Micro language planning in Mandarin-dominated Chinese language education: voices from dialect-background heritage learners in New Zealand

ABSTRACT The field of language policy and planning has seen increasing scholarly attention that explores social actors’ micro language planning towards a given language policy situation at the grassroots level. In educational contexts, teachers and educators are often considered pivotal locals whose agency, when enacted in micro language planning, can contribute to educational language policymaking. However, little attention has been given to learners – those on the receiving end of educational language policy. Situated in the current discourse concerning Mandarin-dominated Chinese language education, within the context of China’s promotion of Mandarin language learning, this study examines dialect-background heritage learners in a tertiary-level Mandarin Chinese language programme in New Zealand. By adopting the concept of micro language planning, this study explores how dialect-background heritage learners perceive and react in a Mandarin-dominated Chinese programme. The findings show that dialect-background learners exercise agency to interpret, implement, reject, and negotiate language policy according to their self-identified language needs. However, their ability to enact agency is also constrained by a series of social realities including socio-economic discourse, language ideology, language proficiency, and technology development. This study offers implications for language policy makers involved with different levels of Chinese language policy.


Introduction
In recent decades, language policy and planning (LPP) research has undergone a significant transformation.It has evolved from its initial focus, which centred on the authority of policy declarations and texts, often referred to as top-down LPP, towards adopting a more inclusive bottom-up perspective, viewing LPP as a dynamic social practice (Levinson & Sutton, 2001;McCarty, 2011).This theoretical advancement has been accompanied by a growing emphasis on social actors and their agency in shaping LPP decisions (Liddicoat & Taylor-Leech, 2021).For instance, Baldauf (2006) asserted that social actors actively exert their influence through micro-level language planning within various facets of language policy, including educational language policy.Notably, 'educational policies remain one of the most powerful mechanisms for implementing, protecting, and maintaining official language policy rules' (Badwan, 2021, p. 99).Consequently, numerous studies have undertaken an exploration of the bottom-up implementation, negotiation, and resistance processes related to language policies within educational contexts.It is worth mentioning that most of these studies have placed teachers and educators at the forefront, acknowledging their central role in shaping educational language policy (Hornberger & Johnson, 2007;Ramanathan, 2005;Ricento & Hornberger, 1996;Stritikus, 2002).However, while teachers' and educators' agency has been well-explored within LPP research, the responses of learners to the given language policy situation have been comparatively less documented.Cheng and Li (2021) aptly highlighted that 'as receivers of language education and final evaluators of LPP, learners, as key actors in LPP, deserve more attention ' (p. 130).
In language education in mainland China, the 'Promotion of Putonghua' policy (hereinafter referred to as 'Mandarin promotion') plays a pivotal role in its historical LPP development.'Chinese', in reference to a language, is an umbrella term covers a large number of language varieties.The standardised national language of China is 'Putonghua', often translated as 'Mandarin' in English (Zhu & Li, 2014).Mandarin is 'associated with the speech of Beijing and its surrounding regions, which for centuries have enjoyed political and cultural significance' (He, 2006, p. 3).Non-Mandarin Chinese varieties (fangyans) are often referred to as 'dialects' in literature regarding Chinese language (Cai & Eisenstein Ebsworth, 2018).Despite sharing the same writing system, many of the Chinese dialects (such as Cantonese and Teochew) are mutually unintelligible to Mandarin speakers, the linguistic distance between them is similar to English and Dutch (Chao, 1943;Erbaugh, 1995;Wong & Xiao, 2010).Established in the 1950s and implemented to this day, China's Mandarin language promotion aims to break language barriers among dialect-speaking communities and reinforce national unity.
In China, Mandarin is adopted as the language medium in schools, which is one of the approaches used to implement Mandarin promotion in the education system.This has increased the usage of Mandarin nation-wide.The number of people who can speak Mandarin in China has reached 79% in 2019 (State Language Commission, 2019) and is expected to grow to 85% in 2025 (State Council of PRC, 2020).Duff (2021) argued that the more linguistically homogeneous Chinese population tends to produce advantages and opportunities for upward social mobility, though it may also cause a series of issues, such as cultural and traditional heritage loss in regional minorities and social disparities.
While extensive research has delved into the Mandarin promotion policy within mainland China, there has been relatively limited discussion regarding its effects beyond these borders.One prominent consequence of this policy is the substantial emphasis placed on Mandarin Chinese within language teaching contexts, as evidenced in studies such as Leung and Wu (2011).The growing politico-economic influence of China has led to a substantial influx of learners into the rapidly expanding Mandarin Chinese language programmes, including many heritage learners from various Chinese diasporas (Ding, 2013).Within the group of heritage learners, dialect-background learners, referring to students with a cultural or familial connection to a specific Chinese variety that is largely or entirely unintelligible to Mandarin, often constitute more than 50% of the student body (Chen & Wang, 2023a).
Dialect-background learners represent a distinct subset of students in Mandarin Chinese programmes, with their home language differing significantly from Mandarin (Wong & Xiao, 2010).Due to their unique background, Wu et al. (2014) argued that considering Mandarin as the true heritage language for dialect-background learners may lead to a 'mismatch ' (p. 19) of their learning needs.Despite being a significant student cohort in overseas Chinese language education, limited research has explored the experiences of these dialect-background learners and their interactions within the context of this mismatched discourse.To this end, this study seeks to bridge this gap by investigating how dialect-background learners perceive and respond to Mandarin-dominated Chinese language programmes through the lens of micro language planning.

Micro language planning and agency
From an ecological perspective, LPP is believed to be a multi-layer construction (Baldauf, 2006;Kaplan & Baldauf, 1997;Ricento & Hornberger, 1996).Ricento and Hornberger (1996) broke down the multi-layer construction of LPP to the national level (e.g., official government policies), the institutional level (e.g., policies in schools or at home), and the interpersonal level (e.g., individual interaction).In addition, they foregrounded the agency of individuals and local communities in implementing or resisting policy initiatives.To emphasise the power of individuals and communities in the LPP process is to interpret LPP as a social practice from a bottom-up sociocultural perspective (Levinson & Sutton, 2001;McCarty, 2011).Different from the top-down tradition that linearly divides language policy formation and implementation and considers that the latter happens passively after the former is made, seeing language policy as a social practice is to investigate 'how people make, interpret, and otherwise engage with the policy process' (Levinson & Sutton, 2001, p. 4).
The attention to bottom-up LPP has led to the consideration of micro language planning and agency (Baldauf, 2006;Canagarajah, 2005;Liddicoat & Taylor-Leech, 2021).Micro language planning reflects the voices from the local policy agents.Canagarajah (2005) demonstrated how local policy actors, such as teachers and learners could inform or reclaim language policy by exercising their agency to meet their own needs.Liddicoat and Taylor-Leech (2021) interpreted the agency of social actors in the LPP context as the power for diverse actors to respond to language policy with free will instead of implementing it unconditionally.However, Liddicoat and Taylor-Leech (2021) also reasserted the necessity of examining the balance between agency and structure, and emphasised the free will of agency is within the constraints of external factors.This echoed Ahearn's (2001) interpretation of agency as 'the socioculturally mediated capacity to act ' (p. 112).
Given agency is placed at the heart of micro language planning, a number of authors have discussed the forms of agency in LPP.Focusing on multilingual education, Liddicoat and Taylor-Leech (2014) described four contexts in which local agency may be enacted: local practices such as curriculum designing or teacher training are conducted to implement multilingual language policy; local resistance such as grassroots counter-hegemonic action are taken to contest the macro policy; local needs are addressed when they are neglected by the macro language policy; and local behaviours to create multilingualism possibilities for themselves when it is not offered at the macro-level.Hornberger and Johnson (2007) noted 'varying local interpretations, implementations, negotiations, and perhaps resistance can pry open implementational and ideological spaces for multilingual language education' (p.511).Therefore, based on the multiple forms of agency, examining locals' micro language planning is to explore 'the agency of those that have traditionally been positioned merely as policy implementers and repositions them as active policy interpreters, appropriators, and creators' (Johnson, 2013, p. 2) In the educational level, there has been increasing attention towards re-examining teachers and educators as pivotal locals that can contribute to the shaping of educational language policy through micro language planning (Hornberger & Johnson, 2007;Ramanathan, 2005;Ricento & Hornberger, 1996;Stritikus, 2002).For example, Stritikus (2002) revealed that teachers sometimes implement Proposition 227 (an English-only oriented proposition in California, United States) with negotiation to meet the actual needs of class.However, compared to the attention given to teachers and educators, learners' agentive action towards educational language policy are rarely researched.Some studies (for example, Brown, 2015;De Costa, 2010) have called attention to the importance of examining learners as agentive locals, as learners are the receiving end of language policies and can be 'insightful and responsible education policy makers in their own learning situations' (Phyak & Bui, 2014, p. 113).As a result, a handful of studies have begun to investigate learners and their micro language planning.
Through exploring students' language attitudes, Vennela and Kandharaja (2021) found that students' feelings towards the use of English in India is various and agentive, and they highlight the importance of considering social contexts and cultural realities when examining agency.Payne and Almansour (2014) revealed that social actors negotiate with the given language policy while implementing it by showing how students create multilingual learning opportunities for themselves under the English-dominated foreign language policy in Saudi Arabia.Willoughby (2014) investigated students' decision to not enol in a heritage language programme at a high school in Australia, which demonstrated that local agents can reject a language policy even if it is being embraced in the meso-and macro-level.Her work also raised concerns that 'local agency does not always have a positive effect on multilingual education and the actions of some local agents can constrain possibilities' (Liddicoat & Taylor-Leech, 2014, p. 239).
In the context of Chinese educational language policy, there is research discussing learners' micro language planning at the education discourse in China, such as Han et al. (2019) which examined ethnic minority students and their agency enaction towards Mandarin education.However, there is still a dearth of research which explores overseas learners' micro language planning in the context of Chinese educational language policy, even though Chinese learning has gained increasing attention outside of China due to China's rising politico-economic power.

China's Mandarin promotion and responses from below
The promotion of standardised Chinese began in the 1950s.Putonghua or Mandarin Chinese 'was officially adopted as the common speech for the PRC at the beginning of 1956' (Guo, 2004, p. 45).In that year, the State Council of the People's Republic of China (PRC) launched 'The Instructions Concerning the Promotion of Putonghua' (Li, 2013) with the aim of dismantling language barriers among dialect speakers and advancing the construction of a new China.In 2000, the adoption of 'The Law of the People's Republic of China on the Standard Spoken and Written Chinese Language' (MoE, 2000) further solidified Mandarin's role as the common language of China and promoted its use among the Chinese population.
Under the promotion of Mandarin, in mainland China, dialects are often banned in practice at school by the local government (Lvqiu, 2018).In the home sphere, diglossia families tend to mix their use of Mandarin and their dialect(s) (Sun, 2017;Wang & Curdt-Christiansen, 2019).After half a century of effort in promoting Mandarin, dialects in China show decline in both linguistic and communication aspects (Li, 2017).Facing the decline of dialects, bottom-up dialect preservation campaigns emerged from social actors of different LPP levels.For example, Guangzhou netizens initiated the online 'Protecting Cantonese Movement' in 2010, in which Cantonese speakers emphasised the status of Cantonese 'out of anger and fear of Cantonese losing its status as a regional lingua franca' (X.Gao, 2012, p. 450).To negotiate with Mandarin as the medium of instruction in educational language policy, the Shanghai government began the 'Shanghainese Heritage Project' in 2013 to encourage the use of Shanghainese in kindergartens (Shen, 2016).Furthermore, there has been a call from academics within China who propose a coexisting relationship between Mandarin and dialects and an extension of space in which dialects could be practised (for example, Cao, 2017;You, 2006;Zhang, 2018).
Outside of China, the Chinese language ecology has also experienced the impact of Mandarin promotion.Mandarin has not only become the de facto language in China's media industry, gaining popularity abroad through films and TV shows, but it is also promoted as THE 'Chinese language' in overseas Chinese language programmes (MoE, 2000).The number of Mandarin learners outside of China has rapidly increased in tandem with China's expanding global economic influence.For example, in countries that are included in China's 'Belt and Road' Initiative, Mandarin has been recognised as a crucial trade language, drawing a substantial number of learners (Y.Gao, 2020).
The language ecology of Chinese communities outside of China has also undergone changes within such macro language policies.For instance, Singapore launched the 'Speak Mandarin Campaign' in 1979, discouraging the use of dialects within the Chinese population.This policy effectively led to the transition of the younger Chinese generation in Singapore into Mandarin speakers (Chua, 2010).Beyond Asia, early Chinese migrants in Anglophone countries are also traditionally speakers of dialects, particularly southern dialects such as Cantonese (Li & Zhu, 2010).However, due to the increasing numbers of Mandarin-speaking new migrants coming from Mainland China in recent decades, Mandarin has become 'the new diasporic standard' (Huang, 2021, p. 165).As a result, Chinese heritage language programmes began to be dominated by Mandarin, similar to how it is in programmes teaching Mandarin as a foreign language.Some studies have examined local Chinese communities' responses towards the rise of Mandarin.The majority of the Chinese diasporas seem to accept the prestigious status of Mandarin and acknowledge the necessity of using one unified Chinese variety, that is, Mandarin, to represent Chinese heritage language (Li & Zhu, 2010;Wiley et al., 2008).However, in the educational level, teachers of Chinese heritage language programmes may complain about the decline of institutional resources given to dialect teaching (Zhu & Li, 2014).Sometimes, it is the student body who have questioned the Mandarin-dominated language policy in Chinese language programme.For example, at a kindergarten-to-eighth-grade school in the United States, Leung and Wu's research explored the attitudes of heritage learners in its Chinese programme in which over 80% of students were dialect speakers (Leung & Wu, 2011;Wu et al., 2014).They discovered that learners express a strong desire to learn their home dialects instead of Mandarin, because they want to connect with dialect-speaking peers and reinforce family bonds.
Currently, studies such as Leung and Wu's (2011) that pay attention to dialect-background learners' perception about the shifted variety emphasis in Chinese language education are relatively few in number, even if this group of learners is most likely to be influenced by such changing discourse.Regardless of the large proportion of dialectspeaking heritage learners within the general heritage learner population, these learners were rarely considered separately but rather examined only together with Mandarinbackground learners in overseas Chinese (referring to Mandarin) education (Wong & Xiao, 2010).Few studies have discussed them as focal group by clarifying their different linguistic and sociocultural backgrounds from the Mandarin-background heritage learners.Some studies argued whether it is appropriate to define them as heritage learners of Mandarin (for example, Wiley, 2008;Wu et al., 2014).Some studies pointed out the placement dilemma regarding whether posit them in heritage track or track for foreign learners (for example, Kelleher, 2008); and others have highlighted their learning difficulties and identity issues compared to Mandarin-background learners (Chen & Wang, 2023b;Wong & Xiao, 2010).However, there is a lack of research examining dialect-background learners through the lens of LPP.Given their large numbers and their important position in building educational equality in Chinese language education, we argue that it is worth investigating how dialect-background learners respond to Mandarin-dominated educational language policies by the means of bottom-up micro language planning.

The New Zealand context
This study focuses on a university-level Mandarin Chinese language programme in New Zealand, a country that has a history of Chinese settlement dating back to the 1860s (Ip, 2003).Similar to Chinese communities in many other countries, Chinese immigrants in New Zealand typically spoke Cantonese and hailed from various rural towns outside Guangzhou during the New Zealand gold rush era.As extensively documented by Ip (2003), these Chinese migrants faced legalised discrimination through measures like the Poll Tax, which was imposed on them in 1881, and a series of discriminatory laws and restrictions that not only limited their entry into New Zealand but also affected their lives within the country.While the Chinese community has made efforts to preserve its linguistic and cultural heritage since the 1950s, the dominance of assimilation policies has exerted significant pressure on their language choice, resulting in the forced transition of many Chinese heritage children into becoming monolingual English speakers (Wang, 2023).
However, the local Chinese language landscape began to change after the implementation of the 1987 Business Immigration Policy and the 1991 point system for immigration assessment.Both policies aimed to attract skilled immigrants and investments, which brought about a vast number of Mandarin speakers from Taiwan, Singapore, and more recently from mainland China.By tHe, 2006 census, Mandarin speakers first outnumbered Cantonese speakers, and the number of Mandarin speakers overall almost doubled between 2001 and 2013 (Stats NZ, 2013).Today, Chinese migrants are the third largest ethnic community in New Zealand after New Zealand Europeans and Māori (Stats NZ, 2020).Based on the rapid increase of Mandarin speakers, Wang (2021) argued that the 'Chinese community is divided by particular Chinese varieties they speak at home, even though Mandarin is often used as the lingua franca in the wider Chinese community' (p.172).
The shifting hierarchy in Chinese language ecology and China's emphasis on Mandarin learning have extended their influence into educational language policy in New Zealand's Chinese education.Most Chinese community schools have changed their focus from teaching Cantonese (before 1987) to teaching Mandarin (Ho & Wang, 2016;Wang, 2021;Yao, 2017).With the strengthening political-economic relations between China and New Zealand (by 2017, China had become New Zealand's largest trade partner) and the increasing demand for Chinese language acquisition, most universities have offered Chinese programmes which teach Mandarin.These tertiarylevel Chinese programmes cater to not only learners who learn Mandarin as a foreign language but also heritage learners who are members of the local Chinese communities.Among these heritage learners, dialect speakers occupy a vital part.For example, the University of Auckland (2020) reported the enrolment of heritage students in its beginner Chinese class was 22.5% in 2020, and more than half of those are dialect speakers.
Despite the significant Chinese population and the high demand for Chinese language acquisition, there has been relatively little international attention to the status and overall development of Chinese language education in New Zealand as a whole, especially when compared to the extensive literature on the subject in other Anglophone countries (Wang & East, 2023).There has been limited research dedicated specifically on Chinese heritage learners and their changing identity and perspectives towards learning Chinese to connect with their roots.At present, no study has been undertaken with a specific focus on dialect-background learners within Mandarin Chinese programmes.This study will fill this gap.

The study
This study is a part of a larger research project examining the Mandarin learning experience ofdialect-background CHL learners at a university-level Chinese language programme in New Zealand.The project has been approved by the University of Auckland Human Participants Ethics Committee under reference number 023842.While the main project includes multiple sources of data including questionnaires, interviews, and written narratives for the purpose of triangulation, the present article focusses on data elicited from interviews.In order to explore the interactive relations between the dialect-background heritage learners and the Mandarin-dominated educational language policy in Chinese education, the research questions guiding this study are as follows: (1) How do dialect-background heritage learners respond to the Mandarin-dominated educational language policy in Chinese education through the lens of micro language planning?(2) What are the constraints of external factors that influence their responses to the given language policy situation?

Setting and participants
This study was conducted at a large university located in Auckland, New Zealand.Participants were all recruited from the Chinese language programme at this university.Although this programme teaches Mandarin as a foreign language subject, it still attracts a large number of heritage Chinese learners every year.Among all heritage learners, dialect speakers usually outnumber heritage Mandarin speakers.This Chinese programme offers comprehensive Mandarin learning and is divided into three levels: beginner, intermediate, and advanced.As shown in Table 1, the study included seven students with four different dialect backgrounds, each assigned a pseudonym.
Their average age was 21.Except for Meg, who had already graduated from the focal university, all participants were undergraduate students pursuing various majors at the same institution.They had enrolled in the Chinese courses either as part of their general education requirements or as part of a double major program paired with another subject.Among them, there were participants with mixed Chinese heritage.For example, Guimei has both Chinese and Caucasian heritage, while Macy is a second-generation Cambodian Chinese.All participants spoke a dialect at least to a conversational level, and they reported that their dialect was the primary language spoken at home, besides English.

Data collection and analysis
For the first stage, a questionnaire which collected data on participants' background, selfreported language proficiency and self-reported language use were administered to students studying in different levels of the Chinese programme.Of the 56 participants who completed the questionnaire, the seven focal participants of this study were selected by employing a judgement sampling method (Milroy & Gordon, 2003).Based on the principle of the judgment sampling method to 'seek out a quota of speakers who fit the specified categories' (p.30), the sample of the focal participants was representative with respect to the diversity of participants' home dialects and Mandarin competence, while taking account of various possible students' backgrounds (e.g., racially mixed dialect speaker).
Two rounds of in-depth, semi-structured interviews were conducted by using prepared interview protocols (please see Appendix).Interviews were carried out in either English or Mandarin based on the participants' preference.The first round of interviews investigated how participants perceive and respond to the Chinese language policies they encountered at home, in the university's Chinese programme, and the macro societal level.Questions in this section included, for example, 'Do you think the university should offer (participant's dialect) course?'The second round of interviews deepened the information collected from the first interview, extending the conversation based on our understanding of the participant from our first interview.
After transcribing the interviews from audio-recordings, the transcript texts were imported into NVivo, a qualitative analysis software.To analyse the dataset, both deductive and inductive approaches were employed.While deductive analysis helps the authors interpret certain data within the predetermined analytical frameworkmicro language planning and agency, inductive analysis provides an overall description for the entire dataset and can support authors to be 'open and naturalistic in pursuing other aspects of the program' (Patton, 2015, p. 483).The procedure of data analysis was guided by Braun and Clarke's (2006) six phases of thematic analysis.They include familiarising oneself with the data, generating initial codes, searching for themes, reviewing themes, defining and naming themes, and producing the report.

Findings
Accepting Mandarin-dominated Chinese language education All participants tend to interpret the trend of Chinese programmes being Mandarindominated positively, which is shown in their great passion for learning Mandarin in the university Chinese programme and their stated needs of Mandarin education.
They expect to learn Mandarin in Chinese language programmes offered by institutions, as they are unlikely to learn it from home.Different from diglossia families in China (Sun, 2017;Wang & Curdt-Christiansen, 2019), family language use does not appear to shift to Mandarin in overseas Chinese dialect-speaking communities.According to the participants, English and dialects are the main languages spoken at their homes, whereas Mandarin is limited or never used.Most of their parents have a conversational level of Mandarin proficiency but they prefer to speak their dialect with their children.Emily, a Shanghainese speaker, shared her experience of attempting to speak Mandarin at home when she was younger: I used to want to learn Mandarin and asked her [Emily's mother] to speak to me in Mandarin.She will say a few words and she'd stop and say 'No, Mandarin is just Mandarin but Shanghainese is more special.'I was just like 'Oh, fine.'That's pretty much one of the reasons my Mandarin is so bad.
Emily's parents were very proud of their Shanghainese identity and believe that one should be able to speak the Shanghainese language in order to be considered a Shanghainese person.As a result, Shanghainese was the only Chinese language Emily was exposed to at home when growing up.In other cases, it may not be pride, but rather insufficient language proficiency which stops participants' parents from speaking or teaching Mandarin to them.For example, third-generation Cantonese speaker Guimei mentioned her father has rarely taught her either Cantonese or Mandarin.Her dad did not teach her Cantonese because he was worried Guimei may sound like a '金山阿爸' ('Gold mountain daddy', a playful way to address early male Chinese immigrants who worked in mining job) if he taught her Cantonese.Guimei's Cantonese-speaking father seemed to not be confident enough to teach her Cantonese, let alone Mandarin.
Participants embrace Mandarin as something to be taught at school also because of their need and desire to learn Mandarin systematically and proficiently.They expressed passion for learning Mandarin for several reasons.One, oft-mentioned, is the economic value associated with being able to speak Mandarin.Many participants are keen to obtain working experience in China, which was motivated by the strong trading bond between China and New Zealand.Others mentioned Mandarin could help them to improve competitiveness in local job market, given the increasing Mandarin-speaking Chinese population in New Zealand.By commenting 'Mandarin is the future', Taishanese speaker Andrew looked forward to working in Shanghai for his family business as a bridge between China and New Zealand, as he felt fascinated about China's hi-tech industry and business potential.Andrew was aware that knowing Mandarin would be necessary to be able to work in Shanghai.
Another reason that participants desire to grasp Mandarin is that Mandarin helps them to better communicate with other Chinese people.New Zealand is a country with a large number of Chinese immigrants speaking different dialects that are mutually unintelligible.Compared to Chinese youth growing up in China, overseas-born Chinese heritage youth have less exposure to Chinese varieties other than Mandarin and their home dialects.As a result, when hearing people speaking other dialects, they may be aware that it is one kind of Chinese but would have weaker ability to understand it compared to youth in China.Therefore, the use of a common tongue becomes helpful to break communication barriers among Chinese communities.Furthermore, along with the growth of Mandarin speakers in New Zealand, most of participants reported their encounters with senior Mandarin speakers who asked them for directions or bus lines.In this scenario, they need Mandarin to communicate with the seniors who can neither speak English nor understand participants' dialects.
Participants are interested in learning Mandarin also because they expect to possess cultural capital delivered in the form of Mandarin.For example, one participant demonstrated a level of passion for learning about Chinese culture.Guimei gave an excellent metaphor when asked what Mandarin means for her: The 'door' and the 'key'.There are two layers of meanings, one is that it's a door, the other is it offers the key.There are so many sources are delivered in Mandarin, if one only knows Cantonese, then s/he cannot access to many things, can't achieve them, can't understand them.
The 'door' described by Guimei represents the closed opportunity for dialect speakers to access Mandarin sources, such as films and radio programmes.In contrast to their diaspora counterparts, dialect-speaking Chinese youth living in China are not blocked by such a 'door'.In China, because of the tremendous exposure of Mandarin, local youth in even the most dialect-dominated provinces can comprehend Mandarin despite their potentially weak oral competence.However, Mandarin for overseas dialect speakers is also the 'key', as it unties them from solely knowing their own regional culture and enables them to explore the culture and knowledge of other Chinese communities.
In summary, participants accept Mandarin-dominated Chinese education due to their perceived necessity for Mandarin competence.However, fulfilling such needs through home-based Mandarin learning is challenging, especially when compared to Mandarin-background youth who can learn Mandarin directly from their parents.

Resisting potential dialect learning opportunity
Drawing on the study of Wu et al. (2014), we created the same hypothetical scenario for participants and asked them what they would choose if the university offered both a Mandarin language course and the participants' dialect course.All participants expressed their firm determination to continue learning Mandarin still.Such determination can be spotted from their wording, such as 'stick with' Mandarin, 'concentrate on' Mandarin, and 'it has to be' Mandarin.
Participants offered different rationales behind their language learning choice.From the perspective of choosing the Mandarin course, participants reaffirmed the perceived value brought about by Mandarin competence, which includes benefits for career, communication, and cultural capital.From the perspective of resisting the potential dialect course, some participants explained that it is based on their confidence about their dialect proficiency, like Andrew noted 'I think I know enough [Taishanese] to get by so I wouldn't need it'.Other participants resisted it because they believe their dialect could be learned from families at home, which was noted by another Taishanese speaker Eddie, 'all my families speak it [Taishanese], so I don't feel like there is anything to learn [from institution].'In fact, all participants expressed their belief that dialects should be spoken, taught, and maintained in the family domain.Participants consider maintaining dialects an obligation of parents, as Cantonese speaker Meg put it, 'I feel like the parents play a very important role in trying to maintain the language and trying to keep using it at home.' Whereas Wu et al. (2014) found that middle school aged dialect-background learners expressed positive aspirations for learning dialects (if offered by school) and showed little interest in learning Mandarin, participants in this study held opposite preference.In Wu et al. (2014), one participant explained his hope of learning Cantonese instead of Mandarin by saying 'Because most of the schools that I have been to, the boys and girls, the Chinese … most of them speak Cantonese' (p.29).The tertiary-level participants in this study hope to maintain such social network in their dialect-speaking community as well.However, at the average age of 21, they also hope to broaden the social network to a wider range of Chinese people instead of only speakers of their dialects.Moreover, some participants believe being able to speak Mandarin is even beneficial for building up social network with people from other cultures.As Eddie stated, I guess the way that most of the people wanna communicate with Chinese people is gonna be through the standard language, because then you can reach more people.
In her statement, Mandarin has gone beyond the common tongue among Chinese people to be the standard language representing China at the world stage.Another participant, Guimei, also enjoyed the wider social network learning Mandarin brought for her.She mentioned it is a good opportunity to meet more people who are non-heritage learners in her Mandarin class at the university.
Creating 'implementational spaces' for dialect learning When Chinese educational language policy is Mandarin-dominated and institutional support of dialect teaching has declined, dialect students' dialect learning outside of home may seem less feasible.However, we found participants in our study are not passive implementers of this language policy trend, instead they actively exercise their agency to negotiate with it, and to create an 'implementational space' (Hornberger & Johnson, 2007) for dialect learning while learning Mandarin.Macy, a Teochew speaker explained, Um … I feel like, in some way, learning Mandarin has also helped me improve my 潮州话 (Teochew).Because I think the grammar structures are pretty much the same, but compared to English, it's very different.So, sometimes, I catch myself speaking 潮州话 (Teochew), but using English grammar, it's not grammatically correct in 潮州话 (Teochew) … learning Chinese and learning the proper language structure has helped me to speak more fluently, I guess.
Macy's words displayed a neglected possibility, that is, Mandarin acquisition could, to some degree, promote dialect development.Such a possibility has been less frequently acknowledged in previous studies, compared to the predominant discussion of the potential benefits of prior knowledge of dialects for learning Mandarin (e.g., Wu et al., 2014).From Macy's case, we can see that dialect-background learners may have the awareness that they could make use of the Mandarin classes to further refine their proficiency of their dialect when no institutional support is given to their dialects.
Other than the attempt to learn dialects through learning Mandarin, we also noticed participants' active engagement in dialect learning resources outside of the home and school domains.Isabelle, for example, believes that most of her Cantonese was learned from watching the TVB channel, which is a Cantonese-speaking TV channel broadcast from Hong Kong.
I think I mostly watch it to make sure I maintain the language, because in my town, at my school, there were only two Chinese people in my year, there were only me and this other girl, and she didn't come until three years ago, because she's an international student.I'm like surrounded by a lot of English-speaking people.TVB was just … you know, help me retain the language and help me refresh … you know, people talk and stuff.
Isabelle is a special case that comes from a region (Rotorua) with relatively small Chinese population in New Zealand.In Isabelle's environment growing up, there was very limited institutional support for Mandarin education, and less institutional learning opportunities for other Chinese dialects.In addition to this, the small Chinese population in Rotorua renders Isabelle no opportunity to speak either Mandarin or Cantonese to people outside of the family.Despite such disadvantage, Isabelle still managed to retain her Cantonese to a rather high proficiency.From her wording ' … to make sure I maintain the language', it would appear that her learning action using TVB programming is an intentional effort to cope with the local educational language policy which has neglected Chinese language education.
It also revealed that social media plays an important role as a resource for dialect language learning.Shanghainese speaker Emily was especially interested in Shanghainese slang, and she reported a habit of watching factual videos on Tiktok (a short video application) that introduce Shanghainese slang.By watching these videos, she learned that 'there is a lot of that [Shanghainese slangs] you can only say in Shanghainese but if you translate it to Mandarin, it's completely different.'Meanwhile, the existence of social media also provides chances for learners to explore dialect-related regional culture.Taishanese speaker Andrew, who reported being curious about the migration history of the Taishan people, joined a Taishanese community on Facebook to learn about the origin of Taishan people in mainland China.Participants of this study affirmed Payne and Almansour's (2014) finding that learners would exercise agency to create language learning opportunities outside of the classroom when language education is not offered by the school system, especially given the internet and modern technology make it more possible today.

Discussion and implications
Despite Mandarin-dominated educational language policies in Chinese education, this study found that dialect-background heritage learners in New Zealand are by no means passive policy implementers.Instead, they initiate bottom-up micro language planning by exercising their own language agency, which demonstrates the social practice nature of LPP (Levinson & Sutton, 2001;McCarty, 2011).Embedded with free will (Liddicoat & Taylor-Leech, 2021), dialect-background learners exhibit various interpretations, implementations, negotiations, and resistance towards the given language policy situation.They interpreted the Mandarin-language Chinese programme positively, not only because they embrace the economic and cultural capital brought about by learning Mandarin, but also because it is difficult for them to learn Mandarin from their parents at home.By contrast, they did not express interest in the hypothetical opportunities of learning dialects at university.Having said that, their implementation and engagement in the Mandarin-dominated Chinese programme is interwoven with negotiation, that is, learning dialects when engaged with Mandarin classes and external learning sources.
Despite the agency enaction of dialect-background learners in the micro language planning process, their responses to the given language policy situation were shaped by various social constraints.First, their desire for Mandarin learning was based on the instrumental value Mandarin competence is perceived to bring them.Such desire was not found in New Zealand in previous times.Like Wang (2021) commented, 'Chinese was not a popular language for New Zealand schools until China emerged as an economic superpower' (p.173).Then, they responded to Mandarin-dominated Chinese programmes positively also because of the changing language ideology that Mandarin becoming 'the new diasporic standard' (Huang, 2021, p. 165) and the constraints embedded in their social backgrounds (such as parents being not able to teach Mandarin).Finally, other than the acceptance of Mandarin education, it is noteworthy that their negotiation of learning dialects outside of classroom and home was only made possible along with the development of internet and social media.To sum up, dialect-background learners' micro language planning and their exercise of their agency were shaped by various social realities including socio-economic discourse, language ideology, language proficiency, and technology development.This affirms that agency is indeed 'the socioculturally mediated capacity to act' (Ahearn's, 2001, p. 112).
To address the inadequacy in the literature, this study contributed evidence that shows learners to be social actors of micro language planning at the educational level.Compared to Leung and Wu's studies (Leung & Wu, 2011;Wu et al., 2014) in which younger participants show little interest in Mandarin learning, the tertiary-level participants in this study interpreted the Mandarin-dominated Chinese programme positively.This is because participants in the two studies were situated in different social contexts including life stage, social needs in that life stage, and life pursuit in that life stage.Participants in this study hope to better prepare themselves for entering the job market and building broader social networks with Chinese people outside of their own community.Therefore, they hold different attitudes towards Mandarin learning compared to younger learners.This comparison supported Vennela and Kandharaja's (2021) argument that learners' interpretation of language policy varies, and it is important to consider learners' social contexts when examining their agency.Furthermore, participants in this study showed a similar resistance to the heritage language (dialects, in this study) learning opportunity offered by institution as in Willoughby's (2014) study.In contrast, this study further revealed that their rejection of learning dialects at university is based on their strategical and predetermined plan of learning dialects at home or other non-institutional ways, such as the use of social media.Therefore, this study demonstrates that participants' resistance towards multilingual education is not necessarily an action of constraint, but rather an agentive construction reflecting their free will of choice making.
This study provides implications for language policy makers in different level of Chinese language policy.First, it confirms that home is a critical domain for dialect maintenance (Spolsky, 2012).Family language policy makers of Chinese diasporas should pay more attention to children's dialect acquisition, as children may consider it to be their parents' obligation to teach dialects and they therefore expect to learn dialects at home.Then, this study displays the positive influence of the Mandarin-dominated educational policy in overseas Chinese education.Mandarin programmes ran by institution offer opportunities for dialect speakers to learn Mandarin, which helps them to gain the new identity of Mandarin speaker and thereby obtain economic and cultural capital carried by this identity.Wong and Xiao (2010) suggested 'we should assist them in formulating new identities that embrace their global aspirations' (p.171).Having said that, it is necessary that teachers and educators can implement the Mandarin-dominated Chinese programmes with the awareness of learners' multilingual backgrounds (Leung & Wu, 2011) and turn Mandarin classroom to be 'a site where multiple Chinese are valued' (Wu et al., 2014, p. 31).Finally, for the macro-level Chinese language policymaking, this study supports the coexisting relationship proposed by previous studies (Cao, 2017;You, 2006;Zhang, 2018).We argue that it is important to expand the domains of dialect usage, not only in recreational contexts such as films and songs but also on digital learning platforms like language learning apps.These platforms can serve as valuable informal learning resources for dialect speakers seeking to learn dialects through non-institutional methods.

Ethics statement
This study is part of a larger body of research examining dialect-background CHL learners at a university-level Mandarin Chinese language programme in New Zealand.The conduction of this research was approved by the University of Auckland Human Participants Ethics Committee under reference number 023842.