DEFINITIONS

Complicating the investigation of HLE policy are the multiple definitions of the term heritage language that operate in the literature. Even when restricted to the research included in this review, we encounter a formidable list of terms often positioned as synonymous with heritage: aboriginal, ancestral, autochthonous, (ex-)colonial, community, critical, diasporic, endoglossic, ethnic, foreign, geopolitical, home, immigrant, indigenous, language other than English, local, migrant, minority, mother tongue, refugee, regional, and strategic.Footnote ¹ Certainly, no single term can (or perhaps should) encompass the widely divergent histories, contexts, and perspectives that such a list indexes. In their discussion of the applicability of heritage language to the African context, Brutt-Griffler and Makoni (Reference Brutt-Griffler and Makoni2005) identified two inherent risks in labeling such a wide array of language and language contexts: Either a term misconstrues the specific local dynamics at play, or it is so broad that it loses any meaning. The work of Wiley (Reference Wiley2005) reminds us that at the heart of this terminological debate is the ethical tension between researcher and researched; that is, the terms that applied linguists assign to languages are often not in concert with how speakers of those languages identify or perceive their own language(s).

In the United States, heritage language can refer to all languages other than English (Hornberger & Wang, Reference Hornberger, Wang, Brinton, Kagan and Bauckus2008). Fishman (Reference Fishman, Peyton, Ranard and McGinnis2001, Reference Fishman, Valdés, Fishman, Chávez and Pérez2006) categorized heritage languages in the United States as indigenous, colonial, or immigrant languages. The primary example of this typology is Spanish: It was introduced to the western hemisphere as a colonial language; thereafter, it was indigenous to territories annexed by the United States in the 19th century; and of course it constitutes the largest immigrant language in the country. Cummins (Reference Cummins2005) and Wiley (Reference Wiley2005) added to this typology the category of refugee languages as well.

However broadly heritage language is defined, the term has evoked a number of criticisms. The most persistent is that the term implies, even romanticizes, past language use, rather than focusing on actual and future language practice (e.g., García, Reference García2005; McCarty, Reference McCarty2008; Wiley, Reference Wiley2005). García (Reference García2005) juxtaposed the emergence of HLE as a field of study with the constriction of social space for the practice of bilingual education. García argued:

Based on an international comparison of terms used to describe minority languages, Wiley (Reference Wiley2005) suggested that heritage and community languages more accurately reflect contemporary language practice, while McCarty (Reference McCarty2008) preferred heritage mother tongue to refer to indigenous languages.

The implications of labeling are even more acute in deliberations over who counts as a heritage language speaker or learner. In general, the debate has centered on the question of proficiency in the language. One approach defines heritage language speakers in terms of their individual or collective affiliation with a specific ethnolinguistic group (Fishman, Reference Fishman, Peyton, Ranard and McGinnis2001; McCarty, Reference McCarty2008). Another defines heritage language speakers in terms of their ability to speak or understand the language, no matter how limited that proficiency may be (Valdés, Reference Valdés, Peyton, Ranard and McGinnis2001). One example of how this distinction operates is to consider the example of an African American student in the United States in an introductory-level course in Hausa at university. An identity-based definition of heritage language learner might include such a student, assuming her studies were part of the student's efforts to rediscover and cultivate ancestral ties to western Africa. Assuming this student was born in the United States and raised with English only, a proficiency-based definition of heritage language learner would exclude the student for having no experience in the language at all.

The implications of this distinction are not insignificant. Certainly, a proficiency-based definition of heritage language learner has more immediate pedagogical consequences for the language education classroom. In fact, the vast majority of the studies reviewed here tacitly or explicitly adopted a proficiency-based definition of heritage language vis-à-vis HLE policy. Conversely, insisting on proficiency in defining heritage language speakers risks excluding those individuals and communities who maintain salient ethnolinguistic identities, especially when the ancestral language(s) experiences language shift or even extinction. Hornberger (Reference Hornberger2005) considered this position in terms of agency, that heritage learners decide for themselves whether or not they belong to a given heritage language community. This article does not take a position in this debate, other than to recognize, as mentioned earlier, that most of the HLE policy research reviewed implied or adopted a proficiency-based definition of heritage language speaker or learner.

The final definitional consideration of heritage language concerns the context in which such languages are used and taught. Thus far, I have used the United States as the touchstone for elaborating terminological definitions and debates. However, the origin of the term has been ascribed to Canada (Hornberger & Wang, Reference Hornberger, Wang, Brinton, Kagan and Bauckus2008; Wiley, Reference Wiley2005). In a recent overview of HLE in Canada, Duff (Reference Duff, Brinton, Kagan and Bauckus2008) stated that heritage in fact refers to a category of languages distinct from Canada's official languages (i.e., French and English) and aboriginal languages. In both Australia and the United Kingdom, the term community language is used more frequently (Wiley, Reference Wiley2005), although not without its own problems. Mercurio and Scarino (Reference Mercurio and Scarino2005) challenged the breadth of the term and whether it indeed fits disparate sociolinguistic profiles in Australia. They also recognized that speakers of Aboriginal languages in Australia tend to prefer terms such as Australian indigenous languages or Australian languages.

Beyond these contexts, uses of the term heritage in reference to language tend to be in the colloquial sense of the word, rather than in relation to a defined body of applied linguistic research or to a model of language education. Instead, a number of labels are ascribed to languages to distinguish between dominant, ex-colonial, official, and national languages on the one hand, and minoritized,Footnote ² indeed perhaps threatened, indigenous, diasporic, regional, and/or immigrant languages on the other. In the African context, for example, Brutt-Griffler and Makoni (Reference Brutt-Griffler and Makoni2005) stated that terms such as home, ancestral, and indigenous are more frequently employed to refer to languages other than colonial languages. De Bot and Gorter (Reference de Bot and Gorter2005) maintained that heritage language has little currency in Europe, where instead reference is made to regional and immigrant minority languages.Footnote ³ In addition, mother tongue is a term widely used in the sub-Saharan African, South Asian, and Southeast Asian contexts. It is also worth noting that much literature simply does not engage the terminological complexities, referring instead only to the languages in question by name. This is particularly true in HLE policy research on Irish, Scottish Gaelic, and Welsh in the Republic of Ireland and the United Kingdom.

Following Hornberger and Wang (Reference Hornberger, Wang, Brinton, Kagan and Bauckus2008), I adopt a broad definition of heritage languages to review recent HLE policy research. Their definition is based on an ecological approach to HLE that considers an individual's ancestral ties to the heritage language; individual agency in identifying as a heritage language learner or speaker; and how the heritage language and its speakers are positioned in social, political, and economic terms. Specifically, this review includes research that investigates HLE policy in contexts where English is the dominant, official, national, or ex-colonial language. The decision to limit the research to English-dominant contexts was driven solely by the logistical constraints of a single article; similar reviews in contexts with other dominant and ex-colonial languages are certainly needed. To be clear, my intention here is not to insist that such a broad conceptualization of heritage language define the field. Instead, because the purpose of research reviews is to synthesize a broad base of literature, we stand to gain by comparing the contexts, no matter how varied, in which HLE policies unfold.

APPROACHES TO ANALYSIS

This article makes two assumptions in discussing recent HLE policy research. The first recalls the onion metaphor proposed by Ricento and Hornberger (Reference Ricento and Hornberger1996). This metaphor understands language policy not simply as legislative processes or formal policy texts operating at the national or supranational level. Rather, language policy at this formalized level is merely the outer layer of a much more complicated process in which policy actors not only implement a given policy but also appropriate it in ways that make the policy their own. This appropriation process assumes policy to be less about formal documents established by governing authorities, instead focusing of policy as social practice (Levinson & Sutton, Reference Levinson, Sutton, Sutton and Levinson2001). The inner layers of the onion include institutional and interpersonal levels of policy formation and appropriation. Ricento and Hornberger identified many institutions in which language policy is made. However, for the purposes of this review, the focus is limited to educational institutions, for example, government-sponsored schools, community language programs, and nongovernmental organizations. Interpersonal policy processes refer to teachers, their interactions with students, and practices that reflect policymaking in a more localized context.

These layers of the onion are not merely markers of where language policy is made and contested but also suggest that the notion of policy has expanded greatly in the field of language planning and policy (LPP). A full discussion of such is beyond the scope of this article (see Ricento, Reference Ricento2006, for a comprehensive overview). However, the HLE policy research reviewed here includes studies of policy as formal texts, as discourse, as ideology, as a process of legitimacy, and less frequently as specific educator and student practices of language education. These varied policy domains also reflect that methods of HLE policy scholarship have expanded far beyond content or discourse analysis of formal texts to include quantitative sociolinguistic analyses, historiography, conventional qualitative research methods, and full-scale ethnographies of the policy appropriation process. As much as possible, discussion of the research reviewed the following sections identifies both the level of and the approach to HLE policy analysis.

The second organizational principle for this review is geographical. As mentioned earlier, this review encompasses HLE policy research in English-dominant contexts, or where English is the ex-colonial language. To honor the differences in how non-English languages are labeled across these contexts, and to acknowledge that there does seem to be consensus within specific regions as to how label non-English languages, the review is organized by geographical region and country. The discussion includes Africa (Botswana, Kenya, Malawi, Nigeria, Uganda, and Zimbabwe); Hong Kong, Malaysia and Singapore; India; North America (Canada and the United States); Oceania (Aotearoa/New Zealand and Australia); the Republic of Ireland, and the United Kingdom. With respect to terminology, the article will use HLE policy throughout, but to refer to specific languages, it will adopt the term employed by the study at hand.

THE AFRICAN CONTEXT

Bamgbose (Reference Bamgbose2004) and Djité (Reference Djité2008, especially chap. 2) presented two comprehensive overviews of language-in-education policy in postcolonial Africa, including the role that heritage languages play in each country. They reported a language-in-education scheme, fairly consistent across Africa, which implements a 1953 UNESCO policy statement in which an African language is the medium of instruction at the primary level, generally in grades 1 to 3. Thereafter, English becomes the medium, and the African language(s) is taught as a subject. In general, HLE policy in the African context tends to reflect conflicts over which language(s) should be the media of instruction in formal, government-sponsored schooling; which should be taught as subjects; and which should be taught or maintained elsewhere in society. This is an instance in which the research assumes a proficiency-based definition of heritage language (HL) learners: the issue is not so much language maintenance (i.e., using language-in-education policy to revitalize a heritage language) as it is about aligning linguistic practices at school with those at home and in the community.

The work of Kamwendo and Mooko (Reference Kamwendo and Mooko2006) is fairly representative of much of the HLE policy research included in this discussion in terms of the structure of their analysis. After situating contemporary language policies against the history of colonial language practices in Botswana and Malawi, Kamwendo and Mooko conducted a comparative review of HLE policy at the national level in each country, identifying divergences in how indigenous medium-of-instruction policies have been legislated. They found that the dominance of English in official domains beyond education has been a major negative factor impacting HLE policies governing indigenous media of instruction in school. In an exhaustive overview of historical and contemporary language planning in Nigeria, Adegbija (Reference Adegbija2004) traced the history of three HLE policies initiated in the 1970s. The author enumerated a series of factors that have inhibited their success: lack of coordinated planning between national and local authorities, lack of suitable curricular materials in the indigenous language(s), lack of educators trained to teach the language(s), and no systematic assessment of the programs or student achievement. Similar inhibiting factors were reported in research in the contemporary Nigerian context (Iyamu & Aduwa Ogiegbaen, Reference Iyamu and Ogiegbaen2007), as well as in Kenya (Musau, Reference Musau2003).

More often, however, HLE policy research in the African context explains the challenges facing indigenous medium-of-instruction policies in historical, sociopolitical, or ideological terms. For example, in a study focused on supranational levels of policymaking, Omoniyi (Reference Omoniyi2003) enumerated a series of political, economic, and cultural forces at work across Africa and globally that have functioned to negatively impact the realization of HLE policies. Omoniyi tied these global forces to an analysis of how they have prevented the acquisition of indigenous literacy and argued, “The liberation of the continent from mass ignorance will require that the indigenous languages be equipped to convey the bodies of knowledge that exist in modes and media that are presently inaccessible to a vast majority of people” (p. 148). Similarly, Matiki (Reference Matiki2006) argued that HLE policies regarding indigenous languages taught in the primary grades in Malawi should be assessed in terms of the extent to which they foster literacy and socioeconomic development of the country. However, he drew a different conclusion than Omoniyi. Especially considering that Malawian HLE policy has begun to incorporate additional indigenous languages beyond Chichewa, Matiki questioned whether instruction in these languages would indeed lead to successful literacy and socioeconomic development given Malawi's linguistic diversity. He justified this question in relation to Malawian attitudes: “While Malawians accept that literacy empowers, they contend that it is only literacy in English that pays” (p. 251).

In fact, policy stakeholders’ attitudes and ideologies were cited most often in discussions of the challenges in realizing HLE policy for African languages as media of instruction. Musau (Reference Musau2003) was unique in exploring these attitudes at the national level. Adopting a framework of linguistic human rights, Musau reviewed general language policy in Kenya, as well as HLE policies in Kenyan schools. He stated, “Negative attitudes toward indigenous languages by the African élite are also a major obstacle in the implementation of linguistic justice” (p. 162) in the form of indigenous medium-of-instruction policies in Kenyan schools. He specified that these negative attitudes have framed indigenous medium-of-instruction programs as too expensive to implement or consider the languages themselves as not scientific, international, or neutral enough to be of valid use. Most HLE policy research on attitudes and ideologies of policy stakeholders, however, focuses on the institutional or interpersonal level. For example, Iyamu and Aduwa Ogiegbaen (Reference Iyamu and Ogiegbaen2007) surveyed 1,000 literate parents and 1,500 primary school teachers in Nigeria about their attitude toward mother tongue policies for Hausa-, Yoruba- and Igbo-medium primary school programs. They found that parents and teachers indeed understood the value of mother tongue instruction in terms of promoting cultural heritage. However, parents reported more frequently than teachers that mother tongue education would not lead to improved learning overall. The authors concluded, “While mother-tongue medium policy is viewed as a major revolution in the history of education in Nigeria, opinions remain divided on its wisdom” (p. 99). Tembe and Norton (Reference Tembe and Norton2008) explored attitudes among community stakeholders in rural and urban Uganda with respect to a 1992 HLE policy. The policy calls for the medium in grades 1 through 4 to be a relevant local language, switching to English from grade 5 and adding Kiswahili as a subject from grade 4. The authors conducted interviews and focus groups to assess the extent of (dis)agreement among rural and urban community members over the policy. They found that although members in both communities were aware of the policy shift, in general they expressed a desire for students to learn English as early as possible. In fact, some rural participants found the policy a “regressive step toward the past” (p. 50).

Importantly, recent HLE policy research in Africa highlights that power imbalances, whether perceived or lived, implicate African languages, not just ex-colonial languages, as the dominant, threatening language. While the focus of analysis in Makoni, Dube, and Mashiri (Reference Makoni, Dube and Mashiri2006) is much broader than HLE policy as defined in this article, their overview of language planning in Zimbabwe is the most comprehensive in interrogating the indigenous-colonial language binary that frames much HLE policy analysis in the African context. Their exhaustive monograph documented the extent to which colonization and missionary work led to the artificial demarcation of one African language from another, and how this process privileged some newly “created” Zimbabwean languages, such as Shona and Ndebele, over others. Makoni, Makoni, and Nyika (Reference Makoni, Makoni and Nyika2008) applied this type of analysis in their description of successful bottom-up language planning among the Tonga in Zimbabwe. The study analyzed efforts by Tonga leaders to reposition their language not as a minority but as an indigenous language in Zimbabwe, and the impact this reclassification would have on medium-of-instruction practices in local schools. The authors reported that Tonga concerns over the status of their language in Zimbabwe relate less to its status against English as against Shona and Ndebele. Igboanusi and Peter (Reference Igboanusi and Peter2004) analyzed a similar dynamic in Nigeria. As with Makoni et al. (Reference Makoni, Dube and Mashiri2006), their analysis is a broader sociolinguistic analysis: The study reported survey data collected among northern and southern minorities and their attitudes toward Hausa, the dominant national language in northern Nigeria, and English. Their data indicated that participants were as concerned with Hausa as with English as the cause of minority language loss.

The final note about HLE policy research in Africa concerns the prominence of nongovernmental organizations in the promotion of indigenous medium-of-instruction policies. Kamwendo (Reference Kamwendo2005) recounted the role played by the Chitumbuka Language and Culture Association (CLACA) in raising the status of Chitumbuka in northern Malawi. Although the analysis of CLACA's advocacy goes beyond school-based HLE policy advocacy, Kamwendo identified the group's efforts to standardize Chitumbuka orthography and create textbooks as an essential step in implementing the language as the medium of instruction in northern Malawi schools. The study of Tonga as medium of instruction in Zimbabwe schools (Makoni et al., Reference Makoni, Makoni and Nyika2008) focused on the critical role played by the Tonga Language and Cultural Association. The organization focused its advocacy to change the status of Tonga from a minority to an indigenous language. This had the effect of positioning Tonga for use as medium of instruction in primary grades rather than simply a subject to be studied alongside a national language and English. Both studies labeled this sort of advocacy bottom-up language planning; however, both describe the language and cultural associations involved as representing the elite and/or elder members of each community. The elite status ascribed to each organization calls into question the researchers’ classification of bottom-up, that is, whether this status obtains in relation to national institutions and authorities, or to the local community.

HONG KONG, MALAYSIA, AND SINGAPORE

A considerable amount of language policy research in recent years has addressed Southeast Asian contexts in which English is dominant or an ex-colonial language. Recent research on Malaysian language and language-in-education policy has tended to focus on the ideological and practical consequences of a 2002 policy mandating English as medium of instruction for science and math instruction at all levels (cf. David & Govindasamy, Reference David, Govindasamy, Tsui and Tollefson2007; Gill, Reference Gill2005; Martin, Reference Martin, Lin and Martin2005). In Singapore, although medium-of-instruction policy is officially labeled as bilingual, it in fact centers on English with one of three official mother tongues (e.g., Malay, Mandarin, or Tamil) taught in addition as a subject. However, only in the case of Malay students does the official mother tongue policy correspond to the home language. In other words, Mandarin is not the home language for the majority of Chinese students in Singapore, nor is Tamil the home language for the majority of Indian students. In fact, recent language policy research in Singapore has tended to focus on initiatives such as the Speak Mandarin and Speak Good English campaigns (cf. Chew, Reference Chew, Tsui and Tollefson2007; Dixon, Reference Dixon2009; Rubdy, Reference Rubdy, Lin and Martin2005). In both cases, because the focus of the policy research is more on English than the heritage language per se, this body of research falls somewhat outside the purview of this article.

Recent research about Hong Kong HLE policy has focused on various aspects of the region's formal language policy of trilingualism (in Cantonese, English, and Mandarin or Putonghua) and biliteracy (in Chinese and English). This policy emerged after the British “transfer” of Hong Kong to China in 1997, at which point Hong Kong was declared a Special Administrative Region (SAR). Hopkins (Reference Hopkins2006) reported 2001 census data citing 7 million residents of Hong Kong, 95% of whom were ethnically Southern Chinese and who claimed to use only Cantonese on a daily basis. In 1998 the government implemented a mother tongue education policy, which mandated that hitherto English-medium schools must switch to Cantonese as medium unless they could meet three criteria: 85% of their students must be in the top quintile academically; teachers must demonstrate sufficient proficiency in English; and the school must demonstrate that it has structures in place to support students struggling with English (Tse, Shum, Ki, & Chan, Reference Tse, Shum, Ki and Chan2007; Tsui, Reference Tsui, Tsui and Tollefson2007). Tse et al. (Reference Tse, Shum, Ki and Chan2007) reported the impact of this HLE policy for mother tongue education: “Strict enforcement of this policy caused more than 300 secondary schools to switch from English to Chinese as the teaching medium. Out of the 430 secondary schools, only 113 were granted permission to teach lessons in English” (p. 138).

Two empirical studies have explored this policy dynamic in terms of teacher and student attitudes toward each language. The first study presented results from a survey of Hong Kong educators regarding HLE policy for Cantonese- and English-medium instruction (Tse et al., Reference Tse, Shum, Ki and Chan2007). Of Hong Kong's 430 government-supported schools, 334 participated, including 284 school administrators and 2,880 teachers. The response rate from Cantonese- and English-medium schools was the same, at 73%. Participants in both contexts reported that a school's medium of instruction should be determined by students’ ability and “linguistic potential” (p. 153). In addition, participants from both contexts were in favor of a policy in which the medium depended on the subject of instruction, not a schoolwide fixed policy. There was a discrepancy, however, in support for Cantonese-medium secondary programs. Participants in Cantonese-medium schools supported consistent Cantonese-medium schooling, while participants in English-medium schools supported a switch to English from the secondary level on. The researchers concluded that in order for HLE policies regarding medium of instruction to be effective, policymakers could not take a one-size-fits-all approach for all schools and contexts. The second study explored similar issues, but from a student perspective (Lai, Reference Lai2005). Within the framework of instrumental versus integrative attitudes toward language study and use, Lai administered surveys to 1,048 secondary students (i.e., 15–17 years old). Respondents indicated in general positive instrumental attitudes to all three languages (i.e. Cantonese, Putonghua, and English). However, in terms of integrative attitudes, Lai stated that respondents “remain emotionally rather detached” from native speakers of Putonghua (p. 376). Based on the findings, Lai called into question the feasibility of official language policies aimed at trilingualism.

Other examples of recent research on HLE policy in Hong Kong consider the issue from ideological and critical-historical perspectives. Tsui (Reference Tsui, Tsui and Tollefson2007) situated Hong Kong HLE policy regarding medium of instruction in the general project of “resinicization” (p. 133), in which Chinese and Hong Kong officials have actively sought to recreate historical and cultural ties between China and Hong Kong, and to create a “Chinese cultural heritage and national identity [which] was perceived as lacking in Hong Kong. This identity was referred to as a new identity that Hong Kong people should take pride in” (p. 134, emphasis in original). Despite these overt efforts, HLE policy research has consistently found that the 1998 medium of instruction policy has not affected social stratification or the hegemony of English in Hong Kong prior to SAR status. For example, Tsui (Reference Tsui, Tsui and Tollefson2007) concluded that “the [Hong Kong] community's apprehension about losing the international flavor and competitive edge . . . because of reunification with China resulted in stronger than ever demands for more English and more English education” (p. 138). Lin (Reference Lin, Lin and Martin2005) reviewed a number of critical ethnographic studies of Hong Kong language policies from the late 1990s that have identified a continued stratification between working-class Cantonese speaking students, for whom “English remains something beyond their reach” (p. 50) and an English-speaking elite. She concluded that:

INDIA

As with much of the literature on Africa, recent research on HLE policies in India has focused either on the mismatch between monolingual assumptions behind medium-of-instruction policy in a profoundly multilingual context, or on the divergent, contested attitudes toward the role of English in Indian schools. Many examples of Indian HLE policy analysis begin with historical overviews of language and education predating and during British colonization, as well as with discussion of the legal background to contemporary language policy (Agnihotri, Reference Agnihotri, Tsui and Tollefson2007; Annamalai, Reference Annamalai, Lin and Martin2005; Khubchandani, Reference Khubchandani2003; A. K. Mohanty, Reference Mohanty, García, Skutnabb-Kangas and Torres-Guzmán2006; P. Mohanty, Reference Mohanty2002). Currently, there are 22 constitutionally recognized languages in India. Moreover, the 1,652 languages (plus varieties of these languages) spoken in India have been classified into 300–400 languages belonging to five language families (A. K. Mohanty, Reference Mohanty, García, Skutnabb-Kangas and Torres-Guzmán2006). Of these languages, 47 are used as media of instruction across India (Vaish, Reference Vaish2005), although A. K. Mohanty (Reference Mohanty, García, Skutnabb-Kangas and Torres-Guzmán2006) reported that this number has declined, especially the number of minority languages (i.e., languages other than Hindi or the regional language of the respective state). This linguistic landscape maps onto official HLE policy in India known as the “three languages formula,” which calls for a regional mother tongue as medium of instruction for the first 5 years of school; Hindi (in non-Hindi states) or another Indian language (in Hindi states) to be taught as a school subject in lower secondary years; and English as a third language, introduced from the third year. In addition, where minority languages are spoken by 10% of the population, those languages may be used in lieu of the regional language as medium of instruction in the early primary years (Agnihotri, Reference Agnihotri, Tsui and Tollefson2007; A. K. Mohanty, Reference Mohanty, García, Skutnabb-Kangas and Torres-Guzmán2006).

A major focus of recent HLE policy research in India has been on the gap between official policies, such as the three languages formula, and actual practice in Indian schools. Descriptions of this gap have led to critical reexaminations of the usefulness of medium of instruction policies. For example, Khubchandani (Reference Khubchandani2003) described a tripartite model of language practice in Indian schools based on the following distinctions: (a) “passive and active media” contexts, in which students receive instruction in one language while writing and responding in another; (b) “formal and informal media” contexts, in which formal teaching occurs in one language while informal explanations use another; and (c) “multi-tier media” contexts, in which different languages are taught in a succession from one grade or grade band to the next (p. 244). Khubchandani ascribed this gap between policy and practice in part to the imposition of standardized urban varieties of Indian languages that do not match the varieties or languages spoken in students’ homes and communities. Agnihotri (Reference Agnihotri, Tsui and Tollefson2007) situated this gap between policy and practice in the historical context in which postcolonial Indian language policy was conceived. He argued that HLE policies regarding media of instruction in India have imposed monolingual categories onto a fluid multilingual context:

An additional layer to Indian HLE policy consists of the role of English in education. As was the case in the literature on African HLE policies, recent research on India has taken various, indeed, competing perspectives on English and the extent to which it threatens Indian languages or constitutes a tool of postcolonial empowerment. These divergent perspectives are tied somewhat to differences in analytical framework adopted in each study. Annamalai (Reference Annamalai, Lin and Martin2005) considered HLE policy and English in India in terms of national development. He did not pit Indian languages against English, arguing that neither is necessarily useful for engendering social or political cohesion. Rather, he insisted that the metric for assessing medium-of-instruction policies should be the degree to which they allow for decolonization of Indian education that combines “inculcation of self-pride with self-criticism, of cultural rootedness with cosmopolitanism, of modernization with tradition” (p. 36). Such decolonized education is the key for Annamalai in national development and progress. Vaish (Reference Vaish2005) is the most critical of applied linguistic research that conceives of English as an oppressing language. The author is the only among those reviewed here to ground the discussion on ethnographic work in schools investigating language practice. Vaish marshaled that research to argue that English is in fact used by the poorest and most disenfranchised students in Delhi as a means of upward mobility. A. K. Mohanty (Reference Mohanty, García, Skutnabb-Kangas and Torres-Guzmán2006) was also concerned with alleviating inequality but took a more critical position toward English. He stated that English is the most pervasive medium of instruction in the country, but that disparities in access and resources, both linguistic and material, have “provided the basis for a new social stratification of the literate groups” that places privileged students educated in expensive private English-medium schools at the top and subordinates the rest, including poorer students educated in “low cost English-medium” schools, below them (pp. 276–277). A. K. Mohanty argued for multilingual education in the mother tongue, not as an either-or proposition with respect to Hindi or English, but in conjunction with them, as the most effective means of realizing social justice.

NORTH AMERICA

Despite the fact that the term heritage language was coined in Canada, preparations for this article found almost no recent research on Canadian HLE policy. Duff (Reference Duff, Brinton, Kagan and Bauckus2008) indicated that most of the significant overviews of policies and programs in Canada were published in the 1980s and 1990s. She offered an explanation as to why so little research has been conducted since: “The dearth of scholarship on HL education in Canada since the early 1990s probably reflects the decline in federal funding for programs and research in the 1990s” (p. 72).

By contrast, a considerable amount of HLE policy research in the United States has appeared. Its emergence in the wake of the events of September 11, 2001, suggests that the focus for most analysis has been the relationship between HLE and national security. The U.S.-based policy research adopts almost universally a resource-based orientation (Ruiz, Reference Ruiz1984); that is, in general, recent HLE policy research has identified heritage languages as resources for national security, for strengthening language teacher preparation programs, and for expanding multilingual competency generally. The following discussion takes each in turn.

Recent HLE policy research has focused overwhelmingly on the connections between heritage languages and national security. This connection is also invoked in more general analyses of whether the United States needs a formal language policy, and if so, what that policy should look like. Edwards (Reference Edwards2004) acknowledged the increased attention paid to HLE post-9/11 and made the most explicit connection between increased language capacity and preventing terrorism. He referred specifically to Sen. Richard Shelby's (R-AL) comments in a 2002 interview with the New York Times, namely, that greater foreign language capacity might have prevented the events of September 11, and thus the intelligence community must recruit more heritage speakers of languages such as Arabic and Farsi. Brecht (Reference Brecht2007) positioned HLE as a crucial component in a general language pipeline running from primary through higher education. He argued that HLE as part of such a system could meet three critical policy needs:

Additional research takes a language-specific approach to HLE policy. McGinnis (Reference McGinnis2005) reviewed the status of Chinese language instruction and found that the HL sector accounts for the largest enrollments, upwards of 150,000 HL learners across community, K–12, and university programs. Al-Batal (Reference Al-Batal2007) and Sehlaoui (Reference Sehlaoui2008) presented similar overviews for Arabic language education, while Shinge (Reference Shinge2008) considered the impact of the National Security Language Initiative of 2006 on Hindi programs.

A complementary set of HLE policy research interrogates the policy connections between national security and heritage languages. Scollon (Reference Scollon2004) distinguished between state power, which privileges monolingual practice, and actual social power, which he argued is rooted in dynamic multilingual practice. Scollon maintained that HLE policies positioning multilingualism in the service of the state have pitted both power bases against one another, leading to an insurmountable paradox. Wiley (Reference Wiley2007a, Reference Wiley2007b) reviewed contradictory historical policy responses that frame heritage languages at once as a threat and an expedient resource. He argued that policies tying HLE to security terms in fact represent a “crisis of monolingualist ideology” (Reference Wiley2007a, p. 255). Allen (Reference Allen2007) raised his critique with respect to Arabic language education specifically and questioned the effectiveness of crisis-driven HLE policy. Lo Bianco (Reference Lo Bianco2008) conducted a close textual analysis of two policy documents positioning HLE in the service of national security. He finds that security-based policy “raises subtle and intangible matters of trust, loyalty and social categorization,” which can ultimately undermine stated policy intentions (p. 159). Bale (Reference Bale2008) reported similar findings of self-defeating policy in an historical interpretive analysis of Title VI, a long-standing federal language policy tied to national defense and economic competitiveness, and its impact on Arabic as a heritage language.

Despite the overwhelming focus on HLE policy and national security, additional examples of recent scholarship position HLE as a resource for meeting other critical social and educational needs. With respect to preparing future language teachers, HL speakers are identified as a more effective source of language teacher candidates, and HL communities are described as an ideal venue for teacher candidate placements (Magnan, Reference Magnan2008; Zimmer-Loew, Reference Zimmer-Loew2008). Met (Reference Met2003) and Roca (Reference Roca2003) also framed their analysis of improving foreign language education in the United States by identifying the essential role that HL learners play in programs that lead to advanced proficiency. Wang (Reference Wang2007) connected her policy analysis of Chinese as a HL with a deeper discussion of developing biliteracy in the United States. She made several recommendations for improving Chinese HLE, with a focus on highly qualified teachers and expanded research, but also positioned Chinese HLE as a resource for expanding what she called “biliterate capital” that can be used in community, professional, and personal domains. Also at a broader ideological level, both García (Reference García2005) and Hornberger (Reference Hornberger2005) framed their analyses of HLE policy in terms of the opportunities HLE offers to break through the fossilized debate over bilingual education in the United States, to create more openness toward education in non-English languages, and to realize a more multilingual society.

The category of languages left out of HLE policy connections to national security and the existing foreign language education system are indigenous languages in the United States. This is not to say that HLE policy research has been silent on the issue. For example, McCarty (Reference McCarty2008) presented an overview of indigenous languages as heritage languages in the United States and described three policy frameworks—tribal sovereignty, medium of instruction, and self-determination—that have been used to inform and implement indigenous language programs. Warhol (Reference Warhol2009) presented the first comprehensive analysis of the Native American Languages Act of 1990 (NALA). She found two competing interpretations among policy actors involved with its formation and implementation. On the one hand, policy actors understood NALA as being too little, too late to have any significant impact on reversing native language shift; on the other hand, policy actors see NALA as integral to the realization of tribal sovereignty.

This review of HLE policy in the United States has generally excluded important recent research on bilingual education policy, which has focused largely on the silencing of bilingual education effected by the No Child Left Behind Act of 2001 (NCLB). This exclusion is based on the fact that most of this research frames language minority students as English language learners rather than heritage or bilingual students (see Menken, Reference Menken2008b, and the recent special issue of Language Policy on NCLB [Menken, Reference Menken2008a]).

The final world about HLE policy in the United States relates to intriguing theoretical and analytical developments in the field. Lomawaima and McCarty (Reference Lomawaima and McCarty2006) proposed safety zone theory in their reading of historical indigenous language education policy. Safety zone theory is an effort to explain, not merely describe, policy shifts over time. Its central premise concerns how dominant national interests and identities (ab)use notions of safety and threat to endorse or marginalize linguistic and cultural difference through policy. Johnson (Reference Johnson2009) used his ethnographic study of bilingual education policy and practice in Philadelphia schools to propose a multilevel heuristic with which to analyze policy. The purpose of this analytic tool is to use ethnographic methods for understanding the policy appropriation process at micro and macro levels simultaneously.

OCEANIA

Although language revitalization efforts among the Maori in Aotearoa/New Zealand are both well documented and emulated, limited recent HLE policy analysis on these efforts were found. May (Reference May, Tollefson and Tsui2004) and Benton (Reference Benton, Tsui and Tollefson2007) provide comprehensive overviews of the history of language policy in Aotearoa/New Zealand.Footnote ⁴ May (Reference May, Tollefson and Tsui2004) focused specifically on the development of language nests as a means of Maori language revitalization beginning in 1982. Language nests are immersion programs at the early childhood level that simulate the extended family in the schooling context to facilitate Maori language development. May's analysis highlighted three pressing issues not only for Maori efforts but also for indigenous language revitalization efforts generally. First, formal education programs cannot compensate for language attitudes and practice in society more broadly; thus revitalization efforts need to look beyond the classroom. May also addressed the issue of romanticizing past language use in revitalization efforts and stressed that Maori-medium instruction allows the Maori to pursue cultural and linguistic change on their own terms. Finally, May raised the paradox of state funding for indigenous language revitalization programs. One the one hand, state funding allows for expansion of such programs, but on the other, it may lead to state encroachment on community control.

Historical perspectives on Aboriginal and community languages in Australia form the background to several examples of recent research (Baldauf, Reference Baldauf2005; Clyne & Kipp, Reference Clyne and Kipp2006; de Courcy, Reference de Courcy2005; Smolicz & Secombe, Reference Smolicz and Secombe2003). These overviews have traced the development of promotional HLE policy in Australia in support of heritage languages on three levels: as subjects in formal education, in Saturday community programs, and in after-hours programs auxiliary to the regular school day. In each case, the authors commented on the decline of state support for community languages. De Courcy (Reference de Courcy2005) argued, “we seem to have gone from a position in this country where we were the envy of other countries for our language policies to a position where the country now operates without a languages policy at all” (pp. 179–180). In addition, these overviews identified a connection between HLE policy and national interest concerns somewhat similar to those raised in the United States. Clyne and Kipp (Reference Clyne and Kipp2006), for example, analyzed HLE policy in Australia since the 1990s and found that official support for community, especially Asian, languages has often been framed in terms of their use for economic development and competition. They noted, however, that:

The authors questioned the effectiveness of tying support for community languages to instrumentalist aims.

Liu and Lo Bianco (Reference Liu and Lo Bianco2007) also pursued this theme of geopolitical rationales driving policy support for Asian languages in Australia. They described this support as connected to political discourses in Australia in the 1980s and 1990s that began to frame Asia as “our region” (p. 96) and raise the question of whether Australia were in fact an Asian nation. Concurrent to this discourse shift, education in Asian languages expanded greatly: “In a relatively short period of time key Asian languages came from the margins of education provision to be the most widely taught second languages at all levels of education, replacing the historical dominance of European languages” (p. 97). While the authors stated that Japanese has in fact been the main beneficiary of this expansion, their analysis comprised a case study of HLE policy toward Chinese. Their discussion focused largely on the mismatch between diverse types of Chinese language learners and the structure of Chinese language programs across the education system and in community schools.

Additional HLE policy in Australia research has focused on specific case studies of community languages or states. Smolicz and Secombe (Reference Smolicz and Secombe2003) included in their extensive overview of HLE policy a discussion of Cambodians in South Australia. Their discussion was based on data regarding Cambodians, mostly of Chinese heritage, who were university students at the time the research was conducted. They found that these students successfully took advantage of community schools to develop literacy in Khmer while they acquired English language and literacy in public school. Smolicz and Secombe noted that a South Australian policy acknowledging Khmer as a subject for university entrance exams has benefited the students, but they also suggested that students have pursued literacy in Khmer for noninstrumentalist reasons. Mercurio and Scarino (Reference Mercurio and Scarino2005) took a wider perspective on the South Australian policy of acknowledging community languages for secondary graduation and university entrance exam purposes. They framed policy as a process of legitimatization in which heritage languages are “grafted” (p. 145) onto socially accepted structures such as school administrations and curricula. The authors acknowledged the symbolic value of the South Australian policy, but they questioned its ultimate impact by analyzing data revealing low enrollments community languages in the state. Nicholls (Reference Nicholls2005) conducted an historical case study of Aboriginal HLE policy in the Northern Territory between 1972 and 1998. She recounted the rise and fall of bilingual Aboriginal-English programs and how state authorities’ perceptions that English skills were in decline led to the dismantling of formal bilingual education policy. Beyond state-level analyses, research in Australia has considered HLE policy at the institutional level as well. For example, Hatoss (Reference Hatoss2006) conducted a micro-policy analysis of Hungarian language maintenance in Australia and found that nongovernmental community organizations have played a critical role in leveraging state financial support for community languages. The author concluded that micro-level HL planning, far from being merely local implementations of state or federal policy, is a “necessary complement” (p. 288) to macro-level efforts.

Baldauf (Reference Baldauf2005) and de Courcy (Reference de Courcy2005) conducted case studies in separate contexts but came to similar conclusions. The former study examined HLE policy vis-à-vis Aboriginal languages in New South Wales; the latter was a multiple case study of HLE policy in the Northern Territory toward Auslan, a deaf language, and Italian. In both articles, the authors identified a number of practical issues complicating the realization of HLE policy. Many factors related to teachers in HL programs. Baldauf (Reference Baldauf2005) acknowledged that, although teachers in community schools have brought extensive language and cultural skills, the spectrum of proficiency in the HL has varied significantly. Both authors stated that teachers in HL programs needed more extensive training in language teaching; de Courcy (Reference de Courcy2005) stressed that training for primary level education is also needed. Additional factors impacting the effectiveness of HL programs auxiliary to public schools included HLs not counting toward credit within formal school accreditation practices, waning state funding, and sociolinguistic factors such as prestige attached to the HL.

THE REPUBLIC OF IRELAND AND THE UNITED KINGDOM

Recent research on HLE policy in the Republic of Ireland and the United Kingdom has focused almost exclusively on policies related to Irish, Scottish Gaelic, and Welsh. Wei (Reference Wei2006) did review the history and contemporary status of complementary schools for immigrant and ethnic minority children in the United Kingdom, but policy was not a significant focus of the discussion. Jones and Martin-Jones (Reference Jones, Martin-Jones, Tollefson and Tsui2004) and Smith (Reference Smith2003) both provided sociolinguistic overviews of Welsh and Scottish Gaelic, respectively, in their discussion of language revitalization efforts. Jones and Martin-Jones combined a critical-historical overview of national policy toward Welsh with a discussion of an ethnographic study of bilingual Welsh-English mathematics classrooms. They maintained that although policy support for Welsh-medium instruction has been strong, the language still faces formidable obstacles. First, despite a general increase in Welsh-medium programs, Welsh acquisition at home has continued to decline. Second, the shift among secondary students has continued to increase. And finally, there has been a greater diversity in Welsh proficiency among students entering school, straining the ability of Welsh-medium programs to cater to students’ linguistic needs. They called for further ethnographic study to better understand classroom-level appropriation of Welsh HLE policy. While Smith's discussion was more focused on the legal framework behind Gaelic HLE policy, she cited similar challenges facing Gaelic. One specific factor she mentioned is the unwillingness of native speakers of Gaelic to become teachers in Gaelic-English bilingual programs. Nevertheless, she concluded that “the law is unequivocal in its support for mother tongue education” in Scotland (p. 143).

The policy status of Irish in Northern Ireland schools is examined in English- and Irish-medium contexts (McKendry, Reference McKendry2007; Ó Baoill, Reference Ó Baoill2007). After presenting an overview of the Irish language and HLE policy toward it in Northern Ireland, McKendry (Reference McKendry2007) analyzed current education provisions for the language in English-medium schools. He raised doubts as to the future of the language in English-medium schools based on the fact that it has been most often offered in a “foreign language” structure, and that limited training has been provided through teacher preparation programs to teach the language. Ó Baoill (Reference Ó Baoill2007) also provided an historical overview of Irish in Northern Ireland with a specific focus on the central role played by Irish-medium HLE policy in revitalizing the language. He argued, however, that the limited amount of “pedagogical, educational, linguistic, and sociolinguistic research” (p. 423) on Irish immersion education has inhibited coordination across programs in Northern Ireland. In addition, he identified similar issues facing HLE policy for Irish-medium instruction, including lack of articulation across program levels and of well-trained teachers with sufficient competence in the language.

Laoire (Reference Laoire2005) provided an extensive historical and sociolinguistic overview of Irish in the Republic of Ireland. His discussion emphasized that “since the inception of the Irish Free State in 1922, the education system has been seen as a cornerstone of the movement for the revitalization of Irish. In fact, the espoused planning policies for revitalization devolved almost entirely on the schools” (p. 261). The focus of much of this language planning has been in the Gaeltacht in western Ireland, a region the government considers to be Irish speaking. Immersion approaches to Irish education in the Gaeltacht date back to the 1930s, and the number of immersion programs began to rise again in the 1980s and 1990s. However, the majority of students across the Republic now study Irish as a school subject. Despite improvements in how Irish is studied—for example, the use of communicative syllabi focusing on “everyday” rather than “school” Irish (p. 276) and the fact that Irish has gained in prestige since becoming a working language of the European Union—Laoire (Reference Laoire2005) argued that the language still faces important challenges. In addition to factors identified in other HLE policy research, Laoire added that immigration to Ireland, which has lead to an increase in students in school who speak neither English nor Irish, has further complicated the efforts to meet the needs of all its students while developing proficiency in Irish.

Two additional HLE policy studies in the Irish context shift their focus from national to institutional and interpersonal levels. Chríost (Reference Chríost2006) analyzed census data from 2002 and found that there were more daily users of Irish in Greater Dublin than in the Gaeltacht, despite decades of formal HLE policy focusing on Irish acquisition in the latter region. The data also reflected that school has been the primary site of Irish acquisition, not the home. Chríost argued that the discrepancy between formal policy and actual language practice suggests that micro-level language planning at the community level, modeled in fact on Welsh efforts, might provide better results in revitalizing the language. Ó hIfearnáin (Reference Ó hIfearnáin2007) drew similar conclusions based on his fieldwork in specific communities in the Gaeltacht region of Ireland. He surveyed residents and their attitudes toward formal HLE policy, which calls for children to be raised exclusively in Irish and for Irish-medium instruction at school. Participants reported general support for Irish use at home, although responses describing their practice reflected divergence from their stated opinion. Participants also generally supported Irish-medium instruction, but claimed that their children were nevertheless English-dominant. Moreover, some 25% of respondents objected to Irish-medium schooling. Ó hIfearnáin identified a gap between official HLE policy that considers the Gaeltacht to be Irish speaking, when in fact it is a bilingual region, and argued:

He concluded that what appears as lack of support for Irish-medium instruction is not a rejection of Irish per se but rather a preference for Irish-English bilingualism.

COMMON THREADS AND FUTURE RESEARCH

Before highlighting the common themes emerging from this broad review of HLE policy research, it is important to recall the caution referenced earlier about collapsing too many disparate linguistic, historical, cultural, and social contexts under one analytical term such as heritage language. Even while treading lightly with this caution in mind, the frequency with which similar logistical and programmatic concerns are cited in the research reviewed here is remarkable. In wealthy nations and developing ones alike, in rural contexts as in urban ones, and with respect to languages carrying the wide array of labels listed in the introduction to this article, researchers have consistently cited similar key factors impacting the success of HLE policies. These factors include linguistically and pedagogically competent teachers for HL programs, curricular materials in the target HL that are age and culturally appropriate, issues of language standardization and corpus planning, HL assessment, and finding ways to integrate multiple venues for HLE into existing administrative and accreditation systems.

While researchers consistently cite school-based factors impacting HLE, they also frequently conclude that formal schooling cannot be the lone site of HL revitalization or education programs. This conclusion holds not only in areas where schools have played a critical role in HLE and language revitalization, such as Aotearoa/New Zealand, Ireland, and the United Kingdom, but also where parental, educator, and community opinion is conflicted over where the HL should be taught (i.e., at home or in school).

The preceding discussion, however, uncovered two important tensions in the literature. The first relates to the historical and intellectual development of LPP overall. As LPP developed as a field of applied linguistic research, it often concerned itself with language “problems” faced by newly independent nations. Policy decisions regarding official languages, media of instruction, and language education generally were often tied closely to questions of national development. Those concerns still dominate much HLE policy research, as evidenced by the literature regarding HLE as a solution to “problems” of national security, economic competition, and socioeconomic development. This language-as-problem approach is deeply intertwined with assumptions about the role languages can, or better should, play in building, maintaining, and expanding nation-state power. The tension obtains inasmuch as HLE policy research increasingly adopts critical or postmodern theoretical frameworks that situate power and power conflicts at more localized levels in society that transcend the nation-state as a unit of analysis.

The second tension consistent across the research reviewed here is the contested relationship between HLs and English. Whether English is framed as a threat, as a nativized language, or a tool for empowerment and integration into a globalized world varied dramatically across the research reviewed here. Moreover, and particularly in postcolonial contexts, there is the additional issue where a regional language that is neither English nor the local HL was perceived by researcher and research participants as a linguistic threat. It is unlikely that consensus will be achieved with respect to whether English, or a national or regional language, is a threat or a tool of postcolonial identity construction and empowerment. Taken together, however, these common themes and tensions undergird another conclusion argued in much of the literature reviewed here, namely, that in order for HLE policy to be effective, there cannot be a one-size-fits-all approach.

In addition to these common themes, the literature reviewed here suggests multiple areas and questions in need of scholarly attention. The majority of the work cited here comprised single-nation case studies in which the bulk of the discussion was dedicated to an historical overview of language policy in precolonial, colonial, and postcolonial times. Another typical approach to analysis was to square HLE policies against sociolinguistic analysis of demographics and census data that included figures regarding language use and/or heritage. In some cases, the literature reviewed here was based on empirical research, and it seems that the most frequent approach included survey, questionnaire, and focus group data assessing attitudes and language ideologies of a given policy-relevant population. The two approaches to HLE policy analysis least frequently applied in the research reviewed here were historiography and ethnography. In each case, however, these approaches to HLE policy analysis stand to provide critical insights into how HLE policy works and has worked, especially when there seems to be such consensus that there can be no universal approach to HLE policy and practice.