Family Language Policy and Childhood Bilingualism: A Multidimensional Theoretical Analysis

This paper focuses on the micro-level of language policy and aims to point out the critical role of family language policy (FLP) in language maintenance/shift of minority/heritage home language(s) and childhood bilingualism. FLP could resist broader language ideologies by transforming parents’ language ideologies and attitudes into language practices and language management that support the development of active or additional childhood bilingualism. Through a research into the interdisciplinary components of FLP, this paper aims to illustrate aspects of multilevel and dynamic relationships between each of these core components. An introduction to FLP as a research field is included as well as some of the studies that spotlighted the way parental agency in regard to each of the three FLP components could shape, explicitly or implicitly, language use and planning for the minority/heritage home language(s).


Introduction
Family is considered as a distinct sociolinguistic domain that shapes a significant sociocultural context. Research on language practices of family context, where macroand micro-sociolinguistic realities intertwine, substantially contributes to the development of theories for children's language socialization and language acquisition. Families could also contribute to efforts for reversing language shift and support language maintenance (Griva & Stamou, 2014: 38-40). In relation to this, Fishman (1991: 87-109) highlighted the determinant role of family for the intergenerational transmission of the minority/ heritage home language(s).
Diverse family backgrounds -relating to the cultural origin of parents and/or home language(s) -indicate that children will be socialized into at least two distinct "communities of practice". As a "community of practice", family members could follow different norms in terms of language use and language culture rather than follow the dominant/majority language norms (Lanza, 2007: 47). More specifically, in case one of the parents or both parents' language(s) is not the language of the wider community ("foster bilingualism"), "family bilingualism" arises. Especially when social bilingualism or multilingualism is not the mainstream language norm, those families face daily challenges in their attempts to support the development of childhood bi-or multilingualism and maintain home language(s) which could be minority or heritage language(s) (Lanza, 2007).
In this article, childhood bilingualism is viewed as functional bilingualism, a definition which supports that the sociolinguistic environment affects the way bilingual children use and switch between their languages. Fishman's (1965) functional bilingualism examines the language use and language choices of the speaker according to the circumstances, the recipients of communication and the sociolinguistic spaces (domains), where communication takes place. The sociolinguistic environment also contributes to the level of language acquisition (Sella-Mazi, 2001: 44-45). Receptive bilingual is a category of speaker who acquires a level of comprehension in a language but little or no active command of it. Active bilingualism refers to speakers who acquire varied levels of different domains of language use in two different languages. The question that arises here is not whether a person is bilingual or not, but to what extent s/he is bilingual (Triarchi-Herrmann, 2000: 46).
De Houwer (2011: 223-232) also refers to two different types of childhood language acquisition in relation to speaker's age as an important environmental factor: (a) "Early Second Language Acquisition" (ESLA); (b) "Bilingual First Language Acquisition" (BFLA). In the first case, monolingual children have been exposed to a second language before the age of six, whereas in the second case children have been exposed to two different languages from birth. The dominant language of a community and a home (minority or heritage) language could be the two different languages that children have been exposed to and use in case of family bilingualism.

Family language policy
Family language policy (FLP) is a growing and useful research field, as it bridges the gap between, draws from and contributes to other research fields like (education) language policy, child language acquisition and language socialization. FLP refers to parents' explicit and/or implicit language planning for language use between family members (King, Fogle & Logan-Terry, 2008: 909). According to Spolsky (2004), like language policy, FLP consists of three basic components: language practices, language management or planning and language ideologies. In this interdisciplinary field, research could focus to one or more of the FLP components, how these components intertwine, what family-external and/or familyinternal factors shape FLP and are relevant to children's mental and bilingual development, children's language use, school performance and, ultimately, the maintenance of the minority/heritage language(s). Curdt-Christiansen (2009: 354) developed a model that demonstrates the interconnection of FLP components in the micro-and macro-level of family context. Yamamoto (2001: 19) also illustrated how familial, sociocultural and linguistic environmental factors are linked to different FLP components (e.g. parents' language choice and attitudes towards bilingualism and/or languages), the development of childhood bilingualism and children's language use respectively.
Therefore, in order to study the dynamic structure of FLP, it is necessary, on the one hand, to analyze and decode parents' language use and strategies which also reflect their language ideologies, personal experiences and the influence of the wider social environment and, on the other hand, to examine how these language strategies are implemented in face-to-face language interactions between family members and in which terms children's functional bilingualism is shaped through FLP.

Models of family language use
Language practices refer to patterns of language use within the family, through which family members realize, negotiate and modify their FLP in face-to-face communication. According to Hamers and Blanc (2004: 63), the linguistic context produced by parental interactions determines children's language productions to a significant extent; this is a scientific observation also confirmed by De Houwer's (2011) studies. Barron-Hauwaert (2004: 163-178) analyzed seven types of language use within the family context: (a) "OPOL-ML" (one parent, one language -majority language), (b) "OPOL-mL" (one parent, one language -minority language), (c) "Minority-Language at Home" (mL@H), (d) "Trilingual or multilingual strategy", (e) "Mixed strategy", (f) "Time and Place strategy", and (g) "Artificial" or "Non-Native strategy". In this unit, we will describe in detail five out of seven types of family language use.
In OPOL-ML families, parents have different mother-tongues and one of the two languages is the dominant language of the community, where the family lives. Each parent uses their mother-tongue in parent-child(ren) interactions while, in this model, parents usually communicate with each other in the dominant language and the language of the community gradually dominates the language choices of the family. In this model, the minority language is likely to be maintained when the mother of the family interacts with her children in the minority language, especially during the early years of their lives. Similarly, in OPOL-mL families, each parent uses their mother tongue in parent-child(ren) interactions, although parents use the minority language to communicate with each other. OPOL-mL model implies a higher level of exposure to the minority language and encourages children to use that language, though also requires one of the parents to be, at least, functional bilingual to the minority language of the family. In the Minority-Language at Home model, children are more likely to develop active/productive rather than receptive bilingualism, in case parents share the same minority language, or trilingualism when parents speak different minority languages and use the OPOL model. In some cases, one of the minority languages of the family dominates the other.
In the case of "Mixed strategy", switching languages within the family in daily interactions, which usually reflects wider language practices in the community, indicates a different approach toward language functionality. OPOL families regularly shift to this model when children have satisfactorily mastered languages and switch them depending on the context of interactions and/or the topic of discussion with their parents. Similarly, "Time and Place Strategy" could serve as a transitional stage leading to OPOL-mL, Mixed or Minority-Language at Home approaches. In this "Time and Place Strategy", each language links to specific family activities or routines in time and/or space.
According to Barron-Hauwaert (2004), no model stands as the most appropriate for all the cases of family bilingualism as the bilingual development of children also depends on other factors. Further to this, although in some cases parents state that they apply OPOL, they may alternate the prearranged OPOL interaction patterns in face-to-face communication with their children. Some parents gradually disengage family from the OPOL scheme as long as children have grown up and, according to parents, satisfactorily acquired two (or more) languages or they are very comfortable in switching these languages (Döpke, 1998). De Houwer's (2007: 419-421) findings highlight Minority-Language at Home and OPOL-mL as the two most successful models for children's minority language use; in these models, both parents primarily use the minority language at home or both parents speak the minority language at home and one of them uses the dominant language at the same time. De Houwer's (2007) findings also showcased that the use of the dominant language at home does not necessarily act as an obstacle for the transmission of the minority language. Especially in cases where mixed-lingual families are demographically isolated, Döpke (1998) argues that OPOL models are necessary so these families can maintain the minority language.
Family synthesis is another factor which shapes family language patterns. Döpke's (1992) findings show that younger siblings in mixed families developed receptive bilingualism in contrast to active bilingualism of the first child, but also that siblings mainly use the school language in their interactions. Despite this fact, according to Barron-Hauwaert (2004), one advantage is that siblings could increase the usage of a minority language, especially in relation to the minority-speaking parent. Furthermore, parents could change their FLP for later-born children into more or even less strict rules for minority language use. Adult time spent with later-born children is also another variable related to language input. Parents could better control the language interaction patterns with one child; however, many siblings could result in older siblings becoming the main linguistic models for the younger ones and thus parental language control weakens. Consequently, according to the above, results in relation to later-born siblings' language skills are varied, especially in OPOL families. Finally, grandparents of the minority-speaking parent could also affect parental language choices and children's minority language development through the so-called "language bath" either during their stay or distance communication with their grandchildren.
De Houwer (2011) refers to 'language input environment' as an overall term for different aspects of language input, which include but are not limited to: (a) the number of utterances children hear in each of the languages; (b) the length of time children hear a language; (c) the way language stimuli are distributed in case of family bilingualism; (d) parents' responses to children's language choices. The first three aspects are related to "input frequency", which De Houwer considers as the most important environmental factor that most likely affects bilingual acquisition in terms of oral language production. Based on her studies, the researcher attributes the differences in different domains of the ESLA (Early Second Language Acquisition) children's language use mainly to the child's age of language exposure and time of length the child hears the language(s). Concerning the BFLA (Bilingual First Language Acquisition) group of children, differences among speakers of this group are mainly attributed to the model of language use among the family members (and also between parents) and the ways parents involve and engage their children in meaningful language learning where children are motivated to use their home language(s).
However, Carroll (2015) raises concerns about another factor which could affect the way bilingual children interpret these language stimuli. The researcher supports that there is no direct, relational causality between the quantity or quality of language stimuli and children's language production, encouraging researchers to explore the field further. According to Carroll (2015: 4-5), decrease in children's minority language use or lack of use may not be entirely linked to the decreased 'input frequency' but also to children's language choices and attitudes towards each of home languages. Parents' language strategies encourage the development of bilingualism, though bilingual children also make their own language choices.

Family language management
Family language management (FLM) is defined as the implicit/explicit and subconscious/conscious parental involvement towards the establishment of those language conditions which support language learning and literacy acquisition of the minority language(s) at home and/or community settings (Curdt-Christiansen & La Morgia, 2018: 179). This definition completes the theoretical framework of Spolsky (2004), considering family literacy practices as part of FLM. According to Schwartz (2010), two main trends in FLM are present: (a) parental language choices on which language(s) to use in parent-child(ren) interactions, discourse strategies that parents adopt, more or less consistently, in language interactions with their child(ren) and home literacy practices (internal control for FLP); (b) parental agency in search of heteroglossic spaces towards the development of childhood bilingualism and/or biliteracy and the maintenance of the minority language(s) (external control for FLP). Lanza (2007) showcased a link between OPOL language strategies and children's switching languages or language codes. Code-switching or switching from the "established" language to the other one is referred to as 'mixing'. The researcher identified five types of strategies illustrated in a linguistic continuum in which the left end represents the monolingual strategies and the other one the bilingual strategies integrated into parent-child(ren) interactions: Lanza's analyses led to the conclusion that childhood active bilingualism is more likely to develop, especially when the parent who uses the minority language applies strategies that approach the monolingual end. In relation to this, it may not be enough for the minority-speaking parent to speak the minority language only but to interactionally co-construct a monolingual context, so the child(ren) are socialized into maintaining this monolingual context with the minority-speaking parent. In addition, each of the parents could follow different language strategies. Therefore, researchers need to analyze not only practices in dyadic interaction (mother-child or father-child) but also in triadic interactions (mother-child-father). Languageswitching or code-switching may be based on the specific context or content of family discussions ("discourse-related switching") and also on participants ("participantrelated switching") or the presence of more than one child in the family.

Discourse strategy Description
Minimal Grasp Strategy The adult does not seem to understand the child when the child uses the other language. Expressed Guess Strategy The adult asks a question to which the child answers with "yes" or "no", using the other language. Adult Repetition The adult repeats what has been said by the child, using the other language.

Move On Strategy
The adult allows the conversation to continue, indicating that s/he comprehends and accepts the child's mixing.

Code-Switching
The adult himself uses both languages or employs intra-sentential code-switching. Table 1. Description of five parent-child discourse strategies (Lanza, 2007) Döpke (1992), who studied the interactive strategies between parents and children, argued that the quality of parent-child interaction is more important for the development of active bilingualism compared to the number of stimuli in the minority language. More specifically, Döpke's analyses revealed that the more the parent speaking the minority language engage the child(ren) in child-centered speech during playful activities the greater the chance the child will become an active bilingual. Moreover, parental language strategies as the above, according to Lanza (2007), are related to individual personality and performance which may also be influenced by societal and cultural norms on gender and family roles.
Finally, these discourse strategies that make parents' preferences over one language explicit could have a success in securing children's subsequent choice of that specific language in parent-child(ren) interactions. Conversely, a different approach, called "Happylingual", is adopted by some parents, so the family maintains the minority language without further pressure on children; in this case, family perceives language-switching and the bilingual phenomenon as a "qualification" (Kapeliovich, 2013).
Despite the success that discourse strategies, such as the above, may have on children's active bilingualism, the same language strategies may not be enough for the development of literacy in the minority language. Active bilingualism in terms of oral language production is not equal to additive bilingualism which includes biliteracy. Biliteracy in the minority and the dominant language involves any form of interaction taking place in these two languages through a written text or in relation to a written text ("... in and around the written text") (Hornberger, 1990: 213). Schwartz (2008) highlighted the impact of family literacy practices on the higher level of performance achieved by children in tests in the minority language. Parent-child shared reading, child independent reading, explicit language learning during parent-child shared reading, parental supervision during writing activities in the minority language but also literacy practices which include cultural content are some of the literacy activities that could support the enrichment of home literacy environment in regard to home languages. Further to this, parental engagement in creative language activities and children's active reading in the minority language link to extended knowledge of vocabulary. According to Zhang and Koda (2011: 14), minority language literacy acquisition could mostly benefit from parent-child(ren) shared or child independent practice on printed literacy resources three to four times per week in addition to the minority school's writing activities. Finally, Hashimoto and Lee (2011: 165) in their qualitative study of three immigrant families of Japanese origin, residing in the USA, revealed that parents modified their practices, resources and materials to stimulate children's interest and enhance the context-related functional biliteracy (similarly to functional bilingualism) in the everyday life of their family. Thus, parents engaged to activities such as active reading between the older and the younger siblings, parent-child shared reading, leaving notes or writing messages to the parent, who was absent from home, using the Japanese language but also family discussions on manga comic books and their captivating illustrations, recipe books, gardening diaries, maps, sending letters and cards to friends and extended families, etc.
Regarding the external control of FLM, parents could explore bilingual education programs, bilingual schools or minority/community/complementary schools for the maintenance of children's non-dominant home language(s). According to a study by Leung

Family language ideologies and parental attitudes
The family model of language use could reflect broader ideologies and practices in relation to language(s) as well as parental attitudes on child rearing and bilingual development respectively. Instead of "language ideologies", Schiffman (2006: 112) uses the term "linguistic culture" which he defines as "the sum totality of ideas, values, beliefs, attitudes, prejudices, myths, religious strictures, and all the other cultural baggage that speakers bring to their dealings with language from their culture". Further to the above, Myers-Scotton (2006: 110) distinguishes attitudes from ideologies; attitudes are more unconscious assessments of the social usefulness of a language or language variety, while ideologies are more constructed assessments. Family-external factors which are connected to historical, political or cross-cultural factors could affect parents' language ideologies and attitudes towards language(s) and/or bilingualism but also family-internal factors which are related to parents' aspirations, (language) experiences and/or emotional attachment to their mothertongues, could also affect their attitudes towards home language(s) and their specific language choices.
Canagarajah (2008: 170) claims that family is not a self-contained community of practice as it can be influenced by historical, social and institutional forces. In his ethnographic study of Sri Lankan Tamil migrant and diaspora communities in English-speaking host communities, he explored the family-internal and -external factors that appear to be related to the language shift from native (Tamil) language to English. Analysis of language trends, patterns and attitudes across the three generations of migrants indicate the dominant use of English, especially in the third generation, and very rapid language attrition for Tamil. Parents' positive attitudes towards English could be related to historical factors (UK colonialism in Sri Lanka and English as a superior language), political factors (illegal immigration to host communities or refugee status and preference over 'Canadian' or 'British' citizenship), sociocultural factors (escaping caste inequalities and stigma by acquiring English and showing disinterest in Tamil culture), sociolinguistic factors (first and second generation's bilingualism due to prior colonial experience and languageswitching to English). Despite the governmental support in host countries for the maintenance of Tamil language, parents' language ideologies and attitudes on the socioeconomic status of English and Tamil respectively lead to the dominance of English as their preferred family language model and a decline on children's proficiency of Tamil.
Mills's (2005) ethnographic research focused on second-generation immigrant mothers of Pakistani origin living in the UK and highlighted the relationship between language, identity, culture and citizenship. Through the narratives of Mills's research subjects, it seems that the two languages (English and the language of origin respectively) hold distinct roles in the life of the participants and that the use of the heritage language is not an obstacle for self-identification with the dominant Englishspeaking community. More specifically, the participants, on the one hand, perceive English as the language of educational and professional success; on the other hand, the language of origin symbolizes the relation with the immigrant community, the country of origin and their religion. Although English dominates language practices of their families, these mothers maintain a symbolic and emotional relationship thus support positive attitudes towards their language of origin. Their language ideologies and attitudes have a significant impact on their FLP and their language planning (e.g. long family trips to the country of origin, satellite for children's language exposure and identification with their language of origin and culture, etc.).
Curdt-Christiansen (2009) examined how values, beliefs and practices as well as power differences in a minority context shape language ideologies and relevant language practices of immigrant Chinese parents in Quebec of Canada. Parents' negative experiences on immigration and the socioeconomic status of English and French do not seem to discourage positive attitudes towards their non-dominant home language. Chinese language as a cultural core value and parental aspirations for children's multilingual development, which parents link to advanced educational and professional opportunities for children, are the driving force for families' language planning which supports the development of Chinese language literacy and multilingual home practices. Yamamoto (2001) studied bilingual families in Japan and showcased that the international importance of the English language as well as the high status of English in the Japanese educational system encourage parents' positive attitudes towards the bilingual development of their children. On the contrary, parents who express negative attitudes towards bilingualism and discourage the development of the minority language at home attempt to eliminate the linguistic, social and cultural distance with the dominant Japanese-speaking community. Except the above family-external factors and the socioeconomic status of languages, parental beliefs and attitudes towards family language planning may be influenced by public discourses upon children's bilingual development (media, school, etc.) and specific aspects of parenting (or parenting in the host country's dominant culture), although parents may rely on their language experiences and selectively draw information from expert advice and popular literature (press, internet, textbooks, articles on bilingual development and education, etc.) (King, Fogle & Logan-Terry, 2006).
Similarly to Curdt-Christiansen's (2009) sample of parents, where Chinese also play an important role as a cultural core value for parental identity and a key language towards multilingualism, some studies mainly focus on parents' background which could affect parental agency over home language planning. In several cases, parents perceive their language as a symbol for their cultural heritage, in which they try to socialize their children through the minority/heritage language use (Griva, Kiliari & Stamou, 2017), but also as a tool for maintaining family cohesion (Schwartz & Verschik, 2013: 6). Tannenbaum (2005) emphasized on the parents' emotional attachment to their mother-tongues as a factor that could shape their attitudes towards language(s) and language choices. Parents' language repertoire and immigration and/or language experiences but also their perceptions of the value of each language seem to significantly define parents' expectations and language management, as found in the survey of Kirch and Gogonas (2018), who studied the FLP of two Greek families in Luxembourg. Parents' identity shapes their language ideologies and relevant language practices in a different way; in one case, parents adopt a "European citizen" or "citizen of the world", multilingual approach which is reflected in their language planning and family literacy practices. In the other case, parents emphasize family cohesion and their identification with the Greek identity so relevant patterns of language use and literacy practices are followed. Kirch (2012) explored the language expectations of seven Luxembourgish mothers residing in England and Scotland. These mothers' FLP strives to support their children's bilingual development (Luxembourgish and English). Although mothers recognize the low functionality of Luxembourgish in the dominant English-speaking community, they feel emotionally connected to their mother-tongue in which they recognize a natural, authentic way to interact with their children. Additionally, they consider multilingualism a core value of Luxembourgish identity. Therefore, through their personal bilingual and/or multilingual learning experiences, they have developed positive attitudes towards bilingualism and multilingualism as well as high expectations for their children's bilingual or multilingual development.
A different group of research focuses on parental beliefs and attitudes which are related to children's language acquisition. De Houwer (1999: 85) represented parental attitudes towards a particular language, bilingual development or specific language choices and strategies in a continuum (negative/neutral/positive attitudes) and distinguished them from parental impact beliefs. Impact beliefs are related to parental perceptions of how parents view themselves as (un)capable of shaping and monitoring their children's bilingual development. De Houwer (1999) also represented strong and weak parental impact beliefs in a continuum.
An innovative concept for FLP is presented by Purkarthofer (2017), who illustrates parents' language ideologies and their representations of future FLP for their unborn children's language development. In her study, parents are encouraged to depict their connection to languages using language portraits and their future aspirations using Lego structures to contextualize social spaces in relation to language regimes, language use and mobility. The three couples in this study also highlight child agency as a decisive factor in FLP. Their perceptions are linked to their status, educational background and prior life experiences.
Finally, parents' language ideologies and positive attitudes towards languages or bilingual development are not always transformed into relevant language practices and language management that contributes to active or additional childhood bilingualism (Curdt-Christiansen, 2016). These FLP inconsistencies may be the result of different combinations of family-internal and/or -external factors which challenge parents in their everyday efforts to maintain minority/heritage home language(s).

Family language policy and child agency
Although parents' language ideologies might be the driving force of FLP, Schwartz (2008) showcased that children's practice in reading in the minority language, parents' language practices and children's positive attitudes towards the minority language are the strongest factors in acquiring the vocabulary in the minority language and that parental ideologies had a minor impact on children's command of the minority language. According to Luykx (2005), research on children's language socialization usually attributes a passive role to them. On the contrary, the researcher highlighted three cases of social conditions in which children may act as active subjects who linguistically and culturally "socialize" adult members of their family so that children negotiate and co-construct FLP in: (a) the case of immigrant families, where children are socialized into the dominant community and acquire a higher level of proficiency in the new language than their parents; (b) the case where immigrant parents adapt their language repertoire to meet their children's new needs; (c) the case where parents practice their skills in the dominant language at home or during their supervision of children's school assessments. Gafaranga (2010) also refers to the phenomenon of "medium request"; in this case, children impose the use of language of their preference through parent-child discourse strategies such as the "Move On" strategy.

Conclusion and further perspectives
The effect of FLP on children's bilingual skills is not unidirectional and linear but dynamic and multifaceted. Although parents' language ideologies might be the driving force of FLP, it seems that language use and practices have a more direct impact on children's language use and bilingual skills. Family-internal and -external variables could affect FLP and consequently childhood bilingual acquisition. FLP may be affected and modified by factors related to the family-internal environment (e.g. family synthesis, parental observations on their children's language development, children's language choices and motivation for language learning, daily family plan, future family planning, etc.). Family-external factors (e.g. governmental support for minority languages and education language policy on minority languages, parents' social network, broader ideologies and the status of languages, expert advice, public discourses on bilingualism, etc.) could also affect the implementation of FLP and possibly provide insight for any inconsistencies between positive language attitudes towards the minority/heritage languages and confirmed family language use and/or management. In many cases, parents strive to create the conditions or the heteroglossic spaces for childhood bilingualism to flourish and their efforts result in receptive bilingualism for children, a situation that indicates a tendency to language shift but also could be seen as a kind of language maintenance. According to Gafaranga (2010), language maintenance and language shift could be considered as the two sides of the same coin; the bilingual phenomenon includes both these processes. Also, in other cases, parental practices related to an in-culture policy seem to attempt to compensate for children's low proficiency in the minority/heritage language(s) or monolingualism in the dominant language.
Future studies need to illustrate successful parental language strategies, which support the development of active bilingualism or biliteracy but also include and extensively investigate child agency towards language use as part of the formation of FLP and its relevant outcomes for children's bilingual skills. Moreover, research on FLP could be a valuable resource and tool so policy-makers or schools are enlightened and support the maintenance of minority/heritage/ community languages, the development of childhood bi-or multilingualism and language learning through language education policy and, for teachers, culturally sensitive and well-structured methods in classroom teaching.