Societal acceptance increases Muslim-Gay identity integration for highly religious individuals… but only when the ingroup status is stable

ABSTRACT

Reconciling religious beliefs with a sexual minority identity can be challenging. After coming out, many gay men report to renounce their religious identity or experience increased identity conflict between their religious and sexual identities. Giving up one’s own identity or identity conflict are known to predict negative wellbeing, and it is important to find ways to reduce this conflict and increase identity integration. In this experiment, we conceptualized identity integration as a social creativity strategy, and we examined whether societal acceptance (vs rejection) and ingroup experience (e.g., whether gay community feels stability or improvement about their status) would alter one’s level of Muslim-gay identity integration. We found that Muslim-gay identity integration was highest among highly religious gay men when societal acceptance was present and ingroup experience was stable. Overall, we discuss our findings by drawing parallels between social identity theory and bicultural identity integration framework, and provide implications to increase identity integration for those with multiple conflicting identities.

KEYWORDS:

• Identity integration
• multiple identities
• social creativity
• social identity theory
• gay
• muslim

View correction statement:

CORRECTION NOTICE

Introduction

People belong to multiple social groups (e.g., work, family, nation, ethnicity) and they are motivated to maintain these group memberships (Tajfel & Turner, 1979), because group-based identification can provide a sense of belonging, comfort, meaning, and purpose, which are crucial determinants of psychological wellbeing (Jetten et al., 2017). The social cure approach – which acknowledges the multiple group memberships people have – suggests an additive effect, whereby identifying with multiple groups can be increasingly protective against negative wellbeing outcomes (Jetten et al., 2017). However, this not only depends on the relevance of this identity to one’s self-concept (Turner et al., 1987), but it also depends on how compatible these identities are (Anderson & Koc, 2020; Koc & Vignoles, 2016). People often feel “caught” between two identities (Benet-Martínez et al., 2002), which may result in an identity conflict. The nature of this conflict depends largely on the content of the identities at stake and their social contexts, and this conflict may inhibit identification thereby reducing the positive gains these social identities offer (Koc & Vignoles, 2016). However, identity integration, as opposed to identity conflict, is known to predict positive health and wellbeing outcomes (Chen et al., 2008; Ferrari et al., 2015). Moreover, emphasis on a single identity may be linked to negative outcomes for the other identity. For instance, among Christian gay men, higher religious identification was linked to shame and guilt regarding one’s gay identity. However, religious-gay identity integration attenuated the negative impacts of religious identity on shame and guilt (Anderson & Koc, 2020). Therefore, it is important to understand how multiple identities relate to one another and how their integration can be achieved and maintained. In this study, we focus on the incompatibility of religious and sexual identities in the context of Turkey – a predominantly Muslim country where most people identify as religious and live under the societal expectations of highly intertwined religious and cultural norms.

Social identity approach to multiple identities

According to the social identity approach, comprised of both social identity theory (Tajfel & Turner, 1979) and self-categorization theory (Turner et al., 1987), social identities refer to how people categorize themselves in relation to others and how they see themselves through the lens of their group memberships. Individuals can use individual traits to define themselves (i.e., “I” and “me”), or they can use features from their group memberships (i.e., “we” and “us”). The latter particularly relies on drawing similarities to other group members with which the individual identifies with (i.e., ingroup members) and making oneself different from whom the individual does not share group membership (i.e., outgroup members). For a long time, the social identity approach has been used to study intergroup phenomena (e.g., Ellemers et al., 2002), yet over the last decade the importance of identification on individual health and wellbeing has gained significant recognition and has been well documented (e.g., S. A. Haslam et al., 2009; Jetten et al., 2012, 2017). In particular, the social cure research suggests that social identification with groups improves wellbeing through the processes of finding meaning, belonging, self-esteem, efficacy, and control in life (e.g., Cruwys et al., 2014; Greenaway et al., 2015; Jetten et al., 2017).

However, social identification with a certain category or group can be a complicated process for individuals with two or more group memberships because one’s multiple identities are not simply independent of each other, but also that they intersect each other in various ways (Van Dommelen et al., 2015). In this case, the combination of these dual or multiple identities might be compatible, or they may be in conflict with one another. In particular, multiple social identities might differentially provide advantaged or disadvantaged positions in society to the individual depending on the context (Case et al., 2012). For instance, while being a male can be easily perceived as compatible with being heterosexual, it may be seen as incompatible with homosexuality in various cultural contexts including Turkey where this study was conducted (Connell, 2005; Eslen-Ziya & Koc, 2016; Koc & Vignoles, 2016).

Although the social cure research has studied the link between multiple identifications and wellbeing (e.g., C. Haslam et al., 2008; Jetten et al., 2015; Kinsella et al., 2020), not much attention has been paid to the compatibility between these identifications. One notable exception is the paper by Iyer et al. (2009), which mostly focuses on adjustments after life transitions such as transition to a university. In this case, compatibility between the identities was conceptualized through the idea of identity continuity (e.g., Sani, 2008). In particular, Iyer and colleagues stated that they were interested in how one’s new identity is perceived to fit with the network of their own existing or old identities. Accordingly, the emphasis was on an existing identity network and whether the new group memberships can be added to this network. However, we argue that a bicultural identity framework might provide better insight into the study of the integration between two existing social identities (e.g., Benet‐Martínez & Haritatos, 2005) by focusing on the compatibility or otherwise of these social identities at a time.

Bicultural identity integration framework

Bicultural identity integration (BII; Benet-Martínez et al., 2002) was developed to address individuals’ subjective experiences of managing dual identities. BII incorporates two bipolar components of identity integration: the first one is blendedness versus compartmentalization, capturing the “degree of dissociation versus overlap perceived between the two cultural orientations” (QL. Huynh et al., 2011, p. 830), and the second one is harmony versus conflict, capturing the “degree of tension or clash versus compatibility between the two cultures” (p. 830). Similar to the social cure research, a high level of identity integration is found to be associated with various outcomes such as higher self-esteem, lower anxiety, and greater life satisfaction in bicultural individuals (see the meta-analysis; Nguyen & Benet-Martinez, 2012). In recent years, research using this framework went beyond the cultural identities and focused on the applications to different forms of social identities such as work and old age (Manzi et al., 2019), sexual and gender (Koc & Vignoles, 2016, 2018), and sexual and religious identities (Anderson & Koc, 2020). In this study, we take BII as a framework to understand the experiences of gay men negotiating their sexual and religious identities in Turkey.

Going beyond merely adding a new identity in the network of existing identities as in the case of the social cure research, BII framework also recognizes how individuals experience the identity confusion between their incompatible identities, to what extent they manage to respond to the dual expectations from both sides, and finally how they negotiate the value clashes caused by the incompatibility (Benet‐Martínez & Haritatos, 2005). Moreover, individuals who are bicultural often experience multiple stereotypes and different pressures from both the communities they are involved in terms of expressing their loyalties and behaviors (LaFromboise et al., 1993). This not only increases the stigma experienced related to both identities in society, it also creates a unique pressure trying to negotiate these separate expectations and pressures. Finally, identity integration can be seen as a “combined, third, emerging culture” (p. 1019; Benet‐Martínez & Haritatos, 2005). This is where we believe a social identity approach can enhance the BII framework.

Drawing links between social identity approach and bicultural identity integration

As discussed earlier, when individuals categorize themselves as parts of groups, they create ingroup and outgroup memberships based on similarities and differences drawn. For instance, in the case of conflicting sexual and religious identities for a Muslim-gay man, the ingroup members typically include people who also identify as Muslim-gay. On the other hand, although Muslim-gay men share the same identities with Muslim people or gay people respectively, these groups individually might constitute two separate outgroups for Muslim-gay men. Therefore, a Muslim gay man is not necessarily a part of the Muslim community or the gay community because the intersection of these two identities creates different norm and value expectations from Muslim gay men (i.e., a third culture) especially when this identity is salient. Moreover, the social cure effect was found to be moderated by group esteem (i.e., how groups are perceived and valued in society; DeMarco & Newheiser, 2019). Specifically, high-esteem group memberships enhanced wellbeing whereas low-esteem group memberships undermined wellbeing. Similarly, in a given context, when one identity is a privileged or high-esteem one (e.g., the Muslim identity in a religious country) and the other one is a disadvantaged or low esteem one (e.g., the gay identity in the face of cultural and religious stigma), we argue that the identity could be threatened more severely, and the associated identity conflict could be even stronger (see Case et al., 2012). Therefore, it is important to understand how Muslim gay men negotiate their identities when each identity (and their integration) may require different cognition, affect, and behavior from them.

Ways of coping with the identity threat

According to social identity theory, when a person’s identity is threatened, they might engage in certain identity management strategies to deal with the negative social identity (Tajfel & Turner, 1979). The chosen strategy depends on the perceived legitimacy of the intergroup situation, perceived stability of group status, and the perceived permeability of group boundaries (Tajfel, 1982). One can engage in individual mobility, when group boundaries are permeable and the group members can exit the group or can achieve passing by making others believe that they are something that they actually are not when the stigma is less visible. However, research shows that self-group distancing is negatively related to self-esteem and life satisfaction among gay men (Bourguignon et al., 2020), and it is not an ideal way of coping.

When group boundaries are not permeable, individuals engage in other coping strategies. One such strategy is social competition, which involves challenging the status quo and engaging in collective action to improve the status of the disadvantaged group. This is likely to happen when the intergroup situation is perceived to be illegitimate and unstable. For instance, Velez and Moradi (2016) found that engaging in high levels of collective action helped attenuate the negative link between discrimination and wellbeing among sexual minority individuals in the US.

However, when the intergroup situation is perceived legitimate and stable, one can engage in social creativity. We argue that identity integration can be conceptualized as a social creativity strategy. The aim of social creativity strategies is not to change the social structures, rather restore the threatened social identity (Tajfel & Turner, 1979). The three commonly examined strategies of social creativity involves selecting another group as a comparison target (e.g., another disadvantaged group in a lower status), redefining the comparison dimension with other groups (Jetten et al., 2005), or changing the perceived importance of group attributes (Becker & Wright, 2011). One thing all these strategies have in common is that they change some aspect of social comparison, but they still continue to endorse the given identity within its limits. Identity integration is therefore similar to social creativity in that those high in identity integration know how to negotiate through different dimensions of their multiple identities and manage to perceive them as compatible, harmonious, and cohesive. However, those low in identity integration struggle with different dimensions of their identities and find it hard to deal with the specific tensions created by them. Therefore, social creativity in the case of multiple identities could be a strategy to negotiate different dimensions of these multiple (conflicting) identities and find ways to (internally) integrate them. As a coping strategy, perhaps individuals choose new comparison dimensions from each identity or they update the values attached to each identity (see also Breakwell, 2015). Ultimately, this helps them see these identities more integrated rather than conflicting as an outcome of this strategy. In this study, by focusing on the basic principles of social identity theory, we aim to improve our understanding of Muslim gay identity integration as a social creativity strategy.

Research on the incompatibility of religious and sexual identities

Research shows that religion and attitudes toward homosexuality correlate negatively (Anderson et al., 2018). LGBTQ+ individuals tend to internalize such beliefs and those who were raised with religious beliefs or in a religious family often struggle to reconcile their sexual and religious identities (Anderton et al., 2011; Subhi & Geelan, 2012). Due to the tenuous nature of the relationship between these identities, they experience stigma from both groups. For instance, LGBTQ+ individuals not only hide their sexual identities in religious circles, they also hide their religious identities among other LGBTQ+ individuals (Beagan & Hattie, 2015; Mitha et al., 2021; O’Brien, 2004). This in turn leads to feelings of guilt, shame, or defectiveness (Ritter & O’Neill, 1989), and might be one of the explanations of high rates of depression, anxiety, suicidal ideation, substance use and self-harm within the community (Pietkiewicz et al., 2016). Rodriguez and Ouellette (2000) identified four possible ways to deal with this conflict similar to the identity management strategies of social identity approach: reject either the religious identity or the gay identity (i.e., disidentification or individual mobility), compartmentalize them by maintaining rigid barriers between them and denying affiliations with each identity in the context of the other, or finally, achieve identity integration by synthesizing these identities into new self-concept. As expected, identity integration was seen as the best coping strategy to manage this conflict. Therefore, we focused on ways to achieve and maintain identity integration.

The context of the current study

Turkey is officially a “secular” country. Its constitution requires the separation of religious and state affairs. Unlike some other Muslim countries like Malaysia or Saudi Arabia, where there is the rule of Sharia law, legal decisions in Turkey are not supposed to be based on religious rules or traditions. For instance, the status of homosexuality has always been ambiguous in Turkey: it has never been illegal nor has it ever been officially recognized. This may seem like a positive condition in a predominantly Muslim society; however, in practice the status of the gay community and how they are treated by the society and the government are quite different. For instance, the Pride parade and relevant activities in Turkey have been canceled in the last few years by the governors, and the non-governmental LGBTQ+ organizations have been raided and closed down by the police. Moreover, the Presidency of Religious Affairs, a state funded organization, occasionally makes stigmatizing public statements about homosexuality. Recently, while answering a question on its website regarding the state of homosexuality in Islam, Presidency of Religious Affairs stated that homosexuality is the absolute limit, it involves sinful acts, homosexual people are condemned by God, and a good Muslim must refrain from such ugly deeds persistently, and if they committed such a sin (i.e., homosexuality), they must repent in order not to commit again. This inevitably affects how the society perceives the LGBTQ+ community as well as creating identity conflict for LGBTQ+ people who also identify as Muslim.

Research shows that anti-gay attitudes toward gay men and lesbians in Turkey are driven by extrinsic religiosity orientation and religious fundamentalism (Anderson & Koc, 2015), and strong religious identification is related to lower wellbeing for gay men (Koc & Vignoles, 2016). This is because most Muslims perceive homosexuality to be incompatible with Islam (Kugle, 2010). Ultimately, such negative societal perceptions are internalized by gay men, and undermine their sense of belonging, and are linked to negative wellbeing (Eslen-Ziya & Koc, 2016; Hinton et al., 2019). However, data from the World Value Survey shows that societal acceptance of homosexuality has been steadily increasing across the world even though it is slower in Muslim countries (Anderson et al., 2021). However, there is a high prevalence of pluralistic ignorance regarding same-sex issues such that people overestimate the societal disapproval of homosexuality (Eisner et al., 2020). Therefore, gay communities should be reminded that the societal acceptance is increasing, and the consequences of this acceptance should be investigated.

Overview of the present study and hypotheses

There are a few studies that have attempted to increase identity integration. For instance, Koc and Vignoles (2018) primed participants with global identification to increase their gay-male identity integration. This study found that increased identity integration was linked to higher wellbeing and access to other ingroups members who accepted the participants as both gay and male (e.g., conceptualized as gay-affirmative social spaces). However, in this study, instead of focusing on other resources participants may have (e.g., participant’s other identities), we decided to focus on a societal variable: societal acceptance (vs. rejection). Previous research shows that family acceptance positively predicts gay identity development and psychological adjustment among Israeli gay men (Elizur & Ziv, 2001). We believe societal acceptance can have similar consequences and might validate the identity of Turkish Muslim gay men and help increase Muslim gay identity integration.

Moreover, people engage in social creativity when the disadvantaged group status is stable. Therefore, we opted to manipulate stability of the ingroup status of the gay community as stable vs. changing. Given the recent events (e.g., Pride March banning), activism in Turkey has become dangerous and nearly impossible. The path for social change is very limited. Therefore, gay men might rather choose to integrate their Muslim gay identity if the gay community is experiencing stability in their status.

Finally, research shows that high identifiers tend to choose social creativity and social competition strategies over individual mobility, which is often preferred by the low identifiers (e.g., Ellemers et al., 1997). Therefore, we test the moderating effect of religious identification.

Accordingly, we investigated how societal acceptance (vs. rejection) and status stability (vs. change) impact Muslim-gay identity integration among Turkish Muslim gay men. Moreover, we tested the moderating effect of participants’ own religiosity. We hypothesized a three-way interaction. In particular, we expected highly religious participants to have higher Muslim-gay identity integration when there is societal acceptance of homosexuality rather than societal rejection. The rationale behind this prediction is that participants may find it easier to integrate these identities when there is a rewarding match between societal religious norms and their personal religious convictions. Moreover, we only expect this interaction to work in the ingroup stability condition, because perceiving their status to be stable will force participants to integrate these potentially incompatible identities whereas perceiving their status to be unstable (i.e., possible to change) might lead them to other identity maintenance strategies like social change.

Method

Participants

Participants were recruited online in Turkey through an advertisement in the gay dating application “Hornet”, a commonly used method to reach men having sex with men samples (Koc, 2016). The study was advertised as a study on experiences of being gay in Turkey and the ad on the application led them to an online survey created with Qualtrics. The recruitment message and all study materials were translated from English into Turkish by the first author, who is fluent in both languages.

The current sample initially included 250 Muslim participants. However, only 187 participants identified as gay (74.8%), and 56 participants identified as bisexual (22.4%), 5 participants identified as heterosexual (2%) and 2 participants identified differently (0.8%). Since previous studies found that the experiences of gay men and bisexual men are different particularly in the context of Turkey (e.g., Koc & Vignoles, 2018), we only retained gay-identifying men. The ages of the final 187 participants ranged between 16 and 57 years (M = 28.70, SD = 8.66).

Fifty six percent were students, and the remainder had various occupations such as teachers, managers, and mechanics. We measured socio-economic status (SES) by asking participants to place themselves on the rungs of a ladder from 1 to 10 (higher scores indicating higher status) in comparison with other people in Turkish society (Adler et al., 2000), and the average subjective SES was 5.28 (SD = 1.92). Eighty three percent were single, 9% were in committed relationships, 6% were in open relationships, and 2% did not disclose their relationship status. Using the Outness Inventory (Mohr & Fassinger, 2000), we also measured participants’ level of “outness” on a 7-point scale from 1 to 7 higher scores indicating more outness – how much they are “out” to a number of other groups about their sexual identity. Participants were most out to friends (M = 4.12), then their work colleagues (M = 2.64) and least out to their family (M = 2.06).

The study received ethics approval from Ethics Committee of Psychology at University of Groningen, and the data are available upon request from the corresponding author.

Design and Procedure

This study has an experimental design. After giving informed consent, participants were asked general demographic questions about their age, gender, sexuality and religion. Then the participants were randomly assigned to one of the four manipulation conditions by the online survey tool Qualtrics, all depicting different newspaper articles. The four articles represented each possible combination of societal acceptance vs rejection and disadvantaged group’s status of stability vs change. Participants had to read the article to continue the study. Afterward, they responded to Muslim-gay identity integration scale. Finally, participants received a debriefing text explaining the aims of the study and thanking them for participating.

The societal acceptance (vs rejection manipulation) included results from a fake survey about to what extent Turkish society is ready to have gay neighbors. This was based on the measures used in the World Value Survey where attitudes toward outgroups are measured by the question of wanting them as neighbors. In the acceptance condition, we stated that more and more people had positive attitudes toward gay people and a lot of people would be accepting having gay neighbors, whereas, in the rejection condition, we stated that less and less people had positive attitudes toward gay people, and a lot of people would be rejecting having gay neighbors. This was followed by a fake interview with a gay-men called Mehmet. In the stability condition, Mehmet stated that no change was recognizable within the gay community and they still faced negative attitudes, whereas in the unstable condition, he stated that things were changing for the gay community, and they felt that their status in society was improving.

Materials

Religious identification

We measured religious identification using two items tapping on importance and frequency of identification (Koc & Vignoles, 2018): “How important is it for you to identify as Muslim?” (1 = not at all, 5 = very important), and “How often do you think of yourself as Muslim?” (1 = never, 5 = always). Participants responded to these two items on a 5-point scale, with higher scores indicating higher religious identification. The correlation between the two items was strong (r = .61).

Muslim-Gay identity integration

We used an adapted version of the bicultural identity integration scale (Q. L. Huynh et al., 2018) to measure the integration between Muslim and gay identities. This adapted version contains 14 items and previously used in different identity contexts such as gay-male (Koc & Vignoles, 2016, 2018), and gay-Christian (Anderson & Koc, 2020). Participants responded to these 14 items on a 6-point scale from strongly disagree (1) to strongly agree (6). An example item is: “My gay and religious identities are complementary.”

Since this scale was adapted to gay and Muslim identities for the first time, we could replicate the original factor structure. We modeled two correlated latent factors for harmony vs. conflict and blendedness vs. compartmentalization dimensions, as well as an uncorrelated method factor that loaded onto every item fixed at 1 to account for acquiescent responding (Welkenhuysen-Gybels et al., 2003). The model fit was assessed by the following cutoff values: CFI ≥ .93, RMSEA ≤ .08, and SRMR ≤ .08 (Kline, 2005). However, at first the model was not a good fit to the data; χ²(75) = 212.49, p < .001, CFI = .88, RMSEA = .10 (90% CI = .08, .12), SRMR = .18. After examining modification indices, added three covariances between the items within the same factor to account for common antecedents, and reran the analysis. This final model was an acceptable fit to the data; χ²(85) = 151.46, p < .001, CFI = .93, RMSEA = .08, (90% CI = .06, .10), SRMR = .08, and all items loaded significantly onto their respective factors (|βs| ≥ .27, p = .008). After reverse scoring the negatively worded items, we computed a mean score with 14 items and used it in all our analysis. The scale yielded an acceptable estimate of reliability (α = .84)

Results

To test our hypothesis, we used JAMOVI (Version 1.6) with general linear model (Gallucci, 2019). We ran a 2 × 2 between-subjects ANOVA with a religiosity as the continuous moderator (centered), and Muslim-gay identity integration as the dependent variable. Homogeneity of variance assumption was met: F(3,170) = 2.42, p = .068.

As expected, the ANOVA revealed no significant main effects, Fs < 0.00 and ps > .197, nor two-way interactions except for a marginal interaction between societal acceptance and ingroup stability conditions: F(1,165) = 3.81, p = .053, ηp² = .02. However, no simple effects were significant, ps > .637. As expected, the three-way interaction was significant: F(1,165) = 5.57, p = .019, ηp² = .03. Simple effects analysis showed that the only significant effects were in ingroup stability condition. As can be seen in Figure 1, when the participants had higher levels of religiosity (+1SD), they had higher levels of identity integration if they were accepted by the outgroup (M = 4.52, SE = 0.25) as compared to if they were rejected by the outgroup (M = 3.72, SE = 0.21), b = 0.80, SE = 0.33, t(166) = 2.43, p = .016. On the other hand, when the participants had lower levels of religiosity (−1SD), they had higher levels of identity integration if homosexuality was rejected by the outgroup (M = 4.25, SE = 0.21) as compared to if it was accepted by the outgroup (M = 3.61, SE = 0.22), b = −0.64, SE = 0.31, t(166) = −2.08, p = .039. In the ingroup instability condition (Figure 2), there were no significant interactions.

Figure 1. The three-way interaction showing the ingroup stability condition.

Figure 2. The three-way interaction showing the ingroup change condition.

Discussion

In this study, we conceptualized identity integration as a type of social creativity strategy, and focused on conflicting Muslim and gay identities. Using the principles of social identity theory, we examined how societal acceptance and ingroup experience of stability influenced Muslim-gay identity integration among Muslim gay men in Turkey, and how participants’ own religious identification moderated this relationship. We found support for the hypothesized three-way interaction. In the ingroup stability condition, identity integration was highest for high religious identifiers if their sexuality was accepted by society (compared to being rejected). This underlines the importance of finding a match between religious societal norms and one’s own religious identification on the integration of these conflicting identities when there is no opportunity for challenging the societal structures. In the ingroup change condition, there were no significant effects.

Theoretical implications

Although the social cure recognizes the multiple identifications, the social identity approach traditionally has focused on single identifications at a time, and hence identity management strategies aim at restoring that single identity (Brown, 2019). Among those, social creativity does not aim to change the social structures, instead restore the threatened social identity (Tajfel & Turner, 1979) at least in three different ways: selecting another group as a comparison target, redefining the comparison dimension with other groups, or changing the perceived importance of group attributes. We argue that identity integration can be seen as a new social creativity strategy whereby individuals try to negotiate their conflicting dual identities. Instead of focusing on conflicting dimensions, Muslim-gay men might redefine comparison dimensions between the two identities (i.e., Muslim and gay) and alter the importance of each identity attributes to create a new third culture. The process of redefining and altering the importance could be the coping strategy, and creating this third culture (e.g., identity integration) could be as the outcome of this process. Future research should try to examine how the process and the outcome of this coping strategy are related.

Moreover, we argue that this new third culture both benefits from the identifications with each separate culture, but also creates norms and values within this culture. Therefore, we conceptualized Muslim-gay as the ingroup, and separate Muslim and gay cultures as potential outgroups. On the other hand, other theoretical frameworks like identity process theory (Breakwell, 2015) suggests that identity threat occurs, when the values attached to the content of the identity is not compatible with the self-concept – that is, individuals may try to change the values attached to the identities to circumvent the threat. In this study, it is plausible that our participants were trying to change the values assigned to attributes of each identity (as they both constitute separate outgroups) so that they are no longer incompatible with one another. Further research should investigate this possibility.

Earlier we argued that compatibility of the identities depends on the content and social context of these identities. There is ample evidence to show that religious and gay identities are in conflict, and gay men struggle to reconcile the two (Anderton et al., 2011). As people with dual identities need to respond to the expectations from both groups, they are likely to experience conflicting demands from each outgroup. However, some identities are in a more privileged position than others (Case et al., 2012), and in this case, religious identity can be seen as more important (Anderson & Koc, 2020) especially in the context of Turkey. Therefore, in this study, we examined the role of acceptance from the privileged outgroup (i.e., Muslim). When there is an acceptance from them, people who strongly identify with this identity (i.e., high religious identifiers) are more likely to benefit from this acceptance and increase their integration. Moreover, for those who less strongly identify with this identity (i.e., low religious identifiers) managed to have the highest integration under conditions of low societal acceptance. Again, this underlies the importance of the effect of the match between one’s strength of identification and the social context on identity integration. Further research needs to investigate the effects of mismatch (as well as match) on identity integration and replicate these findings across different incompatible identities.

In regards to multiple identities, religious identities have been shown to usually take the position of the privileged, more central identity part (Anderson & Koc, 2020). Thus, any other identity (that seems to be added to the self-concept) will be adjusted or incorporated into the religious identity. As the study was conducted in a Muslim-majority country, individuals who identify as strongly Muslim was expected to strive for approval of the majority, and we found this to be true. We believe this is because when someone internalizes religious beliefs and lives in a religious society, what the society thinks about parts of their identity matters. This is also in line with other research that emphasizes the importance of societal support for successful identity integration (Briones, 1998).

Finally, there is an increasing interest in revisiting and extending research on social creativity (see review, Van Bezouw et al. 2020). These studies suggest that there are new social creativity strategies emerging than previously identified. However, they continue to focus on the intergroup dimensions, and not the multiple identities. We believe our study adds to these strategies. Moreover, Van Bezouw et al. 2020 identify social creativity within the dynamic social structure such that some social creativity strategies go beyond merely restoring the threat within the existing social structures, but these strategies also questions stability and legitimacy. Accordingly, further studies should investigate whether identity integration would undermine or foster collective action.

Limitations and future research

The findings of the present study reveal factors that come into play in identity integration; however, to generalize our findings to other contexts, further research is necessary. The sampled participants were inhabitants of Turkey, a Muslim majority country. It should be further explored if these findings are specific to Islam or can be found in other religious majority contexts, such as Christian or Jewish majority. Moreover, we recruited our participants through a gay dating application. Although this might raise concerns about representations, previous research shows that using dating apps provide access to more diverse samples in comparison to recruiting on sites such as gay community centers or bars (Koc, 2016). Especially in a country like Turkey, where it is hard to be out for gay people and very difficult for researchers to have access to them, we believe this recruitment method is indeed a strength of this research.

Typically, in social identity approach, the constructs of legitimacy and stability are used interchangeably. However, in this study, we specifically focused on the stability experience of the ingroup and directly manipulated it by using an ingroup member speaking on behalf of the gay community in the experimental manipulation. On the other hand, we did not manipulate legitimacy because we thought it is nearly impossible to convince people that the intergroup situation is not legitimate in a country where there is an extremely high level of stigma. However, recent research shows that attitudes toward sexual minorities improve after there is a legalization process in the given country such as same-sex marriage (Tankard & Paluck, 2017). Further research could manipulate legitimacy of the intergroup situation by introducing a possible change in the law to undermine legitimacy and measure its effects on identity integration (although such debates can also escalate mental health issues among the sexual minorities and could be done cautiously; Anderson et al., 2020).

Moreover, it is imperative to further explore this process in diverse contexts: In theory, any identity could become privileged by reflecting the societal majority. Thus, research focussing on contexts that privilege other sexual, gender, or national identities could help solidify what the current study has brought to light. For instance, in the Western context where homosexuality is more privileged than Islam, gay men with multiple identities might face different threats to their identity, and perhaps acceptance from the gay community could be more important for identity integration than from the religious community (Koc, 2021; Mitha et al., 2021).

Implications

Recent research shows that perceived discrimination and toleration among LGBTQ+ people in Turkey undermines their wellbeing through threatening their identity needs such as self-esteem, belonging, and meaning (Bagci et al., 2020). Moreover, people tend to overestimate the rejection regarding their identities by society (Eisner et al., 2020). However, as our study showed, even a simple reminder of societal acceptance can be useful in helping people reduce the conflict between their incompatible identities. Therefore, LGBTQ+ groups should highlight positive experiences of their community within society. However, we also acknowledge that this may not always be possible. Despite the small steady increase in acceptance of homosexuality, there is still a lot of stigma around gay identity. In other research, subgroups of society where LGBTQ+ people can feel accepted for their other conflicting identities (i.e., gay-male) were also found to be beneficial for wellbeing (Koc & Vignoles, 2018). Therefore, LGBTQ+ groups can try to provide affirmative contexts where Muslim gay men are accepted and their dual identity is affirmed. These affirmative contexts can be small friendship circles where individuals can experience acceptance and inclusion, and members with similar experiences share their stories for others to feel they are not alone and solutions are possible (e.g., ingroup oriented action, see Stroebe et al., 2019). Moreover, it does not have to be through established organizations or community centers. Indeed, especially for sexual minority members who have visibility concerns and are afraid to be outed, it might be even additional pressure to become members of certain LGBTQ+ organizations and participate in events (Mitha et al., 2021). Therefore, LGBTQ+ groups should aim to provide opportunities for inclusion and for support rather than prescribe them.

Moreover, identity integration is known to be related to a number of positive outcomes (see Nguyen & Benet-Martinez, 2012). According to Rodriguez and Ouellette (2000), identity integration was also named as the best coping strategy to manage the conflict between religious and sexual identities instead of rejecting one of the identities (or both) or compartmentalizing them. Accordingly, these findings have some clinical and community implications. For instance, practitioners working with gay men (as well as the other members of the LGBTQ+ community) should not ignore the religious identity of their clients. Research shows that negative religious coping (e.g., giving up religion or being angry at God) is linked to higher distressed and lower wellbeing (Shilo et al., 2015). And our study shows that identity integration is possible, and it is a better coping strategy. Therefore, practitioners should introduce their clients more progressive interpretations of the religion which accepts and affirms homosexuality (see Jamal, 2001). Moreover, existing gay community centers and activists in the field should welcome religious gay men into their communities and behave with caution that they might be perceived as an outgroup.

Conclusion

In this paper, we attempted to bring together the social identity approach and bicultural identity integration framework to better understand the integration of two conflicting non-cultural identities. This is the first paper to conceptualize identity integration as a social creativity strategy, and we brought the first experimental evidence to increase Muslim-gay identity integration using the social identity principles. Although we believe social change is the only way to improve the status of disadvantaged groups, social creativity is equally important when the structures are stable and do not offer opportunities for change for the minorities. Our findings not only provide novel theoretical insights, but also practical implications for of gay men regarding possible interventions that might enhance their wellbeing. Future studies can aim to use this information to manipulate relevant social contexts and ingroup vs outgroup domains to help increase identity integration.

Disclosure of potential conflicts of interest

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

Additional information

Funding

This work was supported by the Australian Research Council [Australian Research Council (LP190100865)].