Corps de l’article


This paper provides a theoretical discussion with point of departure in the case of Denmark of some of the theoretical issues concerning the relation liberal states may have to religion in general and religious minorities in particular. In liberal political philosophy, it has long been taken for granted that a liberal state has to be a religiously neutral state. The ideal liberal state has to be institutionally somewhat similar to the US; with regard to religion, liberal political philosophy is to a large extent a theoretical interpretation of the US constitution’s First Amendment’s anti-establishment and free exercise clauses.

There are many complicated theoretical discussions of the notion of neutrality, including criticisms of its coherence or desirability (e.g. Sher, 1997) and whether neutrality really requires non-establishment (Holtug, 2009), which I will not even attempt to go into here. Instead I want to ask: how can and should we understand theoretically what a liberal state is with respect to religion and religious minorities if it is not a strictly religiously neutral state with full separation of church and state and of religion and politics?

This question can be understood in several ways. If one shares the theoretical concerns about the requirement of liberal neutrality, then the question concerns an alternative formulation of what liberalism means in theory with regard to religion. If one still endorses the ideal of liberal neutrality in some form, then my question is rather how we should understand non-ideal real-world liberal states, which are often far from strictly neutral, i.e. what it is about non-neutral states that still makes them recognisably liberal? I will sketch some relevant considerations in relation to this question through an investigation of a particular case of an arguably reasonably liberal state, namely the Danish state. My question is how one might understand its relations to religion and what it might teach us about the possible relation between religion and liberal states.

The Danish case of state-religion relations is chosen as focus for a number of reasons: First, the Danish state is in many crucial respects a distinctively liberal state; it is a democratic and (in a sense to be explained) non-confessional state with wide-ranging religious freedom and other liberal rights. Secondly, it is not a neutral state with separation of church and state; the Danish state is rather what has been called a “moderately secular” state, which is arguably neutral in some senses and respects, but not in others. This is not a peculiar phenomenon since most, if not all, European states either have established churches and state support for organized religion or regulate and uphold various links to organized religion, even in the French case (Bowen, 2007; Modood, 2007); the Danish state is merely a particularly stark example of the general European combination of religious neutrality and non-neutrality, relative separation and institutional involvement.

The theoretical question then is: a) how should we understand state-religion relations in such cases, i.e. which theoretical concepts should we use to describe these relations if we cannot simply invoke neutrality and separation, and b) how do we relate normatively to state-religion relations if a state can be liberal without being religiously neutral, either in principle (if one accepts the theoretical criticisms of the idea of neutrality) or at least in non-ideal practice?

The paper will present the Danish case and show how complicated it is to apply standard theoretical concepts such as neutrality, separation, toleration, respect and recognition to it. My focus will partly be on the constitutionally defined role of the Evangelical-Lutheran church as the “People’s Church”, partly on the way the Danish state relates to religious communities other than that of the Lutheran church.

The paper discusses some central liberal theoretical concepts in relation to this case. The first is neutrality. I argue that even though the Danish state is not strictly religiously neutral, we should distinguish between neutrality in different respects and dimensions. It then becomes less clear that the religious non-neutrality of the Danish state makes it especially illiberal.

The paper proceeds to consider whether the Danish state, since it is not strictly religiously neutral, should rather be understood as being tolerant towards religious minorities. The paper shows, first, that there are a number of conceptual complications in applying the (attitudinal) concept of toleration to the state in general and to the relation between the Danish state and religious minorities in particular. It is argued that one cannot in any straightforward manner conclude from the existence of state support for an established church that the state is opposed to other religious communities in the way required for (traditional) concepts of toleration to apply to it.

The paper moves on to further theoretical issues deriving from the debates on multiculturalism, where traditional liberal ideals of neutrality and toleration are often rejected in favor of some form of public recognition of minorities by the state. I argue that the concept of recognition is better suited than notions such as neutrality and toleration to describe Danish state-religion relations and that this need not in itself make them illiberal. But the paper also notes a number of problems for proponents of policies of recognition, namely that recognition may involve misrecognition and inequality.

My aim is not to find the “real” historical rationale for the Danish arrangements, which are complex products of historical evolution influenced by many different ideas and forces, and therefore unlikely to embody one single or coherent idea or rationale. I rather try to use such a messy real life case to illustrate some challenges faced by our theoretical models. I take it that we want our theoretical models to help us both describe and assess real life cases. If we are liberals, we want to have theoretical concepts and normative standards to guide us in assessing the liberality of states and help us understand what their liberality consists in. The paper uses the Danish case in two ways: On the one hand it uses it to point out how commonly invoked concepts like neutrality and toleration may simply be descriptively inaccurate in relation to states like the Danish one. On the other hand it uses the case to question formulations of liberalism, especially as requiring religious state neutrality; if a state can be reasonably liberal without being strictly religiously neutral perhaps religious neutrality has to be qualified as a central requirement of liberalism?

My conclusion will neither amount to an all things considered assessment of the case or a general view about how liberalism should be understood. I will more modestly try to illustrate both how a non-neutral state might reasonably be characterized as liberal, how its relationship to religions challenges concepts of religious neutrality, toleration and recognition, while indicating some remaining liberal normative worries about various aspects of the case.

Moderate secularism in practice: The Danish Case

Denmark provides an interesting case for present purposes because the Danish institutionalisation of the state-religion relationship exemplifies what Tariq Modood has termed “moderate secularism”. This is due to the existence of an established church sponsored by the state, with substantial inequalities in the statuses enjoyed by different religious minorities or communities (Modood, 2007; Laegaard, 2008). Moderate secularism captures the actual types of state-religion relations in Europe, which are not characterized by absolute separation demanded by more radical forms of “ideological secularism”. European states do not ignore or remain aloof in relation to religion in general or religious groups in particular, but recognise and regulate religion and religious groups in a variety of ways. This is the “moderate” aspect, which consists in the ways they depart from the traditional liberal idea of secularism as separation and neutrality. Although not upholding absolute separation, moderate secular states nevertheless secure a degree of mutual autonomy of state and religious communities, which is why they still qualify as “secular”. So Modood’s idea is that secularism is not an either-or, as the traditional liberal interpretation in terms of religious neutrality suggests, but a matter of degree. Furthermore, he claims that the real or genuine sort of secularism is not necessarily the most radical forms, but that there are reasons why the more moderate versions are actually preferable. These reasons have to do with the fact that the moderate state-religion relationship allows states to actively accommodate religious minorities and treat organised religion as a public good (Modood, 2010: 5-6).

In this section I will sketch the institutional relations between the Danish state and religious communities with a point of departure in the Danish state’s own official presentation of this scheme (Ministry of Ecclesiastical Affairs, no date). The point of the sketch is to function as a basis of a discussion of how different theoretical (conceptual and normative) models apply or fail to apply to the case. These are all models that have some theoretical connection with liberalism, so in this sense this is a discussion of in what ways the Danish state is liberal in its relations to religious communities.

The main characteristic of the state-religion relationship in Denmark is the fact that, according to § 4 of the constitution, the Evangelical-Lutheran church is the so-called “Danish People’s Church” (“Folkekirken”) or National Church and “is supported as such by the state”. The Danish head of state, which, since Denmark is a limited constitutional monarchy, means the king or queen, has to be a member of the national church. While the national church is formally distinct from the state (hence the designation as a “National” or “People’s” church rather than a state church) it is described in § 4 of the constitution as a fourth pillar of the state besides the three standard powers described in §§ 2 and 3, namely the executive (personified by the monarch, in whose name the government rules), the legislature and the courts (Christoffersen, 2010, p. 147). So the church is part of the state while simultaneously being distinct from all other organs of the state. In practice, state and church are connected in a variety of respects. The state is the legal subject of the national church, whose highest authority is the national parliament and which is administered by the Ministry of Ecclesiastical Affairs. The state also funds the church, partly by way of church tax collected as part of the ordinary taxation from all members of the national church. But the state also funds church buildings, education of priest and the Ministry of Ecclesiastical Affairs by way of the ordinary state taxation levied on all residents whether members of the National Church or not. In 2008, 82% of the Danish population were members of the National Church, although the share is continually dropping.

So Denmark has a fairly strong form of established religion. In practice there is nevertheless a significant degree of separation of national church and state; the doctrinal creed and similar issues of religious content of the national church are not subject to ordinary parliamentary regulation since they are constitutional issues; the Ministry of Ecclesiastical Affairs only makes administrative, non-theological decisions; day-to-day affairs of the church are run at the local level by elected parish councils.

In political practice, the non-separation of church and state means that the church is dominated by the state, which has never implemented legal rules for internal decision making within the church as promised in the constitution of the Danish state (§ 66), probably because that would allow for the church to form its own opinions on religious or political matters. So the relationship is asymmetrical; the state is separated from the church in the sense that the church has no way of affecting national politics (in this way the Danish national church is in a weaker position than, e.g., the Anglican church in Britain), but the church is not separated from the state, since the state holds ultimate authority over the church (in this respect the Danish national church is in a peculiar and probably uniquely weak position among organised churches, whether established or not, since it has no internal decision making body whatsoever). But at the same time the status of the national church strongly signals that Denmark is a Christian country, and that the state is in some sense a Christian state.

Because of its history as a direct institutional continuation of the church of the absolutist state predating the 1849 constitution, the national church not only performs religious and symbolic functions, but also carries out important executive and administrative state functions. The most important example is that civil registration and the issuing of birth certificates are taken care of by the parish offices of the national church. This means that the records of all people born in Denmark (with the exception of Southern Jutland, which was only reunited with the Kingdom of Denmark in 1920, and people associated with recognised religious communities, cf. below) are kept by the national church and that all parents (until recently, when online registration become an option) had to register their newborn babies at their local parish office. Such purely administrative practicalities have a symbolic effect reinforcing the signals sent by the general church-state relationship in Denmark.

The fact that Denmark has an established church already implies that there is not religious equality in Denmark; the national church is the only church supported by the state. There is religious freedom in Denmark, however, which is both guaranteed by § 67 of the constitution, according to which citizens have the right to unite to worship God according to their convictions as long as this is compatible with decency and public order, and § 9 of the European Convention on Human Rights, which has been incorporated into Danish law. But religious inequality is not simply a matter of all forms of religion deviating from the Lutheran Christianity of the national church being second rated to an equal degree; there is also a complicated gradient of the statuses assigned to religious communities deviating from the national church. Officially, there are three classes of religious communities in Denmark other than that of the national church (Ministry of Ecclesiastical Affairs, no date; Simonsen, 2002).

First, there are the so-called “recognised” religious communities. This is the designation of the status assigned to religious communities other than the national church until 1970 (when a new marriage law came into effect). Recognition was bestowed by royal decree (in practice a legislative act of the parliament) and was extended to 11 religious communities, including the Jewish Mosaic religious community, the Catholic Church in Denmark, reformed churches, the Methodist Church, the Russian-Orthodox Church in Copenhagen and the Baptist Church.

Secondly, there are the so-called “approved” religious communities. Approval is the status assigned to religious communities by the state from 1970 onwards on the basis of the provisions of the Marriage Act. Applications for approval are reviewed by an advisory committee consisting of independent academic experts on religion and law with a view to establishing whether the applying group fulfils non-evaluative criteria concerning its religious nature (belief in transcendent powers), practice (written creed and rituals) and organisational structure (written rules, assigned representatives and membership) (Advisory Committee, 2010, p. 7). Over 100 religious communities, including over 20 Islamic communities, have received approval.

Thirdly, there are so-called “religious societies”. This is the designation of religious communities that have not applied for or been granted approval but operate under ordinary religious and associational freedom, e.g. because they do not conduct marriage ceremonies. These communities may still apply for status as charitable non-profit associations, which makes them eligible for the same kinds of tax-deduction and other tax benefits also shared by recognised and approved communities.

Recognised and approved religious communities have a number of rights and privileges in common, including the right to perform marriage ceremonies with legal effect under the Marriage Act, the right to residence permits for foreign preachers under the Aliens Act, the noted tax benefits, and the right to establish their own cemeteries under the Cemeteries Act. There are some differences in rights, however. Religious communities recognised by royal decree before 1970 continue to have ministers approved by royal decree, they may name and baptise children with legal effect, they keep their own church registers and may transcribe certificates on the basis of such registers. This means that ministers of recognised religious communities perform some of the same executive functions that are performed for the state by the National Church, most importantly the civil registration of newborn babies (Simonsen, 2002, p. 22). Recognised religious communities are accordingly not entirely private associations, but perform some public functions that approved religious communities do not. Curiously, this part of the Danish institutional scheme is reminiscent of corporatist models of society; a feature otherwise quite foreign to Danish society with its strong unitary and universalistic welfare state.


While Danish society is one of the most strongly secularised societies in the world in terms of indicators like church attendance and prominence of religion in people’s daily lives, it is obvious from the above sketch that the Danish state is not religiously neutral and that there is not separation of church and state. I nevertheless described the Danish state as a liberal and non-confessional state in the introduction. I want to defend this characterization (which is not necessarily equivalent to an apology or endorsement of the system thus characterized) by bringing more nuances into the discourse about state neutrality.

Recent liberal political philosophy routinely discusses neutrality as a relationship between the state and persons with regard to their conceptions of the good or ethical doctrines, moves on to distinguish between neutrality of effects, aims and justifications, and often defends a more or less expansive form of the latter as a requirement for state legitimacy on the basis of some form of deontological argument about respect for persons (Gaus, 2009, pp. 81-82). For present purposes, I want to suggest the relevance of a number of further distinctions.

First, I am only concerned with religious neutrality, not with the general rejection of perfectionist reasoning in politics. While this delimitation may avoid complicated discussions about whether the justification of political neutrality itself has to and can be neutral, it may also be problematic, since religious neutrality is usually considered to be a defining feature of liberalism enjoying greater intuitive plausibility than broader doctrines of state neutrality generalised on the basis of it.

Secondly, there is reason not to speak only of state neutrality, since this elides some important distinctions between different levels of state activity. Liberal neutralists in fact disagree about what the requirement of neutrality concerns. Some are concerned with what might be called legislative neutrality, e.g. because it is through legislation that states authorize coercion of citizens and regulate their freedom (Waldron, 1989, p. 71). Others limit the requirement of (something like) neutrality to more fundamental constitutional issues (cf. Rawls’ recurring reference to “constitutional essentials and matters of basic justice”, e.g. 1996, pp. 44, 214-15, 228). Something like neutrality can also be required in relation to the practical execution, implementation and administration of law. This is what is captured in classic calls for equality before the law, non-discrimination and the rule of law; people should be treated the same in relation to a given law unless the law itself specifies that some difference between them should make a difference for how they are treated.

Thirdly, in addition to the distinction between the different levels of state activities, it seems important to distinguish between those activities that are coercive and those which are not. I will not attempt to provide any precise definition of coercion but will for present purposes rely on a common sense idea according to which taxation, the penal system and similar sanctions are clear cases of state coercion, whereas what I will call “symbolic” or merely “expressive” acts of the state need not be.

My point is simply that some state activities at some levels can be religiously non-neutral (in whatever precise sense one is interested in) while other activities at other levels are religiously neutral. My further claim is that non-neutrality is arguably more problematic from a liberal point of view when it concerns coercive acts than when it concerns merely symbolic acts. Together this implies that it may not be problematic if a state is non-neutral in some respects; it is not as if any deviation from neutrality contaminates the entire state and makes it illiberal.

In relation to the Danish case, these distinctions both help explain and justify my characterization of the Danish state as reasonably liberal despite the obvious absence of strict separation and full religious neutrality. As already described, the religious non-neutrality of the Danish state is mainly constitutional (the setting up of the evangelical-Lutheran church as the people’s church and the support of it by the state as such) and administrative (the administration of the people’s church by the Ministry of Ecclesiastical Affairs). Almost all other state activities are in practice unaffected by these specific forms of religious non-neutrality and constitutional, legislative and administrative mechanisms are in place to prevent religious non-neutrality from spilling over into other domains (e.g. § 70 of the Constitution, which specifies that no one can be deprived of access to the full enjoyment of civil or political rights on grounds of his or her faith, as well as various constitutional and human rights protections of religious freedom and provisions against religious discrimination). While the danish state is constitutionally non-neutral in relation to religion, and supports and administers the People’s Church, all other aspects of state activities are as religiously neutral as in other liberal states (which is not to say that they are neutral in other respects). Most importantly, the state is religiously neutral with respect to basic rights, legislation and the implementation and administration of law, and the constitutional non-neutrality is, so to speak, insulated from having effects on other aspects of the state.

Exceptions to non-neutrality are mainly symbolic, i.e. not coercive: The declaration that the evangelical-Lutheran church is the People’s Church is primary expressive and is not in itself a basis of state coercion. The same is the case for the various uses by the state of Christian symbolism. The only real exception to this rule is the small part of the financial support from the state to the church which is not covered by the church tax levied only on members of the church. This is both an economic spill-over effect (the funds used to support the Church cannot be used for something else) and one which is coercively imposed by the state, which is arguably problematic.


Given that the Danish state is not strictly religiously neutral, how can the state-religion relationship be understood theoretically? From the point of view of liberal political philosophy, an obvious idea is that a non-neutral state like the Danish one might rather be described as a tolerant state. This reading is based on a traditional conception of toleration as requiring a) the presence in an agent of a negative attitude, e.g. some form of dislike or disapproval, to some object, which disposes the agent to interfere with the object, b) that the agent has the power to interfere with the object, and c) that the agent nevertheless, e.g. for principled reasons, refrains from thus interfering (Forst, 2008). Toleration thus understood is different from neutrality precisely because of the first condition, which ascribes a negative attitude and resultant disposition to interfere to the agent, which for that reason cannot count as neutral (at least not with regard to aims). This is especially clear in cases where the object of disapproval is religiously defined, e.g. a religious belief or practice, and the reason for the disapproval is itself a religious belief.

The Danish state as described might be understood as tolerant for a number of reasons. First, the very fact that makes it non-neutral in relation to religion, i.e. the constitutional establishment of the Evangelical-Lutheran church as the National Church and the support by the state for it as such, might provide a reason for ascribing religious beliefs to the Danish state that could ground the kind of negative attitude to religious beliefs and practices deviating from those definitive of the national church. Secondly, the state clearly fulfils the second condition of power. Thirdly, since the state nevertheless upholds freedom of religion, it would seem to fulfil the third condition of non-interference. Together this theoretical interpretation of the relation of the Danish state to minority religions seems to fit well with how toleration and neutrality is often discussed in political philosophy; toleration is a kind of half-way house between principled neutrality and religious persecution and oppression, which retains the negative attitude to religious differences, but refrains from acting on these attitudes.

In this section I will argue, however, that the Danish state does not obviously fulfil the negative attitude condition and that it is furthermore questionable whether it makes sense to apply the concept of toleration in the suggested way. This means that, even though the Danish state is not strictly religiously neutral, it is not tolerant either, at least not in the traditional sense usually invoked in liberal political philosophy. The question then is how we might describe it instead? In the next section I will discuss how we might instead interpret the Danish state-religion relation on the basis of consideration of the way in which the application of the non-interference condition also seems to fail to capture the state-religion relation.

The negative attitude condition for toleration includes several sub-components, all of which might be problematic, either on their own or as applied to the Danish state-religion relation: First, it is assumed that the state can have religious beliefs in some sense; second, it is assumed that the having of religious beliefs implies a negative attitude of disapproval or dislike towards different religious beliefs, practices or groups having these different beliefs or practices; and third, it is assumed that such negative attitudes dispose the state to interfere with the objects of dislike or disapproval.

As to the assumptions that institutional actors like states can have beliefs and attitudes, this is in itself problematic (see Laegaard, 2010, for a more thorough discussion of this point). A modern state is not identical to any particular individual person or groups hereof, e.g. government ministers, members of parliament, civil servants, or citizens; a modern state is rather a complex political organisation defined by rules, most notably by the rules laid down in the constitution. Even though the rules can include statements such as “the evangelical-Lutheran church is the national church”, these are constitutive of certain institutional relations rather than expressions of beliefs that the institutions can be said to have. States have decision procedures, e.g. elected legislatures, and representatives who can carry out the decisions made in these procedures, e.g. the various executive branches of government, and can therefore act. These actions may even be responsive to reasons. But the fact that states can act on the basis of decisions responsive to reasons does not imply that the state acts on the basis of beliefs or attitudes that are its own; its acts are rather explained as the aggregate effects of individuals influencing its decision procedures as shaped by the institutional rules and various other factors, e.g. relations of power.

This means that if one ascribes beliefs and attitudes to institutional actors like states, these beliefs and attitudes must either refer to (aggregates of) beliefs or motives of individual persons somewhere in the state’s internal decision procedures, or are really statements about the outwards actions of the state. But in the first case, the beliefs and attitudes are not beliefs and attitudes of the state and in the second case the “beliefs” and “attitudes” are not motivational or causally explanatory factors but empirical re-descriptions of the state’s behaviour. So the assumptions that states can have beliefs and attitudes are problematic in general. In the specific case, this means that one cannot infer from the fact that the Danish state constitutionally and practically supports the national church that the state subscribes to the Evangelical-Lutheran confessional doctrines constituting the church’s creed (whatever this might mean). So it is far from clear that the state can be ascribed Lutheran beliefs on the basis of the fact that it constitutes and supports the National Church, which is part of my reason for describing it as non-confessional.

One response to this reasoning might be to say that it is sufficient for interpretations of states as tolerant that they can be ascribed beliefs and attitudes in more indirect ways. Rather than saying that the Danish state has religious beliefs, one might then simply say that it has a certain institutional relation to the Lutheran church, which is defined by a certain religious creed, and that this is equivalent to having the negative attitude to other religions required for describing the state as tolerant. What matters in order for the state to be described as tolerant is not the having of beliefs but of any feature equivalent to a negative attitude disposing the state to interfere with certain religious minorities, which can then be overridden by reasons for toleration.

On the strictly dispositional reading, the question then is whether the Danish state’s support for the national church implies the equivalent of a negative attitude towards other religious groups? While the Danish state cannot straightforwardly be ascribed religious beliefs it can reasonably be said to prefer the national church, and perhaps thereby its religious doctrine and creed. The acts and policies of the Danish state are not based on beliefs or assertions about religious truth (compare Modood, 2010, p. 8) but they do explicitly, formally, and substantively show preference for a church defined by Lutheran beliefs. The question is whether one can infer from this fact that the third assumption holds, i.e. that the state is disposed to interfere with other religions? This is not a logical entailment; it simply does not follow from the fact that an agent shows preference for one thing that the agent is disposed to interfere negatively with other things. Unless the things in question are practically incompossible, support for one does not necessarily involve interference with the other. If different religions can co-exist, support for one is not tantamount to interference with another. But if the inference is not one of logic, how might it be understood?

One possibility might be that members of the government or prominent members of parliament might express negative views about minority religions and argue in favour of using the state’s power to interfere with these religious minorities. But as already noted, the beliefs and attitudes of individuals cannot be equated with beliefs and attitudes of the state, and even if the former might causally influence how the state acts or is disposed to act, the resultant acts of the state cannot be understood as expressions of the beliefs of individuals. So this possibility only results in the required kind of disposition to interfere if the state has been captured by political actors with such a disposition. But then it is not the constitutional support for the Lutheran Church that makes the state intolerant but the fact that it has been taken over by intolerant people.

Another possibility is that both the ascription of a negative attitude to, and of a disposition to interfere with, religions deviating from that of the national church are based on actual acts of interference. The reasoning would then be comparative: since the state regulates minority religions in ways that are more restrictive than its regulation of the national church this amounts to a form of negative interference with minority religions, and this can furthermore be understood as an expression of a general negative attitude towards and disposition to interfere with minority religions. There are several problems with this proposal.

First, toleration is about acts of non-interference of a certain kind, not about acts of interference. So one has to be able to justify the ascription of a more general negative attitude and disposition to interfere on the basis of actual acts of interference.

Secondly, it is not clear that the Danish state fulfils the empirical condition of actually being more restrictive towards religious minorities than towards the national church; in fact, freedom of religion is more extensive outside the national church than inside it (inside the national church there are limits to what beliefs and practices are acceptable, which are ultimately (although rarely) enforced by the state), and general rules limiting acts of religious minorities also limit the national church. It is clear that the state actively supports the national church, but the absence of support for minority religions is not in itself an act of interference.

The main difference between the state’s regulation of the national church and religious minorities therefore concerns the noted aspects of state regulation of religion that involve delegation of executive state powers to recognised and approved religious communities. Recognition and approval, as well as the accompanying rights, are premised on conceptions of the role of a religious community heavily influences by and biased towards the Lutheran Christianity of the national church to such an extent that the main criteria for being approved as a religious community concern the similarity in organisation and structure between the religious community and the local parishes of the national church. So the state can be said to impose a partial Lutheran conception of religion on minority religions in these respects. The problem is that this does not (or not primarily) amount to interference with religious minorities in any obvious sense; in fact, these are rather examples of how the state actively recognises and empowers religious minorities in certain ways (albeit in ways premised on a partial, Lutheran inspired, conception of religion).

So despite its religious non-neutrality it turns out to be quite hard to justify the ascription to the Danish state of a general disposition to interfere with minority religions, let alone of such a disposition based on a general negative attitude to minority religions. This undermines the interpretation of the Danish state-religion relation in terms of toleration. But the problem does not stop at the already discussed attitudinal condition for toleration; as already touched upon, the Danish state is in fact involved in positive recognition and empowerment of religious minorities that not only make the description of it as interfering negatively inaccurate, but also suggest that its general relation to religious minorities is not one of negative toleration but some more form of positive engagement. I will therefore turn to the concept of recognition as a third theoretical model for understanding the relation between the Danish state and minority religions.


In debates about liberalism and multiculturalism the term “recognition” has come to denote certain kinds of policies towards minority groups that many advocates of multiculturalism as a normative view think liberalism is either not able to justify or provide the wrong kind of justification of. Charles Taylor’s classic statement of the idea of a politics of recognition (1994) was formulated as a critique of liberal neutrality, and multiculturalists such as Bhikhu Parekh (2006) and Tariq Modood (2007) are critical of attempts, such as Will Kymlicka’s, to ground multicultural policies on liberal premises. The idea is that whereas liberalism is concerned with individual equality and uniform rights, multiculturalism is concerned with recognition of collectives and group differentiated rights. Recognition is furthermore thought to be an alternative to neutrality and toleration as a model for how the state should relate to minorities; whereas neutrality for instance may require the state not to take any stand on the value of different ways of life and to consider everyone only in their capacity as citizens, and toleration involves non-interference with certain disliked or disapproved of differences, recognition is generally thought to involve a positive, acknowledging, affirmative and accommodative relation to ways of life, groups and differences.

I will not attempt to recapitulate or assess to extensive body of theoretical work on these issues. I will rather use the Danish case to illustrate both a positive and a negative thesis about the sketched standard understanding of the relation between liberalism and multiculturalism and between neutrality, toleration and recognition. The positive thesis involves a factual-interpretative claim and an evaluative claim: the former is that recognition is in fact a more precise description of the actual relations between a state like the Danish one and its religious minorities than neutrality and toleration, and the latter is that the relevant kind of recognition is not obviously illiberal. The negative thesis is that recognition is not only a positive relation that is good for the groups receiving recognition, since positive recognition is both compatible with inequalities in recognition and involves indirect forms of misrecognition or recognitive forms of discrimination or even oppression.

The factual part of the first thesis has already been argued in the above discussion of how neutrality and toleration are not precise characterizations of how the Danish state relates to religion in general and minority religions in particular. It only remains to be argued that the state-religion relations are in fact relations of recognition in something like the sense at stake in the debates about multicultural recognition. Recognition can be thought of as an act carried out by one actor (here, the state) which both accommodates or empowers the receiver of recognition (here, religious communities, which are granted rights, privileges and executive powers) and publicly expresses some affirmative attitude or message about the receiver (here, that the religious communities have the status of national church, or recognized or approved religious communities, respectively, with the positive acknowledgement by the state that go with these statuses).

Multicultural recognition is furthermore often thought to concern groups rather than individuals and to ground rights that are either group rights or individual rights that are differentiated on the basis of individuals’ membership of such groups. This might be understood in a communitarian sense, e.g. as elevating groups over their members, but it need not; group differentiated rights may be rights of individuals that are merely differently placed in virtue of their membership of different groups, and even rights held by groups can be understood as collective rights based on the aggregate interests of the individual members of the group rather than as corporate rights based on irreducible features of the group (Jones, 2009). What matters for present purposes is that, whatever their theoretical understanding and justification, policies of recognition are somehow group directed or shaped. This is also true of the Danish case, since the religious communities are ascribed rights at the collective level.

Recognition is finally often thought by multiculturalists and proponents of “politics of difference” to concern aspects of collectives or groups which differentiate them from other collectives or groups (Parekh, 2006). This is arguably also the case here, particularly in the case of the recognition of the Evangelical-Lutheran church as the national church, but to some extent also in the recognition and approval of religious communities, which at least serves to recognize them as religious associations and to distinguish them from other private associations. So the state-religion relation seems to be a relation of recognition in more or less the standard sense usually presupposed in debates about multicultural recognition.

Does this mean that the Danish state is illiberal, at least in this respect, because it engages in recognition of religious communities? If one equates liberalism with the requirement of religious neutrality and strict separation, this is of course the case. But if one relaxes the conception of liberalism to primarily denote protection of equal civil, political and social rights, for present purposes especially including religious freedom and non-discrimination on religious grounds, it becomes quite hard to see that the practice of having an established national church and of recognizing and approving religious communities is illiberal. There are some reasons for concern in addition to the noted fact that even non-members are compelled to contribute to the funding of the national church. The law of ethnic equal treatment, which is the discrimination ban applying to public institutions, does not prohibit discrimination on religious grounds in order not to undermine the possibility of the national church to include religious criteria in employment decisions. But in general the actual effects on effective equal rights of these illiberal practices are negligible; they primarily have symbolic significance. My point is that even though the state’s recognition of religion is a form of multicultural recognition, it apparently does not have any of the problematic consequences usually cited as reasons why liberals should be wary of policies of recognition; the recognition in question does not undermine individual rights, does not in itself significantly detract from or subvert policies of redistribution, and arguably does not affect potential internal oppression in religious minorities. This is not to say that liberals should be happy with Danish practices of religious recognition, but the problems involved fall far short of human rights abuses, civil rights violations or socio-economic inequality, and do not seem to be due to the state’s involvement in policies of recognition as such.

If one grants the positive thesis, is there any problem with the Danish practice of religious recognition? Especially from a (liberal) multiculturalist viewpoint, isn’t it good news that policies of recognition are so firmly established and relatively unproblematic? My negative thesis is that things do not look so bright after all. The thesis can also be divided into several claims. One is that recognition is not just positive and accommodating but also involves the imposition of certain conceptions and frameworks on the receivers of recognition. Another is that recognition might actually also involve misrecognition. And a third is that recognition can be unequal. Rather than arguing for these claims in the abstract I will again try to illustrate them by reference to the Danish case.

The recognition by the Danish state of the Evangelical-Lutheran church as the national church and the approval of Muslim associations and congregations as religious communities might appear to be purely positive things. If the standard of comparison is the general freedom of religion protected by the Danish state, these acts of recognition may seem to be improvements, both in terms of the material opportunities provided, legal rights granted, and symbolic messages sent. But this is not the case. As already noted, the privileged position of the national church comes at the price of a relation of domination by the state over the church and some limitation of religious freedom within the national church. Furthermore, the approval of religious communities is premised on a Lutheran inspired conception of religion and especially on a particular Danish understanding of the role and place of religion; religious communities are approved on the basis of having a dogmatic and organizational infrastructure more or less like protestant congregations, and the rights they are granted are interpreted on this basis. Approval might therefore be thought of as expressing not only affirmation and accommodation but also the imposition of an essentialist view of religion.

Some might think this in itself problematic, but proponents of multicultural recognition have a further reason to be worried, namely that the recognition by the state of religious communities for this reason covers a deeper misrecognition: Even though religious communities are approved for the purpose of conducting marriage ceremonies, getting tax deductions and receiving preachers from abroad, they are precisely not recognized in other ways. If one by “recognition” understands the affirmation of some group’s “identity” as understood by its members, approval by the Danish state is in many cases not an act of recognition: the communities are recognized, but not necessarily as what their members understand them to be.

If one furthermore understands recognition as involving special concern and rights, e.g. rights against religious defamation, the Danish practice of approval may also prove disappointing; the latter was illustrated during the Danish cartoons controversy where Muslim representatives reported the newspaper publishing the Mohammed cartoons for breach of the so-called blasphemy clause of the Danish penal code and were surprised, and offended, when the public prosecutor decided that the cartoons were not in breach of this clause (Laegaard, 2007).

Finally, even though the Danish state clearly recognizes a number of religious communities, there are obvious inequalities in the recognition granted, not only between the national church and other communities, but also between recognized and approved communities and between these and mere religious societies. The question is whether this is problematic or even unjust? This is clearly the case if one accepts a requirement of religious neutrality, but then the problem is recognition as such, not the inequality in recognition. If one rejects the requirement of religious neutrality and endorses some form of moderate secularism, the question is whether one also advocates some demand for equal recognition that would make unequal recognition problematic or unjust? Without going deep into the complicated theoretical debates about recognition, I will merely suggest that this question may actually be quite difficult for proponents of multicultural recognition to answer. The answer depends on what theoretical understanding and justification of policies of recognition one accepts. Some proponents of recognition advocate a “contextualist” approach according to which recognition should not be assessed and justified on the basis of abstract theoretical principles but should rather be negotiated in particular contexts of recognition (Parekh, 2006; Modood, 2007). On such views, which are also explicitly coupled to an endorsement of moderate secularism, one cannot condemn inequalities of recognition as such on the basis of some theoretical principle of equality. But if recognition rather has to be negotiated, there is no guarantee or even reason to expect that the result will be some form of equality, since the inequalities characterizing the context are also bound to influence the outcome of negotiations (Laegaard, 2008).

So the picture is quite muddled from the perspective of recognition. On the one hand, recognition in fact seems to be the best theoretical model for capturing the state-religion relations of moderately secular states like Denmark and the practices of recognition of the Danish state are not particularly illiberal. On the other hand, recognition is not all positive, but involves domination and imposition, as well as misrecognition and unequal recognition, which at least “contextualist” theories of recognition may not immediately have the resources to handle in plausible ways. A liberal theory based on some principle of equality might, on the other hand, point to ways of resolving some of the problems with recognition. I have suggested that it is not the acts of recognition as such that are problematic from a liberal point of view, but the domination and inequality inherent in them. The most problematic aspect of Danish state-religion relations is probably the coercive taxation imposed on all residents to fund part of the National Church. This might be disbanded without challenging the establishment as such. The misrecognition potentially involved in the Lutheran inspired conception of religion on which recognition is based might also be softened; the Advisory Committee has already argued for a more “objective” conception of religious communities, and the potential mismatch between the recognition given and the self-conception of members of religious communities might be avoided by publicly explaining the type of recognition given. The final inequality between the symbolic status of the National Church and all other religious communities is inevitable given establishment, but are not unjust as such if on does not equate liberalism with strict neutrality.


In this paper I have sketched different theoretical models for understanding how liberal states relate to religion in general and religious minorities in particular. I have done this in relation to a particular case of an arguably quite liberal but not strictly religiously neutral state, namely Denmark. My question has been how we should understand what it means to be a liberal state in these respects if liberalism is not equated with strict religious neutrality and separation. I have argued that neutrality is not an all or nothing affair, which allows for the characterization of the Danish state as liberal. The fact of non-neutrality furthermore does not imply that the state is tolerant, since the concept of toleration is both problematic to apply to states for theoretical reasons and because its application in the particular case presupposes a number of empirical conditions and assumptions that may not hold. I then argued that the concept of recognition provides a better theoretical model for understanding state-religion relations in Denmark. I have finally suggested that this result poses a number of challenges for the traditional understanding of recognition in debates between liberalism and multiculturalism; recognition is not necessarily illiberal, but it is also not obviously as good a thing as many proponents of multicultural recognition tend to assume, and standard theories of recognition are arguably not immediately equipped to evaluate these problematic aspects and provide practical guidance in relation to them.

The reciprocal illumination of the theoretical models and the Danish case suggests some tentative lessons, which might be the object of further systematic discussion elsewhere. One such lesson is that liberalism might be understood as a more plural normative standpoint, not necessarily in the sense of fundamental value pluralism, but in the sense that there are several things liberals are concerned with, even with regard to a relatively narrowly circumscribed issue such as religion. Liberalism is certainly not just about state neutrality. Whether or not the state should be religiously neutral, it seems more important whether and how it secures freedom of religion and non-discrimination on religious grounds. Returning to the parallel between liberal theory and the First Amendment, one might argue that “free exercise” is more important from a liberal point of view than “non-establishment”. Neutrality may be important, but some forms of religious non-neutrality may be unobjectionable and non-neutrality is generally of greater concern in relation to coercive aspects of state activity than merely symbolic acts of state. Symbolic acts are also important, but even here it is not an all or nothing affair. I have suggested that the concept of recognition is well suited to capture some of the symbolic aspects of state acts, but also that such forms of recognition are often theoretically complex mixed blessings. All of this needs further discussion and systematic defense, which I cannot even begin to provide here. For now I merely hope that the present discussion has illustrated the relevance of these types of considerations.