“Learning Cantonese will help us”: Elementary school students' perceptions of dual language education

Introduction

Advocates of dual language immersion (DLI) programs cite the benefits of language minority children accessing classroom content and developing ‘cognitive academic proficiency’ (Cummins, 1979), wherein learning the home/heritage language at school also assists in developing L2 academic language. DLI programs also bring together English-dominant students and students who speak English as a partner language¹, an intentional move that ‘gives a middle-class orientation to what might have been low-income schools’ (Valdes, 1997, p. 412). DLI programs also aim to counter the types of elite bilingual education that tend to benefit language-majority children and instead celebrates and incorporates immigrant bilingualism, with goals including bilingualism, biliteracy, cross-cultural understanding, and high academic achievement (Perez, 2004; Palmer, 2007, among others).

Many DLI programs in the United States have three goals – (a) academic achievement, (b) bilingualism and biliteracy, and (c) the development of student cross-cultural competence. However, this third goal has not received much scholarly attention (Feinauer & Howard, 2014; Parkes, Ruth, Anberg-Espinoza & de Jong, 2009). Of the limited studies, much of the work has been done with Spanish DLI programs and with secondary school students. Follow-up studies with secondary students who had attended elementary DLI programs show that positive cross-cultural attitudes towards the target language groups, as well as towards self-esteem and motivation remain even after graduating (Bearse & de Jong, 2008; de Jong & Bearse, 2011; Lindholm-Leary & Borsato, 2001). Studies also show differences in attitude depending on student’s ethnicity, with Anglo students seeing bilingualism as a positive attribute for various reasons but feeling that they have not become bicultural, while Latino students note that DLI schooling fostered biculturalism (Bearse & de Jong, 2008; de Jong & Bearse, 2011). In contrast, other studies have found evidence of less Spanish use and weaker linkages to cultural identity for Latino students from Spanish-speaking-homes as they grew older (Potowski, 2004, 2007).

More research is needed on language attitudes of children enrolled in DLI programs and how their attitudes are associated with their identity and achievement. Furthermore, more research is required when it comes to looking at children enrolled in other non-Spanish DLI programs, such as Chinese (both Mandarin and Cantonese) as much of the work in the past has been conducted with students in Spanish DLI programs (e.g. Linton, 2004; Potowski, 2007; Quintanar-Sarellana, 2004). This study aims to examine the attitudes towards dual language education of younger Cantonese speakers/learners and their ideas about their bi/multilingual identities. This work has implications for home and educational practices, particularly for educators working in DLI programs.

Dual Language Immersion Programs

DLI teachers teach in both languages, usually using one of two main models: the 50:50 model, where all grade levels are exposed to an even balance of English and the minority language, or the 90:10 model, where in kindergarten the students are exposed 90% of the time to the minority language and 10% to English, until 4^th or 5^th grade where exposure becomes 50–50 (Baker, 2006). Lee, Hill-Bonnet, and Raley (2011) argue that this continual shifting of linguistic contexts in DLI classrooms in terms of classroom subject content help to create a more equitable environment wherein students of various linguistic abilities are able to take on the role of language broker.

Studies have shown that attending DLI elementary school programs lead to higher longitudinal high stakes test results in reading, writing and mathematics than their peers in mainstream classrooms (Cobb, Vega, & Kronauge, 2006). Freeman (2008) reported that while language majority (English L1) students and language minority (Spanish L1) students did not reach equal bilingual and biliterate standards despite being in the same DLI program, both groups still outperformed their monolingual peers by the time they reached upper elementary grades. Similar outcomes have been reported for DLI students performing at the same or surpassing their peers in traditional bilingual education programs (Castillo, 2001; Coy & Literland, 2000; Stipek, Ryan, & Alacon, 2001) as well as performing on par with native English speakers in English-only classrooms (Lindholm-Leary, 2004).

At the same time, scholars have also pointed out that mainstream dominant discourses of race and power, and how these discourses map on to language, serve as hindrances in DLI programs being able to fulfill their mission of educational equity for all (de Jong, 2014; García, 2005; Palmer, 2007; Reyes, 2005). This can manifest in a type of diglossia (Potowski, 2004; Tarone, & Swain, 1995) where English is the favored language for social functions at school, while the minority language is limited to academic contexts and with the teacher. Majority-minority group tensions can also result in segregation in administration, teachers, and parental involvement (Lee & Jeong, 2013; Palmer, 2007; Valdes, 1997). Indeed, Alanis and Rodriguez (2008) report that features of successful DLI programs incorporate an active parent-home collaboration, qualified bilingual teachers, pedagogical equity, and knowledgeable leadership.

Cantonese Speakers

Though DLI has been an increasingly popular form of bilingual education in the U.S., much of the documented work has been on Spanish-English (Linton, 2004; Potowski, 2007; Quintanar-Sarellana, 2004; Senesac, 2002; among others) and some on Korean-English (Bae, 2007; Lee & Jeong, 2013; Sohn & Merrill, 2008), and far less work has looked at Cantonese-English bilinguals in the U.S. This is somewhat surprising and perplexing, since Cantonese-English bilingual education has been taking place in the U.S. since the 1900s with the arrival of the first wave of Chinese immigrants; it was also parents of Cantonese-speaking students who advocated for the right to linguistically appropriate accommodations in the 1974 Lau v. Nichols civil rights court case in which the Court unanimously decided that the lack of supplemental language instruction in public school for students with limited English proficiency violated the Civil Rights Act of 1964. Although the majority of English language learners (ELLs) in the U.S. are Spanish speakers, Chinese speakers are second (Park, O’Toole, & Katsiaficas, 2017; U.S. Census Bureau, 2015). Between 20–36% of all Chinese Americans (the percentages range depending on the variety of Chinese spoken at home) report not speaking English well or at all (Cooc & Leung, 2017). The meager work that focuses on Cantonese-English bilingual students in the U.S. and Canada almost singularly deals with their acquisition of English, not their acquisition and/or maintenance of Cantonese or their development of identity and/or cross-cultural competence; work on children learning Cantonese as a foreign language (FL) in the U.S. is nearly non-existent beyond case studies, though some research has been done in Hong Kong (Boyle, 1997; Li & Richards, 1995). As such, due to the perception based on the model minority myth that Cantonese-heritage students are successful in school and easily assimilate, the gains that Cantonese-English bilinguals have achieved from being a part of bilingual programs, even if it is as obvious as their having been able to maintain Cantonese or develop cross-cultural competence, have been overlooked and understudied by researchers.

However, Cantonese speakers have had a continued presence in locations like the San Francisco Bay Area, where they represent 34.61% of the total ELL population of San Francisco Unified School District. Researchers have noted the value and ties between Mandarin Chinese and budding linguistic and economic markets (Curdt-Christiansen, 2009; Li & Hua, 2010), but few current studies have documented how a Chinese language like Cantonese, whose speakers still comprise a formidable ELL population in the U.S., is acquired, developed, and maintained. Additionally, it should be noted that students who enroll in Cantonese DLI programs are not just heritage speakers of Cantonese, but include other students of color and White students who(se families) aspire to attain various forms of social and linguistic capital. This type of decision at a school-wide level is exemplified in Wu’s (2017) research at a northeastern U.S. city in an urban charter middle school, where Mandarin was taught to all students as a means to empower students to become global citizens vis-à-vis a “world language.” Thus, the documentation of the Cantonese language development of these different groups of Cantonese-English bilingual children and their shared, as well as differentiated, language learning motivations and trajectories has implications for Cantonese linguistics, foreign language instruction, and heritage language maintenance and education.

Language Attitudes and Identity

In earlier discussions of language maintenance in minoritized communities, Fishman (1991) optimistically implies that if a minority group is united and willing enough to protect and preserve their ethnocultural boundaries, intergenerational mother tongue transmission can be possible regardless of the group’s size. However, Canagarajah (2008) notes that Fishman’s “romantic” line of thinking that local communities can somehow single-handedly tackle the “superhuman struggle” for language maintenance minimizes macro-level constraints such as material inequality, ideological domination, and the role of history in language maintenance (p. 172). These links illustrate the connections among language, language attitudes, and identity construction. Cangaragah’s own work with diasporic Tamil communities in Canada, the U.S., and Britain shows how hybrid identities are continually constructed to make room for multilingualism, “adopting diverse practices to ‘perform’ identity as contextually relevant” (p. 170). Such practices directly influence how diasporic Tamil families view their own responsibilities towards language and cultural maintenance. As Pennycook (2007) also notes, “language identities are performed in the doing rather than reflecting a prior set of fixed options” (p. 115) and that “the ways in which languages are being mixed and changed presents new possibilities for identities that have little to do with national identifications” (2010, p. 5). For the purposes of this research, that is to say that there are ways in which learners of Cantonese, both of Cantonese heritage or not, can use and celebrate their learned and inherited language(s) and cultures that also simultaneously reconciles their identities as “American.”

Language learning and attitudes about language learning are not limited to the confines of the classroom; children’s beliefs about what they are learning and why are also heavily influenced by family members. Howard (2008) points out that oftentimes, adults who themselves had been subjected to “draconian language policies” that did not allow certain languages to be used in school perceive the educational difficulties that they encountered stemmed from their own families’ use of those certain languages (p. 191). In hopes that their own children do not encounter such educational roadblocks, the parents may decide to adopt a monolingual family language policy that matches closely to the school’s language policies. Curdt-Christiansen’s (2009) ethnographic inquiry into ten Chinese immigrant families living in Montreal shows that parental perceptions about their children’s educational and linguistic trajectories were also grounded in the parents’ own experiences and values, as well as the sense of “blocked opportunities” they felt as immigrants (p. 360). Many of the parents considered multilingualism to be an asset, with values ascribed to different languages correlating to their perceived market values of the languages in the linguistic marketplace. These beliefs of striving towards multilingualism were also tied to the parents’ aspirations and perceptions of successful parenting.

With regards to the linguistic marketplace, Li and Hua (2010), examined the perceived hierarchies among the varieties of Chinese from Chinese diasporic communities in Britain, Australia, and Singapore. They found the “Chinese as Mandarin” ideology to be prevalent (p. 158) and noted that language shift “is happening very rapidly in the Chinese diasporas” (p. 165). They state that it is “clear that Cantonese, Hakka and Hokkien are losing their place to Putonghua” (p. 170). Along a similar vein, when Weger-Guntharp (2008) describes a U.S. university-level Mandarin classroom as one where Cantonese played a role for half the class, she writes that Cantonese heritage students find their Cantonese skills to be “not useful” (p. 223). Though convenient as metaphors, Li and Hua’s dichotomy of Chinese languages as “winning or losing out” against each other and Weger-Guntharp’s participants’ dichotomy of “useful or useless” languages critically overlooks the possibility of multiple languages being included in one’s range of linguistic resources. Rymes’ (2010) notion of a communicative repertoire, or “the collection of ways individuals use language and literacy and other means of communication (gestures, dress, posture, or accessories) to function effectively in the multiple communities in which they participate” (p. 528) shows us how multidirectional inclusion of repertoires by interlocutors can foster pride in cultural and linguistic differences. A language-as-resource (cf. Ruiz 2010) orientation to multilingualism reminds us that all languages can be seen as advantageous because they occupy various roles in speakers’ and learners’ lives, beyond simply utilitarian values. Ruiz further (2010) argues that while many language policies are driven by the economy of language and do not necessarily aim to promote cultural democracy and social justice, the issue is how researchers “can accommodate it without having it define the entire effort” (p. 162) while giving guidance on how to promote the use of minority languages. We see these macro-language planning issues to be connected to micro-level ideologies and attitudes about language and language use, including how these beliefs are connected to neoliberal and classed discourses circulating about “useful” languages. While much work has examined the spread of English as a de facto language policy (cf., Piller & Cho, 2013) or how Mandarin Chinese has become the de facto “Chinese” language globally because of the economic rise of China, our work looks at how learners and heritage speakers of a Chinese language that enjoys less cachet compared to Mandarin make sense of their language learning experiences and build up positive beliefs about its utility and “competitiveness” in the imagined marketplace. We also illustrate how these differences in beliefs manifest between those learning Cantonese as a foreign language and those learning it as a heritage language.

Present Study

Given existing research on Spanish DLI (Senesac, 2002; Lindholm-Leary, 2014; Linton, 2004; Quintanar-Sarellana, 2004; Potowski, 2007; Palmer, 2007), Korean DLI (Bae, 2007; Ha, 2001; Sohn and Merrill, 2008; Lee and Jeong, 2013) and Mandarin DLI (Lindholm-Leary, 2011; 2016), this project examines a language background that has been spoken in the U.S. since the late 1800s but still is sorely lacking in the language learning literature in the U.S. context: Cantonese. For research on less commonly taught languages, Lee and Jeong (2013) write that there is still much to be investigated when it comes to DLI’s impact on various stakeholders, especially children and their language learning experiences, which has direct implications for school policies and practices. Lee and Jeong further state that many existing studies on DLI do not provide information about students’ perceptions and experiences, and as such, they designed a research study collecting information about various stakeholders’ thoughts about Korean-English DLI education. Doing so illuminated the multiple perspectives that underpinned participants’ understanding of language and academic outcomes of DLI education.

Along the same vein as Lee and Jeong (2013), we look at the linguistic gains, socioemotional impact, and pedagogical achievements and challenges at a DLI elementary school that teaches Cantonese and English. With a student population that is approximately one-third L1 English speakers learning Cantonese as a foreign language (CFL), one-third Chinese heritage students who speak Cantonese as their L1 and are learning English as their L2, and one-third Cantonese-English bilinguals learning Cantonese as a heritage language (CHL), this DLI school is unlike many documented DLI programs that have roughly 50% White Anglophone children and 50% first-generation language minority (usually Spanish) speakers. Notably, when this school was established, it made the explicit choice to reflect the minority language speakers of its surrounding community and to teach Cantonese as the target language instead of Mandarin, what many would label an arguably more ‘economically solvent’ Chinese language. While Mandarin is considered a Less Commonly Taught Language within heritage language learning arenas, Cantonese is considered a ‘Truly Less Commonly Taught Language’ (TLCTL, cf. Gambir, 2001). As TLCTLs do not receive much investigation in research domains, it is worth further examining this type of Cantonese-English DLI program. This paper is part of a larger project that investigates the gains and challenges for students attending a Cantonese DLI school, with data drawing from student focus group interviews, student surveys, and classroom observations. For the purposes of this paper, only the focus group interviews describing student perceptions of their DLI schooling will be discussed.

Research Questions

Our study aimed to answer the following overarching research question: what are the attitudes towards dual language education of younger Cantonese speakers/learners and their ideas about their bi/multilingual identities in their own words?

To address this question, we conducted two 60-minute focus groups with 4^th and 5^th graders over two days in May of 2016. This specific age group was chosen because as upper elementary school students, they were most likely to be able to articulate their viewpoints most clearly. Additionally, since many of the students at this school began as kindergarteners, 4^th and 5^th graders had also experienced DLI the longest.

Methodology

Participants

Informed consent forms in both Chinese and English were sent home to all fourth and fifth graders, and a total of 60 students (35 fourth grade, 25 fifth grade; 29 boys, 31 girls) returned guardian consent forms and participated in the larger study. The majority of participants (N=43) were Chinese Americans, followed by students of non-Asian backgrounds that included white, African American, and Hispanic (N=10), and other Asian backgrounds (N=7). Additionally, the majority of the participants (88.3%) had attended the school since kindergarten.

All children were enrolled in a public K-5 elementary school in an urban school district in California. In this school district, Cantonese-speakers represent over 30% of the total ELL population. The focal school was chosen because they offered a Cantonese-English dual immersion program. There were approximately 400 children enrolled in the school, with 66.9% identifying as Asian American, 14.1% as white, 10.9% as two or more races, and 1.2% as African American. About 28% of the students were classified as English learners and 15% were classified as being socioeconomically disadvantaged.

Fifteen out of the 60 students who participated in the survey were invited to take part in two focus groups, which took place during lunch time (total one hour each) in a classroom at the school for two consecutive days. Each student attended only one focus group meeting. The 15 focus group children were chosen based on the survey responses. We aimed to choose equal numbers by grade, by gender, home language, and by language learning status (Cantonese as a heritage versus foreign language). Children were informed that they were allowed to stop participation at any point during the focus group sessions. Of those 15 students, one decided to opt out, and in total, there were 14 students who participated in the focus groups, equally split between boys and girls. There were 8 fourth graders and 6 fifth graders. Table 1 details each of the participants and their self-reported ethnicities and language backgrounds.²

Table 1.

Description of Focus Group Students

Pseudonyms	Self-identified ethnicity	Self-reported language background
Xander	Not Asian	Native English speaker
Jackson	Chinese/Chinese American	Native Chinese Speaker, Chinese- English bilingual
Aaron	Chinese/Chinese American	Chinese-English bilingual
Allison	Not Asian	Native English speaker
Erica	Other Asian (not Chinese)	Chinese-English bilingual, native English speaker
Melody	Chinese/Chinese American	Native Chinese speaker, Chinese- English bilingual, Chinese heritage learner
Sherise	Chinese/Chinese American	Native English Speaker
Colin	Chinese/ Chinese American	Chinese-English bilingual
Jason	Chinese/ Chinese American	Native Chinese speaker, Chinese heritage learner, Chinese-English bilingual
Frederick	Not Asian	Chinese-English bilingual
Lorie	Other Asian (not Chinese)	Native English speaker
Cora	Chinese/Chinese American	Chinese heritage learner, Chinese- English bilingual, Native English speaker
Neal	Chinese/Chinese American	Chinese heritage learner; native English speaker
Melissa	Not Asian	Native English speaker

Note: All pseudonyms. Self-identified ethnicity choices were: Chinese/American, Other Asian (not Chinese), Not Asian. Self-reported language background choices were: native Chinese speaker, Chinese heritage learner, Chinese-English bilingual, native English speaker.

Measures

We used the following questions, which were adapted from Lee and Jeong’s (2013) study with Korean-English DLI students, as a skeleton to drive the focus groups:

• Do you want to be bilingual? Why? What do your parents think about being bilingual?

• Do you think you should know how to read and write in Chinese (Cantonese), or is just knowing how to speak it enough?

• How are you doing in school? What is easy? What is hard? What do you like about your school? What do you not like about your school?

• Do you want to attend a DLI program again next year? If you had your choice and if programs existed, would you go to a DLI program all the way to high school?

• Name your three favorite and three least favorite activities you do in Chinese (Cantonese), in English, or in both languages

All questions were covered, though not necessarily in the same order for each focus group. Focus group sessions were led by the first author, a Cantonese-speaking Chinese American, and conducted with the help of two trained research assistants, one a heritage speaker of two varieties of Cantonese and one a heritage speaker of Chamorro. As focus groups go, participants were very talkative, often engaging one another in extended conversations and not merely answering the interviewer’s questions one-by-one, which made it a rather seamless process answering all of the questions on our interview guide. Responses were elicited in more of a popcorn approach, with special attention to elicit responses from more reticent students if/when necessary. Focus group data were transcribed and verified by the same two trained research assistants. Beyond a sprinkling of Cantonese words and phrases, the focus groups were conducted in English; the Cantonese words were checked by the Cantonese-speaking research assistant and first author. Next, members of the research team read through the transcripts doing open coding, writing memos and discussion, and then continued with more focused coding (Emerson, Fretz, & Shaw, 1995). In our initial rounds of open coding, the research team noticed participants’ articulations of language ideologies related to language learning. The first author and second author, a bilingual speaker of English and another Asian language, then went through all the data looking for these themes as they related to the students’ overall schooling experiences. Finally, third author, a Chinese American Cantonese speaker, then went through the themes and quotes again. We feel our research team’s overlapping identities with our student participants give us a unique perspective into this specific student population. At the same time, we recognize the limitations that these overlapping identities have in shaping the ways we view the data, which we have attempted to mitigate through our iterative process of data analysis and multiple rounds of discussions to talk through the data.

Results

Below we report three of the main themes from the focus groups: utilitarian and intrinsic ideologies, the need for contextualizing language learning, and their general perceptions about their schooling experiences and building communicative repertoires.

Utilitarian and Intrinsic Ideologies

To begin, the 14 students were exceptionally well-versed when it came to articulating their thoughts about being bi/multilingual. Many revoiced their parents when it came to why knowing more than one language was important, and others discussed benefits they personally experienced when using a language other than English. This was evidenced in both of the groups (heritage language versus non-heritage language), supporting past literature with Spanish DLI students (Lindholm-Leary, 2016).

For example, Erica, who was a non-Chinese Asian and self-reported as both a Chinese-English bilingual and a native English speaker, mentioned, ‘I want to be bilingual because my parents said you’ll get a better job if you are. Then you can go to other countries where stuff is cheaper, and then you can buy stuff from there using the language’ (FG1, #1). Erica furthers, ‘My parents, whenever we go to dim sum, they make me order in Chinese and then we get free food’ (FG1, #1). Sherise, who self-reported as a native-English speaker, said, ‘if you ever travel, it would be easier to talk like if there’s a different language that they speak. Then, you can speak with their language’ (FG2, #3). Additionally, Frederick, who is non-Asian and self-reported as a Chinese-English bilingual, adds, ‘My parents don’t, they only speak English, but they think that being bilingual is really good because it can help you get jobs, and meet other people that don’t speak English’ (FG2, #8).

This ‘us versus them’ dichotomy is evident in Melissa’s (who is non-Asian and self-reported as a native English speaker) response as well, ‘Cause I would like to be bilingual because I think that it’s helpful for people who don’t know your language and you could just talk to them in their language’ (FG2, #7). Scholars have noted how pronoun use illustrates how various speakers position themselves socially and in relation to others. The selection of pronouns and other stylistic devices is not random and can represent speakers’ status differentials and/or explicit or implicit ideologies (Wodak & Meyer, 2011). Through Melissa’s use of pronouns, it is clear that the additional language, in this case, Cantonese, is still seen as secondary to these students’ primary language (English). Thus, while it seems that bi/multilingualism is definitely seen as a positive attribute for various reasons (e.g., better job attainment, free food when deploying Cantonese at a Chinese restaurant), for some students (including both heritage speakers), it was still seen as the language of ‘the other.’ This may be due to the fact that on average, these children were novice Cantonese learners living in the U.S. As a result, because they speak English as their primary language in school and at home, Cantonese is seen as a desirable ‘add-on’ language to be bilingual in. We see that ‘bilingual’ is being used as a sort of social persona, recognizable because of the stereotypically recognizable forms of discourse used to discuss its benefits (cf., Agha, 2004). In other words, even at the 4^th and 5^th grade levels, students are able to articulate and position themselves as specific types of social personae who partake in bi/multilingual practices, affording them with stereotypically recognizable assets (e.g., better jobs, heightened access across different communities and while traveling).

In addition to the more general bilingual persona that is described, there was also a more specific ‘Cantonese bilingual’ persona that was linked more directly to heritage language maintenance that Cantonese-heritage Chinese Americans were able to articulate. Specifically, Neal stated, ‘I wanna be bilingual because my dad, he came from China and he, and at my grandparents’ house, they only know how to speak Chinese [Cantonese]. And I want to respect their culture’ (FG2, #2). Colin also said, ‘I think I would like to be bilingual because my family. My grandma came from China. And then, since she only speaks Chinese [Cantonese], I guess I, the only way I could respond is by speaking in Chinese [Cantonese]’ (FG2, #6). Thus, for ethnic Chinese students at this school, being bilingual means being able to access family communication by necessity through a common language, as many who have written on heritage language loss (including Wong Fillmore, 2000, and others) have also iterated and that heritage languages have multifaceted values defined along intellectual, aesthetic, cultural, economic, social, and citizenship planes (Ruiz, 2010).

In many ways, the students in the focus group came from exceptionally diverse and multilingual backgrounds, including Spanish, varieties of Chinese, Vietnamese, Korean, Italian; as such, students were especially very aware of the complexities of bi/multilingualism across the lifespan, especially when it came to language attrition. Colin stated, ‘So my dad speaks English. He used to speak Mandarin, but he never practiced it, so he kinda stopped speaking Mandarin. And then, my mom was born in Hong Kong, so she speaks, so she speaks Mandarin, Cantonese, a little bit of Spanish, and English’ (FG2, #6).

When it came to language policy at home, Jason mentioned, ‘I think my mom and dad sometimes disagree with being bilingual, because, my grandparents think that we always argue if you should be doing English or Chinese, and I don’t want that to happen. Because that can cause a big commotion (FG2, #1).’ How language was discussed at home also impacted school learning too, as Sherise discussed, ‘For Chinese [Cantonese] sometimes we have to do reading. We’re supposed to record our readings, and I usually need help with that. Even though my parents can speak, it’s just that since it’s been so long since the last time they’ve actually used it a lot. They can’t really tell me how to write words or say them. And my brother, he sort of knows, but now that he’s been learning it in, all these like Cantonese characters in Mandarin, he can’t give, he tells me the pronunciation or how to enunciate in Mandarin so I get a little confused’ (FG2, #3). Thus, even at the 4^th and 5^th grade levels, Cantonese-heritage students were able to articulate the very complex mechanisms that went into acquiring Cantonese spoken language and literacy and the ways in which family could and could not assist in this process.

Contextualizing Language Learning

The second theme that we will report from our student focus group findings is the importance of (as well as contextualizing the need for) reading, writing, and speaking, not just the teaching/use of one isolated skill. Responding to the question of whether it was important to read and write in Chinese, Erica stated, ‘I think you should be able to read and write it because if you were in China and you don’t and if you like, you’re at a restaurant then you, you’re looking at a menu ((students giggle)) sometimes there’s like no English translations, and then you’re not gonna know even though you know how to speak Chinese’ (FG1, #1). Thus, alluding again to international travel and the need to read a menu to order food, Colin brings up the scenario of being abroad and needing to read a menu without English. Similarly, Colin mentions the situation where spoken language is not enough: ‘If you know how to speak it [Cantonese], and then you’re able to ask people, but some people can’t give the best explanations. And the people, so if you can’t read it, read the signs or the numbers at Chinese restaurants, you’ll, you basically shouldn’t, you shouldn’t be there because, then, you wouldn’t know what you’re gonna be doing’ (FG2, #6). Again, in the fictive scenario of the restaurant and needing to order, this student brings up the need to not only speak in Cantonese but to also read Chinese characters. Interestingly, this student also takes the more extreme stance that if one does not know how to read characters, ‘you shouldn’t be there.’ Thus, linking spoken and written literacies together as integrated skills that must be acquired to be a successful Cantonese-speaking person, this student seems to be articulating that isolated skills are not enough for real life language use in context. Not only this, this student is also taking a stance to legitimize who ‘belongs’ as a Cantonese-speaking person abroad and who does not. We argue that this falls under more of the utilitarian scope of language persona (i.e., using language effectively in order to achieve the goal of ordering food) as opposed to the intrinsic scope of communicating intergenerationally.

The focus groups were overwhelmingly dominated by English, and participants were still using English even when discussing reading the menus and ordering in the imagined restaurant setting. We argue that students discussing the need for integrated (and utilitarian) language skills hearkens back to the first finding discussed earlier, where these integrated skills are still linked to recognizable assets, in this case, being accepted as a successful traveller abroad or knowledgeable culinary connoisseur. Interestingly, many of these examples where students commented upon acceptable and unacceptable language-using behaviors took place within a fictive context of traveling abroad or reading a menu at a Chinese restaurant, and in this imagined context, the place students are going to and the restaurant staff they are interacting with are all Cantonese-speaking. This relates to Bauman’s (1993) notion that eliciting any kind of story establishes a platform for, and elicits a performance by, the speaker. Also inherent to any performance is the concept of voice, which constructs speakers’ identities as well as juxtaposing them against each other. We use voice to refer to how utterances index the various ways of speaking that are “associated, by virtue of linguistic ideologies with different character types, professions, genders, social statuses, kinship roles, moral stances, ideological systems, age groups, ethnicities, and so forth” (Keane, 2001, p. 269). The juxtaposition of voices and selves at various times and places during the performance is what Goffman (1959) calls role distance, which is similar to what he later calls footing and what others may call speaker’s stance. Theorists have noted that this distance helps to distinguish performance roles and selves (Haviland 1996; Levinson 1988). Thus, we argue that role distance here, namely, students’ perceived distance from (as well as closeness to, because of their educational training at this school) Cantonese makes it possible for respondents to comment favorably upon Cantonese, its relevance in the sociolinguistic marketplace, and the need for integration of skills when it comes to its use.

Classroom observations of actual language use, which were conducted but were beyond the scope of this paper, would also help further unpack these language ideologies and the ways students understand and deploy flexible bilingualism (Creese & Blackledge, 2010) or translanguaging (García, 2009).

Positive Schooling Experiences and Building Communicative Repertories

The third theme that resonated from the focus groups is the strong connections students seemed to have with the school and the Cantonese-specific content linking directly to Chinese American contextualized histories. As many of the teachers and school administrators also identified as Chinese American, this solidified the interpersonal connections students had with the content they were learning from their teachers. This mirrors findings from Lindholm-Leary’s work (2016) with Two-Way Immersion (TWI) Mandarin programs, where both the heritage students and native English speakers were also glad to be in the program.

When asked what students liked about what they learned in school, Jackson responded, ‘That it is Can-Cantonese’ (FG1, #5). For those who were not of Cantonese heritage, they spoke of learning a code that their families could not understand, as Erica stated, ‘Sometimes at my house, me and my siblings get to like talk in a secret language [Chinese] because my parents don’t know what it means (FG1, #1). Others noted specific curricular activities, such as ‘Singing in Chinese’ (FG2, #3, Sherise) and ‘I like to read about the transcontinental railroad (FG1, #7, Xander).’ Still others talked about school administrators (‘It has a good principal’ [FG1, #7, Xander]) and the perceived quality of teaching (‘They have good education’ [FG1, #5, Jackson]). Overall, overwhelmingly students stated that they had positive experiences at the school.

Thus it seems that DLI schooling fostered positive interactions and engagements with Chinese American histories and publics, especially for those who are from those communities, which is one of the tenets advocates for the teaching of ethnic studies at the K-12 levels have been promoting (Cammarota, 2011; Sleeter, 2011). This mirrors the literature on culturally responsive pedagogy (Ladson-Billings, 1994) and the importance of ethnic minority representation in U.S. classrooms as ways to mitigate the racial mismatch and stereotype threats (McGrady and Reynolds, 2012). In other words, regardless of whether students are of Cantonese heritage backgrounds, Cantonese DLI education seems to be serving as a way to build up communicative repertoires for all students as well as developing a language-as-resource view towards acquiring Cantonese for these students. Thus, through viewing Cantonese as a “secret language” that other family members do not know or as a language to sing in, these students are orienting themselves towards viewing Cantonese positively, and along with this, positively experiencing Cantonese language education and their larger schooling experiences in meaningful ways. To this end, we would like to argue that this type of orientation moves beyond a mere linguistic proficiency view of language acquisition into a more multidirectional inclusion of repertoires, allowing for the building up of pride in cultural and linguistic similarities as well as differences.

Discussion and Implications: Learning Cantonese Can ‘Help’ – But in What Ways?

As described in the findings above, for many of the students in the focus group, many of whom already come from relatively multilingual backgrounds, DLI education is viewed as an ‘add-value’ setting, wherein they are able to access various stereotypical privileges attributed to the social persona of a bilingual person. Cantonese, for most, is an ‘add-on’ to English, and in some ways, might easily be replaced with any other non-English language, and also operates in somewhat of a diglossic nature. Yet for those who are of Cantonese heritage, the fact that it is Cantonese that they are being immersed in is of particular salience, as it might be one of the few venues to learn the language of their family in institutional settings. Particularly for the heritage speakers, DLI teachers could encourage them to interview their families about their home countries, including family histories and traditions, and provide them with opportunities to introduce them to non-heritage speakers of the class so that all students can find more contextually relevant ways to use Cantonese beyond restaurant or traveling situations.

We would be remiss if we were to not to acknowledge that these discourses are classed and neoliberal in nature (Park, 2010; Shin, 2016). While much of the work on neoliberalism looks at how it operates in the spread of English globally as de facto language policy (cf., Piller & Cho, 2013), there is less existing literature on how these neoliberal discourses operate on languages not typically thought of as economically powerful. Cantonese arguably does not have the same prestige as English or Mandarin in global markets and is oftentimes talked about as a “dying” language in media reports when compared to Mandarin (Semple, 2009), but our focus group findings demonstrate how, for these students, learning Cantonese can engender positive beliefs about this language’s utility and “competitiveness.” To be sure, these beliefs do not operate in isolation and are mediated by the students’ parents and other family members’ own beliefs about language(s), evidenced by the many times students revoiced their parents’ words in our study. Using DLI as a vehicle towards “middle” (and arguably upper-middle) class orientations (Valdes, 1997) for parents and their children, as well as how parental beliefs are transferred and mediated through their children, and how this translates into students seeing and talking about themselves as “competitive” actors in the market, is a strategy that is worth investigating further. This research also suggests that teachers can point out the values in learning a second language, introducing to the students the various ways bilingualism is an advantage in their future careers. Beyond neoliberal and classed discourses, this work sheds light on the need to better understand the complex ways learners see themselves as bi/multilingual individuals and their various intrinsic and instrumental perceptions of language learning and use.

Specific to the DLI context, seeing that these are the voices of the students themselves, there is a need to address these types of bilingual personae in DLI settings and to fold in more of the critical sociolinguistic elements of language learning beyond ‘us versus them’ and language use in a restaurant setting. Doing so could enhance understanding of the place of Cantonese within the U.S. historical context. Pedagogically, this might include field trips to historical sites with direct connections to Cantonese Chinese immigration, including Angel Island Immigration Station, where nearly one million Asian immigrants, most of the Cantonese-speaking Chinese people, were processed and restricted from entering the U.S. because of the Chinese Exclusion Act from 1910–1940. A similar site would be Locke, an agricultural town established by Chinese immigrants (and the only one of its kind that still stands). Thus, using these historical sites as backdrops for contextualizing the relevance of Cantonese in California might help to establish the fact that Cantonese is not just a foreign or heritage language for students, that it is also a Californian language, chipping away the hegemonic ideologies of English in the U.S. Introducing and discussing books on Cantonese people and culture or books written by Cantonese-speaking writers may also add to their dual language immersion experience and to their social persona and their identities. Hiring and providing professional development for Cantonese-English bilingual teachers is also crucial in creating these positive schooling experiences and building communicative repertories.

Data from these focus groups illustrate, from the DLI students’ own voices, the ways in which they understand their Cantonese language learning as well as the ways they view what it means to be a bi/multilingual person in the contemporary, multilingual world. Though this paper draws only from a small set of students’ views, it still is worth investigating the complex interactions between students’ discussions of the perceived utility of language(s) and their actual performance in the classroom, as well as engaging in further member checking with other students who attend the DLI program.

We recognize that one of the limitations of this paper is that we as the research team, though many of us share bilingual identities as bilingual Cantonese-English speakers and/or heritage speakers of minoritized languages, do not speak for the participants themselves. As such, additional work, particularly longitudinal, ethnographic work, that investigates the different ways DLI programs impact Cantonese-heritage students and their non Chinese-heritage peers would also be useful in understanding how DLI programs serve ethnic and linguistic minorities in contexts like California, which is a minority-majority state. Furthermore, longitudinal qualitative and quantitative data following students beyond the elementary school level, as well as their interactions with family and community members outside of classroom contexts, would help to illuminate how students from DLI programs perform across their lifespan.

Few studies have examined the development of student cross-cultural competence and on the language attitudes of children enrolled in DLI programs and how their attitudes are associated with their identity, particularly with Cantonese-English bilinguals in the U.S. The present results, using the DLI students’ own voices, are only a first step in investigating their language attitudes and identities. More work in this area is crucial to provide the knowledge needed to inform DLI schools, as well as families who are trying to raise their children to become bilingual and bicultural in the midst of sociolinguistically and sociopolitically complex educational and global terrains.