1With the historical secularization of French society, religious beliefs have gradually weakened throughout the population and social behaviours are no longer constrained by the norms and prohibitions imposed by religious institutions. But what about the most intimate areas of life, such as sexuality, traditionally bound by strict rules of conduct? Drawing on data from three surveys on sexuality in France, Marion Maudet reveals the major transformations in attitudes to sexuality since the 1970s across all layers of society, but also the disparities between men and women according to their degree of religious commitment. Her analysis shows that it is not actual sexual practices, but rather the level of tolerance towards different forms of sexuality that distinguishes believers and non-believers, Catholics and Muslims.
2France, like many other European countries, is in a process of secularization. But the transformation is less marked by the disappearance of religiosity than by its reorganization (Portier, 2012b). The strong positions adopted by Catholics but also by Muslims, in the debate over the legalization of same-sex marriage in 2013 reflects the continued importance of religion in the French public sphere on issues of sexuality. It also sets France apart from countries such as Belgium, the Netherlands and the UK, where similar laws have been passed without provoking such a vigorous response (Paternotte et al., 2016). Drawing on quantitative data predating the legalization of same-sex marriage in France, this article explores the combined influence of gender relations and religion on sexuality, without focusing solely on the question of homosexuality.
3The dynamics that have transformed religion and sexuality have converged substantially in recent decades. Affective and sexual trajectories have diversified, as have religious ones, while practices have become individualized and some moral norms have shifted. The religious landscape in France was marked by major upheavals in the twentieth century. Catholicism was confronted by “the end of a world” (Hervieu-Léger, 2003), as the number of observant Catholics fell sharply, the practice of religion dwindled and “modes of belief” took new forms (Hervieu-Léger, 1999; Raison du Cleuziou, 2014). These developments have been accompanied by a considerable increase in religious non-affiliation, though in many different guises (Bullivant, 2012). In addition, Islam has gained ground in recent times and has become a more important component in the lives of many Muslims (Dargent, 2010).
4Islam, like Catholicism, asserts that the sexes are not equal and attributes clearly defined gender roles to men and women, considered as complementary (Fortier, 2013; Rochefort and Sanna, 2013). The Catholic Church may well have softened its stance on sexuality, but it remains firmly in favour of the two-parent heterosexual couple (Carnac, 2013). For Muslims, despite a much less hierarchical institutional structure and the variability of Islamic sexual norms (Fortier, 2013), virginity remains an important standard of reference, particularly that of women (Ben Dridi, 2013), as does conjugal, reproductive and heterosexual sexuality for both sexes.
5The existence of religious norms governing many aspects of sexuality and relationships between men and women does not mean that they are necessarily followed to the letter; indeed, people are sometimes unaware of their very existence (Portier, 2012a). Attitudes to sexuality among Catholics (Béraud, 2016) and Muslims (Larisse, 2016) are diverse. The same applies to sexual norms. Individuals have entered a “new normative system” characterized by a distancing from major intangible principles imposed from the outside and an internal adaptation of behaviours (Bozon, 2004).
6To understand the sexuality of Catholics and Muslims, the two groups need to be placed in historical perspective. The number of people reporting an affiliation with the Catholic religion has fallen sharply: according to the European Values Survey (EVS), 70% of the French population said they were Catholic in 1981, while just 42% did so in 2008 (Bréchon, 2009). Yet Catholicism is still often described as the French religion “of the people” (Hervieu-Léger, 2003, p. 7).
7In contrast, the number of people reporting an affiliation to the Muslim religion in surveys has risen sharply since the end of the 1990s. According to the Cevipof surveys, the share of Muslims increased from under 1% in the 1990s to 2.6% in 2004 and 5% in 2007 (Dargent, 2010). Based on data from the Trajectories and Origins survey (TeO), Patrick Simon and Vincent Tiberj (2016) estimate that there are around 2.4 million Muslims in France. According to the 2008 EVS survey, 4.5% of the French population said they were Muslim in 2008, though the real figure is no doubt higher (Bréchon, 2009). The minority position of Islam in France may lead to a stigmatization of Muslims (Deltombe, 2007) that confines them to naturalized and falsely homogeneous categories, often rooted in objections to a male sexuality seen as violent and domineering, and a female sexuality that veers between passiveness and lasciviousness (Fassin and Fassin, 2009; Guénif-Souilamas, 2002). The attachment of young people to Islam is also on the rise, particularly among descendants of immigrants. This may result in part from the desire to avoid various forms of social demotion (educational, occupational), with Islam offering individuals the possibility of “symbolic reclassification” (Kakpo, 2007). The migration status of part of the population also needs to be taken into account. Since the 1970s, “the main immigration countries are no longer in the orbit of Catholicism” but more often in that of Islam, countries where “religion continues to pervade everyday life and, in some of them, holds an official position recognized by public institutions” (Simon and Tiberj, 2016). This article will not address the issue of migration because the surveys on sexuality were not designed to shed light on migration origins and trajectories (the samples were not large enough, making it difficult to differentiate migration trajectories by the respondents’ countries of origin).
8It is particularly interesting, then, to explore the way in which religious questions, intertwined with gender relations, are linked to changes in contemporary sexuality. To what extent can Catholic and Muslim men and women be distinguished from people with no religion in terms of sexuality in an environment of increasingly individualized religious, affective and sexual trajectories? The article draws on three large quantitative surveys on sexuality in France, in which the respondents were also asked about the place of religion in their life. This study focuses on Catholicism and Islam owing to their importance in the French religious landscape; the comparison cannot be extended to other religions because their numbers are too few in the respective surveys.
I – Research methodology
1 – Description of the surveys
9The so-called Simon Survey on the sexual behaviour of the French population, coordinated by Pierre Simon in 1970, interviewed 2,625 people aged 20 and over (Table 1). The survey addressed the realities of “ordinary” sexuality as practiced by married couples. It shed light on changes in sexuality over the generations at a time when medical contraception was first becoming widely available. Oral contraception was legalized in France in 1967, [1] counter to the position of the Catholic Church as asserted by the Vatican in 1968 with the Humanae Vitae Encyclical. The survey showed that the imperative of procreation was growing weaker from one generation to the next, to the benefit of sexual pleasure for its own sake. The raw survey data were lost and our analysis is based on the data published in the “Report on the sexual behaviour of the French population” dating from 1972 in which the questions asked to respondents were cross-referenced with a set of socio-demographic variables, the “criteria of systematic analysis”, including age, sex, place of residence, educational level, occupation and religious practice. The analyses presented below are thus strictly dependent upon the choices made by the survey organizers when they processed and published the data. Religious practice was studied for two age groups only, ages 20-49 and 50 and over, so it is impossible to refine the study by using narrower age groups.
Description of the three surveys on sexuality in France

Description of the three surveys on sexuality in France
10The second survey, “Analysis of sexual behaviour in France” (ACSF), coordinated in 1992 by Alfred Spira and Nathalie Bajos, focused on the issues of AIDS, the propagation of the disease and its prevention (Spira et al., 1993). The approach was multidisciplinary; epidemiologists, sociologists, demographers, economists, psychologists, psycho-sociologists and psychoanalysts helped to design the questionnaire. Responding to a call for tender from the National AIDS Research Agency, ANRS, the survey sought to improve understanding of risk factors, but also to shed light on the behaviour of individuals and the way they handled the risk of AIDS, the aim being to develop effective prevention strategies. The analysis is based on a broad definition of sexuality, notably including a psychological dimension (such as the individual’s attitude to death and their ability to control events), a relational dimension and a series of representations of sexuality.
11The last survey, the “Context of sexuality in France” (CSF), was coordinated in 2006 by Nathalie Bajos and Michel Bozon. It takes account of the increasing diversity of affective and conjugal trajectories (longer youth period, greater conjugal mobility, longer sexual life), the power relations between men and women that inform their sexuality, and the material living conditions that determine the way in which individuals experience their sexuality (Bajos and Bozon, 2008).
12As we were using data on sexuality from several surveys, we had to consider their comparability. The wording of the questions varied from one survey to the next, as did the meaning of the categories used, obliging us to reason in terms of indicators by finding “functional equivalents” (Hubert, 1998) for purposes of comparison.
2 – An indicator of religiosity
13The “religiosity” indicator groups information both on religious affiliation (Catholic, Muslim or no religion) and the individuals’ attitude to their religion. The 1970 survey considered religious practice and the two following surveys the “importance” given to religion in life. The Simon survey of 1972 distinguishes between “regular practitioners” who attend religious services at least once a month, “occasional practitioners” who attend services less than once a month and individuals reporting no religious practice. The two following surveys define “observant” Catholics and Muslims as those who report that religion is “important” or “very important” in their life and “indifferent” Catholics and Muslims as those who consider religion to be “not very important” or “not important”. People “with no religion” report no religious affiliation.
14Consequently, we need to be cautious when comparing the religiosity indicators in 1970 with those in the following surveys, the first measuring a practice and the next two the place of religion in the life of an individual. But the comparison is justified by our analytical approach, as we compare differences in sexual practices between men and women according to their degree of religiosity and never compare in an absolute manner the level of practice between two surveys. The analysis of changes in sexuality between 1970 and 2006 is carried out only for Catholics and people with no religion, the number of Muslim respondents being too low in surveys before that of 2006.
15Between 1970 and 2006, the number of respondents reporting an affiliation with the Catholic religion remained relatively stable between 1970 and 1992 then fell more sharply between 1992 and 2006, whatever the degree of importance of religion in their lives (Table 2). In 1992 and 2006 the differences between women and men were very limited among indifferent Catholics, but the share of observant Catholic women was higher than that of observant Catholic men by nearly eight percentage points. The group of people with no religion contains more men, as it has done since 1970.
Breakdown of Catholics and people with no religion by sex in 1970, 1992 and 2006 (%)

Breakdown of Catholics and people with no religion by sex in 1970, 1992 and 2006 (%)
* The “Other” category groups all other reported religions, the “Other religion” category and non-response. It is not analysed in what follows.Interpretation: In 1970, 28.5% of women reported regular religious practice, at least once a month.
Coverage: Individuals aged 20-49.
3 – Indicators of sexual practices and representations
16The following sexual practices were selected: early first sexual intercourse, [2] extensive consumption of erotic books – five or more in a lifetime (Simon, 1970) – or pornographic films (ACSF, 1992; CSF, 2006); having masturbated; having had heterosexual anal sex; and having used contraception at first sexual intercourse. While these five practices obviously do not cover the entire range of sexual practices, they do reflect changes in contemporary sexuality (Bozon, 2013).
17Age at first sexual intercourse decreased substantially in the second half of the twentieth century and became more similar for men and women. Pornographic films became “everyday cultural products” (Bozon, 2008, p. 279; Gagnon and Simon, 1973), the practice of masturbation became much more widespread – though remains less reported by women than men – and contemporary sexuality may be described as “contracepted” sexuality (Bajos et al., 2012; Bozon and Leridon, 1993, p. 1180).
18These practices are also subject to a powerful normative discourse on the part of religious institutions. The magisterium of the Catholic Church condemns masturbation (Catechism of the Catholic Church, 1999 §2352) [3], pornography (ibid., §2354), artificial contraception (Humanae Vitae Encyclical, Paul VI, 1968; Catechism of the Catholic Church, 1999, §2370) and sexual relations outside marriage (ibid., §2353). While anal penetration is no longer specifically mentioned in the catechism of the Catholic Church, the practice is seen as contrary to reproductive sexuality and therefore often censured.
19In Islam, illicit sexuality is referred to as zina, meaning adultery or fornication (Fidolini, 2015; Lagrange, 2008). Muslim law (fiqh), far from being immutable, is based on a pragmatic approach (Fortier, 2010b) and depends on the interpretations of doctors in Islamic jurisprudence. But it is based on principles common to the Koran and Sunna (comprising the sayings, or hadith, of the prophet), whereby virginity before marriage is required and both pornography and anal penetration are forbidden, as is masturbation where performed alone rather than with a spouse (Fortier, 2010a). Positions on contraception are more nuanced, with coitus interruptus generally tolerated, together with the use of condoms and the pill by married couples (Fortier, 2010b).
20For representations of sexuality, we use four indicators: fidelity, pornography, prostitution, and the idea that regular sexual relations are necessary to “wellbeing” (the wording in the 2006 survey). The first three figure largely in Catholic and Muslim conceptions of sexuality, while the fourth is characteristic of the new place of sexuality in the lives of individuals, having become a central aspect of conjugality in the last century (Bozon, 1991).
21We could have chosen other practices and representations, but those studied here have two advantages: they are available in the various surveys so can be analysed over a long time span, and they provide a reflection of observable trends for an entire set of other practices and representations relative to sexuality (oral sex, talking about sexual problems, having already been in love in one’s life, or having had four or more partners in one’s life).
II – Changes in the sexuality of Catholics from 1970 to 2006
1 – Sexual practices: growing similarities between Catholics and non-religious people
22In 1970, the sexuality of Catholics was highly different to that of people with no religion (Table 3). First, they had first sexual intercourse at a later age. This was especially true for men, with just 7.8% of 20-49 year-old regular church attenders having had early first sexual intercourse compared with 28.6% of non-religious men in the same age group. The share of those having had “other sexual experiences” (in the words of the survey [4]) was lower than that of people with no religion. The proportion of male regular church attenders having experimented anal penetration was considerably lower (11.5%) than that of men with no religion, (32.3%).
Sexual practices by sex and Catholic religious affiliation in the three surveys (1970, 1992 and 2006)

Sexual practices by sex and Catholic religious affiliation in the three surveys (1970, 1992 and 2006)
NA: question not asked in survey.Statistical significance levels for differences between men and women: * 10%; ** 5%; *** 1%; ns: non-significant. Significance reads along rows for each practice. The significance of the results of the 1970 survey cannot be calculated owing to the lack of raw data.
(1) First sexual intercourse is early if it occurs at age 18 or earlier for women and age 16 or earlier for men in 1970; at age 16 or earlier for women and age 15 or earlier for men in 1992 and 2006.
(2) 1970: Has read “5 erotic books or more in his/her lifetime”; 1992: Has “often” or “sometimes” seen a “pornographic film or show in his/her lifetime”; 2006: Has “often” or “sometimes” seen a “pornographic film in his/her lifetime”.
(3) “A man (woman) may couple via the rectum (anus). Have you ever had this type of sexual relation with a woman (man)?”
Interpretation: In 1970, 10.6% of regularly practicing Catholic women had first sexual intercourse at age 18 or earlier. Coverage: Individuals aged 20-49 (n = 2,625 for Simon; n = 3,909 for ACSF; n = 7,658 for CSF).
23The differences between women by religious practice were generally less substantial in 1970 than those between men, as regards having read at least five erotic books during one’s lifetime, having masturbated at least once and having already practiced anal penetration. [5] This may be the result of closer proximity between gender norms and religious norms at the time, which tended to attenuate the effect of religious practice on women’s sexuality.
24In 1992 and 2006, the sexual practices of Catholics and people describing themselves as having no religion became much more similar. For men, differences between observant Catholics and individuals with no religion were minor or even non-significant for all the sexual practices under consideration, apart from anal penetration in 2006 (24.5% and 31.1%, respectively). The results were more nuanced for women. In 1992, the practices of Catholic women and women with no religion were similar in terms of pornography but considerably different in terms of masturbation (20 percentage points), although more similar in 2006. However, in 1992 as in 2006, the age at first sexual intercourse continued to differ according to the individual’s religious commitment, particularly among women. In 1992, only 20% of observant Catholic women had had first sexual intercourse at age 16 or earlier, compared with 37.7% for women with no religion. The situation was much the same in 2006.
25Differences between the sexual practices of Catholics and people with no religion have narrowed over time, except for age at first sexual intercourse, which, as recently as 2006, was higher among Catholics than people with no religion, particularly for women. In 1970, the influence of religious practice on sexuality was substantial for men for certain practices, but less so for women due to the proximity between social and religious norms. This was no longer the case in 2006.
2 – Representations of sexuality: an enduring divide
26In general, the attitudes to sexuality of men and women reporting religious practice in 1970 were less liberal than those of individuals who reported no religion (Table 4). Practicing Catholics were often more critical of infidelity on the part of men and women than people with no religion, while pornography and prostitution were less accepted and regular sexual relations more rarely considered as “necessary to well-being”. Male attitudes to the sale of erotic books and to prostitution differed between religious and non-religious people. The proportion of practicing Catholic men considering “prostitution intolerable” was twice as high as that of men with no religion (at 36% and 15%, respectively). For women, the greatest differences between those with and without a religion were found in opinions on the prohibition of pornography and the idea that “regular sexual relations are necessary to well-being”. Lastly, while there was little difference in 1970 between women with and without a religion in terms of the importance placed on fidelity in marriage, opinions on the subject became more divergent in the following decades.
Representations of sexuality by sex and Catholic religious affiliation in the three surveys (1970, 1992 and 2006)

Representations of sexuality by sex and Catholic religious affiliation in the three surveys (1970, 1992 and 2006)
NA: question not asked in survey.Statistical significance levels for differences between men and women: * 10%; ** 5%; *** 1%; ns: non-significant. Significance reads along rows for each practice. The significance of the results of the 1970 survey cannot be calculated owing to the lack of raw data.
(1) 1970: “A woman should consider the occasional infidelity of her husband as unforgivable”; 1992: “It is entirely unacceptable for a man to have sex outside of marriage”; 2006: Identifies with the statement “Faithful for life”.
(2) 1970: “A man should consider the occasional infidelity of his wife as unforgivable”; 1992: “It is entirely unacceptable for a woman to have sex outside of marriage”; 2006: Identifies with the statement “Faithful for life”.
(3) 1970: “The sale of erotic books should be forbidden, regardless of the age of the person seeking to buy them”; 2006: “Pornography should be banned”. In 1970, given the high non-response rate, the results are provided only for individuals having answered the question. The non-response rates are as follows, for women and men, respectively: 29% and 22% for regular practitioners,
29% and 23% for occasional practitioners, and 24% and 23% for non-practitioners.
(4) 1970: “Prostitution is intolerable. We need to work to eradicate it at all costs”; 2006: “Prostitution should be curbed”.
(5) Given the high non-response rate, the results are provided only for individuals having answered the question. The high non-response rates are as follows, for women and men, respectively: 37% and 16% for regular practitioners, 28% and 26% for occasional practitioners, and 31% and 23% for non-practitioners.
Interpretation: In 1992, 38.5% of observant Catholic men considered that it was entirely unacceptable for a man to have sex outside of marriage.
Coverage: 20-49 year-olds (n = 2,625 for Simon; n = 3,909 for ACSF; n = 7,658 for CSF).
27Between 1970 and 2006, the difference between observant Catholics and people with no religion decreased substantially for certain sexual representations, including prostitution and the importance of sexual relations to well-being. This latter representation reflects the shift in sexual norms whereby sexual relations came to be seen as indispensable to individual and conjugal wellbeing (Bajos et al., 2008; Bozon, 2013). However, opinions diverged on fidelity and pornography, with Catholics much more attached to the idea of fidelity than people with no religion.
28The study of changes in sexual practices and attitudes appears to show that, apart from age at first sexual intercourse, it is in their opinions on sexuality that individuals now present the greatest differences by level of religiosity. To verify this point, we focused on the 18-29 year-old Catholics and Muslims interviewed in the most recent survey (CSF, 2006). These were the first to be concerned by the “deep-seated reorganization of norms on the transition to adult sexuality” (Bozon, 2004), namely the individualization and interiorization of social and sexual norms, which individuals no longer perceive as the product of institutions exterior to themselves.
III – The sexual practices of young Catholics and Muslims in 2006
29In 2006, sexual behaviour varied for young people by sex and religiosity, [6] suggesting that gender relations and religion should be analysed together. Apart from age at first sexual intercourse for women, there are few differences between young Catholics and young people with no religion in all other sexual practices (Table 5). Observant Catholic women have early first sexual intercourse less often than women with no religion (25.4% and 40%, respectively), confirming for young people the situation observed among 20-49 year-olds. The same is true for observant Muslim women. The significance associated with the first sexual intercourse varies by degree of religiosity. For more than half of observant Muslim women, the first sexual partner is considered as a “spouse” or “future spouse”, while this is the case for just 24% of observant Catholic women and 16% of women with no religion (results not presented). For women, first sexual intercourse and its significance are structured by their religious commitment, but the same is not true of men, reflecting the gender differential in the interplay between religion and sexuality. Observant Muslim women also less frequently say they have engaged in oral sex in the last 12 months (46.2%, compared with 60.3% for women with no religion). This may mean that Muslim women practice oral sex less often, but it may also signify that they find it harder to report such a practice (Lahire, 1998).
Sexual practices by religiosity in 2006 among men and women aged 18-29 (%)

Sexual practices by religiosity in 2006 among men and women aged 18-29 (%)
(1) The earliness of the first sexual intercourse is determined on the basis of the earliest age at first sexual inter- course in the first quartile. The age of early first sexual intercourse was 15 or earlier for men and 16 or earlier for women.Statistical significance levels for differences between men and women: * 10%; ** 5%; *** 1%; ns: non-significant. Significance reads along rows for each practice.
Interpretation: 22.4% of observant catholic women aged 18-29 have often or sometimes watched pornographic films in the last 12 months.
Coverage: Individuals aged 18-29 interviewed using long questionnaire (n = 2,819).
30For men, the greatest differences between Muslims and those with no religion concern masturbation and the use of online dating sites. Observant Muslim men report masturbation less frequently than non-religious men and the former say they make more use of online dating sites. The more frequent use of dating sites by young Muslim men can be attributed to a preference for «people like themselves” and to the use of “specialized” sites, notably targeting specific ethnic or religious groups (Bergström, 2011), as well as a desire to escape the control of peers, friends and family and to engage in a form of seduction that does not necessarily involve an actual physical encounter (Bergström, 2012; Bozon, 2013). Also, as Muslims have a greater experience of discrimination than the other religious or non-religious groups (Brinbaum et al., 2016), the use of these sites may help them to avoid situations of exclusion in physical meeting places.
31Lastly, among the youngest people interviewed in 2006, statements on sexual practices are very similar between Catholics and individuals with no religion, apart from age at first sexual intercourse for women. But the sexuality of Muslims is very different from that of people with no religion, with first sexual intercourse occurring later and more often being associated with a conjugal commitment. Sexual practices (oral sex, masturbation, pornography) are reported less often by Muslims, especially women, and Muslim men make greater use of online dating sites. So while the sexual behaviour of young Catholics is similar to that of young people with no religion, that of 18-29 year-old Muslims is different from both these groups, with a later sexual debut indissociable from conjugal commitment for young women, and with an early sexual debut but a more limited sexual repertoire for young men.
IV – Young people’s attachment to the heterosexual family
32Exploring individual representations of sexuality provides a more comprehensive picture of what people consider to be “good” sexuality. This can be addressed from a number of angles (Table 6). First of all, it is important to consider representations that are not directly linked to sexuality but more broadly concern formal gender equality (how to raise boys and girls, representations associated with the types of work that can or cannot be done by men and women). Second, the aim is to identify attitudes to sexuality (fidelity, pornography, prostitution). Last, representations of sexuality are also based on attitudes to conjugal heteronormativity (representations of homosexual relations, same-sex parenting and parenthood).
33Regarding attitudes to sexuality, two main positions can be distinguished among Catholics: representations close to those of non-religious people concerning prostitution, and more divergent attitudes in relation to pornography (Catholics say they watch it less) and fidelity (to which Catholics are more attached). For Muslims, however, the differences are considerable for all three representations. Observant Muslims of both sexes say that “prostitution should be curbed” and “pornographic films should be banned” more frequently than Catholics and people with no religion. The percentage of Muslim men categorically opposed to pornography is more than four times higher than that of non-religious men (39.3% and 8.3%, respectively).
Representations of sexuality by sex and attitude to religion in 2006 (%)

Representations of sexuality by sex and attitude to religion in 2006 (%)
Statistical significance levels for differences between men and women: * 10%; ** 5%; *** 1%; ns: nonsignificant. Significance reads along rows for each practice.Interpretation: 62.8% of women with no religion consider that a woman cannot, or cannot without difficulty, succeed in life without having a child.
Note: Details of questions and response categories are given in Appendix A.1.
Coverage: Individuals aged 18-29 interviewed using long questionnaire (n = 2,819).
34Observant Catholics and Muslims have more similar opinions on fidelity in a couple and are more attached to fidelity than people with no religion. Attachment is particularly strong for observant Catholic men, 53.2% of whom identify with the words “faithful for life” compared with 32.1% of non-religious men. While young generations as a whole identify more with the idea of “faithful while together” than “faithful for life” (Bajos et al., 2008), this is not the case for the majority of young observant Catholic men, for whom fidelity remains the rule.
35The third group of representations includes those signifying a strong attachment to conjugal heteronormativity. As defined by Cynthia Kraus (Kraus, 2005), heteronormativity refers to “the gender system, asymmetrical or binary, that tolerates two and only two sexes, in which gender is in perfect accord with sex (to the masculine gender the male sex, to the feminine gender the female sex) and (reproductive) heterosexuality is mandatory, or at least desirable and appropriate” (p. 24). These are representations that promote, consolidate and render invisible the power of this asymmetrical system as well as the heterosexual order. They concern homosexuality, same-sex parenting and the place of procreation in women’s lives.
36As already demonstrated by a number of researchers (Bajos and Beltzer 2008; Rault, 2016; Rault and Hamel, 2014), representations of homosexuality are highly gendered, with women generally more tolerant than men. But they are also diverse, and a range of indicators need to be taken into account to understand individual attitudes in this respect (Rault, 2016). Broadly speaking, men are less tolerant of homosexuality than women, whatever the indicator, and observant Catholics and Muslims are less tolerant than non-religious people and indifferent Catholics (Table 6).
37While among observant Catholics homosexuality is not seen as a problem by the majority of women and a little under half of the men, tolerance of homosexuality is much lower than among people with no religion. A logistical regression based on this question attests to the influence, all other things being equal, of being an observant Catholic or Muslim on negative representations of homosexuality (Appendix Table A. 1).
38The social category of the respondents, based in this case on the occupational category of their parents, also influences the perception that homosexual relations are a problem, but this varies by sex. For women and men alike, having a mother in an intermediate occupation reduces the likelihood of considering homosexual relations as a problem. For men, having parents who are unskilled manual workers increases this likelihood.
39Belonging to the working class thus appears to have a negative effect among men on the perception of homosexuality. But it should be noted that this is an approval or disapproval “in principle” as opposed to a ‘”pragmatic” approval (for example, accepting that two people of the same sex raise a child together), which is more frequent among the working classes (Rault, 2016, p. 53).
40The divide between Catholics, Muslims and people with no religion is even more pronounced in attitudes to same-sex parenting (Table 6). Even though, here again and independently of religiosity, women are more tolerant than men, the majority of Catholics and Muslims are opposed to the idea of two people of the same sex raising a child together, unlike non-religious people (one-third of whom nevertheless report being hostile to same-sex parenting, particularly for male couples). Observant Catholics and Muslims, mainly women, tend to place more value on motherhood than people with no religion, with three-quarters of them considering that a woman cannot “fully succeed in life without having a child”.
41Most Catholics and Muslims differ markedly from non-religious individuals in their attachment to conjugal heteronormativity, as regards representations of homosexuality, same-sex parenting and the role of reproduction for women. This provides a partial explanation for the existence of “identity dissonance” or “cognitive dissonance” among homosexual Christians and Muslims (Brault, 2013 ; Gross, 2008). The root of this dissonance is to be found not just in the discourse of religious institutions but in that of the religious practitioners themselves.
42The results on attitudes to sexuality and opinions on conjugal heteronormativity are of special interest, given that representations relative to formal gender equality vary little if at all between Catholics, Muslims and people with no religion. The share of “egalitarian attitudes” is relatively similar across the three groups, mainly among women. Nearly half of 18-29 year-old men and women agree or totally agree with the statement that “There are jobs that can only be done well by men or by women”, even though Catholic women are over-represented. Moreover, fewer than 10% of Catholic women and nonreligious women consider it “a mistake to raise girls and boys in the same way”. The proportions are higher for observant Catholic men and Muslims of both sexes, however. These similarities regarding formal equality thus exist side by side with divergent positions on attitudes to sexuality and conjugal heteronormativity.
Conclusion
43Understanding contemporary sexuality involves observing the sexual practices and representations of individuals, but also their social trajectories and situations beyond the domain of sexuality itself (Bessin, 2009). This article explores the combined influence of religion and gender on sexuality. Defining oneself as Catholic or Muslim is more closely linked to attitudes towards sexuality than to sexual practices themselves. It is above all through their positioning in normative terms relative to sexuality that individuals construct a part of their religious otherness.
44By focusing on the sexual practices and representations of individuals we do not assume in advance that religious norms have an effect on their behaviour and attitudes of religious practicitioners; instead we observe the actual similarities and/or differences between observant Catholics and Muslims and people with no religion. The use of quantitative data enabled us to observe in detail numerous aspect of sexuality over four decades while taking account of two groups often neglected in research on religion, namely indifferent Catholics and people with no religion. [7] We have highlighted that for most sexual practices, the behaviour of Catholic men and women today is similar to that of non-religious men and women, including for highly gendered practices such as masturbation, reading erotic books and watching pornographic films. This was not the case in 1970. At that time, being a religious practitioner was associated, mainly for men, with later first sexual intercourse and less varied sexual practices than was the case for people with no religion. The growing similarity between the sexual practices of Catholics and non-religious people is a finding consistent with earlier research on the rapid dissemination of oral contraception among Catholics (Sevegrand, 1995; Toulemon and Leridon, 1992), while here it encompasses other sexual practices.
45The only remaining difference between Catholic and non-religious women is age at first sexual intercourse. It may be presumed that while most young Catholic women do not remain virgins until marriage, the norm of virginity remains a moral compass for some of them. This could be referred to as “moral virginity” as opposed to effective virginity. Moreover, the injunction to “wait for the right time” or the fear of sullying one’s reputation, especially for young Catholic girls who still live at home, may explain the lower presence of young observant Catholic girls among those having had first sexual intercourse at an early age. Similar findings have been obtained in research on the children of immigrants (Tersigni, 2016) and young women from working-class backgrounds (Clair, 2008), but there is less data on contemporary Catholicism.
46The results concerning the youngest individuals in the 2006 survey show a specific link between Muslim religious affiliation and sexuality. In general, the sexuality of Muslim men and women is less diverse than that of nonreligious people, with the former practicing oral sex, watching pornographic films and masturbating less frequently than the latter. But the influence of Muslim religiosity has a contrasting impact on age at first sexual intercourse, which is later for Muslim women and often early for Muslim men.
47The influence of religion can also be seen, and often more clearly, in sexual representations. While religion has little or no influence on opinions concerning “formal gender equality”, the results are more nuanced when it comes to “attitudes to sexuality”. In this respect, Catholics have relatively similar attitudes to people with no religion, but Muslims are different from both, reporting more conservative stances on pornography and prostitution.
48Differences by religiosity are most clearly expressed in the issue of conjugal heteronormativity. This is characterized less by a rejection of homosexuality in itself – infrequent among Catholics, even observant Catholics – than by a strong attachment on the part of most religious people to the heterosexual family. This is especially true for observant Muslims, though this does not mean that Catholics and Muslims form homogeneous groups characterized by the rejection of non-heterosexual sexuality, as shown by Céline Béraud and Philippe Portier (2015) in their work on “Catholic metamorphoses”. Rather, it reveals the need for a more detailed analysis of the social situations and the contexts of socialization.
49In conclusion, this opposition to same-sex parenting, and the reluctance to accept women’s roles that diverge from the requirement to procreate and to accept traditional family norms, shed new light on the demonstrations against same-sex marriage in France in 2012 and 2013, which were more about defending heterosexual filial relations than opposing same-sex marriage.
Appendix A.1. Questions and response categories used in the survey on the Context of Sexuality in France (2006). Results given in Table 6
Group 1. Formal equality
501. Some people consider that there are jobs that can only be done well by men or by women. Do you agree?
- □ totally agree
- □ agree
- □ disagree
- □ totally disagree
512. Girls and boys today are increasingly raised in the same way. In your opinion…
- □ it’s a good thing
- □ it’s a good thing, but it shouldn’t be taken too far
- □ it’s a mistake
Group 2. Attitude to sexuality
523. What do you think about prostition?
- □ prostitution is necessary
- □ it is inevitable, but needs to be regulated
- □ it should be curbed
534. What do you think about pornographic films?
- □ they should be banned
- □ they are shocking but should not be banned
- □ pornographic films are not particularly shocking
545. For a couple in a relationship, what is the attitude that most closely resembles your own?
- □ they should remain faithful for life
- □ they should remain faithful so long as they are together
- □ they should remain faithful, but free to have an affair
Group 3. Conjugal heteronormativity
556. Regarding homosexual relations, would you say that…
- □ they are a form of sexuality like any other
- □ they are the sign of unrestrained sexuality
- □ they are the sign of a psychological problem
- □ they are a a sign of unnatural sexuality
567. Two men can raise a child together.
- □ totally agree
- □ agree
- □ disagree
- □ totally disagree
578. Two women can raise a child together.
- □ totally agree
- □ agree
- □ disagree
- □ totally disagree
589. To childless respondents: If you had children and one of them told you he/ she was homosexual…
- □ that wouldn’t be a problem for you
- □ that would be a problem for you
- □ you wouldn’t accept it
5910. To respondents with children: If one of your children told you he/she was homosexual…
- □ that wouldn’t be a problem for you
- □ that would be a problem for you
- □ you wouldn’t accept it
6011. In your opinion, can a woman fully succeed in life without having a child?
Odds ratios of considering homosexual relations as the “sign of a psychological problem” or “the sign of unnatural sexuality”
Women | Men | |||
---|---|---|---|---|
Odds ratio | p | Odds ratio | p | |
Constant | 0.25 | *** | 0.68 | *** |
Age | ||||
18-24 (Ref.) | 1 | 1 | ||
25-29 | 1.12 | ns | 0.77 | *** |
Religion | ||||
Observant Catholic | 2.51 | *** | 1.81 | *** |
Indifferent Catholic | 1.12 | ns | 1.30 | ** |
Observant Muslim | 5.98 | *** | 6.45 | *** |
Indifferent Muslim | 0.98 | ns | 1.83 | ns |
Other religion | 6.62 | *** | 2.73 | *** |
No religion (Ref.) | 1 | 1 | ||
In a couple | ||||
Yes (Ref.) | 1 | 1 | ||
No | 0.96 | ns | 0.92 | ns |
Occupation of father | ||||
Farmer | 1.27 | ns | 0.90 | ns |
Self-employed | 1.20 | ns | 0.86 | ns |
Higher-level occupation | 0.94 | ns | 0.75 | ns |
Intermediate occupation | 1.36 | ns | 0.76 | ns |
Clerical / sales worker (Ref.) | 1 | 1 | ||
Skilled manual worker | 1.16 | ns | 1.25 | ns |
Unskilled manual worker | 1.70 | ** | 1.98 | ns |
Occupation of mother | ||||
Farmer | 0.64 | ns | 1.34 | ns |
Self-employed | 0.76 | ns | 0.81 | ns |
Higher-level occupation | 0.51 | *** | 0.72 | ns |
Intermediate occupation | 0.57 | ** | 0.63 | * |
Clerical / sales worker | 0.67 | * | 0.85 | ns |
Skilled manual worker | 0.94 | ns | 0.92 | ns |
Unskilled manual worker | 0.60 | * | 0.91 | ns |
Never worked (Ref.) | 1 | 1 |

Odds ratios of considering homosexual relations as the “sign of a psychological problem” or “the sign of unnatural sexuality”
Significance levels: * < 10%; ** < 5%; *** < 1%; ns: non-significant.Interpretation: The probability for an observant Catholic woman of considering homosexual relations as a problem is 2.5 times higher than for a non-religious woman, all other things being equal.
Coverage: Individuals aged 18-29 interviewed using long questionnaire (n = 2,819).
Notes
-
[*]
Institution de recherche interdisciplinaire sur les enjeux sociaux (IRIS), French Institute for Demographic Studies (INED).
Correspondence: Marion Maudet, Institut national d’études démographiques, 133 bd Davout, 75020 Paris, email: marion.maudet@ined.fr -
[1]
The implementing decrees of the law date from 1972.
-
[2]
First sexual intercourse is defined as “early” when it takes place at the highest age of the first quartile for both women and men. In 1970, the calculation was based on published data. The age of 16 was set for men (21.6% of 20-49 year-old men having had sexual intercourse before age 17), while 28% of women had first sexual intercourse at age 18 or earlier. In 1992 and 2006, early age at first sexual intercourse was set at age 16 or earlier for women (27.1% of women in 1992, 24.4% in 2006) and at age 15 or earlier for men (22.2% of men in 1992 and 22.5% in 2006).
-
[3]
Catechism of the Catholic Church, first edition of 1992. Modified version dating from 1999. [URL : http://www.vatican.va/archive/FRA0013/_INDEX.HTM].
-
[4]
) Having read at least five erotic books, having masturbated at least once, having used contraception for first sexual intercourse, having practiced heterosexual anal penetration.
-
[5]
Regarding masturbation, it may be supposed that women under-report the practice (Béjin, 1993). But as the levels reported are very low for women independently of their religious affiliation, the degree of under-reporting may be relatively similar for Catholic women and women with no religion.
-
[6]
Among Muslims, only “observant” individuals are taken into account here owing to the small size of the sample of Muslims saying that religion is “not very” or “not at all” important in their life. But this does not necessarily mean that this group is homogeneous (see notably the work of Vladislav Serikov (2014) on Muslims in Frankfurt). As the samples of observant Muslims were also small, the results advanced should be read with caution.
-
[7]
For example, the work on non-religious people (“nones”) by Lois Lee (2011), the founder of the Nonreligion and Secularity Research Network (URL: http://nsrn.net/), and Stephen Bullivant (2012).