Needs and demands for heritage language support in Australia: results from a nationwide survey

ABSTRACT

This paper reports on a survey conducted in Australia among parents raising their children with a heritage language (HL). We found strong awareness of the importance of HL maintenance (95%), but only moderate levels of engagement (37–44%) with community initiatives for HL support. There were significantly more primary school-aged children reported as enrolled in community language schools than pre-school aged children, but attendance at informal initiatives, such as language social groups, was higher for pre-schoolers. Less than 20% of respondents indicated that there was enough HL support available, with significantly less support reported for pre-school aged children. However, parents of children who attended HL initiatives during pre-school years reported a significantly higher HL maintenance success rate. Most parents who did not bring their children to HL initiatives stated that it was because they did not know about them. Parents of school-aged children had more positive opinions about available HL support than parents of pre-schoolers and most parents said that if more initiatives were made available they would commit to attending them on a regular basis. Consequently, we suggest a formal language extension programme at the pre-school level as the solution for improving multilingual families’ experience.

KEYWORDS:

• Heritage language maintenance
• survey
• Australia
• pre-school and school-aged children
• bilingualism

Introduction

Almost half of Australians have at least one parent born overseas; and one in five speaks one of 300 languages other than English at home (ABS 2017). Despite its substantial multilingual capacity, Australia lags globally in harvesting the significant gains to be made from speaking more than one language and from maintaining heritage languages (HLs). An HL can be defined as ‘an immigrant, indigenous, or ancestral language that a speaker has a personal relevance and desire to (re) connect with’ (He 2010, 66), constituting ‘the language associated with one's cultural background’ (Cho, Cho, and Tse 1997, 106). It has been shown that in societies where a HL is the minority language, there is a ‘high risk for it to be replaced by the majority language from the second generation and to be lost in the third generation’ (Tran et al. 2021, 2). This has been argued to be especially the case when English is the majority language (Oriyama 2012). This study investigates the challenges faced by parents raising their children as HL learners in Australia, with a particular focus on the differences between parents with pre-school versus primary school-aged children.

Within Australia’s federal system of government, the level of official commitment to language maintenance and education varies. The Commonwealth (or national) government of Australia, for instance, supports language education for young children through the Early Learning Languages Australia (ELLA) programme (Kaufman et al. 2017), which provides language exposure through digital apps in pre-schools and in the first year of primary school. However, while this programme is welcome in an ancillary function, it is not intended to replace direct human (including peer-to-peer) input. There is also significant State and Commonwealth government support for community language schools which usually operate on weekends (Cardona, Noble, and Di Biase 2008). However, although recognised as important for families and their communities, community language schools been reported to have limited effectiveness in maintaining HLs (Kondo 1997; Lee 2002; Tse 2001; Xia, 2016; cited in Tran et al. 2021). Furthermore, State and Commonwealth support for pre-school aged children has been more limited, since community language schools normally only cater to school-aged children.

This study reports on a nationwide survey with 488 respondents, of which we focus on a subset of 294 parents raising their children with an HL in Australia. Focusing on five key questions from the survey, we present the most salient needs and demands that emerged from parents’ responses about raising multilingual children. In the final section, we propose a formal language extension programme for pre-schools as a solution to addressing these needs and demands, which is already underway in a pre-school in Sydney (see Escudero et al. In Preparation for an overview) and, to the best of our knowledge, is the only overtly bilingual pre-school in that city. This is because our initial survey findings, along with current research, strongly suggest that to enable bilingualism, HL support should be provided from an early age and in the community, outside the home environment (Sun et al. 2020; Escudero et al. 2020).

The bilingual advantage

It was once commonplace to think that HL use had a negative impact on children’s language and reading skills in the majority language, such as e.g. English in the Australian context. However, this subtractive bilingual view is erroneous, as current evidence suggests higher HL exposure and use in fact promotes these skills (Papastefanou, Powell, and Marinis 2019), as well as promoting healthy cognitive development and the potential to improve attention and memory (Adesope et al. 2010; Pino Escobar et al. 2018). Bilinguals also have a distinct cognitive advantage when it comes to language learning, including faster learning of new languages (Grey et al. 2018).

Children can learn to build on their existing language proficiency in one language to acquire another, such as an HL, so acquisition does not necessarily need to be simultaneous. However, frequent, cumulative and sustained input in the HL is known to be crucial to the development of bilingualism (Hurtado et al. 2014; Place and Hoff 2016; Sun et al. 2020; Unsworth 2016) and the quantity of language input strongly influences language proficiency from an early age (Soderstrom and Wittebolle 2013). For babies, parents generally find it easier to use infant-directed speech in their dominant language, and this kind of speech facilitates sustained input.¹

The ‘bilingual advantage’ is by now increasingly understood amongst the general public and can be a ‘powerful language ideology in family language policy decisions’ (Piller and Gerber 2021, 623). However, cognitive advantages of bilingualism may be less pervasive than ‘economic’ advantages, meaning minority languages may be deemed less advantageous for children to learn, particularly when faced with the pressure to learn English as a majority language (Piller and Gerber 2021, 627; see also Romanowski 2021). Furthermore, families’ appreciation of the bilingual advantage and the desire for HL maintenance might not translate into practice: there exists a ‘gap between explicit discourses about language use and actual socialisation practices’ (Guardado 2009, 2018; cited in Romanowski 2021, 1216).

Challenges to bilingual parenting

A particular challenge to bilingual parenting is that ‘[bilingual] educational support outside the home is, by and large, insufficient at all levels of the education system, starting from early childhood education’ (Benz 2017; cited in Piller and Gerber 2021, 623). Consequently, bilingual language development has come to be primarily regarded as a parental responsibility (e.g. Nicholas 2015; Pacini-Ketchabaw and Armstrong De Almeida 2006; cited in Piller and Gerber 2021, 623) and as an activity that occurs primarily within the home (Liu 2021, 123).

De Houwer (2017) has shown that bilingualism needs to be ‘harmonious’ and that many migrant parents, because of low proficiency in the majority language, feel insecure in their parenting role, resulting in an adverse effect on their children’s development. Insecurity can also arise when there is low proficiency in the HL. Saravanan (2001), for example, found that fathers in Singapore whose proficiency in the HL was limited tended to avoid using the HL when speaking to their children. This provides a case for community-level support of HLs, through such mechanisms as HL-medium playgroups, pre-schools, kindergartens and community schools, since conditions at home may not always be the most conducive to HL maintenance, despite the best intentions of one or both parents (see Lising 2022; Poyatos Matas and CuatroNochez 2011; Romanowski 2021). Encouraging children to enhance their HL through, for instance, the medium of drama (Charalambous and Yerosimou 2020) or play in early childhood education settings (Liu 2021) can create many opportunities for children to learn from educators and peers. Other positive effects of HL maintenance outside the home include prevention of early language shift to the dominant language (Fishman 1991), and the development of children’s general language and literacy skills (Papastefanou, Powell, and Marinis 2019). Furthermore, children can benefit from peer-to-peer learning in the HL and establish trust with HL-speaking educators (Liu 2021).

Conversely, incomplete acquisition can result in reduced connections with the broader HL-speaking community and at home. Recent studies of HL maintenance have shown that it can be encouraged even in settings where HL and L2 learners are mixed within the same cohort, such as the English-Korean bilingual early childhood centre in Melbourne described in Liu (2021, 126), which is attended by children of Korean, but also Chinese, Indian, Japanese and Saudi Arabian descent.

Heritage language maintenance in the Australian context

Australia’s language ecology is often characterised as still being caught in the grip of the ‘monolingual mindset’ (Clyne 2007; cited in Eisenchlas and Schalley 2019 and Hajek and Slaughter 2014). As Australia has become increasingly multilingual, some researchers report that policies have become more assimilationist, with policy orientations becoming even more monolingual (Schalley, Guillemin, and Eisenchlas 2016). Language policing is commonplace in Australia, with migrants being led to believe that it is their ‘civic duty’ to learn English, and to do anything other than learn English would render them undeserving of residence and citizenship in the country (Piller 2003). Proficiency in languages other than English is often considered ‘at best an incidental benefit and at worst even a burden’ (Piller and Gerber 2021, 623). Motivations to make the most of the ‘bilingual advantage’ may be set aside by a desire for children to be proficient in English by the time they start primary school (Piller and Gerber 2021).

As a result, Australia is at high risk of language attrition and loss among its migrant population. This is a phenomenon which has already occurred on a large scale with indigenous languages, where less than 40 of the 250 previously spoken indigenous languages are still being used, all of which are highly endangered, and with only 12 of these being learned by children (Bromham et al. 2021). This has led Australia to be known as one of the principal global language endangerment hotspots in the world (Amano et al. 2014; Anderson 2011) and a ‘graveyard for languages’ (Piller, as cited in Fukui 2019). This is despite research suggesting that Indigenous children learning their own indigenous HLs are happier than their peers who are not and there is some evidence that reversing HL shift may help reduce the rate of self-harm and suicide among indigenous youth (Anderson and Kowal 2012; Biddle and Swee 2012; Hallett, Chandler, and Lalonde 2007; see also Forrest 2018).

Little is known about the challenges faced by multilingual families in Australia in accessing external support for HL maintenance. However, there have been studies investigating the attitudes and emotions experienced by multilingual families, both in terms of their experiences of trying to maintain their languages, and in their daily experiences with the majority language, English. Eisenchlas and Schalley (2019), for example, reported that migrant and refugee parents considering raising their children in a bilingual context in Australia, including their HL alongside English, often express fear and hesitation to do so. Their fears include impeding their children’s English acquisition and thus damaging their academic prospects. This reticence is not helped by the receipt of ‘well-intentioned but misguided advice’ from schools and healthcare workers to switch to English as the household language (Eisenchlas and Schalley 2019, 564; see also Willoughby 2015).

Following a series of workshops for migrant parents, Eisenchlas and Schalley (2019, 571) found that only a minority of non-English speaking background parents give much weight to maintaining their HL. They found that even when parents are interested in the matter of bilingualism and HL maintenance, they are frequently unaware of the affective and social impacts of bi/multilingualism or its absence – for both the children and the family as a whole (Eisenchlas and Schalley 2019, 571). For some parents, bilingualism becomes the ‘scapegoat’ – the reason their child has behavioural issues and perceived speech delays and/or disorders, which in any case tend to be over-diagnosed in bilingual children (Eisenchlas and Schalley 2019, 572). Crucially, strong negative effects were found when there was one majority language-speaking parent in the household, where the use of the HL could be reduced or even prohibited (Eisenchlas and Schalley 2019, 572). This finding calls into question the popular One Parent One Language method, as ideally, HL maintenance should be the result of a combined effort of both parents, extended family, school and the wider community. It is in this context that over fifteen years ago, Clyne (2005) called for a collaborative approach involving government, schools, universities, communities, ethnic schools and families, but unfortunately these synergies have yet to be seen playing out on a widespread level in the Australian context.

Tran et al. (2021) examined Vietnamese-Australian parents’ language use and attitudes towards HL maintenance and associated factors including demographic factors (e.g. education, length of stay in English-speaking countries), language practice factors (e.g. language proficiency, language use), language ideology factors (e.g. beliefs in the importance of home language maintenance), and language management factors (e.g. language policies, intention of future residence in the home country). They found that parents used more Vietnamese with their child when they believed in the importance of HL maintenance and of maintaining Vietnamese culture, values and language (Tran et al. 2021, 8–9). This aligns somewhat with findings from Australian-based studies by Fernandez and Clyne (2007) and Perera (2015) for speakers of Tamil, and Sinhala and Tamil, respectively, whereby religious identity was found to be of importance for HL maintenance.

Tran et al. (2021) found that the most significant factor associated with parents’ HL use with their child was parents’ language use in social situations, i.e. with different people, in different situations, and with different media. Conversely, people who had fewer opportunities to use Vietnamese in different contexts used it less with their child (12). These findings highlight the importance of different domains of use for HLs, particularly the role of community events in the promotion of HL maintenance. However, Tran et al. (2021) did not investigate whether children were participating in formal or semi-formal activities, such as community language schools or HL-medium playgroups. Furthermore, the children in the study were mostly above pre-school age (M = 10.28 years, SD = 4.58).

Piller and Gerber (2021) examined attitudes towards bilingualism in an Australian online parenting forum. They found that individual bilingualism in the abstract is viewed positively and as a ‘gift’ for children. However, they also found that beliefs and attitudes do not necessarily transform into practice, with many forum participants expressing concern that a second language in the early years might jeopardise their child’s English proficiency and hence their success in school. Similarly, Lising (2022) reports on a qualitative study of HL maintenance among five Filipino families residing in Sydney, drawing on data from a sociodemographic questionnaire and a semi-structured interview. The families had mostly shifted to using English at home, but the parents expressed regret about this, and wished they had instilled bilingualism in their families from the outset. However, these desires were also contradicted by concerns of the parents, e.g. one of the fathers had trouble speaking English at work when he moved to Australia, so he decided he would benefit from practising speaking English at home as the family language (Lising 2022, 563).

The ‘One Parent One Language’ (OPOL) policy was promoted by the parents studied by Piller and Gerber (2021) as a way for children to ‘compartmentalise’ their two languages and still have enough English to succeed upon entering primary school. However, many of the posters in the forum were English-speaking mothers with partners speaking a Language Other Than English (LOTE), and they found themselves policing ‘slippages’ into English by their partners and in-laws, meaning that OPOL was in fact not resulting in the kind of simultaneous bilingualism they desired. The authors conclude that families do not actually constitute an ideal locus for advocacy and that advocacy ‘needs to continue to target institutions and particularly schools’ (Piller and Gerber 2021, 633). This contrasts with findings by Bissoonauth (2018) for Hindi speakers in Australia, where the conclusion was that grandparents and non-working mothers were pivotal in maintaining the language at home. Lising (2022, 53), however, also cites the lack of community schools for HL Filipino/Tagalog as one of the reasons why the five families investigated had shifted to English-dominance by the second generation.

Romanowski (2021) studied family language policies for HL maintenance via questionnaires and interviews with 26 Polish-speaking parents in Melbourne. The families studied had children ranging in age from 8 to 18. In all cases, one of the parents was Polish-speaking and the other a speaker of English, and in some cases another language as well. Results showed what the author termed ‘discrepancies’ between declared HL maintenance policies and actual practice, whereby parents showed a commitment to an OPOL Policy but often diverged to speaking English, or a mix of English and Polish, in the home. It also emerged that without parents’ reinforcement and establishment of the HL as a default means of communication at home, ‘children suffer from lower productive skills’ (Romanowski 2021, 1214). All of the children were attending Polish Saturday schools (also known as community language schools), but this was not an aspect explored in the study.

Few studies focus on the amount of input bilingual children receive outside of their home in early childhood education settings, or on the amount of output they produce in their HL in the early years. One exception to this is Escudero et al. (2020), who studied the quantity and type of language input received by toddlers and pre-schoolers in a long day care pre-school in Sydney. The pre-school has an open bilingual policy, where two languages (HL Spanish and English) are used in naturalistic interactions between children and educators. Findings showed that HL input was higher than English for toddlers (aged 0–2 years), while pre-schoolers (aged 3–6 years) receiving less HL input exhibited signs of English becoming more dominant. This suggests that, without deliberate intervention, HL input may wane as children become older and as pressures from the majority language become greater, causing HL shift and loss during early childhood (Anderson 2012). The implications of potential early language attrition and subtractive bilingualism in the Australian context remain under-researched. Even less researched are perceptions of needs for HL maintenance of the Australian population. It is imperative therefore to consider the affected population and investigate precisely what they need to successfully maintain their HL in their children. The present study addresses this gap by presenting partial results from an Australia-wide online survey of HL maintenance.

Research questions and predictions

The present study set out to investigate the following research questions:

1. How important is the maintenance of HLs for parents raising their children multilingually in Australia?

2. What is the level of engagement with both formal and informal HL initiatives by parents raising their children multilingually in Australia?

3. What kinds of obstacles are faced by parents raising their children multilingually in Australia?

4. How ready and willing are parents to commit to community initiatives designed to support HLs (and L2s) in Australia?

Our two predictions at the time of survey release to the public were that (1) parents would be interested in maintaining their HL in their children, but (2) they would report lacking sufficient community support to enable the maintenance of their HLs, particularly for their pre-school age children.

Method

We present the findings of our large-scale online survey of HL maintenance, launched across Australia on 2nd November 2020. The survey, hosted on Qualtrics (Qualtrics 2005), had attracted a total of 488 respondents by the date we exported the data for analysis. A total of 294 eligible responses were analysed in the present study.² To prevent attrition, most questions were optional rather than compulsory, and therefore some analyses include samples smaller than 294.

The survey investigated the perceptions, challenges and needs of HL speakers when transmitting and maintaining their minority language in their children, with 71 questions divided into six sections, including: (1) the range and frequency of languages spoken in households; (2) HL maintenance policies and strategies, such as enrolling children in bilingual day care and pre-school programmes, or video chatting with overseas relatives and (3) attitudes towards language maintenance, such as positive evaluations of its cognitive benefits, and negative evaluations of its potential to ‘confuse’ children. The present study focuses on 5 questions concerning the needs and demands for HL support.

Recruitment and data collection

The study was approved by the Western Sydney University Human Research Ethics Committee under project number H11022. Participation was anonymous and voluntary, with respondents asked to provide signed consent before progressing to the survey.

Participants were recruited through multiple methods, such as mailing lists to community language schools, childcare centres, including the pre-school in Sydney where a formal language extension programme was underway (Escudero et al. In Preparation), and research partners. There were posts about the survey on the authors’ and their respective universities’ social media accounts (WhatsApp, Facebook and Twitter), and radio interviews on Australian Broadcast Radio (ABC). The survey was also featured in an article in The Conversation (Diskin-Holdaway and Escudero, 2021). Over 40% of respondents filled out the survey after the Conversation article was released, with almost 20% explicitly stating they found the survey via the article. However, an initial examination showed that there were no significant differences in the responses of those who had found the survey via the Conversation article and those who had not.

Participants

Eligible respondents were (1) parents of at least one child; (2) living in a household where languages other than English were spoken and (3) residing in Australia at the time of filling out the survey. 244 participants included their Australian postcode, with 73% residing in New South Wales, 16% in Victoria, 5.7% in Queensland, 3.3% in Western Australia, 1.2% in South Australia and 0.8% in Northern Territory.

As shown in Table 1, the most reported first/home languages were Chinese – Mandarin (4.08%), French (4.76%), Spanish (18.37%) and Vietnamese (5.10%). Three of these languages are listed among the most commonly spoken community languages reported in the 2016 Australian census, which include Mandarin (2.5%), Arabic (1.4%), Cantonese (1.2%) and Vietnamese (1.2%) (ABS 2016). Our ‘other’ category (11.9%) included self-reported diverse language combinations or groups (e.g. ‘English-Greek’, ‘English-Farsi’, ‘Bosnian-Croatian-Serbian’).

Table 1. First/home languages as reported by survey respondents.

Which are your first/home languages (i.e. the ones you have spoken from birth)?
	n	%			n	%
Arabic	7	2.38		Other	35	11.90
Armenian	4	1.36		Other (not specified)	38	12.93
Bengali/Bangla	2	0.68		Persian	1	0.34
Bosnian	1	0.34		Polish	2	0.68
Chinese – Cantonese	8	2.72		Portuguese	5	1.70
Chinese – Mandarin	12	4.08		Punjabi/Lahnda	2	0.68
Chinese-Other	5	1.70		Russian	7	2.38
Croatian	3	1.02		Serbian	4	1.36
Dutch	4	1.36		Sinhala	2	0.68
English	28	9.52		Spanish	54	18.37
French	14	4.76		Tamil	4	1.36
German	9	3.06		Telugu	2	0.68
Greek	5	1.70		Turkish	1	0.34
Hindi	3	1.02		Urdu	2	0.68
Italian	7	2.38		Vietnamese	15	5.10
Japanese	8	2.72		Total	294	100.00^3

From the 249 respondents who answered the length of residence question, 1.2% had been in Australia for 1 year or less, 10.4% for 2–5 years, 18.1% for 6–9 years, 27.3% for 10–15 years, 6% for 16–20 years and 36.9% for more than 21 years. Thus, most respondents (over 70%) had been in Australia for ten years or longer. Most respondents were also highly educated: 44.7% with a postgraduate degree, 39% with an undergraduate degree, 13.4% with a vocational education or professional trade qualification, and only 2.8% with secondary education.

Respondents reported to have between one and five children, with two children as the median and a total of 534 children: 108 had one child, 141 had two children; 37 had 3, 7 had 4 and one respondent reported 5 children. Only the ages of 512 children were provided, with a mean age of 6.72 years (SDage = 5.83 years, range = 0.1–36 years): 292 were of pre-school age (Mage = 2.89 years, SDage = 1.52 years, range = 0.1–5 years), 187 were of school age (Mage = 10.01 years, SDage = 3.29 years, range = 6–17 years) and 33 were adults (Mage = 21.94 years, SDage = 4.79 years, range = 18–36 years).

Results

Data for 294 respondents who met the eligibility criteria mentioned above were exported from Qualtrics into Microsoft Excel for preliminary review. We used paired samples analyses when comparing age groups (i.e. preschool vs school years), as parents with older children could provide responses from memory of previous experiences from their children’s earlier years, resulting in responses for multiple age groups from some respondents. Analyses were conducted using R (RStudio Team 2020).

Here we discuss the findings of a selection of five survey questions, renumbered for presentation here. We first show the results for Question 1 (Q1) regarding the importance of HL use, followed by questions focusing on involvement in HL community initiatives comparing pre-school and school-aged children.

Importance of speaking HL to children

Table 2 shows that parents, in general and before breaking up their responses into different age groups, find it very important to speak to their children in their HL, with 95% of the sample selecting ‘very important’ or ‘somewhat important’ for this question.

Table 2. Responses to Q1.

Q1. How important is it for you to speak your home languages to your children?	n	%
Very important	247	84
Somewhat important	32	11
Neutral	12	4
Not very important	3	1
Total	294	100

Community initiatives promoting HL maintenance during pre-school and school years

Table 3 shows that attendance at community initiatives for HL maintenance was slightly lower for parents with preschool-aged than school-aged children (37% versus 44%). Although a paired sample t-test revealed no significant difference between age groups (t = −0.294, df = 150, p = 0.769, Cohen’s d = −0.024), attending HL community initiatives during pre-school years predicted higher HL maintenance success rates than attending similar initiatives during school years, as estimated by a linear regression model (F (2, 148) = 5.783, p = .004; see Appendix 1).

Table 3. Did your children attend initiatives promoting heritage language maintenance during pre-school (Q2) or school years (Q3)?

	Q2. Pre-school years		Q3. School years
	n	%	n	%
Yes	100	37	67	44
No	170	63	84	56
Total	270	100	151	100

We also examined the uptake of HL support in pre-school versus school years (Q2 and Q3) as a function of the children’s main or second language. We found some indication of less support in pre-school for specific languages, but responses per language were low due to the many different languages reported (see the Participants section). For example, for children exposed to Mandarin as a main or second language, 6 out of 14 attended HL support initiatives in the preschool years, while 4 out of 5 attended school-age initiatives. Importantly, Mandarin had one of the highest rates of HL community support uptake overall, which contrasts with children exposed to Cantonese, with only one out of 11 attending HL support in preschool and two out of five in school years.

To examine attendance at specific community initiatives, a ‘yes’ response to Q2 directed participants to provide information on their children’s frequency of attendance at a list of community-based activities (Q2a, Q3a). The responses in Table 4 show that for the shared categories across age groups, pre-school and primary school children differed in their attendance at community language schools (CLS), which are state-funded language schools that operate outside of mainstream school hours, while they seemed to have comparable attendance at social groups and in bilingual education settings.

Table 4. Frequency of attendance for HL community initiatives during pre-school and school years. (Responses are displayed in percentages, with counts between parentheses).

	Daily	Weekly	Monthly	Less than monthly	Never	Total
Q2a. Pre-school years
1. Language playgroups	5.1% (5)	42.9% (42)	12.2% (12)	8.2% (8)	31.6% (31)	100% (98)
2. Language social groups	3.1% (3)	23.7% (23)	22.7% (22)	23.7% (23)	26.8% (26)	100% (97)
3. Community language schools	4.2% (4)	33.3% (32)	4.2% (4)	3.1% (3)	55.2% (53)	100% (96)
4. Bilingual day care or pre-schools	9.4% (9)	17.7% (17)	2.1% (2)	4.2% (4)	66.7% (64)	100% (96)
5. Other	3.7% (2)	18.5% (10)	3.7% (2)	3.7% (2)	70.4% (38)	100% (54)
Q3a. School years
1. Community language schools (primary)	9.1% (6)	54.5% (36)	1.5% (1)	6.1% (4)	28.8% (19)	100% (66)
2. Language social groups	3.2% (2)	22.6% (14)	14.5% (9)	25.8% (16)	33.9% (21)	100% (62)
3. Community language schools (secondary)	7.9% (5)	25.4% (16)	3.2% (2)	1.6% (1)	61.9% (39)	100% (63)
4. Bilingual schools	14.5% (9)	4.8% (3)	0.0% (0)	3.2% (2)	77.4% (48)	100% (62)
5. After-school activities	0.0% (0)	10.0% (6)	1.7% (1)	6.7% (4)	81.7% (49)	100% (60)
6. Other	0.0% (0)	3.8% (1)	3.8% (1)	7.7% (2)	84.6% (22)	100% (26)

For statistical analysis, Table 4 data were recoded into two global categories, namely ‘never attended’ = 0 and ‘attended at any frequency’ = 1. We used the non-parametric Wilcoxon rank sum test to compare age groups on this categorical data. As we predicted for a formal initiative such as CLS, frequency of attendance was higher for primary school (M = 0.7121) than for pre-school children (M = 0.4479), W = 2331, p < 0.001. Secondary school children also had a lower CLS attendance than primary school children (W = 27,675, p < 0.001) and comparable attendance to preschool children (W = 3226.5, p = 0.40). In the survey, we chose to ask participants to report attendance of CLS for children in primary and secondary school separately because enrolments in this activity vary according to children’s age, with 74.9% of all enrolments in NSW corresponding to children aged 5–11 (primary school age), while enrolment for older children up to 18 years amounts to only 0.7% (Cruickshank, Jung, and Li 2022).

In contrast to CLS attendance, we found comparable frequency of attendance for two other activities that were included for both pre-school and school-aged children, namely a formal initiative such as bilingual schooling (W = 3296, p = 0.15) and an informal initiative, such as social groups (W = 3219.5, p = 0.34). The results for social groups indicate that parents encourage attendance at any HL maintenance activity that is available. For bilingual schooling, responses mostly indicate a lack of attendance. This is likely because bilingual schooling is unavailable to most families in Australia. Importantly, and as predicted, pre-school children had higher attendance at informal than formal initiatives, as shown by a comparison between the findings for pre-schoolers’ attendance at language social groups versus at CLS (W = 3333.5, p < 0.0001), while language social groups and CLS were attended to the same extent in primary school children (W = 2150, p = 0.54).

Comparing reasons for not attending HL initiatives between pre-school and school children

Participants who selected ‘no’ for Q2 (pre-school years, n = 170) or Q3 (school years, n = 84) were invited to provide the reasons for not attending community initiatives for HL maintenance by selecting as many options that applied to them in Q2b (pre-school years) or Q3b (school years). Table 5 shows these responses, which as for many other questions, were not compulsory, with 161–165 respondents selecting options for Q2b and 77–79 for Q3b.

Table 5. Statements explaining reasons why pre-school- and school-aged children did not attend language activities or programmes. Counts are shown in brackets.

Statement	Group	Strongly disagree	Somew. disagree	Neither agree/disagree	Somew. agree	Strongly agree
1. I did not know of the existence of these programmes	Pre-school (n = 165)	10.9% (18)	11.5% (19)	20% (33)	27.9% (46)	29.7% (49)
	School (n = 78)	16.7% (13)	11.5% (9)	21.8% (17)	19.2% (15)	30.8% (24)
2. I knew of some of the activities or programmes, but they were located far away from me	Pre-school (n = 164)	28.7% (47)	13.4% (22)	26.2% (43)	18.9% (31)	12.8% (21)
	School (n = 79)	29.1% (23)	12.7% (10)	29.1% (23)	17.7% (14)	11.4% (9)
3. I did not have time to take my children	Pre-school (n = 162)	33.0% (55)	8.0% (13)	27.8% (45)	21.6% (35)	8.6% (14)
	School (n = 77)	29.9% (23)	9.1% (7)	36.4% (28)	15.6% (12)	9.1% (7)
4. I am not interested in these activities or programmes	Pre-school (n = 161)	51.6% (83)	19.3% (31)	18.6% (30)	8.7% (14)	1.9% (3)
	School (n = 77)	55.8% (43)	14.3% (11)	18.2% (14)	7.8% (6)	3.9% (3)

Table 5 shows that responses across age groups were comparable for all four statements. To verify this observation, we re-coded and scaled responses as follows: strongly agree and somewhat agree as ‘agree’ = 3; neither agree nor disagree as ‘neutral’ = 2 and strongly disagree and somewhat disagree as ‘disagree’ = 1. The results of four Wilcoxon rank sum tests showed no difference between groups (see Appendix 2). To compare levels of agreement between statements, we conducted a Kruskal–Wallis test using the same three categories (i.e. agree, neutral and disagree) per statement. We chose this test because it is a non-parametric alternative to a one-way ANOVA, which is used instead of the Wilcoxon test for more than two samples (in this case 4 statements). Kruskal–Wallis Preschool H(3) = 97.549, p < 0.001, School years H(3) = 35.101, p < 0.001. As shown in Table 6, the mean agreement response for statement 1 in both age groups is higher than that of the three other statements, indicating that respondents agreed more with the statement relating to lack of awareness of HL activities than with activities being too far away, not having time to bring children to activities, or not being interested in such activities. Table 6 also shows the results of follow-up Wilcoxon pair-wise tests to ascertain the statistical significance of differences between statements.

Table 6. Mean (M) and Standard Deviation (SD) for Q2b & Q3b statements and follow-up pairwise comparisons. Alfa values adjusted with BH (Benjamini-Hochberg) correction.

Wilcoxon pair-wise comparisons with BH correction for Preschool Statements (Q2b)
Statements	2.	3.	4.
1. No awareness (M = 2.35, SD = 0.83)	p < 0.001	p < 0.001	p < 0.001
2. Too far away (M = 1.89, SD = 0.85)	–	p = 0.9	p < 0.001
3. No time (M = 1.88, SD = 0.84)	p = 0.9	–	p < 0.001
4. Not interested (M = 1.39, SD = 0.67)	p < 0.001	p < 0.001	–
Wilcoxon pair-wise comparisons with BH correction for School Statements (Q3b)
1. No awareness (M = 2.22, SD = 0.86)	p = 0.01	p = 0.01	p < 0.001
2. Too far away (M = 1.87, SD = 0.84)	–	p = 0.94	p < 0.001
3. No time (M = 1.86, SD = 0.79)	p = 0.94	–	p < 0.001
4. Not interested (M = 1.42, SD = 0.69)	p < 0.001	p < 0.001	–

Current support available for maintaining heritage languages

The survey included a question on whether respondents felt there was enough support available from their local community to maintain and develop their HLs (Q4). Participants had to provide a response for pre-school and school-aged children, with some respondents selecting N/A if they did not have a child in this age group or if they wanted to skip this part of the question, resulting in an unequal number of responses in each column. As shown in Table 7, most respondents felt that there was not enough support available. This perceived lack of support was higher for pre-school than school-aged children (52.3% vs. 34.9%), which may be linked to the lower levels of availability/awareness of formal opportunities for HL maintenance during these early years (Table 4).

Table 7. Current support for HL maintenance in pre-school versus school years in percentages, with counts in brackets.

Q4. In your opinion, is there enough support available from your local community to maintain and develop your home language/s?	Pre-school years	School years
No	52.3% (135)	34.9% (90)
Somewhat	24.0% (62)	25.6% (66)
Yes	15.1% (39)	18.6% (48)
N/A	8.5% (22)	20.9% (54)
Total	100% (258)	100% (258)

To verify the statistical significance of the observed difference in opinion of current support across groups, we coded a scale of three responses, where ‘No’ = 1, ‘Somewhat & N/A’ = 2 and ‘Yes’ = 3. This coding scheme assumed that ‘somewhat’ and ‘n/a’ represent doubt or not having children in the group. A paired sample t-test on the responses for school-aged (M = 1.84) and pre-school aged children (M = 1.63) was significant (t = −5.538, df = 257, p < .001), indicating that parents of school-aged children had a more positive opinion of the available support.

Readiness to commit to future community initiatives

Q5 asked participants whether, if given the opportunity to attend community initiatives supporting the maintenance of HLs, they would commit to attending them on a regular basis. Examples of such initiatives included community language playgroups, pre-schools, long day care, community language schools for pre-schoolers, and community language socialising groups. These were the same options as those listed in Q2a and Q3a, with responses for pre-school and school-age children. Table 8 shows that most respondents would commit to attending such initiatives on a regular basis, regardless of their children’s age (pre-school = 62.1% vs. school-aged = 61.7%), highlighting comparable levels of positive attitude towards commitment to community HL initiatives in both age groups. Appendix 3 shows responses to a further question on how much time participants would be willing to commit to HL support initiatives in the community.

Table 8. Commitment to potential HL community initiatives in percentages, with counts in brackets.

Q5. If you could get access to new initiatives that support your HL, would you commit to them on a regular basis?	Pre-school years	School years
Yes	62.1% (157)	61.7% (156)
Maybe	18.2% (46)	23.7% (60)
N/A	15.8% (40)	13.0% (33)
No	4.0% (10)	1.6% (4)
Total	100% (253)	100% (253)

Participants were also asked the open-ended question ‘Are there any additional community initiatives that you would like to have available to encourage the maintenance of your home languages? Please give details’. A total of 97 open-ended responses expressed different needs for community initiatives. These responses mentioned playgroups and parents’ groups (17), activities at local libraries (16) and more books in different languages (7). There was also some mention of the need for more government promotion and support for language classes (11), the creation of more public bilingual schools (5) and more after school activities in languages (4) in various locations. Three respondents also expressed disappointment in Australia’s lack of bilingual support in comparison with European countries (see Appendix 4 for an example). The remainder of the responses asked for more activities and classes for specific languages and cultural festivities.

Discussion

Overall, as we initially predicted (see ‘Research questions and predictions’), our survey has shown highly positive attitudes towards the importance of maintenance of HLs among our respondents, with 95% of our sample rating it as important (Table 2). These findings differ somewhat from those of Eisenchlas and Schalley (2019), who found migrant and refugee parents to be somewhat fearful and anxious of HL maintenance, believing that it may hinder their children’s ability to progress in English. We suspect that demographic differences may be the cause of these discrepancies, since the participants in Eisenchlas and Schalley (2019) were a heterogeneous group including migrants and refugees with more varying levels of socioeconomic advantage and education access, whereas most of our respondents were highly educated. Our findings align more with those of Tran et al. (2021), whose sample also consisted primarily of educated professionals, and who largely indicated that it was important for their child to maintain the ability to speak, understand, read and write Vietnamese (M = 3.38; SD = 1.01, where 1 is not at all important and 5 is extremely important). They also align with the positive views towards bilingualism that emerged in a study of an online Australian parenting forum (Piller and Gerber 2021) and of Polish-speaking migrants in Melbourne (Romanowski 2021).

We note that our respondents were self-selecting and were likely already invested or interested in HL maintenance in some way before completing the survey. This is not surprising given that part of the recruitment strategy was to contact people on mailing lists of e.g. community language schools, and the mailing list of a bilingual pre-school in Sydney where a formal language extension programme was underway (Escudero et al. In Preparation). Some had also read expert advice in our article in The Conversation (Diskin-Holdaway and Escudero 2021), see section ‘Recruitment and data collection’.

Our survey results revealed moderate levels of engagement with community initiatives for HL support (Table 3), with attendance slightly lower for parents with preschool-aged than school-aged children (37% versus 44%). However, we found that parents of children who attended HL initiatives during pre-school years reported a significantly higher HL maintenance success rate (Appendix 1). Of the 37–44% of parents who engaged with HL initiatives, over half were sending their children to formal community initiatives, i.e. 44.8% of pre-school children and 71.2% of primary school-aged children (a significant difference between age groups) had some record of attendance at a community language school (Table 4). Attendance at informal initiatives, such as language social groups, was higher for pre-schoolers. Overall, the parents in our sample who did engage with HL initiatives were clearly motivated and invested in taking part in a variety of activities. The main obstacle recorded by parents for not engaging with initiatives was lack of awareness/availability of such initiatives in the community (see also Lising 2022), with neither time nor distance presenting major obstacles (Tables 5–6).

Furthermore, despite the almost unanimous recognition of the importance of HL maintenance, our results showed that caregivers perceive a lack of sufficient formal support to achieve their goal. Less than 20% of respondents indicated that there was enough HL support available (Table 7) and our prediction that the lack of support would be experienced more acutely by parents of pre-schoolers as compared to primary school children was confirmed. Parents of school-aged children had more positive opinions about available HL support. This does not imply that support for school-aged children is adequate but suggests that pre-school children are in more need.

The attitudinal trends that have emerged from our survey are encouraging, in the sense that they demonstrate genuine interest and participation in HL activities in Australia, with over 60% of parents of both pre-school and primary school-aged children saying that they would commit to attending such initiatives on a regular basis, were they available (Table 8). We believe a formal language extension support programme for early learning/pre-school (Jones Diaz, et al. 2022; Diskin-Holdaway and Escudero 2021; Escudero et al. In Preparation) would address these needs and demands for HL activities, particularly for pre-schoolers, offering a play-based HL experience within early childhood education. This should be a matter of urgency, considering the importance of early establishment of an HL policy for effective longevity (Adesope et al. 2010).

Acknowledgements

The authors wish to thank Dr. Alba Tuninetti and Dr. Criss Jones Diaz for insightful comments during the survey design, and Dr. Criss Jones Diaz for help promoting the survey via her network of community organisations. We also thank Ms Charlotte Casey for help revising and setting up the survey before final launch.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This research was supported by PE’s Australian Research Council (ARC) Future Fellowship (FT160100514) and by the ARC Centre of Excellence for the Dynamics of Language where PE was chief investigator, GPE was PhD student, and CDH and JH were affiliates. It was also supported by the Head of School Investment Fund (HOSIF) provided by the School of Languages and Linguistics at the University of Melbourne.

Notes

1 https://www.careforkids.com.au/child-care-articles/article/999/teaching-multiple-languages-to-under-fives

2 194 respondents were excluded on the following grounds: 88 did not have children, 42 did not reach the question about their children, 16 did not live in Australia and 48 did not answer our target questions on HL maintenance.

3 Percentages have been rounded to the nearest whole number.

Appendices

Appendix 1

A linear regression model was built to predict the success rate in transferring and maintaining home languages and raising children multilingually. Responses to ‘How would you rate your success in transferring and maintaining your home languages and raising your children multilingually?’ were coded numerically as 4 = very successful (n = 103), 3 = moderately successful (n = 79), 2 = somewhat successful (n = 74), 1 = not successful (n = 24), and N/A for a no-answer or left survey (n = 14). Attendance to HL initiatives during preschool (Q2) and during school years (Q3) were entered as predictors. It was found that having attended HL initiatives during pre-school years predicted a higher HL maintenance success rate F (2, 148) = 5.783, p = .004.

Table

Model Summary – rate_success
Model	R	R²	Adjusted R²	RMSE
H₀	0.269	0.072	0.060	1.005

Note. Null model includes Attended_pre-school, Attended_school.

Table

ANOVA	Sum of Squares		df		Mean Square		F		p
Regression	11.674		2		5.837		5.783		0.004
Residual	149.373		148		1.009
Total	161.046		150

Note. Null model includes Attended_pre-school, Attended_school.

Table

Coefficients
						95% CI
Model		Unstandardized	Standard Error	t	p	Lower	Upper
H₀	(Intercept)	2.627	0.119	22.044	< .001	2.392	2.863
	Attended_pre-school (2)	0.378	0.179	2.117	0.036	0.025	0.731
	Attended_school (2)	0.294	0.178	1.654	0.100	−0.057	0.646

^aStandardized coefficients can only be computed for continuous predictors.

Appendix 2

Wilcoxon rank sum test with continuity correction comparing pre-school and school groups in four statements.

Table

Statement	Measure 1 Pre-school M (SD)	Measure 2 School M (SD)	W	p
1. No awareness	2.35 (0.82)	2.22 (0.86)	6971	0.24
2. Too far away	1.89 (0.85)	1.87 (0.84)	6562	0.86
3. No time	1.88 (0.84)	1.86 (0.79)	6307	0.88
4. Not interested	1.39 (0.67)	1.42 (0.69)	6140	0.89

Appendix 3

A total of 242 participants who answered ‘yes’ or ‘maybe’ to Q5 were redirected to a further question, asking them to indicate what a manageable time commitment would be, in case HL community initiatives were available (see table). Participants could select as many options as applied to their situation (n = 389). The table shows comparable responses between weekends (33.2%) or two hours a week (31.1%), with 17% indicating twice a week and 13% indicating five days per week. Under ‘Other’, some parents wrote ‘fortnightly’ (n = 2), ‘monthly’ (n = 3), school holiday programme (n = 2), after school hours (n = 2), and ‘internet’ (n = 1), and three parents wrote that it would depend on the initiative or where it was held.

Amount of anticipated commitment to potential HL community initiatives in percentages, with counts in brackets; respondents could select more than one option.

Table

Q5a. If any new initiatives were made available, what would be a manageable commitment? Select all that apply.
On weekends	33.2% (129)
One day per week (2 h at a time)	31.1% (121)
Two days per week (2 h at a time)	17.2% (67)
Five days per week (e.g. long day care/ pre-school/ school)	13.1% (51)
Other	5.4% (21)
Total responses (out of 242 unique respondents)	100% (389)

The seven participants who answered ‘no’ to Q5 were redirected to an extra question, where they could select reasons from a multiple-choice list for why they would not be able to commit to HL community initiatives, were they to be made available. Two participants selected a reason related to their children’s time constraints (‘My children are too busy already’), while three selected a reason related to their own time constraints (‘I work full time’, or ‘I do not have time during the week/weekend’). The remaining two selected ‘Other’.

Appendix 4

Example of a respondent expressing disappointment in Australia’s lack of bilingual support in comparison with European countries:

‘Back in Europe we used to have parent-initiated masterclasses, festivals and camp activities with quests, games, performances. We had museum groups with guided excursions, we had guided day tours or sports coaches. Here only language tutoring is available on a very low level of everything. It looks like parents feel sort of guilty for any activity in their native language and do it secretly at home somehow’.