Everyday heritaging: Sino-Muslim literacy adaptation and alienation

: What we in this article describe as “ Sino-Muslim heritage literacies ” have existed in China for as long as there have been Muslims in the region (since the 7th century according to the best evidence). The community ’ s religious and heritage literacy practices can incorporate a systematic Arabic representation of Chinese, systems of Chinese characters representing Arabic pronunciation, and more contemporary digitalised manifestations of heritage literacy in everyday life. Using a social practice approach to literacy, this paper reports on multi-generational interviews, artefact collection, and ethnographic observations with two families in Xi ’ an (Shaanxi, China) to explore how heritage literacy practices maintain a presence in Sino-Muslim life through traditional systems of community and religious education and contemporary social and material networks. We discuss what these empirical cases reveal about literacies in Sino-Muslim religious life, with respect to how heritage is adapted or diminished across generations. We also argue that it is crucial to situate Sino-Muslim heritage literacy in spaces beyond rigid and state-defined ethnic and religious discourses which tend to confine the identity of Sino-Muslims into officially designated categories. Doing so, we contend, has useful theoretical and methodological import, and can shed light on inquiry about heritage literacy in other minority settings. Abstract: 本 Regional autonomy is practised in areas where people of minority nationalities live in compact communities; in these areas organs of self-government are established for the ex-erciseoftherightofautonomy.Allthenationalautonomousareasareinalienablepartsofthe People ’ s Republic of China. The people of all nationalities have the freedom to use and develop their own spoken original languages, and to preserve or reform their own ways and customs.


Introduction
In this paper we discuss the first round of data collection for a research project, funded by the Leverhulme Trust, which examines Sino-Muslim heritage literacy in modern China. We offer a critical examination of heritage literacy practices in the everyday lives of two Sino-Muslim families based in Xi'an (Shaanxi Province, China): the Wang family, and the Chen family. We also explore how their religiously expressive heritage literacy practices occur at the interface between an authoritarian state which confines religious practice entirely through minority ethnic identity (shaoshu minzu; 少数民族) and its Muslim minority who have inherited and adapted literacy practices that are situated in heritage-related activities, and which are inherently translingual and transmodal in nature.
The study is located within a tradition of literacy studies that conceptualises literacy to be, above all, a social practice (Gee 2008;Street 1984). The complex nature of ethnic heritage and religion in China also requires an acknowledgement of the complementary fields and sub-literatures which this study must draw from to develop its theoretical foundations. These include not only literacy studies, but also Sino-Muslim studies, and studies of literacy in religious communities and Islamic education, all of which have important implications for our research.
In this paper we begin with a brief overview of the particularities of the Sino-Muslim context in China. Via the interdisciplinary literatures we have drawn on, we then orient the reader to how we have conceptualised heritage and religious literacies within the research, the social practice approach to literacy which guides it, and an account of Sino-Muslim heritage literacy as historical practice. This is followed by two extended excerpts from family interviews. The final sections discuss how it is crucial to situate Sino-Muslim heritage literacy in spaces beyond rigid and state-defined ethnic and religious discourses which tend to confine the identity of Sino-Muslims into officially defined parameters. We argue that such a shift can shed light on heritage literacy inquiry in other settings, and doing so has high practical as well as theoretical potential for literacy studies on minority communities.

Background: Sino-Muslims in China
After the People's Republic of China (henceforth, PRC) was founded, in the 1950s, religion became eschewed as a criterion of social categorisation for China's Muslims. Instead, minority ethnic identities (shaoshu minzu; 少数民族) were ascribed to them by the new government, promoted through social incentives (Mullaney 2011), and an ethno-religious identity assumed for Sino-Muslims on account of historical ancestral commitment to Islam (Frankel 2021). This resulted in religious identity and heritage becoming coterminous with ethnic group (minzu) identity. In particular, with those ethnic groups that identified themselves as 'Muslim' and/or who purported to have a religio-cultural association with Islam.
A subsequent change in the terminology used to describe Chinese Muslims occurred. Prior to the PRC, terms such as 'huihui' were coterminous with 'Muslim', and Islam was described as the 'teaching of the Hui' (huijiao; 回教). These were used as general terms to describe Muslims in China who were of any ethnic background, e.g. Han, Tibetan, Mongolian, and whose antecedents had converted to Islam and intermarried with settled Muslims between the 7th and 14th centuries (Frankel 2021). Among China's Muslim minzu were some communities who became identified through what was deemed their own language, and from which their PRC designated ethnonym was subsequently derived. These communities include the Uighur, Kzakh, and Tajik. Other Muslim communities, such as the Donxiang, Salar and Bao'an, speak a combination of Turkic, Mongolian and Han Chinese dialects and are located largely around the border of Qinghai and Gansu provinces. The label Hui, therefore, became legitimated for those who were perhaps the most elusive to define by the PRC's imposed Stalinist criteria. As Sinophone Muslims, or Sino-Muslims, they could not be identified by region or a separate language and they often lived alongside their Han compatriots. These sinicised followers of huijiao were therefore not regarded as constituting a united ethnic group prior to their minzu identification by the PRC in the 1950s (Gladney 2004).
With regards to the Hui relationship with Islam, the official government position is that they are an ethnic group of putative foreign (i.e. Arab, Persian, Turkic) ancestry who are associated with, but not defined by, Islam (Frankel 2021;Gillette 2000). This aspect becomes central to how we have conceptualised religious and heritage literacy, as outlined in the next section. Since religious distinctions fell under the category of minzu under the PRC they became regulated by state-run institutions as part of China's attempt to honour its constitutional commitment to recognise the distinctiveness of minzu heritages, as outlined in Article 4 of the PRC Constitution: 1 Regional autonomy is practised in areas where people of minority nationalities live in compact communities; in these areas organs of self-government are established for the exercise of the right of autonomy. All the national autonomous areas are inalienable parts of the People's Republic of China. The people of all nationalities have the freedom to use and develop their own spoken original languages, and to preserve or reform their own ways and customs.
A definitive statement on the place of religion in the country is outlined in Article 36: 2 The state protects normal religious activity. No one may make use of religion to engage in activities that disrupt public order, impair the health of citizens or interfere with the educational system of the state. Religious bodies and religious affairs are not subject to any foreign domination.
The phrase that immediately stands out above is 'normal religious activity'. Reading what follows, one could deduce that it relates to practices which fall within the parameters of domestic minzu distinctiveness (as oppose to 'foreign domination'). In the secular PRC, despite the de jure legality of religious expression, how it is tolerated varies from context to context and has become a highly politicised topic rather than a matter of consistently applied state law. In his epilogue, Stroup (2022: 157-165) provides important reflections on how changes in ethnicity policies, characterised by sinicisation and the need to combat perceived "pan-halalifcation" (fan qingzhen hua; 泛清真化), have impacted Islamic communities differently throughout China. Practices related to minzu heritage in some cases become the only outlet for public religious expression. With space for minzu heritage expression as permitted, and sometimes in essentialised ways lauded (Frankel 2021), heritage practices associated with such things as sacred Arabic calligraphy, public prayers at funerals and Eid, and promotion of 'Muslim cuisine' 3 become part and parcel of religiously expressive acts which occur publicly but under the aegis of 'among-the-people' (minjian; 民间) activities. These 'top-down' heritage activities can also be supervised and administered by government intermediaries such as the regional offices of the China Islamic Association 4 who operate under the Bureau of Ethnic and Religious Affairs to mediate between the Chinese state and local Muslim communities. Erie (2016), in an extensive study of Sino-Muslims in Linxia in Northwest China, notes that the notion of minjian allows for a space that is "a middle ground where the Party-State and Hui meet to solve problems, articulate needs" (Erie 2016: 343).
For Sino-Muslims, religious literacy practices exercised through the minjian are often seen as part and parcel of heritage and not just devotional and ritual endeavour. But minjian is deployed as a space in which cultural material is conventionalised and made to associate with an 'ideal' and rather policed Hui identity. It is a site for what Hobsbawm and Ranger (1983) refer to as the "invention of tradition" where practices emerge from "responses to novel situations which take the form of reference to old situations, or which establish their own past by quasiobligatory repetition" (p. 2). While Hobsbawm and Ranger were particularly interested in tradition as it relates to imposed and value-laden rituals, their arguments could also help us understand organic values, practices, and behaviours which may be neither imposed nor ritualistic. Individuals and communities can also engage in new or divergent practices, and invoke personae that are more hybrid and transnational than the minjian space will allow. Importantly for us, therefore, Erie's interpretation of minjian, as between the party-state and the Islamic associations and Imams, is complemented by Veg's (2019) theorisation of minjian as a general term to relate to religious practice which is unofficial, unaffiliated, often self-funded, and grassroots.
In this way, the relationship between heritage and tradition is not so straightforward. Sino-Muslims canand doreinvent new modes of heritage literacy through recourse to tradition. In this research, our intention is to illustrate how some of these practices are invented and re-invented as heritage literacy in particular localities within specific conditions. At the same time, we consider the extent to which different constructions of heritage converge or diverge from officialised minjian.

Literacies, religion and social practice
The theoretical framework for this study brings the ideas of heritage literacy in the Sino-Muslim context together with a social practice approach to literacy (Barton and Hamilton 2012;Baynham and Prinsloo 2009;Street 1984). Sino-Muslim heritage practices inherently involve values, materiality and social relationships, and are often mediated by sacred or religiously themed texts. Thus, conceptualising heritage literacy through a social practice framing means acknowledging that heritage identities change over time and are given shape by the values people hold. Ivanič et al. (2009) have argued that literacy practices involve a complex negotiation of identities which are held together by values and associations. For example, how members of religious communities practice different literacies to stay connected with co-religionists and family will closely relate to their values Sino-Muslim heritage literacy adaptation formed through community activities and religious practice. How and why they stay connected, tells us something about how heritage literacy practices and concepts have become "adapted, adopted, or alienated" (Rumsey 2009: 575) from their everyday use. As we shall come to see in the extracts from our data below, it is identities that affect which literacy practices are maintained, passed on, or relinquished. This is because individuals' interpretations of, and attachments to, heritage literacy practices are formed by their identities and social contexts, which are in turn shaped by their beliefs.
The literacies of religious, including Muslim, communities have been the subject of some important literacy research over the last few decades, even when the focus of the inquiry was not explicitly religious practice. For example Street's (1984) seminal study on Iranian village maktabs (small-scale religious seminaries) found that 'maktab literacy', a close equivalent to the Sino-Muslim "scripture hall education" system discussed below, was an enabler for other forms of literacy later in life, including literacies for commerce and trade. This was despite maktab literacy being focused almost entirely on rote memorisation and religious activity. The majority of children undertaking religious study in maktabs were able to utilise in some capacity what they had learnt and apply it later in their lives when they went out to work in a trade or craft (Street 1984: 179). This argument is significant because it implies that, while the primary goal of the maktab was to teach children the fundamentals of religion, the impact was far beyond the domain of religion, impacting future life, social activity and economic conditions.
Another study in a Muslim population is that of Scribner and Cole (1981) who conducted five years of research on the Vai people, a community in Liberia. Among the aims of their research was to ascertain the effects of forms of literacy related to different functions. For the Vai, as with the Sino-Muslims of our study, literacy was not necessarily associated with formal schooling, and while English was spoken and written in school, indigenous-language literacies were adopted for functions beyond school. The latter was not a 'schooled' literacy but part of community conventions. There were also religious, particularly Quranic literacies which were textually and audibly part of everyday Vai life. They found that formal schooling constitutes only one form of literacy among other possible 'literacies' that individuals in a given society may be socialised into.

Heritage and religious literacies
What do we mean by religious literacy, and heritage literacy in the context of Sino-Muslims in China? What we propose here is a definition which begins with the way religion and heritage are accomplished through practices of literacy. But first the terms 'religious' and 'heritage' as they are used within this research need to be understood.
The sociologist Emile Durkheim (see Durkheim and Fields 1995) argued that practices which define the sphere of religion ultimately serve to differentiate the sacred from the profane through three general functions: Firstly, the function of providing meaning to life through a sacred principle or belief, and fundamental ideas about life, death and existence. Second, a social cohesive function through which definite groups are formed by way of, for example, common ideals, texts, places and people. Religion in this function serves as a socially connecting mechanism. And finally religion's emotive function which allows believers to use religionand its language and ritualto achieve a kind of collective emotional solidarity. Literacy practices related to each of these functions, whether they are as part of (secular) Hui minzu social life or theological conviction, could be related to religion and possible sites of inquiry, therefore.
Furthermore, we consider Kapitzke's (1995) definition of religious literacy as a flexible and supporting analytic tool. She defines religious literacies as "goaldirected sequences of activities organised around sets of beliefs, values, symbols, artefacts, narratives and rituals" (Kapitzke 1995: 3). In our research, however, the term 'religious literacy' becomes problematic when we consider people who regard themselves to be non-religious (in a traditional sense), who are more like 'silent' participants of an ethno-religious community such as the Hui, and yet still maintain practices of heritage as part of minzu affiliation.
In light of this broad and flexible understanding of religious literacy which we sought, at the outset of our research we felt the need to avoid narrowing our scope to such things as 'sacred literacies', 'liturgical literacies' (Rosowsky 2008), and 'faith literacy practices' (Gregory et al. 2013), all of which have been examined in similar previous studies. These terms suggest particular acts of worship within sacred locations, and are therefore much narrower than the literacies we wished to explore. For many Muslims, including Sino-Muslims, sacred literacy is usually associated with Classical Arabic and while this is a feature of our study, we are certainly not confined to it in our exploration of how religious and heritage literacy practices featurein a Durkheimian sensewithin social activities, community bonding and personal growth. Also, for many Sino-Muslims, religious reading practices can be entirely through Chinese characters used to phonetically represent the Arabic and Persian language items central to Muslim prayer, as we outline below.
This brings us to the concept of heritage literacy, a term which we have decided to use throughout the study. The concept of heritage literacy has been defined by Rumsey (2009) as a heuristic for "how people transfer literacy knowledge from generation to generation and how certain practices, tools, and concepts are adapted, adopted, or alienated from use." (p. 575). Rumsey further explored this concept in an auto-ethnographic study in an Amish family and found that heritage literacy practices were centred around individual decision-making processes, which were life-long and constantly transforming as members of particular communities negotiated their intellectual, spiritual, and cultural inheritances in different ways and at different points in their lives. Rumsey also argues that heritage literacy practices are "inherently multimodal" because they utilise multiple sign systems, including, as documented in her account with an Amish family, "cuneiform, hieroglyph, beadwork, wampum, quilts and cooking" (Rumsey 2010: 137).
Building on Rumsey, we will demonstrate that Sino-Muslim heritage literacy practices are inherently translingual and transmodal. Translingualism points to how heritage texts, in the social environments where they are used, necessitate dynamic processes in which Sino-Muslims must strategically navigate use of multiple languages and codes as one complete repertoire (see Canagarajah 2013). Language here is seen as a situated practice rather than a static system (see also Li 2017). Transmodality alerts us to how, in constituent interactions, practices are negotiated across a range of modal transitions (e.g. text-based with oral recital and bodily performance) as part of meaning-making (Cárdenas Curiel and Ponzio 2021). Conceptualising heritage literacy as translingual and transmodal allows us to move beyond the limitations of additive multilingual and multimodal ideologies, which, at their core, distinguish between named languages and modes as inherently separate (García and Li 2014). We use these ideas here to observe the inherent embeddedness of linguistic and modal resources in Sino-Muslim heritage literacy practices. In connection to the related term 'multimodality', both translingualism and transmodality can be captured through Pennycook's (2017) concept of "semiotic assemblage" which refers to the fluid negotiations of language with other modes such as images, bodily movement and sound, and expands the notion of repertoire to encompass "spatial distribution, social practices and material embodiment rather than the individual competence of the sociolinguistic actor" (Pennycook 2018: 47).
We would also add that, in this context, heritage literacy practices are a way to set and maintain 'metalinguistic community' attachments (Avineri 2012). In a study of Jewish communities in the USA, Avineri's account demonstrates how community individuals were socialised into language ideologies related to Hebrew and Yiddish as part of heritage learning. Similarly for us, as outlined below, language is a vehicle for heritage knowledge through a community of positioned social actors, such as literacy sponsors, who engage in discourse around texts and other religio-cultural symbols as part of historical literacy practice.
We therefore continued using the two terms heritage literacy and religious literacy almost interchangeably throughout the research to describe what we regard as goal-directed practices of literacy in which heritage and religion have a role. In this respect, Kapitzke's (1995: 3) definition is the closest to what we sought to deploy in our working definition, alongside Durkheim's privileging of socially bonding practices over beliefs. We accept that this is somewhat of a wide scope, but necessary since our research encompasses informal and mundane interactions sometimes outside the purview of religious institutions, formal education, and state-defined parameters. In the section below, we briefly outline a particular historical system of literacy in Sino-Muslim communities which, we argue, has important implications for their heritage literacy practices today.

Scripture hall education
In the Ming period (1368-1644), Sino-Muslim literati in Shaanxi Province established a linguistically flexible educational system for mosques called 'scripture hall education' (jingtang jiaoyu; 经堂教育), which consisted of an Islamic curriculum of Arabic, Persian, and Chinese texts. As a teaching tool, the system used a Chinese phonetic transliteration system to represent the original orthography of Arabic and Persian passages from classical Islamic texts (Lipman 1997;Petersen 2018). Another form orthographic exchange was the development of xiaojing, 5 a method of transliterating the Chinese language using Arabic script (see Figure 2 as example). This was an Arabised form of Chinese characters or "Pinyin with Arabic" (Qurratulain and Zunnorain 2015: 54) which is said to have been formed by Arab traders and later developed into a system used for purposes of community correspondence and religious learning. With their porous linguistic boundaries, these literacy systems were a way to spread global Arabic Islamic discourses to a growing Sino-Muslim population in historical China. They were often adapted locally by 'ahongs' (Imams) in less standardised forms for use in Sino-Muslim education centres throughout China, and augmented with oral recital, commentary and discussion.
Arguments already outlined regarding how heritage and religious literacies can be applied and adapted to everyday dealings are helpful in our attempt to understand how Sino-Muslim heritage literacy practices have importance beyond fulfilling religious duties and transmission of Hui traditions. How they are also adapted (or relinquished) as a means for participation in a local or global Muslim community, and a crucial linguistic instrument for gaining knowledge and experiential growth are also important areas of inquiry. According to an historical argument made by Hanna (2007), religious literacies in Muslim communities are worthy of a deeper investigation on their own terms. We would argue that this is even more the case in contemporary Sino-Muslim contexts, where long-standing heritage literacy practices have come into direct conflict with State-enforced sinicisation of religion policies 6 which restrict manifestations of Arabesque-Islamic influences in the public domain, such as bans on using Arabic signage to denote qingzhen (halal) food, and the forced removal of Arabic duas (supplications) outside the homes of Sino-Muslims (Ridgeon 2020). This is made further relevant today given China's ambitious Belt and Road Initiative (BRI) which is predicated on a celebrated past of ancient networks of overland and maritime Silk Roads extending China into Central Asia, Eurasia, the Middle East, and Africa. This historical network was largely forged through Muslim involvement with Chinese culture and from which the scripture hall and xiaojing systems of literacy emerged. In many of these current BRI projects, for which the city of Xi'an is an important hub, the PRC will make close contact with Muslim populations and cultures, invoke its Islamic heritage for diplomatic purposes, and allow for the bidirectional flow of goods and people. How heritage literacy emerges in new ways in the midst of such convulsive changes in the social, economic and cultural lives of Sino-Muslims in cities such as Xi'an becomes an important focus of inquiry. It is, therefore, one of the aims of our research.

Methodological approach
The research was conducted by a team of five researchers located across China and the UK, including the two authors of this paper. Participants were recruited through a combination of purposive and snowball sampling, building on contacts made during the pilot and through personal networks and the locally-based researchers in the team. For reasons already highlighted, Xi'an was chosen as one of the focal sites of the research due to its historical connections with the Silk Roads, its localised development of scripture hall literacies, and its contemporary BRI-related cultural shifts. Consistent with our context-sensitive and interpretive approach to literacy, the following methods were used to capture the diversity and richness of the heritage literacy practices of the two families: (i) Group interviews with multiple generations of each family. Discussion prompts adopted a biographic and narrative focus on literacy learning processes in early life, networks of people and locales, and key events around literacy to explore heritage resources, habits, and historical trends. (ii) Between interviews, participants shared documents and samples of artefacts and texts which have significance for heritage literacy. These were often brought out during research encounters. (iii) Where practicable, observations of literacy-related activities were undertaken. As research encounters these included reciting and reading of texts, and were either immediately following interviews or arranged at a later date. . This includes researcher-participant 'Shuhan' who, as shown in the coming extract, recruited her mother, grandmother and wider network as part of the project though was not involved in the data management or writing up for this paper. Interviews were conducted multilingually, with Mandarin as the dominant medium, followed by some elements of Classical Arabic 7 and English where necessary. This was followed by a meticulous process of transcription, translation and editing which involved the entire team in order to clarify translingual religious expressions and features of regional dialects used by respondents. We then integrated the transcripts with other data (images, observation notes) into NVivo and undertook a process of thematic coding which then evolved into a pattern-coded framework of analysis (Saldaña 2013). This enabled us to identify emergent patterns from the data via codes such as translingual-Arabic, transmodal-oral, adaptation, loss and alienation, among others, which we brought together into various analytic themes for this paper. The themes discussed in the sections below relate to heritage literacy loss, coming back to heritage literacy, and disruption and adaption.
Ethical issues relating to this research were fully examined and approved through the institutional ethical review process. 8 Specific challenges emerged which relate to issues of anonymity and identifiable features of participants, and the security risks associated with religious proselytising online (Weixin 2021). In response to these, participants' names and villages have not been identified in the research (though their cities are), and interview discussions and document collection were concerned with issues regarding minzu heritage and practices, rather than proselytisation, and therefore remain within the confines of Chinese law.
In both cases, more than one generation was present in at least one family interview. In the next section, we begin with an extended extract from a research interview involving three generations of the Wang family a Sino-Muslim family in Xi'an. Following this, we contrast the Wang family's account with another Xi'an based Sino-Muslim family, the Chen family, whose experience provides a much different story for our findings.

Three generations of the Wang family
The extract below involves Xiaoming 9 (85 years old; grandma), her daughter Yanyan, and granddaughter Shuhan who is in her late twenties. Shuhan, the interviewer and also a researcher in the project, discusses various aspects of the family religious heritage with both her mother and grandmother and observes the latter reciting some Quranic scripture. The discussion between them covers a range of topics including the role of religious scripture in the family, migration, education, food, and prayer: Xiaoming: I was born and raised in Luoyang and then moved to Kaifeng for marriage, then to Baoji, and then to Xi'an. In Kaifeng, I lived for ten years within a mosque compound and learned religious laws and rites directly from following my husband and the elderly in the community. It was the same routine every day. Yanyan: The young apprentice Imams would enter the mosque in childhood and study and teach. Xiaoming: Some were in their twenties and some in their teens. The mosque admitted them to learn Arabic voluntarily. Some went on to study Arabic abroad. Though I personally did not study Arabic back then. I used to read scripture in Chinese characters instead. I didn't know 8 Approval reference 171_2021 issued by the School of Social Sciences, Education & Social Work at Queen's University Belfast (UK). 9 In line with the guiding principles for ethical research, all names of the participants, and other identifying features, have been changed. the scriptures through Arabic. I looked at the Chinese characters in your grandpa Kaifu's books 10 and learned. I would say combinations of them during worship. Shuhan: Did you think about learning Arabic systematically to better understand the scriptures you were reading, considering that you were in the mosque? Xiaoming: No, because as soon as I went to live at the mosque I gave birth to your mother and her twin, and they kept me too busy. I did not have the time to learn. Yanyan: So she went to live in the mosque after marrying your grandpa. Later, because of the Cultural Revolution and also due to divorce, we went to Baoji. Shuhan: Does Islam discriminate against women when it comes to learning scriptures? Xiaoming: Islam does not discriminate against women. Men and women are equal. Especially in reciting scriptures. In fact, there are female Imams and female-only mosques. 11 Shuhan: Do you want my mother's generation, or even my generation to recite scriptures and learn Arabic? Xiaoming: It's not like that. After all, your uncle is also Han, and your father is also Han. Your eldest cousin, your second cousin are all Han people. Yanyan: You see, when your grandma was young, her parents were more religious, so she learned more. Because we later lived in a Han people's area, grandma was forced by life to be less religious. She did not recite scriptures in Baoji, and we didn't go to the mosque so regularly. So religious knowledge has lessened with each generation. Your parents do not have much religious knowledge, and you have even less. It is getting less and less. But in terms of observing the fasts and Eid, grandma has always insisted on it. Shuhan: Actually, did you also think that we would be a bit resistant to it? Xiaoming: You didn't resist. It was based on all our life conditions. For your mother the conditions were unsuitable because she didn't have a suitable partner for this.

Analysis: letting go of heritage literacy
In the above extract we are given a glimpse into Xiaoming's heritage literacy story spanning three generations, including how much of her heritage literacy learning took place within, and was contingent upon, the religious and self-help systems of her community. These resources included the mosque, the family, and community elders all of whom practiced forms of formal and informal religious education that included reading scripture with Chinese characters (see Figure 1) as well as with Perso-Arabic script (see Figure 2), though Xiaoming's attachment to heritage lay largely with her use of the former. She relied on Chinese characters primarily but also heard the Arabic read aloud simultaneously, allowing her pronunciation to improve through listening. This tells us that as she read the text while listening to it in Arabic, allowing her to 'mouth' her prayers in Arabic through Chinese characters. This helped her memorise the text to then, when ready, physically orient herself towards 'tianfang' (天房; Ar. qibla; the direction of Makkah for prescribed prayer) and 'read' it again with her whole body. The text therefore received multiple renderings in these different ways and, in observation, she recited them to us with great pride and enthusiasm.
Her heritage literacy practices were not only mediated by what Brandt (1998) calls 'sponsors' of literacy, who direct and control which literacy is accessed by whom, for what, and how it is articulated, but also group activities within the community and home. In the face of cultural change, these sponsors of literacy not only circulated resources but also nurtured skills, including reading texts out loud, in ways that fostered spiritual development and heritage awareness. Within the changing economic and political conditions of the last four decades in China, Xiaoming's religious education was, initially, in concentrated sites of sponsorship such as the mosque in Kaifeng and the home. These were deep wells that nurtured a Sino-Islamic form of reading and education that have in some cases maintained a presence in her life till today, although less so in the case of Xiaoming's daughter and granddaughter. The texts from which she learnt (and heard read out) 'live on' every day in her life in daily practice, through a form of 'entextualisation', the dynamic moving of text into other settings and modes. Blommaert (2005) builds on Bauman and Briggs' (1990) use of the concept of entextualisation and describes it as: [T]he process by means of which discourse is successfully decontextualized and recontextualised, and this made into a 'new' discourse. In every phase of the process, discourse is provided with new metapragmatic frames (Blommaert 2005: 252) For Xiaoming, entextualisation is perpetuated by her daily performance which, while limited to her now homely existence, has roots deep in, and is deeply intertwined with, the institutions and activities of her early community life. Another example of Xiaoming's creative and adaptive practices with heritage literacy can be seen through an extract from grandpa Kaifu's book (Figure 2), which she provided. She reports that she learnt from these texts without knowing how to read them, but by hearing them read out and discussed in the mosque and in the family. We have translated an extract as below, with standard Mandarin in square brackets alongside Xiaoming's Henan dialect: The above extract is taken from one of grandpa Kaifu's books that Xiaoming used to learn jiaofa (教法; Ar. fiqh; practical Islamic jurisprudence) in her period of learning in the Henan mosque. Like many Islamic primers of this kind it is based on a question-answer format. It is written in her local Mandarin dialect but through Perso-Arabic script. Most of the text can be read out as Mandarin, as we have shown. Among the dialect features is the word namaz, a commonly used Persian term for prayer instead of libai (礼拜) in standard Mandarin.
Earlier, in Figure 1, we saw how reading out loud was a way for Xiaoming to mouth and repeat texts as learning. Here this was even more necessary as she would learn by having this read out, pointing once again to the permeability of oral and literate modes in situated literacy events. Not only does the oral and literate binary diminish upon scrutiny, but so too does the monolingual and bilingual binary. These everyday lessons and discussions in her early life therefore needed to be both transmodal and translingual encounters in order to make sense for her: she heard these texts read out, expanded their content across modes in particular ways as part of practice-oriented maintenance, and interpreted their meanings through her own everyday actions.
While Xiaoming harnessed heritage literacy for spiritual, relational and moral purposes in the community, she was unable to pass on much to Yanyan except her insistence on observing Eid and fasting. She now directs food-related activities during Eid and Ramadan, placing her as the sole sponsor of heritage literacy in the family. Contemporary social activities which are approved through officialised minjian, such as Eid celebrations and qingzhen cuisine, allow for the entextualisability of select heritage literacy practices. Such practices, however, relate more to minzu commercial placemaking and consumption patterns than religion, as manifested in Xi'an's hugely popular 'huimin jie' (回民街; Muslim street) awash with food vendors selling 'Muslim food'. Other heritage-related domains, such as daily prayers, have been affected by kerbs and regulations (e.g. in many mosques under-eighteens are barred from entry). The effects of this over time are that Yanyan and Shuhan do not pray. Moreover, their engagement with family religious traditions such as Eidand the religiously expressive literacies associated with themis motivated less by theological conviction and more by Hui heritage, and community cohesion and solidarity.
The Wang family extract shows us that heritage literacy practices are distributed across networks of people, locations and artefacts, and not disassociated from social upheaval such as migration, divorce and forced cultural change. Xiaoming recalls how the decline in the family's religious practice is only a matter of adaptation to living in a Han majority area and scarce literacy sponsorship, and that her failure to learn religious literacy more substantially (i.e. through Arabic) was not through lack of dedication and observance but rather that she lacked the time due to housework and childcare commitments. By highlighting women-led mosques, Xiaoming underscores the distinctly localised nature of women's agency within Hui Islamic heritage in Henan and how this vital network of literacy sponsors was not in place in the areas to which she eventually moved. Xi'an, while having a significant local Muslim population, hosts a significant number of Henan Muslims since as far back as the Republic era. There are subsequently a number of local mosques frequently attended by Xi'an Muslims whose ancestral lineage connects to Henan rather than to Shaanxi. Moreover, according to Shui and Jaschok's (2014) research, the tradition of women-led mosques diffused from Central China, notably Henan, to other parts in China, such as Shaanxi. Given that Xiaoming originates from Henan, her background is likely to also have prevented her from being fully integrated into the Xi'an local Muslim community networks. This almost certainly influenced the subsequent intermarriages with Han people (hanmin) on the part of Yanyan and Shuhan, thereby also impacting her (and likely other Hui female emigres') heritage literacy.
This sheds some light on the process of the Wang family's alienation from heritage literacy (see Rumsey 2018), and Xiaoming's attempts at agentive decisionmaking over the years in order to hold on to heritage literacy amid changing social, financial, and political aspects of the family's life. Xiaoming shows a significant acceptance of her circumstances, and according to Rumsey (2018: 99), "perhaps, one of the deepest forms of agencythat is in letting go, in acknowledging, coming to terms, and accepting the losses of various literacy practices". This particular case becomes salient for our analysis in terms of heritage literacy decline and its connections with literacy sponsorship. We later return to this case and contrast the Wangs with another Xi'an based Sino-Muslim family, the Chen family, and analyse their heritage and religious literacies in these terms.

Two generations of the Chen family
As with the Wang family, multiple interviews were also undertaken with the Chen family who originate from a Sino-Muslim enclave in south of Ningxia Hui Autonomous region. In the below extract, Jizhi, Lei's father, begins with an account of his own religious learning followed by Lei's later experience of mosque education: Jizhi: I was born before the Cultural Revolution. Due to this unique historical period, I did not have the opportunity to go to the mosque to systematically study religious knowledge, nor to recite the scriptures in Arabic, etc. When I was seventeen years old, the ahong of our village transcribed for us the Quranic scriptures and prayers related to worship into Chinese characters for us to sound out. We called it 'hanjing' 12 . The main contents of learning were the basics of worship, visiting graves, dua (lit. supplications) and visiting graves. Later, when I was in my thirties, I served as village elder of the mosque. During my tenure, I still had not learned the Arabic alphabet, and hanjing was the mainstay. My study of religion and the Quran was based entirely on this longstanding connection to hanjing. Now I am learning Arabic letters and I am learning everything all over again, this time in Arabic. Lei: My story is different to father's. When I was 7 years old, he sent me to a local mosque to study Islam. This study began with Quranic Arabic reading initially for recitation purposes. The ahong would write on paper or upon cow bone. Mosque education occupied all my holidays every winter, summer, and especially in Ramadhan. After daybreak and before sunset, family members would sit together and talk about religious knowledge.
Languages we used included Persian, Arabic, and Uighur. There were 13 fixed texts for us to learn and study to become ahong. At the start, we learned Quranic pronunciation rules. We then moved on to how to read sentences and grammar, including Arabic morphology. After that, we started to learn the Quranic exegesis and Hanafi fiqh. 13 With the help of the excerpts of scriptures through the teacher's self-made hanjing manuals. Then if you continue to a higher level, it is an education aimed at nurturing you into an ahong. We had poetry, usually Bustan or Gulistan, 14 also 'zan sheng ci' (Ar. dhikr; Islamic meditation chant) through a text called 'Madayeh' by ahong's allocation only. They also fostered our skills of writing Arabic continually. If someone had been learning for one or two years, and they knew the Arabic writing from the books, the ahong will let you teach the new students who come in. Researcher: Why did you leave the course? Lei: Because of going to school and university. Going to university outside the Hui area, it normally means four years away from religious life, so when you come back you feel like you're a complete stranger. It is hard to pray five times a day, so you may find yourself completely detached from religious activity. So Hui university students increasingly talk about the importance of "coming back". As for me, I did not undertake any more religious learning through mosque education. As you know, our knowledge is not publicly validated. If I had questions or doubts, I usually called our family, ahong, or my religious acquaintances. But, instead of becoming an ahong, I went on to study an undergraduate degree is Arabic Linguistics, a Master's then a PhD in Arabic Language. My area is in ethnology, particularly Arabic society, and I went to the Middle East for my fieldwork.
10 Analysis: coming back to heritage literacy In the above extract, we can see a marked shift in religious literacies from Jizhi to Lei's current existence. Jizhi relied entirely on hanjing which was confined to practices of liturgical literacy and not further explanatory texts as used in the more complex system of scripture hall education. What was passed on to Lei was not necessarily a commitment to ritual and rite but rather a kind of metalinguistic attachment to Classical Arabic, and that is what makes this story relevant to our study of heritage literacy. Jizhi's very limited engagement with heritage literacy, through hanjing, was not insignificant. It impacted Lei in many ways, including in two particular ways that are important for our analysis: heritage literacy as a form of 'nurturing', and metalinguistic awareness and attachment.
Heritage literacy education in this context could be described of as an active process of consciousness raising, and not just simply about doing things with culture and passing them on to the next generation. Scripture hall education was designed to orient the student to an Islamic life grounded in a traditional moral and epistemological framework. And to cultivate in mosque students and the general congregation 13 One of the four canonical Sunni schools of Islamic Law. 14 Poetry was a compulsory element in the traditional Islamic madrasa as part of the development of the hermeneutic methods of religious law (see Makdisi 1981). a sense of membership within an Islamo-Arabic "metalinguistic community" in Avineri's (2012) sense of the term. As with the Wangs, some understanding of the Chens' geographical origins are useful in understanding their practices of adaptation and maintenance. Although regulations countering religious expression have tightened up in the past decade for Ningxia Muslim communities, following the Cultural Revolution religious freedoms intimately tied to booming business opportunities with the wider Muslim world allowed heritage norms and informal minjian networks to flourish both locally, translocally, and transnationally. Under these conditions, many of Ningxia's Sino-Muslims viewed Arabic literacy as both a sign of religious piety and a practical skill for business and commerce with the wider Arab and Muslim world. Indeed, Classical Arabic also served as a lingua franca for the primary author and Lei during the research process.
Our data show that Lei and Jizhi, and to some extent Xiaoming, maintained a metalinguistic attachment to Arabic even though it was through the medium of hanjing or scripture hall education, rather than a desire to be communicatively competent. They were taught to valorise Arabic as jing (lit. 'scripture'). While Arabic was a vehicle for religious knowledge and etiquette taught in his formative years, by the time Lei applied for university entrance Jizhi suggested that Lei choose Arabic for his undergraduate. This is because he thought being an Arabic interpreter would earn Lei good money and keep his son connected to the tradition in which he had been raised. In mosque, Arabic was central to status and can lead one to becoming an ahong, but Lei never made it that far. His alternative route took him to study Arabic all the way to PhD and to conduct field work in the Middle East, thereby making a living as an academic in the Arabic language, and also retaining a respect within the Hui community as a person who is connected to jing. Though in a manner that is far removed from the Quranic Arabic of his youth.
Lei also told us that he wants to return to religious study and enrol at the al-Azhar Islamic university in Egypt. This would be a kind of full circle or 'coming back' to the roots of his heritage literacy, through the symbolic association of Arabic with religious authority, as he tells us here: Though I had never led the prayer during childhood, later in life when the villagers found out that I had completed an Arabic language major they wanted me to lead.

Discussion: heritage literacy disruption and adaptation
There are important family orientations to the heritage literacy practices of both sets of participants. Family-oriented events were a key constituent of both Lei's scripture hall education, which was augmented by the oral tradition of the Sino-Muslim community in southern Ningxia, and Xiaoming's religious practices which were mediated by her husband and elders in the mosque in Baoji. Scripture hall education is known for being a mosque-based method of education, but respondents like Lei and Xiaoming have shown that there is symbiosis between the literacy practices of the home, the mosque, and wider networks in the achievement of heritage education, and that it does not take place exclusively in one context or the other. Similarly, Kell (2001) describes two domains of literacy in her ethnographic study of literacy in the context of South Africa. Within one domain, schooled literacy was conceived as generic and associated with passing tests. In the other domain, literacy was dependent upon multiple communicative modalities including oral, gestural, and digital forms of communication and sat uncomfortably alongside practices of schooled literacy. The second domain occupied the largest space in the lives of the participants, and a central tension for school pedagogy was located in the gap between the two. In the formative community lives of Xiaoming and Lei there was a harmony between two similar domains: An interplay of oral-based literacy practices of the home, and incremental study of the texts of mosque education. Heritage education was located in the space between the two, a space that was disrupted for Xiaoming by her migrations, divorce, and the Cultural Revolution.
For Jizhi, literacy was learnt primarily in relation to his relationship with God. It both began and ended there. There was little opportunity, if any, to communicate with the rest of Chinese society, let alone the wider world. Lei, however, drew on heritage literacy as part of getting to know the world, not just beyond Ningxia but into the wider Arab and Muslim world, and then eventually as a language-focussed academic career. Though, as he describes, there was a 'coming back' in adulthood as through learning Arabic he gained respect within the Hui community, compensating for not taking his desired path as an ahong earlier in life. For members of his father's community and for Xiaoming, acquiring literacy to be able to read religious texts, not even for writing in many cases, marks the end of heritage literacy learning. For Lei, it marks the beginning. His literacy world today is digitally-infused and multilingual. It is a modern day "piling up and extending out of literacy" (Brandt 1995: 651) with transformations and influences that reach far beyond the world his father inhabited. He is an active member of Sino-Islamic calligraphy groups and various Arabic language and news groups. Earlier forms of literacy, particularly hanjing and scripture hall education, have either become eclipsed, co-opted or amalgamated into Lei's current world. Modern and digitalised forms of hanjing now feature in his everyday communications (see Figure 3) which index transnational and translingual ways of being Sino-Muslim. In Figure 3, Lei shares an image from a mobile device as an example of how he has used a modern version of hanjing to help young Chinese learn prayer. In this example, he was asked by a younger Sino-Muslim who has not learnt Arabic to check the hanjing text and help with the Arabic formulations necessary for prayer, thereby now placing him as an important sponsor of heritage literacy.
For Lei, undertaking an Arabic language degree and being invited to lead the prayer was the apotheosis of his heritage literacy journey. For Xiaoming's family, only the literacy practices related to Eid and fasting have continued, and there has been very little of the adaptation and extending out that Jizhi's family have experienced. While mosques were the organisational base of communities and served as incubators for Sino-Muslim culture, heritage literacies were not confined Sino-Muslim heritage literacy adaptation to them. Multipurpose sponsors of literacy in Xiaoming's case eventually became fragmented, but in Lei's case become sought out in other places to promote persistence through adaptation.

Concluding remarks
Our consideration of the heritage literacy practices of the Xi'an Sino-Muslims in our study has prompted us to tease apart notions of heritage literacy, religious literacy and other concepts, such as minjian, which relate to the phenomena under study. Sino-Muslim heritage literacies are often religious, but can also be personal and idiosyncratic as they precariously tether Sino-Muslims to Islam, Chinese culture and other languages such as Arabic and Persian. They are borne of experience, shared through narratives, and an important anchor and point of reference as Sino-Muslims attempt to resuscitate sacred-profane distinctions in everyday life.
The Wang family's heritage literacies atrophied and remained confined to cuisine and key events, and for the Chen family a range of opportunities were opened up by traditional mosque education. This serves to remind us that, for our participants, heritage literacies are not just atavisms of bygone traditions made to fit in a contrived minjian space. They are intellectual and cultural inheritances that are translingual, and contingent of social activities that cause them to be expanded across modes in particular ways as part of their fluid maintenance and adaptation within a community that does not have a minority language in which to officialise or protect them. They are drawn upon to enliven something in the present that has recourse to the past, and are not always imbricated with explicit religious faith. Their engagement is also dialogic and performance-oriented, bringing the body to the fore in religiously prescribed ritual.
Our analysis illustrates how the meanings and functions of heritage literacy shifted as ideological, cultural, and religious networks changed, including scripture hall literacies being forced to adapt to the pressures of change during the Cultural Revolution. As more recent measures related to "poverty alleviation by education" (Yong and Zhang 2021) and sinicisation place mass literacycentred exclusively on Standard Mandarin (Xinhua 2019)as a key enabler of gaining economic advantage, Sino-Muslim heritage literacy practices developed new ideological and cultural uses and adaptations. Since Sino-Muslims are considered to be historically sinicised, and with no minority language through which their heritage can be ring-fenced and protected, what remains are only their heritage literacy practices, networks of sponsors, and new adaptations of practice with every generational challenge. At the time of writing this paper, new kerbs have been implemented on religious content in Chinese social media (Weixin 2021).
Such restrictions make new and digitalised renditions of hanjing even more pertinent as a new generation must creatively adapt heritage literacy for current times.
What we therefore see are practices of creation and adaptation of heritage literacy in relation to evolving forms of cultural production and sacred-making in practice: a process of everyday heritaging or "heritage as a verb" (Blommaert 2013: 618). This describes how heritage is subject to "practices of creation and modulation" (ibid) which, while hark back to an established cultural past (real or imposed), are by their very nature hybrid and difficult to pin down as belonging to one idealised language or community. This framing of heritage literacy necessitates, we would argue, looking not just at texts, but also at their trajectories and multiple renditions and adaptations, which themselves become new templates for heritage literacy norms, in ways that transcend the incapaciousness of named languages and ethnic groupings.