Varieties of Chinese as heritage languages: a research synthesis

ABSTRACT

Despite the range of varieties of Han Chinese, Mandarin is the most widely studied variety in research on Chinese as a heritage language (CHL) around the world. To better understand the role of other varieties of Han Chinese in addition to Mandarin, this article presents a synthesis of research on the learning of Chinese varieties in formal and informal contexts worldwide. We reviewed 25 empirical studies in CHL learning worldwide to examine 1) the definition of CHL learning, 2) varieties of Chinese as heritage languages, 3) theoretical frameworks, and 4) research trends in formal CHL learning programs, as well as informal CHL learning in families and communities. Mandarin was found to be the most studied variety, while varieties such as Cantonese, Hakka, and Hokkien are less frequently investigated. The findings highlight research gaps related to the research contexts, language varieties, and theoretical frameworks in CHL learning. This review points to the need for the exploration of CHL learners in non-English-speaking regions, wider representation of Chinese varieties in CHL research, and inclusion of a broader range of formal and informal learning contexts in the CHL research.

KEYWORDS:

• Chinese language varieties
• heritage language education
• language maintenance
• research synthesis

Introduction

As a widespread immigrant language, Chinese has garnered increasing attention in recent years. For instance, in the United States, following English and Spanish, Chinese (i.e. Mandarin and the other varieties of Han Chinese) is the third most spoken language at home for people from Mainland China, Hong Kong, Macau, Taiwan, as well as Chinese diasporas worldwide (U.S. Census Bureau 2013). Likewise, in Indonesia, Chinese linguistic varieties are the most commonly-spoken immigrant heritage languages (Zein 2020). As a result of this increase of Chinese as heritage language varieties around the world, research on the topic is increasing in turn. To this end, this article presents the findings of a research synthesis of empirical studies about the learning of varieties of Han Chinese as heritage languages around the world. Specifically, we examined the empirical studies in terms of their definitions of Chinese Heritage Language (CHL) learners, the target Chinese language varieties, the guiding theoretical frameworks, and the research trends in both the in- and out-of-school contexts.

The gaps in heritage language education research

Originating from Fishman's (1964) work, heritage language education has amassed a rapidly growing body of research in the field of language teaching and learning. Since most of the research is based on the North American context, many scholars have regarded heritage language as a minoritized language compared to English. For instance, Fishman (2001) defined heritage languages as ‘languages other than English’ that ‘have a particular family relevance to the learners’ (81). Moving beyond English-speaking contexts, Montrul (2010) defined speakers of heritage languages as those users of ‘ethnolinguistically minority languages who were exposed to the language in the family since childhood and as adults wish to learn, relearn, or improve their current level of linguistic proficiency in their family language’ (3). On the macro level, viewing heritage languages as national and societal resources echoes the growing recognition of minority rights and valorization of linguistic and cultural diversity (Campbell and Rosenthal 2000). On the micro level, the maintenance of heritage language skills is essential to individual identity construction and communication within families. Research has suggested that exposure to the dominant culture and language variety results in the loss of the use of a minoritized language within three generations (Fishman 1991) and sometimes within two generations (Wiley 2001).

Although a relatively large amount of research has been conducted in the field of heritage language education, two distinct gaps have been identified. First, as Montrul points out (2010), quite a bit of research in heritage language acquisition has focused on describing speakers’ linguistic profiles, whereas much less theoretically driven research has been conducted. Theoretically-driven studies that integrate language acquisition theories and generate implications for research and practices represent a gap that needs to be filled in CHL scholarship. Second, the primary contributions to the knowledge base for heritage language acquisition have come from scholars focusing on Spanish heritage speakers in North America (e.g. Potowski 2004; Montrul 2010; Otheguy and Zentella 2012), while other languages and language varieties, including Chinese, have received far less attention in the research. These two gaps in the CHL scholarship are the central motivation for this research synthesis.

Chinese as a heritage language

The high internal variation of Chinese makes it a special case among heritage languages. In a broad sense, in addition to Han Chinese, there are 55 ethnic minority groups living in China (The Central People’s Government of the People’s Republic of China n.d.). As of 2004, Chinese ethnic minorities speak 72 languages belonging to Sino-Tibetan and other four language families in the world (Ministry of Education of the People’s Republic of China 2004). In a narrow sense, Chinese refers to the language variety mainly spoken by Han, the world's largest ethnic group with approximately 1.4 billion people. In this article, we limit the definition of Chinese to the narrow sense by focusing on the Han Chinese language varieties.

Even though mainly spoken by one ethnic group, Han Chinese showcases high linguistic and regional diversity (DeFrancis 1984). To different Chinese speakers, the concept ‘Chinese as a heritage language’ has very different meanings depending upon their linguistic and cultural backgrounds. For these reasons, viewing Chinese as varieties of languages is a helpful departure point for this research synthesis. According to Kurpaska (2010), the most common classification of Han Chinese divides this umbrella language into seven main varieties: Mandarin, Wu, Yue (Cantonese), Min, Hakka, Xiang, Gan (ranked by the number of the native speakers from the highest to the lowest). Furthermore, high variation exists within each language subgroup. For instance, Teochew and Hokkien are both dialects of Min that differ in terms of vocabulary, grammar, pronunciation, and intonation to some extent. While there is a wealth of research on Mandarin as a heritage language – since it is the lingua franca of Chinese with higher social, cultural, and economical status – other Chinese language varieties have rarely been the focus of heritage language research (e.g. Leung 2012, 2014). Hence, a review of empirical research on varieties of Chinese as heritage languages will shed light on these frequently marginalized varieties of Chinese. To this end, this synthesis examined the research on both formal programs in schools and informal CHL learning through the following questions:

1. How does the research define heritage learners of Chinese?

2. What are the varieties of Chinese language have been studied in the existing research on learners of CHL?

3. What theoretical frameworks have been employed in the existing research on learners of CHL?

4. What are the trends in the research findings related to CHL learning in formal contexts (i.e. schools) and informal contexts (i.e. CHL learning in families and communities)?

Methodology

Data collection

Informed by Norris and Ortega (2006), an integrative approach was employed for this synthesis, ‘looking for patterns and clear moderating factors in the data’ (7). We began our search using the combination of the keywords ‘Chinese’ and ‘heritage language’ or ‘home language’ on three academic searching engines: Scopus, Web of Science, and Linguistics and Language Behavior Abstracts (LLBA). Second, the keyword ‘Chinese’ was employed for scanning the articles published in Heritage Language Journal as a platform providing a pool of rich publications in heritage language education. Third, after removing articles based on the inclusion and exclusion criteria, a backward reference search was conducted using the reference lists of the identified articles.

We included articles that were: (1) peer-reviewed full-text journal articles written in English; (2) primary studies on CHL learners, families, communities, and education programs; and (3) studies conducted all over the world. These criteria helped us capture the empirical research trends with a relatively large database. The exclusion criteria were designed for studies that: (1) involved heritage languages not limited to Chinese varieties; (2) focused on students in bilingual, second, or foreign language contexts because these programs often include predominantly non-heritage learners; (3) specialized in speakers’ linguistic profiles such as their linguistic knowledge of compound sentences (e.g. Zhang 2014). Following these criteria, 25 empirical journal articles published between 2005 and 2021 were included in the analysis.

Data analysis

NVivo 12 was central in the data analysis process. After importing all the selected journal articles into the software, codes were generated with hierarchies to log information about the studies and reference the research questions. NVivo coding hierarchies allowed for the systematic organization of data according to three levels of codes. The first-level codes served as umbrella categories and included: ‘bibliographic features’, ‘background information’, ‘focus’, ‘design’, and ‘results’. The codes in the second level allowed the specification of more detailed codes under a particular first-level code. For instance, the second-level codes under ‘bibliographic features’ were ‘author(s)’, ‘year’, and ‘journal’. The third level of coding included either fixed or open values. Fixed values were set for the codes ‘year’, ‘country’, ‘target language varieties’, ‘data collection instrument’, and ‘data analysis method’. For example, the code ‘data collection instrument’ involves six fixed values, ‘observation’, ‘interview’, ‘recording’, ‘informal conversation’, ‘artifact’, and ‘questionnaire’. Open values were designed for the codes ‘author(s)’, ‘journal’, ‘contexts’, ‘definition of Chinese heritage learners’, ‘focus’, ‘period’, ‘results’ and ‘implications’ since each study involved different information regarding these features. See the values and the organization of the codes in Appendix A.

Findings

Table 1 summarizes the focal details of the studies reviewed, in terms of author, year, journal, country, and the target Chinese language variety involved. The findings are organized according to the four research questions that guided our review. Among the 25 studies, 21 were conducted in the English-dominant contexts (16 in the United States, two in Australia, one in the United Kingdom, and one in Ontario, Canada), two were located in Malaysia where English, Mandarin, and Malay were frequently used in the formal contexts (Ding and Goh 2019; Wang 2017), one took place in Indonesia with Indonesian as the majority language (Budiyana 2017), and one was in Quebec, Canada where French and English were widely spoken (Curdt-Christiansen 2006).

Table 1. Background information of selected studies.

Author and Year	Country/Regions	Target Language Variety
Budiyana (2017)	Indonesia	Mandarin
Chen, Zhou, and Uchikoshi (2021)	The United States	Mandarin and Cantonese
Curdt-Christiansen (2006)	Quebec, Canada	Mandarin
Ding and Goh (2019)	Malaysia	Hakka
Du (2017)	Ontario, Canada	Mandarin
Liang and Shin (2021)	The United States	Mandarin
Li (2005)	The United States	Mandarin and Cantonese
W. Liao and Huang (2020)	Australia	Mandarin
L. Liao, Larke, and Hill-Jackson (2017)	The United States	Mandarin
Luo (2015)	The United States	Mandarin
Luo, Li, and Li (2019)	The United States	Mandarin
Mu (2014)	Australia	Mandarin
Mu and Dooley (2015)	Australia	Mandarin
Park et al. (2012)	The United States	Mandarin and Cantonese
Wang (2017)	Malaysia	Hakka, Mandarin, and Hokkien
Wei and Hua (2019)	The United Kingdom	Mandarin, Cantonese, and Hakka
Wu and Leung (2014)	The United States	Mandarin
Wu and Leung (2020)	The United States	Mandarin, Cantonese, Fujianese, Taishanese, Hakka, or a combination of these languages
Wu, Lee, and Leung (2013)	The United States	Mandarin
Yang (2007)	The United States	Mandarin
D. Zhang (2010)	The United States	Mandarin and Fujianese
D. Zhang (2012)	The United States	Mandarin and Fujianese
D. Zhang and Slaughter-Defoe (2009)	The United States	Mandarin and Fujianese
J. Zhang (2009)	The United States	Mandarin
Zhu, Hopper, and & Kulaixi (2020)	The United States	Mandarin

RQ1: Definitions of ‘Heritage Language’, ‘Heritage Language Learners’, and ‘Heritage Learners of Chinese’

Among the articles reviewed, six did not clearly operationalize CHL education and provided limited demographic information about the participants (Budiyana 2017; Ding and Goh 2019; Li 2005; Wu and Leung 2020; Yang 2007). By contrast, two studies used the term ‘heritage language maintenance’ (W. Liao and Huang 2020; Zhang 2010), while the remaining studies addressed the concepts in different ways.

Some of the studies in Canada, the United States, and Australia defined ‘heritage language’ as an immigrant language other than English or the dominant language. For instance, several studies (Wu and Leung 2014; Wu, Lee, and Leung 2013) in the United States used Hornberger and Wang’s (2008) definition of heritage language learners as ‘individuals with familial or ancestral ties to a language other than English who exert their agency in determining if they are heritage language learners of that language’ (6). Other scholars defined heritage language or heritage language learners in similar ways (Du 2017; L. Liao, Larke, and Hill-Jackson 2017; Mu 2014; D. Zhang 2012; Zhu, Hopper, and & Kulaixi 2020).

In other studies, a clear definition of Chinese heritage learners was not provided; rather, these researchers described their participant inclusion criteria based on the language speakers’ demographic, linguistic, or parental information. For example, Luo (2015) studied Chinese heritage learners ‘who have at least one parent whose native language is a variety of Chinese’ (22). In order to study the impact of linguistic background on Chinese learning anxiety, she further divided the participants into three groups: learners who do not have Chinese language background, learners who have a Mandarin background, and learners who have a non-Mandarin background. While Mu 2014; Mu and Dooley (2015) noted possibly having the largest Chinese–Australian sample – 119 ethnic-Chinese participants aged 18–35 who were born in Australia, Greater China, and other countries including Indonesia, Malaysia, New Zealand, Singapore, Vietnam, and lived in Australia as first-, second-, third-, or even further removed generation immigrants – they did not clearly define Chinese heritage learners. Other scholars (J. Zhang 2009; D. Zhang and Slaughter-Defoe 2009) adopted the concept ‘second-generation’ or ‘1.5-generation’ Chinese American in lieu of ‘Chinese heritage speakers’.

RQ2: Target Chinese language varieties

Minority Chinese varieties have rarely been the explicit focus in heritage language research. The one exception is the study by Ding and Goh (2019), who explored the impact of religions on two Hakka communities’ language usage in Malaysia. As shown in Table 1, Mandarin was the sole target language of study for 15 articles (Budiyana 2017; Curdt-Christiansen 2006; Du 2017; Liang and Shin 2021; W. Liao and Huang 2020; L. Liao, Larke, and Hill-Jackson 2017; Luo 2015; Luo, Li, and Li 2019; Mu 2014; Mu and Dooley 2015; Wu and Leung 2014; Wu, Lee, and Leung 2013; Yang 2007; J. Zhang 2009; Zhu, Hopper, and & Kulaixi 2020) and was studied along with other Chinese language varieties such as Hakka and Fujianese in nine articles (Chen, Zhou, and Uchikoshi 2021; M. Li 2005; Park et al. 2012; Wang 2017; Wei and Hua 2019; Wu and Leung 2020; D. Zhang 2010; D. Zhang 2012; D. Zhang and Slaughter-Defoe 2009).

Among the Mandarin-only studies, three categories emerged in the research on learners’ linguistic backgrounds. In the first category, five studies investigated the language maintenance of Mandarin heritage speakers (Liang and Shin 2021; W. Liao and Huang 2020; Yang 2007; J. Zhang 2009; Zhu, Hopper, and & Kulaixi 2020). For example, J. Zhang's research (2009) in the United States takes a closer look at the 1.5-generation and second-generation immigrant children from mainland China who speak Mandarin as their native language. In the second category, six studies treated the Chinese heritage speaker as a monolithic whole without exploring the maintenance and revitalization of their home language varieties (Budiyana 2017; Curdt-Christiansen 2006; Du 2017; L. Liao, Larke, and Hill-Jackson 2017; Mu 2014; Mu and Dooley 2015). For instance, Curdt-Christiansen (2006) documented students’ Mandarin use in a CHL school even though some of them were from Hong Kong where Cantonese is dominant. Similarly, Budiyana's research (2017) explored parents’ influences on children's Mandarin development without mentioning the high linguistic variation of Chinese Indonesian. As presented by Setijadi (2015), Hakka, Hokkien, Cantonese, Teochew, and Mandarin are the most spoken Chinese language varieties in Indonesia. In the third category, four studies examined minority students’ experiences of learning Mandarin as their assigned heritage language. In Luo's study (2015) for example, ethnic Chinese students without Chinese language background were found to be the most anxious about learning in the Mandarin heritage classroom, followed by those whose mother tongues are not Mandarin, whereas the Mandarin heritage speakers exhibit the least learning anxiety. Together this research revealed the struggles of minority students enrolled in Mandarin heritage language study, their silence in class, and the disconnect they experienced between Mandarin, the focus of study, and their home cultures and varieties of Chinese (Luo 2015; Luo, Li, and Li 2019; Wu and Leung 2014; Wu, Lee, and Leung 2013). The themes related to these three categories of findings suggest important priorities for future research that are addressed in the discussion.

RQ3: Theoretical frameworks of the empirical research

Two categories for theoretical frameworks emerged in the research: sociocultural and cognitive. Broadly speaking, sociocultural theories were predominantly employed by researchers in the field of CHL education.

Most of the studies included in this review focused on Chinese immigrant families and employed various theoretical frameworks. First, a few studies were guided by family language policy theories to explore ideologies, practices, and management of family language practices (W. Liao and Huang 2020; Liang and Shin 2021; Park et al. 2012; Wang 2017). Second, Mu and Dooley's study (2015) applied Bourdieu's notion of ‘habitus’ (1977) to examine the role of family support in CHL learning. Third, two studies were designed based on the theory of language maintenance and language shift to analyze the role of imagination in the language experiences across generations in transnational families (Wei and Hua 2019; D. Zhang 2010). Fourth, Chen et al.'s project (2021) implemented language socialization to assess how language socialization processes such as parental Chinese language use at home and children's exposure to Chinese media predict the children's Chinese proficiency. Finally, two studies were conducted based on language ideology and orientation (Wu, Lee, and Leung 2013; D. Zhang and Slaughter-Defoe 2009).

Research on community heritage language learners has been conducted from multiple perspectives, including cultural-ecological theory (Zhu, Hopper, and & Kulaixi 2020), language of religion (Ding and Goh 2019), social networks (D. Zhang 2012), ethnic identity development theory (L. Liao, Larke, and Hill-Jackson 2017), Bourdieu's sociological notion of capital (Mu 2014), and individual networks of linguistic contact including interpersonal, educational and media contact (Zhang 2009).

As for classroom-based research, language ecology perspective (Wu and Leung 2014), Vygotsky’s (1978) ‘zone of proximal development’ (Curdt-Christiansen 2006), Bakhtin’s (1981) ‘authoritative discourse and internally persuasive discourse’ (Curdt-Christiansen 2006), and multiliteracies framework (Du 2017) were applied. Wu and Leung (2020) investigated a Mandarin heritage classroom using translanguaging as the pedagogical lens. Translanguaging pedagogy encourages students to leverage their full linguistic repertoire for different social and cultural contexts (García and Sylvan 2011; as cited by Wu and Leung 2020). Using this pedagogy, the Mandarin teacher has opened a floor for students who speak minority Chinese varieties like Hokkien, Cantonese, and Hakka to share their cultural and linguistic knowledge and thus created a friendly learning climate. Other studies (Budiyana 2017; Li 2005; Yang 2007), while sociocultural in nature (e. g., strong focus on ideologies, use of in-depth observation), have not provided a clear statement of a particular sociocultural theory.

By contrast, research informed by cognitive theories is scant. Luo's report (2015) on learning anxiety is the only work based on a cognitive perspective. In addition, Luo et al.'s survey of college-level Chinese language programs in the United States does not articulate a clear theoretical framework (2019).

RQ4: Research trends in formal and informal contexts

The studies were divided into two groups: formal CHL programs and informal CHL learning contexts in communities or households. Studies were categorized in this way because these contexts influence CHL development and maintenance in different ways. Among the 24¹ empirical studies reviewed, 15 were conducted in informal contexts. Of the nine studies of formal contexts, eight focused on student learning and experiences in CHL education programs whereas one explored the role of parents in CHL programs (Li 2005).

Formal context

Challenges faced by minoritized CHL learners. Studies depicting challenges faced by CHL learners primarily involved non-Mandarin speakers. For instance, the 171 CHL learners in Luo’s (2015)² survey research in two large universities in the United States revealed differences in anxiety according to students’ experiences with varieties of Chinese at home. In general, students who spoke no varieties of Chinese at home reported the highest anxiety, followed by those who spoke non-Mandarin varieties. Finally, students who spoke Mandarin at home reported the least anxiety. Similarly, Wu and colleagues (Wu and Leung 2014, 2020; Wu, Lee, and Leung 2013) highlighted identity tensions among linguistic minority students enrolled in middle school Mandarin heritage language programs. To overcome these tensions and facilitate the learning of speakers of non-Mandarin varieties, Wu and Leung (2020) described the potential of ‘translanguaging as a pedagogical stance’ (2) to nurture classrooms that are inclusive of the linguistic experiences of these students and strengthen students’ metalinguistic awareness, and challenge the social hierarchy of the Chinese linguistic varieties.

Classroom discourse and pedagogies. Curdt-Christiansen’s (2006) study revealed the authoritative classroom discourse in beginner-level CHL classes in Canada. When the teacher controlled the class through traditional question-answer interactions and memorization tasks such as dictation quizzes and recitation of classical texts, these authoritative discourses seemed less engaging for some students due to the cultural differences between China and Canada. Similar experiences were described in the interview research of L. Liao et al.'s report (2017). The students in this study reported that memorization and recitation of characters and vocabulary were challenging for them. To move beyond these pedagogies and classroom discourses, other research has focused on the role of innovative pedagogies and teachers’ professional development needs. Du (2017) examined how teachers in beginner-level classes in a community language school used gestures and sounds, images and drawing, music, craft-making, and nursery rhymes to teach Chinese in a responsive way. In Luo et al.'s interview research (2019), college CHL instructors stressed the need for classroom activities that were relevant to students and integrated meaningful language use, rather than the tasks provided in their textbooks that were not engaging (e.g. language drills, fill-in-the-blank activities, etc.) and did not connect to students’ lived experiences. Likewise, Li's study (2005) of parental engagement in CHL programs found that parent-teachers employed traditional language teaching strategies like repetitive drills and rote memorization and traditional textbooks, which were not effective in CHL classrooms. To realize innovative practice in CHL will require professional development focused on how to differentiate instruction for students with varying backgrounds, home languages, and Mandarin proficiency within one classroom, as Luo et al.'s study (2019) noted.

Informal contexts

Parental perceptions towards CHL development. In informal contexts, parental perceptions toward children's CHL development were most frequently investigated. In general, most Chinese parents showed positive attitudes toward their children's CHL maintenance (Budiyana 2017; Chen, Zhou, and Uchikoshi 2021; Liang and Shin 2021; W. Liao and Huang 2020; Park et al. 2012; Yang 2007; Zhu, Hopper, and & Kulaixi 2020). According to Yang (2007), parental perceptions about children's CHL development varied among parents of different income levels. The findings revealed that parents who earned less held more positive attitudes toward their children's CHL maintenance than those with higher incomes. In minority Chinese families, parents held differing views of dominant and minoritized varieties. (Wang 2017; D. Zhang and Slaughter-Defoe 2009; D. Zhang 2010, 2012). For instance, in Penang, Malaysia, Wang (2017) found that Hakka-speaking parents placed emphasis on children's development of Mandarin, regarding it as the most important Chinese language variety. Hakka and the dominant local Chinese language variety, Hokkien, however, were not used at home to a great extent. While this research depicted Hakka language revitalization efforts among some of the families, they were not always successful. For example, in one family, parents insisted on the importance of speaking Mandarin and refused Hakka-speaking grandparent's suggestion of using Hakka with their children. At the same time, different language varieties were used for conversations among different generations. The parents used Mandarin with their children, Hakka with the grandparents, and Hokkien with each other, but the children spoke with their grandparents in both Mandarin and Hakka. While these multilingual inputs have equipped these children with passive capacities in several language varieties, Wang's research found that parents believed that Mandarin was the important language variety for Chinese Malaysians.

Three studies in the United States echoed the findings of Wang (2017). This research (D. Zhang and Slaughter-Defoe 2009; D. Zhang 2010, 2012) involved two groups of Chinese immigrant families in the United States: Mandarin-speaking families with higher social status and incomes and Fujianese-speaking families with lower social status and incomes. While the Mandarin-speaking parents recognized the importance of speaking Mandarin and provided direct support for their children learning Mandarin, the Fujianese-speaking parents regarded Mandarin and English as ladders to social advancement and thus did not emphasize the importance of maintaining Fujianese. As a result, children from Mandarin-speaking families maintained a higher level of heritage language than their counterparts from Fujianese-speaking families. Finally, Zhu et al.'s study (2020) depicted inconsistencies between parents’ positive attitude and their actual support of Chinese language maintenance. The parents prioritized the children's schoolwork over their learning Chinese. On the contrary, the Chinese grandparents showed consistency in their positive attitude and efforts regarding the children's Chinese language development, perhaps due to their limited English proficiency and quantity of time spent with children (Zhu, Hopper, and & Kulaixi 2020).

Focusing on imagination as a key factor in language maintenance and shift, Wei and Hua's research in the United Kingdom (2019) investigated the individual and collective imaginations toward the future within some ethnic Chinese families over time. The study depicted the dynamic roles of imagination during the process of language maintenance and shift and the tension between imagination at individual and collective levels. According to the researchers (Wei and Hua 2019), the same type of imagination could result in different attitudes and activities regarding family language use. For instance, families which planned to leave for China in the near future may exhibit high expectations of the children's Chinese language competence or made the best use of their time to develop English proficiency.

Family language practices and management. Chinese parents utilized various explicit and implicit family language practices, including encouragement of children to learn and use Chinese (Mu and Dooley 2015; Park et al. 2012); informal instruction of Chinese at home (Mu and Dooley 2015), maximizing Chinese use with children (Liang and Shin 2021; W. Liao and Huang 2020; Park et al. 2012), employing multimedia materials (Chen, Zhou, and Uchikoshi 2021; W. Liao and Huang 2020; J. Zhang 2009), and sending children to the CHL programs (Chen, Zhou, and Uchikoshi 2021; Liang and Shin 2021; Mu and Dooley 2015; J. Zhang 2009; D. Zhang and Slaughter-Defoe 2009; D. Zhang 2012). Compared with the consistent and direct support of learning CHL from Chinese parents, W. Liao and Huang (2020) found that the non-Chinese parents in cross-cultural marriages provided occasional and indirect support due to their limited Chinese proficiency and less positive perceptions toward CHL development. Furthermore, in Yang's study (2007), parents of children identified as gifted provided significantly less Chinese input than parents of other children, perhaps because of the high expectations toward children's academic achievements held by parents of gifted children.

Research has also documented the ways that family support contributed to children's CHL development. Chen et al.'s two-year study (2021) reported that more parental Chinese language use at home contributed to their children's receptive and expressive vocabulary in Chinese. Similar findings were reported by Park et al. (2012); however, they also found a transactional relationship between parental language use and children's language proficiency. That is, some parents may not use Chinese after noticing weaknesses in their children's Chinese, leading to attrition. Other research highlighted the positive role of language activities and support provided by extended family members such as grandparents (Wang 2017; D. Zhang and Slaughter-Defoe 2009; D. Zhang 2010, 2012; Zhu, Hopper, and & Kulaixi 2020). Finally, ethnic media was another support provided by families, which played a positive role in Chinese literacy in one study (J. Zhang 2009), but was not predictive of CHL proficiency in Chen, Zhou, and Uchikoshi (2021).

In some instances, research described challenges faced by Chinese families in maintaining their heritage linguistic varieties (Budiyana 2017; Liang and Shin 2021; W. Liao and Huang 2020). For instance, parents in Budiyana’s (2017) study noted that the lack of Chinese language exposure in their community in Indonesia limited their children's opportunities to learn Chinese. These findings stand in contrast to Ding and Goh's project (2019) described the rich use of Hakka in religious practices in two Hakka Communities in Malaysia. Other research by Liang and Shin’s (2021) outlined multiple challenges reported by parents, such as inadequate textbooks for CHL, children's limited time and energy due to their busy schedules, and the difficulty in maintaining children's motivation for learning Chinese. Further, W. Liao and Huang reported children's perceptions of learning Chinese as difficult and uninteresting.

Discussion and conclusion

This research synthesis on Chinese as a heritage language (CHL) revealed four key themes related to the research use of varieties of Chinese for in-school and out-of-school contexts worldwide. First, as presented in Table 1, the imbalanced geographical distribution of these studies suggests that more studies in the non-English-dominant regions are warranted. Given the growing and increasingly diversifying Chinese diasporas around the world (e.g. Guo 2022), it is clear that the voices of Chinese heritage speakers worldwide are in need of representation in future research. Second, most of the scholarship has focused on Mandarin heritage language education, whereas only a small portion of minority Chinese varieties are investigated such as Cantonese, Fujianese, and Hakka. This fact is not surprising due to the political and cultural status of Mandarin, as well as its speakers’ large population worldwide. To be more inclusive and representative of the diversity of varieties of Chinese, future research should investigate the learning of a wider variety of Chinese. These efforts are increasingly important for speakers of minoritized varieties of Han Chinese, since heritage language programs function as a critical component of language maintenance for them (e.g. Fishman 1991; Potowski 2013).

Third, while sociocultural theories are widely applied in this research, cognitive perspectives remain seldom employed. As scholars have noted (e.g. Newell et al. 2011; Toth and Davin 2016), cognitive and sociocultural perspectives can be positioned as complementary, rather than contradictory to one other. Sociocultural researchers focus on social interaction and events, revealing how language use in different contexts influences language development (Lantolf and Poehner 2008). Scholars in the cognitive field regard language acquisition as an internalized, cognitive process, and have provided us with deep insights into language learners’ cognitive skills, strategies, knowledge, and schemas (Zuengler and Miller 2006). Beyond the general need for cognitive studies in CHL learning, a thoughtful combination of multiple perspectives, that is, a sociocognitive approach (e.g. Toth and Davin 2016) or perhaps a dynamic systems approach (de Bot, Lowie, and Verspoor 2007) may delineate a more comprehensive understanding of learning CHL research and practices.

Finally, this review has shown that studies in formal Chinese education programs are far fewer than those in informal contexts. In terms of the research in formal contexts, we identified three areas in need of further research. First, tensions in language learning and identity construction of linguistic minority students enrolled in Mandarin education have been identified in several studies (Luo 2015; Wu and Leung 2014, 2020; Wu, Lee, and Leung 2013). More research focusing on the issues related to minority students in Mandarin language classrooms could serve to better understand and overcome these tensions. Second, through observations and interviews, some researchers found that traditional authoritative discourses and teaching strategies such as rote memorization and drill tasks were widespread in CHL classrooms and were not engaging for all students (Curdt-Christiansen 2006; L. Liao, Larke, and Hill-Jackson 2017; Du 2017; Luo, Li, and Li 2019). Cross-cultural perspectives and innovative pedagogies are worthy of further exploration in formal CHL programs. Third, more studies could be conducted on the professional development of CHL educators to help them meet the needs of Chinese minority students (e.g. Luo, Li, and Li 2019). Because formal learning programs are one of the most effective ways to teach a heritage language, especially when the efforts from minority families and communities are limited (Shibata 2010), more research is needed on formal CHL learning programs.

The majority of the studies in informal contexts were conducted in households. Families in a few studies were in need of community support and resources for learning Chinese beyond households (e.g. Budiyana 2017; Liang and Shin 2021). Other recent research in refugee communities has shown that community support, such as workshops on bilingual upbringing, can enable minority families to make informed and appropriate decisions about family language policy (Eisenchlas and Schalley 2017). More research is needed to understand the role of Chinese immigrant communities in children's language maintenance, as well as the orchestration among families, communities, and schools. While parents in the research had overall positive parental attitudes toward maintaining Chinese (especially Mandarin) at home, some discrepancies among parents related to income levels (Yang 2007) and minority Chinese linguistic varieties (Wang 2017; D. Zhang and Slaughter-Defoe 2009; D. Zhang 2010, 2012). In addition, inconsistencies between parental perceptions and their actual language practices and management were identified (Yang 2007). To better understand these attitudes and practices, more empirical research on Chinese immigrant families could be conducted to explore these inconsistencies, reflect the diversity in family language policies of Chinese immigrants, and provide deeper insight into their living conditions.

Disclosure statement

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

Additional information

Notes on contributors

Yizhe Jiang

Yizhe Jiang is a PhD candidate of multilingual language education at The Ohio State University. Her research focuses on multilingual language education, heritage language maintenance, translanguaging, assistive technology for language education, and learning Chinese as a foreign language.

Francis John Troyan

Francis John Troyan is associate professor of multilingual language education at The Ohio State University. His research focuses on the ideologies, beliefs, and languaging practices of pre-service and in-service teachers and students in culturally and linguistically diverse language classrooms in the United States and France.

Notes

1 Mu's survey research (2014) was not subject to this analysis because it focused on the relationships between cultural, social, symbolic, and economic capital in CHL learning.

2 While this research did not describe the types of Chinese language programs where the heritage learners enrolled (heritage or traditional ‘non-heritage’), Luo’s (2015) project is one of only two empirical studies of CHL learners at the tertiary level published in academic journals. Because of its large sample size, it was included in the analysis for this question.

Appendix A

Organization of codes

Table

Code level 1	Code level 2		Code level 3
Bibliographic features	Author(s)		Open
	Journal		Open
	Year		2005–2021
Background information	Country	The United States	The United Kingdom	Canada	Australia	Indonesia	Malaysia
	Target language Varieties	Mandarin		Hakka		Mixed varieties
	Contexts	Formal Chinese language program				Informal Chinese learning contexts
						Community	Family
Focus	Open
Design	Period	Open
	Definition related to CHL learners	Open
	Theoretical framework	Sociocultural		Cognitive	Combination of sociocultural and cognitive
	Period	Open
	Data collection instrument	Observation	Interview	Recording	Informal conversation	Artifact	Questionnaire
	Data analysis	Qualitative		Quantitative		Mixed
Results	Open
Implications	Open