Decolonising Literacy Practices for an Inclusive and Sustainable Model of Literacy Education

Abstract

Despite efforts on the part of institutions, professionals and social agents, the Roma population in Europe still lacks equal access to education. Difficulties in literacy development are at the root of this: Roma learners present lower literacy rates than non-Roma learners and learners in non-segregated schools, preventing them from transitioning to secondary education. This article presents the results of ethnographic research with a group of Roma primary learners in Southern Spain. The aim was to analyse the contexts, interactional spaces, contents and practices of learners’ engagement with literacy in and outside the classroom. Data analysis was carried out using an adaptation of the continua model of biliteracy, useful for analysing literacy practices in contexts with different literacy cultures. Results show that communicative practices that challenged skills-based literacy models helped activate learners’ literacy reservoirs, enhancing their literacy engagement and allowing them to renegotiate their position as Roma learners in a non-Roma institution and as text creators in the classroom. Conclusions point to the need to decolonise classroom practice by identifying learners’ literacy reservoirs and ways to activate these, contributing to a more inclusive and sustainable model of literacy education consistent with the United Nations’ Sustainable Development Goal for quality education.

1. Introduction

In spite of efforts on the part of institutions (Council of Europe, national governments), professionals (educators, social workers) and other social agents (volunteers, NGOs, associations, activists), the Roma population in Europe still lacks equal access to education. In Spain, 43% of Roma children do not complete primary education and 66% leave school before the age of 16 [1]. These are Roma children in structural situations of poverty, marginalisation and social exclusion that face challenges such as school segregation and in-school discrimination. Difficulties in literacy development are at the root of Roma learners’ disadvantaged position to transition to secondary and post-compulsory education: they show comparatively lower literacy rates and reading and writing performance levels than non-Roma learners and learners in non-segregated schools [1,2]. Furthermore, in contexts where there are clashing literacy cultures, that is, differing ways of understanding, talking about, practising and relating to literacy, minoritised practices tend to be subdued and silenced in favour of dominant ones [3]. This divide is likely to occur in cross-class, cross-ethnic and cross-cultural settings, especially when the educational institution becomes a colonising space that represents and transmits the values of the dominant class/group/culture, whereas the learners make up the subaltern group. This is part of the literacy experiences of Roma learners, and navigating between the two becomes an unacknowledged non-curricular skill that they are forced to develop without guidance throughout the school years. This is an unsustainable educational model that results in Roma learners developing skills that are not useful from the point of view of the institution and the formal curriculum, and school failure [4].

This article presents the results of an ethnographic study carried out in collaboration with a group of Roma primary learners in Spain’s lowest-income sub-city district [5]. In the context under study, most learners belong to the gitano community (Spanish Roma, the oldest and largest ethnic minority in Spain) whereas the educational institution is non-Roma. In this context, there is a clash of literacy cultures: non-Roma literacy, centred on the written word and mediated by the verbal sign; and Roma literacy, with emphasis on oral communication and mediated by music and corporeality [6,7]. In this scenario, a New Literacy Studies (NLS) perspective, which considers literacy as a socially situated practice [8], is helpful for laying bare and untangling the tensions between literacy cultures in the classroom. In this way, this study contributes to recent research within the NLS on the connections between out-of-school and in-school literacy practices [9,10,11] and on the use of identity practices for literacy development [3,12,13,14], studies that aim to decolonise literacy practices in institutional settings [15,16] and research on literacy development for social and cultural sustainability [17,18,19,20]. In this regard, this study also works within the lines established in the United Nations’ 2030 Agenda for Sustainable Development and its Goal 4, which aims to achieve universal primary and secondary education for all, universal youth literacy and education for sustainable development (including cultural sustainability) through, among others, effective learning environments and teacher training [21]. To date, research on the formal and informal literacy practices of Roma learners is still scant [22,23]; in Spain, it is virtually non-existent.

The following research questions guided this study: (1) What connections are there between the contexts, interactional spaces, contents and practices through which learners’ engagement with literacy occurs both outside and inside the school setting? (2) Can communicative practices closer to learners’ out-of-school experiences activate their literacy reservoirs and therefore enhance their literacy experiences and engagement in the classroom? (3) What are the effects of these practices on the dynamics of the classroom community? In this way, we explored the connections between what learners did and said about literacy, and how they did it and said it, both outside and inside the school setting, and investigated ways these connections could be exploited for decolonising literacy practices in the English language classroom, which was chosen as the preferred in-school setting for our research due to its focus on orality and communication. There, we researched the effects that these decolonising practices had on learner engagement [24] and on power distribution within the classroom space [25] in order to shed light on different ways literacy education can speak to Roma learners’ literacy needs and build on their own literacy reservoirs. Ultimately, by taking into account how learners constructed and negotiated their literate identities as members of the Roma community learning in a non-Roma institution, this study aims to contribute to a better understanding of how inclusive and sustainable models of literacy education may work.

In order to do so, this article presents the following structure: first, the theoretical underpinnings of the study are introduced in the subsections below, where the two main concepts used in this study (literacy reservoirs and literacy continua) are conceptualised and theorised. Then, the Materials and Methods section includes information about the participants, explicates the intricacies of ethnographic research in a marginalised context and the researchers’ position in the light of critical ethnography and postcolonial theory, and provides a detailed description of the data collection and data analysis processes. The Results section is divided into four parts, following the four levels of nested continua used in our adaptation of the literacy continua model (see below): literacy context, interactional space, literacy content and literacy development. The same fourfold structure is used in the Discussion section, which also contains a subsection on decolonising literacy practices and their effects on classroom dynamics. Finally, the Conclusions section addresses the research questions and points to different strategies in order to work towards an inclusive and sustainable model of literacy education.

1.1. Literacy Reservoirs

Drawing on Bourdieu’s idea of cultural capital [26] and González, Moll and Amanti’s “funds of knowledge” [27], the term “literacy reservoirs” [28] refers to a series of literacy-specific resources that learners bring into the classroom from a variety of experiences (formal and informal, in interaction with other people, objects, spaces), in different contexts (in and outside the school setting, the home, the neighbourhood, in digital environments) and with a variety of agents (teachers, other learners, family, friends, the community). These reservoirs include learners’ communicative resources, cultural legacies, identity-building and identity-affirming attitudes, disposition for meaning construction and negotiation, paralinguistic strategies, expressive mechanisms and mediating tools. Since literacy reservoirs are largely dependent upon each learner’s context, culture, social group and personal experience, they are both manifold and variable. In contexts with clashing literacy cultures, literacy reservoirs originating in the subaltern cultures are rarely intentionally activated, their usefulness for accessing literacy is often disregarded, and they are not assessed. However, these resources are necessary so that learners may establish links between the literacy cultures that come together in the classroom: between oral and written traditions, non-verbal and verbal forms of communication, and informal and formal literacy practices. Indeed, “activating and building students’ background knowledge” [24] (p. 559) has been posited as a key element of effective instruction leading to academic achievement.

In the specific context of this research, learners’ literacy reservoirs mostly originate in their gitano experience and culture and are inextricably connected with oral traditions (for example, mastery of the complex rhythms and linguistic structures of flamenco, expressions of folk wisdom and understanding of local stories); non-verbal communication (expressiveness through body language); personal and emotional relations (strong sense of identity and belonging to the community, family memories, close friendships); out-of-school and out-of-home contexts (communal areas in the neighbourhood such as streets and squares, the homes of other members of the community) [6,7].

1.2. Literacy Continua

Hornberger and Skilton-Sylvester put forth a theoretical framework, the continua of biliteracy [29], for the analysis of literacy practices in bilingual contexts. This framework is a useful tool for understanding literacy practices in contexts with two literacy cultures too. It has been contended above that in cross-class, cross-ethnic and cross-cultural settings, learners have to navigate between differing literacy cultures as they would between languages [30]. For the purposes of this study, Hornberger and Skilton-Sylvester’s model has been adapted to fit the specific context of our research and the literacy practices that occur in it. In this way, the following levels of nested continua serve as the framework for our analysis: literacy context, interactional space, literacy content and literacy development. Within each level, different categories are established: the literacy context can be oral, intercultural and refer to the minority context, or it can be literate, monocultural and refer to the majority group; the interactional space can tend towards fluid, multidirectional and constructed interactions, or towards predetermined, unidirectional and conducted interactions; the literacy content can refer to minority and vernacular subjects that are contextualised, or refer to majority, canonical and decontextualised contents; finally, literacy development can be based on oral, non-standardised and communal processes, or on written, stardardised and individual processes.

One interesting aspect of this framework is that the categories considered within each level are not presented in dichotomous or absolute terms, but rather as continua: depending on the level and category, literacy practices can occupy different spaces along the continua. The level “interactional space” has been included to incorporate a post-materialist concept of space as an interactional and relational construct [9,31,32] that both determines and is determined by practice-interaction. In this way, the interactional space that literacy practices build can be an emancipating space, marked by fluid and multidirectional interactions where meaning is negotiated and co-constructed; or a colonised space, characterised by predetermined, unidirectional and conducted ones. This framework allows for a richer analysis and understanding of literacy practices and provides an insight into how literacy practices in the classroom can be rethought and combined to ensure that they do not colonise but rather emancipate learners, guiding them through the process of navigating differing literacy cultures.

Another interesting aspect of this model is that it draws attention to the status and power traditionally ascribed to certain literacy practices over others in educational institutions. In this way, practices associated with the majority or dominant culture have traditionally been regarded as more relevant in the school context and therefore used as mechanisms to establish and perpetuate power asymmetries between literacy practices and cultures, turning teaching into a colonising enterprise [33]. In this regard, prior research has shown that colonising practices imposed on students from marginalised groups “often reinforced the broader societal patterns of exclusion and discrimination” [24] (p. 562).

2. Materials and Methods

The study was carried out following an ethnographic approach [34], where the focus is on “naturalistic modes of inquiry that privilege the local, and particular, with an attention to unfolding events captured through detailed naturalistic methodologies” [35] (p. 3). The reason for using an ethnographic methodology is that it provides a better understanding of everyday literacy practices, as opposed to psychological or cognitive approaches to literacy. Figure 1 presents the main stages of our research:

2.1. Participants and Building Trust

The sampling method was the selection of critical cases [36,37], which allows for maximum variation sampling when dealing with a small number of cases. In this way, we sought out to work with a group of 14 primary school learners aged between 5 and 7 attending the same classroom in a school in a low-income area in Seville, the capital city of the Southern region of Andalusia, Spain. Whereas the predominant culture in the area is Roma, the majority culture of the institution and the city at large is non-Roma. The learners themselves were Roma or had strong ties with the Roma community; their age and the complexity of their lives in a marginalised context were relevant to our research. Many parents were not fluent readers and writers in the L1 (Spanish), and only one of the mothers had completed compulsory secondary education. There were no books, pencils or notebooks in the homes, and written texts had little relevance in their daily lives.

We became involved with this group of learners through a project funded by the Spanish Ministerio de Ciencia e Innovación. Eduardo was the most experienced researcher, with ample experience in fieldwork and research methodologies. Macarena had a lifelong interest in biliteracy and had previously worked as a school teacher. Yiyi was invested in critical theory and had prior experience teaching Roma children in English immersion summer camps. In spite of this, our first contacts with the community and the learners were challenging. The area is geographically separated from the rest of the city by train tracks and roads, and is accessed via a tunnel. The tunnel functioned as both a physical-real and metaphorical-ideological barrier, foregrounding issues of othering and coloniality [38]. Indeed, although the gitanos are Spanish citizens, they share many of the experiences of colonial/postcolonial communities, such as segregation, persecution and displacement [39]. Despite differences in context, we found commonality in research carried out in the light of postcolonial theory in indigenous communities [16] and communities emerging from diasporic movements [40] and migration [27]. As a result of the colonizer–colonised tensions symbolised by the tunnel, trust took time to build. It has been noted that ethnographic research can aid establish long-term relationships of trust [16], and this happened initially through the school and social worker; with the children, trust was progressively built through shared classroom time, a lot of which was spent playing, conversing and helping them in classroom activities. The written word seemed to be a great source of unease and distrust, so we took steps to decentre it from classroom practice.

2.2. Researcher Positionality

Our research was informed by critical ethnography and postcolonial theory. It was critical because it intended “to address processes of unfairness or injustice within a particular lived domain” and “to make a contribution toward changing those conditions toward greater freedom and equity” [41] (p. 4). Additionally, because by problematising dominant discourses of what literacy is, what it looks like and what manifestations of literacy are institutionally valid we were also critiquing our own epistemological certainties regarding literacy and our research stance [41].

Following on from Spivak [42], we set out to address the question “Can the subaltern speak?” as we explored what learners did with language, the tools they were given and the ones they brought with them. The answer became apparent at an early moment in our research when a learner asked for the “flesh” crayon and was immediately handed a bright pink one. If they could not fully represent themselves with that crayon, how, we wondered, could they speak with our language? Even if they could, would we (the researchers) listen? The question of re-presentation as it interweaves with notions of colonial/postcolonial identity is central to Spivak’s thought, and helped us frame both our research and practice: first, by making us constantly interrogate our position, theories and hypotheses when encountering and interpreting othered manifestations of literacy; second, by impelling us to carefully consider ways in which voice and agency could be restored.

2.3. Data Collection

The fieldwork took place between March 2018 and February 2019 and was carried out in two stages: on the one hand, the exploration stage (March 2018 to October 2018) was intended to shed light on the literacy practices in which learners engaged both outside and inside the school setting. This made it possible to address our first research question regarding the connections between the literacy contexts, interactional spaces, contents, and practices both outside and inside the school setting. It also allowed us to do the diagnosis of the literacy reservoirs that were frequently activated in learners’ out-of-school literacy experiences, but that were left out of regular classroom practice (regular classes in the L1, the English language classroom, the music class, school radio sessions, etc.), and to design the intervention. On the other hand, the intervention stage (October 2018 to February 2019) was intended to include communicative practices closer to learners’ out-of-school experiences as part of the regular development of English language instruction. The intervention allowed us to answer our second and third research questions regarding communicative practices’ capacity to activate learners’ literacy reservoirs and enhance their literacy engagement, as well as the effects of these practices on the dynamics of the classroom community.

Eduardo visited the neighbourhood and the school once a month during the exploration period, obtained the parents’ informed consent, and was in contact with the institution and the community. Eduardo collected data from the out-of-school context through ethnographic walks, structured and unstructured interviews with the headteacher and the social worker and observations of out-of-school literacy practices and events (parent–child interactions, social gatherings, etc.). Data collection from the out-of-school context delivered 5 reports, 25 photographs and 3 audio/video recordings. Macarena visited the school once a week during the exploration period, collected data through participant observation in a variety of in-school settings (including the L1 and English language classrooms and recess), held structured and unstructured interviews with the L1 teachers and conversed with students and participated in different activities (the school radio, reading events, etc.). These rendered 35 reports, 800 photographs, 14 audio/video recordings and 853 written productions, artefacts and drawings. Finally, Yiyi visited the English language classroom once a week during the exploration and intervention stages, collected data through participant observation, held structured and unstructured interviews with the English teacher and carried out the intervention. Data collection during the intervention produced 17 reports, 87 photographs, 47 audio/video recordings and 347 written productions, artefacts and drawings. All the data were collected following the Mosaic approach: it is multi-method (it recognises the different ways in which children communicate), participatory (data are collected in collaboration with children), reflexive (it includes a variety of agents reflecting on meaning and meaning construction), it is focused on children’s lived experiences and embedded into practice [43] (p. 7).

The intervention was carried out in the English language classroom, which was chosen as the preferred in-school setting due to its focus on orality and the use of communicative methodologies and techniques [44]. Our interest in the English language classroom also responded to recent claims that, in Spain, English language learning, otherwise seen as one of the keys to academic and professional success, is acting as a segregating subject for Roma learners [45]. In order not to disrupt the learning environment, the intervention combined features of learners’ out-of-school practices with others more in keeping with traditional classroom practice. A series of methodological premises underlay the intervention:

• Communication: collaborative creation and negotiation of meaning in free interactions.
• Orality: reception before production [46], only English input and flexibility regarding output (including non-verbal and paralinguistic strategies).
• Linguistic experimentation and play.
• Learner-centred and experience-based approach.
• Music, songs, dance, arts and crafts, life stories, storytelling, realia and multimodal materials, identified as appropriate in decolonising practices in primary English language teaching [16].
• Teacher/researcher as prompts/materials provider, instigator of communication and interlocutor.

2.4. Data Analysis

Data collected during the exploration and the intervention periods was analysed using a contextual adaptation of Hornberger and Skilton-Sylvester’s continua model of biliteracy [29]. This framework, applied to a context with two differing literacy cultures, guided the analysis, pointed to relevant contextual conditions to be described, analytic priorities as well as interpretations to be considered. Thus, using theory as a tool, we explored and traced the connections and disconnections between learners’ literacy practices both outside and inside the school setting regarding the four levels of nested continua: literacy context, interactional space, literacy content and literacy development (see Figure 1). This analysis allowed us to determine whether the in-school and out-of-school practices gravitated towards opposite ends of the continua (the case during the exploration) or occupied similar spaces along the continua (the case during the intervention).

Data collected during the intervention was organised and analysed using the following criteria: we determined that learners’ literacy reservoirs had been activated (1) when learners used literacy-specific resources that had not been observed in the school setting during the exploration period, (2) in practices that occupied similar spaces along the continua to those of the out-of-school context and (3) resulting in engagement on the part of the learners. Engagement was determined by learners’ verbal and non-verbal reactions reflecting one or more of the following: interest (in the practice); relevance (of the practice to learner’s life); expectancy (of success); satisfaction (in the outcome) [47] (p. 113). It must be noted that not all the practices had the same effect on all the learners: not all the learners became engaged in all the practices, and not all the practices delivered a product or outcome. Different practices appealed to different sensibilities and tastes, activated a variety of reservoirs and engaged different learners at different times. A nuanced approach including various reactions and levels of engagement are captured in our narrative below. Our aim was not to find patterns but rather to adopt a “baroque orientation” [48] (p. 260), acknowledging and embracing the complexity of our task. Finally, our findings led us to some reflections about the effects of these practices on learners’ construction of their literate identities as learners in a non-Roma institution and on the social configuration of the classroom space.

3. Results

3.1. Literacy Context

Learners’ out-of-school literacy context was made up of friends, family and the gitano community at large. An important part of the literacy practices observed in the out-of-school context happened orally: extended families got together and talked among themselves, teased and told each other stories, sang and danced. Another important part happened through TV, mobile phones and videogames, to which learners had access with other members of the family. In this context, learners’ literacy practices were mostly mediated by oral language and the body; the gitano culture and experience was central to their oral exchanges, whereas non-Roma cultural elements made their way into their practices through ICT and what they heard from adults (Interview with social worker 23 May 2018; Observation 7 March 2018, 5 June 2018, 20 June 2018).

As opposed to this, the school reproduced the literate context of the non-Roma majority culture and operated as a kind of liaison between the education system and the community. Within the school setting, access to information was, to a large extent, mediated by the written word: literacy was based on the development of reading and writing skills using pencil and paper, official curricular content and non-Roma culture and practices. For example, paper had great symbolic power in the classroom, making learners conform to literacy practices that differed greatly from those of the gitano community. This situation often gave rise to contradictory behaviours and attitudes towards literacy: there was an inherent contradiction between the soothing effect that worksheets had on learners, who even demanded this kind of written work, and their rebelling against the written word through lack of interest, tantrums and unwillingness to complete the tasks. Before the intervention, we found learners who had already assimilated the concept of work and literacy of the majority culture: while playing a game in groups, one student insistently asked when we were going to “work.” Indeed, being able to carry out literate practices was considered the most preferred outcome of most classroom work (Interview with English teacher 14 November 2018).

During the intervention, the context of the English language classroom relied primarily on the reception and production of oral input and output, as all the instructions, interactions and classroom practices happened orally. The written word had a rather marginal role in all of these and was only used to accompany oral language when required by the communicative context, or on students’ demand: for example, one student showed the researcher his favourite football team by doing a small graffiti and stamping the name of the team on it (Intervention 5 October 2018). The context of the English language classroom was also an intercultural one: while the large dominant culture of the institution served as a physical container and institutional frame for the other cultures in the classroom, the foreign language (FL) culture and the learners’ culture became non-dominant cultures represented by human agents engaging in communicative interactions, exchanging anecdotes, singing, dancing and clapping to flamenco rhythms. Both the FL and the learners’ cultures were made to interact and, in those interactions, it was the learners, rather than the researcher, who acted as mediators. For example, one of the most withdrawn students insisted on teaching the researcher her own variety of Spanish so she could communicate with the other students; another, more confident, student clarified some instructions for his peers; one previously unengaged learner explained the rules of an online game to his peers and yelled “Wait!” to indicate they had to wait their turn (Intervention 5 October 2018, 30 November 2018, 6 February 2019).

3.2. Interactional Space

In the community, the interactional spaces constructed by learners themselves or in conjunction with other interlocutors flowed quite naturally: conversations with other children were initiated by learners themselves; had a familiar register; were multidirectional; did not follow any pre-established textual conventions (except for basic politeness and turn-taking rules, often stymied by learners’ familiarity and interruptions); relied heavily on non-verbal communication; focused on a variety of topics of interest to them (football, music, family anecdotes, gossip, etc.) (Observation 7 March 2018, 7 May 2018, 5 June 2018, 20 September 2018). As opposed to this, the interactional spaces found in the school before the intervention were quite rigid: the rules that regulated classroom interaction and behaviour followed school policy were clearly established in the classroom (sometimes in written form) and used to discipline or reward students accordingly. Classroom interactions were initiated and conducted by the teacher, unidirectional (teacher–student) and based on a question-and-answer structure. Group work was rarely encouraged; when it was, the focus was on competition rather than collaboration (Interview with English teacher 14 November 2018), which often resulted in the more engaged members of the group monopolising the competition while the less engaged ones stopped participating.

During the intervention, interactional spaces reproduced some of the features of the interactions that happened outside the classroom: they were initiated by students themselves and were allowed to unfold quite freely; they were multidirectional and often involved several interlocutors who were free to join and leave the conversation; they were about taboo or “un-academic” topics (personal issues, how they felt, etc.); they had a familiar register and high tolerance to interruptions; body language (especially gestures) played an important part in them. For example, during a storytelling activity, fake yawning, lying on the floor and pretending to be sleepy was learners’ way to indicate that they wanted to do something else, marking the transition between different interactional spaces. This transition was used by the researcher to do a whole routine on onomatopoeia around sleeping and waking times in which learners engaged actively (Intervention 1 February 2019). Prompts (verbal, visual or aural prompts, realia or even a change of physical space) were used to lead students to initiate many of the interactions that took place in the classroom, and then learners were in charge of self-regulating their practices and interventions: for example, they decided when they were finished talking or when they wanted to change topics, switch activities or places (Intervention 6 February 2019). That is how the interactional space was built. Although teacher–student interaction was common (learners were not used to interacting with each other during classroom activities), student–student interaction was also used: for example, learners were encouraged to ask and tell each other about their families, stories, feelings, etc. (Intervention 25 January 2019, 1 February 2019, 13 February 2019). Group work was based on collaboration rather than competition, which contributed to learner engagement and the creation of new interactional spaces.

3.3. Literacy Content

The content of the literacy practices in the community was based on the gitano vernacular culture and language. On the one hand, the gitano culture was deeply set and widely accepted by the community, and transmitted orally through the examples set by the most influential members of the community (Interview with headteacher 23 April 2018; Interview with social worker 23 May 2018; Observation 7 March 2018). On the other hand, the vernacular language of the gitano community brought together features of Andalusian Spanish (Spanish spoken in the South of Spain, with a series of distinctive phonetic and morphosyntactic features) and Caló (the language of Spanish Roma). Although an endangered language, Caló was known to some members of the community, and some of its vocabulary made it through everyday speech in the homes: learners were exposed to and used Caló words (Interview with English teacher 20 November 2018; Observation 7 May 2018, 20 September 2018). Furthermore, learners’ oral language followed the patterns of their own variety of Andalusian: dropping or aspiration with gemination of final consonants; elision of intervocalic/d/; aspiration of/x/; neutralisation of/l/and/r/before consonant; neutralisation of/s/and/θ/; changes in the quality of some vowels. All this attested to a big gap between their oral language and the written language they were taught at school, deepened their difficulties to access the written word and often resulted in what can be construed as a failure to reach the school’s standards. For example, when asked to write down what they wanted to be when they grew up, learners reproduced the patterns of their oral language: pulisia instead of policía (police officer), with vowel raising and neutralisation of/s/and/θ/; cantao instead of cantaor (flamenco singer), where the final consonant has been dropped. Indeed, a way in which students learned these words and patterns outside the school setting was through songs: a student that found it difficult to string several words together in a sentence could easily sing long songs using complex syntactic structures and a considerable range of vocabulary of relevance to the gitano culture and experience. Most literacy practices that took place outside the school had to do with the here and the now of this experience: covering basic needs, social interaction within the community, entertainment and the socio-political climate (Observation 22 May 2018, 5 June 2018, 20 June 2018, 24 September 2018).

In the school, contents were often standardised and belonged to the majority culture, failing to consider learners’ culture, identities and experiences. Although special occasions and events were used to contextualise contents, lack of planning or adequate materials sometimes undermined the efforts to include learners’ culture in the classroom: for example, the visuals used for the topic of the family featured a standard white middle-class family that differed greatly from learners’ own families. Furthermore, the development of reading and writing skills often led to the neutralisation of the identity traits of learners’ own language variant (considered less prestigious and without a written code) and the “de-gitanisation” of learners’ linguistic identity: one learner wrote bailadora instead of bailaora (flamenco dancer), which is not only an instance of hypercorrection where intervocalic/d/has been restored, but also a non-Roma term referring to a reality that is also non-Roma; also, learners who could read and write sometimes ridiculed peers for sounding “too Roma.”

During the intervention, the students’ culture, as well as the here and the now of their experience, were brought into the classroom and made to converse with more conventional FL contents. In this way, in order to recontextualise content, previous work done on standardised family vocabulary was complemented with the creation of learners’ own family trees, which they also presented in front of their peers. This offered them an opportunity to share their family history, and to decide how they wanted to tell it: they were given the freedom to choose the layout and the elements that they considered central to their family history. For example, one learner placed the grandfather at the roots of the tree, showing the pivotal role that he played in the family; another student placed his baby brother next to the mother; two learners included their dogs within the family structure; finally, guitars were used as key organising elements in two learners’ family trees (Intervention 19 December 2018, 21 December 2018). Since these opportunities for negotiation did not occur often in the regular English language classroom, they were also grounds for some confusion: an otherwise sharp learner, who struggled to understand that her family was the focus of the activity, ended up copying the words in English from the flashcards on the wall, which was indicative that she had assimilated the school’s concept of learning as passive reception and reproduction of knowledge; as opposed to this, other, less confident, learners found the task easy to understand (Intervention 28 November 2018, 19 December 2018).

Family members’ nicknames were also a hot topic in the classroom: learners enjoyed revealing their families’ nicknames and knowing their peers’ too. This allowed them to explore the differences between oral and written genres, as well as sociolinguistic issues: a previously uninterested student revelled in the act of sharing the nicknames of the people in her family, and finally decided on the official names to go in the family tree; as opposed to this, other students included the nicknames, sometimes using phonetic spelling according to their own language variant; finally, other students wrote the words in Spanish, arguing that those were the names they knew these people by (Intervention 28 November 2018, 16 January 2019, 18 January 2019).

3.4. Literacy Development

Literacy development outside the school setting was developed as a communal event involving different people: children learned from adults and other children through casual chatting, storytelling and songs, sometimes in public open spaces. Although these practices were not standardised, the aim was for children to develop the necessary tools to become respectable members of the community (Interview with headteacher 23 April 2018; Interview with social worker 23 May 2018; Observation 7 March 2018, 5 June 2018, 20 June 2018). Additionally, different members of the family often watched TV shows together, shared mobile phones and played videogames; as older siblings and relatives were asked to take care of children, literacy development was not in the hands of only one person or caretaker (Observation 12 June 2018, 20 June 2018).

In the classroom, literacy development before the intervention was mostly an individual process. All intentional work was done through worksheets that learners had to complete on their own, and assessment was based on their capacity to do so, rather than on their ability to use language communicatively. Since most learners could not read and write in their L1, the writing part of the worksheet was rarely completed. Furthermore, the process through which literacy was developed in the school was standardised: the same method was used for all the learners, regardless of their capacity, needs or interests, and orality played no part in this intentional process. Teachers used an adaptation of the global method combined with sign language. As literacy development relied heavily on visual memory, doing the same activity over two different sessions led to learners’ self-distrust: for example, one student could not remember what she had written down the day before and asked several peers to confirm that it was her name written down there. This is indicative that students found it difficult to retrieve information that was only registered in writing.

During the intervention, literacy development was mostly based on communicative principles. Because learners exhibited different levels of confidence and communicative competence, literacy development adapted to these, rather than to pre-established standards. Even if learners’ output in English was very limited, the English language was used communicatively in the classroom. This met some resistance on their part, and it was apparent they struggled to understand sometimes (Intervention 21 December 2018, 18 January 2019); however, communication was made possible by learners using gestures, onomatopoeia and objects (Intervention 14 November 2018, 16 January 2019). At this level, producing one or two words communicatively using English rather than Spanish sounds was considered successful L2 development. For example, one student replied “OK” and “no” (using diphthongs instead of short vowels) when asked a question; another, less confident, student could answer questions reproducing words he had just heard: “Yes, a baby” (Intervention 14 November 2018, 19 December 2018). When a real communicative context was set, several students could say set phrases spontaneously, such as “thank you,” “I love you” and “how are you?” and use isolated English words (“yes,” “mum”) to answer questions (Intervention 30 November 2018, 19 December 2018, 21 December 2018).

Since literacy development was based on communication, it was, for the most part, a collective and shared process. For example, while doing oral storytelling, learners who had not been interested in pencil-and-paper activities sat cosily on the floor and used their coats as blankets, and listened attentively. When it was a guided activity where they were encouraged to explain what was going on in the story or name what they could see on the page, they responded eagerly: previously unengaged learners used three-word phrases as accurately as other students previously identified in the classroom as “more advanced” (Intervention 18 January 2019). Another quiet and unengaged learner made spontaneous one-word comments about what she could see on the page even when a question had not been asked: for example, she said “happy” when she saw the monster’s smiley face (Intervention 1 February 2019). Learners were also encouraged to tell each other stories: when telling each other about their families, students who did not generally produce language in the classroom could name the members of their families in English; when using story cards to tell a story, learners made use of a variety of linguistic and paralinguistic resources (pointing at specific objects on the card, gestures and onomatopoeia), and even proto-sentences (“Nessy family hungry”); when doing shared storytelling, a previously uninterested student corrected another one (“dad” instead of “pap”) (Intervention 25 January 2019, 13 February 2019).

Digital devices and multimodal materials were also used for literacy development: previously disruptive and unengaged learners became engaged in online interactive games and activities, for which they needed to receive and produce a wide range of vocabulary (animals, parts of the body, colours, etc.) while meeting the interactional demands of a real communicative situation; they also worked together and helped each other to complete tasks online (Intervention 1 February 2019, 6 February 2019). Audio and video recordings were used to develop and assess literacy development. These were taken seriously by most learners, who were concerned about what they said and sounded like in the audio recordings, and their appearance and performance in the videos; however, the learners who had assimilated the school’s concept of literacy as mediated by pencil and paper tended to minimise the value of these recordings in the school setting (Intervention 21 December 2018, 19 December 2018).

Other practices introduced to develop literacy included activities that further built on the learners’ literacy reservoirs. For example, practices based on phonetic experimentation and “pretend English,” where learners played with sounds to negotiate meaning: one student said/ju:’glu:/to see if that meant something in English (Intervention 1 February 2019, 28 November 2018, 18 January 2019). Additionally, literacy-through-the-body activities: for example, while playing a game, an otherwise uninvolved student became very motivated when she could guess the answers through lip-reading; another student used dance to complement the information on her family tree (Intervention 18 January 2019, 21 December 2018). Finally, practices where learners could activate affective relations, build on their prior knowledge and use their intuition: a rather withdrawn learner engaged in a very animated conversation when asked why another student was upset; a previously uninterested student could identify the members of a family and provide arguments to support his claims (based on the quality of the photograph, age, likeness, etc.) (Intervention 25 January 2019, 23 November 2018). It was precisely the learners previously identified as “advanced” who initially struggled to activate this kind of intuition: they immediately answered “I don’t know” and looked at the researcher and the flashcards on the wall for the correct answer (Intervention 23 November 2018). This is indicative that they saw the teacher/researcher as the person in charge of literacy development in the classroom and that their way to access information was mostly through linguistic means.

4. Discussion

4.1. Literacy Context

This experience shows that, when literacy is considered a socially situated practice [8,10,49], contextual factors are central to understanding the literacy cultures and practices that come together in the classroom [3,27]. Furthermore, this study shows that the English language classroom becomes a perfect site for the exploration of the different literacy cultures in which learners are immersed and to establish connections between the out-of-school and the classroom contexts. This is so because the FL classroom occupies a sort of in-betweenness, a space that has been claimed to be advantageous for the construction of biliterate identities [33]: it is a pluricultural and plurilingual space [50] where at least two cultures and two languages come into contact and contribute to the construction of learners’ pluricultural and plurilingual identity [51]. As such, it should be a space constructed on the adequate diagnosis of the literacy languages and cultures that are involved, the tensions that may arise between them and the ways in which these can be negotiated through the activation of learners’ literacy reservoirs.

During the intervention, the large dominant culture of the educational institution was represented only by “non-human” elements [52], such as the classroom itself and the display of classroom elements (desks, blackboard and other resources), whereas both the FL and the learners’ cultures were represented by human agents and embodied in “more-than-human” elements [52] that were given special relevance in classroom practices: agents’ voices, bodies, attitudes, gestures, singing and dancing, verbal and non-verbal engagements with other agents in the classroom. Just as it happened in the out-of-school setting, the English language classroom put these elements at the centre of communication. In this way, the intervention was intended to transcend the institutional framework of the dominant culture and the traditional construct of “the classroom” as a purely physical space [32] by giving relevance to the less powerful, non-institutionalised and “more-than-human” elements in the cultural configuration of the classroom. This foregrounded the “immaterial” nature of literacy [53]. Additionally, it helped encourage the use of strategies for intercultural contact (for example, the use of paralinguistic strategies, body language, music, clapping) and foster cultural mediation (when learners spontaneously acted as teachers and interpreters) [50]. Finally, oral texts dominated both the English language classroom and the out-of-school context, as the traditional oral/written divide between everyday classroom interaction and the more specific language of school work disappeared.

All this contributed to the creation of a safe space for cultural- and self-expression, which has been proven helpful to deal with “identity tensions” in classrooms with different literacy cultures [12] (p. 160), and to the destabilisation of institutional discourses of power and competence that accentuated inequalities among different learners and cultures in the classroom.

4.2. Interactional Space

The interactional spaces created in the English language classroom during the intervention brought into the classroom aspects from learners’ out-of-school practices. As stated above, immaterial and more-than-human elements were central to the configuration of new interactional spaces in the English language classroom: there, space became an interactional and relational construct that both shaped and was shaped by the literacy practices in which learners engaged. This draws on the idea that space is the result of the interaction and relations among places, agents and objects [9,31]: in this way, new and different spaces were created as new interactions and relations unfolded. Although there were some pre-established activities, their structures were quite flexible and they were often open-ended, allowing for a variety of literacy practices to take place and a variety of literacy spaces to be created. There was not a fixed time for activities; they lasted for as long as learners were engaged, they could be interrupted and they generally developed quite naturally into different practices, following humanistic approaches to language teaching [54] and baroque approaches to literacy enquiry [48]. Through the use of prompts (verbal, visual or aural prompts related to topics of interest to learners, realia or even a change of physical space) that different learners could find appealing, learners were encouraged to initiate, take up and drop interactions as they saw fit. Learners’ involvement and the number of interactional spaces that activities gave rise to were considered indicators of successful “literacy engagement” [24]. As in the case of onomatopoeia for sleeping and waking up described above, these indicators gave us information about the literacy reservoirs that prompts activated in different learners, which were used to trigger further responses and engagements.

All this bears out the diverse and organic nature of the interactional spaces that were constructed in the English language classroom during the intervention and helps shed light on the importance of opening up spaces for learners to develop agency to decide on what to do with and say about literacy in the classroom, and how to do it and say it, all of which are central to negotiating their literate identities [3,12,13,14,25].

4.3. Literacy Content

This experience shows that content selection for literacy development must take into account learners’ interests, culture, background and experiences, it must be contextualised and include non-dominant sensibilities as well as learners’ vernaculars. This has been attested to by studies engaging in “real literacies” pedagogies [11], which not only imply the use of authentic texts and real uses of the FL in communicative contexts, but the use of texts that draw on learners’ everyday experiences, practices in which learners already engage outside the school, topics that learners understand and are interested in, and learners’ vernaculars [11].

Combining the principles of both constructivist and humanistic approaches to language teaching [54], the intervention relied primarily on contextualisation by making learners’ life stories (who they are, what they like, what they do outside the school), families (where they come from), culture (how they see and do things) and vernaculars (how they say and write things) academic content. Indeed, the English language classroom became a space where the traditional vernacular/canonical divide was blurred for the sake of communication, which is favoured by natural production rather than unnatural reproduction. This helped learners negotiate their identities as members of the gitano community, as learners in a non-Roma institution and as text creators. Only by incorporating these aspects and acknowledging their relevance in students’ learning experiences as well as their effects on how they approach learning in an alien institution may literacy practices become more emancipating for these learners.

4.4. Literacy Development

Since learners were not independent readers and writers in their L1 yet, and in order not to interfere with literacy development in the L1, all literacy work in the English language classroom during the intervention was based on the development of oral language. The oral foundation of literacy has been long acknowledged in the literature, which has drawn attention to ways learners’ out-of-school oral practices can help enrich formal literacy experiences [55]. In the context under study, learners were exposed to and participated actively in a rich socio-cultural environment afforded by families and friends that got together and engaged in communal practices that rarely find their way into the classroom [56]. Replicating some of these helped to draw and build on learners’ literacy reservoirs and contributed to learners’ literacy engagement and development:

• Literacy through stories: storytelling offered great opportunities for learners to produce language through comments and retellings; to put their paralinguistic resources to use and to engage with the written word. Storytelling has proven an engaging and useful tool to develop literacy while negotiating the tensions between differing literate identities [57].
• Literacy through ICT and multimodality: when online interactive practices were introduced, learners used their out-of-school literacy experiences in the classroom, and their concentration span increased. Similar outcomes have been found in other studies [9,31]. Additionally, the use of image and/or sound to aid communication has been shown to contribute positively to literacy development [12,58]. Finally, the use of recordings and videos for assessment purposes foregrounded the intersections of literacy, multimodality and aesthetics in the production (and broadcasting) of a literate identity [59].
• Literacy through phonetic experimentation: phonetic experimentation and “pretend English” allowed learners to explore their own communicative capacity by playing with, manipulating and stretching the communicative possibilities of sounds. This type of practice helped them develop “a feeling of ownership of the language” [12] (p. 155), without which literacy development becomes an alienating process for students that do not belong in the dominant literacy culture.
• Literacy through the body: whereas learners showed little interest in pencil-and-paper activities, they became engaged in activities where they could use gestures, dance and play with space to communicate. In this way, literacy work was not associated with a specific place (the desk) or specific materials (pencil and paper), but with learners’ own bodies. The implication of this was that literacy accompanied them, was within them, while they danced, listened to a story or read someone else’s lips, and hence the power of these practices in the shaping of learners’ literate identities [9,31,52,60].
• Literacy through intuition and affect: learners brought into the classroom information, expectations and intuitions about texts and communicative situations of which they had previous experience [60]. Additionally, personal and emotional rapport with others became a key factor in practices where learners worked together as part of a team and built texts together [3,61,62].

One last relevant aspect of literacy development during the intervention was the use of the English language for communicative purposes at all times. This afforded learners the opportunity to appropriate the FL and establish connections with the FL culture, both of which contribute to “positive identity development” [14] (p. 394): “A child cannot feel connected to a language and culture if they do not interact in the language and have sufficient opportunities to interact extended to them. The child’s self-concept and the way they feel in the language is influenced by the types of interactions they experience in that language” [12] (p. 155).

Furthermore, English-only input has been shown to be, in this context, a source of interest for learners, who became intrigued and challenged by the situation and the need to communicate. However, it also met some resistance at first: learners insisted they did not understand the researcher when, for example, telling a story; later, however, they engaged very actively in retelling the story themselves. Such reactions were forestalled by adopting a flexible approach regarding output: learners were not asked but rather encouraged to use English. Thus, our approach was closer to that of natural methodologies [46], as it relied on comprehensible input while not forcing output until learners felt confident. As the literature suggests, sensitising oneself to learners’ varying degrees of confidence is uppermost: translanguaging and code-switching must be treated, not as transitional stages of language development, but as part of the makeup of learners’ biliterate identities, and as useful strategies for literacy engagement and development [30,61,63].

4.5. Decolonising Literacy Practices

As has been shown, the activation of learners’ literacy reservoirs through communicative practices closer to their out-of-school experiences brought about important changes regarding not only engagement but also the social configuration and distribution of power in the classroom space, an effect that has been noted in other recent studies [3,13,14,24]: a number of students previously unconcerned with literacy practices became involved and active members of the classroom community, even standing out in aspects such as communication, meaning construction, comprehension and production. In this sense, Moses and Kelly underscore the opportunity afforded by identity practices to explore “language learners’ appropriation of practices within the literate community and the identity positions available to these learners within their contexts” [14] (p. 394). Along similar lines, Cummins et al. posit identity affirmation and literacy engagement as key aspects in literacy development [24], whereas González draws attention to these practices’ power to unsettle the institutional and social discourses of deficit that circulate in the classroom [25]. Aiming to decolonise classroom practice allows for previously silenced voices to be heard, subaltern cultures to be explored and brandished, and unacknowledged reservoirs to be activated, while learners develop autonomy and agency in the process of navigating literacy cultures.

At the same time, and this is an aspect that previous studies often fail to acknowledge, these practices also had an effect on the learners who had assimilated the institution’s concept of literacy, learning and work, endorsed the values ascribed to those in the majority culture and adopted its dominant discourses and literacy practices. Initially, these learners were confused by and struggled with practices that, they perceived, deviated from standard classroom practices and compromised their powerful position within the group, namely: the inclusion of non-academic subjects as valid interactional content; practices not based on the use of pencil and paper and traditional print-based encoding and decoding skills; open-ended practices and practices with more than one correct answer; communal practices based on collaboration and shared textual creation. In order to minimise the negative effects that these practices may have on learners and their self-reliance, it is necessary to give them key roles in the classroom [12] and to understand that these more powerful dominant practices are also and already part of these learners’ literacy reservoirs. It has also been pointed out that the tensions between colonising and emancipating pulls in the classroom may be resolved by combining practices found along the continua [64].

5. Conclusions: Towards an Inclusive and Sustainable Model of Literacy Education

This analysis and discussion of the contexts, spaces, contents and practices through which learners’ engagement with literacy occurred both outside and inside the school setting has drawn attention to the facilitating role of learners’ literacy reservoirs when navigating between different literacy cultures in the classroom. First, we sought to answer our first research question: what connections are there between the contexts, interactional spaces, contents and practices through which learners’ engagement with literacy occurs both outside and inside the school setting? An analysis using a contextual adaptation of the continua model of biliteracy [29] revealed that learners out-of-school practices tended towards the less powerful ends of the continua: these practices were often oral and communal, based on the learners’ vernacular language and culture, multidirectional and constructed by learners themselves. As opposed to this, literacy practices in the school setting before the intervention were often standardised individual practices based on the written word that represented the values of the majority non-Roma culture; also, they were unidirectional and conducted by the teacher. The connection between the out-of-school and the in-school contexts was made possible during the intervention through inclusive communicative practices that challenged skills-based literacy models, where learners could participate regardless of their literacy level, and practices closer to the less powerful ends of the literacy continua, resembling learners’ experiences outside the school setting. Regarding our second research question (can communicative practices closer to learners’ out-of-school experiences activate their literacy reservoirs and therefore enhance their literacy experiences and engagement in the classroom?), the intervention revealed that different communicative practices appealed to different sensibilities and tastes, and activated a variety of reservoirs, depending on the student: for some learners, it was their cultural legacy and intuition that became activated; others resorted to identity-affirming strategies; others yet negotiated meaning, even in the foreign language, or used paralinguistic and other expressive mechanisms; etc. As learners’ literacy reservoirs were activated and acknowledged as useful in formal literacy practices, their literacy engagement increased: their verbal and non-verbal reactions showed either interest in the practice, that the activity was relevant to their experiences, expectation of success, or satisfaction in the outcome [47]. Finally, our third research question was addressed: what are the effects of these practices on the dynamics of the classroom community? The decolonising practices implemented during the intervention provided a more inclusive and sustainable framework for literacy development, and effected a redistribution of power in the classroom, helping reshape the institutional and social discourses of deficit around these learners. Indeed, these decolonising practices afforded them a higher degree of agency to self-regulate their engagement and contributions in the classroom and to renegotiate their position as Roma learners in a non-Roma institution and as text creators within the classroom community.

At the same time, these practices also had an effect on the learners who had assimilated the literacy practices of the majority culture. Responding to learners’ need (and right) to identify with and belong to the literate community while ensuring that this does not colonise them, obliterating their own sense of identity and belonging in their own community and literacy culture, is a tricky balance and demands a variety of practices and approaches. In this sense, different studies have pointed to the need to avoid “simple” views of literacy in favour of more complex, rich and demanding approaches to classroom processes and relations [48], to use “hybrid literacy practices” [55] that bring together aspects from the different literacy cultures and traditions that coalesce in the classroom and to create and maintain “percolating spaces” [9] between the out-of-school and the in-school contexts. To these, we add the need for an adequate diagnosis and understanding of learners’ literacy reservoirs and how these can be activated in classroom practices. This is, of course, a time-consuming and demanding task, both for practitioners and researchers. In this regard, and since we are dealing with the literacy practices of young learners, a longitudinal approach throughout different school years would be needed to assess more accurately the capacity of these practices to improve learners’ overall literacy development over time and to effect long-term changes to ensure the sustainability of this model. At the same time, further research is needed into the role of schools in safeguarding social and cultural sustainability through inclusive, culturally responsive and responsible literacy practices. This is especially the case of schools where different literacy cultures coexist, a reality that has become more frequent but that teacher training programmes are still failing to address. This would contribute to the design of teaching programmes to help teachers develop critical cultural and linguistic awareness, observe and do an adequate diagnosis of learners’ literacy reservoirs and design practices that draw on these, bridging the gap between their out-of-school and in-school experiences. This will contribute to a more inclusive and sustainable model of literacy education that not only values learners’ vernacular practices and cultures but, most importantly, equips them better to navigate the tensions between personal, local and subaltern practices on the one hand, and the social, global and institutional demands of formal literacy education on the other. This, we believe, should be one of the pillars to achieve the United Nations’ Sustainable Development Goal for quality education, including universal literacy and education for cultural sustainability.