Citizenship from below: exploring subjective perspectives on German citizenship

ABSTRACT Citizenship is a contested and multifaceted concept. This article introduces a research design to explore individual perspectives on citizenship in Germany. Its emphasis is on capturing broad perspectives on citizenship in a wide array of variations and terms. To achieve this, the article proceeds in two steps: First, a concourse on citizenship is constructed. To populate the concourse, I draw on citizenship theory and the German-speaking Twitter discourse on citizenship. Second, that concourse is used for a further, systematic empirical investigation of perspectives on citizenship in Germany. Results from an online Q-methodological survey with 294 German citizens from autumn 2020 are presented. Four factors emerge from the Q-survey: critical ethno-culturalists, active democrats, liberal democrats, and cosmopolitans. While all associate different concepts of belonging, participation, and equality with citizenship, all four conceptualizations share the view that citizenship is more than a status and connect it to obligation. Obligation takes different forms, namely common tradition, political participation, liberal institutions, and humanist connection. The article combines the most distinctive conflict lines in citizenship studies by drawing on both theoretical and empirical approaches. It thereby contributes a novel methodology and relevant deeper insight into the complexities of citizenship.


Introduction
What does citizenship mean to citizens? In responding to this question, the current state of research offers some puzzle pieces, contradictions, and uncertainties. I argue that expanding our empirical toolkit for identifying individual perspectives on citizenship can contribute to a more comprehensive response. This expansion has the potential to contribute to discussions around how citizenship is understood within both the academic and the public debate. It can also be adapted to explore other contested and complex concepts of political science through a discursive and contextual vantage point.
Bottom-up perspectives on citizenship are understudied. This is an issue in particular because theoretical or elite conceptions of citizenship have been shown to differ from conceptions and practices by citizens themselves (Miller-Idriss 2006). Though investigations of bottom-up conceptualizations of citizenship are receiving increasing attention (e.g. Mader 2016), they observe various inconsistencies. For example, Diehl and Tucci find that prejudice toward foreigners in Germany does not differ between those who hold ethnic and those who hold civic views of citizenship (Diehl and Tucci 2011). In fact, these inconsistencies might simply be an indicator of complexities or ambiguities between different aspects of citizenship. To some extent, such complexities reflect that citizenship is an essentially contested concept (Gallie 1956).
In the following, I propose a research design that captures these complexities to provide a more detailed and differentiated account of citizenship conceptualizations among German citizens. I argue that individual definitions of citizenship should be studied from broader perspectives, and in a wider array of variations and terms. Quantitative analyses of individual-level views on citizenship in particular often rest on the ethnic-civic dichotomy. Typically, attitudes are measured with a closed survey item asking what criteria someone needs to fulfil to be genuinely e.g. German (Helbling and Teney 2015;Lindstam, Mader, and Schoen 2019). While it has been shown that this dichotomy exists empirically (Reeskens and Hooghe 2010), further perspectives might be just as relevant or lead to additional variation. Indeed, investigations into the practices of citizenship have shown that citizens interpret citizenship in various terms, and enact it in a variety of ways (Isin and Saward 2011).
To broaden the strategies we use to study individual perspectives on citizenship, this article introduces a novel research design. This design is discursive, thereby allowing participants to invoke reasons and convey their own definitions of citizenship. It rests on the key concept of subjectivity, understood as individual meaning-making on issues. Subjectivities are always evolving and context-dependent; they are embedded in and emerging from social interactions, shared information, and cultures. A subjective methodology attempts to observe and measure subjectivities, such as through a Q-methodological survey (Stephenson 1986). Such an approach is particularly suitable to the study of citizenship as I define it for the purpose of this article: I follow Isin and Nyers in seeing citizenship as a combination of rights, duties, and performances. Their combinations, situations, and relevant spaces are constantly evolving, and subject to struggles and multiple interpretations (Isin and Nyers 2014).
In this article, I apply a subjective methodology by developing a concourse for citizenship. The concept of concourse originates with the developers of Q-methodology. It describes a universe of statements that captures different aspects and understandings of an issue. How an individual reacts to each statement will depend on experience and context just as much as on beliefs and expertise (Stephenson 1986). In this light, the term concourse is 'to remind us that the concern is with conversational possibilities, not merely informational' (Stephenson 1986, 44). This makes a concourse a suitable approach to studying individuals' views on citizenship. Investigating a concept using a concourse is focused on what most resonates with the individual in their socially constructed context. It has been argued that both collective and individual situated meaning-making are particularly important to the study of contested concepts (Wallaschek 2020). The development of a concourse is a contribution to that end.
Several changes in the relevant social and political context make Germany an interesting case context for studying different understandings of citizenship: First, German citizenship law reforms in 1999/2000 liberalized naturalization requirements, resulting in a changing policy environment. The reforms led to a contested public debate on citizenship and immigration (Miller-Idriss 2006), thereby politicizing citizenship and its inherent conflicts. Second, empirical investigation shows changes in German citizens' perceptions of national pride. These changes indicate heterogeneous, conflictual, and changing understandings of bottom-up perspectives on citizenship (Mader 2016;Miller-Idriss 2006). Third, recent public debate on commonality and obligation has further increased the salience of conflicts between approaches to citizenship: There have been broad discussions around the necessity of a German 'lead culture' ('Leitkultur') as a common set of behavioral rules attached to being German (Pautz 2005), as well as around the policy proposal to introduce a compulsory year of civil service to strengthen social cohesion (Braw 2020).
These developments make it all the more relevant to investigate the changes and ambiguities within bottom-up perspectives on German citizenship. Indeed, the results of this study add nuance to existing observations: I find four distinct bottom-up perspectives on German citizenship. All four place emphasis on obligation as a part of citizenship, albeit with different definitions. Civic obligation is understood as a commitment to democratic institutions, or as a cultural or political practice. The most commonly held perspective on citizenship has some qualities of ethnic citizenship in that its supporters value the protection of German culture and tradition. Meanwhile, the three other conceptualizations all reject cultural conceptions of citizenship to varying degrees. The results show that there are different approaches to defining and practicing non-ethnic conceptualizations of citizenship.
This article proceeds as follows: First, I focus on what a concourse adds to citizenship investigations. Second, I describe how my concourse on citizenship is populated. To construct the concourse, I draw on two sources: citizenship theory and German-speaking tweets on citizenship. The final concourse on citizenship is composed of the conflict lines identified in these sources, grouped into five categories: rights and justice, social and financial justice, individual and social freedom, obligation and participation, and culture and identity. Third, I use the concourse for a systematic investigation of subjective perspectives on citizenship using Q-methodology. I first outline how Q-methodology makes use of the concourse. Then, I present results from an online Q-methodological survey with 294 participants. I analyze and present the results, indicating four emerging factors, or viewpoints, on citizenship: critical ethno-cultural, active, liberal, and cosmopolitan citizens. Thereafter, I discuss the results and their limitations. Finally, I draw conclusions for the further application of a discursive methodology, for understanding subjective perspectives on citizenship, and for everyday and policy discussions on citizenship in Germany.

Populating the concourse
To study subjective perspectives on citizenship, I first populate a concourse. As mentioned above, a concourse aims to capture ways of thinking and, in particular, ways of talking about a concept. This concourse serves as the basis for the following section of this article, an empirical investigation using Q-methodology. Through this design, I investigate subjective perspectives on citizenship in two ways: First, discourses on Twitter allow for an interpretive investigation of subjective perspectives. Second, a Q-methodological survey investigates subjective perspectives in a systematic and structured way (Davis and Michelle 2011), thereby solidifying the findings of this article.
Ideally, a concourse covers the full universe of perspectives on a concept (Van Exel and de Graaf 2005). I approach this as an ideal that motivates a maximum effort to not overlook viewpoints. I follow Sneegas in acknowledging that it is not possible to capture every single perspective, and that, rather, the results of a concourse (or a Q-methodological analysis) will always be based on partial, contingent, and situated knowledge and encounters (Sneegas 2020). In describing the wide variety of conversations that can be had on an issue, 'the concourse is the very stuff of life, from the playful banter of lovers or friends to the heady discussions of philosophers and scientists' (Nikraftar and Shokri 2014, 362). To capture these different settings, I populate the concourse on citizenship by drawing on two sources: citizenship theory and the discourse on citizenship on Twitter.
Drawing on both theory and empirics ensures that different conversations about citizenship in different spaces are reflected in the concourse. In selecting the sources, both mainstream and marginal viewpoints must be included. Equally, it is important that the concourse covers a broad array of issues that are linked to citizenship conceptually, and that these issues are invoked in a way that resonates with people. This will become particularly relevant when applying the concourse to an empirical study: Participants must be able to engage with the concourse. This requires the concourse to be rooted in knowledge or conversations of which participants are aware (Nikraftar and Shokri 2014). In the following, I explain how the sources I select for the concourse contribute to these aims.

Theoretical and empirical conflict lines on citizenship
To populate the concourse, I draw on theory and empirics. Several Q-methodological studies use both theoretical and empirical sources in their concourse development (e.g. Davis and Michelle 2011;Fontein-Kuipers 2016). Both sources contribute to an encompassing concourse that reflects how citizens talk about citizenship 'in the real world'. Each makes a unique contribution: Citizenship theory lends a clear structure to the concourse. It ensures that the concourse covers all relevant topical ground and that concepts historically linked to citizenship are not overlooked. A range of aspects such as rights and duties, commonality and identity, or political practice and belonging provide the basic structure of the concourse.
To represent bottom-up, subjective perspectives in the concourse, the empirical discourse on citizenship is explored. A social network, and Twitter in particular, is conducive to exploring bottom-up perspectives for two main reasons: First, on Twitter participants not only share but also make sense of and interpret information (Maireder and Ausserhofer 2014). Such meaning-making is particularly relevant for capturing subjective perspectives. In the setting of a social network, meaning-making also has an intersubjective component in that it happens in negotiation with others (Maireder and Ausserhofer 2014). It should be noted that only a select sub-group of the population engages in such meaning-making on Twitter. All the while, online activity has been shown to lead to information gains for users and non-users alike (Partheymüller and Faas 2015), meaning that there is an extent to which the common knowledge generated online is carried into broader public debates. Second, Twitter taps into alternative discourses (Kwon et al. 2016), and offers access to a wide range of viewpoints and conversations between them (boyd 2010). This makes it an efficient way to capture discourses and interpretations that are removed from top-down conceptions. Although Twitter does not yield a representative overview of viewpoints, it expands the concourse to a wider range of conversations on citizenship.
To consolidate both sources, I first identify conflict lines on citizenship in citizenship theory. I group those conflict lines into overarching categories. Then, I conduct a social media analysis on German-speaking tweets based on a range of keywords on citizenship. The final concourse on citizenship is a collection of conflicts discussed on Twitter, structured through and enriched by ideas from theory. After presenting this concourse, the article demonstrates an application of the concourse through a Q-methodological survey. The Q-methodological survey explores empirical perspectives on citizenship systematically. The items of the Q-methodological survey are drawn from the concourse, and thus are reflective of the most distinctive conflicts on citizenship. The article then presents the results of the Q-methodological survey in the form of four perspectives on citizenship. Because they are based on the concourse, these perspectives represent encompassing yet differentiated conceptualizations of how citizenship can be understood.

Theoretical conflict lines
In a first step, I identify major conflict lines in citizenship theory. I review the literature on citizenship theories, identify conflicts between normative approaches to citizenship, and group these conflicts into categories. To achieve this, I draw on citizenship theories from classic liberalism, liberal multiculturalism, republicanism, communitarianism, cosmopolitanism, and multiculturalism. Each of these theoretical approaches is encompassing and complex. Focusing on the most relevant conflicts between them clarifies the conversations that they imply.
For example, consider the abovementioned example of defining a German 'lead culture'. A 'lead culture' rests on the idea that citizenship requires citizens to have something in common. Commonality has a thin definition in liberalism, where citizens must only share their subscription to common laws. Republican or communitarian approaches have a much thicker definition that includes common political activity, ways of thinking, or socialization (Taylor 1994). These theoretical approaches define different extents, but also different types of commonality. How commonality is defined depends on the extent to which culture forms part of citizenship definitions, and on whether identity is understood as static or fluid, and as socially or individually construed (Kalu 2003). In turn, understandings of identity and culture are relevant to discussions on the content and application of justice to individuals and communities across the world (Tully 2014). The conflicts that emerge from these approaches are summarized below under the categories 'universal vs. relative justice', 'global and informal justice', 'basis of common life', and 'culture and identity'.
There is a wide range of further core political questions for which normative ideas on citizenship are relevant. The conflict lines highlight where theoretical conceptions define citizenship in fundamentally different ways. In reviewing a wide range of literature on the abovementioned theories, I identify the following conflict lines in eight categories: Universal vs. relative justice: Universal conceptions of justice understand a set of liberal rights to be applicable always and all over the world. Relative justice conceptions take justice to be dependent on factors like regional context, substantive equality for minority groups, and popular opinion (see e.g. Forst 1996;Kiwan 2012;Kymlicka 1995;Lister 1997;Sandel 1999;Tully 2014;Turner 1993).
Global and informal justice: Justice conceptions might advocate for a judicial space beyond the nation-state (see e.g. Carter 2001;Tully 2000). Alternatively, they may rely more heavily on social control and practice than on formalized justice (see e.g. Etzioni 2015).
Socio-economic equality: All approaches understand financial equality as relevant to overall equality between citizens, though different citizenship conceptions define different goals of financial equality (e.g. equal participation [see e.g. Lovett 2010] or individual liberty [see e.g. Marshall 1950]). This implies different extents to which financial equality should be accomplished (see e.g. Jayasuriya 2005).
Non-intervention and non-domination: Liberal citizenship conceptions focus on individual freedom which is achieved through minimal state intervention (see e.g. Zuckert 2007). Republican conceptions focus on the freedom of citizens from each other, with the aim of no citizen dominating another (or having the opportunity to) (see e.g. Grégoire 2014; Pettit 2013).
Individual and societal interests: Liberal citizenship conceptions prioritize the individual and its needs (see e.g. Silva 2012; Wallace 1999), while republican approaches prioritize societal needs (see e.g. Dagger 2002;Pettit 2013). Approaches differ in the extent to which individual pursuits should be limited by societal decisions.
Obligation: While civic obligation is an important virtue and requirement for good democracy in republicanism and communitarianism (see e.g. Arendt 1958;Reiner 2011;Selznick 1995), obligations are minimal and civic participation optional in liberal approaches (see e.g. Galston 2002;Kühler and Jelinek 2010).
Basis of common life: While liberal approaches see endorsing basic laws as the only commonality that citizens share, with individuals pursuing their own and plural ideas of the good life (see e.g. Erez 2017), republican and communitarian approaches aim for a more substantive common ground on which citizens agree to build their lives (see e.g. Thompson 2017). Multicultural approaches endorse commonality through common political practice (see e.g. Parekh 2016; Taylor 1994).
Culture and identity: While identity is fluid and culture multifaceted for some conceptions of citizenship, identity and culture are seen as static in others (see e.g. Kalu 2003;Stevenson 1997Stevenson , 2003. Conceptions differ in terms of the extent to which they view a common culture as relevant for individual identity and common citizenship (see e.g. Forst 1996;Kymlicka 1995;Selznick 1995).

Empirical conflict lines: a discourse analysis on Twitter
In a second step, I conduct a discourse analysis on Twitter to identify conflict lines between subjective perspectives on citizenship. Tweets were downloaded weekly with the help of the Twitter API between March 9th and May 10th, 2020. Each weekly download consisted of the 10,000 most recent tweets that mentioned one of a range of keywords. The search was based on eight keywords, including misspellings. While the keyword definition was informed by the conflict lines from theory, a conscious attempt was made to understand how people themselves shape discussions around citizenship.
To achieve this, the list of keywords consisted of general terms around citizenship, policy areas that mobilize thoughts on citizenship, and two concrete policy proposals from German public debate which can be tied to citizenship. Keywords on general terms around citizenship were citizenship and civic obligation. Selected policy areas that invoke contestation on citizenship were immigration, integration, asylum, and refugees. Two current debates with public traction in Germany were used as keywords: lead culture and compulsory year of service.
The search based on those eight keywords led to a total of 104.759 downloaded tweets (without duplicates). Tweets were filtered based on their relevance to citizenship. This resulted in a total of 1.160 tweets subject to content analysis using a data close reading approach to social media as proposed by Gerbaudo. It is based on close reading typical to interpretive approaches to text analysis, amended to suit the particularities of social media conversations (Gerbaudo 2016). By reading the data in three steps, different layers of meaning are uncovered: first, the topics and language used in the tweet itself; second the context or concrete conversation in which the tweet was sent; third, the broader environment of the discussion on social media. All of these are relevant for understanding a tweet and identifying the conflicts at stake (Gerbaudo 2016). Initially, I read tweets as rows, each devoid of context, to identify topics broadly. Then, tweets are read as parts of conversations by exploring attached links or responses. Finally, I look at the broader context of the tweets by exploring political events around the time (for example, a regional election in the federal state of Bavaria or the onset of the coronavirus pandemic).
Based on close reading of the data, 965 tweets are grouped by relevant conflict lines on citizenship. Those categories can be summarized as follows 1 : Justice concepts: Tweets emphasize the universal and unconditional nature of human rights. Many reference Germany's special responsibility for upholding human rights as a consequence of Nazism. There are many definitions of citizenship as an unconditional and legal status. At the same time, some tweets demand a stronger orientation of German legislation towards public opinion, or for the law to reflect national culture more strongly, thereby invoking concepts of relative justice. Many tweets demand more EU-level policy solutions or global schemes of justice. This is often combined with the idea that national citizenship is fundamentally unfair.
Equality: Many tweets emphasize an economic conditionality for citizenship. This includes tweets that connect citizenship with taking care of oneself and paying taxesregardless of where one lives or works. Economic factors are often connected to viewing migration as a financial threat to the nation. Tweets range from migrants exploiting social welfare systems to demanding that migration criteria be based on job prospects or finances.
Liberal citizenship: Tweets define the role of the citizen as being critical of government action, and being able to think for oneself. Thick civic duties are opposed, often on the grounds of individual autonomy or limiting state intervention. At the same time, ideas from liberal pluralism are invoked, such as the obligation to tolerate different views or to protect basic democratic freedoms.
Obligation and participation: Many tweets endorse civic obligation in connection to two types of activity: First, obligation is connected to participation in elections through voting or volunteering as an election official. Second, obligation is tied to broad civil engagement, like everyday acts of helping others, volunteering, or speaking out against hatred. Other tweets frame obligation more authoritatively: They invoke an obligation to follow the law (and even commands) and to take action when others break it. Some understand keeping public order as a specifically German duty. Other tweets advocate for differentiated civic responsibilities based on individual capacities or economic status.
Cultural conditionality: Tweets put a common culture, or a 'lead culture' at the center of their citizenship definitions, and advocate the protection of German culture. In this context, they often demand the integration or assimilation of migrants and interpret German culture as Christian. Some connect citizenship to loyalty, often in the context of rejecting multiple citizenships. There is quite a broad discussion between tweets about the degree to which citizenship and nationality are separate or whether they should be (more) congruent.
Global and multicultural citizenship: Tweets understand citizenship as geographical belonging, often pointing out that non-nationals already partake in and contribute to the German state. They emphasize European or multicultural values. Citizenship and national identity both are seen as open and dynamic; the concept of a unitary 'lead culture' is criticized. Tweets discuss the difference between nationalism and patriotism, sometimes endorsing patriotism.

Combining theoretical and empirical conflicts into a concourse
In the last step, I draw on the identified conflict lines from citizenship theory and the Twitter discourse to construct the concourse. Conflict lines from citizenship theory provide the basic structure for the concourse. Then, results from the analysis of conversations about citizenship on Twitter are sorted along the conflict lines identified in theoretical approaches. I group the conflict lines into five categories for the overall concourse.
Most of the conflict lines identified in normative definitions are reflected in those which appear in the Twitter discourse. Tweets include statements on the universal or unconditional nature of citizenship (rights and justice), on the extent of loyalty or commitment required to become a citizen (culture and identity), or on the importance of practice (obligation and participation) or financial contribution to the state (social and financial justice). Only the normative conflict between non-domination and non-intervention is added to the concourse without having been found in the Twitter conversations. Table 1 displays the categories and conflict lines of the concourse on citizenship. This is not to say that this is the only possible categorization. In the end, the categories represent one way of structuring the concourse. The main variables of the concourse remain the conflict lines part of each category, and these should be in focus. A different approach to the concourse might warrant a different categorization.
Investigating citizens' perceptions systematically: Q-methodology as an empirical application of the concourse The analysis of social media conversations on citizenship provides a first insight into subjective perspectives on citizenship. These perspectives are reflected in the concourse. To deepen insights into meaningful conflict lines, I now investigate bottom-up perspectives on citizenship more systematically by conducting a Q-methodological survey. Q-methodology is an approach to study subjectivity in a structured and replicable format (Davis and Michelle 2011). Q-methodological surveys draw their items, the so-called Q-statements, from a concourse.
The primary purpose of Q-methodological studies is to identify emerging perspectives, which are composed of several items. Q-methodology 'is primarily concerned not with the attitudinal variables but with the attitudinal patterns within and across individuals' (Pelletier et al. 1999, 139). This makes Q-methodology particularly suitable for the study of a complex and multifaceted concept like citizenship. In the following, I present and analyze the results from an online Q-methodological survey based on the developed concourse displayed in Table 1. The items which make up a Q-methodological survey (the Qstatements) are drawn from the concourse. All of the Q-statements have the same response pattern, making them comparable to each other. In the survey, participants place statements on a quasi-normal distribution. The sorting process asks participants to select statements they agree or disagree with most strongly, but also to sort out those statements that simply do not resonate with them (Watts and Stenner 2014).

Formulating and selecting Q-statements based on the concourse
It is recommended that Q-surveys use a minimum of 25 Q-statements, though it has been pointed out that the exact number will always depend on the study at hand (Watts and Stenner 2014). This survey includes a slightly lower but more manageable number of 24 Q-statements for two key reasons: This survey measures the perspectives of ordinary citizens in Germany. The survey is set up online to reach a wider variety of people. Both of these conditions, having ordinary citizens reflect on the complexities of citizenship, and doing so in an online setting, require significant concentration from participants.
As a first step, I issue several Q-statements based on the concourse on citizenship. This in itself is a lengthy process, because it requires phrasing the conflict lines in everyday language (Theiss-Morse 1993). I draw heavily on how conflicts on citizenship are invoked and discussed on Twitter to compose statements that reflect everyday talk about citizenship. In a second step, I randomly draw 24 Q-statements from the overall set of statements. I apply some limitations to this selection: Statements are drawn to represent the categories equally, in relative terms. The final Q-set includes three to eight conflict lines from each category.
Finally, I put the Q-set to a practical empirical test to evaluate whether the concourse resonates with people, whether it is understandable, and whether it is broad enough to be able to map distinct perspectives on citizenship. To achieve this, I ran an online pre-survey based on the concourse which yielded 33 valid responses. Participants of the pre-test were invited to give feedback on their sorting experience, the statements, and the overall survey. Based on the results, the Q-statements were reviewed again and participants' feedback was incorporated. For example, some expressed uncertainty on how to interpret the statements or noted words or phrases that they did not understand. Some respondents suggested including examples in the statements to help them respond more concretely. This was reflected in the final set of statements, also to address the fact that conflict lines surrounding citizenship are complex and abstract, which may render them difficult to access for ordinary citizens. Third, the revised statements were reviewed with another academic and presented to some participants of the pre-test and ordinary citizens who had not been part of the pre-test. Based on this review, the Q-statements were finalized. The final set of Q-statements is displayed in Table 2.

Design of the Q-methodological survey
The final online survey was conducted between 22nd September -8th October 2020. Participants were recruited via an online access panel, and compensated for their participation. The Q-methodological survey was part of a larger survey of around 30 minutes. Overall, 294 responses included a valid Q-sort. All respondents hold German citizenship, with 3.2% holding an additional citizenship. They are between 18 and 78 years old (with a mean age of 47.8 years) and in relatively equal parts female (52.0%) and male (47.6%). About a quarter has a higher education degree (24.1%), one-third has completed vocational training (30.6%), and the remainder has completed high school. About a quarter (23.5%) has a migration background, 2 and a quarter (24.9%) is from East Germany.
Participants ranked the 24 Q-statements on a Likert scale first and then sorted them in the Q-methodological design displayed in Table 3. This design was implemented via several multiple-choice questions (depending on the column of the normal distribution participants sorted for, they could select one, three, or five statements). The selection of statements was limited for the two outermost columns of the distribution: Only statements that evoked a positive or negative reaction in the Likert-scale rating of all statements were displayed. This prevented participants from becoming overwhelmed by the high number of statements.
In addition, the survey was set up to receive feedback from participants. An avatar named 'Sophie' led participants through the survey and gave instructions. After completing the sort, participants were asked to rate the understandability of the statements and were invited to make additional comments on their sort.

Analysis and results of the Q-methodological survey
The results were analyzed using the qmethod package for R. The analysis is based on a correlation of Q-sorts using the Pearson's coefficient. The data is then reduced by extracting a number of factors through principal component analysis, with each factor designating a similar sorting pattern (Zabala 2014). The extent to which participants subscribe to a factor is calculated through factor loadings (see Appendix 2). Participants German culture must be protected.

Cosmopolitan citizenship
We have to understand ourselves as citizens of the world first and foremost.
(1) Note: A similar table to display the Q-sorting structure is used in Hylton, Kisby, and Goddard (2018). whose Q-sort loads significantly onto an extracted factor are interpreted as subscribing to that factor (Watts and Stenner 2014).
The process of conducting a Q-method analysis also includes selecting an appropriate number of factors. Researchers have emphasized the importance of taking both statistical and content-based markers into account in selecting the final number of factors (Watts and Stenner 2014). The solution using four emerging factors yielded the highest percentage of variance explained by all factors (47.4%). The full table of factor loadings can be found in Appendix 2. Solutions using two and three factors had a slightly higher number of sorts that load significantly onto a factor (25 more sorts for both solutions). I compared the interpretations of the results using two, three, and four factors to select the final result using four factors. Table 4 displays the four resulting factors. These four emerging perspectives on citizenship were given a name summarizing their core position for easier identification. The table shows the average ranking of each Q-statement within the quasi-normal distribution Table 4. Average ranking of the Q-statements of the four emerging factors f1, f2, f3 and f4. f1 f2 f3 f4 Some rights are so generally applicable that they should be implemented all over the world and for every individual.
−1 3 1 2 For some societal groups (e.g. people with different ethnicities or religions) special rights are necessary so that their standpoint is not overlooked by the majority.

−3 −1 −2 −1
It would be good if more decisions were made on the European level. −2 0 0 −1 Germany has a special historical responsibility for protecting human rights.
−2 2 −2 −1 Every person is responsible for their own economic situation.
1 −1 2 −3 I am worried that German citizens have to compete with immigrants for social welfare (e.g. employment benefits or pensions).

−−1 −2
The concept of national citizenship is fundamentally unjust. 0 −2 −2 0 The state too often limits my personal freedom.
1 −3 −3 0 The individual, not the community should be the first priority of our state. 0 −2 −1 0 Citizens always have to be observant and critical towards the state. 1 0 −1 2 If we know our roots, we have a better orientation in the world. 1 0 −1 −2 For me, being a citizen does not only mean having rights but also having obligations toward the state and society. We have to understand ourselves as citizens of the world first and foremost. −1 1 1 3 Note: For each of the four factors (f1, f2, f3, f4), the table indicates the average ranking for each statement. That ranking is obtained by calculating z-scores, which are the weighted average of the scores given to a statement by participants who subscribe to each factor. The z-scores are then rounded towards the discrete values part of the Q-sort grid (see Zabala 2014). In this Q-sort grid, participants could rank statements by endorsing them (i.e. ranking them in position 3, 2, or 1, with 3 indicating the greatest level of endorsement), rejecting them (i.e. ranking them in position −1, −2, or −3, with −3 indicating the greatest level of rejection), or not selecting them as relevant to their definition of citizenship (i.e. ranking them in position 0). Overall, 85 Q-sorts (i.e. participants) load significantly onto Factor 1; 62 on Factor 2; 25 on Factor 3; and 19 on Factor 4. displayed in Table 3, with a ranking of 3 indicating the most preferred and −3 the least preferred statement within the factor. In the subsequent text, the factors are referred to by a name summarizing their position: critical ethno-culturalist (Factor 1), active democrats (Factor 2), liberal democrats (Factor 3), and cosmopolitans (Factor 4). In the following, each of the four perspectives behind the four factors displayed in Table 4 is briefly summarized. The summary draws on the average ranking of each Q-statement within each perspective, which is indicated throughout the text in brackets.
Factor 1: Critical ethno-culturalists' citizenship conception rests strongly on cultural factors. They are the only group to believe that 'German culture' needs to be protected (2), and they strongly demand assimilation from immigrants (3). They are worried about competition for social rights through immigration (2), they do not feel that non-citizens living in Germany belong (−2), and they are the group to most strongly oppose minority rights (−3). In line with this, they oppose further EU integration (−2) and the conception of universal rights generally (−1). They also reject the idea that Germany holds special responsibility for safeguarding human rights (−2). They do not feel connected to people in other countries (−1) and believe that gaining orientation in the world rests on knowing one's roots (1).
This group includes duties in their conceptualization of citizenship (2) and thinks it problematic if not everyone participates politically (−1). At the same time, their point of emphasis for duties does not seem to rest primarily on political or civic participation: the group does not wish for more engagement (0) and does not allocate higher importance to political than to other forms of voluntary engagement (1). This might be connected to the critical distance this group maintains towards the German state: They think of the state as freedom-limiting (1) and believe citizens should be critical towards the state (1). Despite their closeness to German culture, they are not necessarily proud of German democracy (0).
Ethno-culturalists also think of every individual as responsible for their economic situation (1). In their conception of citizenship, they seem to include people who share culture, work, and everyday lifealbeit with distance towards the state, immigrants, and multilateral institutions.
Factor 2: Active democrats are fond of many aspects of German democracy: Universal rights are very important to them (3) and they support Germany's responsibility for safeguarding human rights (2). They strongly reject the idea that the state limits individual freedom (−3) and are proud of German democracy (2). This pride also manifests in the emphasis this group places on political participation: They are the only group to think of political participation as a learning experience (1). They want citizens to be more engaged (1) and value both political and broader civic engagement (1). This group strongly prioritizes common interests above individual ones (−2).
Active democrats are relatively open: They think someone belongs as soon as they live in Germany (1), do not think 'German culture' needs protection (−1), and do not feel that the social welfare system is threatened through immigration (−2). They are the only group that wishes to resolve conflicts based on culture and religion within politics (−1).
Cultural definitions of citizenship mean little to this group. Rather, they think of themselves as world citizens (1). The group also understands economic inequalities as a structural rather than an individual issue (−1). However, they do not consider the concept of national citizenship to be unfair (−2). Again, this may be connected to the fact that they value the German nation-state and its opportunities for political participation.
Factor 3: Liberal democrats most strongly favor a citizenship conception that emphasizes duties (3). However, these duties do not seem to manifest in political participation: The group has no position on whether citizens should be more actively engaged (0) or on the extent of political participation (0). Broad civic participation is just as meaningful to them as political participation (1).
The group seems to place more emphasis on individual private obligation: They favor the idea that each individual is responsible for their economic situation (2) and demand assimilation from immigrants (2). Consequently, simply living in Germany is insufficient for genuine belonging (−1). The group also rejects special minority rights (−2). Beyond this obligation to care for oneself and adapt to society, cultural factors do not play a significant role for this group. They feel just as connected to individuals from other countries as to German citizens (1) and do see themselves as world citizens (1). All the while, they reject the idea that Germany holds special responsibility for human rights (−2) and do not think that national citizenship is unfair (−2). Generally speaking, the group does seem to be open to international cooperation and connectedness, but within the structure of the nation-state.
This may be connected with this group's positive view of the German state. They strongly reject the idea that the state limits their freedom (−3) and are proud of German democracy (1). They do not think citizens should be critical of the state (−1). It seems that this group connects citizenship with individual-level responsibility and activity, and with a (possibly passive) acceptance of the nation-state. They accept both migration and international cooperation, so long as the order of and within the nationstate is maintained.
Factor 4: Cosmopolitans see themselves as citizens of the world first (3). They favor thin commonalities attached to citizenship (1), accept geographical belonging (1), and feel just as connected to people in other countries as to German citizens (1). This group is most strongly opposed to the idea of people needing their roots to find fulfilment (−2). They reject pride in German democracy (−2) and are opposed to protecting 'German culture' (−1).
Cosmopolitans take an open and accepting approach towards others in several ways: they are concerned about structural economic inequality (−3). The group is not worried about immigration leading to competition over social rights (−2). However, they do not necessarily favor institutionalizing minority rights (−1), and consider issues around culture and religion to be private (2). This might be due to their distance from the state: The group associates good citizenship with taking a critical stance towards the state (2). They are against further EU integration (−1) and are opposed to Germany taking special responsibility for human rights (−1)however, they strongly believe in universal rights (2).
Although the individuals in this group think it is a problem if not everyone participates politically (−1), they have no position on other statements on increasing participation. They seem to endorse a liberal cosmopolitanism that connects humanist acceptance with mobility and openness without an institutionalized structure.
Discussion of the results of the Q-methodological survey Q-methodological surveys typically allow the identification of distinguishing statements (that significantly differentiate factors from one another) and consensus statements, positions that the sorts within all factors agree on. The analysis presented here yields no consensus statements, which demonstrates that citizenship definitions are highly contested not just in theory but also in practice.
Nonetheless, some common tendencies can be identified in the sorts: There is relatively large agreement on obligation and participation. Particularly, all four groups understand citizenship as more than a status: Citizenship comprises not only rights but also obligations. The form of obligations differs between perspectives: Obligation means safeguarding culture and way of life for ethnocultural citizens, political participation for active citizens, adherence to laws and the maintenance of institutions for liberal citizens, and a humanist duty towards others for cosmopolitan citizens. Across citizenship perspectives, citizenship is connected with participation and civic engagement, and three of the four groups agree that it is unacceptable if not all citizens participate politically.
There are also some common tendencies on justice and culture: All of the factors indicate a negative reaction towards implementing minority rights for religious or cultural groups. Additionally, two factors (ethnocultural and liberal citizens) favor assimilation by immigrants, with the two factors that entail a more open and inclusive conception of citizenship (active and cosmopolitan citizens) not necessarily opposing assimilation. This implies the importance of shared guidelines for citizenship. These could be in the form of cultural aspects, but it also seems that the validity of universal rights plays a relatively prominent role in German citizens' perspectives.
Additionally, while ethnocultural citizens reject humanist connections across borders, the three remaining groups do also see themselves as citizens of the world. There does seem to be a general notion of feeling responsible for or belonging to a space beyond the nation-state. However, this is not necessarily connected with an endorsement of global institutions: Further EU integration is rejected by cosmopolitan citizens, and not endorsed by active or by liberal citizens. Cosmopolitan citizens may view the German nation-state and the EU as incapable of delivering the interconnected world they envision. Humanist connectedness does not seem to be limited to Europeand there does seem to be a preference for informal rather than institutionalized cooperation.
There are some limitations to the results presented here. Q-methodological surveys are not representative; results can only depict an individual's perspective in the context of the specific time and situation of the investigation. Additionally, this Q-methodological survey is unique in two aspects. First, only a few Q-methodological studies (e.g. Pelletier et al. 1999) have worked with a large number of participants before. A large number of participants leads to inflated Eigenvalues (Zabala, Sandbrook, and Mukherjee 2018), so these were not used for choosing a fitting number of factors. It also led to quite some sorts not loading significantly onto any of the identified factors. This makes it possible that there are further conceptions of citizenship or further complexities that were not fully captured in this investigation.
Second, many scholars conduct Q-methodological surveys face-to-face to gather an impression of how participants engage with and explain their sorting process. This insight was limited in the study presented here. However, 50.0% of participants rated the statements as very understandable, and 49.3% as partially understandable. When asked to comment on their sort, a handful of people said they found the survey a good way to communicate their perspective on citizenship. Some said they had to read statements several times or think about them to form an opinion, especially since some statements offered room for different interpretations. The majority however did not comment on their sort.

Conclusion
In this article, I developed a theoretically and empirically informed concourse on citizenship by drawing on two different sources: citizenship theory and German-speaking tweets on citizenship. In this sense, the concourse serves as a comprehensive overview of conflict lines on citizenship across different spaces. Although a concourse is meant to cover the full range of perspectives on an issue, I acknowledge that it can never be fully exhaustive. There may always be aspects that are overlooked or categories that could have been formed differently. I suggest using this concourse as a starting point to begin work that evolves and adapts through further discussion and research.
The concourse included conversations on citizenship identified on Twitter. The Twitter analysis conducted in this study yielded diverse arguments on core political issues. Some arguments also resonate with key themes from the academic debate on citizenship, indicating a fairly high level of sophistication of social media discussions.
I then explored subjective perspectives on citizenship in Germany systematically via an online Q-methodological survey. The concourse made for an effective basis for an empirical investigation on subjective perspectives. The statements formulated based on the concourse invited distinct and nuanced reactions by participants. The perspectives identified based on the Q-methodological survey made the multifaceted and complex nature of citizenship apparent.
The Q-methodological survey proved to be an effective method to make relatively distinct yet complex perspectives on citizenship visible. Its results resonate with empirical results found elsewhere: The first emerging factor (ethnocultural citizens) aligns with the concept of ethnic citizenship that puts culture, common tradition, and ancestry at the forefront of its definition. Meanwhile, all of the other factors bear some qualities connected with a civic conception of citizenshipall three embrace universal rights, for example. But they do so in very different constellations, and with different points of emphasis on participation, on following laws, and on global connection. This resonates with the previous finding that civic conceptions of citizenship are multifaceted (Wright, Citrin, and Wand 2012). Additionally, the survey results demonstrate that citizenship is understood to be more than a legal status: Something shared between citizens is attached to all perspectives, be it through thick conceptions of shared ancestry, active participation, thin conceptions of shared institutions, or humanist connection.
This article thus contributes to the understanding of citizenship policies and the everyday practice of citizenship in Germany. The results explain the relevance of discussions around commonality (e.g. in the form of the debates around a 'lead culture'), or around obligation (e.g. in the form of the proposal for a compulsory year of service). There is a relatively broad understanding that citizens want connectivity from citizenship. These results tap into a long-standing conflict between commonality and pluralism in modern citizenship (see also Duchesne 2003). In applying these results in policy and practice, it will be necessary to think about ways for citizens to feel connected to a common project in a civic frame. This analysis has shown that citizens themselves have different ideas on this, ranging from thick political participation to a thinner but tangible dedication to institutions. It seems useful to bring these perspectives into conversation with each other to seek out commonalities, compromises, and conflicts. Developing ideas for public spaces or common practices could be a useful starting point for making a unifying experience of citizenship possible in pluralistic societies.
These results differentiate and specify what we know about individual understandings of citizenship. In particular, they demonstrate that it is useful to study contested concepts like citizenship from a broad starting point, granting participants the room to make meaning of their perspectives. Q-methodology and other subjective methods not only offer an expansion of the methodological toolkit on measuring citizenship. By developing a concourse, for example, aspects can emerge that traditional tools capture insufficiently. This methodological approach has shown that a concourse can add to the scope of subaspects of citizenship that we even consider relevant. In these terms, this article's academic contribution is twofold: First, the discursive methods employed here could be applied in other country contexts, or for the exploration of other concepts of political science. Second, its findings can be developed in future studies on citizenship. They imply that it is relevant to further interrogate different definitions and meanings of obligations and to investigate the different variants of civic citizenship.
This study demonstrates that to more fully understand individuals' perspectives on citizenship, we require a multitude of methods. More recently, there have been more bottom-up investigations on citizenship. To deepen their results, we also need to collect more subjective and discursive data to understand how people talk about, practice, and live citizenship.