I. Introduction
In examining how the law talks about people, James Boyd White (Reference White1985) distinguishes between characters and caricatures. While characters, he argues, are ‘believable, full, complex’, caricatures reduce people to ‘single exaggerated aspects, to labels, roles, moments from their lives’ (Reference White1985, pp. 113–114). ‘The law’, he pithily pronounces, ‘is a literature of caricature’ (Reference White1985, p. 114).
Over the past few decades, a number of legal scholars have wrestled with the (in)capacity of the law to do full justice to individuals' complex and shifting identities on the ground (Abrams, Reference Abrams1994, Reference Abrams1996; Karst, Reference Karst1995; Minow, Reference Minow1997; Douzinas, Reference Douzinas2002; Tirosh, Reference Tirosh2007a). One of the main questions animating these debates is how the law, with its ‘discomfort with uncertainty’ and ‘thirst for fixed answers’, can be attentive and responsive to unstable lived experiences (Minow, Reference Minow1997, p. 77).
The tensions and gaps between a real-life person and her ‘legal persona’ are particularly visible in the context of legal proceedings (Douzinas, Reference Douzinas2000, p. 237); the law has its own recognisability terms, that is, the rules that facilitate claimants' recognition in the courtroom. If they want to make sense to courts, claimants have to tell their stories and present themselves in legally recognisable ways. This not only means translating everyday language into legal language (Yovel, Reference Yovel2010, p. 2), turning personal traits into legal arguments (Tirosh, Reference Tirosh2007a, p. 133), transforming stories into rule-oriented narratives (Yovel, Reference Yovel2010, p. 9) and following ‘rules of evidence and procedure’ (Baer, Reference Baer2010, p. 64). Most fundamentally, this may also mean turning particularities into generalities (Douzinas, Reference Douzinas2002, p. 402) narratives into meta-narratives (Tirosh, Reference Tirosh2007b, p. 7), ‘characters’ into ‘caricatures’ (White, Reference White1985).
Applicants' real-life and complex stories may thus get lost in the various layers of judicial translation, including legal counselling, which often turns the personal into the collective or ‘represses the litigant as a distinctly contingent subject in order to reconstruct it in rhetorically effective ways’ (Yovel, Reference Yovel2010, p. 2). Law's need for clarity and certainty may therefore explain why litigants often adopt fixed notions of group identity (Minow, Reference Minow1997, p. 77) in what is known as ‘strategic essentialism’.Footnote 1
The ‘essentialising proclivities of law’ (Cowan, Dembour and Wilson, Reference Cowan, Dembour, Wilson, Cowan, Dembour and Wilson2001, pp. 10–11) are particularly at work in assessments of identity traits – including cultural and religious traits – as individuals inevitably come to courts as part of collectives. In the words of Susanne Baer: ‘Whenever a “culture” or a “religion” claims recognition, we have the problem of reification, in that this suggests that the culture or religion is homogenous’ (Reference Baer2010, p. 59). Or, as stated by the South African Constitutional Court: ‘There is a danger of falling into an antiquated mode of understanding culture as a single unified entity that can be studied and defined from outside.’Footnote 2 In fact, attempts to capture any group commonality often fall prey to charges of ‘essentialism’.Footnote 3 Essentialism has been understood as ‘a belief in the real, true essence of things, the invariable and fixed properties which define the “whatness” of a given entity’ (Fuss, Reference Fuss1989, p. xi). These properties may be attributed to ‘everyone identified with a particular category’ or to the ‘category itself’ (Phillips, Reference Phillips2010, pp. 50, 53). The assumption is that all within the category ‘share the same inherent characteristics’ (Wong, Reference Wong1999, p. 275).
The risks of essentialising, as feminist and other scholars have long noted (Minow, Reference Minow1997; Fraser, Reference Fraser2000; Benhabib, Reference Benhabib2002; Phillips, Reference Phillips2007), include reducing people to ‘one trait, one viewpoint and one stereotype’ (Minow, Reference Minow1997, p. 34), trapping people in categories that deny self-definition (Minow, Reference Minow1997, p. 78), fixing differences among groups and ‘imposing uniformity within them’ (Appiah, Reference Appiah2005, p. 151) and policing group boundaries ‘to regulate internal membership’ (Benhabib, Reference Benhabib2002, p. 68). Another risk of positing one group experience as paradigmatic at the expense of others – which is what essentialism is mostly about (Munro, Reference Munro2006, p. 138) – is that it reinforces hierarchies within groups or categories (Hekman, Reference Hekman2004, p. 3). The experiences of some are thus either ignored or treated as ‘different’ from the ‘norm’ (Harris, Reference Harris1990, p. 615).
Incorporating these insights from anti-essentialist critiques and borrowing tools from critical discourse analysis, this paper scrutinises the ways in which the Strasbourg Court represents individuals' religious and cultural practices in its legal discourse. My analysis starts from the assumption that, given the nature of legal discourse and of cultural and religious claims, some degree of abstraction or generalisation is inevitable. For instance, it is hard to imagine an analysis of Article 9 ECHR (freedom of religion) or Article 14 ECHR (non-discrimination) without the categorisations intrinsic to their operation. It is equally hard to think of cultural and religious claims without individuals' identifications with groups and, as a result, without some level of collectivisation. The challenge is distinguishing inevitable and inoffensive generalisations from more problematic ones. As Anne Phillips (Reference Phillips2010, p. 58) puts it: ‘Essentialism is a way of thinking not always so easily distinguished from more innocent forms of generalisation, and what is wrong with it is often a matter of degree rather than categorical embargo.’
The crucial questions then are when and why the Court's generalisations of applicants and their practices become problematic. These questions are far from minor if one considers that the law ‘imposes itself coercively on the lives of those who come within its embrace’ (Hutchinson, Reference Hutchinson1991, p. 1555). The Court's articulation of a certain notion of religious or cultural identity may thus impose the power of the law behind the designated identity (Minow, Reference Minow1997, p. 61). Moreover, over time, the Court's discourse may sustain ‘a vocabulary of identities’ that may either channel future claimants ‘into recognized identity categories with conventional scripts for behaviour’ (Karst, Reference Karst1995, p. 295) or exclude them from protection. This is because of ‘the sedimenting effects’ produced by the development of case-law, which tends to entrench ideas about what group traits ‘are’ (Johnson, Reference Johnson2010, p. 72).
The thrust of my argument is that the Court's reliance on generalisations becomes problematic when, following the applicant's exclusion (usually by reducing her to and replacing her by one general trait), the Court (i) equates the trait in question with negative stereotypes and (ii) posits the trait in question as the group's ‘paradigmatic’ practice or as representative of the whole group. I further contend that these two sorts of depiction are harmful both for the individuals (actual applicants and those likely to be affected by the Court's rulings in the future) and for their groups because of the inequalities and exclusions they sustain across and within groups.
The paper proceeds as follows. After offering a brief explanation of the methodology, I present the chief research findings of my study and offer an in-depth analysis of the Court's discourse in the areas where I have found problematic depictions. My findings suggest that the Court uses the two problematic modes of reasoning mentioned above most frequently when assessing the practices of Muslim women, Sikhs and Roma Gypsies. The paper continues with a brief comparison with the Court's broader case-law, including gender and sexuality cases, with a view to offering possible explanations for why problematic representations are to be expected in certain groups of cases. Based on the lessons drawn from the paper, I conclude by sketching out the basic elements of an approach capable of mitigating stereotyping and essentialising risks.
II. Methodology
2.1 Selection of cases and scope of the analysis
The analysis offered in this paper looks at the Court's legal reasoning in primarily three areas of its case-law: freedom of religion (Article 9 ECHR),Footnote 4 the right to respect for minority cultural lifestyle (Article 8 ECHR)Footnote 5 and non-discrimination (Article 14 ECHR).Footnote 6 The reason for focusing on these ECHR Articles is because these are the provisions through which applicants have most commonly asked for protection of their religious and cultural practices. The selected cases comprise a mix of ‘high-profile cases’ – which I define as Grand Chamber and widely cited cases in the Court's jurisprudence – and less-known cases. Moreover, the selection includes a mix of Grand Chamber judgments, Chamber judgments and inadmissibility decisions passed over roughly the last fifteen years.Footnote 7 Though the sample is by no means complete, it is substantial enough to allow for meaningful analysis.
The analysis is based on two levels of investigation. The first (and broad) level involves a look into the Court's wider discourse on identity traits – in particular, case-law concerning individuals' religious/cultural practices and case-law involving gender and sexuality.Footnote 8 I compare the Court's language in these different areas of its identity-based case-law in order to identify significant patterns – i.e. which groups of applicants (and their traits) are represented most problematically in Strasbourg – and to formulate possible explanations for why problematic representations are to be expected in certain groups of cases. The second (and specific) level of inquiry involves in-depth discourse analysis of the subsets of cases where problematic patterns have been found. Since this form of analysis only allows space for detailed discussion of a limited number of cases, I focus on four problematic judgments: Leyla Şahin v. Turkey,Footnote 9Dahlab v. Switzerland,Footnote 10Mann Singh v. France,Footnote 11 and Chapman v. the United Kingdom. Footnote 12
2.2 Critical discourse analysis
My analysis of the Court's discourse relies on a set of heuristic tools employed in critical discourse analysis, a form of analysis that views representation in discourse as ‘a constructive practice’ rather than as just neutral communication of events or ideas (Fowler, Reference Fowler1991, p. 25). For critical discourse analysis, the language through which identities are constructed reveals ‘a subject's attitudes and ideologies’ (Benwell and Stokoe, Reference Benwell and Stokoe2006, p. 44). In particular, this type of discourse analysis looks at the role that ‘structures, strategies and other properties of text, talk, verbal interaction or communicative events’ play in the (re)production of inequality in society (Van Dijk, Reference Van dijk, Wetherell, Taylor and Yates2001, p. 300).
Of particular relevance for my analysis are some notions developed by a critical discourse scholar, Theo van Leeuwen (Reference Van leeuwen2008), in his work on the representation of social actors and social actions. Van Leeuwen critically examines the transformations that take place in the recontextualisation of social practices – that is, in discourse (Reference Van leeuwen2008, pp. 3–6). These transformations may include substituting, excluding and adding elements of such practices (e.g. social actors, time, location) by way of different representational choices (Reference Van leeuwen2008, pp. 17–22). Each of these choices, in turn, takes place through ‘specific linguistic or rhetorical realizations’ (Reference Van leeuwen2008, p. 25). Simply put, Van Leeuwen's idea is that, when we represent a social practice, we add new meanings by transforming the actual elements of the practice through different linguistic and rhetorical means.
Van Leeuwen's work is particularly apt to illuminate my task for various reasons. One of them is that representation – more specifically, the representation of facts and of the individuals involved in the facts – is at the heart of the adjudication enterprise (Karst, Reference Karst1995, p. 281). Another reason lies in the purpose of Van Leeuwen's scheme, which coincides with the main purpose of my investigation: scrutinising the ways in which actors and practices are represented in discourse. Moreover, some of the notions from Van Leeuwen's scheme are – per my hypothesis – potentially highly relevant to the selected cases. My study of the Court's discourse has particularly benefited from his notions of the exclusion of social actors, the objectivation of social actions and assimilation of social actors by collectivisation.
On the exclusion of social actors, Van Leeuwen observes (Reference Van leeuwen2008, p. 29): ‘When the relevant actions (e.g., the killing of demonstrators) are included but some or all of the actors involved in them (e.g., the police) are excluded, the exclusion does leave a trace.’ He makes a further distinction between the full exclusion of social actors, which he calls ‘suppression’ (the actors are referred to nowhere in the text), and their less radical exclusion, which he dubs ‘backgrounding’ (the actors ‘may not be mentioned in relation to a given action, but they are mentioned elsewhere in the text’) (Reference Van leeuwen2008, p. 29). The objective is to identify patterns of inclusion/exclusion and to combine them with the ways in which social actors are represented (Reference Van leeuwen2008, p. 32). As regards the ‘objectivation’ of social actions, Van Leeuwen explains that this is one way in which social actions may be ‘deactivated’, that is, ‘represented statically, as though they were entities … rather than dynamic processes’ (Reference Van leeuwen2008, p. 63). In turn, collectivisation of social actors is a form of assimilation that occurs when they are referred to as groups by means of pluralities (e.g. Christians) or by means of ‘a mass noun or a noun denoting a group of people’ (e.g. the community) (Reference Van leeuwen2008, p. 37).
The exclusion of social actors and the objectivation of their social actions may take place through various linguistic forms, including ‘nominalisation’ and ‘passivisation’. Nominalisation consists in ‘turning verbs into nouns’ (Fowler, Hodge, Kress and Trew, Reference Fowler, Hodge, Kress and Trew1979, p. 14) and passivisation in privileging the passive voice over the active voice (Billig, Reference Billig2008, p. 785). An example of nominalisation would be ‘the attack on demonstrators’; an example of passivisation, ‘demonstrators were attacked’ (Billig, Reference Billig2008, p. 785).
The notions borrowed from Van Leeuwen's scheme serve as a starting point for my analysis, which I further supplement with insights from (legal) scholarship on stereotyping and essentialism.
III. Stereotyping and naturalising in the Court's discourse
In this part, I introduce the areas of the Court's case-law where I have most commonly found the two problematic modes of reasoning presented earlier and offer an illustration of their operation by means of in-depth discourse analysis. Moreover, I unpack the harmful consequences of these forms of reasoning both for the individual applicants involved in the particular cases and for their groups.
3.1 Key findings
My first (and broad) level of analysis of the Court's discourse on identity traits reveals a widespread use of collectivisations of the applicants and objectivations of their traits. ‘Transsexualism’,Footnote 13 ‘pregnancy’,Footnote 14 ‘homosexuality’Footnote 15 and the ‘Gypsy way of life’Footnote 16 are but a few examples of the objectivations of applicants' traits. ‘Transsexuals’,Footnote 17 ‘women’ (or ‘a woman’),Footnote 18 ‘sexual minorities’Footnote 19 and ‘Jehovah's Witnesses’Footnote 20 are, in turn, just some instances of collectivisations.
On one side, I have found that sometimes the deployment of objectivations and collectivisations seems inevitable, intrinsic to the operation of the law or legal reasoning. This happens, for example, when the Court has to determine whether a difference in treatment exists by way of comparators in discrimination cases (e.g. same-sex couples and different-sex couples,Footnote 21 men and womenFootnote 22). It also occurs when the Court refers to the content of domestic laws and regulations (e.g. ‘the ban on the wearing of religious symbols in universities’,Footnote 23 under Polish law ‘abortion is lawful where pregnancy poses a threat to the woman's life or health’,Footnote 24 French law opens up the possibility of ‘adoption by a single homosexual ’Footnote 25).
On the other side, my findings also demonstrate that, at other times, collectivised representations of applicants and objectivised depictions of their traits have resulted in two types of problematic portrayal. The first kind of problematic depiction involves negative stereotyping: following the applicant's reduction to and disappearance behind an objectivised trait or a collectivised representation, the Court associates the trait or group in question with a stereotype that ‘assigns difference’ (Cook and Cusack, Reference Cook and Cusack2010, p. 16). This form of stereotyping generally reflects ‘prejudice or bias’ about a group and may exacerbate its subordination (Reference Cook and Cusack2010, p. 16). Briefly put, stereotyping is a form of ‘intergroup bias’, which in turn is a tendency to ‘evaluate one's own membership group (the in-group) or its members more favourably than a non-membership group (the out-group) or its members’ (Dovidio, Hewstone, Glick and Esses, Reference Dovidio, Hewstone, Glick, Esses, Dovidio, Hewstone, Glick and Esses2010, pp. 3, 5).
The second type of problematic representation entails what I will refer to as ‘naturalising’: following the applicant's reduction to and disappearance behind an objectivised trait or a collectivised representation, the Court associates the trait in question with an unchanging core of the applicant's group identity. Two forms of essentialism are at work in this mode of reasoning: the attribution of certain characteristics to some ‘static “essence”’, in a move that ‘naturalises differences that may be historically variant and socially created’ and the treatment of such characteristics ‘as the defining ones for anyone in the category’ (Phillips, Reference Phillips2010, pp. 53, 57).
My examination suggests that these two problematic modes of reasoning feature most frequently in cases concerning members of religious and cultural minorities. In particular, negative stereotypes appear to a seemingly greater extent in cases concerning Muslim women and naturalisations in cases involving Sikhs and Roma Gypsies.
3.2 Negative stereotypes
The cases exhibiting the first form of problematic representation concern Muslim women prohibited from wearing the headscarf.Footnote 26 What these cases have in common is that the Court tends to neglect the applicants and to focus nearly exclusively on their practices or symbols as though they had a separate existence, associating them with negative stereotypes. In other words, the Court talks to symbols and about symbols in stereotypical ways, without regard to the applicants' views or circumstances and without any basis on the evidence of the cases.
I examine the Court's discourse in this first group of cases through the lens of Dahlab v. Switzerland and Leyla Şahin v. Turkey. Dahlab concerns a female Muslim teacher prohibited from wearing the headscarf at a state school. Leyla Şahin concerns a female Muslim student not allowed to wear the headscarf at a state university. The Court declared Dahlab inadmissible and rejected Leyla Şahin on the merits. It examined the two cases primarily under Article 9 ECHR (freedom of religion).Footnote 27
3.2.1 The entrance of ‘the headscarf’
A critical analysis of the Court's representation of the applicants and their religious practices in Dahlab and Leyla Şahin shows that the applicants, though not really suppressed – the judgments include references to them elsewhere – are constantly pushed into the background in the Article 9 ECHR reasoning.
In Dahlab, in determining whether the restriction on the applicant's freedom of religion was necessary in a democratic society, the Court states:
‘The Court accepts that it is very difficult to assess the impact that a powerful external symbol such as the wearing of a headscarf may have on the freedom of conscience and religion of very young children. The applicant's pupils were aged between four and eight, an age at which children wonder about many things and are also more easily influenced than older pupils. In those circumstances, it cannot be denied outright that the wearing of a headscarf might have some kind of proselytising effect, seeing that it appears to be imposed on women by a precept which is laid down in the Koran and which, as the Federal Court noted, is hard to square with the principle of gender equality. It therefore appears difficult to reconcile the wearing of an Islamic headscarf with the message of tolerance, respect for others and, above all, equality and non-discrimination that all teachers in a democratic society must convey to their pupils.’Footnote 28 (emphasis added)
In this passage, the Court backgrounds the applicant through nominalisation. Instead of using an active verb clause with the applicant as the subject, the Court suppresses the applicant and turns the verb ‘to wear’ into the noun ‘the wearing of’ the headscarf. By means of nominalisation, the Court does not only background the agent (the applicant) but also objectivises her action (wearing the headscarf). Indeed, the action is thereby represented statically, as though it were an entity. Moreover, while the applicant is nearly excluded from the text, her supposed victims remain there, albeit in collectivised forms (‘very young children’, ‘applicant's pupils’, ‘children’). In fact, the Court names the applicant only once: when it talks about her alleged victims.
In Leyla Şahin, in turn, in assessing whether interference with the applicant's freedom of religion was necessary in a democratic society, the majority holds:
‘… In addition, like the Constitutional Court … the Court considers that, when examining the question of the Islamic headscarf in the Turkish context, it must be borne in mind the impact which wearing such a symbol, which is presented or perceived as a compulsory religious duty, may have on those who choose not to wear it. As has already been noted (see Karaduman, decision cited above, and Refah Partisi (the Welfare Party) and Others, cited above, § 95), the issues at stake include the protection of the “rights and freedoms of others” and the “maintenance of public order” in a country in which the majority of the population, while professing a strong attachment to the rights of women and a secular way of life, adhere to the Islamic faith. Imposing limitations on freedom in this sphere may, therefore, be regarded as meeting a pressing social need by seeking to achieve those two legitimate aims, especially since, as the Turkish courts stated, … this religious symbol has taken on political significance in Turkey in recent years.’Footnote 29 (emphasis added).
Here, the Court also backgrounds the applicant and objectivises her practice through nominalisation (‘wearing such a symbol’) and other noun phrases that act as either the subject (‘this religious symbol’) or the object (‘the question of the Islamic headscarf in the Turkish context’). As a result, the Court leaves the applicant out and turns ‘the headscarf’ or ‘the symbol’ into the centre of its discourse. Whereas the agency of Leyla Şahin is de-emphasised by way of nominalisations, the agency of the supposed victims is emphasised by the use of the word ‘choose’ in ‘those who choose not to wear it’.Footnote 30 Moreover, while the applicant is fully displaced by an objectivised version of her practice, her alleged victims remain in the text, albeit in a collectivised and somehow indeterminate form.
What is furthermore distinctive about the Court's representation of the headscarf in Leyla Şahin is the deletion of agency through ‘passivisation’. The Court tells us that the symbol is ‘presented or perceived as a compulsory religious duty’ but omits to say who actually ‘perceives’ or ‘presents’ the Islamic headscarf as such. Consciously or not, the fact is that the Court omits agency by leaving unspecified who exactly does the labelling of ‘compulsory’.
One of the reasons why critical discourse scholars have long viewed nominalisations with suspicion is because they facilitate reification (Billig, Reference Billig2008, p. 786). Critical discourse scholar Roger Fowler (Reference Fowler1991, p. 80) explains how, by way of nominalisation, ‘processes and qualities assume the status of things: impersonal, inanimate, capable of being amassed and counted, paraded like possessions’. This is precisely what happens in Dahlab and Leyla Şahin: nominalisation leads to the reification of the applicants' religious practices. By turning verbs into nouns, the Court linguistically creates a ‘thing’. It suggests that ‘the headscarf’ has a real or tangible existence, external to that of the applicants. It gives ‘the headscarf’ a life of its own, while denying the lives of the applicants.
3.2.2 The harmful impact on the case
The representational moves described above entail several negative implications in the particular cases, the majority of which play out in the analysis under Article 9(2) ECHR. This is the stage at which the Court establishes whether the interference with applicants' rights is necessary in a democratic society, most crucially, whether the interference is proportionate to the aim it pursues.
The first negative effect of fading the applicants into the background is the exclusion of their own views and particular circumstances from the proportionality analysis. Indeed, the Court does not just disregard the motivations behind their decisions to wear the headscarf.Footnote 31 It pays no attention to the suffering and loss caused to the applicants by the prohibitions (Evans, Reference Evans2010–2011, p. 330; Rorive, Reference Rorive2009, p. 2683). In this way, the Court excludes all possibility of balancing the importance of the applicants' practices – and the personal/professional/educational costs involved – against the importance of the public interests or rights of others at issue.
Another negative implication implicit in the Court's reification of the applicants' religious practices and the simultaneous obfuscation of their agency is the transformation of ‘the headscarf’ into the agent of the process of threatening others. Thus, in Dahlab, unable to locate the threat in the applicant – more precisely, in the quality and content of her teachingFootnote 32 – the Court searches for a location elsewhere, namely in the headscarf itself. As the Dahlab passage quoted earlier shows, the Court locates the threat in the headscarf by signifying the symbol by reference to: (i) its inherent or essentialist characteristics (‘powerful’, ‘external’ and ‘imposed’) and (ii) the possible reaction of others (young children on whom the symbol may have an impact). Thus, the powerful, visible and imposed symbol, on one side, and the children's tender age, on the other, come together to define the threat that the symbol represents: ‘some kind of proselytising effect’. In Leyla Şahin the Court similarly turns the reified symbol into a threatening agent. As in Dahlab, the threat does not come from the applicant herself. The threat instead comes from a combination of the essentialist attributes of the symbol (its ‘compulsory’ character) and the Turkish context (majority adhering to Islam and extremist political movements seeking to impose their symbols on society).
A third troubling consequence flowing from the Court's objectivation of the applicants' practices is their delegitimation. By reference to the authority of national courts, the Court resorts to what Van Leeuwen calls ‘authorisation’Footnote 33 in order to delegitimise ‘the wearing of the headscarf’. Indeed, many of the attributes that the Court ascribes to the Islamic headscarf come from domestic courts' discourses. In describing the applicants' religious practices, the Court either explicitly refers to these courts' decisions or uses words taken from these sources. For example, ‘like the [Turkish] Constitutional Court’, the Court keeps in mind the impact that a symbol perceived as compulsory may have on others.Footnote 34 Similarly, ‘as the [Swiss] Federal Court noted’, the Court states that the wearing of the headscarf is hard to square with gender equality.Footnote 35
But this is not the only way in which the Court delegitimises ‘the wearing of the headscarf’. The Court employs another form of delegitimation, which Van Leeuwen (Reference Van leeuwen2008, p. 110) dubs ‘moral evaluation’. This sort of delegitimation is based on ‘specific discourses of moral value’ (Reference Van leeuwen2008, p. 110). One explicit value on the basis of which the Court delegitimises the Islamic headscarf is ‘gender equality’, a Council of Europe valueFootnote 36 and, ultimately, a value of any democratic society.Footnote 37 Thus, in Dahlab, it finds it ‘difficult to reconcile the wearing of an Islamic headscarf with the message of tolerance, respect for others and, above all, equality and non-discrimination’.Footnote 38 In Leyla Şahin, and as Judge Tulkens observes in her dissent, the majority considers that wearing the headscarf is ‘synonymous with the alienation of women’.Footnote 39
By failing to ‘see’ the specific circumstances and motivations of the applicants and by relying instead on the domestic courts' preconceptions of ‘the headscarf’, the Court ultimately falls into a kind of essentialism referred to as ‘one of over-generalisation [and] stereotyping’ (Phillips, Reference Phillips2010, p. 50). Broadly speaking, stereotypes are ‘associations and beliefs about the characteristics and attributes of a group and its members that shape how people think about and respond to the group’ (Dovidio et al., Reference Dovidio, Hewstone, Glick, Esses, Dovidio, Hewstone, Glick and Esses2010, p. 8). As Carolyn Evans (Reference Evans2006, p. 73) notes, the Strasbourg Court implicitly advocates two contradicting stereotypes: Muslim women as victims of oppression in need of protection and Muslim women as aggressors from whom everyone needs protection. These generalisations are easily made, without being concretely substantiated with statistics or other evidence in the particular cases (Evans, Reference Evans2006, p. 54).
The Court's reliance on stereotypes about Muslim women harms the particular applicants, for they are denied equal access to benefits (e.g. education or work) based on preconceptions that do not match their actual characteristics, needs and circumstances (Cook and Cusack, Reference Cook and Cusack2010, p. 61). As feminist legal theorists have shown, stereotyping may cause distributional harms to stereotyped group members (Cook and Cusack, Reference Cook and Cusack2010, pp. 59–60; Timmer, Reference Timmer2011, pp. 715–716). The harm of ‘maldistribution’ involves the denial of the resources necessary to participate in social life (Fraser, Reference Fraser2000, pp. 116–118).
3.2.3 The harmful impact beyond the case
The Strasbourg Court is often thought to be ‘one of the most important discoursing machines in the world’ given the ‘pan-European legal framework’ it produces for human rights adjudication (Johnson, Reference Johnson2010, p. 74). This perception gives the Court's representational discourse singular and influential force beyond the particular cases. Indeed, the Court's legal discourse does not just have implications for the parties involved in the specific cases but may affect future applicants and their groups as well.
One of the profound, broader implications of the Court's use of negative stereotypes is the ‘misrecognition’ of the group in question (Muslims in general and Muslim women in particular). The ‘misrecognition’ harm caused through stereotyping (Cook and Cusack, Reference Cook and Cusack2010, pp. 59–60; Timmer, Reference Timmer2011, pp. 715–716) operates by constituting groups ‘as inferior, excluded, wholly other, or simply invisible – in other words, as less than full partners in social interaction’ (Fraser, Reference Fraser2000, p. 113).
The stereotype that Muslim women are oppressed embeds a mix of stereotypical assumptions (gender/religious/orientalist) that constitute Muslims and Muslim women as ‘inferior’ and ‘wholly other’. Thus, this stereotype implies that Muslim men are oppressors, reflecting a historical, colonialist interpretation of gender relations between Muslim men and women, which constructs ‘Muslim men as barbaric oppressors of women, inherently inferior to Western men’ (Vojdik, Reference Vojdik2010, p. 676). Moreover, this stereotype implicitly portrays a religion, ‘Islam’, as ‘barbaric’ and ‘backward’ compared with the ‘West’ (Malik, Reference Malik, Ungureanu and Zucca2012, p. 112). As Marie-Bénédicte Dembour (Reference Dembour2006, p. 213) notes, the Court ‘puts the West on a pedestal and demonizes Islam’.
In negatively stereotyping Muslims and Muslim women, the Court thus implicitly assigns them a lower status vis-à-vis non-Muslim women and non-Muslims, thereby (re)producing hierarchies between groups – or inter-group hierarchies. Stereotypes, as Alexandra Timmer (Reference Timmer2011, p. 715) argues, ‘often serve to maintain existing power relationships’, upholding ‘a symbolic and real hierarchy between “us” and “them”’. The Court (re)creates these hierarchies by implicitly relying on a series of dichotomies (agency/victimisation and reason/culture) and by further associating one group with the ‘positive’ side (agency/reason) and the other group with the ‘negative’ side (victimisation/culture). Thus, whereas Muslim women are assumed to be wholly determined and victimised by their cultures (their religious practices are ‘imposed’ on them by the Koran), non-Muslims are assumed to be rationally choosing agents. As Leti Volpp (Reference Volpp2001, p. 1192) notes in another context, ‘[b]ecause the Western definition of what makes one human depends on the notion of agency and the ability to make rational choices, to thrust some communities into a world where their actions are determined only by culture is deeply dehumanizing’.
Besides the misrecognition implications for the applicants' groups, the Court's stereotyping reasoning has potentially damaging implications for future applicants. The Court's stereotypical constructions of the applicants' religious practices in Leyla Şahin and Dahlab have already turned into principles that, by now, have become well entrenched in the Court's ‘headscarf’ case-law.Footnote 40 Future applicants will therefore most likely have a hard time showing that they do not match the Court's well-established negative image of ‘the Islamic headscarf’ or challenging the image itself.
3.3 Naturalisations
The cases examined in this section involve the second kind of flawed depiction in the Court's legal discourse; the kind that entails equating the trait in question with the group's ‘paradigmatic’ practice / way of life. As I mentioned earlier, I refer to this problem as ‘naturalisation’ because it makes ‘natural’ what in fact is historically contingent or socially constructed (Warne, Reference Warne, Braun and McCutcheon2000, p. 141). This sort of portrayal seems to appear most often in cases concerning SikhsFootnote 41 and Roma Gypsies.Footnote 42
I analyse this kind of representation through the lens of Mann Singh v. France and Chapman v. the United Kingdom. Mann Singh concerns a Sikh man denied the renewal of his driver's licence for refusing to take off his turban for the picture. Chapman deals with the claim of a Gypsy woman evicted from her own land for stationing her caravan there without planning permission. The Court rejected Chapman's alleged violation of her right to respect for home, private and family life (Article 8 ECHR) and dismissed her discrimination complaint (Article 14 ECHR). Mann Singh's claims – including his freedom of religion complaint (Article 9 ECHR) – were all declared inadmissible.
3.3.1 The entrance of ‘the turban’ and ‘the Gypsy way of life’
In both Mann Singh and Chapman, the Court backgrounds the applicants and objectivises their practices, albeit through different representational means. The linguistic move in Mann Singh is passivisation – the use of the passive voice instead of the active voice. In assessing whether Mann Singh's wearing of his turban falls within the scope of Article 9(1) ECHR, the Court says:
‘According to the applicant, the Sikh faith compels its members to wear the turban in all circumstances. It is considered not only at the heart of their religion, but also at the heart of their identity. Therefore, the Court notes that this is an act motivated or inspired by a religion or belief.’Footnote 43 (emphasis added)
In the second sentence of this passage, the Court states that the turban is ‘considered’ to be at the heart of the Sikh religion and identity without saying who actually considers the turban as such. The context indicates that it is Mann Singh who views the turban this way.Footnote 44 However, with the passive construction in ‘[the turban] is considered’ – that is to say, with the deletion of Mann Singh as the subject – the Court separates the turban from its wearer, objectivises his religious practice by reducing it to ‘the turban’ and, ultimately, gives the practice a life of its own, ready to travel around its case-law in the form of a principle.Footnote 45
In Chapman, in turn, the Court's preferred representational form to push the applicant aside is collectivisation. In determining whether Article 8(1) ECHR was at issue, the Court says:
‘The Court considers that the applicant's occupation of her caravan is an integral part of her ethnic identity as a Gypsy, reflecting the long tradition of that minority of following a travelling lifestyle. This is the case even though, under the pressure of development and diverse policies or by their own choice, many Gypsies no longer live a wholly nomadic existence and increasingly settle for long periods in one place in order to facilitate, for example, the education of their children. Measures affecting the applicant's stationing of her caravans therefore have an impact going beyond the right to respect for her home. They also affect her ability to maintain her identity as a Gypsy and to lead her private and family life in accordance with that tradition.’Footnote 46 (emphasis added).
In this passage, the Court first foregrounds the applicant but then leaves her aside, assimilating her to ‘that minority’ and ‘many Gypsies’. By way of collectivisation, therefore, the Court separates the applicant from her group, sending her backstage and bringing her group centre stage.
In another of part of its legal reasoning – more precisely, when setting out the principles necessary to determine if the refusal to let the applicant stay on her land was justified – the Court affirms: ‘[T]here is thus a positive obligation imposed on the Contracting States by virtue of Article 8 to facilitate the Gypsy way of life.’Footnote 47 Here, the Court, first of all, objectivises the applicant's lifestyle by representing it statically rather than dynamically: ‘the Gypsy way of life’.Footnote 48 The Court thereby abstracts the way of life from those who, like the applicant, give life to a nomadic lifestyle. Moreover, the Court reduces the Gypsy way of life to one single trait: nomadism. It implies that there is no other form of living the Gypsy way of life than sticking to nomadism (Farget, Reference Farget2010, p. 250).
Collectivisations and objectivations allow the Court to assess Mann Singh's and Chapman's practices or lifestyles in highly essentialist terms: the Court closely ties their practices to the Sikh and Gypsy identities. At work in the two cases is the kind of essentialism that treats ‘certain characteristics as the defining ones for anyone in the category, as characteristics that cannot be questioned or modified without thereby undermining one's claim to belong to the group’ (Phillips, Reference Phillips2010, p. 57). While in Mann Singh the defining characteristic is the turban, in Chapman the defining trait is travelling.
Now, on what basis does the Court characterise the applicants' practices in these essentialist ways? In Chapman, the Court resorts to history. By recourse to ‘the long tradition’, the Court insists that travelling remains essential to all Gypsies. This is so even when, by the Court's own admission, reality may show that travelling is not practised homogenously within the group (many of them no longer live a ‘wholly nomadic existence’ as a result of either pressure or choice). In this way, the Court ends up freezing the group in time ‘to a retrospective and nostalgic understanding of their identity’ (Eisenberg, Reference Eisenberg2009, p. 131). In Mann Singh, in turn, the Court simply states that the turban ‘is at the heart of’ the Sikh religion and identity. In turning the nomadic lifestyle or the turban into a fixed and ‘natural’ defining group characteristic, the Court obscures the socially created and contingent character of the traits in question.
3.3.2 The impact on the cases
In contrast to Dahlab and Leyla Şahin, the Court's reliance on generalisable and reducible (group) traits in Chapman and Mann Singh serves, to some extent, to better understand the applicants' positions.Footnote 49 Indeed, the essentialist construction of Chapman's and Mann Singh's practices leads to their recognition in the ‘scope’ stage of the analysis. This is the threshold stage at which the Court establishes whether the claim in question attracts the protection of an ECHR provision.
For instance, in Mann Singh, the turban counted as ‘a manifestation’ of the applicant's religion for the purposes of Article 9(1) ECHR,Footnote 50 largely because the practice was viewed at the core of Sikh faith and identity. In Chapman, in turn, the Court's reliance on essentialist views of the applicant's group lifestyle seems to have been instrumental in the expansion of the scope of Article 8 ECHR: the Court recognises that at stake is not just the applicant's right to respect for her home but also her right to lead her private and family life in accordance with her tradition as a Gypsy.
Moreover, in Chapman, the Court's reliance on other generalisable group-based traits such as ‘vulnerability’ results in yet another significant recognition for the applicant: the establishing of a positive obligation to facilitate ‘the Gypsy way of life’, even though the obligation turns out to be limited in scope.Footnote 51 The Court holds:
‘[T]he vulnerable position of Gypsies as a minority means that some special consideration should be given to their needs and their different lifestyle both in the relevant regulatory planning framework and in reaching decisions in particular cases … To this extent, there is thus a positive obligation imposed on the Contracting States by virtue of Article 8 to facilitate the Gypsy way of life.’Footnote 52
In the rest of the reasoning, however, the Court gives lip service to these recognitions. In the proportionality analysis, Mann Singh's and Chapman's essentialised traits are either completely eclipsed by the state's alleged countervailing interests or expressly used to diminish the weight of the applicants' interests. Indeed, in Mann Singh the applicant's essentialised practice plays no role in the proportionality analysis. Mann Singh and what is at stake for him – including the alleged importance initially recognised in the turban for his Sikh identity – is virtually absent in the Court's analysis of whether the interference with his right was justified. The Court looks exclusively at the state's justifications of public order and security and concludes that the obligation to take off the turban for the driver's licence picture was necessary in a democratic society.Footnote 53
In Chapman, on the other hand, the Court's essentialist view expressly serves to reduce the seriousness of what is stake for the applicant in the proportionality analysis. The Court says: ‘[T]he present case is not concerned as such with the traditional itinerant Gypsy lifestyle.’Footnote 54 In the eyes of the Court, the applicant's lifestyle did not fit ‘the Gypsy way of life’ because she no longer lived a nomadic lifestyle. The Court found that she was actually ‘resident on site’ during considerable periods.Footnote 55 The conclusion was therefore that the applicant did not ‘wish to pursue an itinerant lifestyle’.Footnote 56
What follows is that the applicant is no longer viewed as a (‘proper’ or ‘authentic’) group member but just as an individual who ‘chose’ to settle.Footnote 57 Julie Ringelheim (Reference Ringelheim and Brems2012, p. 433) persuasively argues:
‘This reading of the facts appears narrowly individualistic in two ways: first, notwithstanding its acknowledgment that caravan life holds an important place in Gypsy collective identity, at the end of the day the majority discards the cultural dimension of the issue and reduces the wish of Ms. Chapman to live in a caravan to a question of mere individual preference.’
Thus, the essentialist glasses do not allow the Court to see the more complex circumstances in which the applicant found herself. An acknowledgment of such circumstances – more precisely, of the fact that the applicant was pushed into a settled way of life by policies unresponsive to her travelling lifestyle – could have led to a different conclusion. The dissenters, in fact, reached a different conclusion in this regard. They rejected the government's argument that the applicant's intention to settle should detract from the seriousness of the interference.Footnote 58 They noted instead that pressure from UK law ‘has had the effect of inducing many Gypsies to adopt the solution of finding a secure, long-term base for their caravans on their own land’.Footnote 59
There are two troubling consequences flowing from the majority's essentialist approach in the Chapman proportionality analysis. In the first place, in positing one form of lifestyle as ‘the’ group's paradigmatic type, the Court sets a standard against which the applicant's practice falls short. The problem here is thus one of intra-group exclusion and inequality: since the applicant's lifestyle is not as ‘authentic’ as the practices of other group members who have stuck to travelling, her lifestyle is taken less seriously and her group membership called into question. In the second place, and in connection with the first problem, the Court's essentialist reasoning paradoxically serves to strip the applicant's case of the group vulnerability dimensions. This group- and context-stripping approach misses key structural elements that would have allowed for a better appreciation of the vulnerable position in which the applicant found herself. Indeed, one such element was the disadvantageous impact of the planning regulations on the applicant's lifestyle as a member of a particularly vulnerable group. This time, the Court fails to address inter-group exclusion and inequality, as only those group members whose concerns are not taken into account by the regulatory norms suffer detrimental consequences.
It may be argued that it is legitimate for the Court to underline any inconsistencies between Chapman's account of the Gypsy lifestyle and her own actual way of life, as it was the applicant herself who used such an essentialist account to reinforce her claims.Footnote 60 This argument should be rejected on the following basis. The applicant's and the Court's appeals to essentialism cannot be evaluated in the same way given the different positions of power from which essentialist arguments are deployed: the applicant relies on essentialism from a non-dominant position (that of a vulnerable minority) while the Court does it from a dominant one (that of a supranational court). As Annie Bunting (Reference Bunting1993, pp. 11–13, 17, 18) argues in another context, essentialism employed to critique dominant discourses and essentialism employed from dominant positions should be evaluated asymmetrically, as the latter may serve to reinforce exclusion and inequality. Given the authoritative force of the Court's essentialist depictions, it may be problematic to rely on naturalising depictions even when applicants do it themselves.
In summary, the deployment of essentialism is double-edged in Chapman. At the scope level, the Court's essentialist arguments were seemingly instrumental in the recognition of the applicant's right to lead her private and family life in accordance with her traditional lifestyle as a Gypsy. In the proportionality analysis, however, the Court's essentialist arguments served to minimise the seriousness of what was at stake for the applicant and to detach her case ‘from its wider context and from the global difficulties faced, in the whole country, by the minority she belonged to’ (Ringelheim, Reference Ringelheim and Brems2012, p. 432). As with Leyla Şahin and Dahlab, the Court thus reduces the weight of Chapman's interests in the proportionality analysis and fails to appreciate the misrecognition harm implicit in the impugned decision. The underlying rationale is, however, different in the two sets of cases. While in the Leyla Şahin and Dahlab cases the Court undermines what is at stake for the applicants by forcing them into a mould it condemns, in the Chapman case the Court reduces the importance of the applicant's interests by forcing her out of a mould it esteems.
3.3.3 The impact beyond the cases
Contrary to its discourse in Leyla Şahin and Dahlab, the Court's discourse in Chapman and Mann Singh does not go as far as delegitimising the applicants' group practices via negative stereotyping. This is probably one of the most significant differences between the group of cases studied in the previous section and those examined in this section. Indeed, the Court does not deem ‘the wearing of the turban’ contrary to Convention values such as gender equality. Nor does it describe the ‘Gypsy way of life’ as, say, a threat to the rights of those who lead a sedentary lifestyle.
Yet the Court's essentialist discourse results in ‘naturalising’. One of the problems arising from naturalising is that it sets up ‘a standard by which to judge deviation’ (Warne, Reference Warne, Braun and McCutcheon2000, p. 141). The danger of this sort of reasoning therefore lies in the exclusions and inequalities it sustains by deeming some lifestyles or practices ‘natural’ or ‘normal’ and others ‘deviant’. The Court's naturalising language in Chapman and Mann Singh implicitly (re)affirms intra-group exclusions and inequalities: those who do not follow the ‘core’ practices of travelling in a caravan or wearing a turban – or do not follow them strictly – may be regarded as less ‘members’ than others or, simply, as not ‘members’ at all. Briefly put, ‘those who do not fit are in trouble’ (Phillips, Reference Phillips2010, p. 57). Chapman herself is a tragic example.
Moreover, the Court's naturalising reasoning also risks (re)producing inter-group exclusions and inequalities. This kind of risk is illustrated in Horie v. the United Kingdom – a little-known inadmissibility decision concerning a New Traveller who had pursued a nomadic lifestyle for almost three decades.Footnote 61 The Court says obiter dicta that, unlike ‘Romani gypsies’ and ‘Irish Travellers’, ‘New Travellers live a nomadic lifestyle through personal choice and not on account of being born into any ethnic or cultural group’.Footnote 62 The Court hereby reaffirms the natural or immutable status of ‘travelling’ in the Gypsy tradition, albeit by a different criterion – birthFootnote 63 – implying that those who are gypsies by choice are not ‘real’ gypsies. Once the trait is cast in this immutable way, it serves to exclude groups practising itinerant lifestyles such as Ms Horie's from legal recognition.
The reasoning in Chapman, Mann Singh and Horie thus leads to a classic essentialism problem: the policing of group boundaries. Mann Singh may have met the criterion of group membership but Sikh applicants not wearing the turban ‘in all circumstances’ will most likely fail the test, just like Chapman and Horie failed their group membership tests.
In some of its later ‘caravan’ case-law, the Court has adopted more inclusive and socially constructed accounts of applicants' lifestyles (Farget, Reference Farget2012, p. 303). For instance, in Connors v. the United Kingdom, the Court has refused to deploy the sort of generalisations that ‘would identify the nomadic lifestyle as the essence of gypsy life and culture’ (Brems, Reference Brems, Foblets, Gaudreault-DesBiens and Renteln2009, p. 674).Footnote 64 Moreover, in several post-Chapman inadmissibility decisions, the Court has toned down its naturalising discourse by dropping one of the most problematic sentencesFootnote 65 and by accepting that the applicants remained Gypsies even though they had switched to a more sedentary way of life.Footnote 66
All this, however, does not necessarily mean that the problem posed by naturalising depictions of ‘the Gypsy way of life’ no longer exists. First of all, in these inadmissibility decisions the Court, at the end of the day, reaffirms the Chapman rationale by concluding that the applicants' cases did not ultimately concern ‘traditional itinerant gypsy life styles’.Footnote 67 Most importantly, Connors is a Chamber judgment and the others are inadmissibility decisions. Chapman, in contrast, is a Grand Chamber judgment, and therefore remains the authority on the matter. Moreover, Horie, a 2011 case, confirms that problematic essentialist assumptions about Gypsies are not yet fully behind us.
IV. Contrasts with the Court's broader case-law
In other areas of its cultural- and religious-practice jurisprudence, the Strasbourg Court has largely circumvented the problematic depictions discussed in the previous part. The same holds for the case-law concerning gender and sexuality traits.Footnote 68 In this part, I point to four ways in which the Court's wider discourse has mostly avoided stereotyping and naturalising pitfalls.
4.1 Rejecting unfounded generalisations
Aware of the lack of evidence in several cases, the Court has either refrained from making generalisations about the applicants' practices/traits or rejected governments' general assumptions as justifications for restrictions on their rights. Take the case of Eweida and Others v. the United Kingdom, brought by four Christian applicants not allowed to manifest their religion at work – two of them by visibly wearing a cross.Footnote 69 The case of Ms Eweida, a British Airways employee and the only one of the applicants to win the case, is especially illustrative. The airline justified the prohibition alleging the need to protect its corporate image. The Court rejects this argument: ‘There was no evidence that the wearing of other, previously authorised, items of religious clothing, such as turbans and hijabs, by other employees, had any negative impact on British Airways’ brand or image.'Footnote 70
Another example is Lautsi v. Italy, a case concerning a mother's unsuccessful attempt to have crucifixes removed from her children's state school.Footnote 71 The Court's Grand Chamber notes:
‘There is no evidence before the Court that the display of a religious symbol on classroom walls may have an influence on pupils and so it cannot reasonably be asserted that it does or does not have an effect on young persons whose convictions are still in the process of being formed.’Footnote 72
To be sure, the Lautsi and Eweida judgments also rely on characteristics inherently attributed to the symbols at issue. Thus, ‘a crucifix on a wall is an essentially passive symbol’.Footnote 73 And ‘Ms Eweida's cross was discreet and cannot have detracted from her professional appearance’.Footnote 74 The properties of these symbols (‘passive’ and ‘discreet’) are exactly the opposite of the attributes used to characterise ‘the headscarf’ (‘powerful’ and ‘ostentatious’). Moreover, whereas the headscarf's inherent characteristics in Dahlab and Leyla Şahin served to construe the symbol as a threat, the innate properties of Ms Eweida's cross and the crucifix on a wall served to minimise the threat. Either way, the fact is that, unlike in Dahlab and Leyla Şahin, in Lautsi and Eweida the Court additionally takes care to refer to (the absence of) evidence in support of its conclusions.
Jehovah's Witnesses of Moscow v. Russia is another good case in point.Footnote 75 The applicants complained about the dissolution of their religious community. The Russian government argued, among other things, that the dissolution was necessary to protect the followers' health from damages arising from refusals of blood transfusions. In rejecting the government's argument, the Court points to the lack of evidence:
‘[T]he domestic judgments did not identify any member of the applicant community whose health had been harmed or cite any forensic study assessing the extent of the harm and establishing a causal link between that harm and the activities of the applicant community.’Footnote 76
In other cases, the Court has examined governments' allegations of improper proselytism in the light of the available evidence. Indeed, unlike in Dahlab – where the headscarf proselytising effects are assumed rather than proven – in these other cases the Court makes sure to either point to evidenceFootnote 77 or its lack thereofFootnote 78 in order to accept or dismiss governments' reasons to protect others from proselytisers' pressure. The same approach surfaces in various cases concerning claims of religious discrimination in child custody and access disputes.Footnote 79 For instance, in Palau-Martinez v. France, a case in which a Jehovah's Witness mother's custody of her two children was withdrawn, the Court concludes: ‘the Court of Appeal ruled in abstracto and on the basis of general considerations.’Footnote 80
Several examples from the Court's sexual orientation case-law illustrate a similar approach. Emphasising the lack of evidence, the Court has for instance rejected governments' arguments that ‘the mere mention of homosexuality [in public]’ ‘would adversely affect children or “vulnerable adults”’Footnote 81. The Court has likewise noted ‘the lack of concrete evidence to substantiate the alleged damage to morale’ as a result of the presence of homosexuals in the armed forces.Footnote 82 Similarly, it has highlighted ‘the lack of evidence adduced by the Government in order to show that it would be detrimental to the child to be brought up by a same-sex couple’.Footnote 83
4.2 ‘Seeing’ the applicants
Contrary to Dahlab and Leyla Şahin, the Court has ‘seen’ the applicants in several cases. One example is Ahmet Arslan and Others v. Turkey.Footnote 84 At the heart of the controversy was the prosecution of members of the group ‘Aczimendi tarikatı’ for wearing their religious garment in the streets on the occasion of a religious ceremony. The Court observes: ‘[T]here is no indication in the case file that the way in which the applicants manifested their beliefs through certain clothes constituted or was likely to constitute a threat to the public order or pressure on others.’Footnote 85
Here, the Court does not look at ‘the’ black tunic, ‘the’ black turban and ‘the’ stick – the items of clothing at issue in the case – but at the specific way in which they were worn by the applicants. In Leyla Şahin, for example, and as I have shown earlier, the Court fails to see the concrete way in which the applicant manifested her religion. Her claim was precisely that the manner in which she wore her headscarf was ‘neither ostentatious nor intended as a means of protest and did not constitute a form of pressure, provocation or proselytism’.Footnote 86
In Eweida, the Court even acknowledges the importance of what was at stake for the four Christian applicants – winner and losers. For example, in the case of Ms Chaplin, a nurse who unsuccessfully sought to visibly wear a crucifix at a state hospital, the Court holds: ‘[T]he importance for the second applicant of being permitted to manifest her religion by wearing her cross visibly must weigh heavily in the balance.’Footnote 87 Again, the Court's approach in these cases contrasts with the one in Dahlab and Leyla Şahin, where the Court ignores the importance that wearing the headscarf may have had for the applicants.
In its case-law concerning transsexuals, the Court has actually condemned one state for not ‘seeing’ the applicant, more precisely, for substituting its own general assumptions for the applicant's views. In Van Kück v. Germany, a case concerning a transsexual seeking reimbursement of the expenses of a gender reassignment operation, the Court holds: ‘[T]he Court of Appeal, on the basis of general assumptions as to male and female behaviour, substituted its views on the most intimate feelings and experiences for those of the applicant, and this without any medical competence.’Footnote 88
4.3 Limiting generalisations
In a number of cases, the Court has confined generalisations of applicants' traits and experiences to particular contexts and circumstances. In Eweida, for instance, the Court does not assess the impact of ‘the cross’ – or of ‘the wearing of other items of religious clothing’ – on a corporate image in general. Rather, the Court limits the assessment to those items worn by the applicant, Ms Eweida, and ‘by other employees’ and to their impact on a particular corporate image, that of British Airways.Footnote 89
Even in Lautsi, when the symbol in question was worn by no one – it was hanging on a wall – and the levels of objectivation and generalisation were therefore higher, the Court does not just speak of ‘the crucifix’. It also speaks of ‘crucifixes in the classroom’.Footnote 90 There is therefore a significant difference of degree in the generalisations used in Dahlab and Leyla Şahin (the wearing of the headscarf) and those employed in Eweida and Lautsi (the applicant's cross and the display of crucifixes in classrooms).
Let me now briefly turn to two examples of the Court's gender case-law: Opuz v. Turkey and Rantsev v. Cyprus, as they further illustrate how the Court has kept the degree of generalisations confined to specific contexts and circumstances. Relying on extensive background data, the Court has established in Rantsev that ‘a substantial number of foreign women, particularly from the ex USSR, were being trafficked in Cyprus on artistes visas’.Footnote 91 In Opuz, based on reports and statistics, the Court concludes that the highest number of reported domestic violence victims was in Diyarbakir, Turkey, and were all women.Footnote 92 The Court does not affirm that (all) women are trafficked and exploited or that (all) women are subject to domestic violence. The point, rather, is that some women in specific contexts and circumstances are more vulnerable than others to trafficking, exploitation or domestic violence. Moreover, the affirmations are substantiated with ample material such as statistics and reports. This approach contrasts with the implicit unfounded overstatement that ‘(all) Muslim women are oppressed’ made in Dahlab and Leyla Şahin.
4.4 ‘Seeing’ social constructions
Contrary to the approach adopted in Mann Singh and Chapman, the Court has sometimes acknowledged the socially created character of the generalisation in question. The best examples seem to come from the Court's gender and sexuality jurisprudence. For instance, in several cases concerning the lack of legal recognition of post-operative transsexuals, the Court has emphasised the ‘stress’ and ‘alienation’ that ‘a post-operative transsexual’ suffers as a result of ‘a discordance between the position in society … and the status imposed by law’.Footnote 93 This discordance, the Court acknowledges, places ‘the transsexual in an anomalous position, in which he or she may experience feelings of vulnerability, humiliation and anxiety’.Footnote 94
To be sure, there is a collectivised form of representing the applicants – ‘the transsexual’ – and a generalisation of their experiences and feelings. However, the emphasis is on the legal circumstances – lack of legal recognition – that make post-operative transsexuals likely to experience such feelings. The Court does not say that transsexuals are vulnerable, humiliated or anxious. Nor does it hold that alienation is ‘at the heart of’ transsexuals' experience. The socially constructed nature of the attributes ascribed to ‘transsexuals’ is thus implicitly or explicitly recognised in the legal reasoning. In positing these traits as relational (as arising from deficits in legal arrangements) rather than as inherent in transsexuals, the Court treats these generalised experiences as contingent and revisable. A social constructivist approach may therefore eschew the immutability assumptions at the basis of the Court's reasoning in some cases of culture and religion: group membership would not necessarily be conferred by birth or by the fixed nature of certain practices/traits but may be shaped by a range of other factors.
V. In search of explanations
5.1 Negative stereotypes
One possible explanation for why negative stereotypes are most likely to be expected in cases touching upon Muslim women – in particular, when governments invoke arguments based on gender equality and the protection of the rights of others – can be found in the stereotypical images of Muslims dominating contemporary public discourses in Europe. These sorts of stereotypes are so embedded in such discourses that the Court probably does not notice that it is contributing to their perpetuation.
Indeed, there is ample material pointing to the widespread use of negative stereotypes of Muslims and Muslim women in these discourses (Group of Eminent Persons of the Council of Europe, 2011, pp. 15–16; Parliamentary Assembly of the Council of Europe, 2010, p. 1; European Commission against Racism and Intolerance, 2000). For example, Thomas Hammarberg (Reference Hammarberg2011, p. 40), former Council of Europe Commissioner for Human Rights, regrets that ‘in Europe, public discussion of female dress, and the implications of certain attire for the subjugation of women, has almost exclusively focused on what is perceived as Muslim dress’. The Court's discourse in fact meshes noticeably well with wider post-September-11 discourses. As Sherene Razack (Reference Razack2008, p. 5) notes, three kinds of stereotype have come to dominate discourses after September 11: ‘the dangerous Muslim man, the imperilled Muslim woman and the civilized European’. These kinds of representation seem to have a long history (Abu-Lughod, Reference Abu-Lughod2013, p. 6). For instance, in a genealogical study of female subjectivity in international human rights law, Dianne Otto has identified the ‘victim’ as one of the recurring female subjectivities. This subject, Otto (Reference Otto and Orford2006, p. 320) explains, embodies colonial gender narratives and is ‘created by the masculine bearer of “civilization” who rescues “native” women from “barbarian” men’.
Though Islamic rules and practices are certainly not the only victims of the Court's negative stereotypical constructions,Footnote 95 they appear to be the most frequent target. The Strasbourg Court's use of negative stereotypes in the so-called ‘headscarf’ cases seems in fact a symptom of a larger disease. The Court has portrayed other Islamic practices or rules as incompatible with gender equality in two major Grand Chamber judgments. For example, employing the exact same forms of delegitimation – authorisation and moral evaluation – the Court has stated that sharia, with its rules on the legal status of women, ‘clearly diverges from Convention values’.Footnote 96 Similarly, the Court has held in Şerife Yiğit v. Turkey – a discrimination case unsuccessfully brought by a Muslim woman denied surviving spouse benefits because she was religiously but not civilly married:
‘[T]he Court notes that in adopting the Civil Code in 1926, which instituted monogamous civil marriage as a prerequisite for any religious marriage, Turkey aimed to put an end to a marriage tradition which places women at a clear disadvantage, not to say in a situation of dependence and inferiority, compared to men.’Footnote 97
The Court accepts in these terms the legitimacy of the reason invoked by the Turkish government (protection of women) to justify the differential treatment of the applicant's religious marriage. Again, objectivation (‘a marriage tradition’) results in delegitimation (‘places women at a clear disadvantage’). This does not go unnoticed by Judge Kovler who regrets that the majority did not refrain ‘from making any assessment of the complexity of the rules of Islamic marriage, rather than portraying it in a reductive and highly subjective manner’.Footnote 98
The Court's delegitimation of Islamic marriage becomes yet more striking when compared with the judgment in Muñoz Díaz v. Spain, a discrimination case partly won by a Roma woman denied surviving spouse status for social benefits purposes due to the lack of recognition of Roma marriage.Footnote 99 In this case, the Court views the Roma community as exhibiting certain positive characteristics – own, well-established and deeply rooted values in Spanish society – which make the applicant's beliefs worth being taken into consideration in the assessment of her good faith.Footnote 100 Contrary to Islamic marriage, Roma marriage is thereby legitimised.
S.A.S. v. France,Footnote 101 a case currently pending before the Court's Grand Chamber, will be a crucial test on whether the Court is more likely to fall back on negative stereotypes when portraying Muslim women's practices. The case concerns a Muslim woman challenging the so-called ‘burqa ban’ in France. The negative stereotype of Muslim women as oppressed in need of protection has been at the heart of the debates surrounding bans on full-face veils in Europe (Grillo and Shah, Reference Grillo and Shah2012). Indeed, one of the most influential arguments in favour of these bans – usually couched in terms of gender equality – includes the view of this item of clothing as a ‘symbol of patriarchal authority and of female subservience to men’ (Grillo and Shah, Reference Grillo and Shah2012, p. 27). In S.A.S., the French government has actually made the gender equality argument in these terms. Footnote 102
S.A.S. thus offers the kind of elements that have typically led the Court to negatively stereotype Muslim women (or Islamic rules and practices concerning women). However, this time several third-party interveners have submitted empirical studies showing that many of the interviewed women wearing full-face veils in countries such as France and Belgium are not coerced by their male relatives but rather wear it out free choice.Footnote 103 There are two distinctive arguments made by third parties that may allow the Court to break out of the stereotypical constructions that pervade its ‘headscarf’ discourse: (i) ‘the ban did not materially probe the assumption that women are oppressed by wearing the full-face veil’; and (ii) some studies actually demonstrate the opposite.Footnote 104
5.2 Naturalisations
Offering a hypothesis for why the Court is most likely to portray applicants' traits as the unchanging essence of their group identity is in certain cases a more difficult venture. A partial and tentative explanation includes a mix of elements, most notably applicants' arguments, the Court's own ‘assumptions of orthodoxy’ (Beaman, Reference Beaman2012, p. 19) about groups with which it is not familiar, and the assent of those directly or indirectly involved in the case (e.g. governments, religious authorities).
To be sure, this combination partly explains the Court's use of naturalising language in Chapman and Mann Singh: the applicants' naturalised self-representations,Footnote 105 the Court's own assumptions that all Gypsies travel in caravans since birth or that all Sikhs wear the turban,Footnote 106 and the absence of dispute by the governments.Footnote 107 A look at the Court's broader case-law suggests that the lack of dispute by other parties is in fact crucial in making naturalising assumptions more likely in certain cases than in others. Indeed, in several instances where the centrality of a practice to a specific group or tradition has been questioned either from the inside (group authorities)Footnote 108 or from the outside (governments),Footnote 109 the Court has avoided assumptions of orthodoxy.
VI. Conclusion
In unpacking and challenging two major pitfalls arising from the Court's assessment of cultural and religious practices – negative stereotyping and naturalising – this paper hopes to push for a more critical use of group generalisations and categorisations in the Court's freedom of religion and right to respect for cultural lifestyle discourse.
Ironically, the Court's own case-law suggests four ways of keeping ‘a keen eye on generalizations’ (Carbado and Gulati, Reference Carbado and Gulati2003, p. 1776), that is to say, four ways of dispelling the perils of negative stereotyping and naturalising: seeing the (lack of) evidence, seeing the applicants, seeing social constructions and keeping generalisations limited. In making sure to incorporate these levels of inquiry – which can be roughly referred to as evidentiary, individual and contextual – the Court is most likely to keep generalisations ‘under control’.
A supranational court ruling in an increasingly pluralised Europe cannot delegitimise or privilege some group practices over others based on negative stereotypes or presumptions – rather than on demonstrable facts – without risking its own delegitimation.
