Living Islam in Prison: How Gender Affects the Religious Experiences of Female and Male Offenders

: Addressing a signiﬁcant gap in the knowledge of female Muslim prisoners’ religiosity, this paper describes and explains the gendered impact of incarceration on the religiosity of Muslim female and male offenders. Based on quantitative and qualitative data collected in ten prisons, including a male and female prison in England and a male and female prison in Switzerland, the authors show that prison tends to intensify the religiosity of Muslim men and reduce the religiosity of Muslim women. In explanation of this, the authors argue that, at the individual level, the feelings of guilt at the absence of family, the absence of high-status religious forms of gender and feelings of trauma and victimhood impact negatively on Muslim female offenders’ religiosity. At the institutional level, female Muslim prisoners, being a small minority, do not mobilise a powerful shared religious identity and chaplaincy provision—including provision of basic religious services—is patchier for Muslim women than it is for men and often does not take into account the speciﬁc needs of female prisoners.


Introduction
Since the pioneering sociological studies of James Beckford and Sophie Gilliat-Ray on religion in prison (Beckford et al. 1998), the study of religion in prison has become a distinct field in the sociology of religion. Whilst the first studies focused on chaplaincy, the location of religion in prison and prisoners' access to religious provision (Ajouaou and Bernts 2015;Becci et al. 2011Becci et al. , 2017Fabretti 2015;Martínez-Ariño et al. 2015), a second more recent wave of research has concentrated on inmates' religiosities (Becci 2012;Becci and Ghica 2019;Béraud et al. 2016;Sarg 2016; Schneuwly Purdie 2020). As prison populations are predominantly male, research on prisoners' religiosity has mainly focused on male offenders' religious profiles and their individual mobilisation of religion. Rostaing (2017) identified three ideal-typical processes that hinder the knowledge and visibility of women in prison. Firstly, she noted that the principle of formal equality between prisoners led to the negation of differences between incarcerated men and women. Secondly, she highlighted the androcentrism which, due to the numerical superiority of male prisoners, prioritizes the study of the general to the detriment of the individual. Thirdly, she stated that penological studies of women prisoners are often reduced to studies on women and gender, rather than making a contribution to prisons' research more generally. Rostaing's reflections on research on women in prison is echoed in the field of the sociology of religion and it appears that "the relationship between gender and religion in prison constitutes a real blind spot in research" (Béraud et al. 2017, p. 151).

Researching Women and Religion in Prison
Indeed, whereas the above-mentioned research concentrated on men, only a few studies have specifically looked at the practice, meaning and the mobilisation of religion for female prisoners. Among these, some have emphasized how prison is an opportunity for women (as well as for men) to rethink their relationship with God deeply or to experience a new image of God (Schneider and Feltey 2009;Becci and Schneuwly Purdie 2012). Similarly, a recent research by Jang and et al. conducted in South Africa, showed that "religiosity was positively related to a search for and a presence of meaning and purpose in life and virtuous characteristics" (Jang et al. 2019, p. 18) such as forgiveness, gratitude and self-control and that religion produced equally salutary effects for both male and female prisoners in enabling them to cope emotionally with the pains of imprisonment (Johnson and Toch 1983). Others have questioned the functions of religion for incarcerated women. They have showed that religion makes it possible for women to (re)create social bonds and can also be a resource for managing physical ageing, illness and end of life away from families (Aday et al. 2014), as well as being a tool to overcome guilt and forgive themselves. Another study asserted that prison was a place of religious intensification for women more than for men (O'Connor and Duncan 2011). Questioning the intersection of gender and religion in prison, Becci and Schneuwly Purdie (2012) underlined gendered differences in the perception of religion in prison. While religion was seen as a positive resource for women (e.g., promoting social bonds, or to deal with emotions), it was seen as a source of conflict and disorder between men (e.g., generating exclusion and discrimination).

Researching Women and Islam in Prison
Likewise, while research on Muslim male offenders is relatively well-developed, studies which specifically address the experience of Islam by female offenders are few and far between. Research on Muslim male prisoners addresses several topics, such as the construction of the "Muslim prisoner" social category often linked monolithically with extremism (Beckford et al. 2007;De Galembert 2020;Schneuwly Purdie 2011), Muslim chaplaincies and the equal access of Muslim inmates to religious provision (Beckford et al. 2007;Béraud and de Galembert 2019;Gilliat-Ray 2013;Rhazzali 2016;Schneuwly Purdie 2013), their religiosities and the impact of the prison context on their religious identities (Ammar et al. 2004;Khosrokhavar 2004;Spalek andEl-Hassan 2007, Wilkinson et al. 2021b) or the mobilisation of their religious beliefs and practices to cope with the pains of imprisonment (Béraud et al. 2016). Furthermore, a number of studies have specifically addressed religious extremism and the impact of imprisonment on processes of radicalisation (Khosrokhavar 2006;Micheron 2020;Béraud et al. 2017;Schneuwly Purdie 2019).
Only a few studies have concentrated specifically on Muslim female prisoners. Among them is Dix-Richardson's research on the differences between African American women and men in terms of conversion to Islam in prison (Dix-Richardson 2002). Dix-Richardson asserted that African American women converted to Islam far less often than their male counterparts. She noted that institutional factors, such as the place of Muslim chaplaincy in prisons, played an undeniable role in the uptake or not of religion. She also claimed that the perception that gender roles in Islam deny women an active role in leadership was offputting for women who were used to African American models of femininity which value their labour power and ability to financially take care of their families. Dix-Richardson also highlighted the opportunity for leadership roles that black churches give African American women in contrast to the perceived absence of leadership roles for women in mosques. Finally, Dix-Richardson suggested that since there was less prison violence in women's institutions, the need for protection of a group "perceived as dangerous and fearsome" (Khosrokhavar 2004, p. 38) was not as powerful a pull factor as it was for men. Other research has strongly challenged the widely held idea that Muslim males choose Islam primarily for protection, perks and privileges (Wilkinson et al. 2021a).
More recently, two French studies have looked at Muslim men and women's issues in dealing with radicalization in relation to Islam. Béraud et al. (2017) showed that the Islam of women was not framed according to a security logic and that it was not feared in the same way as the Islam of male prisoners. The authors emphasised that the religious demands of Muslim women prisoners are often seen as more authentic than those of men. However, a divergence was found in research on women imprisoned for jihadist links and activities. In this case, the women opposed the system of laïcité not through violence, but through the public performance of purity demands (ablution after menstruation or sexual relations), public displays of Islamic dress, requests for halal food and the conduct of exorcism (ruqayya) (Castel de Bergerac 2020).
Building on the above-mentioned studies in the field of the sociology of religion in prison, this article proposes to reduce the gap between research on the religious praxis of male and female Muslim prisoners. It presents some results of an international research project investigating Muslims in European prisons. Based on a sample of imprisoned Muslim men and women collected in two prisons, this contribution articulates a comparative analysis of the type of religious changes experienced by Muslim prisoners during their incarceration.

Intersectionality
This paper draws on the idea of intersectionality as a key intersectional lens. Intersectionality, as originally conceived by Crenshaw, acknowledges the limits of analysing identity within singular dominant categories such as gender, class, age or ethnicity alone (Crenshaw 1991;Kimberle et al. 1995). Rather, this approach asserts that specific intersections such as faith and gender interact to generate a variety of lived effects, including marginalization and discrimination for some people. For example, racism or sexism are themselves shaped by other forms of oppression such as class-based discrimination (Quraishi and Philburn 2015). In the present context of prisons, the research has enabled a meaningful analysis of the disjuncture between institutionalized categorizations (gender, religion, offender type) and the qualitative ways in which Muslim prisoners conceive and live their religious identities.
The paper therefore serves to de-essentialise the category of the "Muslim prisoner" by showing that Muslim religiosity itself was experienced, understood and performed at the intersection of class, ethnicity, education and gender, the latter of which is the focus of this paper.

Context of the Research and Methodology
The results presented in this article are partial results from an independently funded international research program entitled Understanding Conversion to Islam in Prison (UCIP), conducted between 2018-2020. The focus was on religious change, both within Islam and conversions to Islam, as well as the types of Islamic worldviews developed by the inmates. The research was also interested in the potential of Islam to encourage pro-social behaviour and promote the rehabilitation of prisoners. Five research questions guided the scientific approach:

1.
Who are, in socio-demographic and socio-religious terms, the Muslims who experience religious change in prison? 2.
Why do many prisoners choose to practice Islam in European prisons? 3.
What types of Islam are practiced in prison? 4.
What are the benefits and risks associated with Islam in prison? 5.
How are the processes of religious conversion and change managed by prison authorities and prison chaplaincies?

A Mixed-Methods Study
To comparatively study religious change among Muslim men and women in European prisons required the development of a mixed-methods methodology. More than the addition of quantitative and qualitative materials, our methodology was constructed in four stages. First, the development of a semi-structured interview grid and the conduct of such interviews in two pilot prisons in order to test our research questions and identify relevant quantitative variables to answer them.
Second, based on these initial results, an attitudinal questionnaire of 60 questions was developed. The attitudinal survey addressed several themes such as respondents' religious practices, the importance and role of Islam in their lives before and during incarceration, their educational and socio-demographic backgrounds, their religious worldviews, their involvement with chaplaincy and the quality of life in prison. They were translated and passed among the inmates in English, French and German. The surveys were analysed using, as appropriate, basic descriptive statistics, including frequencies, correlations and chi-squares, principal component factor analysis and linear regressions.
Third, we simultaneously conducted semi-structured interviews coupled with the larger questionnaire in eight other prisons. Semi-structured interviews were conducted in English, French and German and transcribed verbatim in the original languages (English, German or French). Each semi-structured interview was first analysed in its language of origin with a focus on the respondents' own narratives. The interviews were then subjected to inductive, deductive and axial coding via NVivo. The qualitative data (interviews, observation protocols and field notes) were triangulated with the quantitative data. Observations of Friday prayers and Islamic studies classes were particularly useful for understanding the institutional level relationships between prisoners and prison chaplains.
Finally, we observed-and sometimes participated in-religious celebrations such as Friday prayers or Eid el-Fitr (celebrating the end of Ramadan), as well as religious classes (Wilkinson et al. 2021b).

Data Collection
The research project was subject to rigorous ethical evaluation prior to data collection (see Wilkinson et al. 2021a for detailed description of methodology and ethics). The research took place in five English prisons, four Swiss prisons and one French prison in a variety of geographies, holding both sentenced and remand prisoners and covering all prison categories 1 Our research sample included all four security categories used in England and in order to ease the comparison, we adapted the equivalent security categories for use in Switzerland and France. After identifying prisons with significant Muslim populations in each of the three countries, the researchers immersed themselves in the field by spending several days or even weeks in a row in the prisons. These visits allowed the researchers to become familiar with the environment, but also to build relationships between the research team, the chaplains and staff and the inmates. In England, the data were collected by a team of three Muslim researchers: two men and one woman. The men gathered data with the male prisoners and the female researcher conducted most of the interviews and completed most of the questionnaires with the female prisoners. In Switzerland, the data with the prisoners were collected by a non-Muslim female researcher.
Prisoners were recruited through a combination of research announcements via pamphlet distribution and personal invitations on the wings, congregational prayers and religion classes. Chaplaincy teams and management helped to publicize the goals of the research and to encourage inmates and staff to participate.
The quantitative data presented in this article will address the whole quantitative sample of 279 (male = 257; female = 22). However, the qualitative data will concentrate on the interviews conducted in two prisons, HMP Parrett (UK) and La Citadelle (CH) specifically chosen for this paper because they host both male and female inmates.

Muslim Men and Women at HMP Parrett and La Citadelle
HMP Parret (UK) and La Citabelle (CH) were selected to enable meaningful comparative analysis based on gendered differences or similarities since both prisons had side-by-side male and female sections. In order to guarantee as far as possible, the confi-Religions 2021, 12, 298 5 of 17 dentiality of the data and the anonymity of the interviewees, pseudonyms have been given to the prisons as well as to the prisoners. HMP Parrett is located in the east of England. It is a level B security category prison and supervises men and women. With a total capacity of 840 places, some 360 are dedicated to women's detention. It is privately run. There is a shared multi-faith room operated by the chaplaincy unit. In 2019, data were simultaneously collected at the men's and women's sections which in terms of architecture and prison regime are a mirror image of each other.
La Citadelle is a Swiss category A prison. Initially dedicated to female detention, due to overcrowding in male prisons, it was subdivided some 15 years ago into women's (61 places) and men's (28 places) sections. It has a small ecumenical chapel also used by the Christian chaplains for individual interviews with inmates, including Muslims. Religious celebrations with more that 5 inmates take place in a classroom beside the chapel. Note that the research was conducted exclusively in the women's section.
In both countries, women constitute a small minority representing approximately 5% of the prisoners and Muslim women constitute a minority within this minority. According to the site "Women in Prison", the number of female inmates in British prison in March 2020 was 3641 people, which is 5% of the prison population (https://www.womeninprison.org. uk/research/key-facts.php, accessed on 7 June 2020). For Swiss prisons in April 2020, 401 women were incarcerated, which is 5.8% of the total prison population (https://www.bfs. admin.ch/bfs/fr/home/statistiques/criminalite-droit-penal.assetdetail.12667334.html, accessed on 7 June 2020).
Indeed, at the time of the research, HMP Parrett had 20 Muslim female prisoners (i.e., 5.5% of the female prison population). La Citadelle detained only 8 Muslim women (i.e., 13% of the female prison population). Thus, while our sample is small (n = 22), it nevertheless represents 70% and 100% of the Muslim women inmates in each of the institutions, respectively.

Socio-Demographics
The data analysed for this article consists of a sample of 279 questionnaires (257 men and 22 women) and 41 interviews (23 men and 18 women). Female prisoners represent 8% of the total sample, a percentage slightly higher that the total prison population.
Of the respondents, -71% were born Muslim. -29% who had converted to Islam; 22% had done so in prison. - The sample clustered strongly in the 24-36-year-old age bracket. -Regarding ethnicity and national origin, there are differences between the three jurisdictions. For example, prisoners in England are predominantly of Asian (Pakistani and Bangladeshi) or Caribbean origin, with an important percentage of white British converts. Prisoners in Switzerland come from three main regions: the Balkans, North Africa and sub-Saharan Africa. Finally, inmates in France are mainly with Maghreb origin. A second significant difference between the three jurisdictions is that the majority of Muslim prisoners in Switzerland are foreigners by law, whereas in England and France, if the majority are second or third generation immigrants, they have the nationality of the country of incarceration. - The sample's levels of education were also typical of a prison population with 19.3% (n = 45) having no qualifications mirrored exactly by the same number with a university degree 19.3% (n = 45). Half of the sample had either completed compulsory secondary school education (26.6%) or had done some vocational training (24.9%). - The sample's sentence length and the seriousness of their conviction ranged from lesser offences carrying a few months of sentence to high-profile terrorist offenders and sex offenders with multiple life sentences; 10% (n = 29) of the sample were on remand.

Religious Worldviews
It is relevant to articulate the types of Islam, broadly-speaking, that were lived and expressed by our sample. In order to test our participants' basic understanding and practice of Islam, we deployed a non-denominational typology of Islamic Worldviews developed by Wilkinson (2019), which we had tested previously for empirical adequacy through 15 pilot interviews and 30 surveys in one English and one Swiss prison.
Our pilot data showed that Muslim prisoners' practice and understanding mapped broadly onto three main types of Islamic worldviews: The first was mainstream Islam, which is a worldview characterized by "unity-indiversity" and the equality of all human beings before God. It refers to Muslims who regard the acceptance of differences amongst different religions as a key element of Islam.
Mainstream Islam is sub-divided into the traditional worldview and the activist worldview. Traditional Islam expresses the religious practice of those who accept and follow to their best of the ability the basic injunctions of the Qur'an and the Sunna without advocating change, whereas activist Islam describes a worldview which reflects someone's engagement in a religious process through which they hope to achieve individual and, where necessary, societal change.
The second was Islamism, which is the worldview of "Us, Muslim" vs. "Them, non-Muslim". The Islamist construes Islam as a revolutionary political ideology directed at overthrowing existing political structures and replacing them with an Islamic State governed by Sharia Law.
The third was Islamist extremism, which is Islamism as it sharpens antagonistically into a worldview of an absolutely divided, Manichean "Us" vs. "Them". Islamist extremism is sub-divided into two types: non-violent and violent Islamist extremism. The former strips non-Muslims of basic human attributes without foregrounding violence; the latter explicitly foregrounds the use of violence to rid the world of the Infidel (kuffar) and those Muslims who do not support armed struggle to establish an Islamic State.
These worldviews were assessed in terms of the aggregate mean of responses to seven variables (Questions 22 to 28 on our survey): 22 It is part of Islam to treat Muslims more fairly than non-Muslim." 23 "I avoid prisoners who are not Muslim." 24 "It is part of Islam to change things that are unfair in society." 25 "Islam teaches that wisdom can be found in many religions." 26 "Islam teaches that I must follow the law of this country." 27 "Islam teaches that the laws of this country should be replaced by Sharia Law." 28 "Islam teaches me that human life is sacred."

The Religious Worldviews of Male and Female Muslim Prisoners Were Similar
The result of our questionnaire survey showed that, contrary to a widespread idea that prisons are playgrounds for Islamists and (violent) extremists (Hamm 2009;Micheron 2020;Khosrokhavar 2016), 76% of our typical sample of Muslims in prison populated a mainstream Islamic worldview.
Moreover, as Figure 1 shows, the basic Islamic worldviews of our male and female cohorts were similar. The only noticeable difference came in the higher proportion of women (27%) than men (19%) that clustered into the Islamist worldview.
The result of our questionnaire survey showed that, contrary to a widespread idea that prisons are playgrounds for Islamists and (violent) extremists (Hamm 2009;Micheron 2020;Khosrokhavar 2016), 76% of our typical sample of Muslims in prison populated a mainstream Islamic worldview.
Moreover, as Figure 1 shows, the basic Islamic worldviews of our male and female cohorts were similar. The only noticeable difference came in the higher proportion of women (27%) than men (19%) that clustered into the Islamist worldview. This was a surprising finding since at interview the women did not tend to express the politicised Us, Muslim vs., Them, Infidel worldview characteristic of Islamism. However, when we drilled deeper into the component worldview variables, we discovered that there were significant differences between the men and women in response to the variable "Islam teaches that I must follow the law of this country", to which the women had responded significantly (X 2 (3) = 8.382, p = 0.039) more negatively than the men (see Table 1). Although the sample size of women was small, this significant difference had influenced the mean aggregate worldview score. Thus, the "Islamism" of the women appeared, in fact, to be connected with the fact that the women were less inclined than the men to link their faith with the need to obey the law. We explain possible reasons for this later in the paper in Section 6.5.

Impact of Imprisonment on the Religiosity of Male and Female Muslim Prisoners
As well as prisoners' religious worldviews, our questionnaires also tested for significant changes in intensity and type of the religiosity of our prisoner sample. These This was a surprising finding since at interview the women did not tend to express the politicised Us, Muslim vs., Them, Infidel worldview characteristic of Islamism. However, when we drilled deeper into the component worldview variables, we discovered that there were significant differences between the men and women in response to the variable "Islam teaches that I must follow the law of this country", to which the women had responded significantly (X 2 (3) = 8.382, p = 0.039) more negatively than the men (see Table 1). Although the sample size of women was small, this significant difference had influenced the mean aggregate worldview score. Thus, the "Islamism" of the women appeared, in fact, to be connected with the fact that the women were less inclined than the men to link their faith with the need to obey the law. We explain possible reasons for this later in the paper in Section 6.5.

Impact of Imprisonment on the Religiosity of Male and Female Muslim Prisoners
As well as prisoners' religious worldviews, our questionnaires also tested for significant changes in intensity and type of the religiosity of our prisoner sample. These measures were based on the analysis of different quantitative variables such as their perception of the importance of religion, the frequency of individual and collective ritual prayer and fasting during Ramadan before prison and since being in prison. Thus, we aimed to capture prisoners' institutional (Stolz et al. 2016) and individualised religiosity that was ascertained more deeply at interview.
Within the whole project, we developed a typology of religious change experienced by the participants, which drew on existing literature around extra-faith, inter-faith and intra-faith conversion and the idea of conversion as any significant change of religious worldview in terms either of type or intensity within a type. Thus, we identified five main types of conversion to or within Islam ( Intensifying: becoming significantly more devout in terms of participants "strongly agreeing" that their faith was "more important" than before prison and "strongly agreeing" that they were "praying more" in prison, i.e., a form of intra-faith conversion. This was identified though our quantitative data. -Shifting: undergoing significant changes of Islamic worldview in prison, i.e., a form of intra-faith conversion. This was identified though our qualitative data. -Reducing: those becoming significantly less devout in prison in terms of participants "strongly disagreeing" that faith was "more important" than before prison and "strongly disagreeing" that they were "praying more" in prison. This was identified though our quantitative data. -Remaining: unchanging understanding and commitment to Islam in prison. This was identified though our quantitative data.
Although many prisoners' lives exhibited more than one type of significant religious change, such as shifting and intensifying, cases whose experience of Islam in prison predominantly featured one type of change became categorised to that particular type of change.
A Gender Gap in the Prison's Impact on Religiosity: Men Were Intensifiers; Women Were Reducers As previous research on religion in prison suggested was likely (Béraud et al. 2016;Becci and Schneuwly Purdie 2012;O'Connor and Duncan 2011;Sarg and Lamine 2011), the prison experience had produced significant changes in our prisoners' religious practices and beliefs. In particular, prison had generated an intensification of the Muslim men's religiosity and a significant reduction in the Muslim women's religiosity.
As Figure 2 shows, half of the men intensified their religiosity while a third of the women did. By contrast, 7% of the men reduced their religiosity while almost a quarter (23%) of the women did. The differences between men and women were less marked regarding switching and remaining. If we drill down into the variables used to create the type of change factors, we see that: -Islam became (much) more important for 72.2% of male inmates, but for 59.1% of female prisoners; while 69.1% of the men said that they prayed a bit more or much more in prison, only 50% of the women said the same; -While 71% of the men answered that they always fasted in the month of Ramadan, If we drill down into the variables used to create the type of change factors, we see that: -Islam became (much) more important for 72.2% of male inmates, but for 59.1% of female prisoners; -while 69.1% of the men said that they prayed a bit more or much more in prison, only 50% of the women said the same; -While 71% of the men answered that they always fasted in the month of Ramadan, only 29% of the women did so; -while 3.3% of the men mentioned a diminution in the frequency of prayer and 4.1% said that they did not pray, 22% of the women acknowledged that they prayed less today than when they were free and 13% answered that they did not pray at all.

Why Did Prison Have a Different Impact on the Religiosity of Muslim Men and Women Prisoners?
Why did prison have a more intensifying impact on the religiosity of the men in the sample compared to the women? Is this difference explicable through gendered differences in the individual performance of religion between men and women? Or is institutional handling of the religious needs and provisions of Muslim men and women itself responsible? Explanations can be found at the intersection of individual and institutional levels.

Religion Was More Important to the Women before Prison
As the literature on gender and religions shows, contemporary women tend to be more religious than men (Trzebiatowska and Bruce 2012;Woodhead 2012). Our data also pointed in that direction and showed that before prison, Islam was more important for the female respondents than for the male ones. Indeed, while Islam before prison was "very important" for 50% of the women interviewed, it was only "very important" for 36.2% of the male respondents. This higher potential margin for religious intensification for men than for women is reflected in our data which showed that Islam became much more important for 42.3% of the men compared to 36.4% of the women of our sample.

Prison Was a Place of Religious Deprivation for Women
Moreover, women more frequently mentioned that prison was not an appropriate place to be religious: the lack of cleanliness, access to washing facilities and the lack of time to pray, for example, were an obstacle for them. Noma (19, Asian, Reducer, GB) said: "I don't really get the time to pray. Since I've been here, I've wanted to pray, but I haven't got time to. It's like when we clean the cells, I don't know what happens, but within seconds there is dust all over the floor and I don't want to pray on a dirty floor ( . . . ) and you haven't got anything to [perform ablution] with, so I had to keep an empty bottle to wash me".
By stark contrast, many men specifically cited the availability of time and the lack of worldly distractions as factors that facilitated their intensification in Islam (Wilkinson et al. 2021a). Saad (41, Asian Born Muslim, GB), for example, explained that prison gave him time to practice: "I had a lot of time and during my time in my cell obviously I was locked up most of the time, all I could do was just pray and read the Quran and that was just it".

The Gendered Impact of the Absence of Family
Additionally, the women interviewed made regular associations between their religious practice and their family environment. Consequently, we suggest that the separation of women from their families could partially explain the reducing influence of prison on women's religiosity. For example, Dounia (33, French, Switcher, CH) talked about the symbiotic relationship between Islam and family. During the interview, she explained that her conversion to Islam made her feel closer to her daughter who lived with her Muslim father. However, at the same time, she explained how Islam for her was synonymous with gatherings, feasts, Eid, "it is people, family, love". Lyla's comments were similarly directed. This young woman (24, mixed race, Reducer, GB) had converted outside of prison while she was in a romantic relationship with a Muslim man from whom she has been separated since she has been incarcerated. According to her, this man was not a family man: "That's when I had to end it with him because, I didn't think he was much of a family support towards his family with his mum being ill ( . . . ) he wasn't the right man for me, so I ended it with him. It wasn't working". This separation while in prison corresponded with the decline of her religious practice. By separating offenders from their families, prison prevents Muslim men and women from fulfilling some traditional gender and family roles suggested in Islam, such as the breadwinner for men and the homemaker and child-raiser for women (Siraj 2010(Siraj , 2012. Despite the stereotypical dangers of making such divisions, Siraj has shown how performing such traditional gender roles is often an element of self-identification as a Muslim man or women. As such, being a "good Muslim" for a man was connected with his ability to support and provide financially for his family, whereas being a "good Muslim" for a woman was linked to her capacity to give birth, take care of her home and children. Siraj's studies pointed out that expressions and performances of religiosity contribute to the construction of the gender identity of Muslim men and women. Our data seem to echo partially Siraj's argument. Although men did sometimes speak about their anguish over the impact of their imprisonment on various family members, it was noticeable that women often specifically spoke about the suffering they felt as a result of not being able to take care of their children. Women also regularly expressed some feelings of guilt towards their families. Carole's statement (36, British, Switcher, GB) made these kinds of feelings explicit: "It's not a nice place to be (prison), but God can forgive and forget so obviously hopefully God can forgive my sins and see some change for myself. I've got children as well who are obviously going to be affected by this. Yes, I've got seven". This denial of their ability to provide for their family or to look after their families was understood by prisoners such as Carole as a challenge to their (Islamic) masculinity or femininity and thus impacted on their religiosity. In reaction, we witnessed that many women of our sample compensated for the absence of their carer role in the family by caring for their co-prisoners, be they Muslim or not. Alma, a White convert is such an example: "And now I just feel even the girls on my wing they say that they respect me a lot because I am very calm and I'm very wanted when someone arguments with someone I want to sort out the problem for them even I didn't see this. They tell me that I've got, I'm beautiful, I'm beautiful in my heart and I'm beautiful outside, I said 'oh, this is very nice to hear this'. I always been good person." (41, White convert, GB)

The Presence and Absence of High-Status Forms of Gendered Religiosity
In compensation for this loss of aspects of gendered religiosity, some prisoners reaffirmed their gender identity in other ways. For example, both men and women prisoners reinvested in their physical bodies in order to reaffirm their gender as well as their Muslim identity. Thus, the men tended to change their physical appearance to reaffirm their masculinity. Indeed, like the majority of male inmates regardless of their religion, our research noted that forms of hyper-masculinity were expressed through the exercise of sports, in particular weight training, with even slight prisoners often developing significant upperbody musculature. It also highlighted the use of demonstrative forms of piety leading to archetypal forms of Islamic masculinity. Indeed, some Muslim male prisoners purposefully performed exaggerated forms of Muslim piety through the wearing of long robes, long beards and long hair. This "uniform" served to distinguish some prisoners from their fellow worshippers and they were often the most vocal during Qur'anic classes or associa-tion following Friday prayers. Other male prisoners displayed forms of Islamic religiosity interpreting their daily life through a religious discourse using quotes from the Qur'an, Prophetic sayings (hadith) or writings of Islamic jurists in support. Preforming their religious knowledge, they thus adopted a scholarly Islamic masculinity which contrasted with pre-prison street masculinities such as the "gangsta" where knowledge, including religious knowledge, was not valued as a resource for being "a man" (Archer 2001). A male Muslim prisoner who was a mosque orderly explained how powerful these obvious male Islamic identifiers could be: "Yeah, I was mosque orderly, and a few months back, someone approached me and said to me, 'how come there's other brothers who have more wisdom than you?' and I said 'what do you mean by that?' and they said 'well, they're always wearing the... ' They called it their dress, so obviously they don't know the name, the thobe [Islamic robe]." In contrast, for the women in our sample, despite the presence of two female Muslim prison chaplains in HMP Parrett who might have acted as religious role models, similar high status gender models were not so readily available. There was a tendency amongst the women to reify a gender category based on the appearance of their bodies as Dix-Richardson (2002) has also identified. On the one hand, they tried to continue to take care of their bodies as if they were "ouside": they bought cosmetics and perfumes and lent clothes to each other: "I like to dress well. I like to dress up too much, I like to be the beautiful." (Emna, 24, Algerian, CH); on the other hand, they also did not hesitate to oppose the prison system by means of their bodies. For example, a British Muslim female chaplain explained to us that some women, including Muslims, soiled their cells with used sanitary tampons or refused to get dressed and displayed their nudity as a means of protest.

Intersectional Disadvantage, Trauma and Involvement in Crime
Another factor that impacted negatively on women's religiosity was that their involvement in crime was usually intertwined with their own experiences of being victims and, therefore, the women were much less inclined than the men to accept responsibility for their crime. The women's stories included multiple cumulative experiences of trauma and abuse such as: parental substance use; death or loss of close relationships; homelessness; and sexual, emotional and physical abuse of various forms. These experiences were also present in the stories of the men in our study, but not to the same degree. This combination of experience of victimhood and multiple traumas highlights the experiences of intersectional disadvantage often facing Muslim women who are incarcerated. They often experience multiple forms of disadvantage throughout their lives based on a combination of their race, class, faith and gender.
This experience of trauma and its negative impact is highlighted in particular by the story of Brigitte, a Black Caribbean Muslim from Trinidad, who was 44 years old and had been convicted for drug trafficking. Brigitte's father was Hindu and her mother was Muslim; she was brought up as Muslim by her mother as her father was murdered when she was quite young. Brigitte had witnessed the murder and was the one who broke the news to her family. Her brother suffered a mental breakdown on hearing the news and Brigitte held herself responsible for this; she took on the role of caring for him at a very young age. She was also a single parent with several children from multiple partners. One of her children had died in early childhood; Brigitte was too poor at the time to afford the money for the burial and she had to hand over the body to a government morgue. These multiple traumas in her past made her vulnerable to exploitation and she was coerced into carrying drugs into England from Trinidad.
Brigitte was working as a cleaner for a large corporation; amongst their many holdings they also had a car rental dealership. One of Brigitte's friends asked her to rent a car for him in her name; the friend crashed the car and Brigitte was asked to pay for the damages by the car company. When she was unable to pay, the group kidnapped her daughter and told her that a way out of it was if she agreed to carry drugs to England for them. In the face of this pressure, she agreed to carry 1.5 kilos of cocaine for them and was arrested at the airport on her arrival in London. At the time of the interview Brigitte was close to release, at which point she was to be deported back to Trinidad; she was terrified about her release as the mafia who had coerced her would be waiting for her.
This intersection of multiple forms of disadvantage and trauma was repeated in many of the interviews with the women in prison. This also made them less inclined to accept the responsibility for their offences than Muslim men. Indeed, almost half of the Muslim women did not recognize their own responsibility in committing a crime: in their accounts, they were victims and it was self-defence, they were trapped or coerced by others, were associated with relatives or acquaintances who were the actual perpetrators or framed by the police.
Jane (19, White convert, Polish, GB), convicted for the murder of a friend, is another such example. She described how on a night out with some friends, one of the boys in her friendship group tried to rape her. To protect her, one of her girlfriends took a knife and fatally wounded her attacker. Jane had run to the police for help, but her assailant died and she was convicted for murder.
"Yeah, but it was not my fault, I tried to help that guy, even with what he was doing for me, he was trying to rape me this time. He was fighting with me like he fight some boy. Yeah. And they gave me sentence, like life, but I was trying to help the guy." Experiences of traumatic events and post-traumatic stress disorder are also associated with changes in religiosity (Ter Kuile and Ehring 2014). Indeed, religiosity prior to a traumatic event can lead to shattered assumptions and a change in beliefs. Severe posttraumatic stress is also related to decreases in religiosity. Thus, women's experiences of trauma and sense of injustice could be a factor that explains their decreased religiosity in prison.
On the contrary, in the analysed sample, most men acknowledged their responsibility for their offences. Fardeen (20, Asian, GB) explained his path of repentance in these words: "I've done sincere tawba [made repentance] for it. If Allah's forgiven me, Allah's forgiven me but we both know Allah's most forgiving. But whether or not I can be forgiven or not. But the things that pretty much bad as well, in a sense of it, and I've come back to the crime, I'm in for GBH [Grievous Bodily Harm], Section 18 with intent. I beat up a guy.
In correlation with the fact that Muslim men are more religious in prison than Muslim women, this finding suggests that religion can facilitate the acceptance of one's own crime and thus be a tool for the rehabilitation for those who mobilize their faith. While for women post-traumatic stress could inhibit similar forms of coping and rehabilitative change.

Performing a Collective Religious Identity as a Resource
Another set of hypotheses resides at the institutional level. As Dix-Richardson (2002) also has suggested, belonging to a group of Muslim women did not bring the same advantages as belonging to the group of Muslim men. Our observations noted that Muslim religiosities and masculinities were also expressed through an Islamic brotherhood: in all ten of the prisons that we researched in the wider project Muslim men gathered to pray congregationally in cells and to walk together during association. Being seen to belong to the group of Muslims reinforced both their gender identity and their religious identity.
However, Muslim women, unlike their co-religionists, represented a very small minority in prison. According to Mary (28, Albanian, Remainder, GB), most of the other inmates were Christian. An observation shared by Dounia (33,French,Switcher,CH) who thought that the imam did not come and visit the women because "there are not enough of us". Housed in different wings, Muslim women had few opportunities to meet informally and to share an experience of faith. Thus, this minority status did not allow them to create a collective identity that would provide them with affective and emotional support, nor a potentially strong identity that could enable them to respond to some of the privations experienced in prison (goods, services and security). Some Muslim women avoided non-Muslim prisoners owing to concerns of being ostracized. For example, Lyla avoided non-Muslims because she felt like: "they dislike the culture that I am. I will walk around with my headscarf on all the time and people are going to start looking at me and laughing at me ( . . . ) I don't if it's my paranoia." (24, mixed race, GB) Afia said she felt discriminated as a Black Muslim. When an officer told Afia through a Sikh woman to take off her headscarf because of health and security issues, Afia did not wear the headscarf from that moment onwards (21, Black Caribbean, GB). In other words, a moment of institutional discrimination had a lasting reducing impact of her religiosity. A same sense of discrimination was found in Maya's testimony who said: "For a start, when I came into prison, I wore the dupatta (a South Asian traditional dress), I kept the dupatta on, at least a scarf round my neck, or I kept the dress. And I have now had to change my style of dress because I don't want to be judged or be beaten up or treated differently. ( . . . ) They've (the prison) changed me that I can't identify in my own religion, because I've stopped wearing the stuff. It's here, underneath my bed. I have the dupatta, I have the clothes, I stopped wearing them because I want to try to fit, which is bad really. Because they take away, they've taken away my identity and stripped me from my identity and stripped me from who I am because I'm not accepted anywhere." (Maya 60, Asian, Intensifier, GB) These various statements seem to confirm the hypothesis that in a minority environment, the visible performance of Islam is less possible and less meaningful for female Muslim prisoners due to micro moments of discrimination of an intersectional nature, in this case of gender, ethnicity and religion.

The (in)Adequacy of Chaplaincy Activities
As Beckford (2003) suggests, prison can be a place of religious socialisation in which chaplaincy can play a pivotal role. Our data underlined a discrepancy between the religious provision for Muslim men and women prisoners. According to the Muslim women interviewed, their spiritual needs were insufficiently catered for: more than 80 percent wanted more religious classes, 60 percent wanted more collective prayers and more than 70 percent wanted more visits from the chaplain. Although Maya, for example, (60, Asian, GB) had intensified her prayers, she was really critical of chaplaincy at HMP Parrett. When asked if having a religion was helping her, she answered: "I thought having my own religion would mean that I could rely on the chaplaincy and on the chaplaincy leaders. But they are rubbish in every sense, they are nonsense. They actually have too much work to do, too much pressure, they're too overloaded with too much pressure that they actually cannot be bothered. They don't give you any one-to-one, they don't treat you as if you matter. It's like you can just fumble your way through where you're not really getting the help but you want to be able to concentrate on your own religion because that's where you think you'll get the help but you're not getting it." The experience of Islam can be highly mediated through prison chaplaincy. These unfulfilled needs were indicative of great disparities not only between Swiss and British prison chaplaincy provision, but also between the chaplaincy provisions for Muslim women and men within each country (Wilkinson et al. 2021a). Indeed, the chaplaincy team in HMP Parrett consisted of eight people, four being Muslim (two women and two men, all part-time). While any Muslim chaplain could pastorally accompany both men and women prisoners, the teaching of religion and the practice of prayer were gendered. There were important differences between the services provided for men and for women: for example, there was no congregational Friday prayer for women in HMP Parrett. As well as the fact that attendance at the Friday Prayer in Islam is not obligatory for women as it is for men, debates amongst the women led by led in particular by a vocal terrorist offender about to the legality of a woman leading Friday prayer had prevented the chaplain for providing this congregational activity for women, so the women never prayed together on Friday. The chaplain, although she was pastorally accomplished and respected, did not feel that she had the religious education to challenge the TACT offender on theological grounds. However, female Muslim chaplains did organise prayer times with women on other occasions, such as after a religious class.
In La Citadelle, pastoral care was given by the Catholic or the Protestant chaplains to all inmates, irrespective of their religion. The Muslim chaplaincy was almost non-existent. In the male wing of La Citadelle, an imam visited on Fridays to lead the prayer twice a month. In the female wing of La Citadelle at the time of the research, the same imam had suggested visiting the women once a month after male congregational prayers. However, due to a lack of time or miscommunication between him and the prison, it seems that this opportunity for interaction with the women did not take place often. Indeed, (Camilla 43, Franco-Moroccan) said that she had registered for the imam's last visit but that the imam, "saw the men and I think he didn't have time for us afterwards". This was highly indicative of the notion observed in previous research of women in prison as an "after-thought".

The Need for Gender-Responsive Chaplaincy
As well as the institutional disparity in the quantity and regularity of chaplaincy provision for Muslim men and women, one might ask whether the chaplaincy provision for women prisoners effectively meets their religious needs or whether additional spiritual and faith-based activities should be developed in order to fulfil women's specific demands, particularly their need to cope with traumas from their past. In this regard, we encountered female prisoners in particular who had developed personalised forms of religious expression to cope emotionally with feelings of guilt and depression. For example, Emna (Emna, 24, Algerian, born Muslim, CH) explained that she supplemented her five daily prayers with spiritual dancing at night which she found a powerful antidote for depression: "Yes I do the prayer, I pray five times a day. But at night, I dance, sorry to say that, but I dance, because it empties my head and keep me from falling in depression. So, I dance. I put on some music, and I dance. I don't care what they say, I pray, and I dance, and I ask God for forgiveness. Nobody but the wardens can see me, so I don't care, and I dance".
Similarly, Brigitte (41, Black Caribbean, convert, GB) found participating in the prison choir a therapeutic way of coming close to God and coping with suicidal thoughts. "Then they asked me to join choir, so I joined the choir because I love to sing. And within this chance, it made me feel free, it made me feel, you know, singing, even though, it don't matter how I sing. You know, singing songs and praising Allah, even though it's not in Muslim, but singing songs made me feel Allah. Because I wanted to kill myself in the cell, last year." Adding choir to her religious practice as a Muslim was causing problems for Brigitte in prison: the choir was a registered Christian activity and she could only attend services for the religion for which she was formally registered. As a Muslim, she could not attend Qur'an classes and the choir as well. As the prison had forced her to make an exclusive choice, she had picked Islam-only activities, but was sad about being left out of the choir. These examples show how institutional forms of religion and personal expressions of faith exist side-by-side in conditions of incarceration and display syncretism and gendered intersectionality. Brigitte's experience in particular shows how the siloed nature of prison chaplaincy can sometimes deny prisoners nurturing inter-faith experiences which impacts negatively on their religiosity and their ability to process trauma and come to terms with their sentence.

Conclusions
In this paper, we have showed that prison impacts differently on the religiosity of Muslim male and female prisoners. Although our male and female prisoners possessed a similar basic apprehension of what constitutes Islam as a worldview, the religiosity of Muslim men tended to intensify in prison; the religiosity of Muslim women tended to reduce. Muslim men tended to highlight their brotherhood in Islam by praying regularly in congregation, attending religious classes and Friday prayers. Muslim women tended to reduce their religious practice, care for co-prisoners independently of their religion, minimize their Muslim appearance and look for inter-faith or personalised religious experiences.
Moreover, whereas Muslim men tended to develop the type of active religiosity which helped them to accept responsibility for their crime and to aspire towards personal change based on values of their faith; most of the Muslim women had difficulties in accepting responsibility for their crime due to the entanglement of their crime(s) with traumatic experiences of victimhood in the past and therefore did not connect their faith so much with obeying the law and the avoidance of crime.
Our data allowed us to raise possible explanations for these differences, which all merit further research. At the individual level, the presence in prison of meaningful forms of Islamic masculinity, such as "the scholar" for Muslim men and the absence of high-status female ones affected men's religiosity in an intensifying way and women's religiosity in a reducing way.
Both genders suffered from the absence of family, but the absence of family-based roles and relationships in the closed prison environment affected women's religiosity more negatively than men's religiosity.
At the level of the prison institution, the religiosity of women was affected by their minority status in prison which meant that they possessed less influence as a collective and that they were more likely to experience overt discrimination from staff and other prisoners. Furthermore, the provision of chaplaincy services for them was weaker and less aligned with their specific needs than it was for men.
The differences in the experience of religiosity of men and women prisoners and the apparently gendered inequality or inadequacy of religious provision for women in prison merit further research. A finer identification of the similar and different religious needs of Muslim men and women through further research could bring constructive reflections on the implementation of chaplaincy provision adequate to specific religiosities of men and women.