How to Negotiate Class Ambiguity: A Boundary Work Approach.

Research indicates that actors increasingly engage in practices that do not match their class background. This contribution explores how actors negotiate class ambiguity through boundary work. Studying Argentinean middle-class actors participating in Pentecostalism, the article draws attention to boundary conversion as a strategy to manage class ambiguity. Deviating from the middle class with their religious affiliation, the studied Pentecostals convert existing boundaries between the Argentinean middle class and Pentecostalism into internal boundaries within Pentecostalism, creating trenches between "highbrow" middle-class and "lowbrow" mass Pentecostalism. The boundary processes point to the ongoing relevance of class distinction even among those actors that freely engage in practices at odds with their class background. As such, the results underpin the need to study not only what type of dissonant practices actors perform, but equally what boundary strategies they employ to negotiate their class belonging.

Zusammenfassung: Forschung zum Zusammenhang von Lebensstilen und Schichtung verweist darauf, dass Akteure zunehmend Praktiken ausüben, die nicht zu ihrer Klassenposition passen. Dieser Artikel untersucht am Beispiel von Mitglieder der Pfingstbewegung aus der Mittelschicht Argentiniens, wie Akteure diese Klassenambiguität mittels symbolischer Grenzziehungen verhandeln. Die untersuchten Pfingstler*innen weichen mit ihrer Mitgliedschaft in der Pfingstbewegung von vorherrschenden Mittelschichtsstandards in Argentinien ab. Um mit ihrer Klassenambiguität umzugehen, konvertieren sie die externe Grenze zwischen der Pfingstbewegung und der Mittelschicht in eine interne Grenze innerhalb der Pfingstbewegung zwischen einem Mittelschichts- und Unterschichtspfingstlertum. Diese symbolischen Grenzziehungen zeigen auf, dass auch unter jenen Akteuren, die mit ihren Praktiken von ihrer Klassenposition abweichen, Distinktion weiterhin von hoher Relevanz ist. Dementsprechend ist es nicht nur wichtig zu erforschen, welche möglicherweise abweichenden Praktiken Akteure ausüben, sondern welche Grenzziehungsstrategien sie hierbei anwenden, um ihre Klassenzugehörigkeit zu verhandeln.

Keywords: Social Class; Boundary Work; Omnivores; Class Ambiguity; Middle Class; Religion; Cultural Dissonance; Soziale Schichtung; symbolische Grenzziehungen; Klassenambiguität; Mittelschicht; kulturelle Dissonanz

Sociology investigating the relationship between class and culture has highlighted the social function of culture ([11] 1979; [24] & Useem; [114] 2007[1899]; [116] 1972[1921]). Exemplary for this strand is Bourdieu's sociology, which conceives of cultural practices and tastes as social mechanisms for marking and reproducing class positions ([12] 1979; Bourdieu & Passeron 1990).

Though many scholars highlight the ongoing relevance of class distinction ([21] & Roose 2014; [37] 1997; [38] 2005; [40] 2015; [86] & Beckert-Zieglschmid 2002; [87] & Bromberger 2009; [88] 2013; [89] & Lenz 2018; [106] 2005), a considerable corpus of studies points at deviations from exclusive highbrow tastes among upper- and middle-class actors ([14] 1996; [39] & Katz-Gerro 2010; [63] & Skiles 2015; [78] & Kern 1996; [112] 2001; [115] et al. 2007). Studies supporting the "omnivorousness" thesis contend that high-status individuals increasingly appreciate diversity and choose lowbrow options. Even in France, which has been perceived as a stronghold of exclusive higher-class culture (Bourdieu 1979; [56] 1992), [51] (2001, 2005, 2008) reports that upper- and middle-class actors – as well as actors from diverse social backgrounds – divert from the cultural tastes normally associated with their social position.

While the mostly quantitative studies document an increasing "incoherence in the cultural profile" ([35] 2016) of actors, we know little about how these actors deal with this incoherence and, in particular, in what way upper- and middle-class actors practice lowbrow culture. Addressing this topic, this article tackles the question of how actors negotiate class ambiguity. Class ambiguity refers to actors deviating from tastes and lifestyles associated with their social position. This article contends that actors employ boundary work ([31] 1983; [57] 1992, 2001; [117] 2008a, 2008b) to negotiate class ambiguity. They can engage different boundary strategies to do so (e. g. blurring, repositioning). Drawing upon Wimmer's (2008a) taxonomy of boundary strategies, this article introduces the notion of boundary conversion as a strategy to negotiate class ambiguity: it converts the external boundary between two groups (e. g. social classes, religious groups) into an internal boundary within one of these groups. Thereby, the article suggests that combinations of Wimmer's boundary strategies lead to forms of boundary work that make it possible to renegotiate increasing boundary conflicts in times of rising cultural plurality and class ambiguities.

This article illustrates this boundary process in a case study on the Argentinean middle class involvement in Pentecostalism – a religious movement that is considered to be lower class in Argentina ([64] 1999; [71] 2001; [90] 1989; [98] 2004; [124] et al. 1995; [122] 2009). Given that membership in Pentecostalism is not well accepted in middle-class circles, middle-class actors affiliating with the movement face social tension. They thus have to deal with conflicting membership in two groups perceived as mutually exclusive: Pentecostalism and middle class. Argentinean middle-class Pentecostals handle their class ambiguity by engaging in boundary conversion: they seek to convert existing boundaries between the Argentinean middle class and Pentecostalism into internal boundaries within Pentecostalism, creating trenches between "highbrow" middle-class and "lowbrow" mass Pentecostalism. This case allows for a better understanding of class ambiguity. Despite choosing lowbrow culture, middle class actors try to maintain their social profile by generating highbrow forms of their "lowbrow" practices. In order to do so, they engage in boundary work.

The goal of this study is twofold: First, it discusses and illustrates how actors negotiate class ambiguity through boundary work. Second, the article extends existing typologies of boundary processes by introducing an additional process: boundary conversion. Boundary conversion becomes increasingly relevant when class ambiguity increases due to processes of cultural pluralization and growing omnivore tendencies. However, as it offers a strategy to deal with conflicting allegiances to groups and/or practices, it is not limited to class-ambiguity. Actors may employ it in different situations of conflicting allegiances.

The article is structured as follows: It starts with a brief introduction into existing research on class ambiguity and the concept of boundary work, exploring its potential for unfolding class ambiguity. The next section introduces the readers to the relationship between social class and religion in Argentina. The following two sections describe the methods and main results of the case study on Argentinean middle-class Pentecostalism. The conclusion of the article then discusses the notion of boundary conversion and outlines potentials for future research.

Class ambiguity refers to actors that mismatch their class profile in some domains. Examples for this are students from a working class background that attend elite universities or upper class individuals that consume lowbrow culture. Research on cultural omnivores points to the prevalence of class ambiguity by indicating deviations from exclusive highbrow tastes among upper- and middle-class actors (Bryson 1996; Jaeger & Katz-Gerro 2010; [43] 2002; [79] & Kern 1996; van Eijck 2001; Warde et al. 2007). Pioneering this strand, [80] et. al (1992; 1996; 1992) have witnessed a shift from exclusively highbrow musical tastes to omnivorous musical tastes among higher-class groups in the US: their members have become more likely to appreciate and even listen to lowbrow styles of music (e. g. country music). The results suggest that tolerance is favoured over cultural exclusivity among these actors: they can dedicate themselves to lowbrow practices without having to fear social sanctions from peers. All the more, being a cultural omnivore is increasingly regarded as honourable and a sign of competence among these actors ([9] et al. 2010).

While most of the aforementioned studies draw on quantitative surveys, the qualitative interviews of [53] contrast these findings by indicating tensions. He shows that actors drift away in some cultural genres from their class-related cultural patterns. Labeling this phenomenon "cultural dissonance", [54] observes that actors experience tensions resulting from their culturally dissonant behavior. Those who dedicate themselves to lowbrow practices communicate regret and try to distance themselves from this behavior. Aware of the "illegitimacy" of their cultural practices, these actors struggle against their popular "vices" ([52] 2004/2006: 634 f., 2005).

Given these tensions, class-culture patterns do not appear to vanish. Instead, they remain in place while actors engage in dissonant practices. These tensions raise the question of how actors deal with their class ambiguity.

A few studies tackle how actors deal with deviation from established class-culture patterns. Most studies in this area are related to the field of education, centering on how upwardly mobile actors from working class backgrounds cope with their new social environment at elite schools, colleges, and universities ([83] et al. 2001; [84] et al. 2010; [107] 2006). Different strategies have been reported: learning how to dress and behave like other students at elite colleges, students from working class backgrounds mimic their new social environment by displaying similar cultural styles ([33] 1991) and/or gradually adapt to their new environment by transforming their habitus ([62] 2014). Furthermore, some socially upward mobile actors seek to cope with tensions between their background and the new context by actively drawing on lower-class morality, emphasizing hard work and responsibility ([61] 2009). Others that experience a mismatch between themselves and their new social environment drop out despite good academic performance ([60] 2007; [82] 2004).

While the involvement of actors from working-class backgrounds in higher-class contexts has received some academic attention, the opposite is the case with how upper- and middle-class actors engage in lowbrow cultural contexts. Research on middle classes living in working-class cities suggests that middle-class actors tend to evolve distinction strategies, distancing themselves from the surrounding lower-class culture ([15] et al. 2016). Similarly, [77] and van Eijck (2018) find that actors with high cultural capital in the Netherlands seek to perform karaoke in an ironic fashion to distance themselves from its usual lowbrow performers. Unlike the situation of upward social mobility, middle-class actors do not mimic the cultural codes of their new social environment but evolve strategies to highlight their class position.

Not every deviation in the cultural profile will automatically involve tensions. In social contexts that value diversity, lowbrow consumption may even constitute a "badge of honor" for upper and middle class individuals, allowing them to parade their cultural openness (Bennett et al. 2010; [41] & [28] 2017; [113] & Jonvik 2019). Tensions are most likely to emerge when actors feel that their social position is at stake, as they lose control over their peers' recognition. This happens when the cultural consumption of an actor could lead its peers to question the social positioning of the actor and even lead to its exclusion from peer networks. As the examples above illustrate, this is most pertinent when the consumption includes the affiliation with a group (e. g. association, religious community, school) that is regarded as being constituted by members of the other class. Affiliating with this group becomes an act of stepping out of one's own class. Therefore, class ambiguity that involves tensions can be described in terms of conflicting group allegiances. Membership in one group renders membership in another group not only unlikely but culturally mismatching or even morally inappropriate. For instance, the class background of the aforementioned working-class students appears to mismatch the social environment of elite colleges. When the combination of two mismatching group memberships (working-class vs. elite colleges) leads actors to suffer from social and mental tensions, these are likely to develop strategies to deal with the conflicting group allegiance (Lahire 2004/2006: 634 f., 2005).

The notion of "boundary work" makes it possible to explore how actors deal with conflicting group allegiances. This article contends that actors employ boundary work (Gieryn 1983; [58] 1992, 2001) to negotiate class ambiguity whenever the dissonant behavior does not simply constitute a badge of honor, but involves tensions.

"Boundary work" refers to the everyday practice of groups to draw symbolic boundaries between themselves and others (Gieryn 1983; [59] 1992; Lamont & Fournier 1992). By drawing specific boundaries in their everyday practice, actors can display their social class position (Sachweh 2013). These boundaries only have meaning for other actors if they have previously been established as symbolic attributes that generate recognition and respectability within the group. Thus, daily practices of drawing boundaries are based on group values and narratives (e. g. narratives of "middle-class-ness") ([72] 1976; [104] 2004).

Apart from reproducing predominant group narratives, boundary work can also be employed to change established boundaries and reframe the group's attributions (Lamont & Molnár 2002: 186). As boundaries can be redrawn by actors, the concept of boundary work emphasizes agency and conveys flexibility to actors in dealing with existing group boundaries ([30] & Kämpfer 2017; [118] 2008a: 1027, 2009). Moreover, it takes the criticism on "groupism" ([13] 2002) – the tendency to take groups for granted and reify them – into account by focusing on the processes that construct groups through performative boundary drawing processes.

Despite a massive corpus of literature on boundary making and changing processes in various social areas (politics, migration, science, culture etc.), scholarship has remained fragmented consisting mostly of disconnected case-studies and barely offering comprehensive overviews of boundary processes (Lamont & Molnár 2002; [74] et al. 2007; [126] & Woon 1999). Against this background, this article refers to Wimmer's taxonomy of boundary processes which conveys their broadest classification ([119] 2008a, 2008b: 986 ff.). Reflecting on ethnic group boundaries, Wimmer summarizes five strategies of changing existing boundaries: expansion, contraction, inversion, repositioning, and blurring. Applying this classification to the study of class ambiguity, one can distinguish different boundary strategies that actors may employ to deal with class ambiguity:

Boundary expansion shifts class boundaries in a way that allows the previously mismatching group membership to form part of the class. Boundary expansion occurs, for instance, when upper- and middle-class actors become culturally more inclusive approving tastes and practices that were previous excluded (Peterson & Kern 1996). Boundary contraction, by contrast, shifts boundaries into the direction of higher-class exclusivity.

Boundary inversion involves the inversion of the hierarchy of whole groups. Since these inversion processes – which would, for instance, imply that the whole working-class would assume a higher position than other social classes – are only likely in revolutionary scenarios, boundary inversion may refer under more general circumstances to single practices and tastes (but not the whole cultural arsenal) being revalued. Single lowbrow practices may become highbrow such as, for instance, drinking Tequila in Mexico and the US ([29] 2014).

Boundary repositioning refers to actors changing their position vis-à-vis boundaries. As class ambiguity involves two mutually exclusive group memberships, actors may drop out of one of the groups. This strategy is, for instance, illustrated by working-class students dropping out of elite colleges (Lehmann 2007; Quinn 2004), or, in the case of successful upward social mobility, leaving their working-class background behind and culturally adapting to the elite (Lehmann 2014).

Boundary blurring, finally, implies a diminishing importance of boundaries. Class distinctions may become increasingly obsolete: for example, sociologists pointing to individualization processes have claimed that rising voluntarism involves class boundaries diminishing in importance ([7] 1992; [8] & Beck-Gernstein 2002).

Most of these boundary processes involve transformations that small segments of actors can hardly influence if they do not fully control one of the two mismatching groups of which they are members. Only boundary repositioning appears as a viable option, as actors may change their social class position or decide to disengage from the mismatching practice. However, in the case of being unwilling or unable to leave one of the groups, actors have to develop other boundary strategies. This article suggests boundary conversion as an additional strategy.

Boundary conversion refers to the resetting of an existing boundary. Actors convert the external boundary between the two groups (e. g. working class and elite colleges) into an internal boundary within one of these groups (e. g. distinction between working class styles and other styles at elite colleges). In consequence, the boundary changes its position: it simultaneously unites formerly separated groups and creates new trenches between formerly united groups. This involves two sub-processes: (a) blurring the boundary in its present position (e. g. uniting the working class and elite colleges) and (b) redrawing the boundary in a different position (e. g. establishing a working class style at elite colleges).

Using Wimmer's terminology, boundary conversion could also be described as a form of boundary repositioning. However, in contrast to Wimmer's concept of boundary repositioning, boundary conversion does not refer to individuals changing their position vis-à-vis a fixed boundary but to a boundary changing its position vis-à-vis the individuals. In other words: it is not about individuals leaving one of the mismatching groups (e. g. working class individuals leaving elite colleges) but about individuals seeking to relocate the boundaries that separate the groups. They seek to blur the boundary that is the origin of the ambiguity, thereby apparently creating more cultural inclusivity. Nevertheless, boundary conversion is also different from Wimmer's concept of boundary expansion as it involves more than shifting a boundary towards more cultural inclusivity: it draws a boundary within the mismatching cultural practice, thereby creating a distinction between forms of the same cultural practice (e. g. going to elite colleges). This boundary designates which form of the practice is acceptable for the given group (e. g. working class style at elite colleges) and distinguishes it from those regarded as inappropriate for this group (e. g. upper class style at elite colleges). Consequently, boundary conversion extends Wimmer's taxonomy. It stresses the relocation of boundaries, which, so far, has neither been covered by Wimmer's boundary repositioning nor by his concepts of expansion and contraction.

At the same time, boundary conversion indicates that boundary strategies can consist of different sub-processes, In the case of class ambiguity, it involves two sub-processes: Wimmer's boundary blurring and a redrawing of boundaries. Boundary conversion blurs the boundary between a class and a cultural practice and redraws the class boundary within the practice, separating lowbrow forms from highbrow forms of the given cultural practice. Consequently, boundary strategies can consist in combinations of existing boundary processes.

The type of boundary strategies that actors employ depend on the particular setting in which they move. Boundary conversion becomes likely when individuals (a) affiliate with two different groups their social environment perceives as mutually exclusive, (b) need to stick to these mutually exclusive memberships, and (c) control none of the given groups to a full extent.

When members of one of the two involved groups perceive the other group as inappropriate, those openly affiliating with both groups are likely to suffer tensions. In this case, being a member of both groups can place the membership in one of them at risk. For instance, middle class peers may start to question whether the given individual is one of them and exclude the dubious individual from their circles. This creates social and cognitive tensions among those participating in both groups. A pragmatic solution would be to leave one of the groups (e. g. working class leaving elite college). However, if actors are not able or willing to leave one of the groups, they have to search for alternative solutions. In this case, they may conceal their membership in one of the groups (e. g. working class seeking to appear upper class). If this is not possible, the only feasible solution consists in converting one of the groups in a way that it matches better with the other group. Nevertheless, if the individuals lack the full control of the group, they can only transform sections of it in ways that these sections become less conflicting with the other group (e. g. establishing working class styles at elite colleges). This process takes place in the course of boundary conversion.

The aforementioned conditions are particularly relevant in the case of religious affiliations that mismatch the class profile of some members. Members of religious groups usually stick to their religion and in most cases openly state their religious affiliation. As such, middle class members of lower class religions will neither disaffiliate from their religion nor be able to conceal their membership. Moreover, they will also lack the full control over the given religion. As such, they may draw a boundary within the group, thereby creating a sub-group of the religion that blurs the boundaries towards the middle class.

With classical sociologists such as Weber, Marx, Troeltsch, and Niebuhr addressing the relationship between class and religion, the class-religion nexus became a vital subject of sociological inquiry ([68] 1844; [73] 1929; [111] 1923/1977; Weber 1972[1921]). However, as secularization theories grew in importance and the link between religion and class was increasingly questioned, the study of the class-religion nexus was partly abandoned from the 1980s onward ([22] & [103] 1997; [70] 2007b). Today, religion is seldom taken into account in surveys on social class (Bennett et al. 2010; [92] et al. 2013; [93] et al. 2015). Nevertheless, numerous studies show a relationship between social class and religion, thereby indicating its relevance in the drawing of class boundaries ([19] 2002; Darnell & Sherkat 1997; [23] & Pyle 2006; [44] 2008; [45] & Sherkat 2014; [95] 2011; Sherkat & Wilson 1995; [105] & Faris 2005). Moreover, academic debates about religion in modern societies have increasingly questioned the secularization narrative. In many parts of the world, religion appears to be as vital as ever and continues to play a role in the (re)production of social class inequalities ([10] 1999; [69] 2007a). Just as much as other cultural practices (e. g. food, clothing, music, media consumption), religious affiliations and styles are class markers ([49] 2018; McCloud 2007a; [94] 2006). One example for this is Pentecostal membership in Argentina, as will be illustrated in the following paragraphs.

Argentina is notorious for having a longstanding and extensive middle class in Latin America. The cultural imaginary of the Argentinean middle class has developed in distinction from that of the lower class ([1] 2009; [34] 2004; [110] 2006). While the lower class has been portrayed as uncivilized, emotional, uneducated, untidy, and superstitious, the middle class has sought to define itself as its civilized, rational, well-groomed, and educated counterpart. The distinction from the lower class can manifest itself in the disdain shown in pejorative terms such as "negro" or "villero" (slum dweller). Examples of practices associated with the lower class are listening to cumbia music, displaying a naked torso in public, drinking beer and leaving trash in public places, not respecting well administered facilities, employing brute force, and speaking frankly about one's problems in public (Tevik 2006: 114–115, 211–214). Perceived as signaling lower class culture, middle class individuals pursue to avoid these and other lowbrow practices. Several economic crises have placed Argentina's structural class hierarchies under stress in the last decades. Against these insecurities, cultural class markers assume even higher importance. They enable middle class actors under economic stress to continue distinguishing themselves from the lower classes and the "new rich" that managed to improve their economic situation but have little cultural capital. Religion is one of these markers.

Scholarship about religion in Latin America usually describes "popular religion" as a pool of religious beliefs, devotions, and practices that appeal in particular to the lower classes ([75] 1996; [97] 2001a, 1997). Prominent examples are popular Catholicism, Pentecostalism, Afrobrazilian religions, indigenous religion, witchcraft, and sorcery. These share a holistic worldview in which the spiritual and the daily world constantly interact. According to this worldview, the daily wellbeing of individuals depends on spiritual forces (e. g. Holy Spirit, demons, saints, ancestors). Popular religions endeavour to manipulate the spiritual forces in a way that it will positively affect the wellbeing of their followers (e. g. persuading the spirits through offerings).

The educated middle class contrasts this picture. Academic portrayals of the middle class refer to secularization processes often accompanied by sentiments against popular religion ([27] et al. 1998: 296; [96] 2006b). Adopting the European ideology of secularization, middle classes are thought to appreciate rationality and scientific knowledge, as opposed to the allegedly superstitious worldviews of the lower class. Nevertheless, Argentineans with higher education degrees can hardly be described as secularized, given that the vast majority continues to believe in God and is affiliated with the Catholic Church ([18] 2008). Middle class Argentineans tend towards a nominal Catholicism and hold negative views of popular religion ([101] 2006b: 23, 42–44). As such, they interpret the rise of Pentecostalism, which has massively expanded among Latin America's lower classes ([4] 2004: 59, 282; [6] 1997; [16] 1997), as a crisis of modernity.

Pentecostalism is a Christian renewal movement which emerged at the beginning of the 20th century. Typical characteristics of the movement are its focus on the Holy Spirit as well as its gifts (e. g. speaking in tongues), expressive and emotional church services, and an emphasis on missionary expansion (Anderson 2004; [67] 1990; [85] 2004). During the 20th century, it became the most successful religious movement in the world, strongly expanding in the Global South and among its lower classes in particular. Today, its worldwide membership is estimated at up to 600 million ([42] 2013). In Latin America, the movement experienced a massive expansion starting from the 1950s onward ([5] 1994, 1997). Today, Pentecostalism is the main religious competitor of the Catholic Church which has originally dominated the religious field in Latin America ([17] 2003).

In Argentina, first Pentecostals arrived at the beginning of the 20th century while the massive expansion of the Pentecostal movement only started in the 1980s, thus comparatively late ([2] 2010; Saracco 1989; [120] 1999; [125] & Semán 1994). Today, the movement attracts between 8 % ([65] et al. 2015) and 10 % (Pew Research Center 2014) of the Argentinean population. Most of its members are lower class with levels of education and income below the national average (Conicet 2008; [99] 2000, 2004; [123] 2010). Contrasting the aforementioned national average of 8–10 %, Pentecostal membership in slums and lower-class neighborhoods reaches 20 % ([25] et al. 2001; [102] 2010).

Scholars have raised different theories to explain the expansion of the Pentecostal movement among the lower class. In particular, they have stressed that Pentecostalism offers strategies to cope with poverty and poverty-related problems (e. g. lack of access to health and education services) (Chesnut 1997; [66] 1994). Pentecostal churches provide life improvement strategies: they encourage self-discipline and hard work, prohibit alcohol and drug consumption, provide help networks, and offer motivating messages. Other researchers have highlighted the cultural match between the religious worldviews of the lower class (i. e. belief in spirits) and Pentecostalism ([3] 2008; Parker 1996; [100] 2004). Additionally, the stigmatization of Pentecostalism restrains the middle class from participating in the movement ([32] et al. 2008; Semán 2006; [121] 2005). Its portrayal as a religion of the uneducated lower classes starkly contrasts the self-imaginary of the middle class and therefore discourages middle class actors from visiting Pentecostals churches.

Nevertheless, some middle class Argentineans affiliate with the movement, thus allowing us to study how actors deal with class ambiguity. Middle class Pentecostals are often second or third generation Pentecostals. Being descendants of lower class Pentecostals, they experienced social upward mobility and have evolved a style of Pentecostalism that is more appropriate to the middle class. This style and the social networks of middle class Pentecostals have attracted other middle class individuals to their churches, thereby leading to an expansion of middle Pentecostalism.

The case of Pentecostal affiliation differs from that of other cultural practices in at least one regard. While actors may be able to conceal or stop other "inappropriate" cultural practices, they are less likely to do so in the case of practicing Pentecostalism. Pentecostal churches encourage their members to openly express their faith and, whenever possible, even seek to win new souls for the movement. Showing their faith, middle class Pentecostals face tensions in their middle class circles and have to find strategies to deal with their conflicting memberships in the middle class and Pentecostalism.

The results presented in this contribution are part of an investigation about middle-class Pentecostalism in Argentina (for detailed information about the methods, see [48] 2016). To study middle-class Pentecostalism in Argentina, the author of this study triangulated different methods and data sources ([26] 2008). He conducted (a) ethnographic observations in 22 Pentecostal churches, (b) 44 semi-structured interviews with members of the Pentecostal movement, and (c) 10 semi-structured interviews with pastors. The material was collected in different research steps. It started with a broad data collection among lower and middle class Pentecostals and their churches while the last research steps placed a focus on middle class Pentecostalism. The contrastive sampling strategy allowed to determine whether middle-class churches evolve a distinctive style of Pentecostalism ([76] 1990: 169 ff.). As such, lower class churches and interviewees served as a control group that facilitated identifying specific characteristics among middle class Pentecostals.

Ethnographic observations were undertaken in 22 Pentecostal churches in Buenos Aires and its outskirts, involving 12 middle class and 10 lower class churches. A combination of different data sources was employed to identify middle class churches: (a) information from researchers and Pentecostals about potential middle class churches in Buenos Aires, (b) geographical location of the given church in a middle- or lower-class neighborhood, (c) ethnographic interviews with pastors and members of these churches about their class background, (d) observations in and around the churches (e. g. members' clothing, cars, mobile phone devices, use of language). To study the churches, the author participated in its Sunday churches services and sometimes in church groups and other events (e. g. Wednesday night service, missionary events). After the visits, the author summarized the observations in field notes ([47] 2003: 90 ff.; [55] 2005: 613 ff.). With the permission of the pastors, audio and videos of specific activities (e. g. sermon, singing, and exorcisms) were recorded for analysis and their display in the interviews.

Out of the of 44 qualitative semi-structured interviews with members, the author conducted 30 interviews with middle-class Pentecostals and 14 with a comparison group of lower-class Pentecostals. To identify middle-class actors in the Pentecostal movement, the author used demographic characteristics such as job position, economic income, and education above Argentinean average. Interviewees labeled as "middle class" had finished at least secondary degrees (in most cases university degrees and in some cases even doctorates), and earned a minimum monthly household income of 820 Argentinean pesos per household member (the equivalent of 231 US Dollars, based on an exchange of 3.55 Arg. Pesos per US $, in the second quarter of 2009). The interviews usually took place in the church buildings or homes of the interviewees and lasted between one-and-a-half to three hours. Interviews were undertaken in Spanish and entailed: 1) biography, 2) religious likes and dislikes, and 3) the demographic profile of the interviewees (income, age, household size, family background etc.). They also included the display of video sequences from different Pentecostal churches: interviewees were asked to respond to these video sequences with their impressions and opinions about the material.

The 10 interviews with pastors were usually undertaken in their churches. They included questions about the history and social composition of the church, church activities (groups, social projects etc.), practices during the services, and their differences with regard to other Pentecostal churches.

The audio recordings of the interviews were sent in for transcription and analysed with Atlas.ti, a software program for qualitative data analysis ([46] 2004). In analysing the interviews, the author attributed codes to segments of the interviews, inter alia, for the communicated likes and dislikes (e. g. dislike shouting) as well as the practices (e. g. speaking in tongues) of the interviewees. In the interest of preserving anonymity, the interviewee's names have been changed. Though the research took middle-class and lower-class Pentecostals into account, the focus of the following analysis is on the middle class Pentecostals and their boundary conversion strategy. Boundary conversion involves two sub-processes: boundary blurring and boundary redrawing. The description of the boundary redrawing mostly uses material from the interviews, as these enable illustrating the distinction processes against lower-class Pentecostalism.

Contrasting research on the growing acceptance of lowbrow practices among upper- and middle-class actors (Bennett et al. 2010; Jaeger & Katz-Gerro 2010; Katz-Gerro 2002; Peterson & Kern 1996; van Eijck 2001; Warde et al. 2007), being Pentecostal is not well-accepted – or even honorable – in the Argentinean middle class. Many of the middle-class Pentecostals in this study experienced social tensions due to their affiliation with this lowbrow religion. For instance, Alberto reported being labelled as "brainwashed" and being subsequently expelled from his old group of friends after stating his Pentecostal conversion; Andrea, working as a lawyer, said that she was harassed by her boss due to her religious affiliation while Fabian stated that he felt like the evangelical character Ned Flanders from the cartoon series The Simpsons, as colleagues constantly made fun of his Pentecostal affiliation. Rather than being honorable, the involvement in this practice can be shameful and lead to social exclusion from middle-class circles. As has been shown in other studies (Friedman & Kuipers 2013; [36] 2016), having an unclear social position and being confronted with devaluation encourages middle-class actors to undertake boundary processes. Driven by their "desire for self-verification" (Hendley 2016), they renegotiate the value of their activities.

Similar to working-class actors in elite environments (Granfield 1991), the middle-class Pentecostals in this study seek to escape the lower-class stigma. They employ, however, different strategies as they cope with their class ambiguity, not by hiding their class background, but by underpinning it. They renegotiate the appropriateness of their religious affiliation by converting the external boundary between the middle class and Pentecostalism into an internal boundary within Pentecostalism that separates appropriate highbrow Pentecostalism from inappropriate mass Pentecostalism. To this end, they engage in two sub-processes: (a) blurring of boundaries towards the middle class and (b) redrawing of boundaries in opposition to styles of Pentecostalism that are experienced as inappropriate from a middle-class point of view. They undertake these two processes on various grounds: education, levels of expressivity in churches services, and morality.

Middle-class Pentecostals blur boundaries between the middle class and Pentecostalism. The efforts to reduce the solvency of these boundaries become visible in the comments of middle class Pentecostals and the style of Pentecostal Middle-Class Congregations (MCCs), which is shaped by an affinity for tidiness, excellence, education, rationality, and self-control. These characteristics constitute vital elements of the Argentinean middle-class imaginary (Adamovsky 2009; [20] 2007; Guano 2004; [108] 1994, 2001; Tevik 2006). By emphasizing these characteristics, middle-class Pentecostals parade their middle-class-ness and, thereby deconstruct the boundary between Pentecostalism and middle class. The following description illustrates the boundary blurring by summarizing specific middle-class features of the studied MCCs.

Middle-class Pentecostals attribute a fundamental importance to education, as the command over cultural capital allows them to underline their middle-class-ness (Bourdieu 1979; Bourdieu & Passeron 1990). Whereas Pentecostalism is notorious for its focus on the Spirit and its emotionality, middle-class Pentecostals emphasize the need of educational training and rationality. This becomes manifest in the organization and infrastructure of their churches as well as in the style of sermons.

Pastors of MCCs have usually received higher education, with some of them holding PhD degrees, and often work as teachers or directors in some kind of evangelical educational facility. Moreover, middle-class congregations offer a vast array of theological and non-theological courses. Some congregations collaborate closely with bible institutes to develop their religious course program. Finally, some middle-class churches offer school education in the form of church-run primary and, occasionally, secondary schools. Another feature are bookstores, which sell a variety of evangelical books from Anglo-Saxon authors besides bibles, music CDs, and the latest issues of local evangelical newspapers.

An additional element that enables underpinning the level of education are the sermons. While other Pentecostal congregations usually place an emphasis on spiritual practices, the sermons are paramount in MCCs and make up the most important part of the church service by far. They tend to be longer than in other churches and more structured. Moreover, they exhibit a more intellectual style.

Middle-class preachers speak in a clearly articulated Spanish without much of the typical Porteño accent. They use a controlled voice and avoid the shouting that one can find in other churches, thereby transmitting an image of being measured and rational. Underlining their educational skills and affinities, they often use abstract terminology (e. g. loanwords from ancient Greek and Hebrew), deliver extensive background information (e. g. historical context of biblical passages) and sometimes use references to famous scholars or academic theories. In extreme cases, this makes the sermon resemble a university lecture.

The practice of music during the church services relates to middle-class tastes and its ambitions for excellence. Middle-class churches favor styles of music that match the musical affinities of Argentina's middle class. Contrasting the prevalence of Latin American rhythms (e. g. Cumbia) in other Pentecostal churches, Anglo-Saxon pop ballads, classic music, and sometimes Jazz inspire the music in MCCs. A typical type of music resonating well with the middle class is Hillsong, as one interviewee clarifies:

"(...) Hillsong for example. That is what captures the middle class – verses that are coordinated and thought out, with some element of poetry, and music that is much more thoughtful than "tachin, tapum, tachin tapum". That is what most appeals to the middle classes." (Carlos)

This type of music appears as more "coordinated" and "thoughtful" than the "tachin, tapum" of cumbia to middle-class actors. Bands playing during church services consist of trained musicians. They include at least a piano, electro bass, guitar, drums, a singer, and several background singers. Typical are also a saxophone player, a chorus of between seven and twelve singers, and other wind and string instruments.

Aside from the style of music, the importance of education, and the tidiness of the infrastructure addressing the expressivity of Pentecostalism as a substantial boundary towards the middle class is paramount for the boundary blurring processes. Pentecostal church services in Argentina are known for their expressive and emotional style (Pew Research Center 2014: 67). The expressivity and emotionality contribute to their lower-class image whereas the imaginary of the Argentinean middle class is marked by self-control and rationality. In particular, the speaking in tongues, exorcisms, and faith healing practices do not resonate well with Argentina's middle class.

Turning Pentecostalism into a middle-class practice, MCCs emphasize self-control rather than ecstatic spiritual experiences: they show significantly lower levels of expressivity than other Pentecostal churches and ban the aforementioned practices from their church services. Instead of yelling, trembling, or speaking loudly in tongues, they control and limit their physical gestures and expressions and avoid the showing of strong emotions. MCCs establish boundaries for acceptable expressions and closely guard them: outbursts and uncontrolled behavior in the form of Pentecostal practices such as shouting, jumping, speaking in tongues, and other spiritual practices are considered as inappropriate in the church service.

The aforementioned middle-class features do not appear suddenly. Churches have developed this style of Pentecostalism over time, as their membership has become more middle class. An example for this is Asamblea Cristiana. This congregation is one of the oldest Pentecostal churches in Argentina, starting its activities already in 1909. Evolving towards a middle-class church, the congregation has left behind many of the old Pentecostal practices in its church services. According to a pastor, the church originally placed emphasis on religious experience and the Holy Spirit while the church's leadership lacked education. However, over time the social composition of the church has changed: some of the existing members experienced upward mobility while others who newly affiliated with the church had a middle-class background. With the subsequent changes in the class background of members, a new leadership established a stronger focus on education and Bible teaching. Today, its church services show a sober and controlled style. Expressions of emotional involvement are hardly visible. Shoutings of "Gloria a Dios!" or similar exclamations, people dropping on the ground, tumbling in ecstasy, or crying out loud are absent. As members closely control their physical expression in the church services, many participants are reluctant to sing and clap their hands even during the musical praise.

The blurring of boundaries between Pentecostalism and the middle class is a process that takes place as Pentecostal churches become more middle class, thereby evolving a Pentecostalism that is more adapted to what middle-class peers consider appropriate. Pentecostalism becomes more educated, and controlled; in sum: more middle-class.

The deconstruction of boundaries between the middle class and Pentecostalism goes hand in hand with a boundary redrawing process within Pentecostalism. Middle-class Pentecostals draw boundaries between their style of Pentecostalism and that of other Pentecostals. This boundary drawing process becomes manifest in the style of MCCs as well as in the self-presentation of middle-class Pentecostals, who emphasize their distinctiveness vis-à-vis other Pentecostals. MCCs and their members draw boundaries towards other Pentecostals on the basis of their expressivity, education, and morality.

The expressivity and emotionality of church services play a particular role for the boundary process. Watching video sequences from lower class church services, middle-class interviewees criticize these churches for their loud and emotional atmosphere. Distinguishing themselves from the noisy Pentecostals, they portray their own style as calm. The following statement from a middle-class interviewee, Eduardo, illustrates this boundary drawing process:

"I am Pentecostal, but I don't like noisiness. You know, the Pentecostal is characterized by a lot of shouting." (Eduardo)

Characterizing Pentecostals as noisy, Eduardo distinguishes himself from this attribute by stating his dislike of "noisiness" and "shouting." Eduardo reflects the general tendency among middle-class Pentecostals to identify with a less expressive style of Pentecostalism. Moreover, watching videos of other Pentecostal church services during the interviews, middle-class Pentecostals distance themselves from their practices. For instance, Pablo states with regard to a video of faith healing practices in a Pentecostal church:

"It is for ignorant people, those who can believe in all this, people who are needy, who have problems and think that they will find solutions here." (Pablo)

He connects the Pentecostal practice with a specific type of people: those who are "ignorant," "needy" and "have problems." Thereby, he regards this Pentecostalism as representing poorer sections of the population and distances himself from it. Maria and Carlos also share a critical perception of other Pentecostal churches. They refer to the lack of education in Pentecostal churches, which they perceive as a problem:

"If you are going to González Catán [lower class neighborhood outside the city of Buenos Aires], you will find a Pentecostal church every in second block, more or less, everyone is a pastor, without training. You understand? There are no requirements (...) someone opens a place and you see that they are not prepared, not for speaking and much less for knowing really the Word of God." (Maria; parenthesis: author)

"In many Pentecostal churches, and outside the city in the provinces, this occurs a lot. They have experiences with God, and then it's over. They read the Word a little bit, and that's it. So they are becoming cults, forming congregations thinking that God speaks to them and that God controls them (...) and they are really doing a poor job, because when you get there, you realize that they don´t understand anything. They are so far away from the Word of God and the guidance that God offers us in the Bible, because of a lack of knowledge." (Carlos)

Middle-class Pentecostals describe the alleged lack of education in other Pentecostal churches as a crucial problem that inhibits them to really grasp the Word of God. Thereby, they suggest that the middle-class dominion over education facilitates an easier access to the "real" essence of Christian faith.

Additional grounds on which middle-class Pentecostals establish their difference concern the morality of Pentecostal practice (Lamont 1992). They portray other Pentecostals as pragmatically following their interests. Again, middle-class Pentecostals describe the posture of the others as misleading and neglecting the real essence of Christian faith:

"I have talked to a lot of people who go to the Universal Church (...) who are only after economic prosperity, health, to resolve family problems, to find an instantaneous solution to everything, with water or prayer by the pastor, but without making any real effort themselves. They don't understand and don't want to make an effort to forgive or to change their lives and follow the principles of Christ." (Andrea)

In sermons, similar criticisms of other Pentecostals appear from time to time. These criticisms lament that they pursue easy and fast solutions for their daily problems and forget the afterlife. Middle-class Pentecostals reject the spiritual pragmatism and selfishness of other Pentecostals and interpret their own practice as selflessly directed towards God. Middle-class Pentecostals associate the "inappropriate" Pentecostalism with a specific class of actors: those lacking education and suffering from their (material) needs, as they search of fast solutions (in other words: Latin American lower classes). This form of Pentecostalism contrasts with their style of Pentecostalism. By highlighting in what way their practice of Pentecostalism is different from this Pentecostalism, they draw boundaries within Pentecostalism.

This boundary drawing may also take place within congregations. In most cases, middle-class Pentecostals who gather in congregations dominated by the middle class can observe the religious style of lower class Pentecostals from a comfortable distance. However, in some cases, the lower-class Pentecostal is not a distant other, but a co-congregant, as for instance, in the case of Javier: given his family ties, he is bound to a congregation located in a slum that is mostly populated by those coming from the poor neighborhood. He describes the church as a typical congregation for "slum people" and harshly distances himself from the "slum culture" which he sees manifested in the church. Thereby, Javier draws symbolic boundaries between himself as a middle class Pentecostal and his co-congregants:

"(...) it so happens that we come from European roots, from European culture, and this is a church that is really for slum people,(...) – so there are things that shock us sometimes, behaviors that – perhaps the slum people are very disorganized, they have a very chaotic way of being, that we don´t like, and when you belong to a community like that, it´s like they impose it. We, as Europeans, feel discriminated against. It´s very difficult to make friendships, because it's like the slum person doesn´t have the same level of culture that we do." (Javier)

Distinguishing between "slum culture" – which he associates with a lack of organization – and the culture of those with European roots, he draws sharp boundaries in opposition to the Pentecostalism of the "negro," the "slum people," which he regards as inappropriate for people like him. His boundary work reflects usual boundary drawing processes of the Argentinean middle class: the middle class is portrayed as European and civilized in contrast to the non-European "negro" from the lower class, lacking a legitimate command of culture (Adamovsky 2009; [109] 1994).

The creation of boundaries between a "middle-class us" and a "lower-class them" within Pentecostalism involves the construction of an "inappropriate other": middle-class Pentecostals portray their difference from the popular Latin American masses that are conceived of as noisy, uncivilized, non-European, uneducated, and guided by their emotions and material needs. The boundary spanning processes involve a "symbolic exclusion" (Bryson 1996) of illegitimate styles of Pentecostalism. Thereby, it creates a new boundary within Pentecostalism between appropriate, "highbrow" middle-class Pentecostalism and inappropriate mass Pentecostalism. This divide often runs along congregations, separating MCCs from other Pentecostal congregations. However, in some cases, middle class Pentecostals may not affiliate with MCCs, such as the aforementioned Javier. Here, the boundary drawing involves a distancing from the own community and leads to the creation of symbolic boundaries within congregations between "snobbish" middle class Pentecostals and "lowbrow" mass Pentecostals. In particular, Pentecostal mega-churches that attract believers from different backgrounds are likely to produce such internal boundary-drawing processes.

Both boundary processes – boundary blurring and the drawing of new boundaries within Pentecostalism – culminate in a boundary conversion: the external boundary separating middle class and Pentecostalism is converted into an internal boundary within Pentecostalism differing between appropriate middle-class forms of Pentecostalism and mass Pentecostalism.

This contribution shows that class-marking processes do not diminish in importance. Even when middle-class actors increasingly choose lowbrow practices, the drawing of class distinctions remains effective. Contrasting narratives about omnivorousness, class ambiguity appears to encourage class distinction. When class-dissonant behavior does not appear as honorable (Jarness & Friedman 2017), actors are likely to employ boundary work to negotiate the value of their mismatching practices. With rising class ambiguity, there is a need to study these boundary strategies instead of simply diagnosing the disappearance of class distinction.

The article illustrated this by exploring the involvement of middle-class actors in Pentecostalism, a religious practice considered to be lower-class. The studied middle-class Pentecostals evolve strategies to renegotiate the value of their religious practice by engaging in a boundary conversion process: they convert existing boundaries between the middle class and Pentecostalism into internal boundaries within Pentecostalism, between appropriate, "highbrow" and mass, "lowbrow" Pentecostalism. Similar to other middle-class actors that engage in lowbrow practices (Jarness 2015; Peters et al. 2018), they outline their distinctiveness by drawing boundaries against those normally undertaking Pentecostalism. To this end, they generate differences in expressivity, education, and morality. At the same time, they blur the boundaries towards the middle class. They parade a middle-class style of Pentecostalism that employs established symbols of middle-class-ness: tidiness, excellence, education, rationality, and self-control.

While research about the class-culture nexus has mostly focused on other domains (e. g. food, music, clothing), this study illustrates that religion is a valid field for studying class. Therefore, the results of this study are not limited to religion, but provide general insights into class ambiguity and the importance of boundary work.

This study adds to existing research, stressing that the "how" of cultural consumption is becoming increasingly important. Actors do not necessarily mark their class position through "what" they consume but the modes ("how") in which they consume it (Daenekindt & Roose 2014; Jarness 2015; Peters et al. 2018). If actors increasingly transgress established class boundaries, and exclusive links between certain practices and social classes diminish (Bryson 1996; Jaeger & Katz-Gerro 2010; Katz-Gerro 2002; Peterson & Kern 1996; van Eijck 2001; Warde et al. 2007), then the marking of social class may increasingly consist less in choosing lowbrow or highbrow practices, and more in drawing, blurring, or shifting boundaries when conducting (or liking) cultural practices. The rigid highbrow/lowbrow dichotomy becomes substituted by micro-boundary processes of value-negotiation ([91] 2003; Skeggs 2004), redefining in each case what is highbrow and what is lowbrow. In other words, instead of studying what type of practices actors like and undertake, we need to explore what kind of boundary processes they engage in when dedicating themselves to specific practices that apparently cross established class boundaries. Additionally, there is a need to consider that these boundary processes depend on the given contextual settings in which actors move ([50] 2019). As such, actors may engage in the drawing of class distinctions in publicly visible front regions while they abstain from doing so in private back regions.

Taking these boundary processes into account may amplify our knowledge of how class is negotiated on a daily basis, even when it appears as though class exclusivity is diminishing. To this end, an encompassing theory of boundary processes is necessary to classify different types of boundary strategies and the conditions under which they become employed (Wimmer 2008a). This article has contributed to this endeavor by drawing attention to a specific boundary strategy: boundary conversion.

Boundary conversion enables actors that self-identify as middle class but engage in lowbrow consumption to signal "their place": it blurs boundaries between a lowbrow practice and the middle class and redraws class boundaries within the given practice. Thereby, it creates a new middle class style of the given "lowbrow" practice that distances itself from its original "lower class" features. Peters and van Eijck's (2018) research on middle class karaoke indicates similar processes: middle class actors perform karaoke in an ironic fashion to mark distance and cultural superiority, thereby clarifying their own middle class position. In times of rising cultural plurality and omnivorousness, boundary strategies make it possible to deal with the increasing class ambiguities that these processes produce.

As boundary conversion is undertaken to manage conflicting allegiances to groups and/or practices, it is not limited to the negotiation of social class ambiguity. It becomes an option whenever actors engage in culturally dissonant behavior: crossing established group boundaries and facing the threat of losing peers' recognition (e. g. accusations of lacking group loyalty), actors can endeavor to convert the existing external boundaries into internal boundaries. This boundary process becomes likely when alternative strategies are not feasible, as the conflicting allegiance cannot be easily changed. Accordingly, boundary conversion may form a widespread strategy to deal with culturally dissonant behavior. However, it is also a demanding strategy as it involves the blurring of existing boundaries between the two conflicting allegiances and the redrawing of new boundaries within one of them.

Further studies may expand our knowledge about the boundary strategies of culturally dissonant actors and their relative occurrence by revealing alternative boundary strategies and unfolding how, to what extent, and under what specific circumstances specific strategies are undertaken. Ultimately, studying boundary strategies may help us to understand how the maintenance of group boundaries is possible, under rising plurality and apparently increasingly less exclusive behavioral patterns.