Change and variation in family religious language policy in a West African Muslim community

This article examines variation in family religious language policy in a Muslim community in West Africa. Taking an ethnographically grounded case study approach, I situate families’ choices with regards to their children’s religious (language) education within the larger linguistic, social, and cultural context, focusing on new influences on Islamic education and Arabic learning in the community and the wider region during a period of Islamic resurgence. I examine families’ religious education choices with particular attention to what they mean for children’s Arabic learning and what they say about families’ orientations toward Arabic. This article is based on a language socialization study of seven young Fulbe children into three languages (Fulfulde, Arabic, and French) at home, Qur’anic school, and public school in northern Cameroon. Changes in the religious, linguistic, and educational landscape of the region are reflected in the variation within and among four focal families with respect to how they organized and envisioned their children’s religious study and Arabic language learning. In presenting ethnographic research conducted at the intersection of religious language policy, language education policy, and family language policy, my goal is to further our understanding of how private choices are connected to forces in the public sphere, giving rise to changes in the ways faith, language, and learning are practiced and conceptualized.


Religion, family, and language policy
In this article, I examine variation in family religious language policy in a Muslim community in West Africa, taking an ethnographically grounded case study approach. This analysis is located at the intersection of religious language policy (RLP), language education policy (LEP), and family language policy (FLP), a space that has received limited analytic attention from language policy researchers. RLP is an understudied area to begin with, and research in this domain has tended to emphasize public domains and larger-scale policies (Spolsky 2009). LEP researchers have frequently focused on public and secular schooling (Tollefson 2002), with only a very few scholars working in religious educational settings (e.g., Avni 2012;Chew 2013). The growing field of FLP research has been concerned primarily with language minority groups who are making choices concerning heritage languages and the languages dominant in the wider society (Caldas 2012;Curdt-Christiansen 2013;King et al. 2008). Avni (2012) observes that ethnographic research has yielded important insights into how language educational policies are experienced, negotiated, and shaped by historical and ideological forces. Here I examine the religious education choices of four Fulbe families in northern Cameroon, giving particular attention to what these choices mean for children's Arabic learning and what they say about families' orientations toward Arabic. The families' policies with regards to their children's religious (language) education are situated within the larger linguistic, social, and cultural context, focusing on changes in Islamic education and Arabic learning in the community and the wider region during a period of Islamic resurgence. The case studies illustrate how changes in the religious, linguistic, and educational landscape of the region are reflected in the variation within and among four focal families with respect to how they organized and envisioned their children's religious study and Arabic language learning. In presenting ethnographic research conducted at the intersection of religious language policy, language education policy, and family language policy, my goal is to further our understanding of how private choices are connected to forces in the public sphere, giving rise to changes in the ways faith, language, and learning are practiced and conceptualized.
Arabic learning and religious learning have long been deeply intertwined in the Fulbe community in northern Cameroon, which has a long tradition of Qur'anic schooling. Over the past two decades, a new model of Islamic education has gained currency in the region, and this model entails language policies that differ from those that have long been predominant in Fulbe society. These new policies are part of wider changes in northern Cameroon, where new Islamic resurgence movements have been active since the 1990s (Adama 2004; van Santen 2014). There is increasing access to new sources of information about Islam from core Islamic countries (satellite TV, DVDs, print materials, and preachers), making the population less dependent on local religious leaders for Islamic knowledge (van Santen 2014;Seignobos and Nassourou 2000). Private Islamic schools have been established in the region, funded at least in part by Islamic non-governmental organizations based in Arab states (Seignobos and Nassourou 2000).
New media and schools explicitly promoted an Islam free of ''ethnic stains,'' that is, of local practices and beliefs said to be un-Islamic ( van Santen 2012). They also promoted what Liddicoat (2012) refers to as a comprehensibility orientation to Arabic. This orientation emphasizes the use of religious language for the communication of propositional content, between text and reader(s) and/or between speaker and listener(s). Liddicoat contrasts this with what he calls a sacrality orientation, in which emphasis is placed on the reverent and mystical aspects of language use (p. 122). While these two orientations are fundamentally different, Liddicoat makes the point that they are not mutually exclusive and that religious language policy may reflect both orientations.
The Fulbe Islamic tradition has always included both orientations toward Arabic. However, comprehension of Arabic has been the domain of a small scholarly class, while the vast majority of Fulbe learned to reproduce and respond to Arabic texts with reverence and respect but without comprehension of the literal meaning (Santerre 1973;Seignobos and Nassourou 2000). This pattern of Arabic learning can be found in non-Arabophone Muslim communities around the world (Brenner 2001;Spolsky 2003;Boyle 2006). The Islamic resurgence movements also included both orientations toward Arabic, but comprehensibility was promoted as being important for every Muslim. Promotion of the comprehensibility orientation and promotion of Islam without ''ethnic stains'' were explicitly linked: Arabic comprehension gave a Muslim direct access to the Qur'an, the worldwide Muslim community (ummah), and information about Islam from the core Islamic countries, where ''authentic'' Islam was practiced (van Santen 2012). The comprehensibility orientation toward Arabic constituted a challenge to local Islamic beliefs and practices, within which Arabic was a sacred and unequally distributed resource.

The study
This article is based on fieldwork I conducted in a Fulbe community in northern Cameroon in the late 1990s and early 2000s and on my continuing engagement with research participants and other scholars working in the region. Employing language socialization as the central theoretical and methodological orientation (Ochs and Schieffelin 2012), I documented the apprenticeship of seven Fulbe children in the languages of their community: Fulfulde, the primary language of the Fulbe; Arabic, the language of religious schooling and practice; and French, the language of public schooling.
The seven focal children, four boys and three girls, came from four families that lived in six separate households. All of the children were 6 years old at the time the study began, and all attended both public school and Qur'anic school. The households varied in socioeconomic status, family composition, economic activities, and religious and secular educational levels. I worked closely with six Qur'anic educators, three women and three men, who varied in age and levels of Qur'anic and public schooling. All had recited, read, and written the entire Qur'an, and all but one had pursued advanced Qur'anic studies with scholars in Maroua, as was the case for the majority of teachers in Maroua at the time (Seignobos and Nassourou 2000). Two of the teachers had some public schooling and spoke some French with me. Those who had children had sent their sons to public school for at least a few years.
The research combined a video-based longitudinal multiple-case study design with ethnographic study of the children's families, schools, and community. I engaged in (participant) observation, documented the children's writings and drawing, and video recorded naturally-occurring interactions across home and school settings (a total of 90 recorded hours). Interviews and video playback sessions were conducted with family members, public school and Qur'anic school teachers, Islamic scholars, and public education officials. I collected secular and Islamic educational materials that were locally available, including textbooks, teachers' guides, booklets on Islam, and audio-recorded sermons. All video recordings of naturally-occurring interaction and audio recordings of interviews and playback session were transcribed.
In this study of family religious language policy, I sought to answer three questions about the focal families: (1) What choices did they make with respect to their children's religious education? (2) How did these choices shape the children's Arabic language learning experiences? and (3) What did these choices indicate about the families' orientations toward Arabic and their beliefs about Islamic education? My analysis of interviews, naturally-occurring interactions, field notes, and locally collected documents were situated within a broader study of the community in order to understand how families' choices and beliefs related to (shifts in) religious, linguistic, and educational practice and ideology in the community and the wider region. In this I benefitted from the extensive research conducted by other scholars on the Fulbe and their language, northern Cameroon, and public and Qur'anic schooling in the region and beyond.

Arabic and the Maroua Fulbe
Family language policy cannot be understood without paying attention to the language ecology within which families live, and to the meanings and functions of languages in the wider community. The following discussion focuses on the roles and meaning of Arabic in the Maroua Fulbe community, the teaching and learning of Arabic, and the relationship of Arabic to the other languages in the community's repertoire, Fulfulde and French. In Maroua and the wider region, oral proficiency in Fulfulde is a marker of Fulbe and Muslim identities and is essential for participation in the local economy. French is valued for the employment opportunities and the protection from exploitation it can provide, but French is also suspect for many Fulbe because it is associated with ethnic and religious outsiders and a loss of traditional values. Arabic is central to Islamic practice and religious education, and skill in the recitation and transcription of Qur'anic texts is seen as both sign of and means for developing religious knowledge and commitment.
Like Muslims around the world, the Fulbe believe that Arabic is the language of God and that the Qur'an is the word of God, revealed in Arabic to the Prophet Mohammad by the archangel Gabriel (Reichmuth 2009;Santerre 1973;Bassiouney 2009). For the vast majority of Maroua Fulbe, Arabic is not a language for mundane communication, but rather a sacred language that permeates daily life. The recitation of Qur'anic texts is central to Islamic devotional practices and celebrations, including the five daily prayers, observance of Ramadan, and marriage and naming ceremonies. Sermons are built around Classical Arabic texts, with imams speaking solely in Arabic or weaving Arabic into predominantly Fulfulde homilies (Kouega and Baimada 2012;Moore 2013). Qur'anic texts in various forms are central to traditional healing practices (Santerre 1973;cf., Perrino 2002).
Qur'anic texts are believed by Fulbe to have the power to protect, heal, and spiritually transform people, Muslim and non-Muslim alike (Moore 2008). Sacred texts may be translated into Fulfulde but must be in Classical Arabic to have power and ritual validity, what Bassiouney (2009) calls the traditional Muslim position on translation of the Qur'an. While Santerre (1973) describes the attitude of Fulbe in Northern Cameroon toward Arabic as one of veneration; this veneration applies only to the Classical Arabic of the Qur'an and the hadiths, reports of the sayings and deeds of the Prophet Mohammed. Arab Shuwa, a dialect of Arabic spoken by an ethnic group of the same name and used as a local trade language in the northernmost corner of Cameroon, was believed by Fulfulde to lack the sacred and mystical properties of the language of the Qur'an (Echu and Aminou 2009). 1 Modern Standard Arabic was esteemed by many Fulbe but nevertheless was regarded as lesser than Qur'anic Arabic (Echu and Aminou 2009).
The acquisition of Arabic competence has long been inseparable from the acquisition of religious knowledge in the Maroua Fulbe community, as in most non-Arabophone Muslim communities (Seignobos and Nassourou 2000;Brenner 2001). Elementary Qur'anic schooling is dedicated to the recitation, reading, transcription, and memorization of the Qur'an without explanation of its propositional content (Adama 2004), a learning tradition widely referred to as rote learning and which I have reframed elsewhere as guided repetition (Rogoff et al. 2007). Through participation in Qur'anic schooling, most Fulbe develop some reproductive competence in Arabicthat is, fluency and accuracy in reproducing the oral and written Arabic language forms of the Qur'an without comprehension or the ability to generate new, meaningful utterances (Moore 2006). Instruction was individualized, with each child progressing through the curriculum at his or her own pace. Children learned to reproduce Qur'anic texts reverently and accurately, and this process was believed to foster self-control, good moral character, and submission to God's Word (Adama 2004;Moore 2008).
The process of acquiring Qur'anic textual knowledge was seen as transformative and fundamental to becoming a Muslim, but learning to comprehend Arabic texts was not (Moore 2008). Reproductive competence in Arabic was believed to be sufficient for a Muslim to perform his religious obligations and develop proper religious feeling (cf. Brenner 2001). Moreover, it was believed to be the knowledge that was the foundation for all further learning (Denny 1998;Brenner 2001). One could study the literal meaning of the sacred text only after transcribing, reading, and reciting the entire Qur'an, which the Fulbe refer to as ''finishing the Book'' (in Fulfulde timmugu Deftere).
Within the traditional Fulbe Qur'anic education system, knowledge of Arabic was conceptualized as sacred knowledge that could be acquired only through years of apprenticeship with more advanced scholars (Brenner 2001;Santerre 1973). Arabic and Fulfulde were deeply intertwined in this process, as they were in religious observance for the Fulbe and the wider Muslim community in the region. While Arabic is the language of Islam, Fulbe believe that Fulfulde is also a language of Islam (Barreteau and Dieu 2000). Both elementary and advanced instruction are done in Fulfulde, as are sermons in mosques and on the radio (Kouega and Baimada 2012;Tourneux and Iyébi-Mandjek 1994). The Fulbe of northern Cameroon have not adopted a written Fulfulde translation of the Qur'an, but there is a long tradition of Islamic scholars writing in Fulfulde in Arabic script (ajamiiya) and using interpretive works written in Fulfulde (Seignobos and Nassourou 2000).
There has always been variation in how far Maroua Fulbe go in Qur'anic study and the development of Arabic competence. In the recent past, most children attended school only long enough to memorize the few short suras needed for prayer, and the great majority of teachers did not understand Arabic (Santerre 1973;Adama 2004). A small minority finished the Qur'an, and some went on to pursue advanced study and comprehend Arabic texts. A very small elite group, mostly religious professionals (e.g., judges, imams), was conversationally competent in Arabic (Seignobos and Nassourou 2000). Comprehensive measures of Arabic literacy rates in the region are not available; in northern Cameroon, as in most Francophone African countries, ''literacy statistics are concerned only with people literate in French'' (Ouane and Amon-Tanoh 1990).
Whereas Fulfulde and Arabic have long been part of the linguistic repertoire of Maroua Fulbe, only since the late 1980s have a majority of Fulbe come to view French as relevant and useful (Tourneux and Iyébi-Mandjek 1994). In the Far North Region, French literacy rates are 24 % (UNESCO 2009), up from 11 % in the mid 1980s (Iyébi-Mandjek 2000). French has long been seen as the language of modernity and secular schooling, and many Fulbe see it as the language of colonizers past (whites) and present (Christian Cameroonians from the south of the country) (Barreteau and Dieu 2000). French was regarded by some as an un-Islamic language, unsuitable for use in religious educational contexts. Because French was seen as the language of the secular world outside the household and public schooling was widely believed to interfere with the social, moral, and spiritual development of children, some Fulbe argued that Fulbe girls, in particular, did not need to learn the language or attend public school (Moore 2006;cf., Regis 2003). Nonetheless, by the late 1990s, rates of participation in French-language public schooling were on the rise among Fulbe in the region (Iyébi-Mandjek 2000).

Changes in religious education and Arabic learning
During the same period (late 1980s through 1990s), patterns in religious education and Arabic learning in Maroua and the wider region were also changing (Seignobos and Nassourou 2000). Participation in Qur'anic schooling rose dramatically, as did the number of schools. Girls' rates of participation rose, and more children of both genders pursued their Qur'anic studies longer and further than had been typical in past generations. Comparison of Qur'anic teachers' self-reports of Arabic competence in the late 1960s (Santerre 1973) and the late 1990s (Seignobos and Nassourou 2000) suggests that the ability to understand the meaning of Arabic texts-or at least the belief that such ability might be socially desirable-had increased significantly among teachers. Other forms of and pathways to Arabic competence became increasingly accessible. Arab state donors had been promoting Arabic in Sub-Saharan Africa since the Middle East oil boom in the 1970s (Brenner 2001), but in the 1990s northern Cameroon saw a rapid rise in the availability of Arabic language materials and Arabic-language radio and television programming (subsidized by donors) and increased donor support for public and private schools to provide Arabic language instruction (Seignobos and Nassourou 2000).
Many of the private schools were established by Fulbe men who had pursued advanced studies in the Arab states on scholarships from donor nations (Seignobos and Nassourou 2000). Returning home fluent in both Classical Arabic and a modern dialect, they brought new ideas about Arabic education and placed more emphasis much earlier on Arabic comprehension than was traditionally the case (Barreteau and Dieu 2000;Iyébi-Mandjek 2000). Changes they instituted were widely referred to as ''modernization'' (Adama 2004) and included collective Qur'anic instruction in French, a public school-like curriculum, and instruction in Arabic as a second/foreign language from the early grades. Another change was the introduction of standard enrollment fees, which were much higher than the amounts parents typically paid to Qur'anic school teachers to cover the costs of artificial light used during nighttime sessions (Seignobos and Nassourou 2000).
Modernization was controversial, and many Qur'anic teachers spoke out against it. The ''new'' practices introduced in private Islamic primary and secondary schools had long been practiced in the government-run Franco-Arabic schools, which were established in the early twentieth century as part of a wider effort to promote French and contain Arabic, which was associated with Muslim unity and anti-colonial movements such as Pan-Arabism, Mahdism, and Wahhabism (Seignobos and Nassourou 2000). Founders of the new schools argued that the new educational model was Islamic, imported from core Islamic countries. Nonetheless, some Fulbe still found it too similar to janngirde nasaara ('school of the white', or public school) and thus suspect. 2 Most families who sent their children to the new private Islamic primary schools (instead of public schools) also sent them to traditional Qur'anic schools (Moore 2006).
Fulbe participation in public French-language schooling had been low until the mid 1980s, when Fulbe authorities began to encourage attendance after Cameroon's Fulbe president was succeeded by a Christian southerner who ended preferential treatment of Fulbe (Iyébi-Mandjek 2000). Rates of double schooling-participation in both Qur'anic school and public school-rose significantly in the 1990s, and most Qur'anic schools adapted their schedules to accommodate public schooling (Iyébi-Mandjek 2000). However, many Qur'anic teachers still objected to French-language schooling as a dangerous distraction from religious study and socialization into traditional roles and values, particularly for girls (Tourneux and Iyébi-Mandjek 1994). Fulbe girls were far less likely than boys to be sent to public school, but their participation increased too (Seignobos 2000).
Some teachers in ''traditional'' Qur'anic schools were selectively adopting and adapting practices from ''modern'' schools (Moore 2011). While the teaching and learning of Qur'anic texts continued to be the primary activity, some teachers expanded the curriculum to include collective instruction on prayer and the basic tenets of Islam. Other innovations were the teaching of songs and short weekly sermons in which Arabic texts were glossed with Fulfulde and explicitly linked to the children's daily lives (Moore 2013). Several teachers and parents explained to me that these new practices made Qur'anic schooling more appealing and motivating for children, particularly those who participated in public schooling. Thus, while most Maroua Fulbe children did not attend ''modern'' Islamic schools, many attended Qur'anic schools that included ''modern'' elements, including new practices that provided children with information about the propositional content of Arabic texts.

Religious language policies in the focal families
The majority of Fulbe children in Maroua and the region participated in the traditional Qur'anic schooling system rather than the modernized system, as is still the case today. However, variation among these schools was increasing at the time of my fieldwork, and families could choose from and/or combine different forms of Islamic education. The focal families all sent their children to public school and traditional Qur'anic school, and the focal children all started attending Qur'anic school at least a year before they entered public school. While similar in that they all chose double schooling, the families varied in their choices concerning their children's religious (language) education, and this variation makes visible the influence of broader changes in the community and the region with respect to Islamic education, Arabic knowledge, and participation in Qur'anic and public schooling.

Umaru
In terms of household composition and economic activities, Umaru's family was the most traditional focal family: his household was multigenerational; all the men were polygynous; the family kept cows and sheep; and the women of the household sold milk and cotton blankets they wove. Umaru's father had no public schooling but had learned some French in his work. Umaru's mother had completed the second grade and spoke a little French. All of their children who were of school age attended public school.
At the start of my study, Umaru had been studying the Qur'an for almost a year at home. His teacher was his great-aunt Asta, a widow in her sixties who had completed the elementary cycle of Qur'anic school. She began teaching the Qur'an to the children of the household about 10 years before my study began and sometimes taught children from the neighborhood as well. Mal Asta ('Mal' is an abbreviation of mallum, 'teacher', and an honorific used for people who teach the Qur'an) became the family teacher because she had the most free time and was very glad to perform what she regarded as her religious duty.
Such in-home instruction by a family member with few other students was widely regarded as ideal because the child would be more closely supervised by the teacher and would thus make more rapid and correct progress in the learning of the Qur'an (Santerre 1973). Mal Asta's instruction focused exclusively on the accurate reproduction of Qur'anic texts, and she was a strict disciplinarian. She directed children's study with a steady flow of corrections, reproaches, and exhortations to work harder, using a long stick to point on children's texts and to hit children when they were not reciting with sufficient volume and/or accuracy.
Umaru made rapid progress, and the family fully expected him to finish his Qur'an within a couple of years, after which he would re-read the Qur'an several times over the course of his lifetime, committing large portions to memory, as several other members of the family had done.

Hajja, Faariiku, and Adiilu
Hajja, Faariiku, and Adiilu were close relatives and neighbors and attended school together. Hajja's father was a well-to-do local dignitary who had finished the Qur'an and several years of public schooling. Her mother was a woman of means in her own right and also highly educated, having finished the Qur'an and completed the first 3 years of secondary school. Faariku and Adiilu were brothers. Their father had attended Qur'anic school for a few years. He had not attended public school but learned to speak some French in the course of his career as a businessman. The boys' mother had attended Qur'anic school for a few years.
All three children attended a Qur'anic school at the end of their street, where about 10 other children studied. The teachers, mainly retired Mal Didja and her son Mal Buuba, had both finished the Qur'an and pursued some advanced study, specializing in tajwiid, the correct recitation of the Qur'an. Mal Buuba had 3 years of public schooling and sent his sons to public school. When trade kept him from being present during Qur'anic school sessions, his 12-year-old son was put in charge.
The parents of Hajja, Faariiku, and Adiilu sent their children to Mal Buuba's school because it was nearby, because Mal Buuba was of the same patrilineage, and because he was known to be calm and gentle with his students, rarely using corporal punishment. They all said that they wanted their children to be happy to go to Qur'anic school, so they did not want the traditional strict discipline. They also liked that Mal Buuba incorporated short sermons and choral singing in Arabic and Fulfulde.
Faariiku and Adiilu made slow progress in their studies, while Hajja did well, in part because her mother Hawa supervised additional Qur'anic study at home. Hawa wanted Hajja to finish the Qur'an and to go far in public schooling. Hawa saw the two types of schooling as ''balancing'' one another and equipping a girl to be a good Change and variation in family religious language policy in a… 133 wife and mother, able to manage her household well and to support her children in their own studies. Many Fulbe saw public schooling as having a negative effect on moral development, Hawa explained, so a high level of public schooling was a liability for a girl on the marriage market if not balanced by a high level of Qur'anic schooling.

Muuniira and Ladiifa
Muniira and her older sister lived with their paternal aunt Indu, very near their father and several other relatives. Indu was a widow in her early 60s, a businesswoman with agricultural land in the family's ancestral village. She had attended Qur'anic school for a few years. Indu never attended public school and told me that she sent the girls only because their father insisted. Nearby lived Muniira's female cousin Ladiifa. Ladiifa's father was a businessman who lived most of the year in Gabon. He had finished the Qur'an and almost completed secondary school. He had performed the Haj a few years before and continued his religious studies in Gabon. Ladiifa's mother had been tutored at home for a few years in French and mathematics. She had finished the Qur'an and continued her religious studies by reading the Qur'an and attending sermons for women, a phenomenon that was new in Maroua at that time ( van Santen 2014). She regularly supervised her children's Qur'anic study at home.
Ladiifa's middle brother attended the Franco-Arabic primary school, where he was learning Arabic as a second/foreign language. Both her brothers attended a large and highly regarded Qur'anic school some distance from the family compound. Ladiifa and Muniira attended a smaller school very close to home that was led by Mal Fanta, with about 35 students. Mal Fanta, a woman in her forties, had studied tajwid (proper recitation) and tafsir (exegesis) and continued to study with a well-known scholar. Her sermons for women in the neighborhood were highly lauded. All of her children had pursued advanced Qur'anic study, and her two sons teenaged sons were her teaching assistants.
Ladiifa's mother chose Mal Fanta's school for Ladiifa because Mal Fanta was a respected woman scholar who gave both elementary and advanced instruction. This meant that once Ladiifa finished the Qur'an and began advanced studies, there would be no need to find Ladiifa a new teacher once she entered puberty, when most girls stopped studying with unrelated males. Ladiifa's mother liked that there were several older girls studying with Mal Fanta because this provided Ladiifa with female role models of high achievement and opportunities to observe instruction in the meaning of Qur'anic texts. Ladiifa's mother regularly attended Mal Fanta's sermons, and she had begun to bring her daughter with her.
Indu had moved Muuniira to Mal Fanta's school from another where she had been making poor progress. Ladiifa's mother had encouraged this move, arguing that Muuniira would do better under the more scholarly Mal Fanta. Indu told me that she made the move so that Muuniira would be in a school with a close family member, her older sister having recently left the other Qur'anic school. Indu did not expect Muuniira to go far in her Qur'anic studies, but a couple more years would increase her marriageability.
Like the mother of Hajja, Ladiifa's mother wanted her daughter to do well in both religious and secular education and saw both types of schooling as important to making Ladiifa attractive as a wife for a Fulbe man of standing and means. Ladiifa's mother saw religious education as much more important, however, and expected Ladiifa to pursue religious study further than secular education. She hoped that Ladiifa would learn to comprehend Arabic, as reflected in her decision to send Ladiifa to a teacher with a high level of Arabic comprehension and her second son to the Franco-Arabic school.

Maanu
Maanu's family lived in a compound occupied by two brothers and their mothers, wives, and children. Maanu's father worked in a store. He had completed 3 years of Qur'anic and public schooling and had learned French during his years as a professional athlete in the south of Cameroon. Maanu's mother had completed primary school and finished the Qur'an. She read the Qur'an regularly, attended sermons in the neighborhood, and was helping a co-resident sister-in-law to learn to read the Qur'an. The couple had five children, and all of the school-age children attended public school and Qur'anic school.
Maanu's parents had first enrolled him in the same Qur'anic school as his older sisters, but soon thereafter moved him to Mal Jibiri's school, the largest and most ''modernized'' in my study. Mal Jibiri had taken over the Qur'anic school of a retired teacher and incorporated several new practices, including long sermons twice a week, collective teaching of hadiths using a blackboard, and leading and supervision of his students in afternoon prayer. He was the only teacher in my study who taught students to read the Qur'an not only in the Hafs tradition of recitation, the most common in West Africa, but also Warash, the most common in the Maghreb and locally described as reading ''like the Arabs'' (bana aaraabo'en). With over 100 students, he had several teaching assistants, teenaged boys and girls who supervised other students when he was working with individual students or small groups. A man in his thirties, Mal Jibiri had completed 4 years of public school and attended a madrassa, another new kind of school in Maroua in which religion, mathematics, and Arabic were taught via collective instruction.
Maanu's parents were not pleased with the progress he was making at Mal Jibiri's school, but they kept him there because they appreciated the expanded curriculum, in particular the training in prayer and the explicit moral instruction and explanation of Arabic texts provided in the sermons. To make up for what they felt was a lack of individual attention from the teacher, Maanu's mother supervised his additional study at home.

Change and variation in religious language policy
Within and across the families we find variation in how children's religious education was organized and how their Arabic learning trajectories were envisioned. This variation makes visible the influence of broader changes in the community and the region with respect to Islamic education, norms of Arabic learning, and participation in Qur'anic and public schooling. The focal children in this study all attended ''traditional'' schools, but these schools were not homogeneous. All but one had modified the traditional Fulbe model of Qur'anic schooling in diverse ways, and families were sensitive to these modifications in making their decisions about the kind of religious (language) education they wanted for their children and how to support their children's learning at home.
Umaru's family adhered most closely to the Fulbe tradition of Qur'anic schooling and its strong sacrality orientation toward Arabic. His religious education was devoted entirely to the disciplined and accurate reproduction of the sacred text, and comprehension of Arabic was not a goal. The family of Hajja, Faariiku, and Adiilu illustrates a slight shift away from this model. Their parents chose a teacher who was gentle with children and incorporated a few ''modern'' elements that they believed made Qur'anic schooling more enjoyable and informative for their children than an exclusive focus on recitation and transcription. Ladiifa's and Maanu's families show a shift toward a comprehensibility orientation, the parents deliberately choosing Qur'anic schools that provided opportunities for the children to hear Arabic texts explained in Fulfulde.
The focal families were part of the growing number of Fulbe who participated in French-language public schooling, and all the parents expressed the belief that Qur'anic study was an important counterbalance to public schooling (cf., Chew 2013) and Arabic a counterbalance to French. For children in this community, the new forms of Arabic competence were valued as indicators of piety and an Islamic version of modernity (as opposed to a Western/Christian version). This was particularly salient for parents of daughters because Western-style education of girls was considered more problematic than for boys. The mothers of Hajja and Ladiifa, both of whom envisioned high levels of religious (language) learning for their daughters, were explicit that this was important for a girl's marriageability, particularly if she went far in French-language schooling. Both mothers supplemented their daughters' religious education by supervising their study of the Qur'an at home, a growing trend in the community (Adama and Amadou 1998), and they believed it was important to prepare their daughter to do the same once they had children of their own.
In interviews, none of the focal families identified themselves as participants in ''modern'' Islamic schooling, and adult family members expressed ambivalence about such schools. Several objected to collective Qur'anic instruction, pointing out that the teacher could not assure the mastery of each text by every child before the next text was introduced. Another concern was that teaching Arabic ''just like any other language'' (French, in particular) desacralized the language of the Qur'an (Moore 2006(Moore , 2011. While there was some appreciation for the idea of Frenchlanguage education of Islamic inspiration, parents maintained that a child would still need to attend traditional Qur'anic school ''to learn how to be a Muslim/to learn how to pray'' (the verb used, juulgo, means both).
Religious traditions, particularly those with textual traditions, include beliefs about language, and these beliefs can be highly resistant to change (Crystal 1965;Schiffman 1996). However, as a religious tradition changes, so may the associated beliefs about language, change that is observable in religious language policies. In the Maroua Fulbe community, different religious language policies are in competition. In traditional policy, the sacrality of Arabic is emphasized for the great majority, and this orientation sustains local Islamic practices and beliefs. In the new policy, Arabic is both a sacred language and a second language that all Muslims should learn to comprehend and use to access the Qur'an and translocal Islamic discourses.
As Avni (2012) observes, religious educational settings are ''pivotal sites in the construction, maintenance, and reproduction of worldviews and values that reach across individuals, families, communities, and nations'' (p. 170). Families are key sites in which values and expectations regarding religious education and language learning play out and are potentially transformed. While families may not realize that they are participating in language policy and planning, their decisions on the micro level are influenced by language changes planned at the macro level (by government and/or international donors) and may have an impact on the wider language situation (cf., Kaplan and Baldauf 1997).
Only a minority of Fulbe participated in ''modern'' Islamic schooling, but over time this movement may transform the relationship between religious learning and Arabic learning in the Maroua Fulbe community (Moore 2013). Since my fieldwork, private Islamic schools have continued to grow in size and number, with girls outnumbering boys ( van Santen 2014). As the case studies show, even families that did not embrace modernization made choices about religious education that reflected the movement's influence on how religious knowledge and Arabic competence were conceptualized and cultivated. Examining and comparing focal families' religious (language) education policies makes this visible and gives us insight into how broad, public changes in religious language policy manifest in private decisions about what constitutes (adequate) Arabic competence, how it should be acquired, by whom, at what stage in life and religious study, and for what present and future purposes.
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.