Swearing loyalty: Should new citizens pledge allegiance in a naturalisation oath?

Naturalisation oaths are meaningful and (mostly) effective: a rejoinder

Patti Tamara Lenard (University of Ottawa)

Oaths are typically understood to be particularly important and meaningful types of promises. In many countries, immigrants conclude their formal naturalisation journey by swearing a citizenship oath, after which they are full and equal members of a state, which in turn becomes committed to protecting their citizenship rights. In my kick-off piece I defended mandatory oaths for three reasons, all of which I argued contributed to their permissibility.

First, I suggested that although oaths are generally mandatory, they are ultimately voluntary enough to be interpreted as expressing genuine consent to respect the laws and practices of the state that is being joined. Second, I suggested that the taking of an oath is an opportunity for immigrants to publicly accept the responsibilities and obligations that typically attach to democratic citizenship, in exchange for accepting the protection that the state offers. There is an important expressive dimension to highlight as well, which is that for existing citizens this public proclamation of commitment to the new state – which concludes the formal naturalisation process – may well be a signal that incoming citizens can be trusted to take the obligations of citizenship seriously. Finally, I suggested, there is some evidence that the taking on of a new citizenship, via the swearing of an oath, is a meaningful and valuable experience. So long as the oath itself meets certain content requirements, for example by not requiring a commitment to specific religious or cultural values, and so long as the “logistics” for taking the oath are accommodating of the religious and cultural commitments of oath-takers, I argued, mandatory oaths are permissible (but not required) public policy.

My interlocutors contradicted each of these reasons. One repeated objection is that, in my original contribution, I exaggerate or misunderstand the extent to which oath-takers are consenting to taking the oath. A second objection states that mandatory oath-taking does not (and indeed cannot) generate the goods I attribute to it. A third objection worries that the content of the oath is problematically vague, such that oath-takers cannot possibly know what they have committed to do. A fourth and final objection is that the history of oaths of allegiance – as tools of exclusion and stigmatisation – cannot be shed. Throughout all of these counter-arguments, two themes return – one is about the various ways in which mandatory oath-taking for immigrants generates inequalities between naturalised and birthright citizens, and a second is about the availability (or not) of evidence in favour of or against my own claims. Among these objections, the worry that “allegiance” is problematic is most persuasive to me, and I endorse a modification of my original view – offered as suggestions by Rainer Bauböck and Geoffrey Brahm Levey – that it is better to treat the oath as an oath of affiliation rather than allegiance.

Objection 1: Mandatory oaths do not signal consent

Many of my interlocutors are not persuaded that mandatory oaths can be described as demonstrating consent. In particular, some among those who might naturalise are forced migrants, or are stateless, and taking the oath is the only option they have to gain the standard package of citizenship rights. If the oath is the only thing that stands between an individual and citizenship, then it is hardly fair to describe it as consensual. As Ashwini Vasanthakumar puts it: “mandatory requirements in high-stakes contexts have a way of obscuring, rather than expressing, individual intent.”

To properly evaluate this objection, I think it is first important to place the oath in the more general context of naturalisation. Naturalisation is a multi-step, multi-year, process, the requirements of which vary considerably by state and can include linguistic competence, evidence of sufficient income, civic integration as measured for example by participation in citizenship classes and by successfully completing citizenship tests, minimum residency requirements, and sometimes a final citizenship ceremony that includes an oath. These requirements vary in their level of difficulty – for example, in some states, immigrants must prove quite high levels of linguistic competence and in others only very low levels. Each of these naturalisation requirements can be evaluated independently for their permissibility – an earlier forum considered when and whether citizenship tests are permissible, for example. Moreover, taken as a whole, the naturalisation process is both consensual and coercive at the same time: it is consensual in the sense that immigrants have chosen to naturalise and coercive in the sense that each of the requirements must be fulfilled. So the question of whether an oath is problematically coercive is at least in part a function of the fairness of the naturalisation process itself.

Leaving the question of the naturalisation process as a whole aside, however, I do believe taking the oath is voluntary enough to signal consent. Much immigration is at least somewhat voluntary, in the sense that migrants have made choices to move where they could have made alternative decent choices. But choosing to migrate is not what makes naturalisation itself voluntary: what makes the choice to naturalise voluntary is the availability of a decent alternative, which in this case is provided by the availability of long-term resident status, described as “denizenship” by Tomas Hammar and Rainer Bauböck. While it does not, Bauböck says, “fully substitute for citizenship”, the status of legal long-term resident in liberal democratic states is sufficiently good to make the application for citizenship a genuine choice, and therefore taking the oath can be understood as a genuine choice as well. Obviously, forced migration is not a choice – although choices are certainly made even in the context of forced migration. But, having been admitted and granted domicile and the attendant rights, the option to naturalise or remain in a “permanent resident” status is sufficiently good to provide would-be oath-takers with a genuine option.

I did claim that citizenship status is valuable, which Lior Erez suggested raises a puzzle: it is either valuable, in which case there is no choice but to take the oath to gain the relevant value, or it isn’t, in which case the oath is superfluous, since it offers nothing of value to oath-takers. Solving this puzzle requires nothing more than rejecting its binary logic. Citizenship status is valuable – it generally comes with the right to vote and, more importantly, with the right to return and protection against deportation. But, to repeat, denizenship status is sufficiently good that it offers a real choice to would-be oath-takers.

Other interlocutors, including Daniel Sharp and Jelena Džankić suggested instead that the difficulty with my claim did not lie with whether the oath is consensually taken, but rather with the inequality it creates – since naturalising, but not birthright citizens, are required to consent to the authority of the state. The result is that mandatory oaths produce an objectionable inequality between citizens, based on the manner of their acquiring citizenship.

I did suggest in my original contribution that at least some actions taken by citizens by birth can be regarded as equivalent in some way to the oath-taking requested of immigrants. In particular, participation in democratic institutions (in which justifications for policies can be demanded), might offer a way to signal consent among the natural-born citizenry. As Vasanthakumar observed, the difficulty with this position is that not all citizens participate in democratic politics in the first place – and many of those who opt out are marginalised within or alienated from the mainstream democratic institutions that govern them.

I agree that participation in democratic institutions is not a perfect demonstration of consent. Before this consent-generating mechanism is dismissed entirely, however, it is critical to see that we all participate in many more democratic institutions than the narrowly political ones, including the general governing institutions in a state. For example, consider the institution of rule of law and its connection to general public benefits, including health care, welfare, education, transit, and so on. Specifically, I have argued elsewhere that one essential feature of democracies is that their citizens generally abide by the law without direct or forceful coercion; the result is that democracies can direct the resources that might otherwise be spent on enforcement and punishment towards providing shared public goods, in which all citizens can partake. The taking of these goods, in exchange for the general willingness to abide by the law, can reasonably be interpreted as a form of consent-giving to the institutional structures of the state.

Perhaps David Owen is right that “mandatory oath-taking can be defended on grounds of liberal morality, but only when applied to all citizens.” I think, however, that both birthright and incoming citizens are subject to a degree of coercion, and that it is a mistake to treat incoming citizens as though they, alone, are subject to coercion over the course of the naturalisation process, up to and including the oath. Rather, and following Oliviero Angeli’s observations, citizens by birth are subject to extensive “political education [that] is integrated into the school curriculum to promote democratic principles and civic engagement among the younger generation.” As Angeli says, we ought to think of naturalisation as a “fast-track” towards imparting the sort of political education that a state expects and requires of all citizens. So, both birthright and naturalised citizens are subject to a degree of coercion by the state. And both birthright and naturalised citizens demonstrate consent, although in different ways.

Objection 2: Citizenship oaths are ineffective or counter-productive

Above, I listed various requirements that states impose on naturalising citizens and suggested that what matters is that the naturalisation process as a whole is fair, and therefore evaluating the oath as a distinct requirement was difficult. Having said that, to evaluate each of these elements for their permissibility, as well as the entire naturalisation process adopted in specific states, we need to agree, first, on what the goal of naturalisation is, and whether that goal is permissible. One ostensible goal of naturalisation is to prepare immigrants to be successful – that is, to integrate – in the state they are joining. Here, then, the obvious question is whether each specific requirement contributes to supporting integration, and the less obvious question is whether how integration is defined in a particular state is justified. While many scholars agree that “integration” is a justifiable objective (there are notable exceptions, however, including Adrian Favell, Willem Schinkel and Janine Dahinden), there is ongoing and vigorous debate about what “integration” consists in and, more specifically, who is responsible for achieving it, and how much (if anything) incoming citizens are forced to “give up” to integrate into a particular state.

A second and more controversial objective of mandatory oaths is to build trust relations between citizens and immigrants. It is, I have long maintained (and argued at great length in my first book, Trust, Democracy and Multicultural Challenges), a matter of fact that citizens are sometimes wary of the diversity – racial, ethnic, religious – that is created by immigrants. There are lots of reasons for this wariness, and many of them are bad ones – steeped in histories of racism, colonialism, oppression, cultural and religious misunderstandings, and so on. It would be a better world if humans welcomed others, on the assumption that they are trustworthy. But it would be foolish to develop public policy on that basis. It is at least partly for the public demonstration of trustworthiness that I have defended mandatory oaths.

Naturalisation requirements kicking in earlier in the process cannot fulfill this purpose very well, since they do not contain the public and expressive element that oaths do. So, I disagree with a suggestion by Vasanthakumar, Daniel Sharp and Helen Irving that the oath seems “superfluous” or “pointless” or “redundant”, on top of the myriad other requirements of naturalisation, since these ought to be sufficient for reassuring citizens of the host state that newcomers can and should be trusted as new citizens. They do not have a public, expressive element and therefore they cannot play this trust-building role.

I have argued before that democracies’ vigour relies on trust relations among citizens – such relations underpin a shared commitment to democratic practice as a way of decision-making, and at least some willingness to redistribute resources to those in need. It is therefore normatively important that democracies sustain this trust so that they can continue to deliver these goods. The empirical question is open, certainly, and I will return to this question below. But for now, if evidence is available to demonstrate that oath-taking is a good signal to current citizens of the trustworthiness of incoming immigrants, then I think this offers a reason to support this practice, even in cases where it is adopted from a position of wariness about immigrants.

Owen offers a cautious defense of the trust-building that might well be played by oaths – he describes his additional defence as “prudential”, and a piece of “moral hypocrisy” that may be justified if it delivers on its promises. This prudential justification is, I think, congruent with my suggestion that oaths offer signals to natural-born citizens that incoming citizens are trustworthy as fellow citizens – however, I do not think this is moral hypocrisy. Democracies, imperfect as they undoubtedly are, are precious and increasingly at risk as a result of a variety of threats, from inside and outside, as Bauböck outlines. It is not merely prudential that democratic states adopt policies to sustain themselves as democracies – it is morally required.

Do oaths contribute to the building of trust relations, though? One key plank of Liav Orgad’s full-throated rejection of oaths hinges on the apparent lack of “evidence for the positive effects of oaths,” either that they serve to signal trustworthiness or that they serve to shape the identity of incoming citizens in meaningful ways (“we do not know what transformation occurs in the hearts and minds of people taking a loyalty oath”).

Although Orgad is not persuaded, I believe that the first conjecture – about the contribution of oaths to developing trust relations – finds some empirical support from the project Keith Banting outlines in his response. Banting and his colleagues defend the claim that I have also made, namely, that the citizens’ perceptions “are critical to societal support for the full inclusion of newcomers.” The evidence that he and his team have so far marshalled suggests that citizens are responsive to signals that newcomers are committed to the state, giving preliminary support for defending the oath as such a signal. As initial evidence, Banting reports feedback collected by the Canadian government in response to its proposal to continue online oath ceremonies, which began during the covid pandemic: Current citizens (the majority of respondents) expressed very strong support for public citizenship ceremonies.

Not all contributors to this forum are sympathetic to this reason – and indeed, the claim that an anti-immigrant, colonial, racist population ought to be reassured of the trustworthiness of incoming citizens is deeply problematic, as Vasanthakumar emphasises, especially where such citizens are not themselves required to prove that they too are trustworthy or that they consent to abide by the law. As Christine Hobden argues, asking immigrants to swear an oath of loyalty in a society where hostility towards immigrants is pervasive may be equally problematic. Yet, if and where the signalling role is accepted as important, it is better if the state offers incoming citizens an easy and straightforward way to send this signal, rather than requiring them to do the work of figuring out how to do so on their own, and therefore risking getting it wrong. Banting’s team demonstrates that some immigrants do get this signalling wrong, by offering evidence that existing citizens interpret some actions but not others as evidence of commitment to the new state. For example, results from an experiment they conducted show that where immigrants engage in charitable work directed at their new state, existing citizens interpret that as a commitment, whereas “charitable initiatives directed toward an immigrant’s country of origin perversely have the opposite effect.”

Maybe, though, the relevant relations of trust that must be built are not only between new and existing citizens but also between the former and their new state. Zara Gladstone and Avia Pasternak observe that one difficulty with oath-taking is that it is one-sided. Their focus is on whether and when states can legitimately ask for oaths from newcomers, highlighting several cases in which states may simply not have the standing to do so. In agreement with Hobden, they write, “the citizenship ceremony resolves around the newcomers’ oath to the state rather than the other way around.” But, I think, this is a myopic understanding of the oath-taking moment, since it immediately shifts the responsibility of the state to offering oath-takers full and complete protection of their rights. In effect, it is the saying of the oath that brings these state duties into being. As Bauböck says, and I concur, “by welcoming them on this occasion it [the state] promises to protect them [newcomers] and their rights.” Gladstone and Pasternak propose that the state also take on a reciprocal pledge of allegiance to newcomers, and I can see no reason to oppose that proposal; it may be superfluous just as is the taking of the oath itself for those who believe that the completion of all the other naturalisation requirements is sufficient to demonstrate consent to the new state, but it would have the expressive benefit that I attributed to citizenship oaths in the first place.

The second claim that I made about the value produced by oaths is that oath-taking may well be meaningful to incoming citizens. There is at least some evidence from a Norwegian study that many immigrants welcome oath-taking and the celebration that surrounds it, finding it a meaningful and important moment, and more evidence from both Denmark and Australia that a majority of oath-takers experience the same. But, as Sharp observed, the evidence is mixed, and some oath-takers are indifferent or hostile to the requirement. Vasanthakumar too suggests the possibility that the oath-taking is “counter-productive” to the goals it sets out to achieve: “compulsory celebrations that only some people must attend can be something of a buzzkill.” It might be counter-productive in two ways. One way is in generating frustration among newcomers, which as Vasanthakumar reports reflected her own experience.

Another is by creating the idea that newcomers are worthy of suspicion – something citizens might not have considered – an idea triggered by the state’s requirement that the former take an oath before being granted citizenship status. Sharp argues that oaths fundamentally cannot play the trust-building role I ascribe to them because they, fundamentally, pick out naturalising citizens as “presumptively disloyal or disobedient.” Džankić shares this view, arguing that there is an “unspoken presumption of loyalty in the idea of birthright citizenship”, such that asking immigrants to proclaim loyalty is inherently unequal, generating a “dividing line between naturalised and non-naturalised citizens.”

This presumption is particularly problematic in the increasingly anti-immigrant environment that characterises many democratic states – and moreover threatens to do the exact opposite of what I desire, because it may heighten rather than undermine suspicion directed at incoming immigrants. When Hobden describes the decision to mandate oaths in South Africa, this worry is also on her mind. As she explains, South Africa adopted oath-taking as part of its naturalisation procedure in the context of a more generalised anti-immigrant orientation. She writes, “a mandatory oath publicises that naturalised citizens have a different status within the state and so also among their fellow citizens.” However, an optional oath, as Sharp proposes, does not solve the problem. The danger here is the one that Irving identifies, namely that those who choose against it will be viewed with even more suspicion. Where an oath is available, it ought to be made mandatory for this reason.  More generally, it seems likely that whether the oath generates suspicion or alleviates it will depend on the overall environment, rather than the oath in particular. As Yossi Harpaz suggests, “whether oaths act as a positive signal of integration or a negative mark of suspicion will depend to a large extent on the political and media framing around the oath, and naturalisation more broadly.”

Objection 3: The content of a citizenship oath is too vague to be meaningful

While I had conceded in my kickoff essay that there is some fuzziness in defining what is appropriate content for a citizenship oath, I did argue that religious and ethnically specific content was impermissible. In general, the content of the oath should be limited to abiding by the law. Two separate objections arose to these claims: that all oaths are problematically vague and that demands for incoming citizens to abide by the law limits their (moral) right to engage in civil disobedience, once they are citizens.

For example, Ashley Mantha-Hollands objected that content is always vague and imprecise: “what exactly is being asked of the newcomer is unclear.” No oath-taker can reasonably understand to what specifically she is committing herself, based on most citizenship oaths that Mantha-Hollands has studied. Orgad concurs, suggesting that “clarity is required on what obligation they entail, by whom, to whom, when, and where.” Similarly, Erez writes, “it remains unclear what kind of commitment new citizens are making.” This objection seems important: if someone is promising to do something, then it ought to be clear what that something is.

I think this objection is over-stated, for two reasons, however. First, some of the apparent vagueness in oaths is only apparent. As I noted earlier, oaths are typically the final stage of the naturalisation process, and one key part of that process is learning about the obligations that citizenship carries with it. I think it is reasonable to believe that if the naturalisation process is carefully done, then oath-takers will arrive with a clear understanding of what is being asked of them. It is only if the oath is treated as though it stands alone and separated from the general naturalisation process that it will seem unclear to those who take it.

A second reason to believe that the objection is overstated has to do with the implicit assumption that all oaths or promises are, or should be, crystal clear. But that is not the case. When we take wedding vows (an example that Mantha-Hollands raises), we take them in general without knowing exactly what they will entail. When individuals take oaths to uphold the values of their professions, they again do not know precisely what they are committing to do. When friends promise, well, friendship to each other, that promise is made without specifying what exactly that entails, beyond the generally understood duties that are connected to friendship. Where oaths solidify new roles and relations, in other words, it is typically understood that what is undertaken is the set of obligations that are connected to that role, though what precisely one is called upon to do will depend significantly on context. It is truer to the role that oaths play in our lives in general to understand them as setting out a general guiding commitment, rather than a precise list of one’s duties and obligations.

This way of understanding oaths responds to another objection that a commitment to abiding by the law – the thin content that I defended in my initial article – appears, says Vasanthakumar, to deny new citizens the right to civil disobedience when it is called for. Irving concurs, arguing that “the citizen who has sworn allegiance must in principle … subordinate his or her judgement to the new state.” Sharp explains why incoming citizens – but not citizens in general – might be restricted from engaging in civil disobedience: the taking of the oath, he says, gives immigrants a weightier duty to abide by the law (including specifically unjust law).

But, if oaths are understood as offering guiding commitments, then it is open to new citizens – as it is for all citizens – to argue that respecting these commitments (let’s say, to upholding democratic institutions) requires engaging in civil disobedience against policies and proposals that undermine them. Even soldiers, who pledge near absolute obedience to their superiors, are taught that this obedience is required up until they come to believe that they are being asked to do something that is illegal or deeply immoral. In such cases, they may have a duty to disobey rather than a duty to obey. There are necessarily, as Levey asks for, “exempting clauses” that attach to the citizenship oath, as they do to all oaths. Oath-taking is therefore not “counter-democratic”, for denying incoming citizens the right to protest, as Irving suggests. Rather, I concur with Owen when he writes, “such oaths certainly do not require obedience to law when the government is abusing the trust it is charged with upholding.”

One implication of this argument is that the state to which oath-takers are committed may well be unjust – and of course will be true. All democracies are rife with injustices and inequalities, and so the question, which is the central to Goldstone and Pasternak’s response, is whether an unjust state has standing to demand oaths in the first place. Their contribution highlights specific cases, such as when Afghan refugees in the United States are asked to pledge allegiance to the state that is responsible for their emigration in the first place, for example, or where immigrants from Senegal migrate to France because of the ways in which France’s past colonial policies continue to hamper Senegal’s development. Such historically unjust relations with specific immigrants might undermine a state’s right to demand an oath as a condition of naturalisation. These two examples may be extreme cases of injustice, but all democratic states are unjust to some extent. They are nevertheless worth protecting, in a global environment in which democracy is under threat. Moreover, if clear evidence emerges that oaths do signal trustworthiness, by conveying the message that incoming citizens aim to sustain rather than overthrow the state, then exempting those who may have (admittedly reasonable) grudges against the admitting state from the oath will weaken its capacity to do this important – democratic – work.

Given that we live in a world of states, and that movement among them is permitted although controlled, states do need rules for how immigrants can gain access to the rights and protection of citizens. There is therefore a natural asymmetry between the mechanisms by which birthright citizens and immigrant citizens come to access their full package of rights. The existence of this asymmetry is not obviously problematic. It is crucial to distinguish between the manner of gaining citizenship and the rights to which citizens are entitled. As Bauböck pointed out, an asymmetry in the first does not translate automatically into an asymmetry in the second, although many respondents suggested that it does.

Any subsequent inequalities between citizens, based on how they acquired citizenship, are not built into the oath itself, but rather emerge from policies that permit differential treatment. Where denationalisation is permitted, as Irving highlights or as Hobden describes in the South African case, but only for naturalised citizens, that is an objectionable inequality which is connected, by specific states, to the manner of citizenship acquisition. But this specific policy of rendering naturalised citizens less secure in their status, to which I have objected at great length for the ways in which it formalises inequality between citizens, is not permitted or encouraged by the taking of an oath, or even by naturalisation in general. It is adopted by policymakers who are acting undemocratically by legally instantiating problematic inequality among citizens. The taking of a citizenship oath does not generate such inequality.

Objection 4: The dark history of allegiance oaths cannot be overcome

I opened my discussion of oaths with a definition of allegiance, and in so doing triggered Irving’s overview of the history of oaths of allegiance. Both Irving and Orgad emphasise the dark sides of this history, highlighting the frequency with which oaths were used to exclude and intimidate political opponents, and both express scepticism that they can be resurrected for legitimate and valuable purposes, as I and others suggest.

In my initial account, I tried to say something about the permissible content of oaths, highlighting that thinner content focused on a commitment to democratic principles and a willingness to abide by the law was more likely to be permissible than an oath that required a commitment to relatively thick religious or cultural values or uncritical, undivided loyalty. But, says Irving, in defending the former but not the latter, I have failed to understand the “point” of oaths of allegiance which, she says, are necessarily and inevitably “thick” – oaths are, she says, “content laden and coercive”, and therefore “counter-consensual” for the ways they bind new citizens to the state. Similarly, Orgad observes that even in Norway, where the oath meets the thin criteria I have outlined, there are normative difficulties – the Norwegian constitution declares, he points out, a commitment to its Christian heritage. So although the oath is not obviously problematic, the Norwegian constitution remains so.

I think Irving overstates the bond that is created between a state and a naturalising citizen, because of the oath. Naturalising citizens, like all citizens, are free to relinquish their citizenship if they so choose, and this point is one that is also made by Levey. However, I accept the broader challenge that Irving issues, and propose to abandon the concept of allegiance, referring instead to the oath of citizenship as an oath of affiliation.

Re-interpreting oaths of citizenship as focused on affiliation was proposed by both Bauböck and Levey. Levey writes, “I suggest that the oath or ceremonial statement should be limited to conveying one’s new membership and affiliation”, and this strikes me as the right way forward here – given my initial thought that one reason to defend oaths is that they are meaningful to incoming citizens. Whether they are in fact, which some of my respondents questioned, is an empirical question that has not yet been answered satisfactorily. But if they are meaningful, they should be defended because they ritualise a welcome and joyous celebration of joining a new state, of trading a commitment to uphold that state for the rights and protections it offers (on an equal basis) to all citizens.

The oath does not require or celebrate absolute subservience to a state, nor does it commit oath-takers to be more faithful than natural born citizens to respect the law, whether just or not. Rather, it celebrates an expansion of a political community, to individuals who are choosing to affiliate themselves with it. It is, says Bauböck, best understood as a “rite of passage”, which “affirms a change of membership status and proclaims inclusion in a new political community.” The emphasis in such an oath is not on allegiance, but rather on the “consensual nature of membership acquisition”, highlighting the “mutual responsibility” of both new citizens and the state towards each other. In Levey’s words, naturalisation into citizenship is “a decision and transition worthy of being publicly marked in some way.”

Conclusion: Context matters

My kickoff article defended states’ choice to mandate oaths of citizenship as a precondition for attaining citizenship status, in general.  But of course, oaths take place in the real world, in existing democracies, with all their myriad injustices and inequalities. Moreover, many democratic states are confronting significant anti-immigrant voices and are adopting oaths specifically in that context. Hobden’s vivid account of the adoption of oath-taking in South Africa, as just one part of a state-lead effort to reinforce rather than undermine xenophobia, is particularly poignant in that regard. 

But that is not the whole story. Even where states are home to anti-immigrant voices, there are occasions where the adoption of oath-taking is positive rather than negative. In an examination of the use of oaths in South Korea, Jaeeun Kim demonstrates that while oaths of allegiance were used historically in the highly problematic ways that Irving and Orgad describe, for example by the colonial Japanese state to secure the loyalty of South Korean subjects, over time the oath shifted to become an instrument of democratic legitimacy. Offering a counter to Mantha-Holland’s worry that the content of oath is necessarily sticky, the content of the oath in South Korea has changed from demanding pledges of nearly full subjugation to authority towards simply declaring a commitment to fulfilling the duties and responsibilities of citizenship.

More evidence in favour or against oath-taking’s benefit is needed, and luckily Banting‘s team and Harpaz are on the case. For now, taking the two examples of South Africa and South Korea, and the many others offered in the responses to my original defense, together suggests that context matters for whether oaths will be able to do what I argued they can do: provide moments of meaning to incoming citizens, as they adopt a new political affiliation, and signal their commitment to existing citizens, in ways that serve to build trust relations among them. These trust relations are central to the success of democratic institutions, and in this rejoinder I have emphasised the importance of protecting them – as a democratic, not merely prudential, reason to defend naturalisation oaths. Their contribution to sustaining democratic institutions may well be real, even where they are adopted from a position of wariness towards outsiders.