The Role of Age Variables in Family Language Policy

Abstract

Family language policy (FLP) provides a critical framework to explain the planning of language use in the home. It constitutes a dynamic construct that sheds light on variations in the language acquisition of bilingual children, potentially explaining the shifts that may occur in language dominance and preference. The environment and life experiences are thought to shape FLP, yet little is known about the function of age. This study examines the association of FLP with children’s chronological age and the age they become bilingual. Data were collected via questionnaires from parents and their bilingual children (n = 82) aged 5.08–14.08 (M = 8.98, SD = 3.27) speaking English (heritage language) and Hebrew (societal language). Correlations and logistic regressions indicate a relationship between FLP and dimensions of age. Findings reveal that age may have repercussions for parent language beliefs, patterns of language use within the home, and the adoption of language promotion strategies. Younger children and children with a later age of onset of bilingualism are associated with families who lean towards a pro-heritage language FLP. Considering dimensions of age enhances our understanding of FLP and may offer a greater insight into how languages are supported in the bilingual home.

1. Introduction

Children are introduced to language at home, and it is considered vital for language acquisition (Spolsky 2012). Family language policy (FLP) explores implicit and explicit language planning within families used to influence bilingual children’s linguistic proficiency and may change over time (Kaveh 2020; Kaveh and Sandoval 2020; Smith-Christmas 2022b). Two dimensions of time that impact bilingual language acquisition are chronological age (CA) and the age at which a child is introduced to an additional language (age of onset of bilingualism (AOB); Meir 2023; Paradis 2023). Accordingly, the present study’s objective is to investigate FLP as a function of CA and AOB. It attempts to establish whether mechanisms that impact language development also shape FLP.

Children with an AOB from birth may have been exposed to the heritage language from one parent and the societal language from the other parent. In contrast, those with a later AOB are anticipated to be exposed to the heritage language from both parents (De Houwer 2023). Child-rearing practices also change with CA (Orellana et al. 2001). Parents may relax their FLP with older children once they have mastered both languages (Paradowski and Michałowska 2016). Temporal measures may be associated with academic, cultural, and societal pressures. Older children may be susceptible to peer influence and school culture (T. S. Lee 2007). These influences may guide parental FLP decisions (Welsh and Hoff 2021) and impact FLP success (Luo and Wiseman 2000). Nonetheless, the association between FLP in English–Hebrew-speaking bilingual families and their child’s CA and AOB remains unestablished. Investigating FLP may have implications for sociolinguistics and bilingual language acquisition. This study holds the potential to offer practical insights for parents, educators, researchers, and other professionals seeking a deeper understanding of the factors influencing FLP and the approaches to support languages in bilingual homes.

1.1. Context: The Status of English in Israeli Society

In Israel, Hebrew stands as the official language and dominates the public domain (Stavans 2023). However, English has a high status and serves as a primary means of communication internationally and within scientific, academic, and technological spheres (Kopeliovich 2022). English is integrated into everyday life (Stavans 2023). Across numerous sectors, including health, electricity, and telecommunications, companies often offer English as a language option for communication. Television programs and movies are predominantly shown in English, often complemented by Hebrew subtitles. English is also often seen on road signs, shop names, and advertisements.

English is encouraged in Israeli educational institutions (Gordon and Meir 2023). English instruction is compulsory in school. In many state schools, English learning officially commences from third grade. However, students may begin learning English earlier through school or private programs. Third-grade students typically dedicate 2 h per week to studying English as a foreign language. This allocation of time usually increases to 4 h as children get older (see Israel Ministry of Education 2023a). On joining secondary school, the typical number of hours spent studying English is also 4 h per week (see Israel Ministry of Education 2023b). In primary and secondary school, children may be divided according to their English proficiency level. This categorization often distinguishes between students learning English as a second language and those with English as a heritage language. At the end of secondary school, students take a psychometric exam that includes English literacy skills. This is a standardized test that is necessary to be accepted into universities and other higher education institutions.

In Israel, English-speaking immigrants from English-speaking countries make up a small percentage of the population (Zlotnick et al. 2020). English-speaking immigrants often present with higher incomes than other Israeli immigrants, such as those from Russian-speaking backgrounds. They are more likely to find employment opportunities in academic institutions and international companies (Raijman et al. 2015). Moreover, they are less likely to present with proficient Hebrew, attributed to a paucity of social pressure due to the status of English as a lingua franca (Raijman et al. 2015).

1.2. Family Language Policy

Family language policy (FLP) encompasses the language beliefs and planning employed by families to guide language use within the home (Hollebeke et al. 2020). It serves as a link between two fields of study: language policies and bilingual language outcomes of children (Karpava 2022). Subsequently, FLP is considered necessary for heritage language development (Caldas 2012; Dekeyser and Stevens 2019; Romanowski 2021; Spolsky 2004, 2012).

The FLP framework (Spolsky 2012) is frequently adopted to conceptualize language planning within the home. It comprises three distinct yet interrelated key components (Shohamy 2006; Spolsky 2004, 2012). The first component, ideology, focuses on beliefs and attitudes toward the heritage and societal languages (Hollebeke et al. 2022). It is considered the driving force of FLP, formed by many factors (Curdt-Christiansen and Huang 2020). For example, beliefs regarding language as capital (language for professional, academic, and economic advancement), language as connection (language to form relationships), and language as culture (language for relating to the community’s customs and values) are considered crucial aspects of ideology (Seo 2017). Ideology may predict language outcomes in both languages (Ronderos et al. 2022; Rose et al. 2023).

The second component, management, delves into the strategies that the family plans to employ to govern language use within the home. It encompasses how families opt to introduce languages, such as employing strategies like ‘one parent one language’, both parents speaking the heritage language, or parents speaking both languages (Piller 2001). It also includes discourse strategies, which may encourage a monolingual (i.e., heritage language only) or bilingual home (i.e., switching between languages) (De Houwer and Nakamura 2022; Lanza 1997).

The third dimension, practice, pertains to the observed behaviors in the home. Practice encompasses both the strategies employed and language use. It is indicated to influence bilingual language outcomes (Cohen et al. 2021). However, parents may be oblivious to strategies that they employ (De Houwer and Nakamura 2022). Consequently, relying solely on parental interviews may prove insufficient to capture practice. Child interviews may also offer valuable insights (e.g., see Altman et al. 2014; Rose et al. 2023).

Policies in the home may be shaped by internal (e.g., emotions, identity) and external factors (e.g., socioeconomic status, language status, and institutional support) (see Curdt-Christiansen and Huang 2020; De Houwer 2020). The FLP may be modified if it is perceived as unsuccessful (Caldas 2012) and evolve in response to the changing environment (Kaveh 2020; Kaveh and Sandoval 2020; Smith-Christmas 2022b). FLP is potentially subject to the influence exerted by temporal dimensions. The influence may be attributable to the alterations in the environment that occur concomitantly with the developmental changes observed in children (e.g., starting school; Kaveh and Sandoval 2020). Two dimensions of time that are of interest to examine are the chronological age (CA) and age of onset of bilingualism (AOB). It has been established that they distinctly impact bilingual language development (Altman et al. 2022; Armon-Lotem et al. 2021; Meir 2023; Paradis 2023). However, there has been a paucity of extensive inquiry conducted within the context of FLP.

1.3. Chronological Age and Family Language Policy

Chronological age (CA) is an essential demographic variable used to determine children’s developmental expectations, including physical, emotional, and cognitive abilities. The impact of CA on language development is well known. As typically developing children get older, their language abilities improve. However, differences are observed in the heritage and societal languages. The trend is that younger children are more proficient in the heritage language, possibly as they spend more time in the home (Barron-Hauwaert 2004). As children age, and experience more school, they become more dominant in the societal language (Kaveh and Lenz 2022; Montrul 2018). The heritage language becomes secondary, diverging from native speakers’ outcomes (Scontras et al. 2015). However, late introduction of the societal language (i.e., around ages 9–12) may affect the capacity to attain native language levels (Montrul 2023).

Chronological age (CA) may also influence FLP. Parents may have more control over the FLP with younger children. Younger children may adopt the attitudes of those around them, including parents, teachers, and friends. Notably, positive parental attitudes play a pivotal role in cultivating children’s inclination to learn a language (Asmali 2017). However, research is not definitively conclusive as young children have been shown to possess the ability to influence the language used within their home (Smith-Christmas 2016). Nonetheless, it is generally expected that older children have more control over the languages used at home (De Houwer and Nakamura 2022). For instance, when children go to school, they may increase societal language use (Bayley and Schecter 2003; Kaveh 2020). They may even be observed using their own management strategies, encouraging parents to repeat phrases from the heritage language to the societal language (Gafaranga 2010). Thus, the trend is that children’s desire to continue to use their heritage language may decrease with age. Ergo, preference for the societal language may increase over time.

Children’s language shift in language preference may shape FLPs as parents tend to accommodate their children’s choice of communicating in the societal language, thereby reducing conflict (De Houwer 2020; Kheirkhah and Cekaite 2015). However, not all older children choose to communicate in the societal language at home. Children’s perspectives may be influenced by external pressures, such as school policy (B. Y. Lee 2013). The heritage language (English) of the population in this study (English–Hebrew speakers) is supported within Israeli society. However, mere exposure may not be sufficient. Spanish is a globally spoken language and there remains a decline in heritage Spanish (e.g., Montrul 2022; Stavans and Ashkenazi 2022).

Finally, parents’ desire for their children to speak the heritage language could reduce with age and their strategies may relax (Paradowski and Michałowska 2016). This may signify the belief that children have achieved proficiency in both languages, rendering heritage language support no longer essential (Barron-Hauwaert 2004). Parents of older children may perceive the societal language as more essential for accessing activities and meeting academic demands (Bayley and Schecter 2003; Welsh and Hoff 2021). Accordingly, as children get older, the opinions of both parents and children may modify and this may influence the management and use of language within the home.

1.4. Age of Onset of Bilingualism and Family Language Policy

Age of onset of bilingualism (AOB) is a measurement of language exposure that influences linguistic proficiency in the heritage language (Armon-Lotem et al. 2021; Meir 2023; Paradis 2023) and societal language (Altman et al. 2022; Muñoz and Singleton 2011). Presently, there is no agreement on the definition of early or late AOB (Strangmann et al. 2019). In general, the later the AOB, the more advanced the heritage language outcomes (Montrul 2023) and the less proficient the societal language outcomes (Altman et al. 2022; Hamann and Abed Ibrahim 2017). However, the role of AOB seems to endure in heritage language outcomes but not societal language outcomes. Societal language proficiency tends to align with monolingual peers’ proficiency after they receive education in the societal language for a few years (Paradis 2023). Moreover, a strong foundation in the heritage language prior to exposure to the societal language may even facilitate the acquisition of the societal language (Schwartz 2014).

The length of time being exposed to one language may play a crucial role in FLP. It sheds light on the developmental stage that a child is experiencing when bilingualism commences and provides information on the context in which the languages were introduced (Armon-Lotem et al. 2021). For instance, an early AOB may indicate that both languages were introduced in the home. A late AOB may indicate that the child is a sequential bilingual and was exposed to the societal language in more formal settings once the heritage language was established. AOB may also illuminate patterns of parent language choice. Children whose parents adopt the one parent one language approach are likely to have an early AOB and speak the societal language more proficiently. Children with a later AOB are more likely to hear the heritage language from both parents and either achieve balanced bilingualism or attain greater proficiency in the heritage language by the age of 3.5 (De Houwer 2023).

Lastly, there may be a psycho-emotional dimension to consider (Tannenbaum 2012; Tannenbaum and Yitzhaki 2016). Children’s agency influences FLP, as they may react to the introduction of societal language, impacting the success of FLP (Fogle and King 2013; Shen and Jiang 2023; Smith-Christmas 2022a, 2022b). Children under 2 years old have displayed rejection of management strategies to encourage heritage language use when they are introduced to the societal language at daycare (Slavkov 2015). This suggests that a later AOB may change the timing of heritage language rejection. Nonetheless, while the timing of exposure (i.e., AOB) may logically influence FLP, there is little evidence available to support this assertion.

2. Research Questions

Whilst most studies concentrate on defining FLP at a particular moment, it is anticipated that time plays a crucial role in mediating a family’s policies (Lanza and Lomeu Gomes 2020). Within the context of FLP, the dimensions of age (CA and AOB) have rarely been examined, despite the potential they have in shaping parenting ideologies, management strategies, and language practices. Enhancing our understanding of language policy in the home holds implications for refining FLP frameworks. It may have consequences for professionals who support bilingual families, helping to deepen their understanding of the language decision-making processes within the home. Accordingly, this study aims to answer two research questions:

• Is chronological age associated with the family language policy of English–Hebrew-speaking families?

As children get older, parents’ aspirations for their children to use their heritage language may decrease and they may moderate their use of strategies (Paradowski and Michałowska 2016). Furthermore, as children get older and start school, there may be an increase in societal language use (a subcomponent of practice) (Kaveh 2020; Kaveh and Sandoval 2020). This may result in a shift in language preference from the heritage to the societal language, and parents may adjust their FLP to align with their children’s preferred language choice (De Houwer 2020; Kheirkhah and Cekaite 2015). However, English as a heritage language could be unique. There may be no significant differences between the FLP of families with children of different ages as it is taught and promoted in the educational systems (see Israel Ministry of Education 2023a, 2023b). Moreover, the status of English as a lingua franca may reduce the likelihood of a language preference shift as there is less societal pressure to stop speaking English (Raijman et al. 2015). Thus, irrespective of age, it is predicted that English may continue to be supported in the home.

2.

Is age of onset of bilingualism associated with the family language policy of English–Hebrew-speaking families?

The AOB is often adopted as a measurement of bilingual language exposure (e.g., Altman et al. 2022; Paradis 2023). However, it could also provide information on the timing and context of bilingualism (Armon-Lotem et al. 2021). A later AOB indicates that languages were introduced sequentially and that whilst the heritage language was introduced in the home, the societal language may have been introduced in an educational institution. Children with a later AOB tend to encounter the heritage language through exposure from both parents. In contrast, children with an earlier AOB may predominantly receive heritage language exposure from one parent (De Houwer 2023). This could have repercussions for FLP. Furthermore, AOB may have implications for children’s language preferences and attitudes. Children may reject pro-heritage language policies (Slavkov 2015). It is reasonable to posit that the earlier the AOB, the sooner the heritage language might encounter resistance or rejection. Accordingly, it is predicted that families of children with a later AOB are more likely to exhibit a stronger pro-heritage language FLP compared to those with an earlier AOB.

3. Materials and Methods

This study is part of a larger cross-sectional study on the FLPs of English–Hebrew-speaking families (see Rose et al. 2023). The population examined were speakers of English as a heritage language and Hebrew as a societal language. Parents hold an ideology that is pro-bilingual, valuing both languages (Rose et al. 2023). However, in comparison to other immigrant populations where Hebrew serves as the societal language (such as Russian–Hebrew speakers), the importance of Hebrew might be less prominent, mainly due to the pervasive influence of English (Amit 2010; Joffe 2018). Accordingly, most children were exposed to English in the home and Hebrew out of the home (Rose et al. 2023).

In-depth background history was obtained using a parent background questionnaire, adapted from the bilingual parental questionnaire (BIPAQ) (Abutbul-Oz and Armon-Lotem 2022). Maternal educational level, quantified by the number of years of schooling completed (

Table 1. Sociodemographic characteristics of participants.

	All Participants
	n = 82 (41 Female; 41 Male)
	M	SD	R
Chronological age (years)	8.98	3.27	5.08–14.08
Age of onset of bilingualism (years)	1.79	2.25	0–10.92
Maternal education (years)	17.23 a	1.80	13–24

Note. ^a Missing data for 1 participant.

), served as a proxy for socioeconomic status (see Faraj and Hamid 2023; Meir and Armon-Lotem 2017). It was determined that families had a mid-high socioeconomic status. Both parents had resided in Israel for a minimum of 2 years (mothers = 2–40 years (M = 14.86, SD = 9.75); fathers = 2–45 years (M = 19.04, SD = 12.82)). Values are missing for 1 mother and father who immigrated as children. Inclusion criteria specified that at least one immigrant parent was from an English-speaking country (e.g., Australia, UK, and USA). More than 90 percent of mothers and fathers strongly agreed that they were fluent in and understood English. In contrast, only 27 percent of mothers and 54 percent of fathers strongly agreed that they were fluent in Hebrew, while 44 percent of mothers and 59 percent of fathers strongly agreed that they understood Hebrew. Values are missing for 1 mother and 1 father.

Data were collected from 82 (41 females) English–Hebrew bilingual children, with a mean age of 8.98 years (R = 5.08–14.08, SD = 3.27). The children included 36 kindergarten children aged 5.08–6.67 years (M = 5.71, SD = 0.37), 25 primary school children aged 9.0–10.5 years (M = 9.78, SD = 0.34), and 21 secondary school children aged 13.08–14.08 years (M = 13.63, SD = 0.34). These age-related distinctions align with the structure of the educational system in Israel and reflect the diverse academic expectations, literacy levels, and language development stages across three distinct schooling periods (Rose et al. 2023). The decision to include a larger proportion of kindergarteners compared to other age groups was based on the greater variability typically observed among kindergarten-aged children in contrast to older age groups. See Table 1 for sociodemographic data. Mothers’ mean age was 40.94 years (R = 28–54, SD = 4.94, missing 2) and fathers’ mean age was 42.48 years (R = 31–55, SD = 4.73, missing 4). Over 70 percent of the children were born in Israel. They had no known neurodevelopmental, language, emotional, or behavioral disorders. Furthermore, all the children attended Hebrew-speaking kindergartens and schools with a majority of native Hebrew speakers. The mean number of children in each family was 3.88 (range 1–8) and most of the participating children had older siblings. In this study, there were 8 pairs of siblings, with 3 children from one family. To participate in this study, parents’ consent and children’s assent were secured. The study was approved by the Faculty of Humanities’ Institutional Review Board.

3.1. Materials and Procedure

Data were collected by means of two online surveys (one for children and one for parents) that were developed for the purpose of this study, which was part of a larger research project (see Rose et al. 2023). The surveys were based on a questionnaire originally developed by the second and third authors and colleagues for Russian–Hebrew speakers (Altman et al. 2014, 2021), and grounded in studies on sociolinguistics (e.g., Allard and Landry 1986, 1994; Anderson 1996; Lambert 1990; Sachdev and Bourhis 2005). They required participants to evaluate FLP statements on a five-point Likert scale. The complete survey can be found in Appendix A. Descriptive statistics (M and SD) of each FLP statement are provided in Supplementary Table S1. The parent survey was concerned with ideology, management, and language use (a subcomponent of practice) within the home. Following the completion of the survey, parents joined the first author on Zoom and were given the opportunity to ask questions. The child survey focused on observed behaviors in the home, providing an insight into family language practices. Overall, Cronbach alphas ranged from 0.69 to 0.83, all surpassing the acceptable threshold (Babbie 2021). The alphas for ideology in both Hebrew and English were 0.80. Regarding management, the alphas were 0.74 for pro-English and 0.69 for pro-Hebrew strategies. For practice, the alphas were 0.73 and 0.83 for English- and Hebrew-focused practice, respectively.

Children met with an interviewer on Zoom for around 20 min in their preferred language. Online service delivery has been proposed as effective for engaging with children and gathering data (e.g., Hidayat et al. 2022; Manning et al. 2020). Moreover, it has been asserted that establishing rapport with children online, including those aged 5 to 14 years old, is achievable. The flexibility in scheduling and the necessity for assessors to be well prepared and organized may enhance the relationship (Akamoglu et al. 2018). Seventy children were interviewed in English and twelve were interviewed in Hebrew. It is plausible that most children opted for English during the interviews due to their familiarity with communicating in English within their home environment.

Following Altman et al. (2014), the ‘magic ladder’ was adopted to help the children rate on the Likert scale. The ‘magic ladder’ provides children with a visual representation of the Likert scale. Each rung of the ladder is equivalent to a number on the Likert scale and there is a happy face at the top and a sad face at the bottom. The more a child agrees with a statement, the closer their rating is to the smiley face. While the ‘magic ladder’ served as a learning tool for all the children, a subset of them, mostly older ones, consistently demonstrated the ability to use the Likert scale and did not need the support of the ‘magic ladder’ during the actual interview.

3.2. Data Analysis

Two inferential statistics techniques were adopted to enhance our understanding of the association between dimensions of time (AOB and CA) and FLP. Spearman correlations investigated the linear association between ratings (1 = not at all to 5 = very much) of FLP statements and age variables (CA and AOB). Multiple logistic regressions were performed to provide a deeper understanding of the relationships. This choice of statistical analysis is commonly adopted in questionnaire-based research with similar sample sizes, offering valuable insights into relationships between variables and enhancing our understanding of the data (e.g., Faraj and Hamid 2023; Quay and Nakamura 2023). These approaches are also well suited for handling ordinal outcome variables that may be characterized by limited variability and non-normal distributions (Bürkner and Vuorre 2019; Schober et al. 2018).

Both CA and AOB can be categorized in assessments and educational institutions. Supplementary Tables S2 and S3 present the data with CA and AOB as categorical constructs.

4. Results

4.1. Associations between Chronological Age and Family Language Policy

4.1.1. Correlations

Spearman correlation coefficients were performed to investigate the linear relationship between FLP and CA (Table S4). Quartile analysis revealed a strong parental consensus. Notably, most of the ideology components that correlated significantly with CA had quartile 1 values of 4 and quartile 3 values of 5.

A relationship between CA and the three components of FLP was identified. Regarding ideology, the associations mostly indicated a decrease in the importance of English as a child gets older. There were negative correlations between CA and English language as capital, such as the importance of English for school (r = −0.23, p < 0.05) and employment (r = −23, p < 0.05). Furthermore, as the importance of English diminished, the importance of Hebrew increased with CA. For instance, there was a positive correlation between CA and Hebrew language as culture, reflecting the importance of a child’s connection to Israel (r = 0.22, p < 0.05). Nonetheless, English still retains some value as children get older. This was indicated by the positive correlation between CA and the status of English (e.g., parent rating of Israel’s opinion on the status of English (r = 0.26, p < 0.05)).

Management’s association with CA aligns with ideology. As a child gets older, the promotion of English in the home decreases. This is evinced by the negative correlation between CA and the use of TV to promote English (r = −0.27, p < 0.05).

The relationship between practice and CA offers additional evidence of a transition from prioritizing English to emphasizing Hebrew as CA increases. There is a promotion of Hebrew with age, exemplified by the practice of encouraging Hebrew use with siblings (r = 0.28, p < 0.05).

4.1.2. Multiple Ordinal Logistic Regression Analysis

Multiple ordinal regression analysis (

Table 2. Summary of multiple ordinal logistic regression analysis for chronological age predicting ratings of family language policy components (significant models only). Age of onset of bilingualism as a covariate.

Criterion (DV)	Predictors (IVs)	B	SE B	Wald χ2	p	95% CI
Ideology via parent questionnaire
Israel’s opinion (English) a	CA	0.02	0.01	5.62	0.018	0.00	0.03
	AOB	−0.00	0.01	0.07	0.790	−0.02	0.02
School (English) b h	CA	−0.01	0.01	2.44	0.118	−0.02	0.00
	AOB	−0.00	0.01	0.05	0.824	−0.02	0.01
Employment (English) b h	CA	−0.01	0.01	3.53	0.060	−0.02	0.00
	AOB	−0.00	0.01	0.07	0.792	−0.02	0.01
Child sense of belonging (Israel) c h i	CA	0.02	0.01	4.46	0.035	0.00	0.04
	AOB	−0.01	0.01	0.23	0.633	−0.03	0.02
Management via parent questionnaire
One parent one language d i	CA	−0.02	0.01	4.09	0.043	−0.03	−0.00
	AOB	−0.03	0.02	2.07	0.151	−0.06	0.01
TV in English e	CA	−0.01	0.01	5.98	0.014	−0.03	−0.00
	AOB	0.00	0.01	0.16	0.694	−0.01	0.02
Practice via child questionnaire
Everyone speaks English f	CA	0.01	0.01	2.31	0.129	−0.00	0.02
	AOB	0.01	0.01	1.72	0.190	−0.01	0.03
Parent requests clarification in English g i	CA	0.01	0.01	4.64	0.031	0.00	0.03
	AOB	0.00	0.01	0.12	0.732	−0.01	0.02
Hebrew with siblings e i	CA	0.02	0.01	7.96	0.005	0.01	0.03
	AOB	−0.01	0.01	1.08	0.299	−0.02	0.01

Note: Bold text indicates statistically significant results (p < 0.05). In all the models, the model fit was significant, indicating that the final models provide a significant improvement over the baseline intercept-only models (except for ^h). Pearson and Deviance goodness of fit statistics were not significant, indicating that the fits were good. The test of parallel lines was insignificant to ensure the proportional odds assumption was held (except for ⁱ). Consult Table S4 to gain insights into the spread of rating distribution. CA = chronological age; AOB = age of onset of bilingualism; ^a language status; ^b language as capital; ^c language as culture; ^d introducing languages; ^e promoting language strategies; ^f language use; ^g interaction strategies parent/s adopt if child speaks Hebrew at home.

) confirmed a statistically significant association between FLP and CA. The results indicate that in general, English diminished in importance and Hebrew increased in importance with CA. Regarding ideology, the importance of Hebrew as culture was predicted by CA (B = 0.02, Wald χ² = 4.46, p = 0.035, SE B = 0.01, 95% CI [0.00, 0.04]). Focusing on management, parents were less likely to promote English as children got older. Specifically, they were less likely to encourage children to watch TV in English (B = −0.01, Wald χ² = 5.98, p = 0.014, SE B = 0.01, 95% CI [−0.03, −0.00]). Lastly, in practice, parents were more likely to promote Hebrew use with siblings as children aged (B = 0.02, Wald χ² = 7.96, p = 0.005, SE B = 0.01, 95% CI [0.01, 0.03]).

4.2. Associations between Age of Onset of Bilingualism and Family Language Policy

4.2.1. Correlations

Spearman correlation coefficient was calculated to investigate the linear relationship between FLP and AOB (Table S5). Firstly, analyses revealed that as children’s AOB increased, there was a corresponding increase in ratings focusing on English and a decrease in ratings focusing on Hebrew. In terms of ideology, positive correlations emerged between AOB and the importance of English for connecting with people (e.g., siblings and friends (r = 0.29, p < 0.01, r = 0.22, p < 0.05, respectively)). Similarly, positive correlations were found between AOB and ratings reflecting beliefs associated with English culture, encompassing the child’s attachment to and pride in their English-speaking country (r = 0.24, p < 0.05, r = 0.34, p < 0.01, respectively). In contrast, ratings concerning the belief in the importance of Hebrew as a means of connecting with other family members (e.g., grandparents) showed a negative correlation with AOB (r = −0.23, p < 0.05).

A negative correlation between Hebrew and AOB was also observed in management. The adoption of strategies to promote Hebrew, such as encouraging Hebrew use with siblings and family not living at home, decreased as AOB increased (r = −0.24, p < 0.05, r = −0.45, p < 0.001, respectively).

Analyses of the association between practice and AOB aligned with findings examining AOB, ideology, and management. The use of English positively correlated with AOB, including both parent and child ratings of the child’s use of English with parents (r = 0.28, p < 0.01, r = 0.25, p < 0.05, respectively). In contrast, parent and child ratings of the child’s use of Hebrew with parents negatively correlated with AOB (r = −0.35, p < 0.01, r = −0.25, p < 0.05, respectively). Further analyses of language strategy use in practice supported these results. The promotion of English usage with non-resident family members positively correlated with AOB, while encouraging Hebrew usage with non-resident family members displayed a negative correlation with AOB (r = 0.36, p < 0.01, r = −0.32, p < 0.01, respectively).

Lastly, results revealed an association between AOB with the method by which languages were introduced. A rise in AOB showed a negative correlation with management ratings that indicated Hebrew was introduced at home. Consequently, the ratings indicating the introduction of both languages simultaneously, as well as English at home followed by Hebrew at home, and the one parent one language approach exhibited a negative correlation with AOB (r = −0.58, p < 0.01, r = −0.31, p < 0.01, r = −0.26, p < 0.05, respectively). Furthermore, management ratings indicating that Hebrew was introduced outside of the home exhibited a positive correlation with AOB (r = 0.52, p < 0.01).

4.2.2. Multiple Ordinal Logistic Regression Analysis

Multiple ordinal regression analysis revealed a statistically significant association between FLP and AOB (

Table 3. Summary of multiple ordinal logistic regression analysis for age of onset of bilingualism predicting ratings of family language policy components (significant models only). Chronological age as a covariate.

Criterion (DV)	Predictors (IVs)	B	SE B	Wald χ2	p	95% CI OR
Ideology via parent questionnaire
Social benefits a g h	AOB	0.03	0.02	3.17	0.075	−0.00	0.06
	CA	−0.01	0.01	1.71	0.191	−0.03	0.01
Speak with siblings in English b	AOB	0.02	0.01	4.22	0.040	0.00	0.04
	CA	−0.01	0.01	4.24	0.040	−0.03	−0.00
Speak with friends in English b	AOB	0.02	0.01	5.70	0.017	0.00	0.04
	CA	−0.01	0.01	3.03	0.082	−0.02	0.00
Child attachment to English-speaking country c g	AOB	0.01	0.01	2.65	0.103	−0.00	0.03
	CA	−0.00	0.01	0.33	0.568	−0.01	0.01
Child pride for English-speaking country	AOB	0.02	0.01	5.34	0.021	0.00	0.03
	CA	0.00	0.01	0.05	0.820	−0.01	0.01
Speak with other family members in Hebrew b g h	AOB	−0.01	0.01	0.70	0.403	−0.02	0.01
	CA	−0.01	0.01	2.42	0.120	−0.02	0.00
Management via parent questionnaire
Both languages at home together d	AOB	−0.06	0.02	15.77	<0.001	−0.09	−0.03
	CA	0.00	0.01	0.17	0.681	−0.01	0.01
English at home; Hebrew out of the home d	AOB	0.05	0.02	5.58	0.018	0.01	0.08
	CA	0.01	0.01	1.32	0.250	−0.01	0.02
English at home, then Hebrew at home d g h	AOB	−0.02	0.01	2.91	0.088	−0.04	0.00
	CA	−0.01	0.01	0.62	0.432	−0.02	0.01
One parent one language d	AOB	−0.03	0.02	2.07	0.151	−0.06	0.01
	CA	−0.02	0.01	4.09	0.043	−0.03	−0.00
Play with siblings in Hebrew e g h	AOB	−0.01	0.01	0.71	0.400	−0.03	0.01
	CA	−0.01	0.01	0.76	0.385	−0.02	0.01
Hebrew with family not living at home e	AOB	−0.06	0.02	11.19	<0.001	−0.09	−0.02
	CA	−0.01	0.01	1.05	0.306	−0.02	0.01
Language use (subcomponent of practice) via parent questionnaire
Both parents speak English g	AOB	0.04	0.02	3.09	0.079	−0.00	0.08
	CA	0.00	0.01	0.12	0.725	−0.01	0.02
Everyone speaks English g	AOB	0.03	0.01	4.28	0.039	0.00	0.05
	CA	−0.00	0.01	0.02	0.886	−0.01	0.01
Child speaks English with parents g h	AOB	0.01	0.01	1.60	0.206	−0.01	0.03
	CA	0.00	0.01	0.00	0.949	−0.01	0.01
Child speaks Hebrew g h	AOB	−0.01	0.01	3.24	0.072	−0.03	0.00
	CA	0.00	0.01	0.60	0.440	−0.01	0.02
Child speaks Hebrew with parents g h	AOB	−0.01	0.01	1.70	0.193	−0.03	0.01
	CA	−0.00	0.01	0.33	0.566	−0.01	0.01
Language use (subcomponent of practice) via child questionnaire
Both parents speak English g h	AOB	0.02	0.01	2.58	0.109	−0.00	0.04
	CA	0.00	0.01	0.29	0.592	−0.01	0.02
Everyone speaks English	AOB	0.01	0.01	1.72	0.190	−0.01	0.03
	CA	0.01	0.01	2.31	0.129	−0.00	0.02
Child speaks English with parents g	AOB	0.02	0.01	3.16	0.075	−0.00	0.04
	CA	−0.01	0.01	3.45	0.063	−0.02	0.00
Child speaks Hebrew g	AOB	−0.01	0.01	2.00	0.157	−0.03	0.00
	CA	−0.00	0.01	0.14	0.714	−0.01	0.01
Child speaks Hebrew with parents g h	AOB	−0.01	0.01	2.24	0.134	−0.03	0.00
	CA	0.00	0.01	0.04	0.843	−0.01	0.01
Other subcomponents of practice via child questionnaire
Both languages at home together d h	AOB	−0.05	0.01	12.52	<0.001	−0.07	−0.02
	CA	0.02	0.01	5.57	0.018	0.00	0.03
English at home; Hebrew out of the home d g h	AOB	0.02	0.01	2.63	0.105	−0.00	0.04
	CA	−0.01	0.01	0.75	0.386	−0.02	0.01
One parent one language d h	AOB	−0.09	0.03	8.37	0.004	−0.14	−0.03
	CA	−0.01	0.01	2.94	0.086	−0.03	0.00
Talk in Hebrew f g	AOB	−0.01	0.01	1.08	0.300	−0.03	0.01
	CA	−0.00	0.01	0.06	0.805	−0.01	0.01
English with family not living at home e h	AOB	0.04	0.01	6.70	0.010	0.01	0.06
	CA	0.01	0.01	1.11	0.292	−0.01	0.02
Hebrew with family not living at home e g	AOB	−0.02	0.01	4.64	0.031	−0.04	−0.00
	CA	−0.00	0.01	0.02	0.895	−0.01	0.01
Watch TV in Hebrew e	AOB	−0.03	0.01	8.12	0.004	−0.05	−0.01
	CA	0.01	0.01	1.44	0.230	−0.00	0.02

Note: Bold text indicates statistically significant results (p < 0.05). In all the models, the model fit was significant, indicating that the final models provide a significant improvement over the baseline intercept-only models (except for ^g). Pearson and/or Deviance goodness of fit statistics were not significant, indicating that the fits were good. The test of parallel lines was insignificant to ensure the proportional odds assumption was held (except for ^h). Consult supplementary material Table S5 to gain insights into the spread of rating distribution. CA = chronological age; AOB = age of onset of bilingualism; ^a benefits of bilingualism; ^b language as connection; ^c language as culture; ^d introducing languages; ^e promoting language strategies; ^f interaction strategies parent/s adopt if child speaks Hebrew at home.

). Consistent with the findings from correlation analyses, a later AOB was linked to higher ratings pertaining to questions centered around English, while conversely, the opposite pattern emerged concerning Hebrew. For instance, within ideology, a higher AOB predicted a greater emphasis on the importance of speaking English with siblings and friends (B = 0.02, Wald χ² = 4.22, p = 0.040, SE B = 0.01, 95% CI [0.00, 0.04], B = 0.02, Wald χ² = 5.70, p = 0.017, SE B = 0.01, 95% CI [0.00, 0.04], respectively). In terms of management, a later AOB predicted lower ratings in promoting Hebrew with family members residing outside the home (B= −0.06, Wald χ² = 11.19, p < 0.001, SE B = 0.02, 95% CI [−0.09, −0.02]). Lastly, within the realm of practice, English language use was predicted to rise with AOB, such that a later AOB predicted an increase in the parent rating of everyone speaking English at home (B = 0.03, Wald χ² = 4.28, p = 0.039, SE B = 0.01, 95% CI [0.00, 0.05]). Additionally, parents were more inclined to promote English with family not residing at home when AOB was later, whereas the antithesis was observed for Hebrew ((B = 0.04, Wald χ² = 6.70, p = 0.010, SE B = 0.01, 95% CI [0.01, 0.06], B= −0.02, Wald χ² = 4.64, p = 0.031, SE B = 0.01, 95% CI [−0.04, −0.00], respectively).

The AOB also influenced the approach to introducing languages. In terms of management, a later AOB increased the likelihood that Hebrew was introduced out of the home (B = 0.05, Wald χ² = 5.58, p = 0.018, SE B = 0.02, 95% CI [0.01, 0.08]). Moreover, both management and practice ratings revealed that a later AOB reduced the likelihood of introducing Hebrew in the home. For example, introducing both languages simultaneously within the home became less likely with a later AOB (Management; B= −0.06, Wald χ² = 15.77, p < 0.001, SE B = 0.02, 95% CI [−0.09, −0.03]; Practice; B= −0.05, Wald χ² = 12.52, p < 0.001, SE B = 0.01, 95% CI [−0.07, −0.02]).

5. Discussion

5.1. Summary of Findings

The examination of FLP aims to provide a greater understanding of language variations amongst bilingual children. Child internal factors, including CA and AOB, contribute to bilingual language outcomes (Meir 2023; Paradis 2023). However, there is a dearth of knowledge concerning their contribution within the context of FLP. Accordingly, the primary objective of this study was to investigate the relationship between FLP and age-related variables.

Correlations illustrated an association between these age-related variables (CA and AOB) and FLP. Multiple ordinal regression analysis confirmed that CA and AOB also predict FLP. Findings suggested that parents of older children tend to prioritize their child’s integration into the local culture, reflecting ideological shifts. This shift was reflected in management and practice. For example, English may be promoted less, and older children may be encouraged to use Hebrew in the home more than their younger peers. Nonetheless, the importance of maintaining proficiency in English was still recognized by parents of older children, possibly reflecting its prevalence within Israeli society (Stavans 2023).

Regarding AOB, a later onset was linked to heightened importance placed on English as a heritage language. The AOB was also associated with patterns of language use, as per assertions by Armon-Lotem et al. (2021). As anticipated, the likelihood of the societal language being introduced outside of the home increased with a later AOB. Consequently, with a later AOB, it was less likely that both languages were introduced simultaneously. In line with this finding, a later AOB was associated with a greater use of English in the home. Furthermore, English was more likely to be promoted with a later AOB. For example, child ratings for encouraging English language use with family members not residing at home were likely to increase with AOB. Conversely, the promotion of Hebrew exhibited a decline with AOB, as demonstrated by parent ratings for encouraging the use of Hebrew with family not residing at home, and child ratings for watching TV in Hebrew.

5.2. The Relationship between FLP and CA

Children’s CA is associated with developmental expectations including monolingual and bilingual language milestones and linguistic outcomes. It is expected that older bilingual children will demonstrate more proficient linguistic skills in their societal language compared to their heritage language (Kaveh and Lenz 2022; Montrul 2018). Accordingly, CA is purported to shape FLP. A parent’s desire for their child to communicate in the heritage language may decrease (Paradowski and Michałowska 2016). An increase in societal language use at home may also occur (Kaveh 2020; Kaveh and Sandoval 2020). This study supports these assertions, contrary to predictions that CA would not influence FLP due to the unique status of English within Israel. The findings suggested that families of older children exhibited an FLP that was more pro-societal language than those with younger children. They were more likely to encourage siblings to speak in Hebrew and less likely to encourage them to watch TV in English.

It is plausible that parents may perceive their older children as having achieved adequate linguistic proficiency in both languages. Given the support for English in education (Gordon and Meir 2023) and its prevalence in everyday life (Stavans 2023), parents may shift their focus away from furthering its development.

Additionally, expectations within the home may align with child language preferences and strengths, becoming increasingly pro-societal language as they get older. Parents may have reduced pro-heritage language promotion strategies to help ensure that communication at home is harmonious and that family members are able to express themselves easily (Kheirkhah and Cekaite 2015; Nakamura 2018). In line with this proposition, it is reasoned that parents of older children were more likely to ask their child to use English to clarify what they said in Hebrew because older children were using Hebrew more frequently. Furthermore, it is conceivable that English-speaking parents may not comprehend the higher-level Hebrew that older children are capable of using.

5.3. The Relationship between FLP and AOB

The AOB is frequently used as a measurement of bilingual language exposure and is associated with bilingual linguistic outcomes (Altman et al. 2022; Armon-Lotem et al. 2021). It provides information on the length of time a child was monolingual (Armon-Lotem et al. 2021; Meir 2023). The current study determined that AOB may also shape FLP. For example, parents of children with a later AOB were more inclined to introduce their child to the heritage language in the home and the societal language out of the home. It is not surprising that parents of children with an earlier AOB exhibited a greater proclivity to introduce both languages concurrently. Findings support assertions that AOB exerts an influence on the contextual dynamics surrounding language introduction (Armon-Lotem et al. 2021; De Houwer 2023).

In general, it was also observed that families with children characterized by a later AOB were more likely to favor a pro-heritage language FLP. They tended to assign greater importance to heritage language use by children to connect with siblings and friends (ideology). Parents were less likely to adopt societal language promotion strategies. For example, parents with children with a later AOB were less likely to encourage the use of Hebrew with family that did not live at home. Children confirmed that this was observed in practice. Consistent with findings on CA, expectations within the home may correspond to the language preferences and strengths of children. Children who experience a later AOB remain monolingual for a more extended period, with exposure limited exclusively to the heritage language for a greater length of time. Subsequently, they often present with greater heritage language proficiency (Armon-Lotem et al. 2021). The impact of AOB on the heritage language is maintained into adulthood and is observed even when CA is controlled for (Paradis 2023). Moreover, in addition to the association between a later AOB and advanced heritage language outcomes, the advantages attributed to English as a lingua franca and the reduced social pressure to communicate in Hebrew (Raijman et al. 2015) may also partially explain why parents of children with a later AOB are less inclined to prioritize Hebrew.

The observed patterns indicating a shift toward a stronger emphasis on the societal language with an earlier AOB and later CA suggest that variables related to a child’s age may exert influence over FLP. It lends support to prior research, affirming that parents are not merely constructing FLPs in isolation (Shen and Jiang 2023; Smith-Christmas 2022a). They are forming FLPs in response to the environment and attributes of family members.

5.4. Implications

FLP is a dynamic framework, connecting language policy and child development (Karpava 2022). To understand its formation, the sociolinguistic, socioeconomic, sociopolitical, and sociocultural contexts, as well as the home environment, parental background, and economic resources, should be considered (Curdt-Christiansen and Huang 2020). This study suggests that child internal factors, specifically CA and AOB, also have consequences for FLP. Considering these dimensions of time within the field of FLP resonates with the understanding that child agency influences a family’s language planning (Smith-Christmas 2022a, 2022b).

Findings also have implications for models of bilingual language development. The FLP can potentially influence language outcomes (see Rose et al. 2023), in conjunction with the child’s internal and external factors, such as cognitive ability and parent language proficiency (see Paradis 2023). This study indicates that CA and AOB not only impact bilingual language but also have implications for FLP and the language-learning environment. Subsequently, to understand the complex dynamics of bilingual language development, it is imperative that we delve into how influential sources interact (Meir 2023). To deepen our understanding of these interactions, future research endeavors could examine additional potential influencing factors, including parental proficiency levels in both the heritage and societal languages. Additionally, investigating the linguistic experiences of children in greater detail, such as their participation in out-of-school programs focusing on the heritage language, could provide valuable insights. Furthermore, few relationships between ratings on language use and CA and AOB were detected. Financial and time considerations aside, alternative methods to a Likert scale may include video recordings of family conversations (e.g., Kheirkhah and Cekaite 2015; Lomeu Gomes 2022).

Developing our understanding of FLP holds significance for practitioners and educators tasked with supporting bilingual development. It may contribute to the advice given to parents, helping to manage their expectations and plan language use in the home to achieve their family language goals. The emphasis of this study was on thorough assessment of participants, meticulous item examination for questionnaire development, and deeper data understanding. However, to better understand the patterns and relationships among variables investigated and reveal the most relevant questions, future research might consider a larger sample size, which will enable factor analysis. In addition, to determine if the results can generalize to other language pairs and contexts (especially considering the skewed SD on some of the ratings; see Tables S1–S3), further research within Israel and beyond is advised.

6. Conclusions

Family language policy (FLP) is a dynamic phenomenon, influenced by a plethora of factors. This study has portrayed the FLPs of English–Hebrew bilinguals living in Israel as multifaceted and has shown that children’s CA and AOB may play a fundamental role in shaping FLPs. Younger children and those with a later AOB are more inclined to choose a pro-heritage language FLP compared to their older peers and those with an earlier AOB. Ergo, while it is acknowledged that individual experiences of age diverge, it may be expected that CA and AOB might be considered for inferring the context in which languages are experienced in the home. Findings may be used to inform current FLP frameworks and models of bilingual language development. They also bear practical implications. For instance, a deeper understanding of FLP can help professionals, such as educators and speech and language therapists, provide interventions and support strategies that effectively align with the needs of bilingual individuals. Moreover, understanding FLP dynamics can empower and guide parents to modulate their FLP decision-making process.