Othering in sport-related research: How research produces and reproduces images of ‘the immigrant Other’

Abstract

The article explores how sport-related research contributes to the construction and reproduction of immigrants and their descendants as ‘Others’. This process, referred to as ‘Othering’ in this paper, is to be understood as a hegemonic act of ascribing otherness to social groups, marking them as being essentially different, generalising these alleged differences and transferring this alleged otherness into inferiority. This paper elaborates on this process of Othering theoretically and empirically. Qualitative content analysis of sport-related German-language academic publications enables an investigation of how researchers deal with social constructs of difference, such as ‘immigrant’, ‘migrant’ or ‘migrant background’, as well as revealing whether and how Othering occurs in their publications. As a result, this article demonstrates that Othering is found in a substantial number of academic publications. Furthermore, it exemplifies and discusses how the various forms of Othering manifest themselves at different stages in the research process.

Keywords:

• Othering
• integration
• immigrants
• migration
• research ethics

1. Introduction

Over the last 20 years, there has been a surge in research about the role sport can play for immigrants and their descendants. While the respective publications vary in the specific questions they address, they share one commonality: They operate with social constructs of difference that usually refer to the migration experiences of an individual or their family. This commonality also defines the topic of the present article, which addresses processes of Othering. We will take a closer look at knowledge production in the field of sport and migration, as well as how researchers deal with social constructs of difference, such as ‘immigrant’, ‘migrant’ or ‘migrant background’.¹ Our specific research question is whether and how sport-related research contributes to the construction and reproduction of immigrant groups as ‘the Others’. Is the idea that immigrants are fundamentally different questioned, tested, objectified or reinforced by those doing the research? By raising this question, we aim to critically assess and discuss the epistemic foundations of research on sport and migration. We want to (re)raise awareness that we, as researchers, are never fully objective but that we are, and will always be, members of the societies in which we have been socialised and through which we have incorporated a specific set of norms and values (e.g. Homans, 1976; Lacey, 2018; McMullin, 1982). We further argue that such an analysis can be understood as a mirror that helps to draw attention to possible biases and hegemonic patterns in knowledge production, while also serving as one way of combating or preventing Othering in future research.

We will discuss Othering in and through research by focussing on one specific research segment: German-language publications that empirically address the role of sport with regard to ‘immigrant societies’.² We are aware that such a case study is limited in scope and that the results shown here do not necessarily represent how discourses occur in other segments or national contexts of sport-related research. However, this analysis can serve as a starting point for future broader studies on this topic. Furthermore, we argue that such a focus is meaningful, because social categorisations, especially around migration and so-called integration, are most often discussed in a national framework. While theoretical discussions have long pointed towards the limits of methodological nationalism (Wimmer & Glick Schiller, 2003), dominant categories of ‘Otherness’ often differ across national and linguistic contexts due to historical differences in terms of migration, colonialism, citizenship regimes and more (Brubaker, 1990).

A focus on the German context is of scholarly interest and significance in a number of further respects. First, Germany is a country that ranks highly in terms of global national hierarchies, which means that many immigrant groups hold nationalities that are often regarded as ‘lower’ (e.g. Mau et al., 2015; Swindle et al., 2020). Second, Germany has a mixed track record of acknowledging immigration and dealing with related symbolic differences and hierarchies. Current public discussions on migration refer mostly to migratory movements since the start of working migration to Germany in the early 1960s (for a more extensive historical overview, see Bade & Oltmer, 2011). While German society has thus been an immigrant country for decades, self-understanding as an ‘immigrant society’ took a long time to emerge and is often argued to have begun in the 2000s (Bade & Oltmer, 2011). Currently, approximately 26% of the population is made up of first- or second-generation immigrants, many of whom were born with or who have obtained German citizenship (52.4%) and many of whom are foreign nationals (47.6%) (Expert Commission on the Framework Conditions for Integration Potential, 2020). In such a context, it is of importance to ask how researchers are dealing with the respective social constructs of difference, and whether they have paid attention to the fact that the use of these human-made artefacts may carry a risk of overemphasising and essentialising them.

The structure of this paper is as follows: The next section introduces the concept of Othering and explains how Othering processes occur (Part 2). This will be followed by a description of our methodological approach, which is based on a content analysis of sport-related academic publications (Part 3). We then present and discuss the findings of this empirical study, demonstrating that Othering can be found in a substantial number of academic publications on sport in immigrant societies, and revealing that Othering manifests itself in various ways and at various points in the research process (Part 4). In the conclusion, we summarise the findings of our research (Part 5).

2. Theoretical considerations: Understanding Othering

To gain a better understanding of Othering processes in general, and particularly about how immigrant groups are marked and portrayed as ‘Others’ in the German context, it is beneficial to integrate thoughts and references from multiple academic discourses and research contexts.

2.1. Integrating works from different research contexts

Postcolonial studies (e.g. Bhabha, 1994; Chakrabarty, 2008; Said, 1978; Spivak, 1985) examine how colonial hierarchical power relations have continued to shape social inequalities, discourses or representations after the formal end of colonialism. Specifically, Edward Said’s Orientalism (1978) is often regarded as a central work within the postcolonial canon and as having laid essential groundwork enabling the analysis of violent processes of constructing and hierarchising binary forms of difference (Ha, 2016). Deploying Michel Foucault’s notion of discourse, Said examines western styles of representing one’s Self as a civilised authority that defines, reconstructs and fabricates images and clichés of the Orient as the primitive ‘other’, thus detailing ‘the ways in which Eurocentrism not only influences and alters, but actually produces other cultures’ (Ashcroft et al., 2007, p. 85). In the context of our study, another relevant paper is that by Krumer-Nevo and Sidi (2012), addressing the question of how Othering can be avoided in the academic writing process. Krumer-Nevo and Sidi (2012) defined Othering as ‘the risk of portraying the other essentially different, and translating this difference into inferiority’ (Krumer-Nevo & Sidi, 2012, p. 299). Additionally, they highlighted various mechanisms, such as objectification and decontextualisation, that constitute Othering and which will be discussed in more detail in the course of this paper.

While postcolonial studies primarily focus on Eurocentric practices of knowledge production which are crucial in constructing the image of the non-occidental Other through a hegemonic discourse (Said, 1978), research in the fields of migration and racism has mainly addressed intra-societal negotiations in diverse societies. Although these works may not always use the term ‘Othering’, they do address the hegemonic act of producing ‘the Other’ in their alienness and can thus help to shed light on how ‘the immigrant Other’ is constructed. In Germany, this discussion has been driven by researchers following the post-migrant paradigm. These scholars have argued that even though many immigrant groups have lived in Germany for generations and have obtained German citizenship by birth, they continue to be portrayed as strangers who do not entirely belong to Germany and who are thus excluded from the imagined community of the German nation (e.g. Foroutan, 2016). This process of establishing imaginations about differences is also described in academic publications examining racism and racialisation. Here, ascription and the generalisation of alleged differences are seen as foundational acts through which ‘race’ is produced and reproduced (Eggers, 2017; Fassin, 2011; Hall, 1994; Hochman, 2019; Miles & Brown, 2003). Furthermore, works on ethnic boundary making help to understand Othering insofar as they do not conceive of ethnicity ‘as a matter of relations between pre-defined, fixed groups’, but rather as the very result and process ‘of constituting and re-configuring groups by defining the boundaries between them’ (Wimmer, 2008, p. 1027; also see Barth, 1969; Lamont & Molnár, 2002).

In sports studies, explicit references to Othering have been rare ( for exceptions, see Bartsch et al., 2019; Ungruhe, 2014). However, some references can be found in theoretical works on sport and Otherness (Seiberth, 2010; Seiberth & Thiel, 2014) and in international sport sociological studies that deal with the processes of racialising Black athletes. The latter studies have revealed how social constructions of what has often been called ‘race’ are naturalised, as well as demonstrated how racist ascriptions about the athletic superiority and the intellectual inferiority of Black athletes are produced and reproduced in various sport contexts (Carrington, 2010; Engh et al., 2017; McCarthy et al., 2003; van Sterkenburg et al., 2012).

It is important to acknowledge that our thoughts are also influenced by numerous works about the (de)construction of gender and (dis)ability (e.g. Hagemann-White, 1993; Hartmann-Tews et al. 2003; Heckemeyer, 2021). In particular, our understanding has benefitted from Finnern and Thim’s (2013) theoretical study on the ways research contributes to the objectification of disability, highlighting the ‘reification pitfalls’ (p. 160) associated with categorisations. Their work identified crucial points in the research process at which reification occurs, resulting in a framework that we will incorporate into this paper’s own analyses.

2.2. Modes of othering

With reference to the abovementioned discourses, and with regard to the context of our project, we define Othering as a hegemonic act of ascription in which and through which images of ‘the immigrant Other’ are (re)produced over time. The respective publications vary considerably in the concrete aspects discussed and the terminology used. However, bringing together the considerations found in these discussions helps us to create a heuristic framework integral to our subsequent analyses. To this end, we distinguish five modes of Othering, which should be seen as interacting and intricately interwoven elements:

1. Ascription, which may follow the process of categorisation, is a foundational act of Othering (Fassin, 2011). To illustrate this, it should first be emphasised that categories are human-made artefacts resulting from definitions that divide people into subgroups based on certain criteria. A case in point is the category ‘migrant background’, often used in Germany and commonly based on the criterion of whether an individual or their parents have immigrated to Germany or whether an individual, or at least one of their parents, did not acquire German citizenship by birth (for an in-depth discussion, see Will, 2018). Othering occurs when the criteria for these definitions are gradually lost sight of, because they are superseded by alternative attributes ascribed to groups of people. This occurs, for example, if those who are labelled as ‘persons with a migrant background’ are not merely described in their definitional differences (e.g. with regard to their migration experiences) but if further attributes and characteristics ascribed to them (e.g. culture and integration needs) make them appear fundamentally different.

2. Othering is also constituted by marking the differences that have been created by ascription. Marking occurs when emphasis is put on the purported difference, strangeness or peculiarity of immigrant groups. Within this process, the imagined ‘Other’ is labelled and thus demarcated also by language, whereas the ‘Self’ is usually taken as an unquestioned standard and remains unmarked. This is, for example, the case, if white people situate themselves beyond ‘race’ and ‘ethnicity’ while using these social constructs of difference to mark ‘the Others’ (Flintoff & Dowling, 2019).

3. Othering implies hierarchisations and unequal relations of power. There is usually a dominant, hegemonic position that sets the norm counterposed to a marginalised out-group (Engh et al., 2017; Fassin, 2011; Krumer-Nevo & Sidi, 2012; Spivak, 1985). The allegedly different attitudes and behaviours of immigrant groups are not only marked but also linked to moral codes. Differences are construed as deficits, addressed as a problem or portrayed as a deviation from a desirable norm. This, in turn, often results in the demand for immigrants to assimilate to this norm, which is implicitly or explicitly valued as morally superior. This hierarchical process of evaluation may also be expressed in a paternalistic, well-meaning rhetoric about the ‘poor and pitiful immigrant’ (Brons, 2015; Elias & Scotson, 1965; Krumer-Nevo & Sidi, 2012; Said, 1978; Spivak, 1985). As Fassin (2011) puts it in his essay on racialisation: ‘This moral evaluation usually takes a negative form …. But it can also adopt a positive expression: valorisation of difference may then result in paternalistic attitudes and sometimes segregation policies. Whatever moral orientation it follows, ascription is always an abuse of power: in this sense, it is political’ (p. 423).

4. Generalisation is another mode of Othering. In this case, ‘heterogeneous social groups are perceived as homogeneous categories’ (Krumer-Nevo & Sidi, 2012, p. 300), whereas self-groups are described and regarded in a much more nuanced way (Finnern & Thim, 2013). These generalisations occur, for instance, when immigrants are merely considered as representatives of their group (Fassin, 2011; Said, 1978; Seiberth & Thiel, 2014; Ungruhe, 2014) and are then endowed with the social function of illustrating the ‘Otherness’ of this group as a whole. Generalisation can also take the form of decontextualisation, such as when immigrant and non-immigrant groups are portrayed in their differences without considering further subgroups, internal differentiations or intersections (Finnern & Thim, 2013; Krumer-Nevo & Sidi, 2012). Decontextualisation can also mean that potential reasons for differences are not sufficiently considered or that behaviour is ‘abstracted from the context in which this behaviour was developed’ (Krumer-Nevo & Sidi, 2012, p. 300).

5. Typically, the alleged knowledge about ‘the Others’ and about the relevance of certain categories (e.g. migrant background) will be institutionalised over time. Permanently ascribing and marking Otherness, pervasively generalising and continuously hierarchising differences create a veritable ‘dogma of Otherness’ (Foroutan, 2016, p. 234) that is then rarely called into question any more (Krumer-Nevo & Sidi, 2012; Seiberth & Thiel, 2014). Once such patterns of thinking about the Otherness of immigrant groups have become shared knowledge, further theoretical or empirical justifications for the underlying assumptions are often overlooked and even counterexamples fail to invalidate the resulting stereotypes, as they are either ignored or reinterpreted as exceptions to the rule (Foroutan, 2016; Krumer-Nevo & Sidi, 2012; Seiberth & Thiel, 2014). Ungruhe (2014) offers an illustration of this process in his paper on the reception of African football players: ‘For instance, if a Malawian player makes a tactical mistake on the pitch, or if a Cameroonian out-dribbles two opponents within a small space, these actions will often be interpreted with reference to existing ethnic knowledge which ascribes to these players an unsophisticated and immature style of play or, accordingly, an enjoyment of the game. Thereby, certain actions of African footballers on the pitch are explained by referring to this stock of social knowledge of supposed ‘African’ attributes. When these ethnic ascriptions then seem to be continuously verified, the social knowledge reproduces itself and adds to the power of such ascriptions’ (p. 202).

Narratives about ‘the immigrant Others’ are produced, maintained, consolidated or questioned by multiple actors in the field of sport, such as journalists, club officials, athletes, coaches or PE teachers. We argue that researchers are similarly involved in this process as well, and are therefore able to play a crucial role in reifying or in dismantling Othering. Research is often perceived as producing ‘objective knowledge’, and as a result it can—if Othering appears—reinforce invalid imaginations. However, if Othering is directly addressed and discussed, research can also be a strong instrument in combatting and deconstructing metanarratives and imaginations of the ‘immigrant Other’.

3. Material and methods: content analysis of sport-related academic publications

To analyse if, at which points and in which modes Othering occurs in research, we conducted a content analysis of academic German-language publications about the role of sport in immigrant societies.

3.1. Data collection and data basis

To conduct our research, we searched for academic publications that address the topic of sport in combination with keywords such as ‘migration’, ‘integration’, ‘culture’, or ‘racism’. We used the search engines Google Scholar and PRIMUS, and we browsed the tables of contents of the two most renowned sport-related scientific journals in German: Sport und Gesellschaft and German Journal of Exercise and Sport Research. Bibliographical references of the publications found through this procedure were then used as the basis for identifying further publications using the snowball principle. In addition, we specifically searched for publications by authors known to us as researchers in the field of sport in immigrant societies. We established three criteria as a guideline for our search:

1. To ensure that our data represented the current state of research, only articles published since 2010 were included.

2. Assuming that Othering is also manifest in language and with regard to the focus of our case study, which aims to capture a specific segment of research and a specific academic discourse, we focussed on German-language publications only.

3. Grey literature, working papers, evaluation reports and non-edited books were excluded from our analyses.

To ensure that the publications found were thematically relevant, we either read the abstracts of the articles (if available) or scanned the full texts. This second step of selection was relevant in yet another respect. Given our goal of ascertaining the points in the research process where Othering may occur, we selected only publications that used an empirical research approach. Based on these selection criteria, we excluded 17 journal articles (11 of which were non-empirical, 5 of which were not thematically relevant and 1 of which was a portrayal of a sport-related integration project); 26 chapters from edited books with a focus on sport and migration (19 of which were non-empirical, 3 of which were not thematically relevant and 4 of which were portrayals of a specific project, a best-practice-example or a single interview) and 10 chapters from further edited books (7 of which were non-empirical, 1 of which was not thematically relevant and 2 of which were portrayals of sport-related integration projects).

Based on these search and selection criteria, 52 publications remained in our potential sample. Of these, 14 publications were journal articles, 27 were chapters in edited books with an explicit focus on sport in immigrant societies and 11 were chapters in other kinds of edited books. While quantifying the phenomenon of Othering, as opposed to illustrating how Othering occurs at specific points in the research process, is not a key concern of our project, we thought it essential to include a critical mass and suitably wide range of publications in our analyses. As a consequence, we analysed all journal articles (n = 14), and we randomly selected book chapters from the pool. We analysed 75% of all chapters in edited books with an explicit focus on the topic of sport in immigrant societies (n = 21) and 75% of all articles in edited books without an explicit focus on this topic (n = 8). Altogether, our findings are based on the analyses of 43 articles adding up to 699 pages, and thus on a rather large data base for a qualitative content analysis.

3.2. Data analysis

To answer the question at which points and in which modes Othering manifests in the research process, we first reviewed the literature to work out key dimensions of Othering and demarcated the classical parts of research reports. Thus, we developed the first set of provisional coding categories deductively. In the further course of the coding process, these categories were adjusted according to the material analysed (Gläser & Laudel, 2009; Mayring, 2015). The final coding scheme included categories that describe the different modes of Othering, as well as categories that refer to various steps in the research and writing process.

All articles were coded at least twice in this process: once while the coding scheme was being developed and adjusted and once after the coding scheme had been finalised. To keep inter-coder reliability high, coding categories and anchor examples were regularly discussed. Additionally, different coders (both authors of this article) were used for the first and second rounds of coding. Articles by Tina Nobis (author of this article) were coded by a student who was part of our research team in the initial stage of the project but dropped out later on. It should be noted that Othering can also take the form of an omission (e.g. of not addressing certain issues) and that it cannot always be ascertained based on a single text passage; thus, thorough analysis demands that each article must also be considered in its entirety. Towards this end, we wrote a summary for each article describing these hidden aspects of Othering that, while unable to be captured by our coding scheme, were also considered in the data analysis.

We would also like to clarify that the goal of our empirical study was neither discrediting academic works nor discovering which specific researchers are (or are not) engaged in Othering. Instead, we aimed to ascertain whether and how Othering occurs in sport-related research about the role of sport in immigrant societies and—with regard to our results—to raise awareness among researchers and initiate a discussion on how Othering can be prevented in future research. As a consequence, it is neither necessary nor helpful to refer to the real names of the authors whose articles we analysed. We have instead pseudonymised the respective data by referring to each publication by a number from P1 to P43. However, we will share the data corpus upon reasonable request with those who aim to test our study.

4. Presentation and discussion of results: othering in sport-related research

We assumed that Othering can manifest itself in various forms and at various points in the empirical research process. Accordingly, this chapter describes and discusses whether and how Othering occurs in our selected publications by focussing on different steps within the research process.

4.1. What kind of research questions are addressed?

Clustering the 43 publications according to their guiding research questions enabled us to classify them into five thematic complexes: (1) 15 publications dealt with the sport participation of immigrants and their descendants. Issues addressed in these publications included ‘the degree to which immigrants living in Germany actively engage in sport’³ (P1) or ‘how immigrant girls’ and women’s sport club participation can be intensified and what conditions must be met for this to happen’ (P13). (2) 12 publications addressed the potential effects of sport participation at the micro level—exploring, for instance, the ‘psychosocial effects of engaging in sport’ (P6)—or, at the meso level, the integrative capacity of sport clubs. (3) 8 publications dealt with immigrant sport clubs. (4) 6 author collectives explicitly addressed school sport, such as by asking ‘whether intercultural learning and social integration can be promoted through physical education’ (P5). (5) 6 publications reported the results of evaluations of sport-related programmes. Here, the objective could have been to explore how sport programmes (e.g. for refugees) should be designed and analysed ‘to avoid a homogenising or hierarchising ascription to or perpetuation of ethnic and cultural differences’ (P11) or to provide information on the implementation of a specific project, its structure, feasibility, effects and acceptance by the target groups (P35). (6) 8 publications were classified within two of the above-defined thematic fields, and 4 publications could not be classified within any of them.

With regard to the research questions asked and the topics addressed, Othering manifests itself in two ways. Some articles that addressed sport participation and/or the effects of sport participation focussed, as defined by the research question, on immigrants, certain immigrant groups or Muslims only (P1, P5, P6, P24, P27, P34). This is the case, for instance, when factors that influence the sport participation of immigrants were investigated, when the role of religion in sport participation among Muslim girls was explored or when the integrative effects of sport participation on adolescent immigrants were addressed. In most of these publications, no reason was given for this choice of focus. Logically speaking, however, this limitation only makes sense if the authors of the respective papers assumed that the specific groups of interest are somehow different or special, such as with regard to their sport participation rates, the factors that influence their sport participation or their needs to benefit from the potential integrative effects of sport participation. At this point, it becomes clear how assumptions about the Otherness of immigrants have already become shared knowledge among some researchers, as differences are implicitly ascribed to immigrant groups before they have been explored—often with no justification or empirical evidence provided.

Our second finding is that German-language research on sport in immigrant societies primarily adheres to the ‘great sport myth’ (Coakley, 2015), which is the basic idea that sport is good. In most of the publications, the research questions themselves already imply a focus on positively connoted concepts, such as integration, opportunities, participation, support or resources. Only one publication explicitly addressed the risk of ethnic and cultural differences being ascribed and perpetuated (P11). Not a single publication pursued a research question in the thematic field of discrimination or racism. Two publications serve as an illustration here: in both, the research question included the term ‘problem’, but the problems that were discussed had nothing to do with the problems immigrants may face in the field of sport but with problems associated with their inclusion or integration (P34, P42). From our point of view, many of the research questions posed were not problematic per se, nor did we expect each and every researcher to explore a subject (e.g. the role of sport in immigrant societies) in all its facets. The overall picture, however, reveals a bias that is constituted by a specific form of decontextualisation, namely a failure to adequately consider and explore racism and discrimination in the field of sport. We know the field of authors quite well, and can thus conclude that the overwhelming majority of the authors included are white and that hardly any name points to a migration experience. It is thus valid to say that in most cases, white German researchers without a migration experience conducted research about immigrants and their descendants from a privileged hegemonic position. This observation leads us to ask whether one reason for the specific thematic alignment among these German language publications is a form of white privilege, that is, the privilege of not having to deal with racism because one is not directly affected by it (Leonard, 2017).

4.2. What are the theoretical frameworks and concepts?

Othering can become evident in the theoretical frameworks and concepts used. Three arguments support this finding. Roughly half of the publications analysed referred to the theoretical concept of integration, at least marginally (P1, P2, P4, P6, P8, P9, P10, P12, P13, P16, P17, P18, P19, P29, P30, P31, P34, P37, P38, P40, P41, P43). Many of these publications failed to explain why the particular theoretical framework was chosen. Thus, it appears that for some authors, it has become almost self-evident to link the concepts ‘integration’ and ‘migrant background’. This creates the impression that writing about immigrants automatically raises questions about integration, as if it was some kind of shared knowledge that immigrants and their descendants, born and socialised in Germany, have certain integration deficits that need correction. In some publications, the concept of integration is explained, but here again, questions with regard to Othering arise. One publication stated: ‘For one thing, we need to create the societal conditions and prerequisites for immigrants to be integrated into the host society, which also implies the need for society to open up culturally. But the immigrants, for their part, also need to actively participate in the integration effort by, e.g. complying with and communicating legal regulations (e.g. compulsory education) and basic democratic values. At the same time, the norms and values of their country of origin can be maintained and brought into the host society insofar as they are compatible with basic constitutional and democratic positions’ (P7). In this text passage, integration is not understood as a one-way adjustment. However, the specific emphasis on the requirement of the host society to open up culturally already implies that the authors assume that immigrants are culturally different and furthermore leaves us to ask, whether opening up structurally is not that relevant. Furthermore, it is still apparent that immigrants are conceptualised as having different values and norms (which they may maintain under certain conditions) and that it is them who need to actively integrate while the host society is merely expected to create the conditions that facilitate this integration process. Further questions arise from the fact that the integration effort is understood in terms of compliance with legal regulations and basic democratic values: If this is a challenge to be met specifically by immigrants, the impression one gets is that the authors of this article assume that people without a migrant background are per se democratic and law-abiding, whereas persons with a migrant background are not.

While the article quoted above at least elaborated on what is to be understood by integration, many other publications did not provide any such clarification. Often, they either offered no explanation at all. In other cases, they distinguished, at an abstract level, between social, cultural, political, structural and/or affective integration but did not specify the threshold from which a person is considered to be socially, culturally, politically, structurally or affectively integrated. However, when taking a closer look at how the concept of integration was operationalised in empirical studies or how empirical findings were commented on, it becomes evident that some publications operated with an understanding of integration that aligns with the idea of assimilation (P1, P4, P7, P8, P9, P12, P17, P18, P33, P41, P43). This was the case when the social integration of immigrants through sport was measured by interethnic contacts and relationships (P4, P7, P8, P9, P33, P41, P43). Underlying this line of thought is a hidden hierarchisation: It appears to be of higher value if immigrants build up contacts with non-immigrants than if they build up contacts with immigrants, as non-immigrants are the reference standard when assessing the social integration of immigrant groups.

The 15 publications that dealt with issues of sport participation provided another interesting insight. Nine of these publications referred to the theoretical framework of integration, which suggests that sport participation, or sport club membership, is regarded as an indicator of integration into sport. As a result of this stance, and even though this may not be explicitly stated in the publications in question, this also means that lower sport participation rates among immigrants indicate an integration deficit. Looking into further publications about the sport participation among other, non-immigrant population groups (e.g. adolescents with low educational levels), we discovered that they hardly ever refer to integration, but that they primarily explore sport participation with theoretical references to social inequality or socialisation (for a review of research in Germany bearing on sport participation among children and adolescents, see Nobis & Albert, 2018). This contrast suggests a case of Othering by hierarchisation. It becomes apparent that lower sport participation rates among non-immigrant groups are often theorised as a result of disadvantageous socialisation conditions or as an indicator of social inequality, whereas lower rates of sport participation among immigrant groups are often regarded as an indicator of insufficient integration. If we follow this thought one step further, one might argue that immigrants are perceived as a group that has deficits and problems, whereas less privileged non-immigrants are perceived as a group that faces difficulties and problems.

4.3. How are the research questions assessed empirically?

Our content analyses also showed that empirical designs and analyses are points in the research process where Otherness can be (re)produced. Some publications claimed, according to their research question, to explore the difference between immigrants and non-immigrants, but they could not possibly have done so: Their empirical analyses referred exclusively to immigrants or to specific immigrant or Muslim groups and thus lacked a control group (P6, P17, P22, P24, P27, P28, P30, P31, P32, P34; to some extent, also P1 and P13). Often, however, the absence of a control group that could provide empirically substantiated insights into the role of the so-called ‘migrant background’ did not prevent the authors from making statements about the Otherness of immigrants or of specific immigrant groups. A case in point is a publication that analysed qualitative interviews conducted with immigrants who practised sport on an elite level. Although the authors themselves stated that ‘a direct comparison to native peers is missing’ (P32), it is only shortly afterwards that they discussed how ‘migration-related particularities’, such as family ties and gender role orientations, gained meaning for the interviewees. This is one example of how images about the Otherness of certain immigrant groups are reproduced and further institutionalised by publications featuring data that do not allow for such statements.

Another inadequate form of data analysis is to be seen in selectively conducted and presented statistical computations, which can lead to an overemphasis of differences. For example, one author collective computed ‘separate analyses for the group of adolescent migrants and the adolescent Germans’ (P6) but did not explain why the findings reported in the article exclusively refer to the findings for the group of adolescent migrants. The problematic nature of this practice became evident when the authors stated that adolescent immigrants who are members of sport clubs were found to have higher scores with regard to aggressive behaviour than adolescent immigrants who do not engage in sport clubs. However, whether a migrant background plays any role at all in this comparison remains an open question since the readers of this article are not told whether similar correlations were found for the ‘adolescent Germans’. Still, the migrant background becomes specifically marked by this practice, and the very fact that some findings are left unreported can lead to the implicit conclusion that aggressive behaviour is an issue of a particularly high relevance among immigrants. A similar pattern was found in another article, exploring the effects of gender stereotypes on sport club membership via logistic regression models (P28). The authors documented lower rates of sport club membership among adolescent immigrants with traditional gender role understandings—a finding only shown to be statistically significant for one specific subgroup, older girls. One problem here is that the information that only a very small portion of adolescent immigrants adhered to traditional gender role understandings is relegated to a footnote, notwithstanding the fact that this information significantly reduces the overall relevance of the reported finding. Moreover, the regression results showed that the gender role understandings of young immigrants played a secondary role compared to the variables of sex, age, education and place of residence. However, when the findings were summarised in the article’s conclusion, this was not appropriately discussed and contextualised.

Finally, empirical designs may foster processes of Othering when the data analyses are lacking complexity, such as when bivariate analysis is the only method used to explore the link between migrant background and sport participation (P1, P4). Statistical correlations are thus decontextualised, as other factors that usually have an effect on sport participation (e.g. economic background) are not considered. This, too, may contribute to the impression that immigrants are a very special and different group.

4.4. How are the findings presented and interpreted?

Othering also manifests itself in the presentation and interpretation of empirical findings. First, in several publications, the role of the migrant background was given more weight than what is warranted by the empirical data these publications provide. This occurs when overgeneralisation of language is found, as in the following text passage where requirements for successful practical projects are extrapolated from a limited number of qualitative interviews: ‘Muslim mothers and fathers, in particular, permit and support participation in football matches only when they are organised and supervised by the school and on the basis of gender segregation’ (P19). Another mode of inadmissibly far-reaching generalisations was found in some publications where conclusions were drawn from a small number of cases or where the data that the analyses relied on were not further specified (e.g. P1, P9, P12, P19, P29, P30). One publication analysed data that stemmed from a quantitative questionnaire in which 135 out of slightly over a thousand participants were categorised as persons with a migration background. One of the results sections then focussed on determinants for immigrants’ (but not non-immigrants’) sport participation using inter alia differentiation between gender, age, religious affiliation or educational status. Although the number of cases for the subgroups analysed was necessarily very small, the authors’ conclusion claimed that ‘these findings mean that sport offers for the target group of people with a migration background must be adapted to their specific needs by taking gender-, religious- and age-related aspects into account’ (P1).⁴ Another article used a single male case to conclude that ‘familiarity with football is primarily found among boys – as the case of Emanuel also shows – but is uncommon among girls’ (P9). Far-reaching statements such as these are usually not contextualised, nor is their limited validity discussed. As a result, the supposed knowledge about ‘the Others’—here, marked by the abstinence from sport or football among immigrant girls—is further institutionalised.

Second, the respective generalisations were often not neutral in terms of value judgments. Rather, immigrant groups, or groups that were perceived as immigrants, were described as having specific deficits and contrasted with the allegedly ‘normal’ values and behaviours of non-immigrants. A particularly common case found in the publications analysed was the assumed—and usually not sufficiently substantiated—prevalence of traditional and primarily hierarchical and restrictive gender roles among immigrant or Muslim groups. For example, one author collective claimed that ‘becoming a member of a football team’ is an ‘objective of integration’ (P19; similar: P12, P34; see also Part 4.1). Subsequently, the authors of this article recommended that to attain this objective measures such as ‘separate locker rooms for girls’ and ‘respect of religious rules and holidays’ should be taken, and they further highlighted the ‘complex challenges’ that sport clubs have to meet in this respect (P19). This conclusion is somewhat surprising, since gender-separated locker rooms in sport clubs are likely to be the rule rather than the exception in Germany. It is notable that these authors portray it as an extra measure to be taken to address the needs of immigrant girls. Furthermore, it remains unclear why paying attention to religious holidays should be seen as a complex challenge, as sport clubs in Germany have no training on religious holidays such as Christmas or Easter. Conclusions like these not only disregard the hegemonic nature of social positions and the ways they are systematically ingrained in organisations, but they also constitute a specific form of marking differences; the authors give the distinct impression that religion and religious holidays are only relevant for non-Christian immigrants.

These and similar lines of argumentation can also be found in further publications, where ‘the Others’ are easily imagined as backward. One publication noted that ‘especially in schools where many nationalities meet, chaos often rules in the beginning’ (P19; similar: P12, P9). Another team of authors stated: ‘Partly, they [refugee students] still need to learn that living together and learning together can only succeed if they respect each other, as should be the case in sport all over the world. Rituals which are common in Germany and which facilitate living together, are often unknown to newly arrived students in their countries of origin’ (P12). Another author spotted ‘mentality discrepancies’ on the part of immigrant members of sport clubs: ‘namely, comparatively high levels of impulsive and aggressive behaviours in some immigrants, which are likely to contribute to the fact that a disproportionately high number of ethnic sport clubs are known for their teams’ particularly rough playing’ (P17). Passages such as these clearly show that the allegedly different values and norms of immigrants are not only (re)produced and frequently marked as a divergence from what is ‘normal’ but that this supposed Otherness is often translated into inferiority.

This framing of immigrants as a problem group with deficits must be put in context with the fact that, with limited exceptions, the large majority of the publications we analysed did not consider that sport settings might not be as open and welcoming to everybody as they have often been assumed to be. Although the informal closure mechanisms of sport clubs have been addressed in previous German-language theoretical articles (e.g. Seiberth, 2010; Seiberth & Thiel, 2014), these arguments are seldom considered in the empirical publications we analysed. Even in cases where the role of the context (e.g. specific club cultures) is addressed, it is merely mentioned at some point in the article but not further pursued in the empirical analysis or with regard to the interpretation of empirical findings (P34, P36, P39). Experiences of discrimination that immigrants may face in sport clubs or discriminatory practices that may have been inscribed into sport institutions’ policies are only occasionally considered. Cases in point are articles that argue that exclusion experienced in sport clubs may be a motive for immigrants to found their own clubs (P20), describe structural closures against immigrant sport clubs in local sport policy networks (P23) or explain low sport participation rates among young refugees by their uncertain residence status (P10). Much more often, however, the experience of structural exclusion or racism is disregarded and sometimes even reinterpreted. Thus, drawing on empirical data, one article showed that immigrants are underrepresented in German sport clubs, further demonstrating that the question of whether a specific sport club has more or less active immigrant members has much to do with its organisational structure. In their conclusion, however, the authors emphasised the ‘integrative power of German sport clubs’ (P16) and they highlighted the growing number of immigrant members without bringing this argumentation into perspective with their previous statement about these clubs’ exclusivity. In other articles, structural barriers or discriminations, such as a lack of citizenship as an obstacle to engagement in professional sport, are evident from the provided interview material but are nevertheless not considered in the authors’ analysis (P32).

5. Conclusion

In this paper, we analysed the extent to which Othering occurs in German-language empirical research on sport in immigrant societies. Our primary interest was not to quantify this phenomenon, but instead to determine and illustrate whether and at which points in the empirical research process Othering manifests itself. As a result, we discovered Othering processes in a substantial number of the publications analysed. We found that Othering manifests itself in various forms and at various points in the research process: (1) According to the research questions asked, Othering manifests in a specific form of decontextualisation. That is, most publications focus on the benefits and positive implications of sport, with little if any consideration given to the issues of Othering, racism, and discrimination. (2) Othering is also created through the theoretical framing of publications. Specifically, the reference to a concept of integration that subscribes to a deficit-based view of immigrants and the need for them to assimilate can be regarded as a form of Othering by hierarchisation. (3) Some publications feature empirical designs that can lead to an overemphasis of differences. These can be regarded both as a form of ascribing differences and of marking immigrants as ‘the Others’, but also as a way of further institutionalising imaginations about ‘immigrant Others’. (4) In addition, Othering manifests itself in the modes of generalisation and hierarchisation within interpretative frameworks; a considerable number of the publications gives the impression that immigrants are fundamentally different and likely to have specific deficits. Thus, the essence of our analyses is that at present, German-language sport-related research on immigrant societies is a field that tends to adopt, maintain and institutionalise certain stereotypes about the ‘immigrant Other’ but hardly ever discusses topics such as the deconstruction of certain images, the meaningfulness and relevance of categories and the processes of Othering.

In such a context, our research aims to heighten the level of sensitivity required when dealing with the problems associated with using social constructs such as ‘migrant background’ or ‘immigrant’ in research. We argue for increased critical (self-)reflection in the research and writing process, aiming to encourage a stance that considers the fact that working with social constructs of difference comes with the risk of essentialising the categories used. Our aim is not to negate existing and relevant differences. Rather, we claim that assumed differences should not be taken for granted and that it is necessary to empirically explore and adequately contextualise differences, to remain open-minded enough to falsify assumptions about alleged differences, to question and evaluate our own biases as researchers and to keep in mind the vantage points from which we approach a certain subject. This also includes the need to critically reflect on practices of knowledge production in future research. The promotion of early-career and advanced researchers could put more effort into critically assessing whose voices and perspectives have not yet been heard, be it as racialised Others in sport or in academia itself. These arguments gain even more significance when we acknowledge the responsibilities we have as actors who are often perceived as authorities who produce truths and, just as importantly, can give a kind of ‘public life’ (Fassin, 2011, p. 424) to the image of the supposed Other.

Furthermore, we argue that if we want to obtain a more adequate and comprehensive picture of sport in immigrant societies, the research agenda must move away from a sole focus on immigrants in terms of their supposed deviation from the norm. Instead, it must be extended and reaccentuated with empirical studies addressing processes of exclusion and racism in the field of sport, or with research focussing on how the often unmarked and unquestioned hegemonic norm constitutes itself. This is accompanied by a need for a more in-depth discussion of the respective theoretical approaches, and for the introduction of further theoretical approaches (e.g. postcolonial or Critical Race Studies) that can help to shed light upon phenomena that we are not able to capture by studying sport through the theoretical lens of integration only. These underutilised theoretical frameworks may prove essential in helping to approach and interpret old findings in a new light. Having said this, we want to specifically point to two empirical articles, published after we had finalised our analysis, that explicitly focus on postcolonial discourses and othering among PE teachers in Germany: Bartsch et al. (2019) and Bartsch (2020).

The analyses conducted in this paper were limited to current German-language empirical sport-related research, thus covering only a segment of a larger research field. Nevertheless, we believe that the findings on modes of Othering in the research process that we have presented in this paper can also be relevant in other social science fields, other national (and international) discourses and sport-related integration programmes. In particular, the description of the different facets of Othering and the breakdown of our findings according to the various points in the research and writing process may provide a basis for similarly-oriented analyses and discussions that go beyond the field of German-language empirical sport-related research on immigrant societies. Furthermore, international studies that deal with colonial continuities in sport and development programmes (e.g. Darnell, 2015), or analyses that deal with policies and programmes for sports-related integration (Agergaard, 2018), can lead the way in analysing if and how imaginations about ‘the immigrant Other’ have inscribed themselves in sport-related integration programmes.

Disclosure statement

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

Notes

1 When we refer to ‘immigrants’, ‘immigrants and their descendants’ or ‘immigrant groups’ in this paper, we do not follow a clear definition. This lack of conceptual clarity is due to the very issue pursued in this paper: The object of our analysis is publications which use social constructs of difference, it is not the groups of persons that are described by these constructs. Since the terms used in the studies we analysed differ, and because social constructs of difference are not always defined and operationalised in the same way, our own use of these concepts is necessarily inconsistent and vague as well.

2 In Germany, the term immigrant society (sometimes: post-migrant society) is commonly used to point to the fact that a large share of the population in Germany is made up of first- or second-generation immigrants and that societal discourses are shaped by negotiations about migration, participation and integration (e.g., Foroutan, 2016; Mecheril, 2012).

3 These and further quotes from the German language publications were translated by an official translator in a first step and then rechecked by the authors of this paper in a second step. The whole article, quotations included, was proofread afterwards.

4 Even though the authors did state at the very end of their paper that the small sample size led to restrictions with regard to generalisations, they did not further highlight that the conclusions about the Otherness of immigrants should be handled with care; instead, they concentrated on elaborating on the possibility that the small sample size might be one reason that the differences did not reach statistical significance.