We live in an age of “time–space compression” (Reference HarveyHarvey, 1990, passim), characterized by a global, multidirectional, and often instantaneous flow of languages and lives. Whether a blessing or a burden, it is now possible and even necessary for us to live in multiple spaces at once, given technological innovations. If there is something that gives us a reliable sense of coherence and continuity between the past, the present and the future and between the here and the there, it is probably the languages that we speak. When we move to new places and meet new faces, either in person or virtually, and we certainly do much of these nowadays, we may or may not carry our luggage, but we always carry with us our language. But what will happen to the languages that we carry with us when we move from place to place? And what will happen to our sense of self and our sense of the world when the languages we know change their sounds and significance over time?
The languages that are naturally and necessarily attached to us from the past as we build our future in new places are often called “heritage languages” (HLs). HLs accentuate the contingency, hybridity, and indeterminacy against essentialist conceptualizations of identities, communities, race, nation, and culture. They compel us to cope with shifting and sometimes conflicting linguistic and cultural identities; to redefine our sense of linguistic integrity and cultural cohesiveness; to navigate and negotiate communicative borders both real and imagined; and to rethink what it means to acquire or abandon a language and, more fundamentally, what it entails to be a productive participant in the sociocultural space that both unites and transcends the country/culture of our birth and that of our choice.
To an individual, an HL may provide valuable personal and familial resources, or it can become a linguistic and cultural liability. There have been substantive debates (Reference FishmanFishman, 1991, Reference Fishman2001; Reference HeHe, 2010; Reference HornbergerHornberger, 2004; Reference McKay and WongMcKay and Wong, 1996; Reference Wong-FillmoreWong-Fillmore, 1991) at social and political as well as cultural and linguistic levels on whether HLs should be maintained and whether the loss of HLs is part of the price to be paid for becoming acculturated into the mainstream society. Some communities and individuals have taken active and proactive measures to ensure that their HL is passed down from one generation to the next, while others have let theirs disappear gradually or almost abruptly.
What are the decisive factors for the success of HL maintenance? What is the role of family language policy and social networks in shaping the trajectory of HL development? Why is it that we often witness a resistance to HL learning when speakers are young but subsequently an embrace of HL after they come of age? What is the relationship between speaker identity (projected as well as perceived, interactional as well as developmental) and the HL learning outcome? How do Chinese-American immigration experience, societal language ideologies, and racial positionings of Chinese-Americans impact Chinese-American households’ language choice? Does a certain language choice in everyday interactions in the classroom and at home lead to the socialization of certain cultural values and norms? Is literacy a necessary requirement and condition for HL development? In what ways may heritage language maintenance, and multilingualism in general, contribute to a more civil and more just society?
Understanding such complex human conditions necessitates a kind of research perspective that is naturalistic, observational, and descriptive; that traces language and life in time and space; that contextualizes the notion of “heritage language” in the histories and contexts of its use; and that fundamentally assumes that language should be examined as both the venue and the vehicle for human development.
In this book, I offer a narrative-ethnographic, quasi-longitudinal, and interactionally enriched account of language development and language change in the context of immigration, where different languages, cultures, races, and ethnicities come into contact and sometimes conflict. I present a set of stories about language and life from a group of Chinese-American immigrants and their children of mainland China origin who first arrived in the U.S. in the period between the mid-1980s and the mid-1990s. These stories will be told in large part through the original voices of the storytellers, who are college-educated parents who came to the U.S. to pursue graduate studies or to follow their graduate student spouses, grandparents who joined the parents in the U.S. on a temporary basis, teachers from the mainstream elementary schools and weekend Chinese language schools, and children in various age groups who were either born in the U.S. or came with their parents as infants or toddlers. These are stories about the quest for identity, dignity, and opportunity and about growth and change across languages, lives, geographies, and generations.
Before I delve into the stories, it is important to establish a common frame of reference for our discussion. Let us begin with our conceptual approach to heritage languages and the methodology used to examine how heritage languages are perceived and practiced.
1.1 Being and Becoming
Immigrant languages, along with indigenous or ancestral ones, have been conventionally referred to as heritage languages (Reference FishmanFishman, 1991) in ecologies in which another language functions as the dominant vernacular and children of immigrants who speak the immigrant or indigenous language have been conventionally referred to as heritage language speakers (Reference CumminsCummins, 2005; Reference HeHe, 2010; Reference Hornberger, Wang, Kagan, Carriera and ChikHornberger and Wang, 2017; Reference Valdés, Peyton, Ranard and McGinnisValdés, 2001; Van Reference Deusen-SchollDeusen-Scholl, 2003; Reference Wiley, Peyton, Ranard and McGinnisWiley, 2001). This term heritage language has also often been used synonymously with community language, home language, native language, and mother tongue. Like other scholars who find the term inadequate or problematic (see Reference Blackledge and CreeseBlackledge and Creese, 2008; Reference García, Zakharia and OtcuGarcía et al., 2012; Reference Mufwene, Filipović and PützMufwene, 2016), I myself am not comfortable with the “past” connotation of the term “heritage.” “Heritage language” makes immigrant languages sound like repositories of old knowledge, practices, and histories. The term limits our imagination of immigrant languages as agents for language and cultural reproduction and as means of seeing the world that is both deeply grounded and forward-looking (Reference HeHe, 2006, Reference He2013a). However, for want of a more commonly agreed-upon alternative, I will keep the term heritage language (HL) throughout the book.
Heritage speakers, typically, have been exposed to the HL since birth and may have used the HL during the initial years in their life and on and off subsequently, but have never developed the full range of phonological, morphological, syntactic, pragmatic, and discourse patterns which will enable them to use it with the scope and sophistication characteristic of and comparable to native speakers’ usage (Reference Benmamoun, Montrul and PolinskyBenmamoun et al., 2013; Reference Montrul and PolinskyMontrul and Polinsky, 2021). Throughout its history shaped by immigration, the U.S. has always been characterized by rich linguistic diversity and significant language shift. Even though immigrant cultures may survive in some form into the third and fourth generations, immigrant languages generally experience rapid attrition, if not loss (Reference Montrul and PolinskyMontrul and Polonsky, 2021; Reference Potowski, Bayley, Cameron and LucasPotowski, 2013; Reference Rumbaut and MasseyRumbaut and Massey, 2014). Most intriguing, given the largely voluntary nature of immigration, the shift from immigrant languages cannot be adequately explicated by external macro societal forces and pressures only; internal individual agency, i.e., the micro-level, evolving, context-specific purposes and significance that speakers attach to their immigrant languages, appears to play a significant role (Reference Avineri, Wiley, Kreeft-Peyton, Christian, Moore and LiuAvineri, 2014; Reference Curdt-Christiansen and HancockCurdt-Christiansen and Hancock, 2014; Reference FaderFader, 2009; Reference He, Kadar and PanHe, 2012; Reference KroskrityKroskrity, 2016; Reference LiLi, 1994; Reference LoLo, 2004; Reference TsuTsu, 2010; Reference ZentellaZentella, 1997). The trajectory of heritage language development thus casts doubt on existing research on language shift which has generally evoked broadly conceived macro-level variables (such as colonization, industrialization, immigration, globalization, urbanization, assimilation, and national-identity formation) that may accompany and correlate with language shift but are yet to sufficiently explain the individual motivations, circumstances, and outcomes tied to language shift in specific contexts, as has long been noted by scholars such as Reference GalGal (1979), Reference KulickKulick (1997), Reference MufweneMufwene (2017), Reference Makihara and SchieffelinMakihara and Schieffelin (2007), and Reference Sankoff, Trudgill and Schilling-EstesSankoff (2001).
The conceptual approach to language shift that undergirds this book centers on specific, situated communicative and cultural practices; it consists of two fundamental perspectives: (1) language is both a medium and a catalyst for sociocultural processes, and (2) these processes take place in the manner of rhizomes.
1.1.1 Both a Medium and a Catalyst
In research on language shift, the interactional and linguistic micro processes have received relatively scant attention. However, it is these situated, speaker- and setting-specific language choice and use that constitute language and life in immigrant communities; they can lead ultimately to the abandonment or attrition of heritage languages.
This view is inspired by and anchored in language socialization (Reference Duranti, Ochs and SchieffelinDuranti et al., 2011; Reference OchsOchs, 1986, Reference Ochs1993; Reference Ochs, Schieffelin, Duff and HornbergerOchs and Schieffelin, 2008; Schieffelin and Reference OchsOchs, 1986, Reference Ochs, Gumperz and Levinson1996), which, as a branch of linguistic anthropology, focuses on the process of becoming a culturally competent member through language use in social activities. As formulated by Ochs and Schieffelin, language socialization is concerned with (1) how novices are socialized to use language and (2) how novices are socialized to be competent members in the target culture through language use. This line of thinking offers a synthesis between cognitivist and sociocultural approaches that allows a reconsideration of cognition as originating in social interaction and shaped by cultural and social processes, not just mental ones (Reference Duranti, Ochs and SchieffelinDuranti et al., 2011; Reference Watson-GegeoWatson-Gegeo, 2004). Language use and language contact thus become resources for growth and change, as they turn everyday experiences (both at home and in school) into potential sites that foster transformative practices and preferences in participants. It is in those moment-by-moment give-and-take situations between the children/students and their parents/teachers that socialization becomes a vivid, lived family and classroom experience. The quintessential and intrinsic sociocultural nature is particularly salient in the case of HLs. The very notion of HL is a sociocultural construct insofar as it is defined in terms of a group of people who speak it. HLs also have a sociocultural function, both as a means of communication and as a way of identifying and transforming sociocultural groups (Reference HeHe, 2010). In other words, in the heritage language home and classroom, socialization takes place both toward and through constantly and locally re-enacted, redefined and reconstructed meanings of “heritage language and culture.”
For some, heritage as a noun may denote something of the past, something that has a fixed, static, essential property. However, languages change over time, and HLs are no exception. To see an HL as fluid and always in flux with respect to evolving social and cultural conditions is to acknowledge its potential as a critical lens which facilitates our understanding of the construction and reconstruction of identities, communities, and cultures in the human diaspora. Because language indexes culture and identity, as culture and identity acquire newer forms, language also transforms in order to acquire new indexicality, hence creating new values. Given that the heritage culture has a complex, developing, transnational, intercultural, cross-linguistic, and hybrid life, an HL can, like any other language, be thought of as an emergent phenomenon that is constantly engaged in deconstruction and reconstruction processes through the ongoing socialization of its speakers, involving all the participants (Reference He, Duranti, Ochs and SchieffelinHe, 2011a). Therefore, instead of reinforcing or endorsing the idea of heritage language and culture as a set of essentialized practices and concepts, I will explore the transformative potential of HLs (Reference HeHe, 2006) and examine how HLs enable speakers to acquire new meanings; to actively (re)construct themselves as members of a particular ethnicity, nationality, speech community, social rank, and profession; and to (re-)create a new set of familiar and familial, ethnolinguistic indexicalities, while at the same time being transformed through the evolving practices of their speakers.
1.1.2 Like the Bamboo
In Chinese culture, bamboo is a symbol of formidable strength, interconnected roots, resilient adaptability, and unstoppable growth. Bamboos are rhizomes. Their horizontal growth trajectory is not linear, nor predictable. And this is where the notion of “rhizome” can provide some conceptual guidance for our understanding of HL development.
Instead of identifying universal, definitive patterns of causality, my goal is to demonstrate that language shift (in this case, heritage language attrition or abandonment) can be brought about by changes in personal, familial, and communal values, allegiances, and context-specific language practices across the life span. These values, allegiances, and practices are based on speakers’ perceptions of themselves and their world, which in turn may be altered/modified as a result of the unfolding social/linguistic engagements in the context of immigration and globalization. And these altered/evolving perceptions may ultimately be responsible for appropriation or attrition of immigrant languages. Hence in addition to advancing a dynamic, socio-constructivist view of HL, I further propose to examine the development of HL as a rhizomatic system (Reference HeHe, 2013a; Reference Tan and CanagarajahTan, 2017). The rhizome resists a vertical hierarchical structure in order to promote lateral relations. It allows for a lateral comparison, not the perpetuation of hierarchical language rivalries held in place by binary opposites. It is a term from biology that has been appropriated by Deleuze and Guattari to emphasize the principles of relationality, connectivity, and heterogeneity (Reference Deleuze and GuattariDeleuze and Guattari, 2004: 7). Unlike the imagery of the tree, which is centralized and hierarchical, the rhizome is a system of multiplicity that spreads multidirectionally. Thought, for Deleuze and Guattari, is a rhizome. It is an underground stem that is neither an origin nor an end (Reference Deleuze and GuattariDeleueze and Guattari, 2004: 12). It is the middle piece that serves as the initiator, sending out roots and shoots that gradually detach from the source, giving rise to new rhizomes, and fostering a system of proliferations. In embracing the rhizome as a conceptual figure, Deleuze and Guattari portray “thought” as lacking a distinct start or finish, residing in the middle, capable of connecting, engendering new thoughts. Because the rhizome is always in process and spreads in multiple directions and forms a network of communications, it violates any systematic mapping of memory and hierarchical structures of knowledge. And language, according to Deleuze and Guattari, is a typical example of the rhizome. Language, as a means of communication to connect people, communities, and societies, involves multiple sets of heterogeneous elements that are interconnected with other heterogeneous entities. Without a fixed center or a linear path, language evolves and spreads in multiple directions. “Language stabilizes around a parish, a bishopric, a capital. It forms a bulb. It evolves by subterranean stems and flows, along river valleys or train tracks; it spreads like a patch of oil” (Reference Deleuze and GuattariDeleuze and Guattari, 2004: 2).
To demonstrate how the rhizome functions as a model for their extrapolation of the philosophy of thought and language, Reference Deleuze and GuattariDeleuze and Guattari (2004: 7–17) introduce six characteristics of the rhizome in A Thousand Plateaus. They include (1) the principle of connections, (2) the heterogeneity of connections, (3) the principle of multiplicity, (4) the principle of signifying rupture, (5) the concept of mapping/cartography, and (6) the concept of tracing/decalcomania. Setting off new shoots and roots in multiple directions, rhizomes are by nature heterogeneous. Deleuze and Guattari’s orientation is compatible with the “multiplicity hypothesis” I developed (Reference HeHe, 2006, Reference He, Duranti, Ochs and Schieffelin2011a) with regard to HL development, in the sense that neither temporally nor spatially is the HL speaker’s existence singular, unitary, or noncontradictory. The HL learner/speaker inhabits multiple worlds and assumes multiple identities that may be overlapping and/or competing. The degree of HL attrition and loss most likely associates negatively with the ease with which the speaker is able to manage the lifelong differences and discontinuities presented by multiple speech roles in multiple, intersecting communities. In other words, greater capacity to adapt and adjust to multiple discourse worlds can lead to greater likelihood of HL maintenance.
To learn and speak an HL means not merely to command the phonetic and lexico-grammatical forms in both speech and writing and to master a static set of discourse rules and norms; it also means to understand or embrace a set of continually evolving norms, preferences and expectations relating linguistic structures to multifaceted, dynamic, and fluid contexts. Heritage language learners’ acquisition of linguistic forms and structures is thus a developmental process of delineating and organizing complex (and often conflicting) contextual dimensions in continually evolving, culturally appropriate, and meaningful ways. An approach drawing upon both Ochs and Schieffelin’s notion of “language socialization” and Deleuze and Guattari’s notion of “rhizome” views language learners as tuned into certain indexical meanings of linguistic forms that link those forms to, for example, the social identities of interlocutors and the related types of social events. This approach relates a learner/speaker’s use and understanding of linguistic forms to dispositions, preferences, beliefs, and bodies of knowledge that organize how information is communicatively packaged and how speech acts are performed within and across explicable, though unpredictable, contexts. In this view, HL learning involves acquiring repertoires of language forms and functions associated with complex and changing contextual dimensions (e.g., evolving and shifting role relationships, identities, acts, and events) over developmental time and across space (Reference HeHe, 2013a, Reference He2016b), which calls for a serial, narrative, ethnographic mode of inquiry. In the remainder of this introductory chapter, I will specify this research methodology before presenting an outline of the book.
1.2 Serial Narrative Ethnography
Many of the big questions in heritage language research concern cause and effect. How does immigration affect one’s identity and language choices? How does a particular language policy (whether at the state level or within the family) affect one’s ability to use the immigrant language? These questions are difficult to answer due to the lack of comparisons. We do not know what would have happened if a specific individual had not immigrated somewhere else or if that specific language policy were not in place. A quick review of existing research on heritage language learners and learning reveals that many empirical studies use tools such as surveys, questionnaires, interviews, and interactions, which are usually cross-sectional in nature. Using them, researchers analyze data of variables collected at one single given point in time across a sample population or a predefined subset in order to measure factors such as heritage language learner motivation, identity, attitude, aspirations, challenges, family language ideologies, and policies. The research participants in a cross-sectional study are typically chosen from an available population of potential relevance to the research question. There is no prospective or retrospective follow-up of participants over time. While these studies successfully capture the participants’ positions and practices at specific moments and effectively establish preliminary bases for more in-depth research, they can be limited by low response rates (surveys and questionnaires), sampling bias (interviews), and snapshot-based transiency (interactional studies). As a result, it is often difficult for researchers to make a causal inference.
While it is generally acknowledged that modern science began with the introduction of experimental research methods, in social sciences and humanities, due to the complexity of human behavior and the human world, this method brings its challenges. For both methodological and moral reasons, we cannot, for example, randomly assign research participants to different groups with respect to parents’ English language proficiencies or family language policies.
The speakers of Chinese as a heritage language (CHL) and their families presented in this book are the ones I have encountered, observed, and investigated in my ethnographic research over the last two decades. Sequencing these stories along the speakers’ developmental path from early childhood to early adulthood, I will present observational, interview, reported, and audio-/video-recorded data that have been collected in settings and situations that CHL speakers experience across their life spans. I call this qualitative approach to research on heritage language development “serial narrative ethnography” (SNE). It integrates methods of narrative analysis with field-based, interaction-enriched methods of linguistic anthropology and draws inspiration from a number of existing methodologies used in the social sciences (detailed below). In SNE, narrative-ethnographic data are collected from the same demographic group at different time points of the population’s development. At each developmental time point, the researcher takes a different sample (different participants) of the target population, generating aggregate data that are not only “thick” but also “long,” thereby enabling the analysis of the demographic group across space and time. The goal is to draw general and generalizable knowledge from detailed, discrete, and sometimes disparate observations and accounts.
1.2.1 Precursors
Before I detail how SNE is operationalized, let me first review several quasi-experimental research methods sensitive to the particularities of social sciences and humanities that have served as precursors to SNE, including natural experiments, ethnographic experiments, repeated/serial cross-sectional studies, and tracing participants across settings and events.
1.2.1.1 Natural Experiments
Historians of science have demonstrated the diverse range of styles of observation and experimentation (Reference GalavottiGalavotti, 2003) that challenged the laboratory as the exclusive domain for knowledge production. It has been shown that it may be possible to address difficult cause–effect questions using natural experiments, which are observational studies in nature that take advantage of the random or seemingly random assignment of research participants to different groups to address specific research questions. It examines cases in which two otherwise similar groups of people have been distinguished by one particular circumstance. David Card, an economist, for example, has analyzed the labor market effects of minimum wages, immigration, and education. His results showed, among other things, that increasing the minimum wage does not necessarily lead to fewer jobs, that the evidence for the claim that immigrants harm native opportunities is slight, and that the fear that post-1965 immigrants will never assimilate is belied by the educational success of their children (Reference CardCard, 2005). Similarly, the COVID-19 pandemic has presented a huge opportunity for natural experiments in medical sciences (Reference Ernst, Niederer, Werner, Czaja, Mikton, Ong and BeutelErnst et al., 2022) and social sciences (Reference Hawdon, Parti and DeardenHawdon et al., 2020).
Natural experiments are often used to study situations in which controlled experimentation is not possible, for practical and ethical reasons, such as when the experience of global immigration or policy changes cannot be assigned to research participants. Data from a natural experiment can be difficult to interpret, however. For example, extending heritage language classes by an hour each week for one group of students (but not another) may not affect everyone in that group in the same way. Some students would have kept studying their heritage languages anyway and, for them, the value of heritage language education is often not representative of the entire group. So is it even possible to draw any conclusions about the effect of an extra hour of class in the heritage language school? Rather than basing work on models that make large assumptions about human behavior, researchers using natural experiments rely only on empirical data that illuminate causal relationships in society. Another economist, Joshua Angrist, uses natural experiments to study the effectiveness of high schools. While it is not possible or ethical to randomly assign students to different schools, if, in the event that a school district line is redrawn, instantly transferring one group of students to a new school, it will create a natural delineation of cause and effect that isolates the schools’ impact (Reference DizikesDizikes, 2013). For my purposes, my research participants’ immigration from China to the U.S. affords a rich opportunity for a natural experiment on language shift, except that, unlike the cases with the economists mentioned above, the data I am looking for are not quantitative or numerical, but qualitative and ethnographic.
1.2.1.2 Ethnographic Experiments
While ethnography as a research approach is most commonly associated with fieldwork and participant observation, the conventional anthropological fieldwork and participant observation of naturally occurring events can be supported and strengthened by intervention and elicitation, whereby the researcher and the research participants collaboratively identify and articulate research questions and jointly produce ethnographic knowledge that is meaningful to both the researcher and the research participants. The idea of ethnographic experiments gained momentum during the “reflexive turn” in anthropology during the 1980s (Reference Clifford and MarcusClifford and Marcus, 1986, cited in Reference Estalella and CriadoEstalella and Criado, 2017; Reference Marcus and FischerMarcus and Fischer, 1986), which prompted a rethinking of construction and representation of knowledge and a critical assessment of the epistemic duality of experimentation and observation. Consequently, observation and experimentation are considered complementary and not contradictory, as evident in both anthropological written genres and the site of fieldwork itself (Reference MarcusMarcus, 2014, cited in Reference Estalella and CriadoEstalella and Criado, 2017). In my work, as specified below, narrativizing the ethnographic data, including reflections, reports, and recordings of interactions from the participants, will be an important form of intervention in the norm and form of representation. In the spirit of ethnographic experiments, I will bring to the foreground a reconsideration of the role of ethnography and the role of the research participants in the production of knowledge concerning HL practices. I will actively engage not only with the participants’ experiences, but also with their expertise while experimenting with new modes of conceptualization through their perspectives and new modes of presentation through their voices.
1.2.1.3 Repeated, Serial Cross-sectional Study
As my work aims to explore development and change over time, I also draw inspiration from a subtype of cross-sectional study, known as the repeated (or serial) cross-sectional study, which has been effectively employed by researchers in social sciences (e.g., Reference Lebo and WeberLebo and Weber, 2015). Whereas longitudinal studies follow the same sample of research participants over time even when participants move location, cross-sectional studies interview a fresh sample of participants each time the studies are carried out. Cross-sectional studies (e.g., surveys) are repeated at regular (weekly, monthly, quarterly, annually, or longer, such as three-yearly) or irregular intervals so that estimates of changes can be made at the aggregate or population level. Examples include monthly labor force surveys, retail trade surveys, television and radio ratings surveys, and political opinion polls. These surveys are designed to give good estimates for the current population and the changes or movements that have occurred since the last survey or previous surveys. In all cases, repeated cross-sectional data are created where a study is administered to a new sample of participants at successive time points. For an annual survey, for example, this means that respondents in a given year will be different participants from those in the previous year (UK Data Service, 2015). The recruitment of different participants over time provided inspiration for inclusion of different participants at different life stages in my study.
1.2.1.4 Tracing Participants across Events and Settings
In applied linguistics, a number of scholars have sought to enhance validity by reconsidering the issue of interweaving longitudinal dimensions in conversation-analytic studies and by following the same participants across multiple speech events within linguistic anthropological research. In a methodological paper that adds a longitudinal dimension to conversation analysis, Reference MarkeeMarkee (2008) proposed empirically based analyses of classroom discourse that occurs over several days and months, which he calls “learning-behavior tracking” in the context of applying conversation analysis to second-language acquisition research. According to Markee, learning-behavior tracking has two components: learning-object tracking and learning-process tracking. The former tracks how participants deploy potential learning objects within a single conversation and in subsequent speech events. The latter focuses on how participants orient to learning objects as resources for developing language-learning behaviors that occur both in the moment and over time. In a similar vein to strengthen validity, Reference GuardadoGuardado (2018) integrates conversation analysis with language socialization as he explores heritage language trajectories and heritage language learner identities. Reference Wortham and ReyesWortham and Reyes (2020) propose that discourse analysis should extend beyond specific speech events to include the development of successive discourses in linked events over time in order to understand not only the specific discursive practices but also the patterns that emerge across speech events. They argue that, by taking the steps of selecting linked events or mapping narrated events, selecting indexicals or relevant cross-event context, construing indexicals or tracing pathways, configuring indexicals, and inferring social action or broader social processes, researchers will be able to apply this method to a range of data, including ethnographic data, archival data, and new digital media data. Following the example of this body of work, we will examine language choice and language use across interactional moments, across speech events, and across life stages.
Drawing examples and inspirations from the multiple sources mentioned above, I will next present an overview of SNE’s epistemological stances and an exploration of its usefulness for HL research before discussing its methodological orientations. Since the E (ethnography) is a relatively established, known, and familiar research perspective (Reference Duranti, Kiesling and PaulsonDuranti, 2005; Reference HymesHymes, 1964, Reference Hymes, Gilmore and Glatthorn1972), I will focus on the N (narrative) and the S (serial).
1.2.2 Narrative Knowing and/as Scientific Understanding
A narrative refers to an account of a series of related events or experiences across a significant span of life that either have taken place (i.e., are true) or the speaker would like to see happen (i.e., are aspirations and hopes). Broadly speaking, narrative analysis refers to a wide spectrum of research traditions that focus on formal, semantic, or interactional structures of narratives (Reference De Beaugrandede Beaugrande, 1982; Reference Labov and WaletzkyLabov, 1972; Reference Labov and WaletzkyLabov and Waletzky, 1967; Reference Polinsky, Brinton, Kagan and BauckusPolinsky, 2017; Reference SchegloffSchegloff, 1997; Reference Van DijkVan Dijk, 1976), or on the ways in which we use narratives to construct meanings such as personhood, identities, relationships, and communities (Reference Daiute and LightfootDaiute and Lightfoot, 2004; Reference HeathHeath, 1983; Reference NilesNiles, 1999; Reference PolanyiPolanyi, 1989; Reference Scollon and ScollonScollon and Scollon, 1981). Here I focus on the latter. I explore a research perspective within the broader field of qualitative research, which uses stories, interviews, autobiography, journals, field notes, letters, conversations, family histories, and stated and observed life experience as bases of analyses to understand the way people make sense of their lives. This type of “narrative inquiry” has been applied in different ways in a wide range of fields, including sociology, anthropology, sociolinguistics, history, literary criticism, philosophy, communication studies, and education research, among others. In these social sciences and humanities disciplines, “narrative” has been both the phenomenon being studied, such as a narrative of immigration and language adjustment, and the method used in the study, as a way of knowing and understanding (Reference BrunerBruner, 1987; Reference PolkinghornePolkinghorne, 1988) as well as a way of humanizing linguistic justice (Reference ThomasThomas, 2022).
The construction of narrative involves postulating relatively speculative causal connections between well-observed, well-established events/phenomena. A narrative framework and scaffold can help us determine what to include as new, relevant events or phenomena. Where formal theories achieve precision and accuracy, narratives obtain richness as they track complex processes (twists, turns, constraints, possibilities, alternative paths, and opportunities taken or forsaken) in a developmental trajectory over time. However, narrative knowing and scientific understanding need not be antithetical to each other. In combination, they both enable us to understand and explain highly complex phenomena such as heritage language development. A robust scientific theory does not substitute for a rich narrative; conversely, a well-grounded narrative may well lead to a robust theory. What is measurable in quantifiable parameters may not be the same as what is meaningful to the participants. Whereas quantitative data at a large scale involve technologies to capture, store, and analyze, and require normalizing, standardizing, defining, and clustering (thus often stripping the data set of contexts, meaning, and stories), thick narrative-ethnographic data have the capacity to map unknown territory, thereby providing insight and inspiration. The two forms of knowing are not mutually exclusive and may well reinforce each other. Ultimately, explanations derived from rich, deep, significant (but seemingly anecdotal) narratives should complement theories generated by highly structured, formal scientific experiments and inferences.
Narratives in this sense, then, are not merely events strung together. They can serve as the medium for presenting evidence, or as a mode for articulating concepts. They are not merely events or phenomena themselves, but become the means of bringing connection and coherence to otherwise disparate, random events and fragmented, chaotic phenomena, and become the path for unfolding or revealing events or phenomena. Such “narrative knowing” (Reference PolkinghornePolkinghorne, 1988) has the potential to integrate theories of heritage language development and to identify or fill in gaps. It enables us to account for phenomena that can only be revealed, or best understood, in rich contexts and from multiple perspectives.
Narrative-ethnographic inquiry as a qualitative method in language development research integrates text-based methods of narrative analysis with field-based methods of ethnography. The term “narrative ethnography” is relatively new (Reference TedlockTedlock, 1991), but the practice has long been employed by researchers of second-language acquisition (see, for example, the use of learner diaries as data, from early work by Reference Bailey, Scarcella and KrashenBailey (1980) and Reference Campbell, Bailey and NunanCampbell (1996) to more recent work by Reference NortonNorton (2000, Reference Norton2013)) and has gained renewed attention in recent years (Reference Barkhuizen, Benson and ChikBarkhuizen et al., 2013; Reference BensonBenson, 2014; Reference De FinaDe Fina, 2021; Reference PavlenkoPavlenko, 2007). In language-learning research, narrative data are typically collected through interviews, diaries and journals, semi-structured interviews, or e-mail exchanges – sources that are likely to provide the researcher with participant-based, subjective experiences and emic perspectives (Reference Barkhuizen, Fina and GeorgakopoulouBarkhuizen, 2015; Reference LonghiniLonghini, 2002). In this context, the narrative approach has been used as a framework to tell the stories of learners regarding their language-learning experiences. Such stories reveal both actual events as experienced by learners and their reactions to these experiences in the process of learning a new language, as well as learners’ aspirations and expectations. As an emic approach, narratives explain how the language learners themselves make sense of their language development (Reference Duff, Anderson, Ilnyckyj, VanGaya, Wang and YatesDuff et al., 2013) and have the power and potential to explain the key constructs in language development such as motivation across the life span (Reference ChoiChoi, 2017; Reference Simon-MaedaSimon-Maeda, 2011) and changes over multiple scales of micro moments, speech events, and longitudinal time (Reference He, Duff and MayHe, 2017).
A narrative approach becomes most productive when the phenomena in question involve complexity, variety, and contingency, and when events or phenomena need to be carefully “storied” in order to make sense (Reference CreswellCreswell, 2013). Heritage language development is one such case. To gain a full picture of HL development, we need to unravel the complex forces that shape the trajectory of acquisition, attrition, or maintenance; the relationship between heritage language and mainstream societal language; the relationship between different participation structures and interactional routines when HL is used; the relationship between different communicative needs, choices, and styles of different generations; and the relationship between different speech roles (e.g., speaker, hearer, eavesdropper, bystander) required for the same participants in different, sometimes simultaneous, sometimes consecutive, discourse contexts. HL research lends itself naturally to a narrative-ethnographic approach.
1.2.3 The Power and Pitfalls of a Single Narrative
Let us consider the linguistic auto-ethnography by Julie Reference ChoiChoi (2017). In this work, Choi examines her own multilingualism (what she terms “multivocality”) across her life span. She was born and raised in New York, living largely within a Korean-American community and speaking Korean as a heritage language at home, English at school, and a mix of Korean and English in the Korean-American community. Despite New York being the place of birth and upbringing and Korea being her ancestral homeland where she has strong familial ties, she did not have a strong sense of identity either as an American or as a Korean. Subsequently, she spent equal numbers of years in Beijing and Tokyo, from which she found a meaningful identity in Japanese but not in Chinese and acquired a stronger command of Japanese through schooling and work than of her heritage language, Korean. The work was completed and presented as “work in progress” (Reference ChoiChoi, 2017: xxi) as the researcher continues to experience and explore a multilingual life.
In order to make sense of her complex and dynamic changes in life and language, Choi combines narrative studies with ethnography and gathered two major types of data – external and personal (Reference ChoiChoi, 2017: 74–75). The external includes diaries, photographs, letters, cards, teachers’ notes, classroom handouts, song lyrics, screenshots of television shows, plots of popular dramas, paintings, and content and images of books read from the different places where she lived for extended periods of time – New York, Seoul, Beijing, Tokyo, Sydney – which corresponded to the major stages of her life from childhood to adolescence, youth, and adulthood. The personal includes memories in the form of reconstructed dialogues, commentaries, reflections, and analyses. She uses the aforementioned data as “texts of personal experience” (Reference ChoiChoi, 2017: xxxiv) to examine “the multiple forces that crisscross particular events, explore what lies at the intersections, and find ways of broadening possibilities for identity development” (Reference ChoiChoi, 2017: xxxv). Her goal is to explore the “multiple meanings, interpretations, histories, social conditions, dimensions of affect and creative energy” (Reference ChoiChoi, 2017: xxix) that have shaped the trajectory of her language choice and language use. Situating Korean, her heritage language, in the context of a whole range of linguistic resources at her disposal (including also English, Chinese, and Japanese) and documenting and reflecting on the responses and reactions to her language and cultural backgrounds from her peers, friends, teachers, family members, and strangers, Choi argues that her language choices (whether HL or non-HL) and her projected and perceived identity options (whether as American, Korean, Japanese, Chinese, or any combination of these) are highly susceptible to the historical, political, social, and cultural stance, the embodiment of which ranges from macro-level ideologies concerning immigration and globalization to micro-level everyday practices such as different pronunciations or spellings of the “same” name in different spaces by different interlocutors.
As Choi’s work powerfully illustrates, the multiple dimensions to HL speakers’ lives are open to multiple interpretations. Consequently, the HL speaker’s life history is highly unpredictable. Unpredictability does not, however, imply inexplicability. The sense of purpose of a story is not determined a priori but emerges as the narrative takes shape and the meaning and purpose of the linguistic repertoire are reshaped. A speaker’s life history involves not only searching for the significance of the HL, but also creating it. A life span narrative approach instates the meaning and significance of HL as both the results and the antecedents of linguistic practice. In this sense, then, HL learner development is understood as a form of narrative, a form of social life, a form of knowledge and knowing, and a form of communication.
Serious challenges, however, arise when we consider the feasibility of collecting comparable data not from our own lives over the years but from research participants to whom we only have access for a relatively short period of time; it would be extremely unlikely that we can closely follow the same research participants throughout multiple life stages for twenty to thirty years. In this case, how do we determine variable patterns over time? How do we establish cause-and-effect relationships? How do we ensure validity in tracking long-term developmental changes? For, as we have argued previously, HL development is a life span process. Furthermore, there is the “danger of a single story,” as Chimamanda Reference AdichieAdichie (2009) puts it. Whether Choi’s auto-ethnography is a story of growth, possibilities, transformation, frustrations, or challenges, it is but one story. We need to hear multiple voices, from multiple tellers and multiple perspectives, in order to create life span narratives which are not limited to a single time, place, or person. The life span narratives through which we make sense of HL development are not researchers’ nor a single research participant’s introspection or imagination but multiple participants’ lived and reflected experiences. Like identity, instead of being reducible to an essence, life span narratives gathered from different participants at serially different developmental stages make it possible for us to find our own mirror and/or metaphor and to vicariously insert ourselves into the lives of others to seek relevance and references.
A point for comparison here is the quasi-longitudinal approach. With a view to addressing developmental paths in second-language acquisition, researchers decided to carry out a comparison of cross-sectional studies of different groups of learners at different developmental stages, thereby adopting what Reference Huat, Hyland, Huat and HandfordHuat (2012: 197) calls a pseudo-longitudinal approach. The learners’ productions are not from the same learners, hence the use of the “pseudo-” prefix, and the “time” variable is measured by proxies such as age or proficiency level. In such pseudo-longitudinal designs, researchers compare several groups of learners at different levels of proficiency. While pseudo-longitudinal designs do not allow for the analysis of individual development, individual variation within each group can, however, be analyzed, as can group development in pseudo-longitudinal designs.
A serial narrative-ethnographic approach considers micro, discursive processes as underlying language choice in immigrant communities that can lead ultimately to the abandonment or attrition of heritage languages. It enables us to situate the HL learning process in a multiple-voiced, development-oriented, context-enriched, humanistic dimension, as we explore the decisive factors for HL development and maintenance in the long run; the reasons behind shifting attitudes, motivations, and identities across speakers and over time; and the role of cultural and interactional practices at home and in the classroom in shaping the HL development trajectory. An SNE approach underscores the intricate, evolving, rhizomic existence of the heritage culture (Reference He, Montrol and PolinskyHe, 2021). Additionally, it positions the heritage language as an unfolding narrative subject to deconstruction and reconstruction by heritage language speakers and their interlocutors. As such, instead of reinforcing or endorsing the idea of heritage language and culture as a set of essentialized practices and concepts, an SNE approach explores the transformative potential of HLs (Reference HeHe, 2006) – in what ways do they facilitate individuals in actively shaping and shifting their identity within a specific ethnic or national group, speech community, social strata, or professional circle, while also crafting a fresh array of recognizable and personal ethnolinguistic markers? Given its merits, SNE can be particularly productive for a context-dependent field such as language change and language shift over time and space. It holds great potential to advance knowledge about heritage language development and learning.
1.3 Putting Perspectives into Practice
Data that form the basis of this book are drawn from three age groups of CHL speakers. Group A includes child CHL speakers (aged four and a half to ten), their family members, and selected teachers. Group B consists of adolescent CHL speakers (aged thirteen to seventeen) their family members, and selected teachers. Group C comprises young adult CHL speakers (aged eighteen to twenty-two) their family members, and selected teachers. Both the collection and the selection of data for presentation were made possible by my own “linguistic space” for research (Reference Ganassin and HolmesGanassin and Holmes, 2020) with Chinese and English bilingual resources, by my own personal (and often parallel) experience raising two U.S.-born children who are speakers of CHL, as well as my positionality as a participant observer.
I met the participants in Group A at different times and in different cities, largely through the weekend Chinese language schools where I started my research and through the communities where I resided. Within the last twenty years, I followed a total of twelve CHL families with children in this age group at different points in time and for different durations (ranging from one to five years). Data from Group A include informal interviews with family members of targeted CHL speakers, audio- and video-recorded class meetings involving two teachers in two different classes (one beginning-level, one advanced-level) in Chinese language schools, informal interviews with CHL speakers’ elementary-school teachers, and audio- and video-recorded daily routine activities in two CHL households.
I gained access to Group B via personal contacts with children in this age group. While my connections with these families remain continuous and ongoing, data from Group B were collected over a period of approximately one calendar year in the households of four CHL-speaking teenagers (two occasional CHL speakers and two frequent speakers) during their everyday conversations as well as during informal interviews when I asked them questions in ways that were compatible with how we normally communicated with each other (Reference BriggsBriggs, 1986). Given my connection with their parents and following the Chinese cultural practice, CHL speakers in this group called me ayi (aunt), with full knowledge that I am a university professor carrying out research on the language and life of speakers like themselves.
Participants in Group C were CHL students I had the privilege of knowing in my capacity as a university professor. Data from this group include personal communications, informal interviews, and audio-recorded CHL language classes (one beginning-level, one advanced-level) involving two instructors and a total of about fifty students.
In all cases, the targeted CHL speakers’ parents received their undergraduate education in China, came to the U.S. during the period from mid-1980s to the mid-1990s to pursue graduate studies (or to follow their spouses), and subsequently settled down in suburban areas. The targeted CHL speakers were either born in the U.S. or came with their parents to the U.S. as infants or toddlers before age three. Throughout the book, the targeted CHL speakers are all referred to as “second-generation immigrants.” In subsequent chapters, the CHL speakers and their family members and teachers will appear in the developmental order of the life span (from preschool to college), which is not necessarily the chronological order in which data were collected. The real names of all the participants were replaced by fictitious ones. The specific places of their homes, communities, schools, and workplaces were anonymized.
The process of engaging in research guided by SNE included the following steps: (1) focusing on a small number of individuals at varying stages of life from early childhood to young adulthood, (2) gathering various forms of data (stories, documents, conversations, interviews, observations, and so on) about the individuals’ lived and told experiences, (3) shaping the data into a chronological or logical order, (4) analyzing the themes of the data and ways in which they are produced (the telling of stories, the manner of conversations, and so on), (5) locating turning points (Reference DenzinDenzin, 1989) in the data, and (6) drawing inferences and significances. It involved a sewing together of participants, events, stories, interactions, conversations, and interviews for purposes of meaning-making and identity construction (Reference Atkinson, Hammersley, Denzin and LincolnAtkinson and Hammersley, 1994; Reference Barkhuizen, Fina and GeorgakopoulouBarkhuizen, 2015; Reference Holstein and GubriumHolstein and Gubrium, 2000; Reference TedlockTedlock, 1991, Reference Tedlock2004). Throughout the process, particular attention was paid to the nature of research questions, the role of theories, the role of the researcher, field methods, the kinds of data, and the ways in which results are presented.
1.3.1 Questions and Theories
The research questions addressed in the book have a dynamic nature that contrasts with the predetermined, fixed nature of those found in studies in the hypothesis-testing paradigm. Whereas researchers conducting correlational or experimental studies spell out their questions and hypotheses in advance and adhere to them throughout, I asked very broad questions at the onset of the study, such as what happens to language and life when people move from one place to another. I then refined and refocused my questions in the field as the study evolved. The reason is that some of the issues that are important only come to light in the field and cannot be predetermined or foreseen. Like all other ethnographic work, this work is hypothesis-generating rather than hypothesis-testing.
Since new and refined/refocused questions emerge as field research proceeds, the theory or theories guiding the study may also shift or expand. Those that are most foundational for the SNE of HL development typically address sociocultural processes and interactional structures, including language socialization, rhizomatic systems, discourse and conversation analysis, language attitudes and motivations, language and identity, and ethical considerations in language-learning research. Part of the goal of doing SNE is to expand the scope suggested by prior theory in order to identify what may be missing or misleading in existing theory and to search for patterns of language use and their significance that are specific to the situation under study (Reference Copland, Creese, Rock and ShawCopland et al., 2015).
1.3.2 The Researcher and the Researched
In the ethnographic tradition of research on language learning/use (Reference Barkhuizen, Benson and ChikBarkhuizen et al., 2013), I played a dual role, both as a keen observer and reflective participant, and as the narrator providing explanations of events/phenomena in natural settings. While I myself served as a key instrument for data collection, the researcher–narrator dual status afforded me space to establish my credibility and to share with the reader the complexity, unpredictability, surprise, or difficulty that the research participants experienced.
To accomplish the goal of discovering and presenting the participant’s view, I relied primarily on participant observation and interview as field methods. As a parent to two CHL-speaking children as well as an instructor to college CHL students, I participated in most activities and interactions at school and in home settings documented and described in this book. My participation gave me access to what actually happened naturally as I collected and made sense of the participants’ narratives. Being a narrative ethnographer, I also went beyond naturalistic observation to intrude on the natural setting and obtained particular information directly from the participants through various interviews, including life history and key informant interviews (e.g., target CHL learners and their family members in the latter case), ranging in style from semi-structured to informal. Given the emic goal of the research, the interviews were conducted in a way that promoted the natural unfolding of the participant’s narratives and their meanings and perspectives.
In terms of the kinds of data, besides observation and interview data, I also gathered a variety of digital materials such as samples of learners’ homework, community demographic data, and audio and video recordings of actual interactions in the home and in the classroom. While the data themselves do not make the study narrative-ethnographic, the fact that I attended to the context in gathering, interpreting, and reporting narratives does. As Reference ChoiChoi (2017: 131) observes, the narratives are not just told; they are “always told within a specific set of conditions and fused by a practice of a certain kind of telling.” The context can be the “culture” of the family, the community, the classroom, or the larger context of immigration history, language ideology, and language policy. The analysis of narrative-ethnographic data is necessarily inductive, comprehensive, recursive, and interpretive, with the end products being descriptions and interpretations of CHL learning and CHL use that are grounded in both micro and macro contexts.
1.3.3 Presentation of Results
A few guidelines are observed in the presentation of results of narrative-ethnographic analyses. First, sequentiality. Narratives would appear to follow chronological sequences. After all, they tell stories over time. However, much is to be said about the kinds of chronology. Since language development is necessarily longitudinal, involving change over time, some narratives are certainly presented in the sequence of real-time events, such as the developmental narratives of the (heritage) language learner from infanthood to adulthood, the big societal narratives of immigration and language contact across generations, or perhaps the micro narratives of turn-by-turn, moment-by-moment interaction over a sustained period of time. But some narratives can order events and phenomena in logical sequence as well: the sequencing of events/phenomena is determined by the logical relationships between them, such as cause–effect, condition–outcome, circumstances/constraints, and so forth. Ethnographic details, for instance, would be one example of narrating the interconnectedness between events or phenomena that is not exactly temporal in nature. Hence, while narratives often present accounts over time, they are not a mere temporal sequence of events or phenomena; rather, they are also organized around a specifiable, nontemporal connection.
Second, explicability. Since narratives present events or phenomena related to one another in a sequence beyond a temporal listing thereof, they have the power and/or potential not merely to document and describe but also to explain them. Narratives give shape, structure, and significance to the seemingly fragmented events or phenomena and make sense of them in scientific terms. For example, what sets of conditions collectively create the likelihood of HL maintenance in the household? Is the family language policy sufficient to give a credible account of language dynamics? Or do we need also to take into consideration the societal attitudes to the specific HL in order to have a narrative that fully explains the variable outcomes of HL? Narratives attempt to show how things fit together, how they relate to one another, what effect some cause might have, what cause might have produced this effect.
Third, coherence. As narratives strive to establish relations between events or phenomena from various sources and settings that are within the scope of what is to be described or explained, they must also meet the criterion of “coherence.” The narrativizing process may involve different pieces of theories or different conceptual elements. It is a process of coherence making, of showing how disparate elements interrelate, in order to make an account that is coherent in itself, thereby enabling integration and synthesis. This process of narrativization may serve to identify gaps in evidence that might then be filled through the search for further, additional pieces of evidence. The piecing together may also serve to delineate where and why the different accounts and different elements fit or do not fit together. Constructing narratives thus becomes a natural conduit for bringing related, relevant elements into order and for making meaning out of otherwise fragmented events and experiences.
Fourth and finally, perspective. A narrative perspective refers to a set of features shaping a story. A clear sense of perspective gives meaning and significance to narratives and narrativizing. As I present the narrative accounts from the participants, I will draw the reader’s attention to what is being told, what information is included and what is left out, who the storyteller is, who the story is about, from whose point of view the story is told, what are the broad sociopolitical discourses, the local contexts of participants’ lives, and the immediate situational contexts of storytelling.
1.4 The Plan of the Book
The chapters of this book are organized in the developmental order of the CHL speakers’ life stages. Each chapter is written from the viewpoint of some specific participant(s), either someone who is central to the communicative network of a CHL speaker at a given time or the CHL speakers themselves. In an effort to give authentic voice to the participants themselves, the chapters will present ample transcribed and often translated (from Chinese to English) excerpts from recorded interviews, reconstructed conversations from spontaneous informal exchanges, field notes from classrooms and households, and extracts from recorded and transcribed interactions taking place in English, Chinese, or a mixture of both in school and home settings.
All interviews with the first-generation immigrants – parents and grandparents – took place in Mandarin Chinese. Grandparents spoke in Chinese only; their speech was subsequently translated by me into English; in order to keep the size of the book manageable, only the English translation is presented in the book. Parents mostly spoke in Chinese in interviews, occasionally inserting some expressions in English; their speech was translated by me into English as well. Every effort was made in the translation process to remain faithful to the original meaning of the utterances and the original context of the interviews while making sure that the translated texts are readable and accessible to the English reader. No deliberate effort was made to either minimize or amplify the “foreignness” of the original speech (cf. domestication versus foreignization (Reference VenutiVenuti, 1998)).
Chapters 2 and 3 chronicle one family’s journey from China to the U.S. through their own voices. This family has two children, Andrew (aged nine) and Angela (aged seven). Chapter 2 uses Andrew and Angela’s mother’s narrative as a basis to outline the language background of the family and to explore the construction and the characteristics of the bilingual space at home, as well as the history and the forces behind the shift of linguistic repertoire in the household during CHL children’s early years. Chapter 3 presents Andrew and Angela’s maternal grandmother’s perspective on child rearing and home culture. It examines the family’s use of both Chinese and English in a range of routinely occurring speech events.
Chapters 4 and 5 highlight the difficult choices that CHL communities need to make in terms of family language policy and Chinese language school curriculum. Spotlighted are two CHL children, Ben (aged six) and Chris (aged ten), and their respective families. Chapter 4 depicts Ben’s kindergarten classroom and its multicultural and multilingual ideologies and affordances (or lack thereof) from Ben’s schoolteacher’s viewpoint. It further explores Ben’s family language ideology and family language choice as a result of Ben’s schooling, as well as Ben’s parents’ own language experience in the U.S. Chapter 5 takes the reader to the community-based weekend Chinese language school that Chris attends. Drawing upon reflections from Chris’s Chinese language teacher, it delineates the historical complexity of Chinese language traits and cultural values, as well as the challenges in choosing what to impart to CHL children and how to instill a cultural ethos which may be divergent from mainstream culture.
Chapters 6 and 7 center on the views, experiences, and language practices of second-generation Chinese-Americans who are coming of age. Chapter 6 portrays a pair of best friends, Diana (aged fourteen) and Emily (aged fifteen) and their respective families. It situates Diana and Emily’s attitude toward CHL in the contexts of talking about their respective families in terms of values, behavioral patterns, and accents; talking for their families as they interpret and translate from Chinese to English for their parents and polish their parents’ English in everyday social encounters; and talking with their families in digital communication across three generations. Chapter 7 follows a young adult college student, Felix (aged twenty) and his girlfriend Christine (aged twenty-two) as they explore language, life, and race relations during the COVID-19 pandemic, trying to transform the very CHL community they grew up in and transcend the cultural identity that is assigned to them by society.
Chapter 8 concludes the book with a recapitulation of how the serial narrative-ethnographic method sheds light on theories of language shift and cultural adaptation in general and a consideration of some alternative lenses through which to conceptualize the HL linguistic repertoire, as well as a dialectical, dialogical, and ecological take on language shift. It ends with projections regarding the trajectory of the CHL repertoire and a call for shifts from a focus on discrete heritage languages to heritage linguistic repertoires and from static to dynamic views of diaspora with social justice and multilingualism at the core.
