Constructing bilingual input safe spaces at home: rethinking exposure to language from a multilingual and multisite perspective on literacy practices

development of the majority language, and especially in broadening the understanding of the notion of language exposure. Methods: Semi structured interviews with parents and focus groups with children (aged 4–14) were carried out in Munich, Germany, between May 2018 and February 2020. Data from the verbal interactions, in a total of 12h of recorded audio with the parents and 8h with the children, was transcribed and the excerpts were coded in Nvivo. In a ﬁrst moment, a data and theory driven thematic analysis with open data coding was undertaken; subsequently, discourse analysis was used to analyze selected excerpts individually and reconstruct the meaning the participants were attaching to their literacy practices. Three thematic strands were identiﬁed. Results: We make the case that exposure to literacy practices in the majority language is interpreted and (re)appropriated by families and communities to support the development of complex and integrated literacy practices involving the different languages of children’s repertoires. We shed light on how families construct bilingual input safe spaces at home, i.e., spaces where continuities between formal, informal and non-formal literacy practices in both German and Portuguese are assured. Discussion: The results bring a critical perspective on the ecology of literacy practices and highlight the added value of seeing those practices interconnected holistically from a co-developmental viewpoint. Such interconnectedness has a potential positive impact not only in children’s HL but also in the majority language, and possibly in the languages learnt at school.


Introduction
The acquisition and development of Heritage Languages (HL) take place under specific conditions, as HL are typically spoken by communities that have migrated to a new linguistic environment where the dominant language of the host society poses a challenge to HL maintenance and transmission (Torregrossa et al., 2023).In most cases the offer of formal, school-based teaching of HL is restricted, meaning children's exposure to HL use is limited to informal learning environments (such as family, community, and social interactions), so families have to rely on informal learning settings and methods, such as community-led initiatives.
While informal education supports the development of language skills (at least listening and speaking skills, i.e., literacy based on oral competences) and promotes the development of a sociolinguistic identity and the transmission of cultural knowledge, literature has also highlighted that formal education is crucial for the development of literacy (writing and reading in the HL; see Costa Wätzold, 2023, for a review).Reflecting this lack of formal education and development of literacy in the HL, literature has specifically underscored the essential role of literacy practices in the home (Tse, 2001;Caloi and Torregrossa, 2021) and in informal education (Costa Wätzold and Melo-Pfeifer, 2020;Costa Wätzold, 2023) as ways of maintaining HL competencies and developing an identity bound to that language.
Families in general, and parents in particular have an essential role in HL maintenance and transmission (Braun, 2012;Wilson, 2012;Melo-Pfeifer, 2015), providing emotional, cognitive, and interactional support.Families' role in maintenance and transmission of the language is promoted or hindered depending on language ideologies in the host countries, which favor positive or negative attitudes toward (specific) HL.Paradis (2023) highlights the specific role of families' attitudes toward languages in explaining individual differences in HL proficiency, while recognizing a lack of studies on this subject (p.17).Attitude is a positive or negative feeling about some person, object, or issue, acquired through social interaction (Ajzen, 1988).Attitude is the feeling or opinion about something or someone, or a way of behaving that follows from this, and it is not directly observable but rather has to be inferred from behavior.Ajzen (1988) emphasizes that an object predisposes cognitive, affective and conative responses, the sum of which ends up being the attitude toward the object.Attitudes are learnt, which is why parents and education become very influential factors in shaping, for example, children's attitudes toward HL, and their influence is such that attitudes originating in these social milieus are particularly resistant.Other socialization factors to consider are friends, colleagues and the mass media.Lasagabaster (2006) posits that the status of a language, language attitudes and its social functions are closely interrelated topics.Attitudes toward different languages and language varieties and practices reflect perceptions of people in different social categories and how such perceptions influence interaction within and across the boundaries of a speech community.As the concept of attitudes has a high explanatory value in the linguistic and sociolinguistic dimensions, in this study it will be operationalized as a theoretical frame to guide the analysis and understand how families (both parents and the children) see values and social practices linked to HL maintenance.
Another essential aspect in HL maintenance and development is quantitative and qualitative exposure to the HL (Unsworth, 2014;Correia and Flores, 2017;Flores et al., 2017;Paradis, 2023).Exposure, in quantitative terms, refers to the amount of time a child is in contact with the HL, and is measured in terms of age of onset; in qualitative terms, exposure refers to the richness and variety (or lack thereof) of linguistic input, in terms of vocabulary, complexity of syntactic structures, and registers, to give but three examples (Unsworth, 2016).Exposure, even if interacting with other factors such as the occasions to use the HL (Hammer et al., 2014), is essential to understand HL development.Research on exposure has tended to focus more on exposure to a single language (exposure to the HL or, in some cases, to the majority language) or on comparing the quantitative input in both languages (Hammer et al., 2014).Less is therefore known about the interactions emerging from exposure to different languages and in different settings, and how interactions between languages and settings are perceived by both families and children.
As previously stated, attitudes play a role on how families engage in the transmission of the HL.Paradis (2023) recognizes that the role of family's attitudes toward languages is important to explain individual differences in HL proficiency, but she acknowledges a lack of studies on the subject: "Attitudes toward HL maintenance and ethnocultural/ethnolinguistic identities on the part of parents, could, in principle, be a distal environment factor associated with individual variation in children's bilingual development, but this remains a relatively understudied area with respect to [Individual Differences] approaches to child bilingualism" (p.17).In this paper we explicitly address this gap by focusing on parents and bilingual children's attitudes to the majority and the HL.Therefore, this paper focuses on the literacy practices that bilingual families are exposed to and cocreate at home, at school and in a community programme, to understand how these dynamics interact or complement each other.Recognizing the entangled nature of literacy practices in formal (school-based literacy practices) and informal (learning experiences outside the school) education and in several languages, we aim to answer the following research question: RQ: What are parents' and children's attitudes toward the different literacy practices and across the different contexts, and how do they perceive the interaction between these practices and the development of competences in the HL?
We first develop a theoretical section in which we explore the concepts of biliteracy, multiliteracy, and exposure.Based on these theoretical contributions, we present the context of the empirical qualitative study and analyze excerpts from the data collected to explore the limits of the concept of "exposure" when focused solely on the HL.

Theoretical framework
In light of the array of thematic data to be presented in the qualitative study, two concepts will serve as the theoretical framework guiding our analysis.In moving from literacy to biliteracy, we will articulate these two notions with the concept of exposure, which will then guide us in the analysis of the qualitative data.This will enable us to present an extension of the latter term based on a reinterpretation of literacy practices in the context studied.To do this, a key concept used is Hornberger's (2003) biliteracy continuum.In conjunction with this, the concept of exposure, as defined by Paradis (2023), will serve to underpin the theoretical framework that will serve as a lens to describe and interpret the empirical qualitative data.

. From literacy to multilingual literacy
The term literacy stands for reading and writing skills in a broad sense including skills such as text and meaning comprehension, linguistic abstraction, enjoyment of reading, familiarity with books through the competent use of media (Moura and Rojo, 2019).In Webster's dictionary, literacy has the meaning of "the condition of being literate, " and literate is defined as "educated; especially able to read and write."In other words, from an individual point of view, learning to read and write-becoming literate, ceasing to be illiterate, acquiring the "technology" of reading and writing, and becoming involved in the social practices of reading and writing-has consequences for the individual, and alters their state or condition in social, psychological, cultural, political, cognitive, linguistic, and even economic aspects.Implicit in this concept is the idea that reading and writing has social, cultural, political, economic, cognitive, and linguistic consequences, both for the social group in which it is introduced and for the individuals who learn to use it (Soares, 2018, p. 39).The "state" or "condition" that the individual or social group thus acquires, under the impact of these changes, is referred to as literacy.
As Stavans and Hoffmann (2015, p. 268) define it, "literacy practices" is an expression used "to emphasize that reading and writing are activities that take place in specific social and cultural contexts."Even though authors diverge on definitions, a commonality emerges from different accounts of the concept: "literacy is always connected to practice.Such practices are of a social nature and are used with children as a means of socialization.Literacy has an empirical and practical nature and cannot be detached from the contexts in which it is deployed and developed.It depends on individuals, on affordances, on 'ingredients of context, ' on objects under examination" (Helmchen and Melo-Pfeifer, 2018, p. 10).Once the concept of literacy has been defined, the concept of biliteracy needs to be established, since it is not just the sum of literacy in two languages.Hornberger (2003, p. xiii) defines biliteracy as "any and all instances in which communication occurs in two (or more) languages in or around writing, " following from Heath's definition of literacy events as "occasions in which written language is integral to the nature of participants' interactions and their interpretive process and strategies" (Heath, 1982, p. 386).Different from Heath's definition which focuses on the literacy event, Hornberger's definition of biliteracy refers to instances, a term encompassing events, but also biliterate actors, interactions, practices, activities, programmes, situations, societies, sites, worlds, among others (Hornberger, 2003, p. xiii).
The same author presents a model entitled "continua of biliteracy, " which proposes a framework for understanding biliteracy using the notion of continuum to provide an overarching conceptual scheme for describing biliterate contexts, development, and media.Hornberger uses the notions of "intersecting" and "nested continua" to demonstrate the multiple and complex interrelationships between bilingualism and literacy.She points out that although we often characterize dimensions of bilingualism and literacy in terms of opposites, such as first vs.second languages (L1 vs. L2), monolingual vs. bilingual individuals, or oral vs. literate societies, it has become increasingly clear that in each case those supposed opposites represent only edges of a continuum (Hornberger, 2003, p. 5).In order to explain her model, she proposes nine continua around three dimensions: • To illustrate contexts for biliteracy: micro-macro, oral-literate, and monolingual-bilingual.
• To characterize the development of the biliterate individual's communicative repertoire: reception-production, oral language-written language, L1-L2 transfer.
• To characterize the relationships between the media through which biliterate individuals communicate: simultaneoussuccessive exposure, similar-dissimilar structures, and convergent-divergent scripts.
The binaries listed above can be seen as a continuum represented as a line linked by several points.As Hornberger emphasized, it is necessary to focus not only on the two ends of continua but also on all points in between.The notion of the continuum was "intended to convey that although one can identify (and name) points on the continuum, those points are not finite, static, or discrete.There are infinitely many points on the continuum; any single point is inevitably and inextricably related to all other points; and all the points have more in common than not with each other" (Hornberger, 2003, p. 36).According to this model, when we consider biliteracy it becomes clear that these continua are interrelated dimensions of one highly complex system.The important question of how much literacy knowledge and skills in one language can aid or impede the learning of literacy knowledge and skills in another language has been partially answered.Understanding the interrelated and nested nature of these continua allows us to see why there is potential for positive transfer across languages and literacies (Schnoor and Usanova, 2022), since there are a myriad of contextual factors that may promote or prevent such transfer.Hornberger thus highlights that the important point is to recognize what the continua are (a more dynamic paradigm) and understand how they are related to each other (micromacro, oral-literate, and monolingual-bilingual), and concludes that the more the learning context allows learners to draw on all points of the continua the greater the chances for full biliterate development.Our analysis aims to show how the different points of the continuum of biliteracy are perceived by both parents and learners.
. Exposure to language and the development of multilingual literacy Research on the domain of bilinguality has described the HL speaker as unique, with a unique set of competences both in the HL and in the majority language (Valdés, 2001;Carreira, 2004;Carreira and Kagan, 2011).Several factors come together to make each HL speaker unique, from internal (biographic and biologic) and cognitive factors to factors external to the speaker (such as ideologies present in the society about the HL).In particular, exposure to language has been studied as a child-external factor contributing to explain individual variation in HL acquisition (Paradis, 2023).
Bilingual children are a more heterogeneous group than their monolingual counterparts with respect to the sources of variation in their language learning environment and in some of the internal capacities they bring to the task of dual language learning.Bilingual children are learning two languages and so their linguistic input space is divided and can change daily or weekly with respect to the quantity and quality of input and interaction in each language (Paradis, 2023, p. 2).
Paradis (2023) explains that even as an external factor language exposure is dependent upon several proximal and distal factors, which are interconnected, to explain individual variation.These factors include cumulative exposure to the HL and the majority language, degree of use of those languages at home, richness of the HL and majority language environments, literacy and education in the HL, parent proficiency in the HL and majority language, family socioeconomic status, and family attitudes toward HL maintenance, and ethnolinguistic identities and affiliations.
Research shows that the quality and quantity of input is important to explain (and even predict) proficiency in the HL, as well as the distribution of competencies across receptive and productive skills (Unsworth, 2016;Paradis, 2023;Torregrossa et al., 2023).It has been demonstrated that quantity of input usually correlates with quantity of output.Questions related to input quality, such as "who speaks what to whom at home?, " can also be relevant to understand how HL use impacts bilingual development in heritage speakers.The quality of the linguistic environment is related to its richness, meaning "the diverse and complex language children experience through certain activities and interactions" (Paradis, 2023, p. 11).However, quality and quantity of input can be difficult to disentangle: a lot of low-qualitative input may negatively affect linguistic competences in the HL, while sporadic high-quality input may be insufficient to positively impact competencies in the HL.Input quality may be more important to HL development than input quantity (De Cat, 2021).Also, importantly, children can have a high density of rich experiences in both the HL and the majority language.The question therefore arises of to what extent quantity and quality are key in understanding HL development and maintenance, or whether other factors also modulate HL acquisition.Torregrossa et al. (2023) analyzed cross-linguistic influence in HL acquisition as one of those factors, following the premise that "the outcomes of HL acquisition may diverge from those of monolingual language acquisition as a result of crosslinguistic influence from the SL [second language], as has been shown in the domain of reference production" (Torregrossa et al., 2023, p. 180; see also Schnoor and Usanova, 2022).They conclude that bilingual children of Portuguese-German, Portuguese-French, and Portuguese-Italian do not report statistical differences in terms of processing complex items in the HL.However, their study did not cover cross-linguistic influence beyond children's knowledge and exposure to two languages.
Home literacy practices are essential to understand HL maintenance, and include interactive book sharing, language games and maternal responsiveness during such activities (Flores et al., 2022;Paradis, 2023).Nevertheless, research has also made clear that home literacy practices can be insufficient to support the development of skills in the HL, either because they are sporadic or because they tend to naturally diminish with time as children grow to spend less and less time in the parental home, due to integration in other socialization groups (with an exception in the case of large linguistic communities, where children can stay in contact with other speakers of their HL; it is not the case for Portuguese in Bavaria, Germany).For HL development, as with the majority language, richness of input should combine and enmesh home and community factors including "engagement with media (audiovisual and print) and social media, participation in extracurricular activities and cultural events, and language use with friends" (Paradis, 2023, p. 12).
The importance of developing literacy in both languages is that "for heritage bilinguals, by definition, literacy and education in the L2 is taken for granted, but this is not the case for the HL" (Paradis, 2023, p. 13).Because in the scope of our empirical study children and preadolescents are also exposed to a variety of other languages (namely at school), it is important to expand the idea that HL speakers are (emergent) bilinguals of two languages and that young HL speakers develop only literacy in two languages.In Europe, the fact that children start a first foreign language-usually Englishat primary school (starting between the 1st and the 4th year), makes the intersection of literacy practices, languages in contact, and attitudes toward languages much more complex (Usanova and Schnoor, 2021;Schnoor and Usanova, 2022).Importantly, research tends to focus on differences in input in the HL and majority language, thus without considering input in other languages, and mainly taking young children as a research subject (either directly or resorting to parents' reports, the latter method being considered less reliable, according to Torregrossa et al., 2023).
Following Paradis' narrative analysis on exposure (Paradis, 2023), our research is innovative because it considers crosslinguistic and cross-contextual literacy practices that open up spaces for other languages of exposure, namely at school (even if this will not be our main focus of the analysis) and considers both parents' and children's perspectives.The children taking part in the empirical study are not all toddlers or preschool children, meaning a greater age range is considered for the data analysis.

The qualitative study
This qualitative study adopts a within-group design to investigate attitudes and literacy practices, instead of more traditional "monolingual-bilingual between-group comparisons" (Paradis, p.

19). This article presents results from research with
Frontiers in Language Sciences frontiersin.org Costa W ätzold and Melo-Pfeifer ./flang. .ethnographic characteristics carried out with families attending a HL maintenance project called "Mala de Herança" (MH; "Heritage suitcase").The pedagogical proposal is mostly based on reading children's literature as a strategy of HL maintenance: the project therefore highlights the role of reading and storytelling in the development of literacy in the HL. .

Research context
In Bavaria, Germany, the context of this study, the community of Portuguese as Heritage Language (PHL) learners has access to different learning settings: a formal learning context from the courses provided by the Camões Institute (Camões IP), non-formal learning contexts in a variety of grassroots initiatives, and informal learning contexts provided by the family both at home and in other informal settings.Thus, our research context can be characterized as a non-formal/informal HL learning scenario.
The MH project is a socio-educational project initiated in 2012 as a community-based initiative in and around the city of Munich (Costa Wätzold, 2023).MH promotes monthly meetings for Portuguese speaking families, functioning as a community of practice (Wenger, 1998).The project's didactic-pedagogical objectives include supporting families in the transmission and maintenance of PHL, disseminating scientific and objective information about bilingualism, as well as creating opportunities for families to interact in Portuguese.To this aim, the project coordinators make use of didactic resources such as shared book reading, handcraft and musical activities, theater, dance, traditional festivals, and typical cuisine, as well as offering a variety of books that families can take home with them, so that they can continue to have contact with the HL beyond the project.Meetings take place monthly, lasting 3 h, and are free of charge and open to families with Portuguese (as a HL) speaking children, with specific segmentation for different ages.The events are organized and curated by coordinators, in most cases mothers and fathers who volunteer and mobilize other parents by involving them in the organization and implementation of the didactic-pedagogical proposal.These parents, who act as de facto language educators, are not always and not necessarily language teachers.
. Participants, data collection, and data analysis Among the participants (N = 37: 22 parents and 15 children: aged 4-14) are learners who have access to formal, non-formal, and informal education in the HL, while others have access only to nonformal and informal education (N = 14).The parents interviewed Formal learning is by the process of learning in an and structured context leading to a certified qualification.Non-formal learning is similar to formal learning but without certification.Informal learning refers to a learning process that takes place in everyday life, in routine.This can happen at work, in the family circle or in leisure time, i.e., outside of conventional institutions.From the learner's perspective, this may not be desired (European
all had higher education and medium-high socioeconomic status.Most interviewees were mothers, housewives, dedicated almost full-time to raising their children.The families reported using Portuguese and German in their daily lives, and some also used other languages at home, such as English or Spanish.The children who took part in the different stages of the study were born in Germany (with the exception of two who came from Brazil to the country as babies) and are all bilingual (Portuguese-German).Due to the curricular languages, they come into contact with in the German school system, the children participating in this study all have a (potentially) multilingual repertoire.
The data was collected in Munich, Germany, between May 2018 and February 2020 as part of a doctoral research project, with both the supervisor and the doctoral candidate being proficient in Portuguese and German.The primary subjects addressed in the interview with the parents were: (1) the rationale behind their participation in the MH project, (2) the activities conducted at the meetings and the impact of the MH project on the development of PHL, (3) the materials and methodologies employed by the parents to transmit and maintain PHL at home and their evaluation of their children's development in both languages.With the children, the primary subjects were: (1) Their ways to learn Portuguese, (2) Their perception of the MH project, (3) An account of their experience of being Brazilian in Munich.Before the data collection, when parents were invited to take part in the research, they were given a free consent form in which the ethical issues surrounding the research with the children were explained, especially the point that there was no risk to the participants.This form also explained the objectives of the research and provided the researcher's contact details for any clarifications.An ethical approval was not required.Semi-structured interviews were conducted with 22 parents who attended the MH project.Each family was interviewed only once, in some cases both parents, in others only one.The interviews were conducted in Portuguese or German, depending on the parent' preferences, and took between 30 and 90 min, depending on their willingness to provide information.
A total of 15 children (aged between 8 and 14) participated in three focus groups (FG), with drawings used as a starting point for group discussions (Kalaja and Melo-Pfeifer, 2019).The FG format was selected not only for its playful aspect, but also for the possibility of constructing a shared meaning.Three FGs were held on different occasions, with participants of varying ages at each meeting.All three FG were conducted in Portuguese, as all the children had sufficient oral proficiency in Portuguese to participate.Although some of the children occasionally engaged in German-language interactions with one another, they consistently utilized Portuguese orally when addressing the researcher (and in her immediate presence).
Data from the verbal interactions, both with parents and children, were transcribed and the excerpts were initially coded in the original language of the interview (Portuguese).Only relevant excerpts for this article were subsequently translated into English, by the first author, and double checked by the second author.To guarantee participants' anonymity, only their initials are shown in the excerpts from the interviews and FGs.The complete pool of Original Portuguese excerpts are provided in the Appendix.
Frontiers in Language Sciences frontiersin.orgverbal data was categorized based on content analysis (Mayring, 2016).
In terms of codebook, it contains the set of categories related to the theoretical framework we are focusing on in the present article, namely the continua of biliteracy model and the notion of exposure.We first selected the relevant excerpts in terms of content relevance; second, we analyzed them looking for discursive markers that demonstrate how literacy practices in the two languages are reinterpreted and appropriated by parents and their children.
In the first phase of the data analysis, after transcribing the verbal material (12 h of recorded audio with the parents and 8 h with the children) using the software Nvivo, we identified the literacy practices and literacy contexts used to learn the majority and HLs and how these practices and contexts interact with each other.The Nvivo software allows free coding of all verbal material and offers the main advantage of organizing the data into nodes, which is a very practical tool for visualizing themes and rearranging the excerpts into thematic categories as many times as necessary until the essential categories are reached.It is important to mention that, for the present study, in the midst of a verbal material of 100 pages of transcripts of the interviews and the FG, it was possible, with the help of the NVivo nodes, to reduce and organize the data, selecting only the excerpts related to the theme of literacy practices.Therefore, data and theory driven thematic analysis with open data coding was undertaken (Schreier, 2014).In the second phase of data analysis, we adopted an approach based on discourse analysis (Baxter, 2010).Discourse analysis was used to analyze selected excerpts individually and reconstruct the meaning the participants were attaching to their literacy practices, i.e., theoretically driven, looking for thematic reconstruction of "contextually situated" macro discourses.In this way, three thematic strands were identified: the complementarity of development of HL in a grassroot project and at home (Section 4.1); the construction of a bilingual input safe space at home based on literacy practices at school (Section 4.2); and the construction by the families of continuities and connections across spaces and languages for the holistic development of their bilingual children (Section 4.3).Table 1 shows the categories used.

Findings
To answer the proposed research question, we present and discuss a selection of excerpts that allow us to glimpse the literacy practices these families perform to develop the HL, and how they interact with each other.
. Attitudes toward developing HL in a grassroot project and at home: a perceived relationship of collaboration and complementarity The excerpts selected for this sub-section demonstrate how non-formal and informal environments provide a set of affordances aimed at guaranteeing exposure to HL. Parents explain why they participate and engage in the intergenerational transmission of HL afforded by the MH project.These statements show that parents emphasize the significance of literature and picture books reading in HL transmission and literacy practices as a means of increasing exposure and learning in the Portuguese language.
Excerpt 1: "(. . . ) At home we use books a lot, right?I actually looked for the Mala de Herança ('Heritage suitcase'), I found out it was called Mala de Leitura ('Reading suitcase') before, right? . . .because of my need to have books here in Portuguese.So, we use it, it has this very clear role: it's the place we go to get the books, so we can read them at home in Portuguese.I think that's how we use it.And so, it also comes that, from that meeting, from that talk, my children see, 'look this child speaks Portuguese, look, how cool. . ., so many people from Brazil here' , right? . . . it generates another conversation.(. . .)" (CM, Brazilian mother) This mother reveals that the non-formal learning context she attends is an important partner and an input provider for literacy practices in the home.By referring to MH as the "place" where families can be supplied with books, it can be understood that the project is seen as an input space for HL.This mother's perception of the collaborative relationship between the two environments is indicative of her view of the significance of creating bilingual sociolinguistic spaces (Lanza, 2021;Purkarthofer, 2023), valuing and capitalizing on exposure in two languages.As will be demonstrated in the subsequent excerpts, the attempt to establish an input safe space will be evoked by other parents.As she describes, all the other affordances provided by the project's environment are also part of what she perceives as a holistic set, composed of opportunities for exposure to the Portuguese language ("from that talk. . . it generates another conversation"), which generates identification and can influence the socio-affective dimension of HL learning.This shows that literacy is not exclusively about learning to read and write, but also relates to oral capabilities, namely talking about the books and the project.
A similar perception is pointed out by a father when he explains the dynamics of the MH project: Excerpt 2: "Someone or several people read, children's or young people's books, to the whole auditorium, there may be parents reading, there may be older children taking part, what's also good is that there's also celebrations depending on the season, there's also a celebration of what's to be celebrated, for example, the June festival,... the content of the books read is also linked to the season, to the occasion, for example, Brazil's national day in September, carnival, or another subject, right?... which is also good, necessary, as well as those long breaks, cooking breaks,.... people talking to the children in Portuguese as well, in an environment entirely in Portuguese, and I think it's important that the children have the opportunity to practice their Portuguese not only in Brazil during the holidays, but also on specific occasions here, at home." (CK, German father) This father describes that at the center of the project's activities are the shared book readings, in which not only the children but also the adults are involved as an intergenerational activity for the transmission of PHL as a language-culture ("Someone or several people read, children's or young people's books, to the whole auditorium, there may be parents reading, there may be older children taking part").He highlights how literacy practices become a catalyst for developing cultural practices and multiplying opportunities for language use, with different interlocutors and on different subjects, therefore increasing the quality and the quantity of exposure.As this father explains, the readings are not aleatory, but rather connected to the cultural practices that will be recreated at the meetings ("there's also celebrations depending on the season, there's also a celebration of what's to be celebrated, for example, the June festival,... the content of the books read is also linked to the season, to the occasion").In this context, reading can be regarded as a catalyst for other practices, as it forms part of a pedagogicaldidactic plan designed to encourage the other linguistic and social practices that take place in this setting.
Another mother (PP) expresses a similar perception of strategies to develop HL.She emphasizes how literacy practices around reading serve to transmit the HL as a language-culture even if they mean "just playing with other children who are similar, " meaning bilingual in the same languages and having fun through Portuguese.This mother's descriptions allow us to glimpse the role of literacy practices in the transmission of a cultural load and how these cultural and holistic literacy practices can be interconnected to provide other forms of linguistic input: around reading, many other pedagogical resources can be incorporated into the HL transmission dynamics such as dancing, crafts, games, always in a pedagogical way to reach children in their playful universe.As she emphasizes, learning the HL must go beyond what is only transmitted by parents.
During the FG discussion, children highlighted the joint reading practices and the role of an adult who leads these readings in the HL learning process.They also emphasized the role of parents who continue the literacy practices at home, which indicates a relationship of continuity in the generation of input through the practices initiated in the non-formal setting of the MH project.Children therefore express their comprehension of the process of establishing continuities across input spaces that mutually reinforce each other, thus leading to what we could call "exposure overlapping, " i.e., the concomitant exposure to the HL and the majority language.Starting from the explanation of her drawing, which prompted the FG interviews, L describes her pictorial representations in such terms, that they corroborate the meaning attached to the literacy practices described by parents.The interactional co-construction in excerpt 3 depicts shared attitudes by children and parents toward reading and talking about the reading: Excerpt 3: "L: 'reading' , because, I think that by reading I learn a lot, and I learn words that I don't know and then I ask my mother, what do they mean; 'listening to stories' , in fact it's the same thing, and then you read alone, but you can discuss with others, the meaning of a word; and meet children who speak Portuguese, to show that it's not just you in Munich who speaks Portuguese, I speak Portuguese at my school, but it is European Portuguese; and... 'singing songs' , like the ones my mother sang when she was a child, yes, I really like it when she tells how her life was like... and 'eating typical food'.I also think it's important, so you can know more about the culture;... 'talk about indigenous tribes' because they already lived there, I don't know how long... yes... and they also had the things they did, to protect the forest, also talk about deforestation, it is dangerous; . . . to 'speak Brazilian Portuguese' , because at school, as I already said, I only speak Portuguese from Portugal.(...) As the discussion progresses, the children repeatedly emphasize the significance of literacy practices as the core element of the project.The other children describe the activities they engage in within that setting and offer their perceptions of the MH as a learning space: St: (..) most of the time it's different.Or there's a mat on the floor, and then everyone sits together, and the person who's reading is at the front, and the others are in chairs.
JU: We learn.Interviewer: How? JU: Someone says things, and we learn these things... like new words, or... something like that.St: and sometimes readers come to the 'Mala de Herança, ' actors, I don't know if it's right, authors of books... M: Authors, writers, and we read the books they've written.And then there are usually a lot of them, and you can buy the books to take home.
Interviewer: Can you take them home?And then?M: You can read as a family; you can read at night.And some books have already been translated into German.
Interviewer: Do you read them with your mum, or do you read them on your own?M: Oh... I've read a few by myself.We also read; my mum reads to both of us in the afternoon.
J: Oh... in the afternoon?M: in the evening."(FG with children) Children in this FG seem to be very aware of literacy practices as a means to learn and develop the HL, associated with other diverse affordances (such as reading, listening to stories, singing songs, participating in drama activities or eating typical food), that enhance qualitative and quantitative exposure.From the children's discourse, it can infer that the bilingual literacy practices carried out at MH are not merely regarded as a free time activity, but rather as a means of learning the HL mediated by more competent adults, increasing the quality of the input they are exposed to.
Answering the question about what they do at the MH meetings, one of the participants said: "we learn."During the discussion, children highlight the bilingual literacy practices offered by the project with emphasis on reading and discussing books written in both Portuguese and German.This brings us back to the continuum between L1 and L2 and also between monolingualism and bilingualism (Hornberger, 2003).Reading books in both languages makes the intersection of the continua of biliteracy (HL/majority language) into a concrete affordance.

. Perceptions on the construction of a bilingual input safe space at home
In the next excerpts (4 and 5), the awareness of the transition along the points of the continuum between the HL and the majority language also appears in parents' speech when they explain how they carry out their bilingual literacy practices at home in order to generate exposure: Excerpt 4: "Interviewer: And how do you use the 'Mala de Herança' at home?TB: (. . . ) We have a bookshelf with books in both German and Portuguese, right?She can choose, it's there for her to pick up, and then she chooses, one day, in German, another day, in Portuguese, or when she was younger, it's...I would read to her in Portuguese, even if the books were in German, which she didn't know, and we wanted her to know and learn Portuguese as well, right?So, I'd take a book and read it to her and show her the picture, and sometimes I'd tell her the story... we'd tell her the story based on that picture, you know, 'look, here'... and didactically speaking, too, you know... 'Look, the girl is holding a red book'... you know, not just 'the girl went to the park to read' , you know?No, 'the girl is there, in the park, she's looking at the grass,...' well... that's me... (laughs)...I teach and everything, right?I do it that way, reading at home, it works that way for me."(TB, Brazilian mother) The manner in which TB reports her actions in relation to literacy practices demonstrates that she does not perceive the Portuguese and German languages as distinct and incompatible entities that necessitate differential treatment within the domestic sphere.Instead, she affords the child the autonomy to select the language and delineates her own approach to navigating the distinction between written (and in which language) forms and the act of reading (which she terms "teaching, " underscoring the element of intentionality).A comparable dynamic is observed in the family in the following extract, as reported by the mother: Excerpt 5: "P: The girls only watch TV in German.Or, when F comes to explain something about the school, she only explains it to us in German.But as far as 'we' parents are concerned, it's only in Portuguese.Even when my husband reads... books at night to them.That's his role.And he reads them books before bedtime.He likes it, and that's his role.He does simultaneous translation.He doesn't read in German.That's our thing.He doesn't read.We have books in German (. . .)" Interviewer: He translates from German into Portuguese, right?P: He translates.That's his thing, so much so that when he's not at home, I only take books in Portuguese.I can't translate quickly enough so that the children don't get lost in the story.But because he already does it, it was kind of funny at first.Because he started, and then he would translate more slowly, and F says: "go daddy" (...).(PP, Brazilian mother) As PP explains, there seems to be a division of roles in their family language policy, with reading being the father's role.In an attempt to keep the balance in the ecology of languages and create a safe bilingual input space, the family finds a way to take advantage of the affordances provided by a monolingual book (usually in German), expanding it to a bilingual modus.The dynamic they adopt is to allow the oral use of the majority language at home (TV in German is acceptable, it is also acceptable for the child to report school issues in German).It is also possible to identify the intersection of continua in this case: orality in the majority language on the one hand, and HL associated with written literacy practice on the other, even when based on German monolingual reading material.Furthermore, it can be inferred from her speech that there is a constant search for the development of harmonious bilingualism (De Houwer, 2015), which is experienced positively by the whole family.
The next excerpt brings the answer of L, father of the child mentioned by TB in excerpt 4.He draws attention to a point often mentioned by other parents and a topic that has already been studied by other researchers (Szelei et al., 2024): the influence of the school context in the development and maintenance of HL and the challenges that come with it.Literacy practices at home become a safe space for the development of bilingual literacies: Excerpt 6: "L: Well, at home, as we have a lot of books, we try to keep..., it's.... we don't have a lot of discipline when it comes to reading, but whenever possible we take a book, or we let the child choose a book for us to read, and we also encourage them, in the case of K, who is already learning to read, to read when they can, or if they don't want us to, sometimes she wants to read a book in Portuguese, she wants to read it in German, we don't have that rule either, if she's going to read it in Portuguese or in German, we let her choose, but we try to keep her reading at home too, sitting down, before bed, we do that too.(. . . ) L: I think the 'Mala de Herança' meeting is important because, for the children, they have this contact with the Portuguese language, especially at school age, when the child spends the whole week going to school in German."(L, Brazilian father) In this case, the family language policy for the development of HL has no strict rules for the use of the languages in the child's bilingual repertoire, but rather allows the child to develop agency and a sense of autonomy in becoming a reader.When referring to the moment when the child starts to learn to read in the institutional context ("when the child spends the whole week going to school in German"), L perceives the amount of exposure to the majority language as a trigger for also developing literacy in the HL.That's why he highlights the need to balance the ecology of languages at home, especially considering the school routine in the majority language, a theme that will be developed in next Section 4.3.
. Attitudes toward (dis)continuities and (dis)connections: school onset a ecting literacy practices in the HL at home What we have seen so far is that, in acting as language educators and invested in the development of HL, parents refer to a relationship of collaboration and partnership between the informal context (the home) and the non-formal context (the MH project), as micro contexts for the HL acquisition and use.However, when school (as a macro context) becomes a relevant context, the relationship established between the literacy practices and the different learning settings reveals how the respective languages are perceived along the micro/macro continuum.In this section, we describe how parents react to the school onset and try to use it in favor of the HL.
Parents seem to believe in and even construct a hierarchy of situations conducive to developing competences and literacy in both languages.For example, in the following excerpt father T positions school as a privileged context for acquiring literacy.For this reason, since school is the setting dedicated to the majority language, he sees the development of that language as privileged over the development of HL, whose space he restricts to family interactions at home: Excerpt 7: "Interviewer: Do they also write in Portuguese?T: fluid reading, both (referring to his children), in writing naturally they have more difficulty in Portuguese than in German, right? . . .There are many more hours of writing in German compared to Portuguese, but they already have a certain ease.(. . . ) So, there will be some phonetics that they will still have some inherent difficulty with, or the confusion of s and z, or cedilha and ss, and these things we will still see errors, right?I think this is still very difficult without going through a school and a lot of repetition, for you to internalize this type of syntactic knowledge.But they can write."(T, Brazilian father) The father acknowledges that both his children have better receptive than productive competences in the HL.He concedes that the difficulties in HL writing stem from differences in the writing systems, grammar, or familiarity with the language, i.e., cross-linguistic differences.Additionally, he seems to believe that internalizing syntactic knowledge, such as grammar rules and sentence structure, can be challenging without formal education and repetition, leading to a disadvantage to the HL.
In our cohort, the parents' agency for language transmission/acquisition is very much in place by the time children start school.For many parents, the introduction to the mainstream school environment, that is the moment when literacy is initiated in an institutional educational context in the majority language, seems to be a trigger for parent's agency regarding transmission of the HL or development of writing skills also in the HL.As excerpt (7) above illustrates, when parents perceive that the HL has switched to a disadvantaged input position (given the fact that it is not a curricular language) which jeopardizes its development, they assume the role of language educators.Parents often reveal concerns with learning of orthography in both German and Portuguese, focusing on what they consider to be the main areas of difficulty zones, based on their perceptions of crosslinguistic difference ("cedilhas, " phonemes, diphthongs, the writing of "ch" or "sch, " etc).Significantly, parents seem to perceive the concomitant development of written skills in German and Portuguese as potentially problematic and a source of confusion, even though this is an already debunked misconception related to bilingualism.For instance, a Brazilian mother CM, states that "N is literate in German, [she doesn't] want to make a mess in her head." Other parents seem to be less concerned about this concomitant development of writing skills in the two languages, and they create informal moments at home to support the development of written Portuguese.These moments, as Brazilian mother KA explains, are neither systematic nor structured, and are not guided by the adults but instead by the children's interest to "play school."Another mother corroborates this "HL literacy desire" ("she brings the desire to write in Portuguese, " PP, Brazilian mother) and recognizes an intrinsic motivation to start writing in that language: "And she wants to write a message to us in Portuguese, not in German."(PP, Brazilian mother).KA refers to picking up those moments of children's agentivity to consciously select words that might pose problems at the orthographic levels: "so I'm writing some words on the board for her..., recently, I was showing her words that are spelled with ch and x."When thematizing both writing systems, this mother tends to select elements that she perceives as different linguistically, as in the examples discussed previously: "like 'ch' becomes sch... 'lh' becomes lha... she starts asking and as I respond she starts recording that.So, I think she already reads practically all these junctions, so she already knows almost all of them."(KA).
The development of literacy in German at school ("she has a lot of reading activity in German") triggers KA to start developing literacy practices in Portuguese as well.Reflecting an awareness of the need to consider the ecology of languages as a continuum (learning context: school-home), she decides to change her literacy practices strategy becoming a bilingual language educator.For this mother, the HL is "endangered" because it is not learned at school but at home (macro-micro).Conversely, formal education is perceived as more efficient for language development, which once again can be related to exposure, as already pointed out by another participant (excerpt 7).Despite the recognition that she is developing well in Portuguese and German, the mother decides that her daughter needs "tutoring in German, " because "she had a little difficulty."As with mothers quoted previously (KA and CM), bilingualism in general and the development of biscripturality more specifically are seen as sources of confusion that must be solved, and generally in favor of the majority language of the school.Despite parents' agentivity, i.e., intentionality in their practices, and good-will, they recognize the limits to their preparation.Some parents show frustration by the fact that they are not language educators and do not have the experience needed to help their child develop literacy in two languages.Importantly, some non-German parents also express frustration because of their own challenges with the majority language, which negatively impact their competencies as potential bilingual educators.As PP explains: Excerpt 8: "Because at school she has the greatest experience of her life, so her vocabulary is much greater in German, because I don't provide the same amount of...I don't provide... There's no way I can provide her experience in German, in Portuguese.I even try, when she brings activities,... it's very funny, because she does homework with me sometimes, and then, I don't answer in German, I'm reading there, and I explain it in Portuguese, and she understands, and the words, I say, if she needs it, she says it, I'm saying the words in German to her, in Portuguese to her, even though she reads them in German, it's kind of funny.But it is the only exchange of information for Portuguese.So, she has command of German, due to her greater experience."(PP, Brazilian mother.)This excerpt implies that challenges with the majority language impact parenting and the way families interact during literacy events.Like father T (excerpt 7) previously, this mother acknowledges that her child's greatest language learning experience happens at school, where she is extensively exposed to the German language, which is aligned to the literature that refers to the diminishing exposure to the HL when entering primary school.Nevertheless, knowing that she cannot provide the same level of experience in Portuguese as her child receives in German at school, she makes efforts to support her child's learning.She describes helping with homework and explaining concepts in Portuguese when needed, even though the child is reading and learning primarily in German.Such practices mean that the mother is consistently using her parenting time to provide the child with semantic equivalents in two languages, providing opportunities for what we could name "school content and language integrated learning" at home, replicating the content learnt at school in the HL.This practice indicates that mothers attempt to create moments of transmission of content learnt formally at school, therefore enriching the quality and the quantity of the input the child is exposed to in the HL.Such practices counter the tendency to develop only competences in the HL' informal register.
Children also have an implicit awareness of the influence of the level of exposure that school provides and how this affects the development of the majority language in relation to HL.The excerpt below is a part of one FG discussion when the children were questioned about their attitudes toward their bilingualism: As emerges from this excerpt of the FG, the languages of N's repertoire are placed along the continuum L1/L2 or micro/macro.Although N doesn't state that there is one language in which he is more proficient, he refers to two aspects that determine the continuum: the amount of exposure provided by the school and the time of acquisition, i.e., the HL was learned first (at home), while the majority language, which came later in sequence due to the learning context (at school), is now stronger in the continuum.In his own words, he represents the successivesimultaneous continuum which Hornberger (2003) refers to in her model, while showing an awareness of the respective places these languages occupy yet highlighting their entanglement: school for the majority language, associated with more exposure, while the HL is identified as the L1 but starting to get the second position in this dynamic "ranking." In a later moment of the same FG, we can see how school can also be perceived by children as not being the place to use the HL.The HL is perceived as an inadequate or illegitimate linguistic resource in the school context, which is (re)constructed as monolingual: Excerpt 10: "M: in the first and second classes it was a little complicated, because every now and then, it happened to me [to answer in Portuguese] instead of answering in German.
M: ahm...I also imagine it's a bit annoying, because at school there were times when I mixed Portuguese with German.(laughter) (M, 12 years old) The child expresses her experience with language mixing during classes.She mentions the embarrassment caused by her lack of ability to keep both languages separate, and more precisely the minority language out of school.This excerpt thus illustrates the challenges that can arise for children from navigating multiple languages, particularly in educational settings where one may need to select the languages to be used, censoring a part of the linguistic repertoire.

Discussion
Our study was concerned with the identification of literacy practices that children are exposed to and co-develop with peers and adults in different contexts.Our results point to literacy practices in different contexts that complement and support each other in both the heritage and the majority languages.The data analyzed reveal the care taken to create safe zones for the development of literacy skills in both languages at home, with parents taking advantage of the practices, resources and materials used in the MH project and at school to support an integrated learning of the majority and HL languages.In other words, from two potentially monolingual spaces (MH for Portuguese and school for German), they consciously build a synthesized space that supports integrated bilingual development through what we have called "bilingual input safe space."By doing this, the home space becomes a "fusion space, " where monolingual norms are challenged to promote "harmonious bilingualism" (De Houwer, 2015).
The research question aimed at investigating children and parents' attitudes toward those literacy practices across formal, informal and non-formal contexts.The data analyzed points to a great deal of agency on the part of the children, who urge their parents to act in response to their questions about bilingual development.Both children and families focus their efforts on the development of both languages, although parents seem to invest more in supporting the minority language, especially as it is the language in which they feel most competent.After entering school, a period in which HL is weakened due to (perceived) reduced exposure, parents adopt an "if you can't beat them, join them" stance, taking advantage of and transferring their child's exposure to literacy practices in the majority language to the HL/minority language.Indeed, it is noticeable that the insertion in formal regular education influences parents' attitudes and agency to support a balanced bilingualism.In this way, they show that they understand literacy practices in different contexts as potentially disruptive but develop mechanisms that allow their children to be exposed to the same practices in both languages, promoting continuity.Such practices might be seen as an attempt to balance the extremes of the continuum: oral and written language development, on the one hand, and majority and minority language integrated learning, on the other (in line with Hornberger's model).As Hornberger's (2003) theoretical approach points out, there are several points on the continuum that intersect and along which learners and educators can move: reception-production, oral language-written language, or L1-L2 transfer, as well as the micro-macro, oral-literate levels.
It is noteworthy that parents mainly address issues of crosslinguistic interference in learning to write, when they refer to the development of literacy in both languages.In terms of literacy development, they seem to be mainly concerned with areas of orthographic interference, while other aspects related to the macroand meso-structure of texts are not mentioned.This reduction may be related to the fact that their children are still in the early years of primary school, in the 1st years of acquiring writing.Children, on the contrary, do not refer to those orthographic concerns.Like their parents, they realize that there is a clear relationship between the practices started in the MH project and those that are continued at home; less continuity is reported by children concerning practices at school.Children emphasize the role of parents in the continuity of literacy practices at home, reinforcing that these learning environments establish a complementary relationship.

Conclusion
Based on the data presented, we can conclude that literacy practices across formal, informal and non-formal contexts are perceived as a means of maintaining and developing the HL, particularly by the families.The learning settings implied in the development of the HL are interconnected in a complex ecology.The relationship between formal, nonformal and informal environments is understood as one of collaboration and complementarity, even when the formal context could be conducive to disruptions and challenges to HL maintenance.In the equation of the ecology of languages (HL/majority language/curricular languages), children's insertion in the institutional school setting and development of literacy in the majority language (and so the development of biliteracy) becomes a complex and challenging scenario for parents as (bilingual) language educators.
Our research highlights that the development of literacy in HL requires a balance between formal instruction and the promotion of informal language use within the community to ensure the continued vitality of HL in the face of societal language dominance.Literacy in HL needs not be disconnected from the development of literacy in the majority and in the other languages of children's repertoires.It was important to see that parents do identify the possibilities of transferring literacy practices from a formal context for the development of the majority language to capitalize on them at home for the concomitant development of the HL/minority language.In this sense, they are able to identify areas of communication between literacy practices and contexts, rather than focusing only on ruptures and disconnections.From this perspective, this paper suggests that the concept of "exposure, " when used to conceptualize, describe, and predict the development of the HL, should refer to more that the lived contact with the HL, at home and in the community, and encompass the lived contact with other languages, even if they are perceived as concurrent, such as the majority language or even English, as the first foreign language learnt in primary school.This paper makes the case that "exposure" as a concept should be regarded through a multilingual lens, as all the languages of the children's repertoire are said to influence each other.
By expanding the concept of "exposure, " this contribution acknowledges the complex multilingual environments that many families navigate, while coordinating affordances in multiple languages at home, school, and community-led initiatives.Such a holistic view on (multi)literacy practices better mirrors real-world scenarios where children are exposed to multiple languages simultaneously and families engage in multilingual literacy practices.Such a stance also emphasizes the fluidity and permeability of language boundaries (even between languages of different linguistic families, as German and Portuguese) and linguistic practices, therefore proposing a comprehensive framework for understanding literacy practices and HL exposure and development.
Even if this contribution focused on the dynamics of exposure to the majority and the HL, future research should cover the contact with curricular languages in more depth.This paper shows that families (and to a certain extent children) exploit the porosity of exposure to different sources and languages in their literacy practices in a variety of contexts, so this porosity could be studied in contact with other languages and other contexts.Indeed, the flexible use of linguistic resources and the transfer of literacy practices and events from one context to another could be objects of further research.
A final word should be left in terms of the study's limitations.Regarding the generalization of results, it should be pointed out that this study was carried out with a very specific and relatively small audience, a group representative only of very homogeneous Portuguese-speaking families, within the heteronormative fathermother-children model.The group did not cover families made up of a more socioeconomic heterogeneous profile or even more linguistically complex families.All the participants have higher education and medium-high socioeconomic status.The majority of those interviewed were mothers and housewives dedicated to raising their children almost full time.This group can therefore be described as middle class, with an academic background, aligned with the school aspirations.It is therefore important to consider that other families, with a different socio-economic and (ethno)linguistic profile, may perceive and experience the circumstances of a transnational life and the transmission of a HL/minority language in a very different way.The results obtained in this study cannot, therefore, be seen as generalizable outside of groups with these characteristics.It is therefore important, for follow-up studies, to consider the lived experiences of families with different profiles, to make the research more inclusive.
Because of my need to have books here in Portuguese.So, we use it, it has this very clear role: it's the place we go to get the books, so we can read them at home in Portuguese." " What do you think?Is there anything that is difficult in Portuguese for you that is not difficult in German?N: in Portuguese I imitate my father's accent, and now sometimes I sound like him, with his accent.But actually, German is a little easier for me, because now I speak it longer than Portuguese, but I started speaking Portuguese first, so it's a little easier too, but German is easier."(N, 12 years old)