Later Life Sex and Rubin’s ‘Charmed Circle’

Gayle Rubin’s now classic concept of the ‘charmed circle’ has been much used by scholars of sexuality to discuss the ways in which some types of sex are privileged over others. In this paper, I apply the concept of the charmed circle to a new topic—later life—in order both to add to theory about later life sex and to add an older-age lens to thinking about sex hierarchies. Traditional discursive resources around older people’s sexual activities, which treat older people’s sex as inherently beyond the charmed circle, now coexist with new imperatives for older people to remain sexually active as part of a wider project of ‘successful’ or ‘active’ ageing. Drawing on the now-substantial academic literature about later life sex, I discuss some of the ways in which redrawing the charmed circle to include some older people’s sex may paradoxically entail the use of technologies beyond the charmed circle of ‘good, normal, natural, blessed’ sex. Sex in later life also generates some noteworthy inversions in which types of sex are privileged and which treated as less desirable, in relation to marriage and procreation. Ageing may, furthermore, make available new possibilities to redefine what constitutes ‘good’ sex and to refuse compulsory sexuality altogether, without encountering stigma.


Introduction
Gayle Rubin's charmed circle diagram characterises a hierarchy of types of sex, whereby some sex is treated as 'good, normal, natural, blessed' and other sex is treated as 'bad, abnormal, unnatural, damned' (Rubin 1984). The types of sex argued by Rubin to be within the charmed circle are; heterosexual, married, monogamous, procreative, non-commercial, in pairs, in a relationship, same generation, in private, no pornography, bodies only, vanilla. The corresponding types of sex in the 'outer limits' are: homosexual, unmarried, promiscuous, non-procreative, commercial, alone or in groups, casual, cross-generational, in public, pornography, with manufactured objects, sadomasochistic. The diagram, and the wider discussion of the sex hierarchy in which it occurs, were first presented at the legendary Barnard College feminist 'sex wars' conference (Hemmings 2005) in 1982 and then first published in the collection Pleasure and Danger (Vance 1984). In an interview with Judith Butler, Rubin discusses how the paper came partly from a desire to reassert the significance and centrality of sexual activity to the study of sexuality, at a time when gender was more often the privileged lens (Butler 1994). Rubin's paper has since been reprinted numerous times, arguably acquiring canonical status (Love 2011) by its inclusion as the first chapter in The Lesbian and Gay Studies Reader (Abelove et al. 1993). It remains a staple of the undergraduate and post-graduate curriculum for courses focusing on sexuality or sexual identity and has been argued to be especially valuable as a device to enable critical thinking about sexuality (Epstein 2011). It is credited with laying some of the groundwork for Queer Theory (Brickell 2009) through making visible the disciplinary functions of heteronormativity (Chan and Howard 2018). Many commentators have responded to and developed Rubin's work, most notably the collection 'Rethinking Sex' published as a special issue of the journal GLQ in 2011, in which Rubin herself revisited her original paper and described the historical and political context in which it was written (Rubin 2011).
Rubin's paper has been criticised for paying insufficient attention to intersectionality and for assuming a white Western subject (Hoad 2011;Holland 2011;Schueller 2005). It has also been criticised for privileging the regulation of sex in relation to types of sexual activity over the regulation of sex through gender, race, ethnicity, class, national origin, culture, religion and other axes of difference (Ho 2006). Other scholars have focused less on the specifics of how Rubin characterises the sex hierarchy and more on the notion that sex hierarchies exist and serve to regulate sexual activity. With this approach, Rubin's own delineation of which types of sex lie within the inner and outer circles of the charmed circle, while a very important starting point for discussion, is not its greatest benefit. Rather, the notion of a charmed circle enables us to examine how distinctions and boundaries between good and bad sex are negotiated and policed differently for differently sexually and socially positioned individuals and groups. What is included within the charmed circle and what within the outer limits can thus be theorised to be highly locally variable, depending on social, geographical and historical setting, immediate rhetorical purpose and the people involved in the interaction.
Since 1982, the scope for treating a wide range of sexual acts as within the charmed circle has vastly increased, as seen, for example, in changes in public opinion around the acceptability of homosexuality (Watt and Elliot 2017), increases in claims of non-heterosexual identities (Office for National Statistics 2017) and the acceptability of mild forms of BDSM . In this vein, Mullholland (2011) draws on the notion of the charmed circle to explore how it is redrawn to include 'respectable raunch' at a 'SEXPO' retail event, Hoad (2011) explores the historic drawing and redrawing of sex hierarchies in South Africa, and Barker uses the charmed circle as a tool to rethink 'the rules' about sex (Barker 2018) This is the 1 3 approach I take in this paper, focusing on the ways in which the charmed circle is drawn and redrawn in relation to sex in later life.
Despite this long history of responding to and developing the concept of the charmed circle, discussion in relation to later life and older people is lacking. Rubin herself does not mention later life or older people in the original article 1 and later commentators have focused on how the charmed circle plays out for other categories of person e.g. disabled people (McRuer 2011) and South African citizens (Hoad 2011), but not for older people. There had, in fact, already been a flurry of research and comment about older people's sexual activity by the time Rubin was writing, including two influential large-scale empirical studies which both aimed to demonstrate the extent of older people's continuing sexual activity (Starr and Bakur 1981;Verwoedt et al. 1969). An annotated bibliography listing more than a thousand works on older people and sex had already been published (Wharton 1981), as had Sontag's famous commentary on the double standard by which women are seen as sexually undesirable decades before men (Sontag 1978). Since then, the literature on later life sex has become substantial (see DeLamater 2012; Gewirtz-Meydan et al. 2018; Hinchliff and Gott 2016 for recent overviews) but remains relatively undertheorised, with the majority of empirical studies identified as focused on sexual dysfunction and biomedical solutions (Tetley et al. 2018;Tiefer 2007).
Unlike the vast majority of this literature, in this paper I do not aim to generalise about older people's sexual acts or attitudes. Rather, my focus is on mapping out some of the patterns in the discursive resources (Taylor 2006;Willig 2008) commonly invoked for thinking and talking about sex in later life. I draw particularly on Billig's notion that talk and texts are often inherently persuasive and argumentative, designed to persuade real or imagined audiences of their power and correctness (Billig 1996). I draw also on positioning theory van Langenhove and Harré 1999) to explore some of the ways in which invoking a particular position brings with it moral framings and related storylines, and closes off alternatives. I use the findings of UK, North American and Australasian 2 empirical studies of later life sex as my main evidence of these discourses, supplemented by some discussion of two self-help books about sex for older people. 1 'Same generation' and 'cross-generation' in the charmed circle diagram refers to sex between adults and children, not sex between adults of different generations. 2 I focus my discussion on UK, North American and Australasian literature because this literature can legitimately be considered to be a (focal) discourse community (Swales 2016), as evidenced by such features as mutual citation and shared theoretical and political concerns. This makes the task of mapping patterns and rhetorical devices both legitimate and feasible. While much of my discussion could perhaps be extended to other Western locations, I limit my focus in this way to avoid homogenising e.g. Scandinavian or South European sexual discourses into Anglophone ones.

3
Later Life Sex and Rubin's 'Charmed Circle'

Older People's Sex: Still Outside the Charmed Circle?
Researchers focusing on later life sex often claim to be breaking a taboo or refuting a myth that people lose interest in sex as they grow older (e.g. Appel Doll 2012;Duffy 1998;Granville et al. 2018;Kessel 2001). They often cite as evidence phenomena such as older people's sexuality being treated as a source of ridicule (Butler and Lewis 1988), lack of positive media representation (Garrett 2014) and the particular anxieties around older men's sexuality which is often discursively associated with paedophilia and 'dirty old men' (Hughes 2011). Older people with dementia, who are experiencing significant ill-health or disability, or who live in residential care homes are claimed to be particularly likely to be excluded from sexual citizenship (Deacon et al. 1995;Mahieu et al. 2017;Simpson et al. 2017). Exclusion, anxiety and ridicule would seem to indicate that older people's sex remains outside the charmed circle.
However, the very proliferation of this literature claiming to break taboos and refute myths about older people's sexuality suggests that something more complex is going on: can older people's sex really be within the outer limits when so much is written arguing that it should not be? Beyond academia, self-help books, websites and media articles attempting to normalise later life sex also proliferate; for example, in the UK, the National Health Service and the largest older people's charity, Age UK, have webpages and resources devoted to the topic. Some researchers have instead argued that there are now two main competing discourses that are commonly invoked around older people's sexuality-a traditional one focusing on asexuality and a new one focusing on continuing sexual activity, variously characterised as 'the sexy oldie' (Gott 2005), 'sexy seniors' (Marshall 2010) or a 'liberal' storyline (Jones 2002).
Critical gerontologists argue that there are new and continually increasing pressures on older people to remain sexually active as part of a wider project of 'successful', 'active' or 'positive' ageing (Calasanti and King 2005;Gott 2006;Katz and Marshall 2003). Neo-liberal political regimes, which increasingly privatise and voluntarise the provision of social care services (Minkler and Holstein 2008), create new responsibilities for ageing citizens to manage their own health and display their wellbeing (Katz 2000) and indeed to age without ageing (Katz 2001(Katz -2002. Continuing sexual activity thus becomes a moral imperative in order to demonstrate one's ongoing good citizenship. These scholars also argue that continuing sexual activity in later life serves the agendas of late consumer capitalism, opening up new markets for consumption and continuing economic activity (Katz 2001(Katz -2002. For example, self-presentation and continuing pride in appearance and clothing are important aspects of the project of remaining sexually visible but, as Twigg (2007) and Ward and Holland (2011) note in relation to clothing and hair styles respectively, wealth makes a significant difference to older people's ability to maintain these visual signifiers. Feminist critical gerontologists also note the new interest in older people's sex among pharmaceutical companies which, they argue, reinforce phallocentric, heterosexist definitions of sexual activity (Cacchioni 2015;Marshall 2002Marshall , 2010Tiefer 2007). This body of work draws our attention to the ways in which the charmed circle of sexuality is being actively redrawn to include some older people and some types of sex-chiefly heterosexual, penis-in-vagina sex between older people in the 'Third Age' (Laslett 1989). The very existence of a literature critiquing this redrawing suggests its power and prevalence-the redrawing is prevalent enough to have attracted critique.

Redrawing the Charmed Circle
This redrawing of the charmed circle can also be seen in cultural artefacts which themselves both produce and replicate inner and outer circles. Self-help sex advice books have been argued to constitute a particularly important site for the cultural mediation of sex (Barker et al. 2017;Gupta and Cacchioni 2013;Potts 1998) and in this section I briefly discuss two such books targeted at older people, one dating from the 1980s and one more recently, 3 in order to characterise some of the ways in which the charmed circle has been redrawn in recent decades in relation to later life sex.
Living, Loving and Ageing (Greengross and Greengross 1989) was published by a large UK non-governmental organisation, Age Concern England (now part of Age UK) and was one of the first sex advice self-help books for older people to be published in the UK, following on from the similar US-published Love and Sex After Sixty (Butler and Lewis 1988). Living, Loving and Ageing undertakes significant discursive work to establish the terrain of later life sex as legitimate, not least in the framing of sexuality in the wider non-sexual context of 'living, loving and ageing'. The preface imagines a reader who might be shocked or offended by the contents and tries to ward this off by warning 'Many people will not be used to the frank and explicit way in which issues about sexuality are approached. This is in no way intended to cause offence' (p. 9). The preface also imagines a reader who is a carer uncertain whether to support an older person with sexual and emotional needs or whether to 'discourage something that is difficult to accept and sometimes to understand' (p. 8). The authors claim "[i]n this book we take no particular standpoint, but draw attention to needs which some people experience and pinpoint various choices which are open to them" (ibid). In positioning themselves so firmly as neutral in relation to a topic that they treat as contentious, the authors in this preface re-inscribes the 'outer limits' nature of later life sex, despite the wider project of the book apparently being the legitimisation of later life sexual activity.
The cover image of Living, Loving and Ageing shows a white man and woman gazing into each other's eyes, the woman's wedding and engagement rings clearly visible. They look to be in their 60s or early 70s. This cover image is indicative of the contents, insofar as they emphasise sexuality within the context of a relationship, usually marriage, and seem to presume heterosexuality, cisgender and whiteness, with the exception of one brief section about sexual identity, containing two short case studies about older lesbians and gay men (pp. 36-40), and a very brief mention of racism combining with ageism (p. 15). There is a heavy emphasis on the presumed relationship context in which sex is occurring and only very minimal discussion of more 'outer limits' types of sex such as using 'manufactured objects' or masturbation. Thus, the types of sexual activity that are treated as within the charmed circle for older people in this book are predominantly those from within Rubin's original charmed circle.
In contrast, a more recent popular book Naked At Our Age (Price 2011) draws the charmed circle very differently. Price's voice is markedly more liberal and permissive and her advice is much franker and more specific. For example, she talks about testing out vibrators asking questions such as 'Could I figure out the controls without wearing reading glasses?' (p. 52) and has a discussion of the pros and cons of different types of lubricant (pp. 18-22). The cover features a picture of a white lily against a black background-there is no assumption of any relationship context to sex on the cover. Non-heterosexual people are somewhat more normalised in this book. The chapter entitled 'Off the beaten path: Nontraditional sex practices and relationships', which might be anticipated to be a chapter on LGBT older people is, in fact, about non-monogamy, BDSM, internet or phone sex, and older women who have sexual relationships with much younger men. Another chapter discusses paying for sex, in a way that legitimises and normalises doing so. These types of sex would be categorised as 'outer limits' in Rubin's original charmed circle diagram but here Price explicitly includes them as legitimate sexual possibilities. Older lesbians and their sexual experiences are simply included in the main text without comment alongside the many case studies of heterosexual men and women. References to gay and/or bisexual men are much scarcer and are introduced with a little commentary on how few responded to the author's requests for information on a particular topic (p. 241). Despite the greater visibility of some forms of non-heterosexuality, Naked At Our Age is still predominantly a heteronormative vision of later life sex and one based on very conventional notions about gender and sex, such as the idea that women require intimacy to have sex and that men are more focused on the mechanics of sex (e.g. in a letter addressed to a man 'For women, sex doesn't always begin with lust, but instead starts in our hearts and minds', p. 47).
Naked At Our Age contains extensive discussion of the use of 'manufactured objects', such as vibrators and other sex toys and also of the use of hormone replacement therapy and medications for erectile dysfunction, which can be understood as modern versions of Rubin's 'manufactured objects'. The new imperatives to continue to have sex may, for some older people, come with the complication that, in order to have 'good, normal, natural, blessed' sex (and especially penis-in-vagina sex) they may need to use 'outer limits' manufactured objects and technologies. The tension that this creates is evident in the following justification of the use of vibrators: "We don't object to putting moisturizer on our faces or a brace on our kneeso why not use vibrators to keep orgasms coming? […] Orgasms are important for keeping the genitals healthy and the sexual responses coming. Think of it as physical therapy, if that makes it more palatable" (Price, 2011, p. 50) Clearly using vibrators is not uncontentious. Price is doing a lot of rhetorical work here to establish that it is acceptable to use sex toys. She naturalises them by comparing them to common everyday self-care practices for women and to prosaic medical devices. She invokes health and ones's implied responsibility to keep oneself healthy, a key part of the successful ageing agenda (Katz and Marshall 2003). Here, we can see someone actively redrawing the charmed circle for older people to include these kinds of technologies.
Analysis of Naked At Our Age suggests that what the book endeavours to present as constituting 'good' (normal, natural, blessed) sex for older people is broadly the same as what constitutes 'good' sex for younger people. There is the same normalisation of sex toys and legitimisation of mild forms of kink Mulholland 2011). There are also the same concerns with relationships, intimacy, authenticity and self-discovery alongside personal responsibility for working to resolve sexual problems (Gupta and Cacchioni 2013). What is implicitly excluded from the version of the charmed circle drawn by this book includes: fuller forms of BDSM, the sexuality of transpeople, bisexual women, people with other Queer identities and intersex people; and non-consensual sex of any type, especially sex between adults and children, which as Rubin herself notes (Rubin 2011), has become even more outer-limits than when she first created the sex hierarchy.
Thus, in some contexts, older people's sexual activity is increasingly treated as within the charmed circle, but not straightforwardly so. The continuing need that writers, including academics, feel to assert the legitimacy of older people's sexual activity suggests that it is not yet truly part of the charmed circle. It may have moved higher up the sex hierarchy than when Rubin was originally writing but the rhetorical work that is still devoted to justifying and normalising sex in later life suggests that it is still somewhat problematic and suboptimal.

Reversing the Spokes: Procreative Sex and Marriage
I turn now to consider what a focus on ageing adds to our understanding of the nature of the charmed circle. I identify a phenomenon whereby types of sex that would usually be categorised as within the charmed circle are instead treated as 'outer limits' because of the particular characteristics of the people or situations involved. I describe this redrawing of the charmed circle as 'reversing the spokes.' Procreative sex and sex within marriage were placed by Rubin within the charmed circle, and non-procreative sex and sex between people who are not married were placed in the outer limits. Since Rubin was writing, non-procreative sex and sex between people who are not married has become much more widely accepted (Watt and Elliot 2017) and thus are much more commonly included within the charmed circle, but it remains uncommon to find the converse: that procreative sex and sex within marriage are treated as 'bad, abnormal, unnatural and damned'. This reversing of the spokes can, however, be encountered in relation to some particular Later Life Sex and Rubin's 'Charmed Circle' categories of person, including young people, disabled people, and of most interest to this paper, older people.
Procreative sex, even within marriage, is treated as problematic if the parents are 'too old'. Discursive gerontologists have drawn our attention to the high local variability of who is included in categories such as 'older person', 'in later life', 'the elderly' and so on (Jones 2006;Nikander 2002). The case of older motherhood exemplifies particularly clearly the ways in which what counts as being 'old' varies by context. Women aged over 35 having their first babies are generally classified as 'older' mothers, and even as 'elderly' in the medical term 'elderly primigravida'. While the average age of first-time motherhood in the UK, as elsewhere in the developed world, continues to rise, it is still only in the late 20s (Office for National Statistics 2019). Nonetheless, discussion of the dangers and risks of later motherhood is pervasive (Budds et al. 2013). Post-menopausal pregnancy, via In-Vitro Fertilisation (IVF) usually using donor eggs, is rare but receives considerable negative media attention (Parks 1999;Perrier 2012). While procreation in these circumstances is not achieved through traditional sex, the association between procreation and sex justifies considering it in relation to Rubin's charmed circle. Thus the literature on later motherhood suggests that, while 'older' women having sex that leads to procreation may be an increasing occurrence, it is not treated unproblematically as 'good, normal, natural and blessed'.
Both public discourse and academic study of older fathers is much scarcer (Carnoy and Carnoy 1997). In everyday talk, older fatherhood is often treated as evidence of a man's continuing virility, especially in relation to male celebrities such as Michael Douglas, Kevin Costner and Larry King (McCabe 2012). This is in sharp contrast to commentary on older mothers which makes no commensurate reference to admirable fecundity. When commentary on older fathers is negative, it tends to focus on the possibility that older fathers may pass on genetic mutations, such as increased susceptibility to autism and schizophrenia (e.g. Alok 2012) rather than on the parenting or wider social implications of being an older father.
Thus, the example of older people's procreative sex demonstrates how differently the charmed circle can be drawn for people within the category 'older people' according to their gender. It also demonstrates how differently the charmed circle can be drawn once people, especially women, are categorised as 'older'-even when 'older' means aged over 35-thereby reversing the usual privileging of procreative sex.
Turning to marriage, sex within a committed but non-marital relationship is often now treated as equally, or if not, almost as 'good, normal, natural and blessed' as marital sex (Barker 2018). However, in later life, there is additional scope to treat not marrying as morally preferable to marrying. In later life, even people with value systems that generally privilege marriage can treat living 'in sin' as morally preferable to remarrying, thereby reversing another spoke of the charmed circle diagram.
Gerontologists have noted an increasing trend within Europe, North America and Australasia of widowed or divorced older people choosing not to remarry when they form a new different-sex relationship 4 later in life (Levin 2004), either living together or keeping their own homes but being in a relationship, a phenomenon often referred to as a LAT (Living Apart Together) relationship (Jong Gierveld 2002). Researchers have been interested in the reasons for this increase and have identified a range, including financial incentives around pensions and benefits (Moore and Stratton 2002), the desire not to be a carer for a spouse again, and women's wishes for freedom and for more equal relationships, which they see as threatened if they remarry (Davidson 2001). Remarrying is also seen as threatening the 'natural' order of the transmission of wealth down the generations. For example, Jong Gierveld quotes an older man talking about why he does not marry his new female partner: I prefer to be independent…. I have one daughter… and, yes, some money, and she has more children and no money…. A marriage would soon bring us problems. I do prefer to give my money to my daughter and my grandchildren. (Jong Gierveld, 2002, p. 73) In this example, the speaker offers both independence and issues of inheritance as reasons not to remarry. As well as relatively pragmatic reasons such as this, the choice not to remarry can also be framed as 'good, normal, natural and blessed'. Older people can draw on the idea of a normative life course in which they have already accomplished marriage and so have no need to repeat it. Davidson found that older men (age 80 and over) were particularly likely to opine that remarriage would be inappropriate at their age (Davidson 2001). Choosing not to remarry can be framed as morally preferable because it honours a dead spouse who is thus not 'replaced' and as fairer to an imagined new spouse who could not live up to the romance and excitement of a relationship formed in youth (van den Hoonaard 2002). Not remarrying can also be treated as morally preferable because it protects the emotional interests of adult children. Remarrying might threaten to replace the dead parent, whereas 'living in sin' gives the new partner a lesser status. This is not to argue that the spokes are always reversed in this way for older people. Some older people do remarry when they form new relationships in later life, and some claim moral (especially religious) grounds for doing so (Jong Gierveld 2002, p. 70-71). Rather, the point is that growing older troubles the privileged status given to marital sex in Rubin's original version of the charmed circle. While nonmarital sex is nowadays much more commonly treated as 'good, normal, natural and blessed' than when Rubin was writing, the case of later life provides a rare example of the converse-a life situation where choosing not to marry can be treated as morally preferable to marrying.

3
Later Life Sex and Rubin's 'Charmed Circle'

Redefining Sex
This then begs the question: does ageing create possibilities for redefining what constitutes good (normal, natural, blessed) sex? Does, for example, the discursive possibility of treating non-marital, non-procreative sex as morally preferable mean that definitions of sex can become more pleasure-focused? Can any ageing-related difficulties with penis-in-vagina sex lead to a widening of sexual possibilities and a rejection of the coital imperative (McPhillips et al. 2001)? Scholars of disability and sexuality, especially those drawing on crip theory (McRuer 2011), have argued for the radical potential of disability to redefine sex: given increased longevity and the prevalence of late-onset disability, can ageing also help to queer sexual norms, as Gallop (2019) argues? Gerontologists too have argued that ageing should create new possibilities for defining what constitutes sex (Calasanti and King 2005;Deacon et al. 1995) and there is some empirical evidence that it does.
For example, Rhiannon Jones found that, while some of her heterosexual women participants aged 70-83 understood non-penetrative sex to be 'second best' or not sex at all, others gave accounts that defined sex more widely and valued, for example, the pleasures of multiple ways of touching and new underwear (Jones 2011, p. 181;. Rebecca Jones discusses an account from an older woman of a new different-sex relationship in later life: "The whole sexual part of it was absolutely wonderful and like nothing I'd ever experienced before, which is so strange, being that he was also on Warfarin […] so there was no erection, there was no completion as such. […] I don't know, this woman certainly found it enormously satisfying, joyous and satisfying and erm it was wonderful. And we both thought it was wonderful" (Jones, 2002, pp. 137-8) Jones discusses the ambiguity of the phrase 'there was no erection, there was so completion as such', arguing that it might indicate no penetration, no ejaculation or, more radically, no orgasm. Hughes (2011) conducted research with a group of heterosexual men in their late 50s and 60s, most of whom were experiencing significant illness and disability which affected their sexual functioning. He examined the extent to which they continued to draw on heteronormative understandings of sexuality which conceptualise the erect penis at the centre of their sexuality. While some respondents continued to define (hetero)sexuality phallocentrically, some did not. One participant, who stated he had not had an erection since 1994, linked his inability to have penetrative sex with his interest in wearing women's clothes and being spanked or caned, saying 'I didn't get an orgasm but I got satisfaction out of it, it satisfied my brain.' (op cit, p. 98). Such a redefinition of sex, not as orgasm but as 'satisfaction' is found in other accounts of sado-masochism (Langdridge and Barker 2007) but here is linked explicitly to age-related bodily changes. Sandberg discusses accounts from older men which foregrounded intimacy in later life sex and a process of becoming a more considerate lover (Sandberg 2013). A study which solicited participants aged over 60 who had had 'great sex' found that some reported improved sex now that penis-in-vagina sex, and sometimes also orgasm, was no longer possible (Dana Ménard et al. 2015).
It should be noted that all the studies mentioned here, and others with similar findings (e.g. Hinchliff and Gott 2004;Tetley et al. 2018) also found a continuation of phallocentric heteronormative understandings of sexuality, with other forms of sexual expression treated as lesser or as not properly constituting sex. Further exploration of the contexts in which older people are, and are not, able to redefine sex would be of great interest. The 'baby boomer' generation are often characterised as more sexually adventurous than previous generations of older people and so might be expected to be more likely to define sex widely than previous generations but Tetley et al. (2018) found that cohort differences do not seem to account for diverse understandings of the nature of sexual activity. Perhaps local interactional variables-such as the way in which a research encounter as framed, or the ways in which speakers position themselves-have a more significant role in enabling redefinitions of good sex than has yet been examined.
Nonetheless, it is already clear that ageing creates new possibilities for some older people in some contexts to reimagine sex in broader ways than those they have been able to access in earlier years. It is, however, noteworthy that much of the literature on older people's sexual activity assumes heterosexuality (Westwood 2018) and very little has been written about the sexual activity of older lesbians, gay men, bisexual people or transgender people. A few studies consider the effects of issues such as diabetes and menopause on sexual experiences for lesbians, gay men and/or bisexual people (Jowett et al. 2012;Winterich 2003), and some find benefits for older people in same-sex relationships in the form of better communication and wider definitions of sex than their contemporaries in different-sex relationships (Paine et al. 2019;Winterich 2003). However, we do not yet have a significant body of evidence about how ageing might create new possibilities for redefining sex for people who are not heterosexual and cisgender.

Refusing Sex
Later life also provides opportunities for individuals to refuse to have sex at all, and to define themselves as someone for whom having sex would be inappropriate. While people of all ages choose not to have sex at all, making public this choice usually incurs significant social stigma, particularly if this choice is tied to lack of desire, rather than other reasons such as religious celibacy or recovery from a painful relationship break-up (Carrigan 2011). However, in later life a choice not to have sex, and even a declaration of lack of desire or sexuality, can be non-stigmatising. The continued availability of the 'asexual older person' discourse, which places the sexual activity of older people within the 'outer limits' makes it relatively straightforward to refuse sex and sexuality on the grounds of old age-after all, to do anything else would be 'bad, abnormal, unnatural, damned'. While the pressures on older people to remain sexually active as part of a wider project of 'active' ageing are undoubtedly increasing (Katz and Marshall 2003;Marshall 2010), there is still scope to refuse sexual activity on the grounds of advanced age.
Later Life Sex and Rubin's 'Charmed Circle' This may be particularly useful to people who have not enjoyed sex earlier in their lives. Gott and Hinchliff (2003) found that, for older women who had experienced sex as a marital duty that gave them no pleasure, growing older was described as providing a happy liberation from sex. Even people who have previously enjoyed sex may be able to refuse the possibility of sex in later life by drawing on the notion of an inevitable (good, normal, natural, blessed) decline in sexual interest. For example, Jones found that some participants described an abrupt end to their interest in sex, often linked to the menopause, using metaphors such as turning off a tap or turning off a switch (Jones 2017). Invoking biological processes, such as the menopause, naturalises and legitimises having no interest in sex in a way that avoids any personal responsibility for what has happened and so avoids any potential blame or stigma.
The literature on later life sex usually treats the idea that older people are asexual as simply problematic and inaccurate-the stated aim of such work is often to refute this notion. However, it is worth noting that, for individuals who do not wish to have sex in later life, placing the sexual activity of older people within the outer limits of the charmed circle benefits them by naturalising and making unproblematic their lack of sexual desire.

Conclusions
What, then, does applying the concept of the charmed circle to the topic of later life add to our understanding of later life sex and our thinking about the charmed circle itself?
Turing first to what the charmed circle adds to our understanding of later life sex, I have argued that the charmed circle, if theorised as in this paper as locally variable within broader societal norms, allows us to examine the ways in which the boundaries between good and bad sex are policed according to older age status. Older people, writers of self-help resources, academics and other social actors negotiate between possible 'good' and 'bad' versions of sex in later life in complex and nuanced ways. In particular, the notion of the charmed circle helps to explain the co-existence and ongoing tension between the 'asexual' and 'sexy oldie' (Gott 2005) discourses of older people's sexual activity. Academic literature, talk and other cultural products which invoke the discourse of older people's asexuality are positioning older people's sexual activity as beyond the charmed circle. Those that invoke the 'sexy oldie' discourse are endeavouring to place older people's sexual activity within the charmed circle. I have argued that this redrawing is still incomplete, since so much discursive work usually still needs to be undertaken to enable these latter kinds of accounts.
Educational interventions and health and social care services that aim to improve older people's sexual health need to take account of this complex discursive climax if they are to be effective. Simply invoking the 'sexy oldie' discourse is unlikely to empower older people as sexual actors in a context where 'asexual' storylines have previously been dominant. Older people may choose to draw on the idea that pharmaceutical interventions for 'sexual dysfunction' are unnatural 'manufactured objects', in order to resist pressure to be sexually active. Educators and service providers must recognise older people's already-existing enmeshment in these discourses and tailor their interventions in ways that both recognise older people's cultural competence and offer alternative framings to be taken up or rejected.
It is clear that the ongoing redrawing of the charmed circle for older people favours particular kinds of sexual activity and sexual actors. The absence of discussion of older LGBT + people's sexual activity has already been mentioned. Perhaps this silence arises from the desire of researchers and participants to counter reductionist or fetishizing accounts of homosexuality, bisexuality and transgender, which might focus on sexual activity as the defining aspect of these identities and lives. However, the literature on LGBT + ageing is now substantial enough that this danger is surely reduced [see Almack and King (2019), and Fredriksen-Goldsen and Muraco (2010), for UK and US-focused overviews, and see King et al. (2019), for an attempt to redress some of the gaps in this literature]. The current silence suggests that the sexual activity of older non-heterosexual and/or transgender people remains beyond the charmed circle.
As already indicated, the focus in the literature on later life sex is often on addressing the sexual 'problems' of older people caused by common later-life health conditions (Tetley et al. 2018) and this is problematic in terms of perpetuating phallocentric heterosexist conceptions of sex, as well as in uncritically serving consumerist and neo-liberal agendas (Katz and Marshall 2003). DeLamater(2012) argues that this focus on sexual dysfunction perpetuates the notion that later life is a time of sexual decline and difficulty. While I do not disagree, I would, however, argue that this focus on difficulties but also on (some) solutions does at least legitimise some kinds of sexual activity for some people experiencing ill health and disability. By its very existence and prominence, this literature places some kinds of sexual activity of some older people with disabilities and health problems within the charmed circle, in a way that is still seldom the case in relation to the sexual activity of younger people experiencing ill health and disability (Blackburn 2018;Shakespeare and Richardson 2018).
Turning now to the charmed circle, using an older-age lens adds to our thinking about the nature of the charmed circle and of sex hierarchies, and thus has implications for gender and sexuality studies more widely. One conclusion drawn from employing this lens might be that we should add an 'age' spoke to the wheel: the inner charmed circle might be 'sex involving young or middle-aged adults' and the outer limits might be 'sex involving children or old people'. Such an addition would have the benefit of drawing attention to continuing unease about children and older people's sexual activity. It might be accompanied by other additional spokes such as 'sex between able-bodied people' versus 'sex involving disabled people'. However, simply adding an age spoke, or other such spokes, would be to understand the charmed circle as constant and reified and to suggest that sex in later life can never be treated as part of the charmed circle. While it is undoubtedly more difficult to position sex in later life as 'good, normal, natural, blessed sexuality', I have demonstrated that in some contexts, such as academic literature and self-help books for older people, this is at least attempted.
Later Life Sex and Rubin's 'Charmed Circle' The simple addition of more spokes is also inadequate because, for some categories of people (here, older people), the spokes of the wheel can be reversed so that behaviours which would normally be seen as 'inner circle' become 'outer limits'. Older women's procreative sex is often treated as part of the outer limits whereas older men's procreative sex has a more ambiguous position. While married sex can still be treated as part of the charmed circle for older people, there is also new scope to treat the choice not to remarry as morally preferable to the choice to remarry, a framing which is not commonly found for other age groups.
Other redrawings could also be suggested and would further complexify the diagram. For example, this paper has not discussed sex between adults of different generations which has a very marked gender difference. It is acceptable for Catherine Zeta-Jones at thirty to play a love interest for Sean Connery aged sixty-nine in the 1999 film Entrapment, but older women who have sex with younger men are ridiculed (Scodari and Mulvaney 2005). If cross-generational heterosexual adult sex was being drawn as part of the charmed circle, different diagrams would be needed for different genders.Thus, attending to older age status in relation to the charmed circle draws our attention to the continuing significance of gender in the regulation of sexual activity.
Different global contexts are also likely to draw the charmed circle differently. In this overview, I have focused on empirical studies and self-help books based in Britain, North American and Australasia but have not distinguished between these contexts-doubtless more nuanced distinctions between these parts of the world could also be made. Participants in this discourse community (Swales 2016) are already privileged by their use of the globally-dominant English language and by the power and wealth of the countries they inhabit-this paper's focus on this already overprominent discourse community is a limitation. Studies of the disciplining of sexual activity in later life in other parts of the world are urgently needed to redress this balance and may reveal that the charmed circle is drawn very differently. For example, a Turkish study in this journal of experiences of later life sex limited recruitment to married women, perhaps suggesting a local necessity to frame later life sex within a marital context in order to be acceptable to potential participants (Yıldırım Varışoğlu and Yeşiltepe Oskay 2018). Further such studies may have much to add to our understanding of the ways in which the boundaries between 'good' and 'bad' sex are drawn in different global contexts.
Thinking about older people's privileged ability to refuse sexual activity without stigma draws attention to the meta-issue of compulsory sexuality (Gupta 2015) and the way that Rubin's charmed circle and notion of the sex hierarchy do not help us problematise the centrality of sex to current understandings of human nature. It would be interesting, but beyond the scope of this article, to compare the refusal of the possibility of sex in later life with arguments used in the asexual community to legitimate low or non-existent interest in sex (e.g. Carrigan 2011). Such a study would add to our understanding of both ageing and asexuality.
Looking at the example of later life sex makes it particularly clear that how the circle is drawn varies by local context. Attention to this variability is important because it leads to more nuanced and sophisticated theorisations of sexuality. However, although I have argued that the charmed circle should be conceptualised as