Cultural Injustice and Refugee Discrimination

Abstract

Many believe that it is morally impermissible to select refugees applying for resettlement on the basis of religion but morally permissible to do so on the basis of language. In this paper, I challenge this position. I argue that if we oppose selection by religion, then we should also oppose selection by language. I argue that the kind of religious selection proposed by some is demeaning because of a history of cultural injustice, which I examine through the context of colonialism. I show that this account of the wrongness of religious selection fares better than alternative views. Since language played an important role in this history of cultural injustice, I conclude that language selection is demeaning for the same reason. An upshot of my argument is that some kinds of language selection should be viewed as a form of cultural selection. Language, like religion, can be culturally laden in ways that makes it impermissible grounds for the selection of refugees applying for resettlement and indeed would-be immigrants more generally.

Introduction

In 2014, amid the wars in Syria and Iraq, there were demands from some of those living in traditionally Christian countries, such as the UK, to prioritise assisting Christian refugees. For example, Nigel Farage, leader of the UK Independence Party, claimed to ‘particularly feel for the plight of Christians who have got nowhere to go in that region’, adding that he ‘would happily, as a country, take Christian refugees’.¹ In 2017, Donald Trump stated that persecuted Christians would be given priority over other refugees seeking to enter the United States. When he was Senator for Alabama, Jeff Sessions (who would go on to serve as Attorney General in the Trump administration) declared that, as a Christian nation, the US ‘should only be accepting Christian refugees’.² Some people defend this view by arguing that Christian Syrian or Iraqi refugees are especially vulnerable to harm because of their religion.³ Others, such as Sessions, argue that Christians will be more able to assimilate into these countries.⁴

Setting aside the point about vulnerability, this kind of selection by religion strikes most – probably all – political theorists as morally wrong. Religion, like race and ethnicity, is taken to be a morally impermissible ground for selecting migrants.⁵ Yet, many of these same theorists believe that other aspects of cultural selection – that is, selection based on cultural affinity – in which states engage are morally unproblematic. One example is linguistic competence. Political theorists typically assume there is nothing wrong with the use of such a criterion.⁶ Thus, debates on cultural selection routinely treat religion as an obviously morally wrong basis for cultural selection and language requirements as a morally uncontroversial basis for such selection.

In this paper, I challenge this asymmetry, focusing on the selection of refugees for resettlement. I argue that if we oppose selection by religion, then we should also oppose selection by language. My argument is based on the idea of cultural injustice, which I examine through the context of colonialism.⁷ I explain that this cultural injustice encompassed not just religion but also language, and I argue that this historical context of cultural injustice makes both kinds of selection wrong. Thus, the paper shows that language selection is laden with cultural meaning, bringing with it the kinds of problems that afflict other more obvious cases of cultural selection, such as religious selection.⁸ To be clear, this need not be the core wrong of the cultural selection of refugees. Indeed, the material deprivation of being excluded from a country with important economic advantages or the lack of a chance to escape violence or oppression is clearly more significant. My aim is to draw attention to the ways in which language selection can be a kind of cultural selection that is, like religious selection, morally impermissible.

The paper proceeds as follows. In section one, I set out my account of the wrongness of religious selection. On my view, this kind of selection is wrong because it demeans would-be immigrants on the basis of past cultural injustice. In section two, I consider three alternative accounts, showing why my account does a better job of explaining the wrongness of religious selection. In section three, I argue that language selection is often demeaning for the same reason that religious selection is demeaning. Thus, it follows from my argument that if we oppose religious selection, then we should also oppose language selection. Language, like religion, can be culturally laden in ways that make it impermissible grounds for the selection of refugees.

Before I proceed, two clarifications are in order. The first concerns my focus on the selection of refugees and the selection of refugees for resettlement more specifically. A state may engage in religious selection when selecting among not refugees but, say, skilled migrants. This would raise the same concerns that I discuss in the paper. However, I focus on the refugee case as this is what the calls for religious selection of migrants have targeted. For this reason, I take the refugee case to be the more morally urgent one to settle. Moreover, I focus specifically on refugees applying for resettlement. There are two parts to the international refugee system. The first part is what we can call sanctuary, which is the initial access to asylum. Second, we have resettlement. According to the United Nations High Commissioner for Refugees, ‘resettlement is the transfer of refugees from an asylum country to another State, that has agreed to admit them and ultimately grant them permanent residence’.⁹ Resettlement is typically not the initial access to asylum; rather, refugees are usually resettled from the country in which they have been granted initial asylum.

I focus on the selection of refugees for resettlement because the selection of refugees for sanctuary seems much more obviously morally problematic. Any form of discrimination when admitting refugees for sanctuary is generally condemned by political theorists, since it implies that some will be denied access to asylum and safety.¹⁰ By contrast, refugees applying for resettlement are already in their country of initial asylum and so are presumably protected from the dangers in their home country.¹¹ Unlike in the case of sanctuary, we cannot object to resettlement selection on the grounds that it deprives some refugees access to asylum and safety.

Second, it is worth briefly explaining how cultural selection occurs in the process of refugee resettlement. Selecting refugees for resettlement proceeds in two stages. First, UNHCR supports the resettlement of refugees who fall under certain categories of vulnerability. Second, resettlement states consider these submissions and select refugees based on both vulnerability and ‘the refugees’ future integration or, at least, ability to cope in the host society’.¹² For example, Germany assesses the integration potential of refugees using proxies such as educational level and language skills, and Australia uses proxies such as past work experience and ability to speak English.¹³

Religious Selection

It is uncontroversial amongst political theorists that it is wrong to select refugees on the basis of religion, even granting that religion gives rise to cultural affinity. What explains this wrongness?

My claim is that religious selection is wrong because it is demeaning to would-be immigrants. This, as I later argue, also applies to language selection. But what makes a policy demeaning? I draw here on Deborah Hellman’s view, according to which whether a policy or an act is demeaning depends on the context and history of how the targeted group has been treated. This view has been applied by Désirée Lim to skill-based immigration policy.¹⁴ I show that here too it is fruitful to situate cultural selection in a broader context and to examine the message it can be said to communicate. This approach also bears some similarity to Sahar Akhtar’s analysis of ethnic and racial selection as Akhtar grounds her analysis on group status, which is determined on the basis of the wider context.¹⁵ However, I focus less on group status and more on the kind of message immigrant selection policies can be said to communicate.

As Hellman explains, different things will be demeaning in different countries with their own histories. Suppose that a school principal in the US orders black students to sit in the back of the bus and white students in the front.¹⁶ This order, Hellman argues, is demeaning because of the culture and the context in which it is given: ‘separation by race in seating on buses, trains, and so on in our culture is conventionally understood to connote inferiority so that the treatment meted out by this classification is more symbolically loaded’.¹⁷ It is demeaning for the school principal to make black students sit at the back of the bus because racial separation has been historically understood to connote the inferiority of black people.

Moreover, there does not have to be a history of a particular policy being used in demeaning ways for this policy to now count as demeaning. For example, suppose that a lecturer in a university in the US separates the auditorium based on race. This policy is demeaning even if there has not historically been a policy of specifically putting one race on the left or right side of an auditorium. It is sufficient if there is a history of racial separation more generally. This is because, again, separation by race has been historically understood to connote the inferiority of black people.

On this account, whether a policy is demeaning does not depend on the subjective feelings of the targeted individuals. It depends, rather, on context and history. While this might at first glance seem strange, it is in fact an intuitive view. Consider the following example. Suppose that Wendy starts a new job and finds that her male colleagues frequently explain to her aspects of the job about which she is very knowledgeable. Wendy is not aware of the phenomenon of ‘mansplaining’ and does not feel insulted by these incidents. Nonetheless, these incidents are demeaning because they reflect her male colleagues’ sense of superiority over women. Such ‘mansplaining’ incidents count as demeaning even if Wendy herself does not feel demeaned by them. Thus, whether a policy or an act is demeaning does not depend on whether the targeted person feels demeaned. Instead of further defending this view, however, I will examine what it can tell us about the cultural selection of refugees.¹⁸

I suggest that calls for Global North countries to prioritise Christian Syrian or Iraqi refugees on the grounds of cultural affinity are demeaning because of the long history of Global North countries denigrating the culture of countries in the Global South. Such denigration was especially apparent during colonialism, where ‘the local cultures were either completely destroyed (especially in Africa) or coercively changed (as in the case of India); in many cases, elements of the indigenous culture that were fundamental to the national identity of the colonised were marginalised in order to promote the values and customs of the colonisers’.¹⁹

Consider the following example of how colonising countries denigrated the cultures of countries in the Global South. In The Colonial Harem, Malek Alloula examines the postcards that were produced by the photographers who arrived in Algeria with the French colonial armies.²⁰ These postcards were sent back to France and were presented as snapshots of everyday moments in Algeria. In reality, these pictures were staged by the photographers. The aim of these photographs was to present Algerians as in need of ‘civilising’. For example, many photographs of families and couples included ‘reminders of backwardness’ such as ‘the rags of the “parents” and of the “child”; the trachoma of the “baby” … and generally a repugnant squalor.”’²¹ The aim was to portray Algerian culture as backward and barbaric.

There is also a history of specifically religious denigration by colonising states. Colonisers tried to supplant and control ‘native’ religions because they believed them inferior to Christianity and threats to colonial authority. For example, the British, French, and Italian regimes in African countries employed various measures to control Islam, which was viewed as a threat.²² There were policies that limited the ability of Sub-Saharan Muslim groups to connect with the centres of Islam, especially by limiting pilgrimage. The regimes also saw the Islam that was practiced in these regions as an inferior kind of Islam.²³ The religions of the colonised states, then, were also often seen as in need of controlling.

These practices of cultural injustice were especially unjust because they often persuaded the colonised that their cultures were genuinely inferior to those of the colonisers. It led to what Frantz Fanon calls cultural estrangement, where the colonised were led to believe that ‘if the settlers were to leave, they would at once fall back into barbarism, degradation and bestiality’.²⁴ The resultant sense of inferiority is why, as Kok-Chor Tan explains, Fanon believes that ‘one of the central struggles of postcolonialism is the rehabilitation and revitalization of local national cultures that have been eroded by colonialism’.²⁵

On Hellman’s view, practices are demeaning if they have historically or conventionally been understood to connote inferiority. As we have just seen, practices that elevate or prioritize the culture of countries in the Global North have historically reflected degrading views of Global South – in particular, their backwardness and their need to be civilised.²⁶ In light of this, the cultural selection, and in particular the religious selection, of refugees is demeaning as it is the kind of practice that has historically been understood to connote the inferiority of those from the Global South. In other words, it is a policy that repeats practices that have been used to denigrate those in the Global South and is, resultantly, demeaning.

The demeaningness of a policy is importantly independent of the intended aims of the policy. For example, as I mentioned above, religious selection is often justified on the grounds of integration. One might argue that such selection prevents the creation of ‘parallel societies’, ‘whose members have very little contact with those beyond their own community’.²⁷ David Miller, for example, believes that we should avoid having these parallel societies because ‘immigrant communities are likely to be less well endowed with resources of various kinds – physical, human, and social capital – and so their members’ opportunities will be diminished unless they have access to the wider networks that integration would create’.²⁸ One could make the argument that cultural selection, including religious selection, facilitates – and aims at – this valuable social integration and therefore, is not wrong.

We may grant that, in the absence of any context of cultural injustice, it might be permissible for states to engage in some cultural selection to facilitate social integration. But our past wrongs can make it impermissible for us to engage in what would otherwise be permissible behaviour. To illustrate this point, we can return to Hellman’s example of the principal who makes black students sit at the back of the bus. There is nothing inherently wrong with a principal making students sit at the back of the bus. We can imagine all sorts of reasons for the principal to do this. For example, the principal might want to split the bus by class or sports team. But when the criterion is race, it is impermissible to engage in what could otherwise be permissible behaviour because history tells us this has been used in a way that denotes inferiority. The same is true in our context: past wrongs make it wrong for states to engage in what might otherwise be permissible cultural selection.

Two components of this account of the wrongness of religious selection require elaboration. The first is whether the relevant cultural injustice must still persist for these policies to count as demeaning. Of course, in many cases we have good reason to believe that the kinds of cultural injustice under discussion have persisted.²⁹ But even if these cultural injustices have been rectified, policies that prioritise the culture of countries in the Global North over countries in the Global South could still be demeaning. This is because a practice can be demeaning even if there is only a history of problematic relations.

To see this, suppose that I have a colleague who constantly demeans me by making small, teasing ‘jokes’ about me. Assume that the wider context of their frequent repetition suggests that they are not genuine jokes – they are, rather, insulting and demeaning. Now suppose that my colleague realises what they have done and repents, and we then enjoy a healthy relationship. It seems that any return to these comments by this colleague would be demeaning, despite our current healthy relationship, because of how they treated me in the past. These new jokes could be well-intended, but they nevertheless seem demeaning because of my colleague’s past treatment. So, even if problematic dynamics have been rectified, an act or policy with these dynamics can still be demeaning.

The second component of this account of the wrongness of religious selection concerns whether it applies only between previously colonizing states and previously colonised states or between countries in the Global South and countries in the Global North more generally. I think we should prefer the second view. Religious selection connotes inferiority even if it is applied to people from a country in the Global South that has not been colonized by the receiving country in the Global North, nor any other country. To see this, we can return to the ‘mansplaining’ example. Let’s assume that the man interrupting the woman has no history of treating or viewing women as inferior. This cannot exempt him from the charge of demeaning treatment in this case; it is his membership of a group that has tended to have these problematic views of women that renders his actions demeaning.³⁰ I think a similar point applies in the context of states. That countries in the Global North have a history of demeaning the religions of people living in the Global South can make religious selection by a Global North country demeaning even if that particular country has no history of, say, colonialism or crusading.

One limitation of my argument is that it explains the wrongness only of the use of cultural selection by countries in the Global North against refugees from the Global South. It does not tell us anything about the wrongness of cultural selection in the case of, say, refugees from the Global North, or refugees from the Global South being excluded from other Global South countries. This is not to say that such cultural selection might not be wrong, but only that its wrongness is not explained by the view developed here. Nevertheless, as I argue in the next section, my account fares better than alternative views in explaining the wrongness of religious selection, even granting this limitation.

Alternative Accounts

In this section, I consider three alternative explanations of the wrongness of religious selection. While the first two focus on Trump’s Muslim ban, which applied to travellers from certain countries in general rather than involved the religious selection of refugees, they are both nevertheless relevant for our purposes.

First, consider Gillian Brock’s account of the wrongness of religious selection.³¹ On her view, a legitimate state cannot ban Muslims from entering as this undermines the state’s legitimate exercise of power. More specifically, such a ban violates two key legitimacy requirements: the internal criterion and the contribution criterion. Let us begin with the internal criterion, which posits that states must protect the human rights of their citizens. According to Brock, a Muslim ban violates this criterion because it indirectly undermines the human rights of Muslims in the United States by violating the right to freedom of religion. A ban would fuel and vindicate hostility towards Muslims, hindering their ability to freely practice their religion. We might think that selecting refugees for resettlement based on religion undermines a state’s legitimate exercise of power in exactly the same way and that this is why such selection is wrong.

One problem with this line of reasoning is that it does not locate the wrong of a Muslim ban – and in turn the exclusion of refugees by religion – in what is done to the would-be immigrants. Rather, it holds that excluding would-be immigrants by race or ethnicity wrongs members of the targeted ethnicity or race within the state.³² This seems especially odd given that it is the would-be immigrants who are the targets of the ban. It is compatible with my account that such policies also wrong their citizens in this way. But this wrong seems best understood as a concomitant or secondary wrong of religious selection. The primary wrong is the wrong done to those who are the target of the policy, as identified by my account.³³

We can also imagine variations of these cases that do not violate Brock’s internal legitimacy criterion but should still strike us as impermissible. To use Adam Hosein’s example, imagine that these policies are explicitly framed as a problem with foreign Muslims; that a state excludes refugees who are Muslim but makes it very clear that this is because there are ‘some issues with Muslims out there, beyond the border, not Muslims within the country’.³⁴ Suppose that at the same time Muslim citizens do not face discrimination within the state. In this case, it is at the very least unclear whether excluding Muslim refugees will have the effects on the right to freedom of religion that Brock describes. And yet such discrimination still seems impermissible.

Now consider Brock’s contribution criterion. This posits that ‘states must […] contribute appropriately to a justified state system, one capable of sustaining a robust human rights practice’.³⁵ To satisfy this criterion, states must meet several requirements. The ethos requirement holds that states must have ‘a commitment to maintain an ethos conducive to respect for the practice of human rights’.³⁶ According to Brock, central to the practice of human rights is the idea that every person must be treated with equal respect. Brock argues that the Muslim ban clashes with this idea as it fails to treat ‘each person as an individual whose claims deserve fair consideration, in attending closely to the reasons why that person seeks to be admitted and their particular personal characteristics’ (Brock, 2020, p. 83–84).

But it’s not clear that states do fail to treat people as equals in a morally problematic way when they select immigrants on the basis of group membership rather than personal characteristics. Consider freedom of movement within the European Union. This is based on citizenship. For example, an Australian would not be able to travel freely to, say, Spain, whereas an Italian would be able to. This system does treat immigrants on the basis of nationality rather than personal characteristics – it is only nationals of countries in the European Union that have the right to travel freely – but this is not obviously a failure to treat, say, Australians as equals in any morally problematic way. Thus, it is unclear why we should think that the mere fact that Trump’s Muslim ban does not pay attention to personal characteristics amounts to a failure to treat would-be immigrants as equals.³⁷

Adam Hosein offers two objections to the Muslim ban. The first is that refusing to admit Muslims gives American Muslims a second-class status within the United States. Hosein cites Justice Sonia Sotomayor’s dissenting view of the Muslim ban, in which she claims that the ban makes Muslim Americans ‘outsiders, less favored members of the political community’.³⁸ However, this has the same problems that I identified in Brock’s account: intuitively, the wrongness of religious selection is located in what is done to the excluded would-be immigrants themselves.

Hosein’s second objection does locate the wrong in what is done to the excluded would-be immigrants. According to Hosein, we ought to avoid a ‘global order in which some people and nations stand above each other’.³⁹ The United States already ‘stands above’ Muslim-majority countries such as Iraq, Pakistan, Syria, and so on. For example, ‘it influences elections, builds military bases designed to change the balance of power in the region, and so on’.⁴⁰ A Muslim ban – which involves ‘the humiliation of foreign powers and peoples’ – exacerbates these problematic global relations. Again, we can apply this account to the religious selection of refugees applying for resettlement. Here too, religious selection can exacerbate these problematic global relations.

The main problem with this argument is that it rests heavily on the position of the United States as a global superpower. It is because the US has significant influence in these countries that a Muslim ban exacerbates global relations. This argument, however, is much less plausible if, for example, Norway decided to implement a Muslim ban. In this case, we cannot say that a Muslim ban is morally problematic on the grounds that Norway ‘stands above’ these countries since, unlike the US, it does not extensively influence various institutions in Muslim-majority countries. Thus, Hosein’s argument for the wrongness of religious selection is too limited. While I think Hosein is right to examine the broader context in his assessment of these policies, the focus only on the status of the country is too narrow.

Finally, consider David Miller’s rival account.⁴¹ Miller argues that the religious selection of refugees is wrong because it is unfair to disadvantage individuals on the basis of statistical facts about the groups to which they belong. Likewise, he suggests that the cultural selection of refugees is wrong because individual refugees ‘might be eager to throw off their inherited culture or adapt it creatively so as to fit more easily into the society they are joining’ even if members of their group are statistically reluctant to do so⁴² Thus, insofar as states would be engaging in this kind of statistical discrimination by prioritising Christian refugees over Muslim refugees, on the grounds that the former group is more likely to integrate into the receiving state, such prioritisation is wrong.

I’ll grant Miller’s assumption that the likelihood of assimilation is a legitimate criterion for selection. However, it is not always wrong to disadvantage individuals on the basis of statistical facts about a group to which they belong. Consider travel insurance for emergency medical expenses. People over 80 years old usually have to pay more for insurance because they are statistically more likely to make a claim. A very fit 85-year-old might be less likely to make a claim than a less fit 75-year-old; their higher premium reflects facts about others (somewhat) like them, rather than facts about them. It might be unfair to burden the very fit 85-year-old with a higher premium on account of what others their age are like, but such an insurance system nonetheless seems morally permissible. If so, then the argument that selecting refugees by religion is wrong because it disadvantages individuals on the basis of statistical facts is unconvincing: it is not always wrong to rely on statistical facts in this way.⁴³

However, we might think that states have to meet a higher standard of justification than, say, insurance agents if they are to distinguish between refugees. I think one reason why we might have this intuition is that these kinds of decisions that state make are high stakes, which might make the use of statistical information unfair. Indeed, the travel insurance example loses much of its force if we increase the relevant burden. For example, it would be wrong for a country to exclude any would-be visitors over 80 on the grounds that members of that age group often fall ill. This would impose a very high, uncompensated burden because of what others are like. In such a high-stakes case, it is plausible that this policy would be wrong on fairness grounds. The burden imposed in cases of cultural selection of refugees for sanctuary and resettlement is very high. In the resettlement case, which is the case I am focusing on, being forced to stay in a refugee camp on a long-term basis might well count as a human rights violation, since it conflicts with human rights such as those to bodily safety or nutrition. Thus, we might revise Miller’s account to say that a policy is wrong if it unfairly imposes very high, uncompensated burdens on individuals because of what others have done.

So revised, Miller’s account could explain the wrongness of cultural selection in high-stakes resettlement cases. However, its explanatory power is limited to only those high-stakes cases. My account, in contrast, applies even in low-stakes cases. My argument is that cultural selection of refugees is demeaning as a result of past cultural injustice. This wrong obtains regardless of whether excluding a refugee would impose a high cost. My account therefore has greater explanatory power than even a revised version of Miller’s account.

Language Selection

I have offered an account of why the widely condemned practice of selecting refugees by religion is wrong. It is wrong because it is demeaning, in virtue of practices that have historically connoted the inferiority of those in the Global South. I will now argue that this account also shows that the widely endorsed practice of using language as a selection criterion is also wrong.

The treatment of native languages was a significant part of the cultural injustice of colonialism. Fanon makes this point at length in Black Skin, White Masks, drawing upon his experience of growing up in Martinique. Here, the African ‘will be proportionately whiter […] in direct ratio to his mastery of the French language’. Language reflected the ways in which ‘the colonized is elevated above his jungle status in proportion to his adoption of the mother country’s cultural standards’.⁴⁴

There are countless examples of how the language of the colonisers was imposed on the colonised. For example, under British colonial rule, Indians had to learn English in order to apply for more prestigious jobs and positions. This was part of what Lord Macaulay, then-president of the Council on Education in India, called a ‘civilising mission’ that aimed to transform Indians into ‘Indian in blood and colour, but English in taste, in opinions, in morals, and in intellect’.⁴⁵ There are similar examples in African countries. For example, when Tanzania was a British colony, ‘the English language […] was made the main prerequisite for the acquisition of formal education at post-primary level and for employment in white-collar jobs’.⁴⁶ European languages became the official languages in many African countries when they were colonized by European countries.

I argued in section two that the uncontroversial wrongness of states selecting refugees by religion arises because such selection is demeaning against a backdrop of historical cultural injustice. The cultural injustice of colonialism included denigrating the language of the colonised populations and elevating the language of the colonisers, just as it involved the denigrating of religion. The same history and context that make prioritising Christian Syrian or Iraqi refugees demeaning also makes, say, an English requirement demeaning. Language selection is thus the kind of practice that has been historically understood to connote the inferiority of those from the Global South. If we accept my argument for the wrongness of religious selection, then we should accept my claim that using language as a selection criterion is also wrong.

Let me address two objections. The first is that a language criterion might in fact end up advantaging refugees who come from countries that were previously colonized – for example, since many Indians speak English because of British colonialism, requiring proficiency in English will give Indian refugees priority for resettlement. We might think that rather than demeaning refugees, this is a morally good outcome, and perhaps something that former colonial powers owe to those living in former colonies.

Some forms of cultural selection in immigration policy have in fact advantaged immigrants from countries that were formerly colonized.⁴⁷ For example, consider Canada’s points-based immigration system that awards points, amongst other things, for knowledge of English and French. Yasmeen Abu-Laban argues that the French criterion advantages immigrants who comes from countries that were colonised by France such as Haiti, Lebanon, and Senegal.⁴⁸ Moreover, David Scott Fitz-Gerald and David Cook-Martin find that the English-language criteria advantages immigrants from countries formerly colonised by Britain.⁴⁹

However, a practice or policy can be demeaning even if it prioritises those it demeans. For example, suppose an employer decides to hire only women because they believe that women are more obedient and submissive. This is clearly a demeaning practice – it treats women in a way that is conventionally understood to connote inferiority – and the fact that this policy prioritises and advantages women does not make it any less demeaning.⁵⁰ So, the claim that cultural selection prioritises refugees from previously colonised states does not show that such policies are not demeaning.

The second potential challenge is that this account proves too much. To see this, think of an English-speaking university in the Global North that makes English an admissions requirement. Many universities in the Global North have played a significant role in and have benefitted from colonialism.⁵¹ It seems to follow from my argument that it is demeaning for these universities to now require English language skills from applicants from the Global South as this also occurs against a backdrop of historical cultural injustice. This is a problematic implication as there does not seem to be anything wrong with this admission criterion, and we certainly do not think of it as demeaning.

But there are important differences between this case and mine. To start with, for these two cases to be alike, there must be a history of universities – rather than countries as we have been examining – denigrating the cultures of those in the Global South. If there is such a history, then this should matter to the university’s admission procedures. This history might make it wrong to engage in what would otherwise be permissible admission procedures. However, there are always competing factors at play. In this case, it is hard to see how a university could function without a language requirement and, unlike with long-term resettlement, time constraints arising from the length of degree courses might make it impractical for someone to learn a new language in situ in order to undertake their degree. There will therefore sometimes be genuine feasibility considerations that do outweigh the demeaning nature of a policy. However, the university might have to implement other policies in order to offset this wrong, such as scholarships for students from these countries or a somewhat lower language requirement coupled with free language courses upon arrival.

However, it is also important to note the ways in which it is not overinclusive. At first glance, it might appear that my argument implies that the main alternative to language requirements at the point of entry – namely, requiring refugees to take language classes post-arrival – is also demeaning.⁵² It seems that even citizenship tests, which typically include a language test, might be demeaning on the grounds that the state elevates its own culture above that of the immigrant and this, as we have seen, has been historically understood to connote inferiority.

But it does not necessarily follow from my argument that these practices are demeaning. Not all policies that include elements of the culture of the receiving state will be demeaning. For example, in the case of language classes post-arrival, this seems to depend on whether it is compulsory for newly arrived refugees to attend these classes or whether these classes are merely offered to refugees. It seems that only in cases where these classes are compulsory that there is the concern that cultural practices are being imposed in ways that might track cultural injustice of the past. Whether the language component of citizenship tests is demeaning also seems to depend on background factors. For example, it might depend on whether the immigrants were also provided with language classes for their children to enable them to learn the language of their country of origin. Whereas a language requirement pre-entry is imposed on the would-be immigrants, these other practices need not be. This makes an important difference as to whether the policies can be said to impose the new culture in a way that would track the cultural injustices of the past.

Conclusion

Many believe that it is morally impermissible to select refugees on the basis of religion but morally permissible to do so on the basis of language. I have argued that this position is untenable. Language selection is wrong for the same reason that religious selection is wrong: it is demeaning in light of past cultural injustice. This need not be the core wrong of this selection, but it is an important wrong nonetheless. The aim of the paper has been to show that some kinds of language selection should be viewed as a form of cultural selection. I argued that language, like religion, can be culturally laden in ways that makes it impermissible grounds for the selection of refugees applying for resettlement. While I focused on the case of refugees applying for resettlement, this analysis also applies to would-be immigrants more broadly. Language should not be viewed as wholly neutral grounds for selecting among would-be immigrants.