Othering, Alienation and Establishment

This article examines the relationship between religion and the state, focusing on cases of establishment in which one religion is formally recognized. Arguing that religious establishment is wrong if it causes some citizens to feel alienated, we reject the criticism that feelings of alienation are too subjective a foundation for a robust normative case about establishment. We base our argument on an account of collective identities, which may have an ‘inside’ but are also subject to a process of othering in which a dominant group imposes an identity on a subordinate group. The establishment of a religion may contribute to othering, and the othered group may consequently be alienated from the state. However, since establishment does not always cause alienation, it is necessary to seek evidence and engage in a dialogue in order to understand a group’s own account of its experience of its situation.


Introduction
Contrary to the predictions of the classical sociologists, such as Marx and Weber, modern societies have not become fully secularized over time; religion has not gradually faded away as these societies have become more rationalized and urbanized. Indeed, it is often said that today we are living in a post-secular age, one marked by the 'continued existence of religious communities', albeit 'in an increasingly secularized environment' (Habermas, 2008: 19). On behalf of those sociologists, it is fair to say that they did not anticipate the non-European immigration into European societies, which has given religion a new vitality and prominence. The migrants of postwar Europe, including the second and third generations, have been and are more religious than the natives they have joined (Gale and Hopkins, 2009;Modood et al., 1997;Woodhead, 2016). This has also become true of the Others argue, in sharp contrast, that formal connections may be forged and maintained between at least one religious community and state institutions; some say that such connections are appropriate when the religion thus established is closely bound up with the history and identity of that state. As David Miller (2016: 454) puts it, the historic nation may 'legitimately give precedence to the artefacts of a particular religion when decisions about the use of public space are taken'.
In this article, we focus on this second option, usually referred to as religious establishment. According to Sune Laegaard (2013: 125), this is 'an institutional relationship between religious organisations such as churches and the state'. Establishment, thus defined, can take many different forms and degrees depending on the number and strength of the linkages between these organizations. To simplify, we can say that establishment comes with a range of benefits for members of the established religion which members of other religions -and non-religious members of the political community -either do not enjoy to the same degree or do not enjoy at all. Here, it is possible to distinguish between symbolic and material advantages. The favoured religion may play a prominent role in the symbolic life of a state that other religious groups lack; for example, the head of the church may crown monarchs and its symbols may be included on the national flag. It may also enjoy a wide range of material advantages that other groups are denied; for example, the church may be able to raise taxes to pay its ministers' salaries, and the state may provide funds for the maintenance of that church's places of worship, schools or hospitals.
Our aim is to reconstruct and defend one particular argument about religious establishment which contends that this relationship between religion and the state is wrong if it causes some citizens -usually, members of religious minorities -to feel alienated. This argument has been deployed by Rajeev Bhargava, for example, to argue that Muslims in Britain (in contrast to India) are alienated from their national citizenship by the fact of Anglican establishment (Bhargava, 2013). As a matter of fact, Bhargava provides no evidence for his claim, and we believe it to be false (Modood, 2019: 206-207). In fact, since 2008, we would argue that Muslims in India are much more likely to be alienated from their political community than their British counterparts, and it is surprising that Bhargava does not discuss that in his commendation of Indian secularism (Farooqui, 2020;Khan, 2019;Shaban, 2018). Yet the normative character of Bhargava's argument is sound: the existence of alienation does provide a reason to change existing relations between religion and the state. In recent discussions, however, the alienation argument gets mentioned only to be fairly summarily dismissed. A number of critics contend that this argument is subjective since it is based on an account of what some people happen to feel and think about a particular situation. As such, the critics say, an account of alienation is much too fragile a foundation on which to build a robust normative case for or against establishment. These critics then present other purportedly objective arguments independent of what people happen to think in support of their views about state-religion relations. Cécile Laborde (2017: 135-136) defends what is probably the best-known of these arguments, according to which establishment is objectively wrong if it sends a message to some citizens that they are of lower civic standing than their fellows. 1 Against Laborde and other critics, our aim in this article is to defend a version of the alienation argument against religious establishment which is founded on an account of collective identities, and which is aligned with normative concepts of recognition and misrecognition. On this view, identities have a special meaning to their bearers and exist in a context of unequal power relations. It is possible, therefore, that the subordinate group's identity can be distorted by a dominant group failing to understand that minority identity on its own terms, or worse still, by a process of othering in which a dominant group imposes an identity on a subordinate group (Modood, 2013(Modood, [2007: ch. 3). We argue that such a process can create a division between groups defined partly by reference to certain religious markers, but specifically so when race and ethnicity are other factors. In such a racialized or ethnoreligious context, a state's decision to establish or to continue to support a particular religion may be a factor which contributes to this othering. As a result of such a decision, while one group is positively recognized by the state and is therefore able to identify with it, the other group is not given the same sort of recognition by that state and may therefore be alienated from it. While a number of factors should be taken into account when determining what the relationship between religion and the state should be -including the patterns of religiosity in a particular political community, the character of the majority community's religion, the history of that religion's relationship to the state, and the value that a state accords to religious identities -our aim here is to show that alienation cannot be overlooked since its presence or absence is relevant to arguments for or against establishment and its alternatives.
In the next section, we critically analyse and deconstruct the dichotomy between subjectivity and objectivity used by critics to undermine the alienation argument about establishment. In the section that follows, we present our general account of identity and processes of othering, including their racializing form. In the section after that, we show that some forms of othering are processes of alienation, and in particular that religious establishment can cause the alienation of religious minorities. The next section focuses on the role of multicultural dialogue in our account. Only in such a dialogue is it possible to determine whether the groups who we think are alienated really are. In the conclusion, we restate our principal claim that state-religion arrangements should not alienate religious minorities so that, if evidence of such alienation is presented, then appropriate institutional reforms should be undertaken at least to reduce and ideally to eliminate it. While we do not develop the point here, we do want to note that our argument applies to statereligion relations in general, including laïcité and other forms of separation. Or, to put it another way, if establishment can alienate, so can the marginalisation of religions.

Deconstructing the Dichotomy
In contemporary academic debates about state-religion relations, an often-heard criticism of the alienation argument is that feelings of alienation are too subjective a basis for the evaluation of religious establishment. It is argued that an analysis of the communicative meaning of establishment, and in particular the implications of such communication for certain minorities, provides a more objective basis for such an evaluation. These two approaches are presented as if they are exclusive and exhaustive: establishment must be evaluated by reference either to citizens' subjective perceptions or to the objective meaning of the symbols of establishment, and, since the first approach fails, the second must be adopted. In this section, we deconstruct this position, showing both that it presents a highly -and unnecessarily -reductive account of alienation and that its proponents have not successfully shown that their alternative is as objective as they claim. In order to keep this task manageable, we focus in particular on the versions of this argument found in recent work by Laborde and Laegaard (Laborde, 2017;Laborde and Laegaard, 2020;Laegaard, 2017Laegaard, , 2020. Beginning with alienation, Laborde's and Laegaard's account has two features in particular which are worthy of attention. First, they imply that feelings of alienation are without cognitive or normative content; they are simply the affective reactions that people have to events they experience or to objects in their environment. For example, they suggest that an account of feelings of alienation is purely 'descriptive', in no way providing a 'normative' evaluation of the object of such feelings (Laborde and Laegaard, 2020: 179). Yet it is arguable that to make sense of their comments that affective reactions may be 'crucial epistemic indicators that something has gone amiss' (Laborde and Laegaard, 2020: 179), it is necessary to assume that such reactions have an interpretive and normative component. Anger, for example, cannot be understood merely as an explosive negative reaction to an external situation but must be regarded in part as a reaction to its perceived unfairness or injustice (Holmes, 2004). Second, Laborde and Laegaard (2020: 177, 179, 180) downplay -and arguably trivialize -the significance of feelings of alienation by associating them with notions such as displeasure, unhappiness, offence and resentment. 2 As an alternative to the alienation argument, Laborde and Laegaard (2020: 172, 179, 180) recommend focusing on the 'communicative meaning' of the symbols of religious establishment. They argue that such symbols can signal approval for a certain state of affairs, including, for example, 'a general acceptance of religion' (Laborde and Laegaard, 2020: 177). They can also express certain values, such as 'the moderate, liberal values of western democratic states' (Laborde and Laegaard, 2020: 178). Furthermore, a symbol of establishment such as the crucifix can function as part of the putative identity of social groups, implying that Christians are full citizens of a particular political community, while 'non-Christians do not belong' (Laborde and Laegaard, 2020: 186). In addition to focusing on the meaning of symbols, Laborde and Laegaard (2020: 182, 185-186) make the further and stronger claim that establishment and its symbols can have a specific 'function' or produce specific 'effects'. In particular, it can have deleterious effects on what they refer to as 'socially vulnerable groups' (Laborde and Laegaard, 2020: 183). The invocation of such symbols can actually render members of such groups second-class citizens or exclude them from the political community. Hence, Laborde's and Laegaard's (2020: 180) conclusion is that symbolic religious establishment 'is wrong because of its communicative meaning . . . when it aggravates the social vulnerability of minorities'.
Emphasizing the objective standing of their analysis, Laborde and Laegaard (2020: 185-186; and see Laborde, 2017: 135) emphasize that '[e]ven if minorities do not feel alienated . . . symbolic religious establishment can still function as a way of excluding minorities from equal standing as members of the national community'. In other words, the normative status and practical effects of establishment are entirely independent of how establishment is perceived and any thoughts people may have about it.
In the construction of their dichotomy between subjectivity and objectivity, Laborde and Laegaard never explicitly define these terms, and we do not think that their meanings are in any way obvious. To begin with subjectivity, they likely mean that a feeling such as alienation is subjective in the sense that its qualities (its object, intensity, duration, etc.) are dependent on the specific characteristics of the person experiencing it. The fact that they refer to Joel Feinberg's claim that some people may have 'abnormal susceptibility' to offence (Laborde and Laegaard, 2020: 179) supports this interpretation, as does Laegaard's suggestion that people have 'idiosyncratic sensibilities' (Laegaard, 2017: 122, 123, 125). Since the burden of our argument is to show that alienation can reasonably be understood as more than a subjective feeling, and that as a consequence it can play a legitimate role in the normative evaluation of establishment, we shall say no more about this part of Laegaard's and Laborde's analysis at this point.
Turning to the idea of objectivity, Laborde and Laegaard do not provide a clear definition of this term either, and again we do not think that its meaning is at all apparent. First, they could be suggesting that interpretation can be objective. Some of what they say seems to support this reading. For instance, as we have seen, they note that 'many historically Christian signs are now interpreted as the benign symbols of the moderate, liberal values of western democratic states' (Laborde and Laegaard, 2020: 178). But they then claim that since 'liberal values are expressed through religious symbols, the implicit message is that one religion -Christianity -has a privileged relationship -a monopoly, even -over these values' (Laborde and Laegaard, 2020: 181). Here, Laborde and Laegaard claim with confidence that linking Christian symbols to liberal values gives that religion a status which other religions lack, and they reject the rival readings which would deny that unequal status.
Second, Laborde's and Laegaard's suggestion might be that certain sorts of empirical claims can be objective. Thus, they imply that it is possible to impartially determine which groups may be adversely affected by establishment: 'We need to work with a precise definition of which groups are relevant minorities, as measured by objective criteria of collective vulnerability in particular societies' (Laborde and Laegaard, 2020: 180; second emphasis added). They also suggest that certain actions or states of affairs have objectively ascertainable effects: symbolic religious establishment 'might be legitimate if it enjoys broad support among the population, as it is not objectively divisive and exclusive of atheists' (Laborde and Laegaard, 2020: 180; emphasis added).
Third, Laborde and Laegaard could be using objectivity in a normative sense, according to which it is possible to know with certainty whether particular actions, relationships or states of affairs are right or wrong. For example, Laborde (2017: 135;emphasis added) contends that 'what matters is not what government intends to communicate, nor how citizens subjectively perceive it but, rather, whether governmental messages express objectively appropriate attitudes toward people'. In this third sense, then, it is possible to determine whether the attitudes which the state takes to its citizens are right or wrong. 3 In fact, we would suggest that Laborde's and Laegaard's argument relies on all three of these senses of objectivity in conjunction. Consider the following: Symbolic religious establishment is wrong when it communicates that religious identity is a component of civic identity -of what it means to be a citizen of that state -and thereby denies civic status to those who do not endorse that identity, who are then treated as second-class citizens (Laborde, 2017: 135; see Laborde and Laegaard, 2020: 180).
Here, Laborde assumes that it is possible to know objectively (1) what a religious symbol communicates (some people are regarded as members and others are not), (2) what effects that communication has (the latter are treated as second-class citizens), and (3) why those effects are wrong (the argument is premised on a normative principle of civic equality). 4 In response, we want to argue that, while Laborde and Laegaard invoke claims to interpretive, empirical and normative objectivity, they do not offer an account of what objectivity means -'objectively divisive', 'objectively appropriate', and so on -and why they think that their claims possess it while rival claims do not. To make this point, consider an argument opposed to that of Laborde and Laegaard. Let us say that the situation in question is one in which the state orders crucifixes to be displayed in some public buildings. Let us also say that Laborde and Laegaard claim the following: (1) this symbol signals that only Christians are full members of the polity, (2) the effect of its display is to marginalize non-Christians, and (3) it thus violates the principle of equal citizenship. The opposing argument is the following: (1) the public display of the crucifix expresses the liberal and egalitarian values of the political community, (2) thus this symbol expresses a message of civic equality, and (3) it is therefore compatible with the principle of equal citizenship.
The question is: how do we decide between these two arguments and on what grounds? Certainly, it is possible to favour one of these arguments over the other by claiming that there are better reasons on one side than the other. 5 For instance, proponents of the argument opposing that of Laborde and Laegaard may point to the dominant understanding of the religion in question, according to which all individuals are of equal moral worth. This would be a reason in support of the interpretation of crucifix display as the expression of egalitarian values. However, the claim that one argument is objectively true, and the other not, would only make sense if a convincing account of objectivity was presented and applied to the claims in question. Our point is that Laborde and Laegaard fail to do so: to repeat, they do not explain why their position enjoys interpretive, empirical and normative objectivity. In this case, we are forced to conclude that saying that their hermeneutic interpretations, empirical claims and normative judgements are objective does no more than add a certain rhetorical flourish to Laborde's and Laegaard's argument.

Othering and Racialization
Given our critique of Laborde's and Laegaard's reasons for rejecting the alienation argument, and in particular our deconstruction of the dichotomy on which these reasons depend, our task is now to explain why we think that an account of alienation can and should play an important role in the evaluation of different forms of state-religion relations. As we shall explain in the next section, we regard the sort of alienation relevant here as a condition in which a number of citizens, given their particular identity, are unable to identify with their national citizenship. Since this places an idea of social or group identity at the heart of our argument, we begin with an account of such identities. In this section, it should be emphasized, we focus on the processes of interaction between groups in which their identities are formed; in the next section, we emphasize the role that the state can play as the facilitator of -or collaborator in -such processes.
On our account of identity formation, there are no social identities independent of intersubjective agency and the interaction of individuals and groups. Nor do any such identities have fixed boundaries or stable contents since groups are fluid, and the interactive dynamics between groups never cease. For instance, Tariq Modood describes the changes that one British community has made to its self-ascribed identity: Since the 1980s-1990s onwards there seems to have been a constant rise in religious community identification among South Asians to supplement or demote race/ethnic identifications, and in a very few years they moved from 'black' to 'Asian' to Indian/Pakistani/Bangladeshi to Sikh/ Muslim/Hindu (Modood, 2019: 102).
This social constructivist position puts us in opposition to any form of essentialism or quasi-naturalism about social identity. These 'ism's, and others like them, try to fix identities which cannot in practice be fixed, and they often try to do so for their own purposes rather than for the good of those whose identities they seek to fix. In a moment, we shall elaborate on this point by discussing the idea of racialization.
To explain our position in more detail, we hold that group identity is produced by how groups understand and value themselves and by the dynamic interaction of two (or more) such groups. In this process, each group ascribes certain features to others, and those others ascribe certain features to it. Each group has a conception of itself, produced partly in response to others' ascriptions. In other words, identity is shaped and maintained from the 'outside in' and 'inside out' at the same time. We would emphasize that this is the normal way in which social identities are formed and constantly reformed. Since there is no other way in which such identities can be constructed, such a process is not, in itself, a malign operation of power which should be in some way resisted.
This normal pattern of identity formation can, however, take the pathological form of othering. According to Modood: 'othering' sees a minority in terms of how a dominant group negatively and stereotypically imagines that minority as something 'other', as inferior or threatening, and to be excluded. Indeed, the dominant group typically projects its own fears and anxieties onto the minority (Modood, 2019: 78).
Assuming a simple scenario in which only two groups interact with one another, othering can happen when one group is relatively dominant and the other relatively subordinate. In these circumstances, while the dominant group has the power to impose its account of the other onto that group, the subordinate group lacks the equivalent power to resist that ascription and to describe itself on its own terms -in an 'inside out' way. Many factors can enable one group to become dominant, including its level of financial resources, dominance of social elites, control of mass media, ability to influence the state's policies and to shape its institutions, and so on. 6 The group lacking those resources and powers is then rendered subordinate.
In processes of othering, the dominant group can attribute strongly negative characteristics to all individual members of the subordinate group. Those individual members will find it difficult to reject such attributions and to replace them with their own descriptions of themselves. Such a process of othering takes a racialized form if it latches onto what are regarded as distinctive physical features of a group or beliefs about ancestry in order to impute negative characteristics to members of this group. Beyond this process of racialization can be a further racialization of the cultural community aspects of a racially othered group. The assertion that 'All Muslims are fanatics' would be a case in point (Modood, 2019: 80). As a result of othering, the targeted group is more likely to be considered an 'other', placed below or beyond what is normatively acceptable, stigmatized, harassed, blamed and even hated.
Having said this, it is never the case that the subordinate group is entirely powerless in the face of othering; it is never merely the passive object of such a process. This group is always able to respond to some degree to the othering to which it is subjected. We describe this as a process of 'de-othering'. In some cases, the subordinate group may be able to reject the dominant group's ascription, denying it has the characteristic attributed to it. For instance, Muslims may be able to effectively counter the accusation that they are religious fanatics. In other cases, the subordinate group may accept but revalue certain attributions, acknowledging that it does have a particular characteristic -but claiming this to be positive rather than negative. 7 To continue with our example, British Muslims may agree that they are socially conservative, but argue that this is a positive virtue. If, furthermore, the dominant group comes to recognize the subordinate group's description of itself, then this is another way in which the latter group can be de-othered. However, despite this possibility of resistance, it is important to emphasize its limits: the greater the degree of dominance of one group over another, the stronger its ability to impose its ascriptions onto that other, and the weaker the other's ability to resist.
With this social constructivist account of the production of collective identities in mind, it should be clear that we think such identities are neither wholly subjective or objective in the senses in which the critics of the alienation argument use those terms. On the one hand, collective identities are not merely subjective; having such an identity is not reducible to a matter of sharing a certain emotional state or psychological condition. A social group in our sense still exists even if its members do not all feel the same way about a particular situation. Idiosyncratic variations in individual sensibilities do not imply that a collective identity is weak or absent. On the other hand, a collective identity is not purely objective; an account of such an identity cannot do without reference to group subjectivity, to the self-understanding of that group. Such an experiential dimension of identity is not reducible to a matter of how individual members of the group happen to feel; rather, it consists of ways of thinking and acting that constitute that group in its social location, including therefore its vulnerabilities.

Alienation and Establishment
So far, by focusing on the interaction of groups, we have explained how social identities are normally formed and how this can take the pathological form of othering, including in some cases racialization. Such processes can take on many different forms, drawing on many different types of power and having many different effects on those involved in them. On this last point, the othered group could have its opportunities curtailed, its rights restricted, its political voice diminished or its membership threatened. In this section, we focus on one particular form of othering, which results in a state of alienation for the targeted group. We contend, more specifically, that the state's decisions regarding religion may contribute significantly to a process of othering which alienates religious minorities from their polity.
In making this argument, one of our concerns is to demonstrate that it is worth using the specific idea of alienation when seeking to describe and evaluate some forms of othering. In other words, we want to show that alienation is a distinctive idea with an indispensable role to play in our argument, and that it is not merely a fancy synonym for other notions such as injustice, unfairness, inequality or misrecognition (although it is not unrelated to some of these notions). The idea of alienation has a long and complex history and is most closely associated with the Hegelian, Marxian, and, to a lesser extent, existential traditions in philosophy. 8 We cannot provide a detailed account of this diverse history here. What we can do instead is to draw on the literature in order to identify a generic concept of alienation as the disruption, damaging or destruction of a relationship between two or more parties that could be, should be, or was previously in a state of harmony. 9 The concept of alienation, thus defined, could be applied to a wide range of experiences, relationships and situations. For example, it could describe religion's role in alienating humanity from its own essence (Feuerbach, 2008(Feuerbach, [1841), the proletariat's experience of the capitalist labour process, or the relationship of every individual to every other. For our current purposes, we need a more specific conception of this concept, which identifies the relevant parties, the potentially harmonious relationship between them, and the presently disrupted state of that relationship. 10 According to our more specific conception, the relevant parties are citizens and their polity, and those citizens are alienated if they are unable to identify with that polity. To put these elements together: citizens who are not able to identify with their polity are alienated from it, and in effect they are alienated from their national citizenship too.
In order to justify this particular conception of alienation, we need to speak more about the normative standard at its heart -namely, the assertion that citizens should not be alienated from their polity. Laborde and Laegaard propose a different standard, according to which all members of a political community should enjoy 'equal civic status' (Laborde, 2017: 135; see Laborde and Laegaard, 2020: 175;Laegaard, 2017: 124-125). By this standard, as we have seen, religious establishment is wrong if it sends the message that some groups of citizens are more worthy of the state's concern and respect than others. For us, the relevant normative standard is significantly different. Our claim is that citizens, regardless of their religious affiliations, should not be alienated from their state. Laborde herself seems to have endorsed an argument of this kind prior to the publication of Liberalism's Religion in 2017. In an article published four years earlier, she describes 'a conception of citizenship which postulates that all citizens should be able not to feel alienated by their political institutions in light of their deepest beliefs, and that institutions should consequently be framed with that aim in mind' (Laborde, 2013: 84). Commenting on Laborde's position, Modood (2019: 206) says that he holds 'a stronger version of this duty of symbolic recognition: not only must the state not alienate', it 'must make positive efforts to ensure that all citizens are able to feel a sense of belonging'. 11 We are sympathetic to that view, but here insist only that the state has a duty to not alienate citizens by misrecognizing or othering them.
Having shown why we think that alienation -understood as citizens' lack of ability to identify with the polity -is wrong, we shall now explain how the state's approach to religion may result in a form of alienation. One way is when the state gives greater value to non-religious (e.g. gender, sexual, linguistic) than to religious identities, alienating citizens for whom their religious identity is of prime importance. The case we are particularly interested here is how the state's establishment of religion may contribute to a process of othering which causes such alienation. Othering, as we have explained, is a process by which a dominant group imposes an identity onto a subordinate group. Focusing on the cases of concern to us here, two possibilities may be distinguished. First, where an existing form of establishment continues to be supported by the state, the disparity between two religious groups can be perpetuated, enabling the group which is already dominant to continue to impose an identity of not belonging to the state on the subordinate group. Second, if the state decides to establish a particular religion when none previously were, then this could enable one religious group to become dominant (or more dominant than it already was) and could make another an outsider or subordinate group (or more subordinate than it already was). This second possibility harks back to our account of collective identity by showing that the groups in question are not already-existing entities which may be affected in different ways by the state's decisions about what sort of relationship to have to religion. Rather, these groups are actually formed in part by the decisions that the state makes about how religion should be treated (Asad, 2003).
In order to explain further how establishment might contribute to a process of othering, we shall focus on a particular case. In England, the Church of England is the established faith, 12 and this status gives it a number of symbolic and material advantages which are not enjoyed -or not enjoyed to a proportionate degree -by other faiths. 13 Here are two examples. The first is the funding of faith schools. According to The 36th Report of British Social Attitudes, in 2018, 12% of the British population were Church of England, 7% Roman Catholic and 6% Muslim (Voas and Bruce, 2019). While it is difficult to find data on faith schools for all of Britain, statistics are available specifically for England. Combining primary and secondary figures, 22.6% of schools are Church of England, educating 13.3% of pupils; 9.7% are Roman Catholic, educating 9.3% of pupils; and 0.2% schools are Muslim, educating 0.2% of pupils (Long and Danechi, 2019). If these two data sets were strictly comparable, we could say that Anglicans and Catholics enjoy a slightly more than proportionate slice of faith school funding, while Muslims receive a significantly less than proportionate slice, as if their faith was less worthy of state support. 14 The second example concerns political representation. Twenty-six Anglican bishops sit in the House of Lords as the 'Lords Spiritual'. The only other religious leader is the Chief Rabbi, who is a 'Lord Temporal'. There is no place for representatives of any other religious communities, including, for example, Roman Catholics. In this way, the Church of England enjoys a degree of formal political representation which no other faith community does, and hence its ability to make its voice heard in public debate is much greater than that of other faith communities.
What effect does establishment have on the relationship between different religious communities and the state? The dominant religious group, which becomes or remains the established faith, finds that its group identity is closely aligned with the religious identity of the state. In the case with which we are concerned, the state singles out the Anglican church for special recognition: it declares that it is in virtue of Anglicans' specific religious identity that their privileged status is deserved. It is very likely that, in these circumstances, this dominant religious group will be able to identify with or feel strongly connected to the state, which it is able to see as a place in which its own identity is reflected. 15 At the same time, the subordinate religious group, which is denied the benefits of establishment, finds that its religious identity is not aligned with that of the state. As we have seen, British Muslims do not receive the symbolic and material benefits which Anglicans enjoy to the extent that their growing numbers warrant. In these circumstances, members of this subordinate group may be alienated from the state: it may not feel like their state, not a place in which they can feel at home. To emphasize the contingent character of our argument, we are not claiming that religious establishment is always wrong because it is always alienating. Rather, our claim is that if establishment causes some religious minorities to be alienated, then this provides one reason against such establishment. If, by contrast, religious minorities do not feel alienated by establishment, then their feelings do not count against it, and the suggestion that establishment devalues their citizenship is misplaced.
We end this section -as we ended the previous section -by emphasizing that our account is neither subjective nor objective in the way that Laborde and Laegaard seem to define those terms. On our account, alienation is a collective experience which is partly shaped at a number of sites, and by a number of relationships, of power. Within this nexus of power, people are classified into groups of various kinds, and individuals are placed into these groups. These processes help to determine each individual's place in their social world, thus shaping their experience of it. Focusing on religious establishment, some versions of this relationship between religion and state may prevent groups of citizens from being able to identify with the state. As we have shown, however, this lack of ability is neither subjective nor objective: it is not reducible to individuals' affective states, nor is it a condition that could be described without any reference to group sensibilities. As we shall now show, it follows that it cannot be claimed a priori that establishment is alienating to a religious minority like British Muslims, without giving that group a voice.

Multicultural Dialogue
In the second section of this article, while criticizing Laborde's and Laegaard's claim that their arguments are objective, we did not deny that some arguments are better than others because they are supported by better reasons. Now we want to show that engagement in a multicultural dialogue provides an indispensable reason for determining whether a particular group is alienated or not. Some reasons for arguing that a particular case of majority establishment causes minority alienation may be empirical in character, derived from descriptions of processes of othering, the effects of these processes on particular groups and so on. However, although reasons of this kind can support the hypothesis that a group is alienated, they cannot by themselves prove that hypothesis. For us, one reason which must be provided as part of the claim that a group is alienated is derived from an account of that group's collective subjectivity, where such an account is built up from targeted research, organizational statements and debates, publications by faith leaders and public intellectuals, discussions in inter-faith fora, political lobbying, protest and mobilization, and so on. To emphasize: no claim that a group is alienated and has had its civic status damaged is complete without that group's own account of its situation, and that account must result from a dialogue with that group, and not simply by a purportedly 'objective' analysis of the message that establishment is sending.
To see why we insist on this point, let us first describe the dangers that may be attendant on claiming that a group has a particular view of the situation in which it finds itself without asking members of that group itself. Here, we shall draw on an example which concerns disability rather than religiosity. According to its website, the Oregon Health Authority (n.d.) designed the Oregon Health Plan with the aim of providing 'health care coverage for low-income Oregonians from all walks of life'. However, when this plan was being drawn up in 1990, it attracted considerable controversy. According to Anita Silvers, the public policy makers charged with drafting the plan judged 'that having an irreparable disability should be sufficient ground for being denied health care, and even life itself'. This judgement 'was arrived at through a telephone survey of able-bodied individuals who surmised, for example, that they would rather be dead than confined to a wheelchair'. In seeking to understand why they made this judgement, Silvers (1995: 35-36) argued that these individuals' 'suicidal fantasies' were imposed 'on disabled individuals, thereby diminishing their access to health care and denying them recognition as self-determining moral beings'.
The process of public consultation about the Oregon Health Plan provides a vivid example of the dangers of the external evaluation of a group's experience. In the terms we have been using in this article, the process of consultation used in the Oregon case permitted and legitimized a form of othering in which able-bodied individuals projected what they thought it was like to be disabled onto disabled individuals, an othering which the latter group had no opportunity to contest or reject. This is, to be sure, an extreme example of the sort of danger about which we are concerned. Our claim, nevertheless, is that Laborde's (2017: 85, 135) invocation of Justice O'Connor's 'reasonable observer' test runs just the same sort of risk, even if not to the same degree. 16 In the present context, the lesson to be drawn is that any argument which claims that religious establishment causes alienation must ask those who are said to be alienated whether they experience the situation in this way or not. In other words, as part of the process of seeking to determine whether a particular group is alienated, there must be a moment of dialogue between (at least) the dominant group and the subordinate group. We are not claiming that a group is a monolith or speaks with one voice; it is better understood as 'a democratic constellation', consisting of several asymmetrical yet patterned points rather than a single mass (Dobbernack, 2019;Modood, 2013Modood, [2007). The aim of such a dialogue, then, is to give all groups and their various factions -in particular, those which are subordinate -the opportunity to generate and share accounts of collective experience so that some degree of mutual understanding can be achieved.
Consider, for example, the first publication of Salman Rushdie's novel The Satanic Verses in September 1988. Soon afterwards, British Muslims began to express anger and feelings of outrage about the book, by, for example, burning copies of it in a mass protest in front of Bradford City Hall. At that time, most of British society could not understand why Muslims were so angry. However, as a result of public dialogue, rightly including religious and non-religious citizens, many came to understand that anger and how it was the manifestation of a social vulnerability. As a result, unlike their continental European counterparts, no British media outlet reproduced the Danish Cartoons of the Prophet Muhammed, and in 2006 the Parliament passed the Racial and Religious Hatred Act which proscribed incitement to religious hatred (Modood, 2019). To generalize, without the emergence of a shared understanding through public dialogue, members of mainstream society -legislators, judges, lawyers, juries, scholars, members of the public and so on -are likely to misunderstand what is, and what is not, alienating for members of a subordinate group and to see them as an incomprehensible 'other'.
Another example of how a dominant group can be oppressive simply by assuming that a minority thinks in the same way as it does is the case of liberals' and left feminists' views of Muslim women's modest dress. Well-intentioned feminists, especially in France, until recently used to -indeed some still do -denounce the wearing of a headscarf (misleadingly called the veil in France), and especially the full face-covering (niqab or burqa), worn by some Muslim women as patriarchal and oppressive, something imposed upon them by Muslim men (Göle and Billaud, 2012;Scott, 2009). The problem is that some people hold this view without feeling any need to hear what Muslim women themselves think, and, when asked, they usually say they are following the will of God (rather than the will of a male relative or imam). A notorious instance of such an approach informing public policy is France's Stasi Commission's conclusion that the headscarf should be banned in French state schools -a conclusion that it reached without looking at any research on the wearing of modest dress or taking evidence from any wearers (Modood and Kastoryano, 2006). 17 This example is also particularly apt in suggesting that it is not only establishment that can alienate minorities; majoritarian secularism can also, mutatis mutandis, to be considered in the same way.
Although there is a lot more to say about the nature of multicultural dialogue, we only have enough space to indicate some of its key features here. First, in multicultural dialogue, groups make claims, assert their identities, explain what things are like from their particular social location and hopefully come to understand each other. Second, such dialogue must be reasonable in the sense that there must be limits to what can be said. There is, for example, an important difference between expressions of Islamophobia and reasonable criticisms of Muslims (Modood, 2019: 85-86;Modood, 2020). Third, a reasonable dialogue is one conducted in accordance with the operative public values prevailing in a particular place at a particular time. These are the 'constitutional, legal and civic values' which 'represent society's public culture' and 'give shape and substance to its inevitably vague conception of the good life' (Parekh, 2005(Parekh, [2000: 269). Fourth, principles of social justice themselves are arrived at through dialogue and negotiation in which all groups have a right to speak for themselves and to assert their vision of the whole. Fifth, multiculturalists recognize that the existing public/political space in which dialogue occurs is strongly skewed in favour of already-dominant groups, hence their commitment to ensuring that such groups can be included, and their voices heard. Sixth, more broadly, citizenship itself may be regarded as a continuous multilogue of ongoing modulations and contestations. The ethics of such citizenship is based on the commitment to include all groups -especially the marginalized -in this multilogue (Modood, 2013(Modood, [2007: ch. 6). 18

Conclusion
In this article, we have claimed that othering is a process in which a dominant group imposes a negatively valued identity onto a subordinate group, where the latter finds it difficult to resist that imposition. We have then argued that, as a result of some processes of othering, only the former group can feel at home in their social world, while the latter cannot. More specifically, we have claimed that, in some cases of religious establishment, the religious majority is able to identify with the state, while the minority is alienated from it. Finally, we have suggested that multicultural dialogue is a vital part of an investigation into alienation. Without seeking to understand how the minority in question processes its experience, investigators cannot determine whether it is really alienated or not, and what might be done to reduce or eliminate that condition. In the case that we have discussed throughout this article -that of British Muslims and the Anglican establishment -we actually think that if political theorists were to follow up on their claims of alienation, they would not find the necessary evidence, although they might find some support for extending establishment to recognize other faiths (Modood, 2019: 12-13, 205-209). Most religious minorities do not seem to be alienated from their state due to state-religion arrangements that may be described as establishment or that approximate to it (Perez et al., 2017). Nevertheless, it is the case that in many parts of the world majoritarian nationalism is on the rise, and it often takes a religious character (Soper and Fetzer, 2018). One country in which this is most evident is India, whose state and democratic institutions are being redesigned away from religious inclusivity in favour of Hindutva, or Hindu nationhood, and where there is growing othering and persecution of Muslims (Farooqui, 2020;Khan, 2019;Shaban, 2018). While there can be various grounds for criticizing such developments, in our opinion a central one would be the alienation of Muslims.
We must emphasize for one last time that, by making the experience of a group -in particular the group's own understanding of its experience -a vital part of the evidence about their situation, our account is not merely subjective in the sense that Laborde and Laegaard define that concept. 19 This is because we do not claim that group members' reports of their feelings taken in isolation provide adequate evidence to determine whether or not they are alienated. We do deny, however, that such a determination can be reached without any attempt to understand a minority on its own terms. As we have argued, if political theorists, and their surrogate reasonable observers, do not test their own potential biases by seeking to understand what a group thinks and feels, they run a significant risk of misunderstanding that group's situation. Avoiding both these options, we have argued that group experience -including the experience of alienation -is intimately connected to the group's sense of itself and the situation in which it finds itself, including the structures of power in which it is located, and the normatively understood group relations which shape how that group thinks and feels.
As we have just reiterated, our overarching purpose in this article has been to argue that establishment may be one means by which a process of othering results in the alienation of a particular group. For this reason, we have suggested that, when assessing the justifiability of specific cases of establishment, whether a group is alienated or not by that particular state-religion relation is one important factor that needs to be taken into account. This requires empirical research and dialogical engagement with the minorities in question, and not just deontological or 'reasonable person' or 'objective' ratiocination. In passing, we have also suggested that our argument is not specific to establishment but has the scope, suitably adapted, to extend to all state-religion arrangements, including strict separation. What we have not attempted to do here is to outline the sort of relationship between the state and religion which we think would be desirable in light of our concerns about alienation. In closing, however, we would like to suggest that one form that such a relationship might take would be plural or multiple establishment. If alienation is a condition in which a group is unable to identify with the state, and if one of the state's duties is to try to make such identification possible, then we believe that in some circumstances more than one religious community should receive a significant degree of formal recognition from the state. Defending this further argument is a task we must undertake elsewhere.
Secularism and the Governance of Religion: Bringing together European and Asian Perspectives (GREASE) project, funded by the European Union's Horizon 2020 research and innovation programme under grant agreement number 770640.