelligence, strength and the like …’ (Rawls 1971:12). This view of justice entails that legitimate political institutions must bracket precisely the characteristics which were essential to feudal legitimacy. The liberal selves are ‘unencumbered’ (Sandel 1982), or indifferent to specificity, and the state must be ‘neutral’ in order to be legitimate. It is not that telos disappears entirely from the liberal interpretative horizon but becomes separated from the public realm. As observed by Hayek (1960), we must distinguish between political systems (composed of anonymous individuals thrown together by chance) and organizations, which are deliberately created for specific purposes. Telos becomes confined to the private realm of organizations.

This fundamental social change of interpretative horizon reflects (and legitimizes) the rise of the bourgeoisie. The private world of commerce works according to laws of economics which, like gravity, know no exceptions. As argued by Weber, it is a world that functions best when it conforms to principles of calculated rationality, one in which precise calculability of profit makes reinvestment possible. Of course profits were made in the feudal world, but they were not derived from this form of precise instrumental rationality—from fortuna, not double entry book-keeping.

In feudalism what could be bought and sold was frequently tied to status, while in capitalism all that counts is the ability to pay. This indifference to the identity applies to the internal structure of the firm, which works according to bureaucratic instrumental rationality. The good bureaucrat applies means-ends rationality to all things and people irrespective of ‘irrelevant considerations.’ People are like things that, in principle, are interchangeable (numbers in a file) and are the conceptual equivalents of the unencumbered selves of liberalism.

As has been argued by Spruyt (1994), an essential element in the triumph of the bourgeoisie was the emergence of the sovereign territorial state, which made the world of commerce easier. The interests of the early bourgeoisie were fundamentally at odds with the confused hierarchies of the pre-modern world. They wanted one legal system where goods could not arbitrarily be seized or taxed based upon ancient feudal privilege and desired one monetary system and single set of weights and measures.

The emergence of a centralized state is essential to liberalism. Hobbes (1914 [1651]) is regarded as one of the founders of modern liberalism (although he believed in absolute sovereignty) because he argued that all authority should be vested in a centralized state, in contrast with the diffused power structures of feudalism. The reason that the liberal state claims a monopoly on violence is that it is also the sole source of political authority.

While the unencumbered self of liberalism is inherently rational, it has to be created through socialization. The world of interchangeable individuals presupposes that they are relatively similar. While the feudal world presupposed socialization that made classes dissimilar, in contrast, large contiguous industrialized territorial states are premised upon a relatively homogeneous people. As Gellner has argued (1983), the only method of achieving this outcome is by moving socialization from the home to the state. In essence, a common educational system functions as a form of mass state-controlled socialization.

When an employee is hired, their personal biography, lineage and descent become irrelevant. A common education ensures that their social practices are relatively predictable in accordance with established norms (‘lateness,’ ‘negligence’ and ‘disobedience’ etc. should have been eradicated: Foucault 1979: 178) and their educational qualifications can tell you exactly what the person is capable of. However, this mass socialization through education makes culture a political issue. For instance, a local dialect or language has to be chosen as ‘standard.’ While the Danish peasant was content to speak Danish at home and the local lords to converse in French, suddenly everyone had to speak Danish and, of course, this raises the question, whose Danish or which dialect? the answer was, unsurprisingly, Danish as spoken by the bourgeoisie of Copenhagen. Dictionaries were compiled, grammars written, which became the basis for ‘school’ Danish. Slowly Danish dialects were eradicated, except in Norway where it was decided to standardize different dialects and call the result Norwegian. Similarly, in all the major European sovereign territorial states (Spain, France, Italy, Germany and Britain) a single dialect becomes standardized as the language of the nation. The language that you speak, or your children are compulsorily socialized into, becomes a political issue and nationalism is born.

State control over education is not only functional to capitalism but reflects both the liberal view of individuals as essentially equivalent and the nationalist perception of a people or nation who share common socialization. Common education is also central to legitimating capitalism. Individuals who share common state-monitored education are the essential starting point for meritocracy. In theory (although frequently not in practice), effort should be rewarded and educational qualifications reflect this. The fact that theory and practice do not always correspond results in temporary localized perceptions of illegitimacy but even this failure leads to renewed efforts to use education as a tool of meritocracy, which is central to liberalism. Individuals come into the world unencumbered and, entirely by their own efforts, succeed. For this reason so much of A Theory of Justice (Rawls 1971) is devoted to this subject.

Nationalists do not present the standardization of a particular local culture as a form of homogenization but as a return to traditional community. This dual aspect of nationalism led Gellner to argue that there is an essential deception at the core of nationalism whereby it is really ‘the establishment of an anonymous, impersonal society, with mutually substitutable atomized individuals …’ which presents itself as a return to the peasant virtues of the volk (Gellner 1983: 57). In essence it is a Gesellschaft presenting itself as a Gemeinschaft (Gellner 1997: 74). However, as has been argued by Delanty and O’Mahony (2002: 73), this type of false consciousness argument is theoretically unsatisfactory because it reduces social agents to dupes who are mistaken about their beliefs.

The founders of modern sociology (Marx, Durkheim and Weber) shared the common Enlightenment misconception that modernity was a move from Gemeinschaft to Gesellschaft, from traditional communities to societies governed by abstract reason. In all probability, this influenced Gellner into believing that, as a modernizing force, nationalism had to be a Gesellschaft, irrespective of what nationalists believe. However, I would argue that in the transition to modernity Gemeinschaft does not disappear, or is not overcome, but becomes transformed, even if it appears counterintuitive that the disenchanted, individualistic and de-essentialized modern world could be fertile soil for a nationalist Gemeinschaft.

Following Beck (1992) and Giddens (1990), modernity entails an increase in levels of reflexivity In traditional societies social structures are taken as given and tradition confers sanctity upon them. In contrast, in a reflexive society actors are constantly expected to justify social structures, which, I would argue (unlike Beck and Giddens), is deeply problematic (Haugaard 2002: 122-37). The reproduction of structure by individual social agents, or structuration (Giddens 1984), presupposes knowledge of social life which, when tacit, is unproblematic. However, if it is reflexive, it is no longer routine. If actors are called upon reflexively to justify everyday social practices, this entails a massive transfer of taken-for-granted knowledge into discursive consciousness. However, while such a transfer may be critically liberating, it can lead to massive ontological insecurity, whereby actors feel ill at ease with their being-in-the-world. Competent social agency entails viewing social convention with a ‘natural attitude,’ whereby social life remains as a given. This is not simply because the complexity of social life demands that most social knowledge remains practical consciousness but because there is a real danger that the structures of social life dissolve once converted into discursive reason, as they are largely arbitrary social constructions that cannot be justified through reason.

Giddens observes that modern reflexivity entails that your life history is of your own making (1991), while in traditional societies, careers, sexual orientation, gender and so on were largely predetermined. While Giddens views this reflexivity as inherently empowering, I would argue that this invitation to construct your own history has the potential to be quite the opposite: an invitation to meaninglessness. Constructing self, coupled with the realization that social structures are an arbitrary convention, results in ontological insecurity. While most social actors are forced to choose careers within the industrial system of modernity, they do not have a desire to choose their identity in its entirety, especially in the knowledge that such a choice is between arbitrary conventions. Hence, they may embrace the idea that while being a carpenter or high court judge is not predetermined by family lineage, there are other aspects of identity that should remain foreclosed—you simply are Irish, Japanese or Norwegian. Which is why we may ask a child of a carpenter if they wish to become a firefighter or judge but it would be most unusual to enquire of an English child if they wished to become Japanese or Norwegian.

In the minds of most people nationality is part of their inherent being-in-the-world. Of course, people can change nationality but this is often viewed as suspect. As has been argued by Bauman (1989: 52-6), the Nazis felt more threatened by Jews who were indistinguishable from other Germans than they did by Orthodox Jews. The point is that the former demonstrated, through their social competence, that ‘German-ness’ was a social construction, which they could choose to adopt. Consequently, they became an essential reminder of the arbitrariness of nations, one that had to be eliminated.

Nationalism entails that identity is a given and, as a form of Gemeinschaft it means that there is something beyond the self securing the structures of social life from degeneration into arbitrary conventionality. Following Durkheim’s Suicide (1952), Gemeinschaft, or what he termed ‘mechanical solidarity’ (I am subsuming the two), entails that the self depends for its existence upon an external whole in which community is more than the sum of its parts. This totality is not seen as an arbitrary construct but is reified; consequently everything that is embodied by the nation becomes more than mere convention. As nationalism embodies the socialization of a people, this entails that the entire way of life of a people becomes deconventionalized and, thus, saved from the inherently deconstructive tendencies of modern rationalism.

The reifications used are complex and varied. The doctrine that the nation is created by God is a relatively common one. It is most obviously manifest in the doctrine of ‘God’s chosen people,’ which is central to Zionism but also found in the nationalism of Dutch settlers in South Africa, Ulster Protestantism (Akenson 1991), American and Basque nationalism.

Nature has traditionally always been considered beyond convention. So, nations are frequently naturalized and it is claimed that belonging to nations is some kind of natural inclination, which is common to all humankind. Primordialism affirms the naturalness of nations by indefinite extension into the past. A diluted version of the primordialist thesis, which nonetheless preserves this form of reification, is the doctrine that nations are essentially equivalent to the tribes of traditional societies.

Science is another way of reifying nationalism. Biological genetic theories of race are the most obvious and enjoyed a vogue in the nineteenth century but generally have fallen out of favour since 1945. A different mode of making nationalism scientific is Darwinian evolutionary theories, which claim that nationalism is a manifestation of needs that are essential for the survival of the species. For instance, van de Berghe (1981) argues that the nation is a natural extension of the family.

The nation can be reified by claiming that it has a unique civilizing role. As a ‘higher civilization,’ a national culture is not some arbitrary form of social convention but a unique achievement, which transcends convention in some deeply metaphysical way—for example, French, English, Italian, Japanese and Norwegian nationalism. A variant on this is the claim that, even though the national culture is not uniquely civilized at the moment, there was a past ‘Golden Age.’ Decline was forced upon the nation but national awakening contains the promise of a return—for example, Irish nationalism. The Hegelian notion of the nation representing the march of reason would be another variant on claiming a unique civilization.

What all these claims have in common is the reification of the nation as something other than arbitrary convention. The ontological security that this provides is coupled with an attempt to render the nation sacred through ritual. The flag becomes iconic and the yearly calendar punctuated by collective rituals, such as the national day, and more routine and everyday rituals, which include diplomatic and sporting events—banal nationalism (Billig 1995). These share more with religious belief than they do with instrumental disenchanted logic. Large crowds come together and chant encouragement at ‘their’ national team, while they wave the national flag, which serves a similar function to a sacred totem in primitive religions. Supporters frequently consume large quantities of alcohol, which has the effect of intensifying feelings of excitement and consequent enchantment (Durkheim 1995: 228).

In essence what is being argued is that while liberalism is consonant with Enlightenment rationality, the latter threatens the ontological security of everyday life by showing cultural practices and identity as part of essentially arbitrary cultural conventions, which become disenchanted. In contrast, the appeal of nationalism is precisely that it does the opposite. The practical consciousness of the nation becomes other than arbitrary convention, which can be cherished by the state and, when placed in the context of banal rituals, becomes enchanted. The functionality of nationalism to modernity stems from its homogenizing effects while it appeals to social actors because it provides ontological security and enchantment in a social world where reflexivity tends to undermine it. This explains why nationalists find constructivist accounts of nationalism threatening. It is not that they do not understand that social constructivists take ‘the social’ seriously (Gellner 1997) or that ‘imagined’ means ‘imaginary’—Anderson—but they intuit that the appeal of nationalism lies in its capacity to overcome the ontological insecurity entailed by constructivism.

If ontological security is rooted in practical consciousness knowledge, it is logical that nationalism should be a less discursive ideology than liberalism. Nationalism is part of the practical knowledge of competent social agency and the reifications, which make nations non-conventional and sacred, entail that the nation does not require intellectualization. The nation just ‘is’ and to theorize it in the language of sociology is irrelevant and smacks of profanation.

The perception of the nation as non-conventional and sacred gives externality to the social actor which renders their identity enduring. For the liberal self, the autonomous self has no meaning outside itself, hence it makes no sense for the self to be sacrificed for a collective social construction that is arbitrary convention. In contrast, for the nationalist the nation is what gives meaning to the self as a socialized being. Socialization makes them part of the nation, which is made real by the belief that it is beyond convention and part of the enchanted world. To such an actor the ultimate self-sacrifice, dying for the flag, represents a union between the self and the real. As Durkheim argued (1952), such self-sacrifice is not self-annihilation but self-realization. The dead soldier or suicide bomber lives on through their fusion with the nation. Dead liberals no longer count, but dead nationalists are still real as long as the nation lives on.

To nationalists, betraying their dead is both a betrayal of the nation and a betrayal of the self. In negotiations with Sinn Fein/IRA it is noticeable that the views of all those who died for Ireland are considered particularly significant. Within this the views of those who died on hunger-strike in 1983 are given particular weight. Suicide for the cause renders these members ‘real’ through ‘fusion’ with the history of the nation.

If the hypothesis is correct that nationalism is essentially a modern Gemeinschaft, while liberalism is a Gesellschaft, the question has to be raised: how is it possible that modern social actors subscribe to both ideologies? The easy, although sociologically unsatisfactory, answer might be that they represent different people—liberal moderns and nationalist moderns. However, I would argue that the majority of social actors are nationalists and liberals simultaneously. In modernity there is a latent promise of a great fusion and singularity of interpretative horizon—once reason develops, all knowledge will become integrated. Yet, the reality is that modernity has frequently resulted in fragmentation (Delanty 1999). Indeed, fragmentation is central to competent social agency in the modern world. Consider Weber’s characterization of the modern bureaucrat with the capacity to switch between instrumental purposive rationality at work, to affective rationality at home. Of course, there are moments of conflict (for example, when a much-loved relation is the number on a file), but the routine functioning of a bureaucratic machine is premised upon the ability of social agents to switch interpretative horizon. In Goffman (1969), the competent social agent can switch from ‘backstage’ to ‘frontstage’ monitoring of their behaviour. Similarly, modern social agents have both a liberal and nationalist interpretative horizon, which alternate as circumstances demand. Modernization theory predicted the advance of scientific rationality would result in the demise of religious belief but, contrary to expectations, many social actors find it possible to be both technically competent and religious. It is not that some consistency between science and religion has been discovered, rather that social actors switch, depending upon circumstances, between scientific and religious interpretative horizons. I am not claiming that this discontinuity does not create moments of internal conflict when it is not clear which interpretative horizon is appropriate to the circumstances. In Sartre’s famous conundrum: in a state of war, should a son take care of his sick mother or defend the nation? The irresolvable conflict is caused by the fact that the question can be answered either through a nationalist or an affective interpretative viewpoint. Similar conundrums are raised between nationalism and liberalism: in a state of war, should the state do everything in the interests of defending the nation, including curbing freedom of speech? However, both are scalar concepts and when there is a conflict between these ideologies, degrees of intensity are crucial to determining which ideology wins.

Normative Considerations

While nationalism and liberalism are functionally compatible with modernity, I would argue that they are normatively theoretically incommensurable. The basic premise of liberalism is the singular individual who, in considerations of politics, has the capacity methodologically to bracket their particularities. In contrast, the basic unit of nationalism is a self who realizes themselves through a collectivity which embodies significant aspects of ‘their’ socialization. For the liberal, the ultimate good is the autonomy of self, while for the nationalist the ultimate end is the autonomy of the nation. For the liberal the legitimate state is a neutral one, while for the nationalist the legitimate state has a telos—the flourishing of the nation. The liberal individual critically subjects their beliefs to discursive examination, while the nationalist believes that the nation is a natural entity and practical consciousness.

This is not to argue that a liberal nationalist theory cannot be constructed, which may be internally consistent, but at the expense of the very characteristics which are central to the success of nationalism. For instance, Tamir (1993) attempts such a fusion by premising her liberal nationalism on the liberal self. Tamir’s self is an autonomous chooser who chooses nations in the same way as they select a career which, of course, misses precisely the attraction of nationalism in a uncertain world, that is, part of your identity is not subject to radical reflexivity Miller (1995) also begins from liberal premises but his nationalism is of a weak variety, which amounts to little more than a slight preference for co-nationals over non-nationals in regard to welfare provisions.

Moore (2001) is unique among these attempted fusions in working from nationalist premises to liberalism—a nationalist liberalism. Her opening premises are: first, that national identification is a legitimate constitutive part of the self and, secondly, that in practice liberal states are never entirely neutral. When these elements are placed together it appears that legitimacy demands that the nationality of the self and the state be congruent. After all, if the state represents one particular nation, it is an injustice if that state has jurisdiction over a different (minority) nation. In those circumstances secession becomes legitimate—either full or partial depending upon the practicalities. However, her second premise, concerning the neutrality of the state, is flawed because she treats liberalism in absolute terms, not as a scalar concept. Of course it is correct that no state can be entirely neutral but this is not the same as arguing that there is not a significant difference between a state that aims at neutrality as an ideal and one that considers its legitimacy inherently linked to its ability to promote the will of the nation. Liberal states are not liberal by virtue of their ability to be entirely neutral; rather they are more and less neutral. To take an instance, in the 1937 Irish constitution, the Catholic Church was given a special status that reflected the (supposed) views of the majority nation, although other churches were allowed to continue to practise. In 1972 this clause was removed and there began the slow process of separating church and state—a process that is still ongoing. In this case we would argue that the Irish state became more liberal with the removal of the clause privileging the Catholic Church. This is a meaningful statement even though it was not entirely illiberal prior to the removal of the constitutional clause: after all, other religions were allowed freedom of worship. Given that it is possible for the state to be more or less neutral, the fact that it is impossible for a state to be entirely neutral (in every respect) does not entail that claims to state neutrality are bogus, even if they are relative. A state with high levels of impartiality may, in practice, be substantially more just with reference to competing ethnic claims than secession into avowedly nationalist states—especially if we take account of the fact that these states will have jurisdiction over minorities of ‘the other nation,’ other cultures and, indeed, individuals who may not wish to be defined in nationalist terms.

While a fusion of nationalism and liberalism into one ideology can be achieved only at the expense of removing what is central to the ideological success of nationalism, this does not lead to the conclusion that one is to be jettisoned at the expense of the other. I would argue that the ability to switch between interpretative horizons for different conceptual problems is intrinsic to modernity. This parallels Mouffe’s (2000) argument concerning democracy. What we call democracy is actually made up of two elements, which are frequently in conflict with each other. On the one hand there is the liberal tradition that emphasizes the rights of the individual and, on the other, the democratic participatory perspective with emphasis upon collective decision-making. The answer to a conflict between the two is not to try to subsume one into the other but to view this as a constructive agonistic tension, which is central to democratic debate. Similarly, I would argue that the conflict between nationalism and liberalism is a constructive force. Even if Tamir, Miller and Moore do not succeed in constructing a single ideology, part of the initial plausibility of their argument derives from the fact that, in practice, liberalism leans heavily upon nationalism. For instance, developed to their logical outcome, the premises of liberalism lead to international cosmopolitanism. Yet, all liberal states exclude some people from the citizen body. Nationalism provides a ready-made solution to ‘who are the people?’ However, such a solution can never be made consistent with liberalism. Instead of attempting to reconcile these ideas, we should view them as in constructive tension. The liberal state is less liberal because of this exclusion but, as it is a scalar concept, this does not entail that it is entirely illiberal.

The result of living with this kind of constructive tension is similar to Walzer’s (1983) description of spheres of justice. Walzer argues, for instance, that some issues are decided by principles of the market while others are decided by meritocratic principles or principles of need. A doctor is appointed on meritocratic principles and we consider it wrong that such a position could be bought or given to the person based upon need. In contrast, we think it correct that medicine be distributed upon principles of need. Of course, not all issues come with a ready-made decision as to which interpretative horizon is appropriate but this ambiguity is constitutive of democratic debate. When we define citizenship, there may well be a creative tension between applying liberal principles and wider or narrower definitions of the nation.

Viewing the tension between liberalism and nationalism as agonistic and seeing both as scalar concepts, allows us to accept that the workings of liberal societies presuppose some nationalist premises. However, recognizing the tension, in place of opting for one over the other, allows us to avoid certain liberal and nationalist excesses.

There are four principle issue areas in which liberal practice presupposes nationalism. The first is relatively self-explanatory, nationalism answers the question: ‘Who are the people?’ Secondly, as argued by Miller, nationalism both limits the numbers of welfare recipients of the liberal welfare state to practical levels and facilitates redistributive justice (1995: 92-4). Without nationalism, liberal welfare states would find it difficult to justify giving welfare to co-nationals over those in greater need in distant places. How could the Swedes ever justify their welfare state when people in Ethiopia are in so much greater need? Nationalism also makes redistribution easier. The idea of the nation as a community, which is ‘ours,’ entails that any redistributive welfare taxation is giving from ‘me’ to ‘us.’ Even if this is reprehensible to a committed liberal, the belief that welfare is going to one of ‘us’ may be necessary for the welfare state to function. It is a necessary condition, but not a sufficient one—it does not follow that nationalist states are invariably generous in their welfare provisions.

Thirdly, nationalism contributes to liberal practice by providing sufficient common culture for the effective functioning of the type of reflexive debate central to liberalism. On its own, liberal debate is premised upon a kind of Kantian pure rationality. However, as has been argued by Fichte (1922), our ability to reason is not simply an unencumbered abstract one.

Reasoning without meaning is a conceptual impossibility. This is not simply J. S. Mills’s practical observation that democracies function better with a common language (Mill 1972 [1861]: 359-66) but the more profound point that the ability to reach consensus through disputation presupposes convergence of interpretative horizons. When Rawls wrote A Theory of Justice (1971), he argued that his liberal principles were based upon universal premises, but with Political Liberalism (1993), he had come to see that an ‘overlapping consensus’ was necessary for liberalism. I would argue that part of this ‘overlapping consensus’ is not only a certain civic spirit (following de Tocqueville) but also a certain commonality in systems of meaning. While I would agree with Habermas that liberalism presupposes something like an ideal speech situation, in which the most convincing argument wins on its own merits, like Ackerman (1980), I would argue that this ideal is only possible between actors who share certain fundamental ways of seeing the world, presupposing a substantive degree of shared culture. Consequently, while the homogenizing task of ‘nation-building’ would be anathema to any serious liberal, the outcome of such a process is a community that shares sufficient cultural similarity to allow persuasion by the force of better argument.

Fourthly, nationalism provides the conditions for liberalism by providing the state with citizens who are willing to undergo significant sacrifice in its defence. An entirely rational liberal is unlikely to put their life at stake for the sake of the state, while the altruistic Gemeinschaft nationalist is willing to die for the flag.

In all these instances, in which nationalism creates the conditions for liberal states, it is the case that these contributions come from a nationalist interpretative horizon that, in many respects, is antithetical to liberalism. Defining ‘we the people’ in strongly nationalist terms can lead to xenophobia in citizenship and welfare policies. Nation-building can result in illiberal enforced assimilation of minority cultures and the nationalist desire of self-sacrifice can lead to belligerence and ‘suicide bombing.’ However, I would argue that this is less likely to take place if we recognize the essential tension that exists between nationalism and liberalism, rather than blind our critical faculties through some kind of enforced fusion which, in any case, will not become a popular ideology since it has lost the Gemeinschaft characteristics that have made nationalism an ideologically successful outcome of the conditions of modernity.