Resisting assimilation – ethnic boundary maintenance among Jews in Sweden

ABSTRACT This article evaluates Andreas Wimmer’s theory of ethnic boundary making by applying it to the maintenance of Jewish ethnic identification in Sweden, as expressed in interviews with Swedish Jews. Wimmer proposes that ethnic conflict routinizes and entrenches perceptions of ethnic difference; we argue that the antisemitic persecutions of the twentieth century have entrenched the perception of the ethnic distinctiveness of Jews among Jews themselves. These persecutions also contribute to alienation from Swedish society, which does not share the same frames of understanding. These factors motivate the interviewees to maintain the ethnic boundary between Swedes and Jews and guard it against assimilation. We propose a nuancing of the debate between instrumentalist and primordialist conceptions of ethnic identity by arguing that while our interviewees express a taken-for-granted view of their ethnic identities, they advance ethnic discourse strategically in order to protect the Jewish community from losing its distinctness, especially through assimilation.


Introduction
The academic debate on ethnicity over the last few decades has largely focused on the opposition between essentialism and constructivism, the latter paradigm now being almost entirely dominant. But while the constructivist literature (for a review, see Brubaker 2009) has emphasized that ethnicity is malleable, situationally defined, and constructed on the micro level in day-to-day life, it has often failed to place such social construction in a historical context and explain why we see such large variation in how much, and how, ethnicity matters for individuals and societies. Andreas Wimmer (2013, 3) accuses this literature, inspired by post-modernism, of being 'hyperconstructivist', overstating the malleability and situatedness of ethnic identity. To remedy this, Wimmer offers a comparative analytic of ethnic boundary making that attempts to understand why ethnicity is sometimes fickle and changeable, but sometimes stable and taken for granted. Wimmer claims to provide a deeper understanding of the relationship between culture, ethnic self-understanding, and the historical background of ethnic conflict. In this study, we will evaluate this approach by applying it to the case of Jewish identity in Sweden. We will focus on evaluating one of the most contentious aspects of Wimmer's theory and of the boundary making paradigm in general: the suggestion that ethnic boundary making ought to be analysed as strategic action.
Jews in Sweden are generally socially, culturally and economically well integrated into mainstream society (Dencik 2009). In many cases, they are not a visually identifiable minority (Nylund-Skog 2006), meaning that for many Jews it would be possible to fully assimilate, i.e. to be absorbed by the mainstream, an option that is not open to all ethnic minorities, neither in Sweden (Bursell 2012) nor in other contexts (e.g. Lamont et al. 2016;Mizrachi and Zawdu 2012). Yet, Jews across contexts have been shown to hold on to their Jewish identities (cf. Rapaport 1997;Buckser 1999). The subject of how European Jews view their ethnicity and their place in the national community has, however, not been given extensive attention by social scientists in the constructivist tradition, which tends to focus on more stigmatized or socioeconomically disadvantaged categories. The studies on Jewish identity that have been carried out, moreover, predate some prominent changes in Europe's ethnic dynamics including significant immigration of groups that have a history of antagonism with Israel, the continued development of discourses of multiculturalism, and the resurgence of far-right movements; a situation which has resulted in dramatic increases in perceived levels of threat among Jews in Sweden as well as in other contexts (FRA 2018). The purpose of this study, then, is twofold: to evaluate and expand upon Wimmer's ethnic boundary making approach while at the same time shedding light on the ethnic boundary work of Swedish Jews. Our primary research questions are the following: . How can we understand the maintenance of Jewish ethnicity in Sweden by applying Wimmer's boundary making perspective? . How can we refine this perspective by studying the case of Swedish Jews?
Through addressing these questions, we make: (i) a theoretical contribution concerning the applicability of the strategic mode of analysis of ethnic boundary maintenance, as well as (ii) adding to the understanding of Swedish Jews' ethnic boundary work today.
The paper is organized as follows: first, we outline our theoretical approach. We then account for previous research on Jews in Sweden and Jewish identity and outline the methodology. Thereafter, we present the empirical analysis and the concluding discussion.

Theoretical perspectives
In this section, we will first summarize the boundary making paradigm, particularly as it is formulated by Wimmer (2013). We will then discuss the debate surrounding the strategic interpretation of boundary making, and present Wimmer's response to criticism of that interpretation.

Ethnic boundary making
In the constructivist literature on ethnicity, the concept of boundary making strategies has often been employed to talk about the processes through which ethnic groups are cognitively and socially delineated (Eriksen 2010;Zolberg and Woon 1999;Alba 2006). Andreas Wimmer (2013), drawing on Fredrik Barth (1969), argues that to understand ethnicity and ethnic identity, we must focus on the boundaries which these phenomena imply: Each identification ('I am Swiss') obviously implies a categorical boundary (the non-Swiss); each corresponding action (e.g. helping another Swiss to find an apartment in Los Angeles) implies discriminating against those on the other side of the divide (i.e. not helping someone from Sweden). Focusing on social and categorical boundaries allows us to study the formation and dissolution of ethnic groups with more precision than standard sociological approaches that take the existence and continuity of such groups and categories for granted. (Wimmer 2013, 3) This approach, then, takes seriously the connection between individual identity and group boundaries. It argues that an 'individual' identity as, say, Jewish, only makes sense on the assumption that the group Jews is somehow important, and in turn that self-identification reinforces the boundaries of that group.
Ethnic boundaries contain both cognitive, or 'categorical' aspects, and 'social or behavioural' aspects; 'The former refers to acts of social classification and collective representation, the latter to everyday networks of relationships that result from individual acts of connecting and distancing' (Wimmer 2013, 9). Moreover, Wimmer argues for a distinction between different modes of making categorical distinctions between people, and the means of making one's 'vision of the legitimate divisions of society relevant' (Wimmer 2013, 63), that is, to attempt to change, or preserve, society so that it will be structured in accordance with the ethnic divisions one considers legitimate and important. When boundary making is successful, it produces a social world structured according to one's preferred categorical distinctions, and, we should emphasize, a world in which these distinctions are shared by all and considered 'natural' and reflecting the inner 'essence' of the persons they sort. The modes of boundary making are: attempts to either shift an existing boundary to a more inclusive ('Expansion') or more exclusive level ('Contraction'), 'Transvaluation'to change the hierarchical ordering of ethnic groups, 'Positional moves'to change positions in an existing ethnic order, and 'Blurring'to de-emphasize ethnic boundaries altogether (Wimmer 2013, 50-58). The means of boundary making are: (1) discourse and symbols; (2) discrimination; (3) political mobilization; and (4) coercion and violence. Wimmer argues that violent conflict is the most effective way of sharpening ethnic boundaries and entrenching perceptions of difference. 1

Boundary making as strategic action
There has been much debate about whether it makes sense to view ethnic self-identification as strategic in the way that Wimmer and others do, rather than taken for granted and expressed in 'good faith'. Barth (1969) went a long way in the direction of rationalist instrumentalism, viewing ethnic classifications as strategies undertaken by rational actors with clear goals in mind. This aspect of the boundary making paradigm has been criticized (cf. Gil-White 1999;Eriksen 2010, 63-9). Lamont (2014, 816f) argues similarly that the strategic component of Wimmer's approach is misguided since individuals 'are rarely conscious that they inhabit categorization systems. Instead, they tend to use schemas that are largely taken for granted and made available by the national cultural repertoires that surround them'. But Wimmer's position is not reductively rationalist. Rather than establishing a priori either that ethnic identification is open for choice or is always set in stone from birth, Wimmer incorporates the continuum between these two extremes in his theoretical model: sometimes ethnic distinctions are 'taken for granted, routinized and "constitutive of reality"'; but sometimes 'they represent classificatory elements to which individuals maintain considerable reflective distance'. Moreover, 'even when an ethnic or racial boundary has solidified into a taken-for-granted schema, […] individuals negotiate strategically what it exactly means, in each micro-minutiae of an encounter ' (2014, 840) to belong to a particular group. It should also be emphasized at this point that the goals sought in strategies of boundary work need not be exogenously given, that is, independent of previous boundary struggles. Once we come to identify with a certain group, we may, of course, be willing to 'incur high costs to defend the culture and honour of [our] community' (Wimmer 2013, 104;cf. Gil-White 1999, 805-807). Far from being a 'rational choice' theory, then, Wimmer's conception of strategic action is much closer to Bourdieu's account of strategies for 'symbolic profit' (e.g. Bourdieu 2010), suggesting a feedback process between the classificatory struggles forming the categories through which we understand ourselves and the rewards sought through such struggles. At the end of this paper, we will argue that our data suggest a way of adding nuance to this debate while still retaining the framework of strategic analysis.

Jews in Sweden and previous research on Jewish identity
Jewish immigration to Sweden goes back to the 1770s. 2 At that time, immigration was restricted to well-to-do Jews. Jews were not granted full citizenship until 1870, after which a considerable number of Russian and Eastern European Jews came to Sweden fleeing poverty and pogroms. The refugee wave caused a considerable gulf between the newly arrived poor Eastern European Jews and the well-established and integrated Jews from Germany. During the 18th and 19th centuries, the strategy had been assimilation, to reduce Jewishness to the religious sphere, with the goal to present the Jews as 'Swedes of Mosaic beliefs'. But the Eastern Jews, in general, shared none of these intentions. It was in this group that Zionism, the movement to create a Jewish state, first manifested itself in Sweden, advocating that Jews should be seen as a nationality in their own right. But if Jewishness was a matter of nationality, this seemed to imply that Jews were not fully Swedish. For this reason, many leading Jews, including the leadership of Stockholm's Mosaic Congregation, fiercely opposed Zionism, fearing it would roll back the progress made in making Jews Swedish. It was not until the Nazi rise to power in Germany that the Congregation and many assimilated, 'western' Jews in Sweden realized that national loyalty and patriotism had been unsuccessful in hindering antisemitism. In this group, too, it became more common to embrace Zionism. Sweden's role as a safe haven for Jews and other refugees fleeing from the Nazis during World War II is contested. Sweden did not open its borders for Jews until the end of the war; in 1943 for Scandinavian Jews, and in 1944 for European Jews. Critics argue that this dramatic change in refugee policy was a pragmatic response to signs of the Nazi's losing the war. Others argue that it was a result of an increased awareness of the situation of Jews due to Germany's persecution of Jews in neighbouring Norway and Denmark (Kvist Geverts 2008).
Some 4000 Jews who came to Sweden as refugees stayed after the war. Since then, the biggest wave of Jewish immigration has been from Poland, due to its antisemitic campaign in 1968. Ilicki (1988) describes the changes in ethnic identity for this group. In Poland, many Jews had been heavily assimilated but when antisemitism flared up, they realized that they were not considered Poles by the surrounding society. In Sweden, a majority of these refugees became much more invested in their Jewish identity and took up Jewish cultural practices that had been unknown to them before migration. More recent survey studies performed with members of Stockholm's Jewish congregation suggest that Jews in Sweden in general view their Jewishness as mainly a matter of ethnicity and not religion (Fischer 1996;Dencik 2007), a conception of Jewishness that they share with Jews in other parts of Europe and North America (e.g. Chervyakov, Gitelman, and Shapiro 1997;Meyer 1990;Webber 1997). It will also be noted that since Jews have arrived in Sweden from different countries and through very different contexts of migration, Swedish Jews should not be seen as one unambiguously delineated ethnic grouprather, Jews in Sweden make up a diverse group who only have in common their connection, in some way, to the global Jewish community (cf. the arguments against 'groupism' in Brubaker, Loveman, and Stamatov 2004) However, since Swedish Jews are primarily of European descent, it is likely that they are of Ashkenazi rather than Sephardic or Mizrahi descent (cf. Nylund-Skog 2006).
The political dynamics of antisemitism and Zionism have played a large part in defining not only the meaning of Jewishness in Sweden, but in the modern world at large. 3 Drawing on in-depth interviews with Jews in Germany, Rapaport (1997) emphasizes that ethnicity is created through people's ethnicized distinctions and perceptions. She emphasizes the importance of the Holocaust in the construction of Jewishness: 'the Holocaust plays a central role in contemporary Jewish life in Germany, indeed, […] it is the element on which ethnicity is based' (Rapaport 1997, 14). Rapaport's participants expressed a great deal of mistrust in and alienation from German society, by viewing the 'German character' as cruel and conformist. According to Rapaport, then, the mistrust and fear that has followed the Holocaust is the principle perpetuating the German-Jewish boundary. However, Buckser (1999), studying Jewish identity in Denmark, emphasizes that the tension inherent in identifying both with Denmark and Jewishness, leads to individualized historical narratives of the relationship between Jews and Danes during the war.
The position of Jews in modern multicultural Sweden Sweden has made the transition from a comparatively ethnically homogenous to a country of immigration in a fairly short period of time. In 1940, only 1% of Sweden's population was foreign born. Today, 20% of the population are foreign born, and 5% are native born to two foreign born parents. The largest countries of origin among the foreign born are Syria, Iraq, Finland, Poland, Iran, Afghanistan, Somalia, former Yugoslavia and Bosnia-Herzegovina. In addition, Sweden has five national minorities: Jews, Roma, Sami, Tornedalers and Swedish Finns. These groups have a long history in Sweden, and are thus not identifiable in Swedish registers, as data is only collected on the inhabitants' country of birth; not on ethnicity, race or religion. Estimates, however, suggest that there are 20,000 Jews in Sweden. Thus, Jews constitute a fairly small ethnoreligious minority in multi-ethnic Sweden.
Multiculturalism was incorporated into the Swedish political model in the mid-1970s, replacing an assimilationist approach, in part as a result of arguments forwarded by the Estonian and Jewish diasporas (Wickström 2015). Integration policies have since then revolved around promoting economic and legal equality and the opportunity for minorities to retain their original cultural identity (Hammar and Lindby 1979, 10). There is thus no institutional pressure to assimilate culturally in Sweden. The national minorities, Jews included, have enhanced rights in terms of the preservation of culture and language, including the right to speak the minority language in school, in contacts with officials, in health and elderly care.
Since there are no official records of ethnicity or religion, it is difficult to provide reliable evidence on how Jews have fared in terms of political, cultural, social and economic integration. The information available suggests, however, that Jews in Sweden are participating in all of these societal spheres (Dencik 2009). This is in line with the general pattern that individuals with a background in Europe or Western regions tend to integrate rather smoothly, while those from non-Western regions are overrepresented in the segments of the population that are worst-off socioeconomically (Aldén and Hammarstedt 2014).
As mentioned above, Swedish integration policy aims at integration, not assimilation. In terms of retaining original culture, it is, again, difficult to find objective measures on to what extent Jews in Sweden maintain Jewish culture. Criteria such as the use of a shared language does not capture the vitality of Jewish culture that well, since Swedish Jews have origins in many different countries, and there is no data available on intermarriage rates. While participation in religious activities is also a crude measure of group identification, 6000 of Sweden's Jews are members of the Swedish Jewish Congregation.
The Swedish population has for decades stood as the most tolerant population in Europe, in terms of support for multiculturalism as well as solidary immigration policies. A compilation of several different opinion polls indicates, however, that attitudes to immigration and multiculturalism have turned increasingly negative in the last decade (Höjer 2020). This attitude change is also reflected in political elections. A right-wing extremist party with roots in neo-Nazi movements, the Sweden Democrats, was elected to the Swedish parliament in 2010, reaching 17% of the popular vote in the 2018 election. While the anti-immigrant rhetoric in Sweden is primarily directed toward Muslims (Mulinari and Neergaard 2012), antisemitism is also growing in Sweden (FRA 2018) and so is reported antisemitic hate crimes (Brå 2019). Today's antisemitism is practiced by right-wing extremists as well as Islamic extremists (Brå 2019).
To sum up, Jews are an established national minority in Sweden with a long history in the country. They are of predominantly Eastern European descent and are well integrated into Sweden's political, cultural and economic life. At the same time, they are subject to antisemitism, threats and hate crimes.

Methodology
The data consist of 16 semi-open/thematic interviews that were carried out in 2015-2017 and lasted between 30 min and 2 h. The first three interviews were entirely inductive in character. These pilot interviews influenced further theoretical readings and gave a clue about what to expect in further interviews, but care was taken not to let these expectations unduly influence the questions asked. The interviews, conducted in Swedish, followed a series of predetermined themes such as the interviewees' views of: (i) what Jews have in common, (ii) integration/assimilation 4 ; (iii) antisemitism (iv) the situation of Jews in Sweden; and (v) their relationships to Jews and non-Jews. Most interviews started with the purposefully vague question: 'What does it mean for you to be Jewish?'. The aim was to follow the course of conversation and allow the interviewees' conceptions of Jewishness to emerge rather than those of the interviewer. The interviewer (one of the authors) is Jewish, and some interviewees made it clear that this made them more comfortable in telling him about some of their experiences.
The first set of interviewees was found by posting requests in Jewish-related Facebook groups. 5 Thereafter, snowball sampling, with several starting points, was employed. The ambition was to find a heterogeneous set of respondents with respect to age, gender, family histories, degrees of religious and community involvement, and places of origin. Thus, the sampling has aimed for the greatest possible variation rather than proportional representativeness (Boeije 2010, 36f). Our objective is to understand how ethnic boundaries are maintainedwe are not claiming to gauge the level of intensity of ethnic feeling of Stockholm Jews as a group, or to make inferences as to the proportions of Stockholm's Jews that hold certain views. If that had been our goal, our current way of finding interviewees would have constituted a case of sampling on the dependent variable, since we have sampled our interviewees on the basis that they identify as Jewish. We have rather attempted to find out: where Jewish self-identification does manifest itself, how can we understand it? It may also be noted that it is beyond the scope of this study to directly address the question of how the Jew-Swede boundary is co-constructed by non-Jews.
Analysis of the data was carried out during the data collection process, in a reiterative process between theory development, coding and data collection (Boeije 2010;Aspers 2011). This also facilitated a theoretical sampling: as the interview transcripts were coded and compared to each other, it became clearer along what dimensions it may be important to look for variation in cases (Boeije 2010, 112f). The interview transcripts were initially subjected to an inductive coding. A coding scheme was then developed by a combination of the most salient codes from the open coding and codes derived from the theoretical framework (cf. Boeije 2010). The most relevant codes turned out to be; history, experiences of antisemitism, Swedish identity, relationships to non-Jews, intergenerational transmission, and the relationship to Israel. These codes do not correspond to the themes of the interview guide, since care was taken to ask open questions which did not simply confirm our preconceived notions. Below, we interpret the predominant themes in the interviews through the lens of Wimmer's boundary making approach.
To elaborate on how we have thought about these interviews; recalling Wimmer's taxonomy of means of boundary making (page 3 above), it can also be seen that the interviews are useful in two quite distinct ways; first, the interviewees' accounts are themselves examples of the discursive construction of boundaries. But we also use the interviews to gain substantial information about things the participants have experienced or otherwise have knowledge of, including other means of boundary making, e.g. by learning how the participants have been influenced by their backgrounds. We have, then, seen these interviews both as examples of boundary making discourse, and as sources of various relevant facts.

Analysis
Transvaluation through pictures of the past It was common among the interviewees that they understand their own Jewishness as intimately tied to the history of their family, which was in most cases a history of oppression. Aaron, aged 38, for example, views with pride his family's tenacity in On this theme, Ben, a 45-year-old actor and comedian, made many references to the 'bond' between Jews that is a result of their historic experience of being a heavily persecuted minority. Oppression is here reinterpreted as something that builds resilience, character and solidarity. Numerous interviewees even explicitly cited the memory of the Holocaust as a reason to maintain and pass on their Jewish identities and to honour Jewish customs and culture. Many felt that doing so constituted a symbolic victory against the Nazi genocidaires and that it honoured those who had been killed for being Jewish. For Aviva, aged 40, this took the form of a feeling of duty towards those murdered among her family members, and a feeling of guilt when she did not observe Jewish customs and rules. When asked about what her Jewishness means for her, she immediately began to talk about the Holocaust, moving remarkably quickly from 'belonging' to 'duty', 'coercion' and 'guilt': I think it's a really big part of my identity. But I'm not religious, so it's, like, a feeling of belonging, but also some kind of feeling of duty. (…) I have very little family left because most were put in concentration camps, so there were not many who survived. So, I have probably always felt that a large part of the Jewish identity is some kind of duty towards those who [did not survive] that I need to be a good person and to maintain Jewishness, that it would be messed up to just let it go, towards them, even if they in practice perhaps would not … no one would actually be angry, I think, but it still feels like I have to. There is some kind of duty, some kind of coercion, in Jewishness for me now.
The commitment to fallen family members was cited as a motivation to not only maintain a Jewish identity, but to even feel obligation towards religious precepts, even by individuals who explicitly described themselves as non-religious.
It seems, then, that there is a great need for many of the interviewees to engage with the past by interpreting the stories of their predecessors as heroic and inspirational, in a way that fits well with what Wimmer calls transvaluation; 'reinterpret[ing] historical defeat and subjugation into a heroic struggle against injustice and domination' (2013,58). Jewish resistance and resilience in the face of oppression is honoured; persecuted predecessors are revered for their tenacity and strength. Later, we will argue that one of the factors that causes Jews to feel somewhat alienated from Swedish society and motivates them to emphasize their Jewishness rather than their Swedishness is that it is only among other Jews that there is sympathy for this need to engage with the past, whereas the surrounding society by comparison seems indifferent. The stories in this section mostly concern Jews' relationships not to Swedes but to majority populations of countries other than Sweden, in which Jews have been persecuted. And yet they contribute to the boundary between Jews and Swedes, since they are a part of why the interviewees understand themselves as different from Swedes, who lack this background.
The interviewees' discussions of their family histories also illustrate well Wimmer's thesis that ethnic conflict, especially in the form of oppression, stabilizes ethnic boundaries and makes them appear natural. For the interviewees, Jewishness is viewed as the defining element of family history, leading them to view their family history as a reason to affirm their Jewishness. The main cause, in turn, for viewing Jewishness as the defining element of family history in this way is antisemitic persecution, because of which the Jewishness of the interviewees' families has been entirely central to their destiny. We have also seen how Jewish identity may come with significant feelings of guilt and pain.
Boundary salience: Swedish extenuation of the current antisemitic threat and the shadow of the Holocaust The interviewees had varying experiences with antisemitism, ranging from none to severebut with few exceptions, they confirmed that there is a general feeling among Jews of an increased threat of violence towards Swedish Jews, a feeling that, as mentioned above, corresponds with current research on this topic (FRA 2018; Brå 2019). The experiences of antisemitism among the interviewees included violent bullying by neo-Nazi youth, antisemitic provocations in school or at work, as well as hostile looks and insinuations in public places. They had experienced antisemitism from ethnic Swedes as well as from members of other ethnic minorities. Some were especially worried about antisemitism in Sweden's Muslim communities, while others had not noticed any such problems. A large share of the interviewees described being careful not to divulge that they were Jews to new acquaintances until they had come to trust them.
Several Jewish commentators (e.g. Bachner 2011; Josephi 2015; Nudel 2017) have noted that when antisemitism is publicly discussed in Sweden, it is often used as a historical lesson about how to combat the kinds of racism that exist today, like anti-Muslim racism, with the implication that antisemitism is no longer a pressing problem. Samuel, a 32-year-old medical researcher, thought that the state did not do enough to combat antisemitism. When asked what he thinks about the general non-Jewish Swede's knowledge about antisemitism, he answered; I think there's some consciousness of the issue amongst those who take a particular interest in it, but most people don't care. I mean, there are like 20 000 Jews in Sweden. That's about the same number as there are Mandaeans in Swedenhow much do you know and care about their situation? Swedes' perceived ignorance and tendency to extenuate the severity of modern antisemitism was perceived as a lack of recognition of the interviewees' perception of reality. These diverging views on the status of modern antisemitism seem to have contributed to social distance between the interviewees and 'Swedish society', strengthening their feeling of community with other Jews. This feeling was often related to a perceived antisemitic threat; knowing that other Jews will show them solidarity and understanding in the face of antisemitismwhich most interviewees do not trust most non-Jews or the Swedish state to do. Thus, Ben said that there is a 'silent code' amongst Jews that 'when the shit hits the fan, it's you and me'.
Antisemitism, or at least the lack of recognition of antisemitism, was sometimes connected to a perception of Swedes and Swedish institutions as having an anti-Israel bias. Sweden has, historically, been comparatively supportive of the Palestinian position in the Israeli-Palestinian conflict, a position that has on several occasions upset representatives of the Swedish Jewish community. For instance, Sweden and Iceland are the only Western European countries that have recognized the Palestinian state. Johanna mentioned the possibility that a person may hold anti-Israeli sentiment as a source of trepidation when meeting new people; If I were to meet a completely new person, that I didn't know before, getting to know a non-Jew, then it's important for me to know-, I might be attentive towards anything this person might say that's negative about Jews. This happens very seldom, of course, that is, some kind of hidden antisemitism. The other thing is if the person is very negative towards the state of Israel. That's also a problem for me. If someone, for example, is very critical and questions the existence of the state of Israel, perhaps even starts talking about Palestinians' rights and such, that person I would not want to have anything to do with, since it runs so directly against-I would experience that as a threat, since the country of Israel gives me security, with my history.
Johanna, then, is made to feel personally unsafe when someone criticizes the country which for her is an important sanctuary, mentioning in this connection her history as the child of Holocaust survivors. This sentiment was very common among the interviewees (although going so far as explicitly denouncing Palestinians' rights was not). The perception of Swedes' tendency to not recognize modern antisemitism, or to be critical of Israel, is thus connected to a feeling of fundamental insecurity, that the position of Jews in Swedish society rests on fragile grounds. As expressed by Aviva; I feel that I could be sold out. If a great power came and said, 'give us all the Jews', then I am incredibly unsure that Sweden would not just say 'sure, you're welcome, here, take them'. I don't think they have learned anything.
Thus, the interviewees express concern about a perceived increase of antisemitism in Sweden, a development that they understand in the light of historical persecutionespecially the Holocaust. Antisemitism on the part of non-Jews of course plays a crucial part in maintaining the ethnic boundary between Jews and others. But even Swedes not seen as antisemitic themselves are perceived by the interviewees as nonchalant and ignorant about antisemitism, and as uncommitted to the safety of Jews. This hurtful perceived lack of recognition of the Jewish situation, as well as the need to connect with others who share these preoccupations and concernsnamely other Jewsall constitute incentives for drawing boundaries against Swedes, emphasizing instead the boundaries that unite all Jews.

Boundary-shifting: Swedish identity and lack thereof
We now turn to the question of whether, and in what respects, the interviewees identify as 'Swedes'; a key issue in ethnic minority boundary work. In terms of Wimmer's theory, this is a question of expanding or contracting boundaries to include or exclude non-Jewish Swedes. As we saw in the review of previous research on Swedish Jews, there has historically been conflict and division between an assimilationist and an anti-assimilationist camp. Rapaport's respondents tended to 'see Jews and Germans as mutually exclusive dichotomous categories ' (1997, 146). This was not as clear cut among our interviewees; while some of them said they were not Swedish, several interviewees claimed that there is no contradiction between being Swedish and being Jewish. But when asked to elaborate on this, it turned out that to the extent that the interviewees did regard themselves as 'Swedish', many viewed themselves as 'Swedish citizens' but not as 'ethnic Swedes'. Johanna, a 64-year-old teacher, exemplified this tendency: Well, Swedishness … I am Swedish, regardless; I was born here, I am Swedish. At the same time, I'm Jewish. There is no contradiction. I'm a Swedish Jew. I don't know why people often think … {Pausing, then changing tone of voice as when making a realization:} Oh, you mean nationally … well … It's a very difficult question, about nation. (…) The homeland of the Jews is Israel. Historically that has not been the case. And that has been a problem for the Jews in a hostile environment, and I can say that it gives me a sense of security to know that we have a national homeland, given the history we have as Jews. I mean, I'm a very clear example, I'm the child of a [Holocaust] survivor. And if things would turn as bad as one often thinks, as the child of a survivor, and that's not impossible -I mean, we see persecution in every countryand would a situation obtain where Jews are persecuted, I know I have security in the country of Israel. I'm not going to be persecuted in my own country.
This seems to exemplify a civic, institutional, view of her Swedishness and an ethnonational view of her Jewishness, closely connected to feeling safe from persecution. It also exemplifies how the country of Israel is seen as an expression of Jewish peoplehood and a sanctuary for Jews. Such an image was shared by almost all interviewees, and what many interviewees saw as an anti-Israel bias in Swedish media and among Swedes and Swedish politicians constituted a major source of distrust and alienation. Some interviewees took active part in pro-Israel advocacy online, which should be seen as a form of ethnopolitical mobilization (one of Wimmer's [2013, 69-70] means of boundary making) that itself strengthens ethnic boundaries by asking for loyalty to the Jewish nation.
Several interviewees expressed that it was difficult to identify with Swedishness because they perceived it as mainstream and devoid of content. For example, although Ben had a Swedish father and had always lived in Sweden, he said he saw Swedishness as 'too broad' to identify with -'It's like saying you're "European" in general'. Ben also told a story about when he and his family were watching the popular televised Swedish musical contest Melodifestivalen (in which artists compete to qualify for the Eurovision Song Contest): his son asked if they were going to call in to vote for a contestant. When Ben said that he usually does not, his son asked, 'Is that because we're Jews?' Ben told this story with obvious pride and said that it illustrates that '[as a Jew], you realize early on that you are not like everybody else (…)we are distinct!'. This is an example of transvaluation, an attempt to reinterpret marginalization as authenticity; it also illustrates how such transvaluation is used strategically in the boundary maintaining process of discouraging one's children from boundary-crossing through assimilation, which will be further discussed below. For many participants, belonging to a minority was an important part of their self-conception. There were, however, other times when the interviewees emphasized that they are a part of the Swedish people: Judaism/Jewishness is thus sometimes described as a personal attribute compatible with full membership in 'the Swedish people'.

Boundary maintenance: social closure and exogamy
Wimmer's theory views social closure as the behavioural component of boundary maintaining and predicts that an ethnic boundary will be porous and ill-defined if it remains at the cognitive level and is not accompanied by social closure. It is, then, relevant to ask to what extent the participants close their networks towards non-Jews. Some of them do. Anna, who was born in the Ukraine, a severely anti-Semitic context where Jews must hide their Jewishness in order to survive, discovered her Jewish identity when moving to Sweden. She describes leaving her non-Jewish husband immediately upon returning from her first visit to Israel.
However, while all of the interviewees are concerned to maintain the ethnic distinctiveness of Jews (and thereby implicitly a categorical boundary between Swede and Jew), most of them are, unlike Anna, not prepared to limit their social interaction, including choice of romantic partners, to Jews. They are, however, generally aware of the contradiction inherent in wanting to maintain the Jewish identity of themselves and their children, while at the same time not wanting to make ethnicity a factor when finding a spouse. This contradiction is resolved in various ways. When Elsa, now a 51-yearold lawyer, met her non-Jewish husband, she realized that she had to lay out some unconditional requirements before they formed a family; I was very clear with [my husband] (…) that; 'I love you, but this is a prerequisite … ' (…)and this was very strange of course, since we had no plans to have children at that point, but still, you have to be pragmatic as hell -'if you really feel that you will stand up for this, and really work to make sure that our children (…) receive a Jewish upbringing, that if we get a son he will be circumcised, and we'll have Jewish traditions-, and if this is not OK with you all the way, then this ends here'.
Ben had also struggled with how to think about these issues. When he met his first wife, he was not as invested in his Jewish identity as he is now and did not care that she was not Jewish. After he and his wife had children, he realized that he wanted to pass on Jewish identity and traditions to them. This turned out to sometimes be an uphill battle in a 'mixed' family: If every Jew were to decide tomorrow to marry a non-Jew, then it's gonethat's the downside of assimilation. Especially if you're living in Sweden where it's like, 'aren't you Jewish? What is it you celebrate again? What do you call it, Hanukkah? Hm, should we do that or Christmas? I think we should just celebrate Christmas, so the kids won't feel left out at school and we won't have to get time off work for Yom Kippur'and then it's gone. So, it's a struggle to be Jewish.
When Ben later met his current wife, she converted to Judaism, and they are now raising their children with a strong sense of the significance of being Jewish. Thus, convincing a non-Jewish partner to convert to Judaism is not an unusual way of resolving the tension among the interviewees.
Theoretically, we would expect the will to maintain ethnic distinctiveness to sit uneasily with a universalist outlook on partner choice. This is confirmed by several interviewees, who had to struggle to reconcile these two positions; in Wimmer's terms, a discursive, categorical boundary between Jews and others; but a reluctance to discriminate based on ethnicity when choosing a partner. Thus, while several of the interviewees did choose a non-Jewish partner, it had generally not been an easy decision. Related to this, we have seen that maintenance of Jewish identity in a 'mixed' family requires pervasive effort. More generally, many interviewees view assimilation as something that easily happens 'automatically' if one is not vigilant to maintain a sense of Jewish identity; this is a central point of our argument to which we now turn.

Boundary crossing: guarding against assimilation
The possibility of assimilation, that is, the absorption of Jews into majority society and the disappearance of Jewish identity, has long been seen by many as the main existential threat to the survival of the Jewish people. This concern was shared by the respondents. This was especially evident when it came to the topic of raising children. Raising a Jewish child invariably requires parents to take a stand on what it means for them to be Jewish. To end up with a child who proudly identifies as a Jew and finds meaning in Jewish culture and history is no small feat, and those who had achieved it were quite proud of it, as exemplified here by Elsa: Interviewer: And how has it gone, how much have they [the children] taken to heart? Elsa: They've taken to the whole kit! [She goes on to list some ways her children are active in the Jewish community]. And that's wonderful. (…) They have received that whole feeling of community. But they also know that it's not easy. Interviewer: What is it that isn't easy? Elsa: It's not easy to belong to a minority. I don't think it matters if you're Jewish or if you belong to another minority. Being a minority always means, for it not to disappear-, it's so easy to assimilate without even thinking about it, since the things that are not Jewish are all around you in your everyday life in a way that requires a certain knowledge to reflect on. So, you have to cherish itit's a responsibility, it's a job, and it's not easy, darn it! (…) I'm proud and happy about it.
Elsa, then, sees assimilation as something that happens automatically if it is not actively resisted. As we saw above, this picture was shared by Ben, describing the plight to maintain Jewish tradition in mixed families.
For many minority groups, the path to assimilation is, however, closed and heavily guarded by the members of the dominant ethnic group (e.g. Bursell 2012). A lack of the required cultural resources or visual ethnosomatic differences may hinder inconspicuous inclusion in the majority group, depending on the barriers erected by the latter. For many of Jewish background in Sweden, however, the path of least resistance seems to be to adapt to the majority. If relying on the accounts of the interviewees, it is not, in general, Swedes that block the path into Swedishness for Jews, but rather Jews themselves who resist assimilation. Guarding against assimilation, then, becomes a matter of boundary work, drawing firm boundaries around 'the Jewish people' and advocating identification with that entity, to maintain its integrity and dignity. By passing on a 'Jewish consciousness', the interviewees maintain this ethnic boundary by encouraging their children to see themselves as ethnically distinct from majority Swedes. Placing significance in one's Jewishness by itself constitutes an act of boundary maintenance which is more central as an aspect of resisting assimilation than any particular outward cultural trait; although such boundary maintenance motivates the respondents to learn more about Jewish culture and history.
It is crucial at this point to emphasize the following; that the path of least resistance for Jews is to assimilate, and that they are not hindered from assimilation from majority Swedes does not mean that majority Swedes do not draw an ethnic boundary between Swedes and Jews. This is because the acknowledgment of a boundary is compatible with accepting or encouraging the crossing of that boundary. Consider for example the controversial statement by far-right politician Björn Söder that Jews must leave behind their Jewish identities to become Swedes (Orrenius 2014). This statement clearly is a case of maintaining an ethnic boundary between Swedishness and Jewishness; at the same time, it encourages individuals to cross this boundary by eschewing Jewishness and assimilating into Swedishness. Thus, while we claim that many majority Swedes accept Jewish assimilation into Swedishness, on the basis of the evidence from the interviews presented above, we do not mean to say that majority Swedes do not perceive an ethnic boundary between Swedes and Jews, or that they do not engage in antisemitism. Indeed, some may view a pressure from majority society to assimilate as a quite insidious form of antisemitism in itself.
The question of why parents find it important to pass on a Jewish identity to their children is of course tightly connected to the question of why it is seen as important to maintain the ethnic distinctiveness of the Jewish people at all. Many of the interviewees tend to view Jewishness as an essential aspect of the self, denial of which means a life of selfdenial and inauthenticity. As Johanna put it, failure to 'affirm your background' means that you 'lose out on what it means to be yourself'. On the one hand, most participants share an essentialist view of ethnicity, which sees Jewishness as being transmitted through 'blood'. On the other hand, there is a fear that Jewishness can easily disappear through assimilation if this is not actively guarded against. Those that assimilate are still seen as Jewish, but Jews who have 'forgotten who they are'; the life of an assimilated Jew is viewed as incomplete or inauthentic. Several interviewees also pointed out that it is good to be aware of one's heritage since it makes one better prepared against antisemitism. Assimilation is also seen to endanger the entire Jewish community by weakening the bonds of solidarity that are required to guard against the next, inevitable attack on Jews. While an essentialist view of ethnicity is taken, strategic action is, then, still required to discourage assimilation; this has important theoretical implications which we will discuss at some length in the conclusion.

Making sense of Jewishness by discovering Jewish culture
A key point of the boundary making perspective has been to suggest that cultural differences do not track, or explain, ethnic boundaries. Do our results support this assertion?
There are examples of when the interviewees experience themselves as ethnically different because of differences in cultural practices, such as cuisine, names, dress, language, and so on. But this is not the entire story of the relationship between culture and ethnic boundaries. Another source of feeling of ethnic difference in this case, as we have argued, is the memory of persecution, and, crucially; the feeling of ethnic pride and alienation that results from that persecution often itself motivates Jews to seek out 'Jewish culture' which they previously had no knowledge of. In these cases, ethnic difference cannot be explained by cultural differences, since it is the feeling of ethnic differentness which is the very reason to seek out the 'different' culture. There seems, then, to be an important feedback mechanism at work here; both cultural differences and other causes for ethnic differentiation (as we saw in the section on the interviewees' views of history, some explicitly stated that the memory of the Holocaust motivated them to keep Jewish culture alive) induce ethnic solidarity in people, motivating them to invest more heavily in what they consider to be the culture proper to the group.
Several of the participants had, at some point in their lives, found a new interest in Jewish identity and culturemuch like the Polish-Jewish immigrants described by Ilicki (1988), whose new-found Jewish cultural orientation was a result of their exclusion from the Polish national community that they had previously regarded themselves as a part of. In some cases, the participants in the study had gone from being almost totally indifferent and ignorant of Jewish customs to finding them immensely important. In such cases, their interest for those customs and cultural practices cannot be explained by socialization or habit. Rather, there was clearly some reason other than familiarity with the cultural practices themselves that made the participants consciously seek out Jewish culture.
Consider, to illustrate this, the case of Anna. Anna was born in Lviv in the Ukrainian SSR of the Soviet Union in 1974. As Anna grew up, she and her family faced incessant antisemitism, including physical violence, harassment and blatant discrimination. This made Anna keenly aware that she was Jewisheven though her upbringing was completely devoid of Judaism, Jewish culture and celebration of Jewish identity, her family always attempting to conceal that they were Jews. Some time after moving to Sweden, Anna felt a need to learn more about Judaism. She became deeply committed to what she sees as the values of Judaisma transition she found to be 'incredibly natural' as she already on some level knew about these values 'under her skin' and 'subconsciously' because she has 'Jewish blood'; Anna: For me, Judaism is not so much a religion as it is something you carry with you when you are born. I mean, you're not any less Jewish if you don't go to the synagogue (…) For me, Judaism is much more than a religion, it is a way of being (…). Since Anna's time in the Ukraine contained no Jewish culture and no pride in Jewish identity, her consciousness of her Jewishness seems to be solely a result of the harsh boundary enforcement she faced as a Jew in the Ukrainian SSR. As Wimmer's model suggests, harsh ethnic conflict will make boundaries appear sharp and make people take them for granted. This seems to apply well to Anna's situation. Having realized that 'Jew' was the ethnic category which was most meaningful for her, she then set out to fill that category with cultural 'stuff'. For the purposes of discussions of 'assimilation', this should caution us against a view that knowledge of Jewish culture is present from the start, and then gradually withers away in the new cultural context. Assimilation, as we have used the concept, can then not be measured by an objective list of outward cultural signs, although there is a relationship of mutual reinforcement between one's sense of the significance of one's Jewishness and one's knowledge of Jewish culture.

Conclusion
We have argued that the antisemitic persecutions of the twentieth century have entrenched the perception of Jews' ethnic distinctiveness among the respondents. These persecutions also contribute to the alienation of the respondents from Swedish society, which does not share the same historical identity and frames of understanding. They also constitute incentives for maintaining cohesion in the Jewish community, lest that community be poorly prepared for the resurgence of antisemitism. These factors in turn motivate the interviewees to maintain the ethnic boundary between Swedes and Jews and guard it against assimilation. We have seen how the interviewees actively struggle to maintain Jewish peoplehood, by keeping Jewish culture alive; by teaching offspring to take pride in being Jewish; and by mobilizing support for what is seen as the manifestation of Jewish peoplehood, the state of Israel. A perception of a firm difference of ethnic or even racial essence lives on in some of the descendants of the survivors of persecution. At the same time, we have argued that the discursive ethnic boundaries drawn by the participants may be analysed as strategic action. It is crucial, then, that the latter notion is not understood in a narrow way. As we have seen, the embracing of Jewish identity may include a great deal of pain, mourning and survivors' guilt, as well as a partial alienation from the surrounding society. Moreover, it does not offer any apparent benefits in terms of social position. It seems, rather, that one embraces Jewish identity, and all the pain that comes with it, because one thinks that doing otherwise would be to live inauthentically. Identifying as Jewish, then, is not done calculatingly but rather reflects assumptions about reality.
However, a strategic analysis still applies; assimilation into the mainstream and the loss of authenticity is perceived as an ever-present threat. Active struggle is thus needed to maintain the ethnic boundary between Jews and othersconfirming the applicability of a strategic mode of analysiseven though Jewishness is perceived as a matter of inner essence rather than freely chosen. This argument for the applicability of a strategic mode of analysis of ethnic boundary maintenance is the main theoretical contribution of this article. While the interviewees in many cases view the ethnic boundary between Jews and Swedes as natural and reflective of the inner essence of Jews and Swedes, they act strategically in their attempts to prevent that boundary from being crossed by their children and other Jews through assimilation, or erased all together altogether by no longer being viewed as significant.
This seems like a strong argument for advocates of a strategic view of ethnic discourse. At the same time, it nuances such a position; it tells us that ethnic actors are not merely calculating, strategic actors. They act, in this case, from the genuine conviction that there is something objective and primordial about ethnic identity, even when, as we have seen, it is painful. This argument, then, warns us from going too far in the direction of rationalism when analysing ethnic identity, while at the same time giving a lot of credence to the idea that there is some clearly strategic component to how we talk about ethnicity.
The argument also seems to have implications which should be tested on cases morphologically different from the one studied here; if it applies to a minority resisting assimilation, does it also apply to other forms of boundary maintenance? It seems likely, for example, that members of a hegemonic majority might have a taken-for-granted, essentialist view of ethnicity while at the same time strategically employing ethnic discourse to e.g. perpetuate their hegemonic position or to combat boundary-crossing. The theoretical discussion referred to in the beginning of this paper between those (like Wimmer) who say that boundary making is strategic and those (like Lamont) who say it reflects taken for granted assumptions seems to not fully capture this possibility.
We have seen that the participants engage in discourse that makes Jewishness appear attractive, by reinterpreting marginality and persecution as resilience and authenticity and stresses the warmth and solidarity of one's ethnic group in contrast to a cold and faceless mainstream. This bears clear resemblance to the strategies of other minorities; similar narratives are found in the work by Michèle Lamont (e.g. 2000) on North Africans Muslims in France and on African-Americans, as well as for example in work by Mizrachi and Zawdu (2012) on Ethiopian Jews in Israel, or middle-class minorities across different contexts (Schwarz 2016). 6 But we have also seen how this transvaluation is an important part of boundary maintenance itself since it is a crucial tool in strengthening solidarity among Jews and combating assimilation, since transvaluation increases the attractiveness of identifying as Jewish.
While much analysis of the role of memory narratives in the perpetuation of ethnic groups approaches collective memory through a bottom-up discursive analysis which stresses the 'day-to-day' micro-level construction of ethnicity (e.g. Rapaport 1997), we have also attempted to go beyond this approach by analysing not just micro-level discourse, but its historical determinants. As Wimmer argues, his model 'offers a "full circle" of explanation, as specified by Coleman (1990) [and others], leading from macro to micro and back to the macro level again ' (2013, 112).
The purpose of this article was to describe how the ethnic boundary between Jews and non-Jews is maintained by the interviewees; there may also be tendencies for this boundary to blur. This should be kept in mind by the reader to avoid an exaggerated impression of the distinctness and alienation of Swedish Jews. As mentioned in the methods section, our sample is too small to aim for representativeness. Similarly, it has been beyond the scope of the study to directly investigate the degree to which there is consensus or contestation on the meaning of the Jew-Swede boundary. In order to fully explore the construction and maintenance of this boundary work, further research should also ask how non-Jews and assimilationist Jews view this boundary. Wimmer's theoretical framework contains a plausible account of the degree to which consensus on the location and meaning of boundaries is achieved through a process of ongoing negotiation and cultural compromise (Wimmer 2013, 97-101). Notes 1. We use 'boundary making' and 'boundary maintenance' roughly interchangeably. This may seem confusing. To maintain something, it must already exist; and to make something it must not exist to begin with. However, as we note throughout this paper, there is no consensus on whether there is an ethnic boundary between Swedes and Jews. Since an ethnic boundary must exist in the minds of persons to exist at all, this lack of consensus means that there is no straightforward answer to whether there is 'actually' an ethnic boundary between Swede and Jew. There is thus not a clear answer to the question of if the strategies described in this paper should be seen as making or maintaining boundaries; that would seem to depend on the reader's perspective, or possibly on the perspective of actors in the situation described. One and the same action may be seen as maintaining a boundary to the extent that it reinforces a boundary that ego perceives and making a boundary to the extent that it induces a boundary where it was not perceived before. 2. This historical description is based on Valentin (1964), Sjögren (2001) and Bredefeldt (2008). 3. For an extended, albeit dated, overview of research on Jews in Scandinavia, the British Commonwealth and the United States, see Fischer (1996). 4. As mentioned above, Swedish policies aim for integration, not assimilation. Integration is the concept used in the public discussionassimilation has a negative connotation in the Swedish context. But assimilation is the word used by the respondents. When we refer to 'assimilation' in this study, we are not using the word in the tradition of assimilation scholars like Park and Burgess (1969) or Alba and Nee (1997), but in the way that the respondents describe it. As will be evident below, the respondents are not resisting participation in Swedish mainstream culture. When they speak of assimilation, they refer to full assimilationthe complete disappearance of Jewish heritage and culture (group level) and of the sense of Jewish distinctiveness something which they fear will happen if they, as individuals, let go of their Jewishness. 5. For a brief biography of the interviewees, see appendix. 6. See also the special issue of Ethnic and Racial Studies on destigmatization, edited by Lamont and Mizrachi (2012).

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