Relational integration: from integrating migrants to integrating social relations

ABSTRACT

The conventional notion of integration as ‘immigrants becoming part of something’ has been widely criticised for its undesirable normative connotations. In response, scholars either discard the concept altogether, or they strive for a ‘non-normative approach’. In this paper, I argue that both strategies are unsatisfactory and present a third: through ameliorative conceptual analysis, I rethink the concept such that it is useful for both the normative and analytical purposes of investigating inequalities and social boundaries that so often emerge in contexts of immigration. Building on insights from political philosophy, I argue for a conception of integration problems as a subset of relational inequality. Crucially, this framework shifts the site of the integration problem and process from ‘the immigrant’, and a process that ‘immigrants’ go through, to the relations amongst all people within a society, and a process of relational change amongst them.

KEYWORDS:

• Integration
• migration
• relational equality
• political theory
• relational sociology

1. Introduction

This paper responds to an ongoing discussion in migration studies on the use and usefulness of the integration concept. Critics argue that the portrayal of integration as a descriptive concept for how people find their ways after migration often obscures problematic normative assumptions (see Saharso 2019). The standard critique proceeds as follows: successful integration is often operationalised as the extent to which ‘immigrants’ resemble the ‘majority population’ in terms of economic or political structural outcomes, or in terms of social and cultural values. The integration concept as a ‘positive’ phenomenon – as something we consider desirable – then engenders the normative expectation that ‘immigrants’ ought to adjust to ‘non-immigrants’ after their arrival in a country, such that live up to the standards set by ‘the majority population’. These expectations are condemned for being infeasible (there is no clear standard that people could adjust towards), as well as being undesirable (for portraying ‘society’ as a prescriptive norm and ‘unintegrated immigrants’ as an aberration from this norm) (Klarenbeek 2021). Integration, understood as ‘people integrating into a social entity’, can only be achieved through the subordination of these people.

Two distinct general responses to this critique have emerged. The first is to condemn the use of integration as an analytical concept altogether. The second has been to neutralise the concept to make it ‘non-normative’ and ‘purely analytical’. Here, I propose a third strategy: building on the tradition of ameliorative conceptual engineering (Díaz-León 2020; Haslanger 2012), I aim to rethink the concept such that it is useful for the purposes of investigating inequalities and social boundaries in contexts of immigration. I present a relational conception of integration that allows us to think of integration as overcoming problematic forms of social hierarchies, without the subordinating implications that have been, rightly, criticised. On the contrary, ending subordination is a necessary condition for the attainment of, what I call, relational integration.

The framework I propose builds on normative insights from political philosophy on relational egalitarianism: I argue that we should think of integration problems as a subset of relational inequality, pertaining to inegalitarian relations amongst people which constitute social closure, social hierarchies, and generate and justify inequalities in the distribution of freedoms, resources and welfare (Anderson 1999, 312). Such inegalitarian relations are normatively wrong, and hence provide a motivation for investigating integration problems. They also provide a starting-point for the kind of phenomena that we should be discussing under the notion of ‘integration’. Crucially, this relational approach focuses on integration amongst all people, rather than the integration of some ‘into a social entity’, like conventional frameworks tend to do.¹ Such a theorisation should keep us away from undesirable normative expectations towards anyone, while also enabling us to use the concept for the investigation of the inequalities that so often follow immigration.

The article is set up as follows. Section 2 discusses the methodology of my theoretical approach. Here, I further outline my arguments for taking an ameliorative approach to the integration concept, as well as my primary focus on theorising the notion of integration problems, rather than a positive normative ideal of integration. In section 3, I introduce my understanding of the integration problem as a subset of relational inequality. I hone in on the particular relations that I consider of interest for integration problems: inegalitarian membership relations in which some people, often those designated as ‘immigrants’, are regarded as inferior members by their fellow members as well as by the institutions that regulate these membership relations. Based on this understanding of the integration problem, I make inferences about the conception of an integration process (section 4), and about integration as a normatively desirable good or outcome (section 5). I finish with some first suggestions for an alternative integration research practice, which enables us to move away from investigating the ‘integration of (categories of) individuals’, towards a focus on a dynamic, relational process that unfolds amongst members.

2. Theoretical approach

2.1. Ameliorative conceptual engineering

Here, I will outline my theoretical approach of ameliorative conceptual engineering by contrasting it to (generalised versions of) the two strategies that are most commonly used vis a vis the conflation of the descriptive and normative aspects of the integration concept. The first strategy is to reject the concept altogether as a heuristic tool, the second is to clear it of its normative content.

The first strategy, taken by opponents of the concept, pursues the critical deconstruction of its use in academic and political discourse (Favell 2019; Korteweg 2017; Rytter 2018; Schinkel 2018), generally referred to as a discourse of civic integrationism (Mouritsen and Jensen 2019). Several authors have shown how, within this discourse, ‘integration’ has become a nativist political project, used by governments for managing and subduing diversity. Civic integrationist discourses express concerns about unity, security, survival and the protection of allegedly liberal values. Civic integrationist policies aim to discipline people to protect the majority group and culture (Blankvoort et al. 2023), and ‘integration’ has become a euphemism for xenophobic and racist ideals. In short, ‘integration’ is used for the subordination of those who need to be integrated. Therefore, these authors argue, social scientists should refrain from using it as an analytical concept which can help us understand the social world.

I agree that integration research in its current forms tends to reinforce one-sided, subordinating discourses around immigration, and I want to acknowledge the important insights this critical scholarship has generated. However, I do not believe that the integration concept is intrinsically connected to subordination and is therefore, by definition, a normatively objectionable good. Instead, I argue the concept can be useful for investigating problematic forms of separation, namely social closure, social hierarchies and concurrent inequalities, oftentimes found in migration contexts. Not using the integration concept may be a firm statement against the ways it has been abused by others, but it does not solve the existence of these problems. Hence, I do not join these scholars in renouncing the concept altogether.

One could argue that we do not need the word ‘integration’ for investigating social oppositions and inequalities. To avoid the contamination of civic integrationism, we could either come up with a new word, or use a less politically loaded term. Nonetheless, I have chosen to stick to it. First, in contrast to its conventional use in migration studies, the concept leaves semantic space for analysing a relational process of mutual, reciprocal change, taking place amongst people. This relational connotation makes the word more apt than, for example, inclusion or incorporation, which invoke the image of some people needing to be included ‘in’ some pre-existing corpus or body which they are currently not a part of, making it more of a one-sided process.² In its many uses in classical sociological theory, the integration concept always refers to a reciprocal set of relationships between objects or actors (albeit in a more functionalist fashion than I wish to employ it) (Holzner 1967). In political science, ‘European integration’ refers to the reciprocal integration amongst countries of the European Union. In racial studies, the concept has been used to analyse racial relations. In this latter context, it has often been refuted as oppressive (e.g. Freixas and Abott 2019), but there have also been recent attempts to refurbish the concept along relational lines (Anderson 2010; Stanley 2017), which this framework builds on.

Second, many terms are already employed in the field, ranging from incorporation, assimilation to multiculturalism and different variations on the word ‘pluralism’, and much debate is going on about the exact differences between them (Hartmann and Gerteis 2005). Adding more words would only add to this semantic confusion, while also providing further ground for talking past each other. In this sense, my attachment to the word is also a strategic, pragmatic choice, through which I aim to nurture a dialogue with empirical scholars who work on integration and who should, hopefully, be interested in applying the relational framework. Furthermore, it should also increase its resonance outside academia, as the term remains current within wider public discourses as well. By not using the word, one risks side-lining oneself from these conversations.

Rather than accepting that problematic predispositions of civic integrationist discourse have made ‘integration’ into something undesirable, my aim is to rethink the normative foundations on which the integration concept builds. This way, we can investigate and problematise social hierarchies and structural inequalities without feeding into these problematic, subordinating discourses. This project can be seen as an attempt to reclaim the concept of integration from objectionable nativist uses. If scholars are right in emphasising that research practices are part and parcel of general integration discourses, we may as well try to improve both.

The second strategy which my approach differs from, is one taken by advocates of the integration concept. They have taken the route of ‘neutralising’ the concept – cleansing it of its normative connotations (e.g. Grzymala-Kazlowska and Phillimore 2018; Penninx and Garcés-Mascareñas 2016; Spencer and Charlsey 2021). Penninx and Garcés-Mascareñas (2016, 13), for example, advocate a ‘non-normative’, analytical approach to integration which they define as ‘the process of becoming an accepted part of society’ (14). Such an approach, they argue, does not take a stance in the normative debate on whether and to what extent who needs to adapt to whom, in order for such acceptance to come about (Penninx and Garcés-Mascareñas 2016).

Such an approach, however, is at best still normatively loaded, and at worst can have seriously undesirable normative implications. Concepts like integration provide orientations towards social phenomena and an epistemic lens to understand them (see also Díaz-León 2020). Moral and political considerations are an intrinsic part of our analytical understanding of these phenomena. ‘Integration’ is not a social phenomenon that is ‘out there’ in the social world, waiting for scholars to be analysed objectively. To take the abovementioned definition as an example: it matters why one is interested in people ‘becoming an accepted part of society’ – for example, for the ‘functioning of society’ (e.g. concerns of unity, or protection) or for ‘equality within society’. The answer to this question gives purpose to the investigation of integration, and it affects which social phenomena we consider (most) relevant for our research, as well as our evaluations about potential progress or regression in terms of integration. Presenting statistics about immigrant communities as ‘non-normative, descriptive indicators’ of integration obscures the normative assumptions which give an inherent normative direction to the measurement of an analytical concept.

A call for non-normativity not only obscures our own normative assumptions, it also leaves more space for our work to be hijacked for objectionable purposes. Seemingly objective numbers of, say, political participation, national identification, or intercultural marriage, are oftentimes presented as if they are naturally of interest for integration. This has normative implications. Presupposing that ‘immigrant political participation’, to give but an example, is an indicator or even a proxy for political integration, without further theorisation or explanation, reinforces the politically dominant idea of the integration problem of ‘passive immigrants’, who ought to be activated and participate in order to be good democratic citizens. It feeds into a political idea that the passive immigrant may be a threat to the functioning of this allegedly democratic polity. This way, ‘residual civic integrationism’ finds its way into supposedly value-neutral integration research.

Hence, we should not empty our conceptual understanding of integration of its normative content, but rather theorise it thoroughly. A proper analytical framework of integration needs to start from explicit normative foundations, such that it provides a consistent understanding of the wrongs that constitute integration problems. Rather than trying to find a neutral, descriptive definition of integration that fits some seemingly objective reality of integration problems and integration processes, I take an ameliorative conceptual approach (Haslanger 2012, 2020): I ask how we might usefully revise what we mean by integration, such that we can use it for analytical purposes, while staying away from civic integrationism.

2.2. Theorising the problem

For this project, I take a pragmatist approach, set to analyse problems as we find them in our current institutional setup of society. Rather than starting from the conceptual question of what perfect integration entails, it starts from the question of what an integration problem looks like. The framework explicitly combines normative and empirical claims about what constitutes an integration problem. That is, it aims to demarcate specific empirical situations that emerge after immigration, and formulate normative reasons for diagnosing them as a problem. Hence, this conception of integration helps us understand concrete troubles that societies are currently experiencing (see also Roth 2012).

In doing so, I start from the assumption that it is possible to identify situations that are undesirable and think about potential solutions, without knowing exactly what the ideal-type opposite of this situation would look like. Instead of theorising what we should strive for, I theorise what we should move away from. This approach follows a tradition that was set out by Sen (2010) whose aimed ‘to clarify how we can proceed to address questions of enhancing justice and removing injustice, rather than to offer resolutions of questions about the nature of perfect justice’ (ix). Similarly, Anderson (2010) and Wolff (2015) argue that conceptualising egalitarian relations as the absence of hierarchical relations provides enough conceptual foundation to both detect problems of inequality, and to think about potential solutions for these problems: ‘knowledge of the better does not require knowledge of the best’ (Anderson 2010, 3). This may seem less ambitious than providing a positive understanding of an ideally integrated society – it involves taking a step back from defining solutions to defining problems. Nonetheless, I take it to be the most appropriate strategy, for two main reasons.

First, there may not be just one positive model of an integrated society. Integrated relations could be embedded in rather different institutional setups, dealing with the demographic make-up, cultural or religious diversities, and histories of inequalities in different ways. It is therefore more viable to develop a conception of integration problems than one of the integrated society. In addition to this epistemic concern, it is also a strategic move: I expect it will be easier for scholars to agree on what constitutes relevant problems of integration, than on the setup and conditions for a positive ideal of an integrated society. It will hopefully make the framework appealing to a broader range of scholars.

Second, the way we define and frame a problem constitutes the solutions that are appropriate to solve it, since this framing informs who is responsible for the problem, what should be done about it, and by whom (Verloo 2005, 24; see also Bacchi and Goodwin 2016). Being thorough and precise in our depiction of the integration problem therefore inherently provides a foundation for thinking about solutions as well.

Providing a theoretical conception of an integration problem, this framework should be helpful, first and foremost, for determining the subject of integration research: for diagnosing integration problems in the empirical world, as well as for differentiating what kind of social phenomena are relevant for our analysis. From this conception of the problem, we can make some inferences about integration as a process (see 3.4), as well as about the characteristics of integrated relations, as the absence of the depicted problem (see 4.1). In its current form, this framework does not provide a sociological theory of how integration processes happen empirically. This would require the development of a comprehensive relational sociological framework of integration as a form of social change, which is outside the scope of this primarily conceptual endeavour.

3. The relational integration problem

This ameliorative project starts from a core set of phenomena that unite the interests of many migration scholars in integration: structural inequalities in contexts of immigration. These inequalities do not only involve socioeconomic disparities, but also extend to political issues (e.g. democratic representation) and cultural issues (e.g. religious rights) (Alba and Foner 2015; Waters and Pineau 2015). We need a conception of equality here that does not portray the integration problem as ‘immigrants needing to become equal to non-immigrants’ because they are currently ‘lagging behind’ and need to be ‘elevated’. As I have argued elsewhere (Klarenbeek 2021) such conceptions of integration reinforce a one-way understanding of the process with many undesirable implications. I explicitly distance relational integration from notions of ‘immigrants striving for proximity to non-immigrants’, in which ‘well integrated migrants’ are those who are ‘most able to thrive under institutions that are potentially very exclusive’ (Stanley 2017, 37).³ This distances the framework from civic integrationism from the outset.

3.1. Relational (in)equality

Based on theories of relational equality, I propose we think of integration problems as a subset of relational inequality, or of inegalitarian relations amongst members of a society. Unlike frameworks of distributive equality, relational equality is not primarily concerned with the equal distribution of goods but rather with the relationships in which these goods (both material and immaterial, such as opportunities, freedoms, or access to resources) are distributed (Anderson 1999; Fourie 2015; Schemmel 2012).

Relational inequality emerges where inegalitarian relations constitute superior and inferior social positions, while also providing a justification for these inequalities (Anderson 1999, 312; Tilly 2005). Within inegalitarian relations, some people are, for whatever reasons, deemed inferior to others: they are considered less worthy as moral agents, less capable as participants within a community, and less appropriate advocates of critique on, or change of these relations as they currently are. Such relations constitute all kinds of harms that can broadly be summarised under Young’s (2011) five faces of oppression: those of superior rank are thought entitled to ‘dominate, exploit, marginalise, demean and inflict violence upon others’ (Anderson 1999, 213). So, relational inequality not only establishes a foundation for structural inequality in distributive outcomes, it also limits the capacity to change these unequal and unjust relationships, and thereby has self-reinforcing consequences (see Hayward 2000; Kolodny 2014; Stahl 2017; Young 2011).

Relational equality is characterised by relations in which people enjoy social standing as equals: they acknowledge and treat each other as equals. These egalitarian relations are to be established not only in interpersonal contact, but also in formal and informal institutions (Schemmel 2012, 124). There is much discussion about the characteristics of an egalitarian relation, with some theorists focusing mainly on the political, democratic aspect of egalitarian relations (Anderson 1999) and others emphasising more social and emotional aspects (Baker 2015) What most relational egalitarians share is the conviction that relational equality is more demanding than the recognition of people’s equal moral worth, which in many egalitarian theories functions as a ‘floor of basic respect’ below which no-one should fall (Baker 2015, 69). Instead, relational equality encompasses a more positive, more demanding form of mutual respect that is referred to as social value (Lebron 2013), social honour (Weber 1946) or social esteem (Fourie 2015). This is not to say that everyone in a society should be held in the same amount of esteem, but it does require that all members have equal opportunities for acquiring esteem.

It is not my aim to provide a picture of the perfect egalitarian relationship in all its aspects. Relational egalitarians tend to define egalitarian relations as the negation of inegalitarian relations (Wolff 2015, 209). I follow this line of thinking: relational equality is characterised by the absence of structural hierarchies in terms of social esteem and power, through which (categories of) people are oppressed, attributed second-class status, and excluded from access to resources (both material and immaterial).

I propose we use the integration concept to denote problems of inegalitarian or hierarchical relations amongst people that arise in contexts of immigration, and constitute categories of superior and inferior people within a community or society. Such categories are not just mental constructs but consist of socially negotiated boundaries (Tilly 2005, 100). The more consensus there is on the social reality of these boundaries, the more they are revealed in (relatively) stable patterns of behaviour and distribution (Lamont and Molnár 2002, 168). These boundaries have very real consequences: they function as a network of relations that constrain and enable actors, ‘particularly by granting differential access to resources, opportunities, and norm-crafting and rule-making positions of authority’ (Stanley 2017, 9; see also Hayward 2000). Hence, they constitute not only unequal distributions, but also unequal capacities to change these inequalities. Relational integration is thereby not just concerned with inequality as separateness or imbalance, but with oppression and subordination (see also Stanley 2017, 35).

I end this section with some remarks on the distinction between a relational and a distributive approach to integration and inequality. Relational and distributive equality are likely to be empirically, if contingently, linked. Yet, they do not necessarily coincide. We can think of situations in which distributive outcomes are equal, while the relations in which such distributions are being made, are not. We should not, and generally do not, think of such situations as ‘integrated’. First, distributive equality could be subject to exogeneous influences, such as the general condition of an economy. Equal levels of employment may be reached temporarily in times of economic prosperity, but an economic or political crisis may change this equal distribution. If membership relations are not egalitarian, those categorised as immigrants may be the first to lose their jobs or suffer all other kinds of social harms. The equal distribution of before, did not depict integration.

Second, distributive equality could be attained through redistribution, of either opportunities or outcomes, without putting an end to the inegalitarian relations within which material and immaterial goods were distributed in the first place (Schemmel 2012). Imagine a situation in which a category of people achieves equal opportunities in terms of employment through arrangements of affirmative action. As long as their colleagues do not grant them equal social standing, there could still be oppression within the workplace without this being immediately reflected in the employment numbers. In fact, people may grant their colleagues even less social standing if they (suspect they) have been hired through affirmative action schemes, thereby creating a backlash of even deeper relational inequality (see Pierce 2003). The attainment of some state of distributive equality is not necessarily an indication that equality in terms of social status has been established. Addressing problems of distributive equality in such situations does not touch upon the core of the inequality problem but instead on the effects of distributive schemes.

My argument for a relational approach to equality and integration is not to disregard distributive equality altogether – some extent of distributive equality is likely to be a necessary condition for relational equality, and therefore also for relational integration. Also, severe distributive inequality is very likely to be an indicator of relational inequality (Phillips 2004).

Yet, too much of an emphasis on the distribution of opportunities and outcomes, may lead to unwarranted conclusions about the central problem of integration, as well as the solutions to these problems – particularly about their sufficiency for solving the issue. By drawing our attention to issues of power and status, a research focus on the relations in which distributions are taking place provides a more comprehensive outlook on the integration problem.

3.2. Inegalitarian membership relations

Any account of relational inequality needs a specification of the kind of relations that are at stake. For the context of immigration, I take the membership relations that make up a society to be of central importance.

Crucially, we should not consider society a holistic entity that can act as such, but rather as a sum of relational interactions. These membership relations generate certain mechanisms and patterns, such that ‘society’ can be relevant as a level of analysis, without reifying it as an actor or a normative entity. Similarly, membership is not conceptualised as an individual possession or characteristic, as something that one does or does not have, but expresses a relationship between individuals and a larger social entity, as well as amongst the individuals that make up this entity (see Bauböck 2017). These relations give meaning to the very notion of membership, as well as to conceptions of ‘the immigrant’ and ‘the non-immigrant’, which are only meaningful through these relations with other members and the larger social entity.

Membership relations are particularly relevant for immigration contexts since immigration, particularly in our current geopolitical system of national borders and citizenship regimes, by definition triggers a transition in terms of these relations. People who previously did not reside in a territory, did not participate in its institutions and were not subject to the rules of its polity, transition into people who do and who are. (Depending on the rules of the specific border regime, this transition potentially starts before actual immigration.) Sociologically and politically, people inherently become part of these membership relations after immigration.

Empirically, we see that the full benefits of membership are not automatically extended to newcomers, either immediately, or over time (Bloemraad et al. 2019; Bonjour and Block 2016; Sadeghi 2019). This is where the relational integration problem becomes perceptible. Relational integration problems are situated in the hierarchical relations amongst members of a society, in which some, often those categorised as ‘migrants’, do not enjoy standing as equal members. (I elaborate on this mechanism in the next section.) The inequality problem that underpins our concerns for relational integration thereby shifts from ‘the unequal position of unintegrated immigrants in society’ to ‘the inequalities between members of a society, constituted by unintegrated membership relations’.

I confine the members of a society to those who reside with some permanence in a territory, and are, consequentially, part of the same political and social institutions. They are thereby likely to develop meaningful connections with other members and with these institutions. This is in line with Carens’ (2015) theory of social membership (see also Cohen 2018 on membership and time). Hence, the category ‘members’ does not equal ‘citizens’, but instead ‘residents’. Due to the empirical prevalence of the national polity, the relevant level of analysis will often be the residents of a national territory, but the framework is not necessarily limited to national-level integration.⁴

This may be, to some, an over-inclusive understanding of membership, because it does not exclude those who are, in practice, barely engaging with their fellow members, either socially or politically. While the theory of social membership builds on social connections to other members and to institutions as the normative foundation of membership, I follow Carens (2015) in that we should not focus on the actualisation of such connections – who would be the judge of what are ‘enough’ meaningful connections to qualify as a member? It is likely that unfair biases, based on for example gender, class and race would play a role in the assessment of the ‘connectedness’ of potential members (166). Indeed, taking the social connections that people have actualised as a condition for membership, resonates with discourses of civic integrationism and the prescriptive importance it attaches to ‘society’. It invokes questions of what types of connections are considered meaningful for a connection to ‘society’ and what constitutes sufficient engagement with ‘society’. Hence, we just assume that residency functions as a proxy for some meaningful investment in a community and a polity. The potential ‘problem’ of an over-inclusive conception of membership is preferable to the alternative, in which eligibility for membership needs to be qualitatively measured.

Other critics may argue here that this framework is methodologically nationalist (Wimmer and Glick-Schiller 2003). Accepting the notion of membership, particularly when connecting it to national states and their territory, means accepting the idea of bounded national societies with an inside and an outside, which has two alleged problems. First, it is a nationalist, colonialist construction that serves to legitimate the exclusion of everyone who is not born as a ‘national’ and thereby reinforces colossal global inequalities (Shachar 2009). Second, and following this logic, any striving for bounded national equality reinforces global inequality. The conclusion would be that membership is part of the problem, rather than the solution, and cannot be squared with the principles of egalitarianism in the first place.

Both of these critiques are valid and become more salient the more exclusionary membership is. Yet, some form of bounded membership seems to be a necessary condition for any form of a democratically organised, self-governing community. While bounded membership may reinforce global inequalities, discharging all forms of membership does not seem a feasible solution to this problem. For now, I side-line the discussion on the conditions for a just form of borders and bounded membership (for an overview of the debate in political theory, see Song 2018). By doing so, I do not wish to suggest that societies could become more integrated by keeping out ‘undesirable members’. Also, this is not a normative defence of the Westphalian nation-state. Yet, the characteristics of egalitarian relations between members, potential members, and others, is a different question from the one discussed here.

3.3. Social legitimation of membership

So far, I have argued for a normative understanding of the integration problem as a problem of hierarchical membership relations in the context of immigration, constituting superior and inferior positions in society. In this section, I hone in on an important empirical mechanism for the constitution of such relational integration problems: processes of legitimation. This refers to recognition processes (Fraser and Honneth 2003) of the value of an entity (either a person, an action, or a situation), which result in a generalised perception of whether an entity is appropriate or ‘as it should be’ within some socially constructed system of norms, values, beliefs, practices and procedures (Johnson, Dowd, and Ridgeway 2006; Lamont 2012; Suchman 1995, 574; Zelditch 2001, 33).⁵ For example, what makes an object legitimate as ‘a work of art’ is socially constructed, depending on a social system through which it is attributed a certain level of legitimacy.

Similarly, we can observe that people living within a certain territory are ascribed more or less legitimacy as a member of society. Legitimate membership provides a classification scheme for hierarchies amongst members in terms of the rightfulness of their membership, which makes it a significant manifestation of a relational integration problem. To question people's legitimacy as a member means to question their worthiness of membership, their entitlement to the social advantages of this membership, the appropriateness of their participation in the decision-making procedures of the community, or to downgrade the importance of their perspectives and interests in these decision-making processes.

Let me provide some illustrations of such situations. Current research on immigration and welfare solidarity investigates the ‘anti-solidarity effect’, through which people become less supportive towards the welfare state in general when welfare benefits go to ‘immigrants’, whom they consider to be ‘undeserving recipients’ (Burgoon and Rooduijn 2020; Koning 2019). Here, some people are seen as less deserving of the welfare benefits of membership, because of their ‘migration background’. The legitimacy of their membership is questioned.

A different illustration is provided by the discussion around the candidacy of Moroccan-born Khadija Arib as Speaker of the Dutch House of Commons in 2016. Arib's aspirations were openly questioned because of her ‘migration background’, both by populist politicians and liberal commentators. A primary national newspaper (De Volkskrant) questioned whether she could be sufficiently neutral in guiding debates concerning ‘Morocco’⁶ and in the televised political talk show Buitenhof, Arib was asked whether she was endangering the status of the Speaker of the House by ‘giving rise to’ the discussion about her Moroccan background.⁷ Arib's entitlement to hold this democratically important position in the Dutch polity was questioned based on concerns around the legitimacy of her membership.

The legitimacy of membership is to be seen as a scale. For people who enjoy virtually no legitimacy, their sheer presence in a territory will be problematised, while for others, their presence may be tolerated as long as they lay low and do not ask for specific benefits or rights. It is perfectly well possible to be granted the (informal) right to be somewhere, as long as one does not expect social standing as equals. We can think of situations in which the presence of certain communities of labour immigrants is tolerated or even welcomed because they take on the jobs that no one else wants, or the presence of refugees is tolerated as proof of the benevolence and generosity of the receiving society. Yet, people may start to actively question the legitimacy of membership as soon as these people start to find better positions in the labour market, or do not express enough gratitude for being granted asylum and instead voice critiques about the ways in which governments deal with their situations.

3.4. Example: the relational integration problem of language

As I argued in section 2, our conception of a problem also directs our thinking about solutions. Let me illustrate this through the example of language in integration. Language acquisition is often used as the ultimate example of an allegedly ‘purely pragmatic’ skill that newcomers need to attain in order to overcome social boundaries within a society. People who do not speak the dominant language may, for example, not have equal opportunities of finding a job, and consequentially, are more likely to be unemployed. Similarly, political participation is impeded by language barriers. In this reading, the integration problem at hand has nothing to do with the legitimacy of membership. Unequal status, the argument goes, would then be due to the fact that some people lack the skills that are necessary for functioning as an equal member of a community. The solution to this formulation of the integration problem is one-sided: people need to acquire skills (or ‘human capital’) in order to acquire equal status. One just needs to learn ‘the language’.

From a relational point of view, this is a misrepresentation of the role of language in integration. First, because a relational perspective questions the absolute need for a majority language for the realisation of equal standing and equal opportunities in all circumstances.⁸ Not speaking a majority language does not, per definition, obstruct one's chances for a job in all contexts. In many work fields all over the world, for example, English functions as a lingua franca. As for political participation: a plethora of democratic countries have more than one official language, and multilingual institutions. The European Parliament shows that political debates can be held in several languages simultaneously. The attainment of equal democratic standing and opportunities, does not necessarily depend on a shared language (see also Benhabib 2002; Bonotti and Willoughby 2020; Wright 2015).

Second, and related to this, a relational perspective draws our attention to the power dynamics around the standards for what it means to be linguistically proficient, or to be recognised as a skilled speaker of a language (and hence to be suitable for a certain job, or to be considered capable of making comprehensible political arguments) in the first place. Realising equal standing does not depend merely on whether one speaks a language, but also on whether one is being heard. This accounts not only for immigrants’ linguistic comprehension of a language, but also for the ways in which they speak. As put by Heller in her work on language and citizenship (2013):

[O]ne can count as a legitimate speaker without being able to deploy the legitimate language (excuses will be found for you), just as one can speak the legitimate language without counting as a legitimate speaker (and so you will not be heard). (190)

Since some people may feel that they have an interest in maintaining their privileged position as a legitimate speaker over others, these standards could be, and often are, set at inappropriate levels in order to protect their position. Social norms and linguistic prejudice against ‘ethnically marked’ ways of speaking make language acquisition, again an issue of legitimate membership and hierarchy.

I do not claim that language barriers do not raise practical challenges of comprehensibility. From a relational integration perspective however, they should be regarded as exactly this: practical challenges of communication, which can be solved in various ways, depending on the context of the challenge. A relational understanding of integration does not provide any cultural or otherwise principled arguments for demanding newcomers to learn a majority language, nor for minority language rights in the name of equality. From a relational perspective, the key priority is establishing egalitarian communication between members. This requires, first of all, recognition of others as capable communicators, regardless of their language skills. Institutionally, it requires a pragmatic approach which largely coincides with what Kymlicka and Patten (2003) have called the ‘norm-and-accommodation’ approach: a likely predominance of a certain language or languages is supplemented by the possibility of language accommodation if required by the circumstances.

4. The relational integration process

Now that we have a conception of integration problems, we can start to conceptualise the notion of the process, which involves the workings of hierarchies in membership relations. Integration processes play out in the ongoing relationship between members, constituted through (institutional) interactions.⁹ Such processes can be characterised as disintegrative if people increasingly question the legitimacy of their fellow members, such that hierarchies increase. Integrative processes are those in which people increasingly recognise their fellow members as equals, such that hierarchies decrease.¹⁰

Importantly, not only do relational integration processes involve people gaining standing, but also others giving up standing – those aspects of their standing that are constituted by illegitimate claims to superiority over others – as well as privileged access to resources that is connected to this status. A critical reader may respond here that standing is not a limited good, in the sense that there is a fixed or capped amount of standing that can be distributed over members. As a following, integration could involve a general increase in standing, without some people having to give up theirs. I agree that standing is not a limited, nor a fixed good: it can be generally high or generally low in a society. It is, however, a relational good, meaning that inequalities in standing are inherently based on superiority or inferiority vis-à-vis others. For inferior members to gain standing, others will have to lose the privileges of having a superior position.

As a consequence, integration processes are very likely to involve conflict and power struggles, and may very well lead to resistance and backlash. Integration in some aspects of membership relations (e.g. more political equality) may lead to disintegration in others (e.g. social closure). Likewise, the integration between some members of a society may coincide with disintegration between others. Note that, in conventional integration research, such conflicts are mostly seen as contextual factors for the ‘integration of immigrants’. In a relational approach, they are an intrinsic part of what constitutes integration, and therefore they should be an important part of our research focus. Importantly, and in contrast to how they are often portrayed in political discourse, such conflicts are not necessarily indicative of disintegration. Overt conflicts are merely moments in which the power struggles surrounding hierarchical membership relations become visible and tangible.

This relational perspective on the integration process has implications for the subject of integration research. Below I elaborate on three of these: our attention for the recogniser in integration, the need for an intersectional research approach, and the de-migranticization of integration research.

4.1. The recogniser and the recognised

Whereas many research endeavours on integration focus on those who are ‘in need of recognition’, it is crucial that we devote more attention to the question of recognition by whom? Integration processes fundamentally involve these ‘recognisers’, but who are they? This is not an easy question to answer. Note that ‘the recogniser’ and ‘the recognised’ pertain to roles, not to distinguishable groups with attributes that are determined independently from this very relationship. Hence, they are not synonyms for ‘non-migrants’ and ‘migrants’ respectively. Within the relational process of recognition, the same people take on both roles, and we can only understand them through this transactional process, never in isolation from each other (see also Emirbayer 1997, 296).

First, the recogniser is not necessarily an individual or a group of individuals. Membership relations are constituted, reinforced, and transformed through formal and informal institutions, such as norms, policies and symbols, which cannot be reduced to the intentions or actions of individual persons.¹¹ Institutions form an important aspect of any framework of relational equality: institutions can fail in recognising people as equals by treating them with hostility, contempt or neglect (Schemmel 2012, 135). The ways in which institutions constitute (in)egalitarian membership relations would be an important focus of relational integration research.

This is not to say that individual recognisers are unimportant: institutions are inhabited by individuals, and individual interactions can change institutions (see also Lebron 2013). Ideally, integrated membership relations are constituted by all residents recognising all residents as equals. Yet, empirically, the denial of such recognition, also referred to as misrecognition, by one individual may not substantially affect relational integration.

There is no a priori theoretical answer to the question of who exactly needs to recognise a person as equal for them to have equal social standing. First and foremost, the framework provides the theoretical starting point that not recognising others as equals is a potential integration problem. An important part of the relational integration research agenda would be to analyse the ways in which the relations between institutions and individuals affect the overall hierarchies of membership within societies, and the notion of misrecognition would be a starting point for doing so.

Two main concerns would direct the conducting of such empirical investigations. First, the position of the recogniser in terms of (symbolic) power. Some people and institutions have important gate-keeping functions, which makes the impact of their (mis)recognition more significant than of others. People whose own legitimacy as a member is being questioned, on the other hand, are likely to be less influential in their recognition of others than people who are considered fully legitimate members. Second, we should be concerned with the degree to which misrecognition follows a structured pattern. Here, the notion of accumulative harm is useful (Mason 2015). Misrecognition becomes particularly problematic for relational integration when it happens on a structural basis. There is no fixed answer to the question when misrecognition stops being incidental and starts being a pattern. Numerical qualifiers, such as ‘a majority’, do not suffice since small groups of ‘misrecognisers’ may hold very powerful positions in a society through which their misrecognition could still constitute serious problems of relational integration.

4.2. Intersectionality

Hierarchies of legitimate and non-legitimate members can be based on many different characteristics, and are highly contextual and fluid. The ‘legitimate member’ is not per definition a synonym for someone who can be analytically characterised as a ‘non-immigrant’, nor is the ‘non-legitimate member’ automatically an ‘immigrant’.¹² Empirically, we see that hierarchical membership relations are often deeply intersectionally constructed, infused with, inter alia, hierarchies of class, gender and race (see also Yuval-Davis 2007). Research shows how people with a migration background are racialised and othered to different degrees (De Balibar 2007; Erel, Murji, and Nahaboo 2016; Genova 2018), how gender intersects with migration background (Korteweg 2017; Roggeband and Verloo 2007; Yuval-Davis 1993), and how class affects which people are problematised as ‘immigrants’ (Bonjour and Chauvin 2018; Bonjour and Duyvendak 2018).

This makes integration problems more complex problems than a form of mutual adjustment between ‘migrants’ and ‘non-migrants’. Integration processes involve the challenging of norms and hierarchies that may be deeply ingrained in a society, such as ideas of white supremacy, or classist notions of deserving members. Depending on the rigidity of the hierarchies in question, integration processes are very demanding for all members in a community. An intersectional perspective thereby also points out that integration problems do not merely emerge after the arrival of newcomers. When people migrate to a new country, they become part of already existing membership relations, with all their pre-existing inequalities. We should be careful not to depict integration problems as merely a temporary aberration in the functioning of polities that are otherwise democratic, liberal and inclusive, thereby implying that things will go ‘back to normal’ once newcomers have ‘found their place’ in their new societies (see also Stanley 2017, 63).

4.3. De-migranticizing integration processes

The relational integration framework aims to ‘demigranticize’ the subject of inquiry of integration studies, as has been called for by Dahinden (2016). In her seminal article, she argues that we need to question the ‘migrant’ as the primary, a priori research subject of migration studies. This may seem strange, since the distinction between ‘migration’ and ‘non-migration’, and between ‘the migrant’ and ‘the non-migrant’, are ‘the raison-d’être of migration research’ (2210). However, if researchers do not critically reflect on their use of these oppositions, if they just assume that they are inherently relevant for whatever it is they study, these oppositions take on an essentialist character: they reinforce the notion that ‘migration-given differences’ are an inevitable category of difference that are naturally of interest for social sciences. She therefore urges scholars to ‘move away from treating the migration population as the unit of analysis and instead direct the focus on part of the whole population, which obviously includes migrants (11).

A relational conception of integration questions the ‘migrant’ as the primary a priori research subject of integration studies. Relational integration processes are not situated in the ‘unintegrated immigrant’, and are not instigated by the ‘migration-related difference’ that people allegedly bring once they become part of a society. The extent to which integration problems are migration-related problems, and how they intersect with other markers for relational inequality, such as gender, race, class, would need to be part of empirical investigations, rather than a theoretical a priori assumption. Note that this is not the same as taking a difference-blind approach which would deny the experiences and structural positions that can be connected to the empirical phenomenon of migration. However, the concern for migrants in this conceptual understanding of integration is only valid if based on empirical observations that migration and migration background are salient for the foundations of relational inequality. The role of ‘migration’ in relational integration problems is taken as an empirical issue that is currently salient, not as theoretical given that is naturally salient.

5. Towards relational integration

As discussed in section 2, I define the normative good of relational integration as a negation of the integration problem. Integrated membership relations are characterised by an absence of relational integration problems, or, an absence of status hierarchies and corresponding social boundaries of membership in which some members, for example those who are categorised as ‘immigrants’, are degraded to second-class members.

When membership relations are integrated, all members are – on an abstract level – recognised as social equals and are treated accordingly. An absence of hierarchical membership relations requires that all members are seen as equally entitled to use the resources that are offered within a society and to ‘take up space’, that is, they are equally entitled to be present, visible and successful, rather than expected to lay low and to be humble. They are seen as equally deserving of making a good life for themselves and are granted equal access to the resources that are on offer for this, without, for example, receiving accusations of ‘stealing’ these from others who would be more entitled to them. On the other hand, members also get equal space to be ‘unsuccessful’. In times of need, they are seen as equally entitled to, and deserving of, the advantages of membership that are supposed to help members in disadvantaged positions, such as welfare arrangements in cases of unemployment. A ‘migration background’ would not be a motive to ask anyone, either formally or informally, to provide any extra proof that they are deserving of such benefits, nor would they need to display any extra gratitude for receiving them.

Politically, all members are seen as equally entitled to participate in the decision-making processes within a society, and to have their perspectives and interests equally weighed in these decision-making processes.¹³ As put by Bloemraad et al.:

If someone is a member of society, then society belongs to them as much as to anyone else, and the common institutions that govern that society should be as responsive to their interests and perspectives as to anyone else's. (2019, 86)

All members are seen as equally entitled to utter critique on the polity, rather than to be merely grateful for whatever goods and resources the polity grants them. This goes further than formal participation rights: members are seen as equally entitled to strive for change. On the other hand, all members are also equally allowed to not participate. To what extent political passivity is accepted in general will depend on the specific citizenship norms of a country (Dalton 2008), but ‘a migration background’ should not create any extra duty for participation and engagement (Klarenbeek and Weide 2020).

Both the social and political aspects of integrated membership relations depend, to some extent, on legal membership (often in the form of national citizenship), since legal membership generally provides social and political rights that one would not enjoy otherwise. In that sense, it is a necessary condition for integrated membership relations (de Waal 2020). Nevertheless, an extension of formal citizenship is not a sufficient condition for integrated membership relations. Bloemraad et al. (2019), for example, have recently drawn our attention to how the extension of legal citizenship to newcomers has actually coincided with a tightening of ideas of who deserves to benefit from the advantages of legal membership.

6. Towards a relational integration research practice

Revising a concept and developing a concurrent alternative research approach is an extensive academic endeavour, and this article only provides the very first step in such a direction. The main contribution of the presented framework is that it provides a normative understanding of the integration problem and should thereby help empirical scholars in differentiating relevant research questions and social phenomena for their research. Conceptualising the integration problem as a subset of relational inequality requires that we do not focus our research on the ‘integration of (categories of) individuals’, but rather on a dynamic, relational process that unfolds amongst members. This shift has substantial implications for the social phenomena that we investigate to research integration. The field traditionally takes ‘immigrant achievements’, such as language acquisition, political engagement, or employment figures, as primary indicators for integration. A relational research practice involves a shift away from this persistent bias towards ‘the immigrant’, towards the phenomena in which membership relations are manifested. When investigating relational integration, we analyse the emergence, reproduction, or transformation of hierarchical membership relations.

In order to further operationalise integration as the dynamic, interactional social phenomenon I have argued it to be, we would need to dive deeper into the ontological and epistemological assumptions of relational sociology in general, and relational integration specifically. Most scholarship on relational sociology (sometimes also dubbed process sociology) thus far has been devoted to asserting its principles over substantialist and positivist views on the social sciences (Cederman 2005; Crossley 2011; Donati 2010; Van Krieken 2001). More concrete methodological approaches for relational research need to be developed.

However, I do not think that such a relational shift would necessarily require the end of all research practices as we know them. There is a plethora of data and findings that could be reinterpreted in a more relational fashion. Existing bodies of literature provide us insights that can be highly informative of relational integration, even though authors currently do not engage with the concept themselves. Let me finish by providing a few examples that could pave the way for some first research endeavours on relational integration.

Through critical discourse and policy analysis, we can investigate how policies, discourses and norms create, reproduce, or challenge (un)equal standing between members of a society and thereby constitute problems of relational integration. By scrutinising discursive and policy categories, hierarchies in the legitimacy of membership can be exposed (see e.g. Yanow and van der Haar 2013).¹⁴ Such forms of analysis could also be conducted in the field of cultural and media studies, scrutinising categories and hierarchical relations in news outlets and cultural expressions such as TV shows, films and literature.

Social psychology could provide interesting insights in relational inequalities on an interpersonal level, by investigating phenomena such as stereotypes, beliefs and emotions (see e.g. D’Ancona 2014; Pettigrew and Meertens 1995). The German SVR, for example, has been quantitively measuring norms, perceptions, and expectations about immigrants and integration (Beigang and Wittlif 2018).¹⁵ In doing so, they make tangible how various members of society experience their interaction with others as well as their attitudes towards others. From such data too, we can gather information about social hierarchies and unequal standing.

Ethnographic research could provide us with insights in the workings of hierarchical relations and social legitimation in the social interactions between members (see e.g. Wekker 2017). Relational integration through interaction entails more than what is generally referred to as ‘social integration’, as for example defined by Miller (2016) as regularly interaction across a range of social contexts (132). Regular interactions may still be taking place in the context of hierarchical relationships (see e.g. Stanley 2017).¹⁶ Whereas spatial segregation may indeed fuel relational inequality (Anderson 2010), its opposite does not guarantee the realisation of relational equality. Ethnographic research on boundary (un)making can provide insights on the quality of interactions and their effects on hierarchies in membership relations, and hence provides a suitable research practice for the investigation of relational integration processes.

7. Conclusion

I have argued for a conception of the integration problem as a subset of relational inequality, which constitutes the normative wrong that makes integration a relevant subject for migration scholars. More specific to contexts of immigration, I have pointed to hierarchical membership relations as the central issue for the investigation of integration. Immigration is prone to trigger questions around membership, since it inherently changes the composition of membership relations as they were. Empirically, we often see that immigrants and their offspring are not acknowledged as equals in terms of membership, and, accordingly, are not extended the formal and informal rights and benefits that membership entails. Social legitimation process, i.e. the processes in which people ascribe fellow residents various degrees of legitimacy as a member, provide important mechanisms for the constitution of these hierarchical membership relations. These legitimacy processes constitute a relational integration problem: the idea that membership of a society is not a given (one either is a member or is not), but is gradual and conditional, generates a hierarchy in membership relations through which some people are seen as superior members, and some as inferior.

For relational integration to be attained, all members should enjoy social standing as equals, not because they have somehow deserved it, but because they are there. Empirically, the overcoming of underlying assumptions of what makes a legitimate member is likely to involve deeply intersectional process of change. Crucially, the framework of relational integration does not take such dynamics as contextual to the integration problem and process. It does not take discrimination based on descent, race, or class as a problem for the integration of ‘immigrants’. Instead, it conceptualises them as integration problems in and of themselves.

Acknowledgements

Thanks to Richard Alba, Floris Vermeulen, Eric Schliesser, Rainer Forst, Luara Ferracioli, Ilaria Cozzaglio, Natalie Welfens, Fenneke Wekker and Fatiha El-Hajjari for their stimulating conversations on the concept of relational integration in various stages of the development of this framework. Thanks to Enzo Rossi en Wouter Schakel for their thoughtful feedback on previous versions of the manuscript, as well as for their overall moral support. Thanks to the participants of the Normative Orders Seminar at the Goethe University Frankfurt, and the ECPR panel on Relational Equality for their thought-provoking questions. They have furthered my thinking on this concept substantially. Thanks to the anonymous reviewers for their time and their valuable comments.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 I have provided a more elaborate discussion of the ways in which migration scholars use this conventional notion of ‘integration of migrants into society’, in an earlier article (Klarenbeek 2021).

2 So, whereas some opponents have refuted integration for being an organicist term (Schinkel 2017), I would argue that a relational understanding of the concept actually leaves more semantical space to stay away from functionalist and organicist thinking than these alternatives do.

3 See also Gassan Hage’s (2000) critique on the assumptions of white supremacy in many understandings of multiculturalism.

4 From the outset, the resident category will not be clear-cut. First, residency may be temporary because people are mobile. Second, under the influence of globalization, physical residency may not always be the most important determinant of who forms a community, and who is subject to which institutions and political rules (Bauböck and Guiraudon 2009). E-government and digital citizenship (Björklund 2016) provides a situation in which our classic understanding of residency may not be immediately helpful. Further, the category comprises people in very different situations: people with legal citizenship; people with a legal status as resident but no citizenship; people residing in a country without any formal permission to do so, and probably many more variations on these themes. These constitute different forms of relational integration problems with different integration dynamics. For further discussion about the foundations of ‘just membership’, see e.g. Benhabib (2007) and Song (2018).

5 See also Haslanger’s (2014) discussion on what is ‘normatively normal’.

6 https://www.volkskrant.nl/nieuws-achtergrond/dubbele-nationaliteit-achilleshiel-van-zelfverzekerde-arib~beab467f6/.

7 https://www.groene.nl/artikel/bij-gelijke-geschiktheid-kamervoorzitter-arib.

8 Recall that a relational perspective on integration is not concerned with the integration of some people ‘into’ a society, but with the integration between members of a society. Relational integration is thereby, in no way, concerned with requirements of adaptation for the protection of a specific language, as a majority cultural right.

9 This approach is in line with the general starting point of relational sociology (Emirbayer 1997, 287). Elias illustrated the relational approach through the example of a game, which, he argued, does not just consist of players and rules that can be insulated from each other, but: ‘the changing pattern created by the players as a whole, . . . the totality of their dealings in their relationships with each other’ (1978, 130). Similarly, I argue that we should confine our interest in integration processes to insulated individuals or groups, or to specific social outcomes, but instead as a configuration of people, ‘the totality of their dealings in their membership relations’.

10 Note that this recognition element has been brought forward by, for example, Penninx and Garcés-Mascareñas (2016) and Alba and Foner (2015), but always as a condition for ‘the integration of immigrants’, thereby reducing it to an external factor See also Klarenbeek (2021).

11 See also Haslanger’s (2022) discussion on the failures of methodological individualism.

12 Potentially, the framework of relational integration could be used to investigate other forms of relational inequality beyond immigration contexts. There are many other social markers that provide foundations for oppositions between legitimate and non-legitimate members. One can think for example of homeless people, felons, or people with severe intellectual disabilities as categories designated as non-legitimate members of a society. While I focus on immigration contexts in this paper, the reader may find (aspects of) the framework to be more broadly applicable.

13 See also Fraser’s (2010) conception of ‘participatory parity’.

14 Such analysis should not be restricted to policies that are formally designated as ‘integration policies’: policies in all kinds of areas could be scrutinised for their implications for the social standing of members.

15 Although the SVR investigates both people with and without a migration background for this research, they do not take on a relational approach to integration. They use these indicators to measure the ‘integration climate’ as the environment that sets the scene for ‘the integration of immigrants’, rather than it being an aspect of integration in itself.

16 Moreover, the aim of mixing may place an extra burden on people in inferior membership positions because they may lose the safety and opportunities provided by their ‘segregated’ networks. If the socially mixed environment does not provide them the advantages of relational equality, they may therefore be worse off in mixed spaces (Stanley 2017, 167).