Mid- and Later Life Cross-Sex Friendships in Minority Ethnic Contexts: Insights From Scotland

Abstract

This article draws on qualitative research among older adults (50+ years) of South Asian heritage in Scotland to explore what cross-sex friendships reveal about the normative tenor of gender, sexualities, and intimate relationships in minority ethnic contexts. I argue that South Asian cultural norms work against the ‘patterning’ and maintenance of cross-sex friendships. When they do occur, they have to be managed with regard to familial ties and community expectations. The risks are greater for women who must deal with policing of their sexuality even as they age. Such friendships signal social change and agency within the diasporic communities. Yet they require negotiation of gendered and ethnic/cultural scripts and point to the continuing significance of kin and community.

Cross-sex friendships are ‘marginalised’ within the literature on friendship, with studies on cross-sex friendship and older age considered ‘woefully inadequate’ – a scholarly neglect that mirrors real life as friendships between men and women are relatively uncommon (Monsour, 2002: 1, 193). Yet, a focus on cross-sex friendships can contribute to understanding friendship more generally and shed light on distinctive types of friendships, especially as they raise ‘thorny issues’ from which same-sex friendships are exempt (Fehr, 1996: 152). Men and women typically develop more same-sex friendships because enduring cross-sex friendships can become ‘sites of struggle’ over sexual identities, enactments of gender, and the place of friendship relative to romance, which is, the model of cross-sexual intimacy in the heteronormative hierarchy of relationships (Rawlins, 2009: 124). Euro-American contexts place the (heterosexual) couple at the centre of personal life, and cross-sex friendships are constrained because their sexual ambiguity is perceived as a threat to the couple (Cronin, 2015; Jamieson, 1998). Some studies (e.g. Jerrome and Wenger, 1999), however, suggest that cross-sex friendships become more common in later life.

Here, I explore how issues raised by research on cross-sex friendships (focused on the White ethnic majority) play out among ageing adults of South-Asian heritage in Scotland, where: heterosexuality is hegemonic; marriage-as-institution is centred, but not the conjugal relationship; and individuals are embedded in a wider web of kin and community relationships. Within sociology, friendship has received insufficient attention in the context of ethnicity, including among studies of South Asians. In her work among Manchester Pakistanis, Werbner (2002) attributes this to the treatment of friends as ‘quasi-kinsmen’ and so to friendship requiring no further attention as a ‘distinct social phenomenon’ (p. 174). Similarly, writing on South Asian women in East London, Mand (2006) notes the inattention to South Asian women’s networks beyond the ties of kinship, specifically in writing about the elderly because the family is the predominant framework through which South Asian women are studied.

This article thus offers insights on the under-researched topic of cross-sex friendship and even rarer discussion of it beyond the White context and in relation to ageing. Like others I emphasise the need to debunk ethnocentrism in work on intimacy (Jamieson, 2011), to theorise on personal life in relation to cultural diversity and to incorporate transnational families when analysing social change (Smart and Shipman, 2004). As with existing research, most friendships in my study were same-sex. Here, I examine not only what the minority status of cross-sex friendships tells us about the normative tenor of gender, sexualities and intimate relationships in South Asian diasporic contexts but also the space for manoeuvre in creating such relationships. I ask, what makes cross-sex relationships possible, and what works against the ‘patterning’ and maintenance of such relationships?

The first section outlines key ideas in the sociology of friendship on which I build. The second section details methods. The third section discusses cultural assumptions that shape individual decisions around establishing cross-sex friendship. The fourth section explores factors that facilitate friendship formation and maintenance, and asks if there are qualitative differences between same-sex and cross-sex friendships. The final two sections focus on the relational practices involved in management of boundaries between sex/romance and friendship and public perceptions of cross-sex friendships.

Friendship, status divisions, and inequalities

Friendship is ‘the archetypal social relationship of choice’ (Pahl, 2000: 171), a voluntary relationship (Allan, 1989) without a formal institutional structure, either to keep friends together or to sanction duties and obligations (Allan, 1996; Pahl, 2000). Yet, friendship does not result solely from individual preferences, decisions, and commitment but has social significance beyond the ‘dyad alone’. Friendships are ‘patterned’ by wider contexts, defined as ‘conditions external to the development, maintenance, and dissolution of specific friendships’ (Adams and Allan, 1998: 2, 4). Thus, while doing friendship involves ‘interactional agency’ (Allan, 2008: 10), friendship formation results from a ‘complex interaction of factors’ (Spencer and Pahl, 2006: 180) and ‘sources of inequalities’ limit the degrees of freedom enjoyed by individuals (Jamieson et al., 2006). Moreover, although there may be ‘no clear-cut rules governing friendship, there are cultural scripts about [how] friend relationships should be structured’ (Allan, 1996: 99).

That friendships are not entirely a matter of individual volition but shaped by individuals’ structural locations is supported by the observation that most personal relationships are constrained by ‘social similarity’ (of class, race, ethnicity, education, age, religion, attitudes and beliefs) that creates patterns of ‘homophily’ – ‘similarity breeds connection’ (Bottero, 2005: 9; McPherson et al., 2001). All intimate relationships are tied to processes of hierarchy and differentiation because individuals tend to ‘confirm and reproduce the social order’ by forming close relationships with those who are socially similar (Bottero, 2005: 159). The study of friendship, then, highlights inequalities and stratification but also social change, since boundary crossing in friendship has the potential to counter and ‘subvert’ social divisions (Adams and Allan, 1998; Rawlins, 2009: 195).

Studies note shifts in personal life (Jamieson, 1998; Jamieson et al., 2006) with some arguing that wider transformations in late modernity have meant greater ‘flexibility’ in the construction of intimate ties. Allan (2008), for instance, sees the ‘growing acceptance’ of cross-sex friendships, especially among young adults as indicative of this greater ‘flexibility’. He writes, ‘while gender remains a significant status division . . . its power in shaping friendship eligibility is less marked than it was’ (p. 5). Blatterer (2016), however, argues that friendship’s freedoms remain ‘embedded’ in the contemporary gender order and ‘persisting inequalities’ both structure opportunities for friendship and continue to pervade intimate life more generally (pp. 63, 68; also Jamieson, 1998): cross-sex friendships are curtailed because no comparable transformation in cultural norms has occurred, despite structural opportunities for increased cross-sex interaction. Cronin (2015) moves beyond an analysis of the differential impact of gender on cross-sex friendship formation to unpack how gender and (hetero-) sexualities are made and remade through an interaction of friendship and other intimate (couple) practices. She concludes that personal relationships have the potential both to undo or subvert and to reproduce ‘dominant modes of intimacy’ (p. 1179).

In this article, I contribute to debates on intimate relationships and inequalities by examining how ethnicity ‘patterns’ cross-sex friendships and arguing that cultural scripts impinge on and inform ‘contexts’ within which friendships are navigated by determining norms for appropriate interaction that are gendered and remain effective even in later life. A gendered life-course approach assumes particular significance in South Asian contexts where roles and identities are linked to clearly defined life-stages (e.g. for women: marriage, motherhood, widowhood). Ageing is ideally associated with enhanced power, respect and status for men and women alike. For women, family developmental processes and social status transitions at mid-life (rather than chronological age) generally involve the acquisition of increased authority and autonomy (Gardner, 2002; Vatuk, 1992). Constructions of sexuality are also tied to life-course stages. The sexuality of pubescent girls and women of childbearing age is considered ‘dangerous’ and a potential source of ‘disorder’ that is domesticated and made ‘safe’ through marriage (Werbner, 2002: 260, 272–273). The asexuality of post-menopausal women assumed and deemed unthreatening (Hershman, 1977 in Vatuk, 1992: 139), although some question such cultural assumptions about ageing and desexualisation, arguing instead that concerns about women’s sexuality may remain even after they pass the reproductive phase (Vatuk, 1992).

The study and methods

This article draws on my wider research exploring how changes within South-Asian families can inform current theorisations of ageing, care, and relationality. I conducted semi-structured interviews with 40 participants aged between their 50s and 80s (11 Indian Sikh and 11 Pakistani Muslim women, and 8 Indian Sikh and 10 Pakistani Muslim men) from Edinburgh and Glasgow between December 2020 and September 2021. Twenty-seven were currently married (of whom three had previously divorced), seven widowed, and six divorced. Diversity in class status and living arrangements (e.g. living alone, couple household, and multi-generational household) also guided informant selection. They had varied migration trajectories, as post-war, marriage/family, educational or ‘twice’ migrants. Eighteen were UK/Europe-born or had arrived as children as part of family reunification. With one exception, all had either migrated from or had a family history of migration from Indian/Pakistani Punjab. All participants identified as heterosexual.

Covid-19 regulations involved heavy reliance upon others in informant selection and I used multiple channels: charities, mosques and gurudwaras, personal contacts and snowballing to minimise the influence of ‘gatekeepers’. Due to restrictions on face-to-face contact, I conducted interviews (largely) telephonically or on Zoom in English, Urdu/Hindi and Punjabi, as suited the participants. In research, ‘one’s self can’t be left behind’ (Stanley and Wise, 1993: 161), so the research process involved a reflexive engagement with how my positionality (class, gender, ethnicity, age, nationality) influenced both what I could know and my analysis. I am neither Sikh nor Muslim, but I shared a wider Punjabi and South Asian identity, and linguistic commonality with my participants. Nevertheless, I remained aware throughout of the ‘multiple, intersecting and shifting nature’ of both my own and my participants’ social locations and positionalities (Zubair and Victor, 2015: 964).

I developed a broad topic guide for the interviews, starting with contextual issues (age, employment status, migration history, household composition, etc.) before moving to more personal ones (opinions about ageing, health and support needs, quality of relationships, etc.). The interviews became progressively more narratively driven, disclosing the storied character of people’s lives and experiences (Kvale and Brinkmann, 2009) and their connections to wider social processes. Some accounts were much more elaborate than others: interviews ranged from 50 minutes to 7 hours, and were one-off with 11 participants, while the remainder entailed two or more conversations. I transcribed the digital recordings verbatim and thematically analysed the data following the approach of Spencer et al. (2014) that allowed for ‘iterative movement between the original data and the conceptualisation, abstraction and interpretation derived from them’ (p. 292). All participants are anonymised by using pseudonyms and omitting identifying details.

To understand participants’ relational lives, I asked, ‘What do you consider to be your most important relationships?’ This phrasing avoids asking directly about particular relationships and predetermining the contexts people should consider (Spencer and Pahl, 2006: 192). In response, several participants mentioned friends. If they did not, I asked if friendship was important to them. Responses included: ‘yes’, ‘very important’, ‘yes, absolutely’, ‘friends are there for life’ and ‘without friendship, life’s lonely’. Only one of 40 participants said he had no friends and 32 of the remaining 39 had at least one ‘close’, ‘true’, ‘best’, ‘genuine’, or ‘good’ friend, distinct from those described as ‘other’ or ‘just’ friends. Some also distinguished friends, colleagues, acquaintances, and others in their professional networks. Participants used the English term friend, although saheliya (used by women for female friends) and dost (male friend, although women too may describe their female friendships as dosti, a masculine identification) were sometimes mentioned. Their various articulations of what made a friend close revealed that they ‘valued particular attributes’ (Spencer and Pahl, 2006: 59) and included: deep knowing, empathy, understanding, loyalty, respect, honesty, openness, and acceptance. Several mentioned deep emotional attachments with close friends, of ‘being supported through trying times’, ‘talking about old times’, ‘sharing the ups and downs of life’, ‘sharing your soul’, and ‘crying together and finding deep joy together’. Trust, identified as characteristic of modern friendship (Pahl 2000), was especially crucial in defining someone as close. Some participants (mostly women) raised concerns that what was shared should ‘go no further’ because, in close-knit communities, their family’s reputation was at stake. As Allan (1989) noted, ‘friend is not just a categorical label, indicating the social position of each individual relative to the other but is . . . a relational term which signifies something about the quality and character of the relationship involved’ (p. 16). Consequently, I relied on my participants’ own assessments.

A few reflected on cultural understandings of friendship. Two UK-born participants, for instance, noted that women of their mothers’ generation did not use the vocabulary of friendship, a point I address later. Another participant regarded ‘Asian’ definitions of friendship as distinct from ‘Western/White’ ones because there are dedicated social settings in the latter (such as pubs, cafes, etc.) for meeting and defining friendships. For Asians, though, friendship was not limited in this way and could be ‘enjoyed’ at home, over a garden wall, at weddings/family events, at a spiritual centre, and so on (see Allan, 2008, on friendships in Britain becoming less bounded).

Nine participants reported having cross-sex friends in their ‘friendship repertoire’ (Spencer and Pahl, 2006: 54): people with whom they had regular contact, or regarded as ‘good’ or ‘close’ friends. More, however, mentioned some experience of ‘friendly’ cross-sex relationships. The following account focuses on cross-sex friendship, but also draws on my wider empirical data on friendship.

Cultural norms and cross-sex friendship

In British and American contexts, studies focused largely on the White population note ‘taboos’ on cross-sex friendships for married individuals (Allan, 1989), with some describing close cross-sex friends as ‘inappropriate’ post-marriage (Werking, 1997). Likewise, my participant Sabiha (female, Muslim, married) described speaking to other (unrelated) men as ‘taboo’, but attributed this to being Muslim. Zoya (female, Muslim, widow) believed friendships between men and women were possible but cultivating them was risky: ‘we all have our weak moments’, and she had to ‘protect’ herself because Islam permits sex only within marriage. Zoya’s view was consistent with some other participants’ opinion that cross-sex friendships are most acceptable when seeking marriage. This was not the only view, however. While discussing appropriate man–woman relationships, Zubair (male, Muslim, married) recalled a ‘severe beating’ from his father because someone had reported seeing him talking to a girl in the street:

That particular incident took place because my parents didn’t sit us down and say, this is permissible, this is not . . . My father’s fears were that we were young, that we had come into a new environment from Pakistan, these were the early years. They never said you don’t have female friends or White friends, what they said was: we don’t consume alcohol or have sex outside marriage. Although the reaction was brutal, the lesson that was delivered was that: yes, you can have friends but be conscious of this that unless you make a commitment to marry, you cannot engage with them in an intimate way. It didn’t stop me, and I brought female friends home to meet my parents and the confidence grew in them that boys can have friends as long as they know the limits they can’t cross and they were keen to know what friends we had.

With my own daughters, I have not said to them that you cannot have male friends and they do have male friends . . . I can contrast and compare our family to my aunts’ and uncles’ families and distant cousins and there may be subtle differences in sticking to the spiritual line. There are those who are carefree and there are those who are extreme at the other end and who have attempted being rigid and they failed in that so I think that families are learning through their experiences around them . . .

For Zubair, cross-sex friendships could exist ‘within limits’, while Humera (female, Muslim, married) explained: ‘The religion does say that a man and woman should not be alone together’ but friendships with men were ‘taboo’ only in ‘certain [conservative] circles’:

My definition of religion is very different from the majority of Pakistanis. I practice. I do follow the faith. I pray as often as I can and keep the fasts. I do follow the pillars of Islam but my faith is a very liberal faith.

Thus, although they navigated cross-sex friendships within predefined faith-based scripts, each participant had assessed what was acceptable. Zubair also demonstrates that transnational families are not immune to wider social changes and that cultural norms may be adapted and evolve over time. Furthermore, even within ethnic groups, differences within families’ understanding of and deployment of cultural practices exist.

The Sikh participants also had diverse views. Unlike Muslim interviewees they did not explicitly mention culture or faith, but they shared a South Asian cultural ethos that disapproved of sex outside marriage. Preet (female, Sikh, married) reacted with ‘tauba tauba’ [heaven forbid] to my question about man–woman friendships, adding that she herself saw nothing wrong with such friendships, but any relationship with a man other than one’s husband was improper in her community. Any male-female interaction had been forbidden when she was a young married woman, and she operated in largely gender-segregated contexts. Her daughters and daughter-in-law, however, were employed in and socialised in gender-mixed environments yet one-on-one cross-sex friendships remained unacceptable for them too. Preet belonged to the Bhatra community that is more conservative and closely interlinked than other Sikhs. While her case indicates generational change with regard to women’s employment and visibility in gender-mixed spaces, community expectations seemed weighty. For Kuljeet (male, Sikh, married), friendship with a woman implied sexual interest, and, speaking generally about friends Ranbir (male, Sikh, married) said:

In a sense, everybody is my friend, but there are a couple of very close friends. I’ve got other people at my work, people I volunteer with etc. If somebody will say to me [sic], ‘Who is this guy?’, I will say he’s one of my friends. So, it’s just the terminology you use, friend, but that doesn’t mean they’re close.

Shruti:

Are these close friends all male?

Ranbir:

Yes, but I do have others colleagues, etc. who are women, but not close.

Shruti:

Any particular reason why?

Ranbir:

Don’t know, just don’t have them.

What is permissible is undoubtedly socially and culturally scripted, then, but there is space for individual volition in meaning-making. People’s choices, attitudes and individualised practices cannot simply be reduced to ‘Asian-ness’, ‘Muslim-ness’, and ‘Sikh-ness’ or ‘culture’, for these are neither homogeneous nor fixed. Cultural beliefs, identities, and practices necessarily embody the structural forces that affect people’s lives and culture itself is ‘dynamic rather than static’ (Phoenix, 1988: 152) with culture making being ‘a project of social continuity placed within, and contending with, moments of social change’ (Baumann, 1996: 31).

The structuring of cross-sex friendship and the content of relationships

Friendship formation requires individuals to come into contact with each other in ‘spaces of friendship’, including material and visual spaces and places (Bowlby, 2011: 613; Fehr, 1996). My participants formed cross-sex friendships in educational, work, and leisure contexts, the ‘routine environments’ that have enabled more gender-integrated activities than those available to previous generations (Allan, 2008: 5). Some of these friendships were also cross-ethnic. Cross-sex friends are most prevalent in young adulthood (Monsour, 2002; Rawlins, 2009), with married status a ‘barrier’ due to potential spousal suspicion and jealousy (Allan, 1989; O’Meara, 1989). In her study of cross-sex friendship, Werking (1997) reported that only 19 among 100 White, middle-class, heterosexual American adults between 21 and 46 years were married, while Booth and Hess (1974) found cross-sex friendships of married individuals generally entailed less contact and affect than among the unmarried. Of my participants who currently had cross-sex friends, only one said he had forged new friendships post-marriage. The remainder were formed when single (unmarried or divorced). Zubair (male, Muslim, married) believed that marriage changed friendships, cross and same-sex alike, because time and opportunity became limited and responsibilities from other relationships increased, not because of spousal jealousy or a deliberate prioritisation of the couple relationship relative to friendship (as in the White context: see Cronin, 2015). A life-course approach that conceptualises ageing in relational terms, as ‘linked lives’, is important here (Bengston et al., 2012): Zubair drifted away from school and university friends as he formed a family, yet could forge new cross-sex friendships through his wife. Similarly, Zoya (female, Muslim, widow) lost but later reconnected with an old school friend from ‘back home’ [country X not Pakistan]. As both Zoya and Zubair advanced in the life-course and children grew to adulthood, they had more ‘space’ in their lives (Allan, 1989: 153) for friendship. Zoya explained:

Zoya:

I had my friends from my childhood but after I came to the UK, I got so busy I was not able to keep up with them.

Shruti:

So, you recently reconnected with a friend?

Zoya:

Oh my God! He has been trying to locate me for many years. Honestly, I never, never checked messages on my Facebook . . . until last year . . . He was my friend from school . . . Also our community, Pakistani. But at that time, he used to be scared of my brothers because he was from a lower social class, so he kept a distance. His sister, she’s now a colleague of my brother’s . . . I think he had seen photos of me and the kids on Facebook. But there was no other, so he realised I was single.

Shruti:

Is he single as well?

Zoya:

Yeah, he’s single.

Shruti:

If you were married, do you think he would have reached out to you?

Zoya:

I don’t think he would try.

Shruti:

Do you keep in touch with him?

Zoya:

We talk quite often, yeah.

Zoya wanted to remarry and was open to this becoming a romantic relationship: ‘If you are thinking of finding someone, Islam doesn’t stop you. You can at any time, any age, you can get married’, she added. Her case highlights the intersection of life-course stage, marital status and also social class: class homophily, which Zoya and her friend lacked before, becomes an important consideration in marriage and friendship (Bottero, 2005; McPherson et al., 2001). Zoya’s account also indicates how social media facilitate continuity in transnational friendships, including same-sex friendships, with email, WhatsApp and Facebook replacing letter and phone calls as ways of doing friendship over distance.

For Sohail (male, Muslim, divorced) same sex and (non-romantic) cross-sex friendships were qualitatively different (cf. Monsour, 2002; Werking, 1997) and he regarded cross-sex friendships as beneficial because they provided a female perspective (e.g. for dealing with a teenage daughter). Humera (female, Muslim, married) considered gender unimportant and valued qualities in friendship: someone she could phone when she needed to talk, male and female friends who shared her interests and with whom she felt comfortable to talk about ‘everything and anything’ (even about her hysterectomy, for example). Likewise, Nimrat (female, Sikh, divorced) asserted there was no difference between male and female friendships in terms of shared activities: go for a meal, invite them to her house, or sit in the garden, go for a coffee or a walk. Unlike Humera, though, she would not speak to male friends about ‘certain things’ – like menopause but if she needed something, she would treat everybody ‘equally’. Yet, after an operation, it was her female not male friends who provided practical support (getting dressed, dropping food, shopping off, etc.). This was also so for the male Sikh participant I discuss below (who received practical care from a female friend following hospital admission) and in some same-sex female friendships where friends had provided such support during episodes of illness. This agrees with the observation in the wider literature on the gendered expectation that the work of care is women’s work (Morgan, 1996).

Sexual ambiguity in cross-sex friendship

Some participants suggested that certain types of friendships were best avoided. Zoya (female, Muslim, widow) spoke of being particularly cautious with married men:

There is attraction to the opposite sex . . . There is a saying: Ghar ki murgi dal barabar? [we don’t value what we have, but think that what we don’t have is better], do you know it? The people I talk to, of course they have families. I am very, very careful because I can be very attractive. They have their wives which may seem like nothing because they see them every day . . . maybe something can attract them to me and then I will be responsible. I wouldn’t want to wreck another person’s life . . . isn’t it also about your own respect?

In a different conversation, she said:

Earlier, I was without a hijab [headscarf]. I used to do a lot of make-up. I was a very outgoing person and I was quite a fashion icon but after my husband passed away, I could see how men looked at me. I had a few experiences and I decided to put on the hijab because it is a symbol that I am not open for these things, that I am a Muslim, it means stay away from me in that respect.

Nimrat (female, Sikh, divorced), like Zoya, described being approached by married men:

I just ignored them and got on with it . . . And the next time they wouldn’t . . . I think I was judged by other people [in the community] . . . Not now, no. I think everybody knows who I am now . . . After my divorce, I had a lot of hassle from my mum, ‘I want a good girl’ . . . Now, I’m not really classed as, ‘That single woman’ . . . I’m just like everybody’s sister, or everybody’s auntie.

A (married) Muslim participant (hereafter MP) spoke about her cross-sex friends, including an ‘older Indian gentleman’ (in his 70s) whom she met over two decades ago. Both were divorced at the time. She subsequently remarried. She also had friendships with a few White men (some single, some partnered) through volunteering. As a divorced woman, she added, she avoided befriending Pakistani men because they were likely to interpret the friendship as an invitation to sexual intimacy and marriage and her parents, especially her mother, had concerns about the cross-sex friendships she established while divorced:

It depended on who I was being friends with. My dad was okay about me being friends with White guys than with Asians. He probably thought that I wouldn’t have a relationship with a White guy. He’s still a bit conservative. He wanted me to settle down with a good Pakistani boy.

MP decided to forgo friendships with Pakistani men: they could have affected her marriage prospects if they were construed as potentially romantic. In her first marriage, she had exercised ‘choice’ but there were limits that it would not cross (religious) community boundaries, an expectation that also held for any future marriage. So did she have concerns about her Indian or White male friends?

I’m not sure exactly but I get the feeling with men from my own community that they will naturally expect that you are being friendly with them because you are interested in them, on a sexual level. I understand that is a prejudice of mine. Of course men from other communities may think along sexual lines as well, but I guess they will not think of marriage.

This suggests that there is an understanding between cross-sex friends that sexual desire and romantic interest would be shared with somebody else and becomes possible where cultural and religious differences are taken to mean a marriage would be impossible and hence sexuality kept at bay. In her relationships with White and Indian male friends, the boundaries between platonic and romantic relationships seemed clear-cut for MP. Yet the following example demonstrates that these boundaries in relationships with ethnic outsiders were not always as clear-cut as she had suggested. MP mentioned another friendship she formed while divorced: this (White) male became ‘more than a friend’. The relationship ‘was not going to go anywhere’ but she was still ‘friendly’ with him, texting him sometimes, even though things were now different. I asked how her husband felt about her male friends:

MP:

My husband has never said that he is not fine with them. I have invited people to the house . . . With my Indian friend, he doesn’t seem to have an issue because we never had a sexual relationship, but with my White friend, that I had a relationship with, I have not told him that we are still in touch.

Shruti:

Do you think that he wouldn’t be okay with it?

MP:

I don’t really know. I mean, nothing is going to happen. I am probably not ever going to see him again. If he wants to send me a text message now and again, I don’t think there is any harm in that . . . I just think that it is part of my past. I don’t see why I should have to explain myself . . . My past has been quite colourful for an Asian Muslim woman. Within the Asian community you would expect to keep these kinds of things hidden because the spouse could become jealous and upset and to keep the family bond, you don’t want to tell everything and anything . . . In my present, I don’t feel as though I am doing anything wrong . . .

MP specifically notes how men from her own community were likely to see cross-sex friendships in a sexual way (cf. Halatsis and Christakis, 2009; Rawlins, 2009), and she, Zoya and Nimrat all indicate how single women are particularly responsible for preventing relationships becoming sexualised. MP decided to forgo friendships with Pakistani men to avoid being seen as available, while Zoya and Nimrat both suggest that men could be predatory towards lone women. In South Asian contexts, men (fathers, brothers, and husbands) are regarded as the guardians of women’s sexuality. Writing about Sikh patriarchies, Guru (2009) explains,

A respectable woman is one who avoids eye contact with unrelated men, is modest in her attire and does not flaunt her ‘dangerous’ sexuality. Her honour, prestige and reputation (izzat and sharam) are closely tied to the representation of her body and to the male protection that she finds. Women without the defense of men can be left extremely vulnerable and easy sexual targets. (p. 293)

Similarly, Qureshi (2016: 217–222) describes how divorced women become subject to sexual predation in the Pakistani Muslim context. Indeed, gendered notions of honour (izzat) and shame (sharam) continue to resonate across both Sikh and Muslim contexts in my study. For Zoya, being respectable meant not ‘wrecking another person’s life’, an acceptance of sexual standards in which women ‘outside marriage’ embody a disruptive potential that they must themselves police. Nimrat and MP both reveal expectations about morally acceptable behaviour as a daughter (be ‘good’) and wife (‘I’m not doing anything wrong’), and indicate that single women remain under the ‘community’s’ watchful eye. For Nimrat, conformity to acceptable feminine behaviour had gradually rendered her ‘asexual’ (making her a ‘sister’ or ‘auntie’) and she no longer had to worry about being viewed as sexually available. Zoya, however, actively works to challenge this by drawing on her religious repertoire and desexualising herself by donning the hijab to ward off sexual advances: doing ‘appropriate’ femininity, but also showing ‘the agency of the hijab’ and the significance of its visibility (Tarlo, 2007: 131). The hijab makes a woman visible as Muslim, an identity that Zoya asserts, and becomes a ‘powerful constraining moral force’ and a ‘barrier’ by governing who can interact with whom and the nature of these interactions. The hijab protects women from the male gaze and provides a sense of respect (pp. 140–142, 150–151). Cross-sex relationships, then, are ‘patterned’ by similar gender norms for Sikhs and Muslims, although these find explicit religious expression only among Muslims.

Furthermore, we can see how sexual double standards are reproduced in how men and women do intimacy, both in MP’s account and in that of a divorced Sikh male participant who mentioned a friendship where the boundaries between friendship and sex were blurred. During adolescence and early adulthood, his cross-sex relationships had all been romantic or involved ‘casual sex’. He mentioned Karen (White Scottish, divorced) as a friend who had nursed him for two days during an illness, and whom he later described as a ‘previous girlfriend’. When I asked him about her, he said, ‘Let’s just say, friends with benefits’. This example shows that sexual ambiguities in cross-sex friendships may allow men to capitalise upon sexual freedoms and opportunities not granted to women.

Thus, MP speaks of needing to keep her unusually ‘colourful’ past ‘hidden’ to avoid jeopardising her marriage and facing the resulting disapproval of the community at large. By contrast, the Sikh male participant said, ‘For me, for Asian boys, we’re let loose, we can do what we want more or less’. His ‘friends with benefits’ appear as continuations of the sexual freedom he enjoyed as a young single man and highlight men’s taken-for-granted and condoned sexuality and promiscuity (an observation also made about young British South Asian Muslims: Ali et al., 2020: 8).

Managing cross-sex friendship

Cross-sex friendships face an ‘audience challenge’: to present an ‘authentic friendship’ to the ‘relevant audiences’ to promote its acceptance and avoid rumours and misrepresentations (O’Meara, 1989: 537). Outsiders attempt to influence and control cross-sex friendships far more than same-sex ties (Allan, 1989: 83) and cross-sex friends must negotiate the public image of their relationship (Werking, 1997: 120).

The married and cross-sex friendship

Married people may take ‘precautions’, such as limiting cross-sex friends to those met through a partner, or socialising along with their spouse (Fehr, 1996: 149), for a partner’s presence renders such interactions ‘asexual’ (Jamieson, 1998: 104). Some of my participants spoke of regularly socialising with their spouses’ friends and their partners, although their descriptions suggest friendliness rather than friendship. Some, however, did establish close, lasting friendships in the process, almost always forged between two husbands and their wives and not crossing gender boundaries (cf. Allan, 1989: 81). Nonetheless, some exceptions emerged. Humera (female, Muslim, married) said her parents had several such couple friendships and she mentioned one of her father’s friends whose wife was in Pakistan and who regularly visited Humera’s home. Her mother cooked for him and they would chat. Her mother considered him a ‘friend’ and they referred to each other as bhai-behen [brother-sister]. He withdrew, however, once his wife joined him in Scotland because she was uncomfortable with this relationship. Humera added that those of her mother’s generation did not use the term friend, especially in the context of man–woman relationships.

In my study, some participants also described a close (same-sex) friend as being like a ‘brother’ or ‘sister’ demonstrating ‘suffusion’ – ‘a fluid interchange of friend-like and family-like relationships’ that denoted closeness (Spencer and Pahl, 2006: 54). With Humera’s mother, it served to desexualise the relationship and made interaction between the sexes safe and respectable (although in her case, it had failed to ease spousal suspicion). The kinship idiom of relationality is ubiquitous in these contexts (although not unique to South Asians). It works not only to desexualise the tie but also adds an intensity to the friendship by transforming it into one on which people can rely (e.g. in times of need, such as a wedding or funeral, a close friend can assume the responsibilities of a family member).

Cross-sex friendships were not always contentious, however. MP said her husband had no problems with her male friends with whom she had never been sexually involved and she invited them home to meet her husband. Zubair (male, Muslim, married) spoke about friendships with his wife’s colleagues with whom he shared interests in politics, social issues, and the environment:

I have more of a banter and exchange with them on a regular basis [on WhatsApp] than she does: that’s just her personality. I have more female friends than my wife has male friends . . . I am a much more sociable person. I have never spelled out to her what the scope of her friendships should be. She just doesn’t want deep friendships, she has siblings and cousins. We have been married for decades and there has never been an issue between us.

Nevertheless, Zubair mentioned ‘certain cautionary behaviour’ with female (but not male) friends that inhibited ‘deep’ friendships with women. For him, WhatsApp allowed ‘distanced or remote friendships’ with female friends (see Ali et al., 2020: 9–11 on the similar use of social media in dating by young British Muslims). Zubair elaborated,

It isn’t limited by any notion of ring fencing or securing myself or the person whom I regard as a friend. We understand and accept that it is in our spiritual teaching that friendship has a definite scope, a line of friendship exchange and we do not move out of that boundary into an area where the relationship would be defined as ‘more than just friends’. We are alert that there is a risk. I would not put myself in that location, social event, where friendship is going to be at a risk to develop into something else so I have met females at bar settings but I know that there is a certain point at which I need to take my leave.

Zubair reflects widely held assumptions that homosocial relations engender no tensions around sex. He highlights the significance of personality characteristics and preferences in friendship formation and indicates the difficulties in fully escaping the heterosexual script. Yet he does not mention the gendered privilege of relative freedom from scrutiny that enables him to meet friends in bars, for instance.

Single status and cross-sex friendship

Non-marital familial relationships became significant for divorced and widowed women, especially natal kin ties that were enduring sources of support and care (see also Qureshi, 2016). Nimrat (female, Sikh, divorced) described a ‘couple of men’, whom she met through work as ‘best friends’. Of them, one was White (single). Keeping these friendships platonic was not a problem, yet the friendship with her also single White friend required careful navigation:

There have been . . . At first, I was a bit . . . Until everybody got to know my friend, I didn’t want to be driving about with him or anything like that, not being seen with him.

Concerns about propriety loomed large. Nimrat said her behaviour had implications for maintaining the respect of her brothers and also her children and grandchildren. After being widowed, Tavleen (female, Sikh, widow) worried about ‘the community’, the wider Asian as well as Sikh community, and rumours about being seen talking to a man in public, for instance. For some time, she avoided social events and later her granddaughter or sons accompanied her:

I suppose in some ways I kind of covered my back even though I am of a certain age, but there are things that sometimes you just can’t control.

With age women gained respect, Tavleen explained, yet (older) age does not offset gendered expectations of ‘appropriate’ behaviour. My participants neither fixed nor clearly defined the boundaries of community, but they conveyed the ever-present threat of gossip because people tended to know each other or were somehow interconnected. Gossip is well-recognised as a mechanism of social control with the potential to enhance or diminish social prestige and reputation (Rajačić et al., 2020). Here gossip could tarnish the izzat not only of women but also their families. Men were also concerned about gossip. Sohail (male, Muslim, divorced), set boundaries in his friendship with an (Asian) woman. It had grown while spending time organising work events but he did not want his colleagues to think they were romantically involved so he interacted openly with her at work and in other public settings, but discussed personal matters only on the phone.

Fear of gossip, then, meant self-policing: Nimrat avoided being in public alone with her male friend, Tavleen ensured she was chaperoned, Sohail concealed his private one-on-one phone conversations. Men and women, however, differed. Sohail emphasised his self-perception as a ‘good Muslim’ who abided by his ‘moral compass’: ‘not drinking or extra-marital sexual relationships, or diet of pork and stuff’. Similarly, Zubair limited his interactions with women in line with ‘spiritual teachings’. As for Sikh men, while the participant, who speaks about ‘friends with benefits’, inhabits a very different space from others included here, in Ranbir’s account (discussed earlier on) we see boundary work when he describes being friendly but not ‘close’ with women, without reflecting on why this is so. Women, Sikh, and Muslim alike, were predominantly concerned with respectability and reputation.

Conclusion

Returning to how ethnicity ‘patterns’ friendships, my data show that wider South Asian cultural norms (taken-for-granted heterosexuality, the prescription that sexual relations only occur within marriage) inhibit the ‘patterning’ and maintenance of cross-sex friendships, despite increased opportunities for cross-sex interaction. My participants with cross-sex friends struggled with the dynamics of these friendships. For married participants, socialising in a spouse’s presence, using social media or making friendships kin-like enabled the maintenance of cross-sex friendships without threatening the marital tie. For the unattached, cross-sex friendships sometimes became a route to romance, sometimes the boundaries between sex and friendship blurred, and sometimes people took care to maintain the friendships as non-romantic. Emotional support, exchanging confidences, and sharing leisure time generated intimacy in cross-sex friendships and same-sex ones alike. The former, though, required the regulation of romance and sexual engagement to maintain the friendship, a management not only in relation to significant others but also community expectations. Doing friendship meant simultaneously doing gender because cross-sex friendships involved negotiating predefined gendered scripts that overlapped with ethnic/cultural scripts, and women faced greater risks. Similar gender patterns prevailed in the Indian Sikh and Pakistani Muslim contexts, although only Muslims explicitly verbalised how the requirements of faith shaped the navigation of their relationships.

In his work in the American context, Rawlins (2009) optimistically wrote of the ‘subversive potential of cross-sex friendships’ for challenging ‘the enforcement of heteronormative and gendered scripts for male and female relationships’ (p. 195). That cross-sex friendships exist in these diasporic communities and are not numerically insignificant (almost a quarter of participants had cross-sex friends) signals both social change and agency. My data show that within the changing context of intimacies in the South Asian diaspora, with people not necessarily staying in lifelong marriages (Guru 2009; Qureshi 2016) and seeking new relationships in mid- and later life, cross-sex friendships become possible, notwithstanding concerns that they may become sexual. Despite changes in what is possible in their intimate lives, women in particular must contend with community policing of their sexuality, even as they age and acquire more status and respect(ability) within their families and communities. (Older) age does not dispel anxieties around women’s (potentially transgressive) sexuality (Vatuk, 1992). Studies in other cultural contexts, likewise, suggest that gender scripts are not necessarily attenuated in older age. For instance, in her work with White British single women in mid-life attending salsa classes, Milton (2017) found that the classes allowed women to embody a ‘sensual’, ‘independent’, and ‘socially active femininity’ that challenges ageist assumptions. Yet they simultaneously produced the performance of a ‘policed’ and ‘age appropriate’ femininity (p. 151, 155), and women’s accounts also articulated concerns about respectability that were related to age. For my female participants, kin and community contexts were dominant with gender, age, and ethnicity intertwining in powerful ways to ‘pattern’ friendships despite evidence of creativity in making and sustaining relationships. Jamieson et al.’s (2006) observation is pertinent here: ‘Opportunities for intimate practices may be more numerous and more accessible than they once were historically but this does not necessarily mean that they are evenly distributed in the present’ (p. 5).