Blog post by Janna Wessels, Assistant Professor, Amsterdam Centre for Migration & Refugee Law (ACMRL), VU University Amsterdam. This blog post is part of the blog series on feminist theory in legal displacement research.


When one types “feminist theory book” into a search engine, it becomes immediately clear that feminist theory encompasses a wide field of very different approaches, ideas and perspectives. This breadth and complexity is not much reduced if the search term is specified as “feminist legal theory”. The fact that there are multitudinous discussions on how the various ideas interrelate, including fundamental questions such as, for example, whether and how queer theory is part of the family of feminist theories, can seem overwhelming for anyone interested in using a feminist theoretical, conceptual or methodological approach in their work. However, the rich body of theories and angles that feminism provides is also very much an opportunity. Certain approaches may have explanatory value for some specific questions, whereas others are suitable for theorising others.

This blog is part of the series on feminist theory in legal displacement research, which showcases a variety of ways in which feminist theories in the broad sense can shed light on legal research surrounding forced migration. In this particular piece, I want to tell the story of how adopting a particular queer theory lens helped me see certain difficulties surrounding the conceptualisation of the Convention grounds in the refugee definition. I submit that the queer lens makes these challenges visible, and that recognising them enables us to – at least try to – use them productively. In this sense, the theme of the title, “to see or not to see”, applies in a dual manner to this piece. First, it asks how decision-makers “see” (gay) asylum claimants. Secondly, it asks what we “see” when using queer theory as a lens to analyse refugee law doctrine relating to the Convention grounds.

Can asylum seekers be expected to hide their persecuted characteristic?

The refugee definition is laid down in Art 1(A)2 of the Refugee Convention. It states that:

‘[T]he term “refugee” shall apply to any person who … owing to wellfounded fear of being persecuted for reasons of race, religion, nationality, membership of a particular social group or political opinion, is outside the country of his nationality and is unable or, owing to such fear, is unwilling to avail himself of the protection of that country.’

So, to be a refugee, a person has to have well-founded fear of being persecuted for one of the five listed grounds. The problem that arises is that the reasons for persecution (the ‘Convention grounds’) are not necessarily immediately visible. Political opinion, religion, ethnicity in some cases, need to be expressed in one way or another to become discernible to the persecutor (and of course, the persecutor can also wrongly impute those characteristics to an individual). This also goes for certain ‘particular social groups’, and has been vigorously discussed for sexual orientation. 

Any person has at least some discretion regarding what others know – or might assume – about such characteristics. This creates a dilemma: refugee status determination is based on a future-focused analysis, but claimants can influence that future to some extent. Does this mean that claimants can be expected to hide their persecuted characteristics? If not, can claimants at the very least be required to exercise some restraint in order to avoid a threat of persecution?

To illustrate, consider the following two cases. In the case of Krystych v Canada, the claimant’s long-lasting marriages with women were held against him, casting doubts on the veracity of his contention he had engaged in same-sex acts. Conversely, in the Australian case of SZJSL v Minister for Immigration & Anor, a man who had sex with another man in male-only detention applied for asylum. Here, the asylum decision-makers found that the claimant’s acts were due to the context where only men were available and did not properly indicate his sexual orientation. They doubted that he was “actually” gay. 

In both scenarios, the Convention ground (in this case, sexual orientation, as a particular social group) are not established – because of an absence of same-sex acts in the first case, and an absence of gay identity in the second case. The outcome is the same for both, to the extent that the asylum seeker would be denied protection and returned to their country of origin where they would be compelled to hide their sexual orientation from the persecutor. To put it differently, the persecutory environment functions like a forced return to the closet. And this is obviously something queer theory has a lot to say about.

The act/identity dichotomy in queer theory

To better understand these dynamics of inclusion and exclusion based on act and identity reasoning, I turned to queer theory. Indeed, a distinction between acts and identity is at the very core of the conceptualisation of (homo-)sexuality by Michel Foucault. It is connected to his work on language, power and truth and the notion of “discursive formation”. According to Foucault, because the world is constructed socially, its character is not predetermined or pregiven. But once a dominant discourse is formed, it provides parameters of true and false – it simultaneously produces and constrains reality. Michel Foucault explains the invention of homosexuality as one such discursive formation, in a famous passage from The History of Sexuality:

‘Homosexuality appeared as one of the forms of sexuality when it was transposed from the practice of sodomy onto a kind of interior androgyny, a hermaphrodism of the soul. The sodomite had been a temporary aberration; the homosexual was now a species.’

A set of previously unrelated acts becomes a function of an identity, defines the personality structure of ‘the homosexual’ to such an extent that it marks them even if and when they do not engage in sexual acts.

However, as other queer theorists such as Eve Kosofsky Sedgwick, and Janet Halley have later clarified, this does not mean that acts are “transformed” into identity. Acts do not lose their significance. Instead, both discourses remain in place. Sexual orientation is constructed as the coexistence of two simultaneous categories – acts and identities, both mutually constitutive elements of the characteristic. This binarism of acts and identities can function as a double bind. Acts and identity are matched in an unstable relationship, where one can be preferred over the other, but the preferred term in fact depends for its meaning on the submerged one.

Double bind (damned if you do and damned if you don’t)

In refugee status determinations, this act/identity double bind develops a potent incoherence. It creates an ‘anomalous legal situation’ that allows shifting from the regime of acts, which may be allowed or prohibited, to the regime of identity, which may or may not exist, irrespective of any particular acts. This self-contradictory definition has potential for powerful manipulation: depending on context and circumstances one can be prioritised as relevant, whereas the other is submerged and deemed irrelevant. But that relationship can be flipped – the master of the double bind always has somewhere to go.

This is exactly what we saw in the examples discussed above: for the purpose of the refugee definition, the Convention ground is split into an identity and an act element. Depending on context, one or the other will be preferred in a given context. As a result, claimants find themselves in a catch-22 situation, because they can be damned if they commit particular acts (since that does not conclusively indicate their gay identity, and hence the characteristic protected by the Convention ground), and damned if they don’t commit such acts (since that also casts doubt on their sexual orientation and the ways in which it may put them at risk of persecution). The unstable relationship between act and identity creates confusion about what is persecuted in ways that create opportunities for the restriction of the scope of protection. This split and the consequent shifts occur in claims based on sexual orientation, but are not limited to this ground: They have also been observed in cases based on political opinion and religion.

What does the act/identity double bind mean for conceptualising the Convention grounds?

The situation becomes even more complex when we look at it in the wider context of the refugee definition, and specifically, in relation to the persecutory harm feared. According to the refugee definition, a refugee is someone who has well-founded fear of persecution for one of the five Convention grounds. It appears that persecution emanates from the persecutor, and the Convention ground is a characteristic that lies with the claimant. However, it is not that simple. Both the persecutor and the claimants themselves can attribute the characteristic. In that sense, group membership can be established in two ways, either via the attribution of the persecutor or via ‘actual’ membership of the claimant.

Because the claimant has at least notionally some discretion over what others know about their characteristic (i.e., the Convention ground), they can influence the prospect of persecution to some degree. They can (try to) hide from the persecutor, by adjusting their behaviour. On the other hand, no matter whether someone tries to hide, the persecutors will only persecute those individuals that they think have the characteristic, and they may actually be wrong in making that assumption. Refugee law is equipped for this kind of situation, and provides for protection where someone is only “imputed” with a Convention ground, be that political opinion, sexual orientation or religion. But it is very difficult to predict on what basis the persecutor might come to the conclusion that an individual is in fact gay, or of a particular religion etc.

It is here that it becomes complex at the level of principle: does “actual” group membership then even matter at all? Refugee status determination procedures go to great lengths to establish whether someone is actually gay. This has led to various attempts to adduce evidence encroaching on the dignity of the individuals involved, from using psychiatric reports, to presenting videos of sexual acts, to ‘phallometric tests’ (though the most egregious of these measures have thankfully been banned). But the question here is: why does this assessment matter at all if it is in fact entirely the persecutor who decides what, and therefore who, is persecuted? And if it is not about the question of whether the person actually “is” gay – something generally impossible to prove – is it then about their acts, how they will behave, and so, how they may become “visible” to the persecutor, even in if they do not self-identify with the group encompassed by the Convention ground?

This solution also seems unworkable. If what counts is the persecutor’s establishment of a link between the claimant and the persecuted group, irrespective of ‘actual’ group membership, anyone could potentially be imputed to be a group member. That not only seems like a very uncertain assessment, which involves guessing the indicators or signifiers that might lead a potential persecutor to think that someone is gay, but is also unworkable. If no connection is required between the claimant and the Convention ground at all, the future-focused analysis that refugee law requires becomes impossible. If any person in a particular country could potentially be targeted because the persecutor might think they are gay, the whole system breaks down. There must be something that links the asylum seeker and the Convention ground. This link can either be established through actual membership or through membership having already been imputed – where each of these approaches leads to differently constituted protected groups.

Establishing the Convention grounds in light of the act/identity dichotomy

The table below illustrates the different ways of establishing the Convention grounds – either through the focus on the persecutor and ‘perceived’ membership of the group, or through the focus on the claimant (‘actual’ membership) – and the ways in which they interact.

The case is relatively straightforward for those situations where ‘actual’ and ‘perceived’ membership overlaps, i.e., for those claimants who self-identify as gay and have been outed (box D). In this instance, claimants who possess the persecuted characteristic, and where this has already come to the knowledge of the persecutor, constitute largely uncomplicated cases. This type of claims often involves experiences of past persecution.

But the insights from the above analysis become relevant for the more complicated cases classified in Boxes B and C. A claimant who does not self-identify with the persecuted characteristic that has been imputed to them (Box C), would be excluded from protection if the decision-maker adopted an interpretation of the Convention ground that prioritises the identity element and ‘actual’ membership. In this scenario, the claimant could try to shift the discourse and argue that it is the persecutor who decides who is persecuted, regardless of their ‘actual’ identity.  In contrast, a claimant who does self-identity although this has not yet come to the attention of the persecutor (Box B), would be excluded from protection if the act element and ‘perceived’ membership are prioritised by the decision-maker. The best strategy for such a claimant could be to flip the discourse back by emphasising the notion that the Convention protects those characteristics that trigger persecution, and to give them up would undermine the purpose of refugee protection. Finally, claimants who do not self-identify and who have not been imputed with a persecuted characteristic (Box A) constitute the most difficult cases, and would, for the most part, not be protected. There must be something. However, if such a claimant does have reasons to fear possible future imputation based on a context where it is likely that the persecutor might assume that they are members of the persecuted group because of certain behaviour or other indicators that could lead them to presume this, the most promising strategy for their claim is to rely on the persecutor’s perspective.

The inherent tension that the act/identity inscribes into the Convention grounds in the context of the refugee definition means that both approaches (prioritising the act or prioritising the identity) are legally correct interpretations – regardless of the fact that they lead to vastly differently constituted groups (i.e., Convention grounds). This has consequences for differently situated claimants, as described above: Depending on whether they self-identify with the characteristic, or whether it is imputed to them, whether they subscribe to a gay identity, or whether they engage in same-sex behaviour, the way the decision-makers deal with the tension in any given case can mean a wide scope of protection or a very narrow one for them. Ultimately, this produces inconsistency in decision-making. However, this inconsistency cannot be resolved once and for all, by generally adopting one or the other approach. The best way forward for individual claimants is to attempt to use the in-built instability productively in favour of their particular situation. However, this is never a safe strategy. Precisely because of the instabilities inherent in act/identity dichotomy, the discourse always remains vulnerable to being flipped back to limit rather than promote protection.

The power of seeing refugee law through a feminist lens

Queer theory conceptualisations of the dichotomy between identity and acts proves to be a powerful explanatory tool to explore the question of whether and to what extent refugees can be expected to hide their persecuted characteristics from their persecutors. This approach not only sheds light on the ways in which decision-makers “see” gay asylum seekers – shifting between requirements of identity and acts –, but also uncovers possible ways of dealing with the tensions that drive such reasoning. Moreover, and importantly, these findings are not relevant ‘only’ for sexual orientation cases. Rather, they similarly apply to other Convention grounds such as religion or political opinion. In that sense, the queer theory lens helps break new ground for refugee protection overall: an awareness of the dynamics at play when conceptualising the Convention grounds in refugee law doctrine also points to the levers that may be shifted in individual cases. This example illustrates in a very concrete manner the power of feminist theories to help us see the unseen. Looking at refugee law through a feminist lens, using the wide range of approaches and perspectives that it can offer to address specific questions and problems, can lead to discoveries insights that help improve the system for everyone.

For a fuller version of the argument presented above, including detailed case law analyses of various jurisdictions, see: Janna Wessels, ‘The Concealment Controversy – Sexual Orientation, ‘Discretion’ Reasoning and the Scope of Refugee Protection’ (CUP 2021).


The views expressed in this article belong to the author/s and do not necessarily reflect those of the Refugee Law Initiative. We welcome comments and contributions to this blog – please comment below and see here for contribution guidelines.