Keywords

3.1 Introduction

Since 2002, the People’s Republic of China (PRC) has instigated language policies in education that have resulted in schools in ethnic minority areas being required to provide trilingual education. These moves form part of a broader movement initiated by the State Council to enhance and speed up the reform of minority education (State Council 2002). The three languages are the ethnic minority language (L1), standard Chinese (L2) and English (L3). The rationale behind these policies is that the three languages can play their respective roles in both the social and economic development of the country. Briefly, for a minority group, L1 is crucial for maintaining their linguistic and cultural heritage, thus their identity, and for children’s cognitive development (Baker 2011); L2 is equally important for social and economic integration and L3 is helpful for engaging with internationalisation.

Alongside the Han majority (which constitutes approximately 91.6 % of the total population), there are 55 ethnic minority groups that are officially recognized by the Chinese government, and these groups have a total population of around 106 million living in 155 largely resource-rich but economically under-developed ethnically autonomous areas, many of which are located near the country’s frontiers (China National Commission for UNESCO 2004).

Although the treatment of ethnic minority languages in the education system has varied between support and suppression since the founding of the PRC in 1949 (Adamson and Feng 2009; Lam 2005), the equality of ethnic groups is enshrined in law and their languages are protected by state institutions. The emphasis on the equality of citizens and their right to education and to use or study their ethnic language are assured by legislation. In China’s Law on Regional National Autonomy (1984), Article 10 states: “Organs of self-government in autonomous areas responsible for the local protection of all ethnic groups have the freedom to use and develop their own spoken and written language”; China’s Law of the People’s Republic of China on the Standard Spoken and Written Chinese Language (2000), Article 8, states “all nationalities in China have their own language and freedoms and rights”; and the Constitution of the PRC (1982) states: “Each nationality has the freedom to use and develop their own spoken and written language.”

On the other hand, standard forms of written and spoken Chinese, the mother tongue of the Han majority, are strongly promoted in the school curriculum as the language of national unity and communication. The political and economic capital associated with Chinese obviously endows this language with a high social status. The resources and roles (such as its use as the medium of instruction) ascribed to Chinese are pivotal in determining the status of the ethnic minority language in the curriculum. When Chinese receives a significant amount of curriculum time and is used as the medium of instruction, this scenario is very often detrimental to the well-being of the minority language. To a lesser extent, the same is true for English, although it is very rare for this language to be used as the medium of instruction in schools in ethnic minority regions. In summary, trilingual education–if implemented effectively–can reduce the potential marginalisation of ethnic minorities by enabling them to engage in the social and political life of mainstream society, and to accrue economic benefits through national and international trade, while at the same time allowing them to maintain and develop their own cultural identity and language. On the other hand, however, poorly conceived and/or ineffectively implemented trilingual education policies might have the opposite effect, leading to a greater sense of marginalisation and economic deprivation.

English has been included in the primary school curriculum since 2002 from Grade 3 because of the international economic advantages that proficiency in that language can bring (Ministry of Education 2001a, b, c). The benefits accruing to ethnic minority regions include increased tourism, joint ventures, international economic activities (such as the important China-ASEAN Expositions held annually in Guangxi) and other ‘open-door’ activities (Blachford and Jones 2011; Huang 2011; Sunuodula and Feng 2011). The importance ascribed to English in contemporary PRC has “reached unprecedented heights, although fundamental cultural and political tensions remain” (Gil and Adamson 2011: 30). These tensions include the perception that the current expansion of English language education is inevitably widening the economic gap between the Han majority and minority groups and augmenting the educational inequities that minority peoples already face in the traditional system (Beckett and MacPherson 2005).

In the PRC, as elsewhere in the world, the term “trilingual education” is used in different ways. In one, narrow, sense, it refers to the use of three different languages as the media of instruction to teach different subjects across the curriculum: for instance, the use of English to teach mathematics and science, while standard Chinese is used for social science subjects and an ethnic minority language is used to foster cultural and mother tongue literacy. A broader meaning of the term can extend this definition by including the teaching of the three languages in the curriculum, but using only one or two language(s)―usually the mother tongue of the minority group or standard Chinese, or both―as the medium of instruction for most school subjects. This chapter is concerned with trilingual education in the latter sense.

Trilingual education has proved controversial on the grounds of equity and social justice. The approaches to trilingualism, although legislated at the national level, are heavily influenced by regional and local factors. Feng and Sunuodula (2009: 699), for example, in their analysis of language education policies for minority groups found a ‘weak link’ at the regional, prefectural and county levels in Xinjiang with regard to English language provision for the minority students. The decentralisation of policy-making that was enshrined in the educational reforms of 1985 produced a plurality of education systems within the PRC (Lewin et al. 1994). As a result, provinces and autonomous areas now enjoy a high degree of independence in education policy. In regions where there are significant populations of ethnic minorities, this flexibility has produced a variety of responses in terms of language policy. Generally speaking, such language policies reflect the social status of the ethnic minorities in that particular region and their position in relation to the majority groups (Adamson and Feng 2009). Supportive policies in respect of minority languages can help to preserve and revitalise these languages, as well as contribute to the conservation of cultural heritage; suppressive policies can result in linguistic decline, social marginalisation and economic disadvantaging (Edwards 2004).

Controversy also surrounds the linguistic and cognitive demands the policy of trilingual education places on students. Can students cope with learning three languages? There appears to be agreement in the literature that bilinguals are normally better at learning a third language in schools than monolinguals are at learning a second language and that they have a cognitive advantage over the latter (Cenoz and Jessner 2000; Clyne et al. 2004; Hoffmann and Ytsma 2004). Cenoz (2003) and Cenoz and Valencia (1994) demonstrate with empirical evidence that bilingual students (Spanish and Basque in their case) achieved higher proficiency in the English language than monolingual Spanish counterparts learning the same language. However, the European cases of trilingual education are characterised by relatively close linguistic distance between the three languages and relatively rich linguistic environments. As a result, the European model of trilingualism promoted in the literature usually calls for early introduction of the three languages (Ytsma 2001).

However, in the emerging literature on trilingual education in the PRC, despite occasional reports that give support to the hypothesis, many educators and researchers argue that minority students experience various cognitive, cultural and psychological problems in learning the third language, in this case English (e.g., Zhang 2003; Jiang et al. 2007). The European model, therefore, does not seem appropriate for the PRC (Jiang et al. 2007; Zhang 2003). This indicates the need to establish indigenous Chinese models of trilingual education that take into account the range of contexts in which they operate.

This chapter reports on a project that studied the implementation of trilingual education policies across ethnic minority regions in the PRC, including Yunnan, Sichuan, Guangxi, Inner Mongolia, Xinjiang, Jilin, Gansu, Guizhou, Qinghai and Tibet. Using first-hand data collected from each region, the research team examined language policies and curricula, as well as language allocation in the classroom and in the community, and analyzed them in their specific historical, socio-political, demographical, economic, geographical and cultural contexts. Three years of research by the teams yielded substantial data and exciting findings which will be reported in Feng and Adamson (2014). Given the confines of this chapter, we will restrict ourselves to a summary of the findings by answering the following two questions:

  1. 1.

    What are the common and diverse features of language policies promoting trilingualism at the planned (policy-making) and implemented (classroom) levels in different regions of the PRC?

  2. 2.

    What are the main factors that account for these commonalities and diversities?

The team uncovered four models in schools, each incorporating a different view of trilingual education and thus producing students with different levels of trilingual competence. These models have only superficial similarities to the European model, as they have to take account of the complexities of the Chinese context. Analysis of the factors influencing the four models reveals various tensions in the context of policy implementation.

3.2 Complexity of the Context

Implementing the policy intentions to foster trilingualism is complicated by a number of geographical, economic, political and linguistic factors. The relative remoteness and poverty of many regions—with some notable exceptions—where ethnic minority groups live creates problems in resourcing language education policies. A survey conducted in rural areas of western China, home to many minority groups, found that “37.8 % (of the schools) do not have enough desks and stools; 22.3 % have unsafe classrooms or offices; and about 32.5 % do not have enough funds to buy teaching aids, ink, chalk, and other supplies” (Yang 2006: 20). A common problem facing school management is finding sufficient teachers of English who are able to teach through the ethnic minority language. In many areas, primary students have to learn English through Chinese, adding to the cognitive and linguistic demands of mastering an already complex and alien language. English is linguistically-distant from the second language, Chinese, in terms of syntax, vocabulary, pronunciation and written form. Both English and Chinese are often linguistically-distant from the minority language. Under these circumstances, ethnic minority students tend to have a high drop-out rate, while those who stay in school often perform worse than their Han majority counterparts (Hu 2007a; Jiang et al. 2007; Tsung 2009).

The problem of medium of instruction was exacerbated in the past by unfavourable official attitudes towards the minority languages. Despite some periods during the 1950s and 1980s when minority languages and cultures were more respected, for decades the dominant mentality of policy makers and educators has been to standardise China’s education in terms of the curriculum, textbook materials and school activities. A culturally and linguistically homogenous education system oriented towards socialisation was the chief aim (Hansen 1999). Linguistic assimilation among policy makers can be best illustrated by the claim made by the then influential personal secretary of the former all-powerful paramount leader, Mao Zedong, that the central “government must eliminate Han dialects within 10 years and eliminate minority languages after we develop them into Hanyu Pinyin scripts” (Tsung 2009: 88). Such an attitude on the part of the key policy makers meant that minority languages were hardly valued. Minority pupils found it difficult to compete with their majority counterparts as they had to depend on their second language, not their mother tongue, for studying school subjects.

The homogenous approach was what Cummins (2004) describes as a coercive policy that avoided the complications arising from the diversity of ethnic minority languages in the PRC. The ethnic minority languages may be categorised into three types (Zhou 2000, 2001): Type 1 group refers to those that possess both the spoken form and traditional written form of the language of wide usage and have had regular bilingual education since 1949. Type 1 groups include Uyghurs and Kazakhs living mainly in Xinjiang, Tibetans in Tibet and Qinghai, Mongolians in Inner Mongolia, and Koreans in Jilin. Type 2 groups are those that have functional writing systems of only limited usage and have had only occasional bilingual education since 1949. This type consists of Dai, Jingpo, Lisu, Lahu, Miao, Naxi, Va and Yi minority groups living mainly in the south-west of the country. The remaining 42 minority group communities belong to Type 3 which are defined as those that had no fully functional writing systems before 1949 and since that time have had limited or no bilingual education.

Lack of access to languages with high social and economic capital can create marginalisation (Beckett and MacPherson 2005; Adamson and Feng 2009), which can be exacerbated by a perception in the mainstream that ethnic minority groups are difficult to assist, hard to reach in both the geographical and social sense, conservative, culture-bound, and lacking the know-how to help themselves (Abdulrahim 1998). Even today, the life opportunities of many minority pupils depend on ‘preferential policies’ (Feng and Sunuodula 2009; Adamson and Xia 2011). One important policy of this type is to allow minority students to enter higher education institutions with lower marks in the high-stakes National College Entrance Examination. Once attending the university, however, minority students are often found to struggle in order to keep up. When they graduate, they find themselves seriously disadvantaged in the job market.

While many regions where ethnic minorities live are remote and poor, not all minority groups suffer from what Vaughan et al. (2005) label marginalisation through “spatial and economic poverty”. Zhou (2001) reports that some minority groups such as Koreans and Russians have a higher percentage of college degree holders than the national average, including the Han. The Korean and Russian groups are particularly advantaged owing primarily to the fact that South Korea and Russia play important roles in the world economy. There is, for example, a high demand for Korean graduates in neighbouring South Korea and by the companies set up by Koreans inside China (Lin 1997). As a minority group inside China, the Koreans have become empowered not because they are linguistically or culturally assimilated into the mainstream society, but rather because they have demonstrated a strong sense of identity, as well as multilingual and multicultural competence.

Some groups receive more state attention than others. The location of many of the ethnic minority groups on the geographical fringes of the PRC means that some groups have ties of ethnicity and language that stretch across the border outside China, which could be seen as a positive for trade purposes but as a negative in some instances for political reasons if the state feels there is a danger of “separatist” sentiments developing, as in the case of Xinjiang and Tibet. There can be differences in state policies among the minority groups because of these factors. Groups such as the Koreans are in a strong position to negotiate their identity and bargain for linguistic rights, creating a dynamic relationship with the state or regional government (Schluessel 2007). Some groups including Uyghurs, Yi, Dai, and Kazakhs managed to overturn the policy made in the 1950 to change their writing scripts and retained their original ones. During the 1980s, many groups re-established the minority language as the medium of instruction in primary schools (Tsung 2009), which matched the preference of parents for their children’s education (Postiglione 1999; Zhou 2000, 2004). Nowadays, in addition to standard Chinese, the high-stakes national College Entrance Examination is administered in Tibetan, Uyghur, Mongolian, Korean, Kazakh, and Kirghiz (Mackerras 1994), all of which are the languages of minority groups in border areas where there are strategic economic and political considerations to be taken into account.

3.3 The Project

Clearly, the issues surrounding the provision of trilingual education in the PRC are complex and sensitive. However, research into the diverse ways in which trilingual policies have been conceptualized and implemented in different parts of the country has hitherto often been limited to one individual region or education institution (e.g., Huang 2007; Jiang et al. 2007). Feng and Sunuodula’s (2009) research into three minority dominated regions (Xinjiang, Guangxi and Yi Autonomous Prefecture in Sichuan)—their findings are reported in Adamson and Feng (2009), Feng (2007, 2008), Feng and Sunuodula (2009) and Sunuodula and Feng (2011)—was more comprehensive, but largely based on small-scale studies and documentary analysis of policies. Their research found that many minority pupils fail to acquire age-appropriate competence either in their minority home language or the majority language (standard Chinese) and thus are unlikely to avoid the negative consequences of this for their social and economic development; second, while some minority regions have responded to the official 2001 English Curriculum Standards by enhancing English provision, some only seem to pay lip service and their priority remains the further enhancement of the teaching and learning of Chinese. The low standards in English resulting from the (lack of) regional policy have a detrimental effect on the students’ life chances.

To date, little multilevel, comparative work has been carried out in mapping the different forms of policy and their impact across the country. Our earlier study (Adamson and Feng 2009) examined the tensions behind trilingual education policies by comparing the implementation of policies for three minority groups: the Zhuang, the Uyghur and the Yi people. It found that ethnic minority languages are at a disadvantage compared with Chinese and English, and that additive trilingualism (the learning of three languages without mutual detriment) is facilitated by strategies such as supporting the development of not only speaking competence but also literacy in the minority languages and provision of space in the school curriculum for all three languages; while the barriers to additive trilingualism include the low social status ascribed to minority languages because of their lack of associated economic and political capital. However, as already mentioned, this report was not a comprehensive nation-wide study.

Bearing this in mind, our 2009 project started with the establishment of the network of nine research teams across the PRC, in Gansu, Guangdong, Guangxi, Inner Mongolia, Jilin, Qinghai, Sichuan, Xinjiang and Yunnan—teams studying Guizhou and Tibet joined later. The results of these regional projects provide an overview of the policies and the implementation of trilingualism and trilingual education, the forces that shape them, and the implications for social equity, in different parts of the PRC. Multiple-case studies were conducted in three areas in each autonomous region or prefecture covered by the network. To ensure that the data from each region was comparable, the research teams chose three sites that demographically represent the population typology of the region or prefecture, geographically represent the whole region or prefecture in terms of typography and transportation, and economically, represent the region or prefecture in terms of GDP. At each site, four schools (three primary schools and one secondary) were chosen using criteria similar to those used for determining the sites. Primary schools are as representative as possible in terms of resources, history, demography and geographical location. Only one secondary school (an ordinary school attended by minority children) was chosen because minority children from remote village schools tend to go to a secondary school in a town specially catering for minority children. They are unlikely to go to the privileged ‘key schools’ (zhongdian zhongxue) dominated by the Han majority children.

The research teams include members who are proficient in the minority language and Chinese, or other local language(s). The field visits took place over an extended period (several weeks) and included questionnaire surveys, documentary analysis of policies in each region, observation of classroom practice and interviews with stakeholders such as policy makers, and focus groups of teachers, parents and children.

In order to provide coherence across the project, an annual symposium was held, bringing together all the teams and the principal investigators to decide on common approaches to data collection, analysis and presentation, and to report their findings. These symposia facilitated the application of comparative education research methods. The added value of using a comparative perspective is that the contextual factors (geographical, economic, political, linguistic, etc.) are brought into sharp relief by juxtaposition.

3.4 Results

The project distinguished four policy models of trilingual education. All four models were found to be present in some regions such as the Inner Mongolian Autonomous Region (IMAR), while three of them were found among Korean communities in north-eastern parts of the country, and at least two of the models were present in all other regions.

The first model focuses strongly on the ethnic minority language. In a typical school in remote areas in the IMAR, for instance, the 9 years of compulsory education from Grade 1 in primary schools to Grade 3 in junior secondary schools is provided through the medium of Mongolian. A similar situation is also found in many schools in the Yanbian Korean Autonomous Prefecture (Zhang et al. 2014). Chinese and English are taught as subjects in the curriculum. The school environment, as manifested by the slogans on the wall, notices, portraits (including prominent figures in their history), school documents and spoken interactions between the teachers and students emphasizes their own language and culture. Take for example one school situated in a remote region where the vast majority of people are Mongolian. Han students attending the school tended to be bilingual in Chinese and Mongolian (Dong and Narisu 2014). In this model, the ethnic language tends to have strong vitality, being widely used and supported by the community. Where favourable conditions exist with regard to resources and regional policies, the other two languages are equally robustly promoted by being allocated sufficient time in the curriculum for the students to acquire a high degree of competence in Chinese (L2) and a developing competence in English or another foreign language (L3) (Zhang et al. 2014). This model is likely to produce additive trilingualism, by which we refer to complementary competencies in L2 and L3 that pupils acquire in school and in society while maintaining a high standard of their L1 and their ethnic identity. For the purpose of comparison with other models, we call this an Accretive Model.

The second model is a balance between Chinese and the minority language. The balance is evident not only in terms of the medium of instruction but also in terms of the ethnicity of the teachers and students. In one school visited by the research team in Inner Mongolia, the ratio of Han teachers to Mongolian teachers was 30:70, and that of Han students to Mongolian students was 60:40. As well as using both languages as the medium of instruction, the school encourages a bilingual environment through the use of both Chinese and Mongolian notices. The playground language was also bilingual. English is taught as a school subject and the teacher used either Mongolian or Chinese to explain difficult points, depending on the preferred language of the students in the class and the ethnicity of the teacher. The second model is more nuanced than the first model, as the school and the local community display more ethnic diversity. The ethnic language is supported, while the educational needs of the students to learn through a familiar language is respected. The bilingual pupils are likely to perform well in other school subjects, including English, their L3. Again, for comparison, we label this a Balanced Model.

The third model often exists in two different forms. First, in some mixed communities such as towns and cities where there is a substantial minority population (see Dong and Narisu 2014), the medium of instruction can be the reverse of the first model, that is, Chinese is used as the medium of instruction. However, the dominant ethnic minority language in the region is taught as a school subject to all students in the school, irrespective of their own ethnicity or mother tongue. This might be seen as an attempt to maintain the minority language, but there is little role for the minority language to play in later years of schooling. Second, in many remote village schools in which one minority group dominates, the minority language is used as the medium of instruction for the first 2–3 years, with Chinese taught as a major school subject. Starting from Year 3 or Year 4, all school subjects are taught in Chinese. In both cases, English is again taught as a school subject, with Chinese being used when necessary in those lessons.

The two forms are included in Model 3 because they have one feature in common, that is, the transit to using Chinese as the medium to teach all school subjects. The second transitional form of bilingual education is widely seen in many Type-2 and Type-3 communities (Zhou 2001) which possess a weak degree of ethnolinguistic vitality. In some Type-1 communities such as Inner Mongolia and Xinjiang, as the transitional model privileges Chinese, there is concern to preserve the ethnic minority language and to propagate it among non-native speakers. This model is especially applicable in regions where there is a significant Han population but the ethnic minority language also possesses a strong degree of vitality. However, while the cultural value of the ethnic minority language might be acknowledged, its vitality is often insufficient for it to be used as the predominant language in the school. The third model, in these two forms, is therefore termed a Transitional Model.

A fourth model is represented by schools that proclaim to be ethnic minority language schools but, in reality, do not use the minority language as the medium of instruction nor even teach it as a school subject. Such schools also claim to be bilingual, in the sense that Chinese and English are studied as languages in the curriculum and Chinese serves as the medium of instruction. In these cases, the bilingual label reflects the curriculum content, while the trilingual label reflects the ethnic profile of the students. The outcome is almost inevitably the loss of pupils’ L1 and eventually their ethnic identity. This is a Depreciative Model.

Table 3.1 summarises the four models with key features ascribed to each of them. It can be seen that these four models form a continuum, moving from the predominance of the ethnic minority language at one end to the predominance of Chinese at the other. In all the models cited English is taught as a school subject, with some schools in cities or towns providing earlier and better quality than those in remote areas, even though in some parts of the PRC—most notably the richer areas of eastern China—English is used in some schools as the medium of instruction for teaching a certain percentage of school subjects (Hu 2007b, 2008) and immersion programmes in English are also mushrooming in these areas (Qiang et al. 2011). The medium of instruction used to teach English in minority regions follows the line of the continuum.

Table 3.1 Summary of the four models found in the trilingualism-in-China project
Full size table

As the aims and outcomes of different models specified in Table 3.1 indicate, the Accretive and Balanced Models are likely to help pupils develop additive trilingualism, whereas the Transitional and Depreciative Models may result in subtractive trilingualism, that is, the gaining of L2 (plus some competence in L3 if it is taught) at the expense of L1. How can we explain the existence of the four different models? The data from our analysis of policy documents and our interviews with key informants reveal that each model exists within a complex ecology that is determined by geographical, linguistic, pedagogical, historical, economic and political factors. The first model of language policy tends to be found in areas that are more remote and where the ethnic minority group forms the majority of the school population. In these areas, there is also a degree of economic stability that allows schools to recruit a sufficient number of teachers proficient in the ethnic minority language—the model often breaks down when such teachers are unavailable because potential recruits have joined other adults from the local community seeking employment as migrant workers in the big cities.

The vitality and prestige of the particular ethnic minority language is also related to economic factors. For example, Korean, Mongolian and some of the languages found in Yunnan Province have high economic capital because of the opportunities they afford for cross-border trading. Another factor is political: the first model is supported by the local, regional and central governments where the language and the associated minority people are not perceived as representing a threat to the integrity of the PRC—it has become less common to find these schools in Xinjiang, for example, where more measures are being taken to assimilate the minority groups and Mandarin Chinese is being more forcefully promoted than ever before (Tsung and Cruickshank 2009).

The second model tends to be found in areas such as medium-sized cities where there is more of an even balance between minorities and Han Chinese. It can also be located in regions where there is a high degree of historical assimilation by the ethnic minority group into the mainstream Han society, as is the case with the Zhuang in Guangxi Province (Adamson and Feng 2009) and with the Yi in towns and cities in Liangshan, Sichuan Province (Liu et al. 2014). Again, there needs to be a degree of economic stability to ensure a sufficient supply of bilingual teachers, so that the use of both the minority language and Chinese as the medium of instruction can be assured. An interesting insight into the dilemmas and preferences of people living in such areas was provided by the principal of one of the schools visited by the research team in Inner Mongolia. The principal had chosen not to send his son to his own school, preferring instead that he should attend one that offered just Chinese and English, and that used Chinese as the medium of instruction. “It was tough to make a decision whether to send him to my school or a Chinese school. Thinking about his future, Chinese is more useful than Mongolian when he is grown-up,” he said (Dong and Narisu 2014).

As noted above, the two transitional forms of the third model are primarily found in big cities or towns in regions dominated by one ethnic minority group. In these cities, there is often a very large Han population and considerable economic interaction with the rest of the PRC, which requires proficiency in Chinese. Many schools in cities in Inner Mongolia, for example, adopt this model. The transitional model is also widely seen in village and town schools in regions where a minority group (usually a Type-2 or Type 3 community) dominates but only maintains the home language in its spoken form. Pupils in these instances need home language support in the early years of schooling but that language quickly gives way to Mandarin Chinese in late primary school years. Zhuang in Guangxi and smaller ethnic minority groups in many counties in Yunnan, Guizhou, Sichuan and Hunan often adopt this transitional model.

The fourth model is found in areas where there is a low political commitment to the maintenance of the ethnic minority language, or where there is considerable difficulty in recruiting teachers from the ethnic minority community. It also tends to occur in schools where there is a mixture of ethnic minority students, making the provision of teaching in those languages economically inefficient and difficult to accommodate in the curriculum.

In addition to these major models, other practices are found for students from specific contexts—often in those areas that are deemed politically sensitive. For example, in Xinjiang, minority and Han schools are sometimes merged as Min Han Hexiao (Tsung 2009); another practice is Tibetan and Xinjiang Neidiban—classes usually located in schools in relatively developed inland cities in other provinces but attended by Tibetan or Xinjiang secondary students away from their minority communities.

3.5 Conclusion and Discussion

These four models are underpinned by prevailing attitudes towards the three languages—the ethnic minority language, Chinese and English. It appears that in some localities the minority language is barely tolerated, and it is only promoted in contexts where it has high economic capital. Chinese is strongly preferred because of the economic, social and political benefits that are associated with the language. This phenomenon represents a contradiction between national policy and practice at the grassroots, which stems from the political and economic complexities of the local context. There is a further contradiction with regard to English. Although the language is taught in the majority of the schools studied in the project, and its value as a language for international trade and communication, as well as for access to higher education, is recognized, many schools struggle to recruit teachers with a reasonable competence in English or the ability to teach the language through the students’ mother tongue. This logistical problem tends to undermine the efforts of national policy to produce trilingualism in ethnic minority students.

The tentative conclusion of our research is that trilingual education that does not have a detrimental effect on minority pupils is only really flourishing in schools that use the first model and to some extent the second model if it is used appropriately. The effectiveness of the first two models is clearly fully demonstrated by Zhang (2008) and Zhang et al. (2014) in Yanbian where the minority language is powerful in economic and socio-geographical terms and where many pupils are trilingual or multilingual. These strong forms of trilingual education may well serve as good models for minority groups such as Russians and Kazakhs in Xinjiang and Mongols in Inner Mongolia which are areas with potential for easy cross-border trading and communication. More encouragingly, in some areas where the minority language does not seem to have much current economic values, some small-scale but significant work has been done to revitalise the home language and empower the minority pupils with bilingual or trilingual competence (Geary and Pan 2003; Finifrock 2010). Such work demonstrates that for a minority group the value of trilingual education goes beyond mere financial benefits to include greater confidence in one’s own culture and identity and better cognitive development.

Unfortunately, in our nation-wide analysis to date, the first or the second model has rarely been found—most schools studied in the project appear to follow the third and fourth models, which are seen as weak or subtractive by definition (Baker 2011). The implication is that much work remains to be done in the PRC if the national policy of fostering trilingualism in ethnic minority regions is to be achieved. It is important that this work is carried out as issues of social justice are at stake. Failure to support the language of an ethnic minority group runs the risk of weakening their sense of identity and linguistic and cultural maintenance. Research also shows that without a fully-developed mother tongue, minority pupils usually suffer cognitively and academically, which causes them to lag behind their majority counterparts. Inadequate provision of English can affect their life chances, as it reduces access to higher education and lucrative employment, which in turn may affect ethnic harmony and social stability. However, stronger models do exist in specific contexts in China. The challenge—which is familiar to everyone engaged in comparative education —is transferring stronger models to other contexts in ways that are appropriate and effective.