Introduction

In 2013 and 2014 respectively, the Malaysian and Thai governments announced that they would begin using the Common European Framework of Reference for Language (CEFR) in the development and implementation of future language policies. By doing so, both governments appear to be following a worldwide trend, with the framework having long ago transcended European borders and become a globalized language policy instrument (Byram and Parmenter 2012). At the local scale, CEFR has seen widespread use in the design of curricula and in the development of teaching materials and tests. In parallel, the framework has also become increasingly associated with the global influence of major ELT textbook producers and testing organizations, which have made significant use of CEFR in the development and marketing of their products (Littlejohn 2012). Additionally, the framework remains associated with the original agenda underlying its development, that of European integration, though this in itself is a site of struggle between humanist and neoliberal language policy discourses (Krzyzanowski and Wodak 2011).

The aim of this paper is to examine how the recontextualisation of CEFR in non-European contexts takes place against this complex discursive background. The paper begins by considering, from a discursive perspective, how globalization affects language policies. I continue by discussing the trajectory of CEFR, starting with its inception in the European context while also discussing its use in globalized English language teaching products. I then examine the recontextualisation of the framework in Thai and Malaysian language policy, focussing in particular on the key concepts that underpin the policy agendas in which CEFR is integrated in either context. A data-set of recent policy texts from both contexts is analysed with the aim of addressing the following research questions: What local language policy agendas is CEFR integrated into and what are the key concepts and conceptual relations that underpin those agendas? What elements of CEFR are recontextualised and what conceptual relationships are they placed into? How are ideological struggles surrounding CEFR negotiated in policies at the local scale?

Globalized language policies: scale and recontextualisation

Attention in ethnographic and discourse-analytic work in sociolinguistics (see e.g. Heller 2011) has recently shifted to the examination of relationships between different contexts of language use. To aid in the theorisation of such a conceptual refocus, Blommaert (2007) proposes the concept of scale, which he sees as complementing the existing horizontal dimension (represented in concepts like language community) with “a vertical dimension of hierarchical ordering and power differentiation” (2007, p. 4). Issues of power are thus central to analysis of scales, though it is, as pointed out by Canagarajah (2018), overly simplistic to assume that a higher or lower position in such a hierarchy would automatically correspond to a higher or lower level of agency for the actors involved. Instead, the examination of scales continues to have an ethnographic orientation, stressing the need to examine particular nexuses of practice and describe the configurations of power that language mediates within them (e.g. Hult 2010; Pietikäinen 2010; Savski 2018).

The concept of scale has also seen use in language policy research, particularly in studies pursuing either a discursive or ethnographic approach, or a combination thereof (Barakos and Unger 2016; Wodak and Savski 2018). For policy analysis, the concept of scale is seen to add not a complementary dimension but to bring to the forefront the need to interlink various analytical levels in an area which is, in many cases, inherently hierarchical. This is particularly central to the examination of policies associated with state authority, since their creation, interpretation and implementation often takes place in conditions typified by the existence of both agentive opportunities and structural constraints (Jessop 2007; Savski 2016). This means that the trajectory of policies across spatiotemporal scales is often determined by the ways in which specific actors interpret them while negotiating constraints of different types (Hornberger 2005; Johnson 2013). In Savski (2018), I thus examined how actors participating in a committee meeting in the Slovene parliament negotiated subject positions imposed by institutional practices (e.g. committee chair vs. member of the public) as well as those imposed by the broader discourse surrounding language policy in Slovenia (e.g. linguist as expert vs. politician as non-expert) while attempting to agree on a mutually acceptable set of amendments to a language strategy. I showed how such lines both governed specific interactions at the meeting but also how actors were able to subvert and exploit them by switching between different subject positions in order to achieve their goals.

In this paper, I analyse the trajectory of a transnational language policy text as it is interpreted and appropriated by actors at the national level in two contexts. To examine this trajectory, I take a discursive approach by focussing on recontextualisation, which I understand to refer to the creation of intertextual and dialogical relationships through the transfer of specific elements of a given text to another context (Maybin 2017; Reisigl and Wodak 2015; see also Wodak and Fairclough 2010). Such a transfer is seen to be potentially determined both by agentive opportunities and structural constraints, with complex configurations of power determining what elements are transferred across the boundaries between particular scales. Furthermore, as such transfers involve the negotiation of boundaries between the power relations, practices, discourses and ideologies that characterise particular scales, they may lead to shifts in the meaningfulness of the recontextualised elements. Kulsiri (2006) for instance examines how elements of educational policy were recontextualised from the US state of Louisiana to Thailand, highlighting how the values of learner-centredness that dominated the original policy were complemented by elements conforming to the more teacher-centred practices traditionally found in Thai education.

This example also highlights the fact that policies which transcend the borders of a single nation-state and become globalized in the sense that they have become embedded in broader transnational and transcultural flows of people, technology, information, finance and ideas (Appadurai 1990) are not examples of simple transfer but rather of hybridization. The globalization of policy takes the form of a continuous tension between antagonistic tendencies toward greater heterogenization, decentralization and particularization from the local scale and toward more homogenization, centralization and universalization from the global scale (Wodak and Fairclough 2010). Since this antagonism is often dependent upon the specific discourses and political economic conditions of each specific local context, the global spread of policies is often defined by its non-simultaneous and asymmetric nature (Krzyzanowski and Wodak 2009). Such local negotiation of global policies may often lead to unlikely compromises between opposing agendas, as highlighted by the aforementioned example from Kulsiri (2006).

Such agentive negotiation is, however, subject to the structural constraints presented by political-economic relations. A key point to underline is that policies which transcend the national level are often either expressly created with the intention of influencing decisions across a variety of potential polities (as is the case with CEFR) or have become associated with such an agenda despite having originally been created for use in a specific local context [as was the case with the Louisiana curriculum discussed by Kulsiri (2006)]. In other words, language policies associated with processes of globalization are instruments of transnational governance through which powerful global actors attempt to influence local policy (Holzinger and Knill 2005; Roger and Dauvergne 2016). While this may take the form of outright coercion or imposition of reform, as in the case of the imposition of austerity measures on European Union members in the wake of the Eurozone crisis, instruments of ‘soft power’ are the more usual form of transnational governance in fields like education and language policy, with league rankings, independent ratings and ‘good practice’ examples being common catalysts for policy change (De Costa et al. 2019; Rutkowski 2007). CEFR is also such an instrument of transnational governance, though highly specific in the sense that while it was developed for use in a clearly delimited set of polities (European nations) it has since also become associated with broader global agendas. Its trajectory toward globalization is discussed in the following section.

CEFR as a globalized language policy

The Common European Framework of Reference for Languages (CEFR) was published by the Council of Europe (CoE) in 2001, the direct result of a decade-long effort initiated in 1991 at a symposium in Rüschlikon, Switzerland (reported on in North 1992). As summarized by Trim, the main outcome of this symposium was a commitment to develop a framework which would ‘assist learners, teachers, course designers, examining bodies and educational administrators to situate and coordinate their efforts’ (2012, p. 29). Over the following years, the proficiency descriptions that have now become the most familiar element of CEFR were developed through a process in which teacher perceptions were collected and used to create series of descriptors calibrated to the now familiar six levels (with A1 and A2 representing ‘basic’, B1 and B2 ‘intermediate’ and C1 and C2 ‘advanced proficiency’; the calibration process is described in detail by North 2014). While these levels were nominally new, they represent an example of evolution rather than revolution, with several of the levels having been previously described, starting with B1—originally referred to solely as Threshold (Van Ek 1975). The very fact that this level was described first is indicative of the motivation underlying it, with its intention originally having been to describe the abilities an L2 speaker needed in order to survive in a context where that language was in dominant use (ibid.). Such a need stemmed directly from the European post-war period, when the establishment of early organisations such as the European Community for Coal and Steel began to push transnational workforce mobility to the forefront of the policy agenda, with issues related to the linguistic and cultural integration of such economic migrants also becoming relevant (Tabouret-Keller 1991).

While the historical background of CEFR may thus be attributed to a relatively practical set of needs for a set of transnational guidelines for language learning, the framework was at the time of its publication attached to a much broader language policy agenda promoted by CoE. The centre-pieces of this agenda were the twin concepts of plurilingualism and pluriculturalism, which feature heavily throughout parts of the original publication (Council of Europe 2001). In the framework, they are seen as reflecting ‘the fact that as an individual person’s experience of language in its cultural contexts expands, […] he or she does not keep [different] languages and cultures in strictly separated mental compartments’ (ibid., p. 4). In other words, the CoE agenda with which CEFR has become associated is one which argues for a fluid approach to defining language competence, one which focusses on the integration rather than the segregation of languages (for further discussion, see Piccardo 2010, 2013 and Savski 2019). By setting this agenda, CoE and CEFR can thus be firmly positioned at the centre of a significant shift in sociolinguistic thought, one which has emphasized the need to rethink established monolingual models of language and the speaker by, for instance, bringing to the fore concepts such as translanguaging (García 2009). Despite its significance in this respect, however, CEFR has also garnered significant criticism (e.g. Shohamy 2011; Pilkinton-Pihko 2013) for failing to incorporate plurilingualism and pluriculturalism more explicitly into the reference levels which have since become its most prominent part, a shortcoming addressed with the recent publication of a set of new reference level descriptions for mediation and plurilingual and pluricultural competence (Council of Europe 2018).

The impact that CEFR has had at the national level in Europe is significant, though this success may not only be attributable to the influence of the Council of Europe but also to the policy actions of the European Union (EU). Almost immediately after the publication of CEFR, the European Council adopted as part of its resolutions from the 2002 Barcelona summit the policy recommendation that all EU citizens should learn two foreign languages in addition to their first language and also called for the ‘establishment of a linguistic competence indicator’ (European Council 2002, p. 19). It is thus unsurprising that CEFR was soon adopted by the EU and recommended to its members (Jones and Savile 2009), becoming thus associated with a conceptualization of language learning that generally foregrounds its economic rather than cultural benefits (Krzyzanowski and Wodak 2011). The joint influence of both CoE and EU has meant that CEFR has had a broad impact on language education policies across Europe, having in particular a central role in the reform of curriculum development and assessment practices in different nations. What these applications (described in edited collections such as Byram and Parmenter 2012) have highlighted is the flexibility with which CEFR is interpreted by its users in different contexts, a pattern aligned to the intentions of CoE that the framework be used in a flexible and descriptive rather than a rigid and prescriptive manner (Trim 2012).

While the ‘European’ trajectory of CEFR is in itself complex, a study focussing on how the framework impacts policies outside Europe needs to take into account a further type of user of the framework, namely major ELT textbook producers and testing organizations. Since its publication, CEFR has seen increasingly broad use by organizations such as Cambridge English and the British Council as a means of describing the difficulty levels of exams and textbooks, with the framework thus to an extent displacing the previous system of labels like ‘basic’ and ‘lower intermediate’. A significant body of research has developed around such uses, with a variety of studies reporting on the different procedures relied on to achieve alignment between the relatively open-ended descriptions provided by CEFR and the more detailed specifications used in the design and evaluation of test tasks (e.g. Martyniuk 2010). Through this connection, CEFR has come to be associated with an agenda whose key presupposition regarding how language is to be taught and learned are potentially contradictory to the concept of plurilingualism that the framework was intended to advance, with international English exams and textbooks continuing to largely be associated with a monolingual learning model directed toward the acquisition of native-like proficiency in standard English (Hamid 2014; Shohamy 2011).

It is against this history of antagonism that the appropriation of CEFR beyond European borders takes place, though there is as yet little literature examining the ideological underpinnings of such uses. It is perhaps indicative that while some examples of such uses have sought to apply CEFR in explicit support of plurilingual language policies (see e.g. Arnott et al. 2017; Piccardo 2014), most applications have focussed specifically on English education (Byram and Parmenter 2012). The framework has also seen some adaptation, particularly in Asia, with localised versions having been developed for ELT purposes in Japan (Negishi 2012) and China (Jin et al. 2017). The focus of these uses and adaptations on ELT suggests a shift with regard to the ideological underpinnings of the framework, but raises questions regarding the extent to which different features of the discourse surrounding CEFR are recontextualised to the local scale and how they interact with local discourses. These questions are addressed in the following sections.

Methodology for examining CEFR in Thailand and Malaysia

Context and data

This paper presents the results of two case studies of CEFR recontextualization at the local scale. The two case studies, Thailand and Malaysia, exhibit significant parallels: The nations are in the same geographical region, share a border and are also embedded in the same process of regional integration through their membership in the Association of South East Asian Nations (ASEAN), which has gradually seen a shift toward greater mobility in the region. Both are also categorised as nations with developing economies, typical of the region but atypical with regard to context where CEFR had been developed. The way in which the framework has been rolled out in the two also appears to bear many similarities, with the process starting almost concurrently (2013 in Malaysia, 2014 in Thailand) and with several policy texts produced since.

For the purposes of this research, recent educational policy texts in which CEFR was recontextualised were collected, which yielded a data set of nine texts, four from the Thai contextFootnote 1 and five from the Malaysian (see Table 1 for a full list). A broad categorisation of the documents according to form and function indicated that various types of comparison were possible:

Table 1 Overview of analysed policy texts
Full size table
  1. (a)

    During the time frame under examination, the government of each nation had drafted and published a macro-level education strategy. The two texts in question, M-Blueprint and T-Plan, exhibit numerous similarities and parallels and therefore allowed for closest comparison.

  2. (b)

    Each government also produced meso-level policy texts more specifically aimed at English language education, to which CEFR was seen to be relevant. Here, however, significant differences are to be found—while the Malaysian government produced strategies similar in structure to M-Blueprint, the Thai government produced a localised version of CEFR. Here, direct comparisons were less feasible.

  3. (c)

    In each context, a pair of teaching manuals was also created to facilitate the implementation of CEFR. These four texts again bear many similarities and allowed for a direct comparison.

In summary, significant parallels as well as contrasts were observed when examining the data available from both contexts. While differences in how processes of policy development and appropriation unfold in different contexts are natural, a significant departure in this case is in the timelines followed. In the Malaysian case, the policies were developed in a linear sequence, with the top-level strategy published first and the lower-level documents coming after. In Thailand, however, the sequence was inverse, with the teacher manuals being the earliest publications and the overall strategy coming last.Footnote 2 The main effect of this was that the texts contrasted in their use of intertextuality—while the Malaysian texts were generally presented as a top-to-bottom chain, the Thai documents were generally independent of each other.

Analysis

The framework guiding the analysis of the policy texts was the discourse historical approach to critical discourse analysis (DHA; see Reisigl and Wodak 2015). In line with DHA, I examined each text with regard to the discursive strategies used within it, by which I refer to “more or less intentional plan[s] of practice […] adopted to achieve a particular social, political, psychological or linguistic goal” (Reisigl and Wodak 2015, p. 33). In this analysis, discursive strategies were described in terms of the different nexuses they formed between particular concepts, seen as particularly meaningful ideas which become associated with discursive struggle and attain considerable mobilizing power (Koselleck 1982, 2004). In this way, the analysis drew on the analytical framework of conceptual history (Ger. Begriffsgeschichte; Koselleck 2002) and its previous applications in DHA research on policy (Krzyzanowski 2016; Krzyzanowski and Wodak 2011). In line with Begriffsgeschichte, attention was paid to the identification of key concepts in each set of policies and to the establishment of semantic relationships between them. In particular, the following types of relationships were examined:

  1. (a)

    Key socio-political concepts (Ger. Grundbegriffe) which act as ideological cornerstones in a given discourse;

  2. (b)

    Neighbouring or complementary concepts (Ger. Nebenbegriffe) which facilitate the creation of ideological-conceptual networks;

  3. (c)

    Oppositional or counter-concepts (Ger. Gegenbegriffe) which illustrate how the boundaries of ideological-conceptual networks are drawn through exclusion.

While such an approach allows for the identification of key concepts in texts and discourses, it is important to point out that any conceptual relations encoded in text are context-bound and thus likely to shift once that text is transferred to another context, either because the text is thus integrated into discourses where particular concepts have attained a different meaning or because recontextualised concepts are hybridized as they enter into new relationships of complementarity and/or oppositionality (Wodak and Fairclough 2010). Thus, the identification and examination of concepts in this study involved reference to how their meaning was defined by different levels of context. Specifically, the meaningfulness of concepts was examined with regard to (a) intra-textual relations, (b) inter-textual (discursive) relations, (c) features of the situational context in which the text was produced, i.e. the fields of (English) language policy in both Thailand and Malaysia, and (d) the broader socio-political and historical context that determines relations of power in those fields (Wodak 2008). Given the fact that the analysis compared policy environments with often differing features and contrasting histories, such a context-aware approach was vital to avoid oversimplifications or overgeneralisations, with the awareness that the meaningfulness of a concept may vary according to the context in which it is used.

Recontextualisations of CEFR in Thai and Malaysian language policy

Contextualizing CEFR: the bilingual agenda

An important early step in analysing how a globalized language policy is recontextualised into a particular local context is to examine how present-day language policies in that context are embedded into a broader historical trajectory. In the case of Thailand and Malaysia, such an examination foregrounds both points of comparison and difference with regard to how policies have related to the language ecology in either context. A key parallel is the countries’ linguistic diversity, with both being highly linguistically complex with regard to the number of distinct languages spoken and in terms of the variation that may be found within larger languages. In Thailand, for instance, a plethora of indigenous and immigrant languages (e.g. Chinese dialects, Pattani Malay, Burmese, Khmer, Hmong, etc.) is spoken alongside Thai, which is itself highly differentiated, with the standardised Central variety, which bears great influence as a result of its symbolic association with the state, being distinct from the regional languages (dialects) spoken in the periphery (Kosonen 2017; Premsrirat 2011; Smalley 1994). A perhaps even greater level of variation may be found in Malaysia, where indigenous languages (such as Sama, Murut and Thai) are outnumbered by languages linked to historical immigration (Chinese dialects, Tamil). At the same time, the national language, Bahasa Malaysia (below: BM), was only standardised in the aftermath of independence in 1957, with its power being relativized by its perceived lack of neutrality from the perspective of non-Malays, in particular the economically powerful Chinese and Tamil communities (Coluzzi 2017). An additional challenge to the dominance of BM is the continued prominence of English (Gill 2014), both in its standard and indigenized variety (known as Colloquial Malaysian English or ‘Manglish’).

The differing status of the national language and English in the two nations is a consequence of their recent histories and in particular their contrasting positions vis-à-vis the British Empire. Malaysia was gradually colonized by the British from the seventeenth century onward and was therefore governed by an English-speaking elite until relatively recently (Pennycook 1994), which meant that there is no long-standing tradition of systematic use of local languages for official purposes (Gill 2014). Thailand, on the other hand, retained political independence throughout this period, being the only nation in the region not to come under either direct British, French or Dutch rule, a fact exploited by local elites for systematic promotion of centralised ‘national’ authority (Anderson 1998). As this was built on a single national language, little space was left for English to have any official role. However, while this distinction appears clear-cut with regard to Kachru’s (1985) outer and expanding circles, it is in fact highly problematic. Anderson (1998) for instance argues that Thailand was de facto colonised from an economic perspective if not politically. Indeed, English has long been part of the language repertoires of the Thai elite despite its unofficial status (Diller 1988; Sukamolson 1998; Wongsothorn et al. 2002). Conversely, while Malaysia would nominally be classified as an ‘ESL’ nation in Kachru’s model, the diffusion of English among its population was historically low, being limited mainly to locals working in the colonial administration (Pennycook 1994), with little or no proficiency in the language continuing to be typical outside major urban centres (Gill 2014).

Such historical contrasts and similarities provide a broad frame of reference for the interpretation of present-day policies in either context. In Malaysia, the lack of diffusion of English, coupled with tensions between indigenous Malays and immigrant Chinese and Tamils, has produced a number of policy shifts. After independence in 1957, BM was made the national language and gradually replaced English in official functions. While ethnic nationalism thus led to English diminishing in status, recent decades have seen attempts at reversing this policy. In 2002, it was announced that English would replace BM as the medium of instruction in mathematics and science classes at all levels of education, a policy later re-reversed in 2012 after a series of issues hampered its implementation (Gill 2014). This did not, however, signal a shift away from English, as the adoption of a new language policy agenda, named MBMMBI (BM: Memartabatkan Bahasa Malaysia dan Memperkukuhkan Bahasa Inggeris, English: ‘To Uphold Bahasa Malaysia and to Strengthen English’), was announced in the same period. As suggested by its title, this policy sets as its objective the establishment of a bilingual society, the rationale for which also can be seen in the policies analysed as part of this study:

In general, the Ministry has three goals for the learning of languages:

  • Fostering a unique shared identity between Malaysians anchored in the ability to be proficient in the use of a common national language, Bahasa Malaysia;

  • Developing individuals that are equipped to work in a globalised economy where the English language is the international language of communication; and

  • Providing opportunities to learn an additional language. (M-Blueprint, pp. 4–10)

These points are indicative of the language hierarchy constructed by M-Blueprint, a broad government strategy published in 2013. Reflective of historical tendencies (see above), BM was positioned at the top, being exclusively granted the label of ‘national language’, and was associated with concepts like ‘identity’ and ‘uniqueness’, as well as ‘unity’ and ‘nation-building’ elsewhere in the document. The political and cultural nature of how BM is constructed through these concepts stands in contrast to the way in which English, as the second language in the hierarchy, was positioned. English was nearly exclusively constructed as a language of international (rather than national) communication and a (utilitarian) means of achieving economic success, one which bears no significance to identity or culture. At the bottom of the hierarchy is the open-ended category of ‘additional language’, which in fact consisted of a mix of different languages. In some parts of the text, the label was used to refer to the languages of other ethnic groups, in particular Chinese and Tamil. Elsewhere, however, this term was also used to refer to the learning of other world languages such as Spanish, French and Japanese, while also being used to refer to languages of neighbouring countries. The vague nature of this term reflects the extent to which languages other than BM and English are backgrounded throughout the document, a position compounded by references to ‘bilingual proficiency’ as an objective in which no space at all is left for other languages.

Just as in Malaysia, the position of English in language policy in Thailand can be seen to have undergone a series of shifts. While English has been present in Thai education in various ways since the nineteenth century, it has drifted in and out of favour since the 1950s, being at times a required subject for all and at others an elective alongside other European languages, such as French or German, or Asian languages, such as Chinese and Japanese, with Arabic and Pali also offered (Sukamolson 1998). Broadly, however, policy in recent decades have seen the importance of English increase to the extent that it is now de facto the only foreign language most students learn (Baker and Jarunthawatchai 2017). The policies examined in this study reflected this orientation, with English education being extensively referenced as a key objective:

The governmental organizations and related sectors must therefore help develop frameworks and direct the production and development of human resources in different fields so that the country will have the right people for the right job in the market for national development. The curricula for different levels that can give learners skills needed for the 21st century world, especially English, science, and digital skills, should be improved. (T-Plan, p. 100)

Support people of all ages to be able to read and write Thai, their indigenous languages and languages of neighbouring countries. (T-Plan, p. 98)

These extracts illustrate the association between languages and concepts in T-Plan, an educational strategy published by the Thai government in 2017. The pattern most evident is the relatively intensive conceptualization of English, which featured prominently throughout the document, with conceptual associations often mirroring those observed in M-Blueprint. In particular, English was again exclusively constructed through an economic prism, being positioned in this case as one of the ‘twenty-first century skills’ central to the ‘development of human resources’ and being associated with the highly influential concept of ‘national development’ (Hill and Fujita 2012) and elsewhere in the text with the concept of ‘Thailand 4.0’, central to the economic policy of the government under which the policy had been created (Jones and Pimdee 2017). This stands in contrast to the few references made to other languages, which also saw few conceptual associations. Thai saw little attention in the policy, a clear contrast from the intensive conceptualization of BM in M-Blueprint, though this disparity can be seen as a reflection of the unchallenged status of Thai as the national language when compared to BM, which is not only a relatively newly standardised variety but also a code embedded in ethnic struggle (see above). However, the lack of attention to other languages used in and around Thailand, referred to generically throughout the policy, is a reflection of their generally low status (Premsrirat 2011), indicating the existence of an implicit bilingual agenda similar to that explicitly adopted in Malaysia.

Recontextualizing CEFR: a selective endeavour

It is as an integral part of the ‘national language plus English’ bilingual agenda that CEFR was recontextualised into both the Thai and Malaysian context. In both sets of documents, CEFR was primarily positioned as an instrument of English language teaching and learning and was generally not associated with learning other languages. As outlined in the previous section, the focus on positioning English as the dominant second language to be taught in schools is an established part of Thai and Malaysian language policy and also broadly resonates with tendencies observed in the region (e.g. Baldauf et al. 2011; Kirkpatrick 2010, 2017). Comparing this orientation to CEFR, however, indicates a significant reconceptualization, since not only is the framework not intended to be tied to a particular language, it includes a remark that a way of achieving multilingualism might involve “reducing the dominant position of English in international communication” (Council of Europe 2001, p. 4).

Though such a tight conceptual nexus between CEFR and English may appear potentially contradictory when considering the plurilingual/pluricultural agenda promoted by the Council of Europe, both Thai and Malaysian documents avoided such incompatibility by selectively recontextualizing parts of the framework. As discussed above, CEFR has been critiqued for, among other reasons, failing to completely integrate the two key concepts of plurilingualism and pluriculturalism into its different elements. Indeed, among its chapters, only the introduction (pp. 1–8) and those outlining the framework’s broad vision of language learning and teaching (pp. 131–156) and the development of curricula (pp. 168–176) consistently reference plurilingualism/pluriculturalism, whereas the concepts rarely, if at all, appear in the chapters where reference levels are described (pp. 43–130) and where task-based teaching (pp. 157–167) and assessment (pp. 177–196) are described. The latter elements, in particular the reference levels, are instead based on a monolingual construct of language proficiency (Shohamy 2011).

When considering the recontextualization of CEFR in terms of how these different elements were foregrounded and backgrounded as part of its transfer into the Thai and Malaysian contexts, a clear pattern emerges. In the policies examined, the CEFR reference levels were often the only element referred to explicitly while other elements of the framework were generally backgrounded, with plurilingualism and pluriculturalism receiving no mention in any of the documents examined. The Malaysian English language education strategy M-Agenda, for instance, made no reference to the Council of Europe nor to its language policy, with CEFR instead typically positioned as simply ‘international’:

What we need to create is a programme that is simultaneously international, because it is aligned to international standards as represented by the CEFR, and national, because it is carefully tailored to the specific needs of Malaysia. (M-Agenda, p. 14)

This quote is significant because it illustrates the extent to which the foregrounding of the levels was naturalized in these documents, with the meaning of ‘the CEFR’ here implicitly being narrowed down only to its reference levels, since it is to these that a test or curriculum may be seen to be ‘aligned’. Such a focus on the CEFR levels was also evident elsewhere in the Malaysian data: In M-Blueprint, for instance, parts of CEFR pertaining to the reference levels were wholly recontextualised despite the relatively broad nature of the document, with CEFR-specific terminology (‘operational’ and ‘independent proficiency’) used when setting objectives. Such references to CEFR-specific language were also common, with both M-Agenda and M-Roadmap including numerous references to the ‘can do’ approach of the framework. In documents more specifically aimed at implementation at the local scale, namely the two sets of teaching manuals, CEFR was referred to in more detail yet again narrowed in scope to only its six reference levels. Thus, Thai teaching manuals (T-PrimMan and T-SecMan) included copious reference to both the global level descriptions of CEFR and the more specific skill-by-skill ‘illustrative’ descriptions, which were presented in translation according to the level dictated by policy as being relevant (e.g. A1 for final year of primary school). Similar patterns of recontextualising CEFR may also be found in Malaysian documents aimed at teachers (M-PrimMan and M-SecMan).

The focus on levels is also evident when considering FRELE-TH, a unique text in this dataset in that it does not constitute a policy drawing on CEFR but is instead a localised version of the framework produced for use in Thailand. Thus, its main equivalent is the only other ‘national’ CEFR adaptation at the time, the Japanese CEFR-J, to which FRELE-TH also bears greatest similarity in the sense that its key feature is a vertical expansion of the levels. In CEFR-J, this expansion created 12-levels in place of the original six, which was achieved by creating a new bottom level (Pre-A1) and splitting remaining levels (A1 into three sub-levels and A2, B1 and B2 into two sub-levels each). A similar approach was taken in FRELE-TH, in which the original six levels have been expanded to ten with the addition of four so-called ‘plus-levels’ (i.e., A1+, A2+, B1+, B2+).Footnote 3FRELE-TH thus features revised versions of 33 reference level descriptions—all those in the original CEFR and one additional (‘Reading Literature’) drawn from the EAQUALS Bank of Descriptors. It also includes a level-by-level description of language and content topics appropriate for each level (developed on the basis of the Core Inventory for General English) and a vocabulary database (developed from the Word Family Framework). However, while all these resources mark it as a significant adaptation, it is notable that they are all tied to the levels, which were in the case of FRELE-TH presented largely in isolation, with the brief accompanying text again making no reference to plurilingualism or pluriculturalism.

A key feature of how CEFR was integrated into the examined policies was thus the decontextualisation of its six levels, with key accompanying concepts seeing little uptake among policymakers. At the same time, however, CEFR was also placed into new conceptual relationships. Above, I outlined the broad language policy agenda into which they CEFR was integrated, one which sets societal bilingualism in the national language and English as its main objective. However, alongside its contextualization in a discourse about language policy, CEFR was also embedded into a broader discourse about education, one not necessarily centred on language. The key characteristic of this discourse, evident in T-Plan and M-Blueprint as the two overarching strategies, was their reference to instruments of transnational governance in education like the Programme for International Student Assessment (PISA), administered by the Organization for Economic Cooperation and Development (OECD). The policy impact such instruments, PISA in particular, have at the national level is significant (Grek 2009) and it is thus unsurprising that league tables based on them featured prominently in both T-Plan and M-Blueprint. In both cases, aspirational goals were also set, supported not by context-specific analyses but largely on transnational comparisons. Thus, M-Blueprint highlighted traditionally high-performing nations like Singapore and South Korea, describing them as those which ‘Malaysia seeks to compete against in today’s knowledge economy’ (pp. 3–9).

Such comparisons also played a key role in T-Plan and it is here that an extreme example of how CEFR fits into this PISA-centric discourse may be found. Table 2 presents part of a section in which key indicators of ‘Quality’ were presented in reference to the 20-year plan outlined by the document. Here, future policy targets were set according to 5-year periods, with expectations regarding students’ English ability expected to progressively rise until the target proficiencies were set at B2 for both junior and senior high school students and C1+ for university students.Footnote 4 What is most significant about this table, however, is the juxtaposition of CEFR and PISA, which is an indicator of how CEFR levels may be interpreted when combined with elements of a global discourse about education. When juxtaposed with PISA, CEFR is positioned as an objective international standard which drives policy change by forcing comparisons between ‘high-‘and ‘low-performing’ educational systems on the basis of the CEFR levels they set. While this is an extreme example unique to T-Plan, it should be remarked that the consistent positioning of CEFR as an ‘international standard’ to which an educational system must be ‘aligned’ (exemplified by the above extract from M-Agenda but ubiquitous across the data-set) can also be broadly seen as reflective of such a perception.

Table 2 Juxtaposition of CEFR and PISA (T-Plan, pp. 83–84)
Full size table

CEFR in the hands of global and local agendas: ideological struggle or symbiosis?

As indicated in the previous sections, the recontextualization of CEFR in Thai and Malaysian language policy often followed parallel trajectories. Such textual trajectories are, as pointed out, determined by a variety of factors, such as the balance of power between actors in a particular nexus of practices or by the ideological struggles that characterise many discourses. Such struggles could also be observed when interrogating the ideologies underlying the interpretation of CEFR in Thailand and Malaysia, with two ideologies in particular vying for dominance.

Through its association with English, CEFR could be seen as a vehicle of neoliberal ideology in language policy, one which views language teaching and learning, and indeed all education, as an instrument of workforce production and profit (Ng 2018). In such a neoliberal imaginary, individuals positioned as agents whose language learning goals are driven by their quest for achievement and their need to remain competitive in a job market defined by flexibility and mobility (Block et al. 2012; Flores 2014; Kubota 2014, 2016). This orientation can be seen with regard to how English was positioned in the Thai and Malaysian policies above, where it is consistently presented in close association with concepts from the economic field such as ‘employability’ and ‘skills’, being in this way clearly differentiated from other languages in its ecology. The construction of such language hierarchies is a key part of such a neoliberal agenda in language policy, with languages with perceived high value for employability being prioritised over those which are deemed to have low value (Lorente and Tupas 2013). Most often, the ‘high value’ language favoured by such an agenda is English, with local languages in contrast facing exclusion (e.g. Kirkpatrick 2017).

A further characteristic of neoliberalism in language policy is its focus on the transnational scale, particularly on the establishment of transnational regimes of governance (De Costa et al. 2019). Here, CEFR plays a key role, since it does not only mediate local developmental agendas but has also been appropriated by powerful global institutions which seek to profit from its implementation. Here, I refer to institutions like Cambridge English and the British Council, which are seen by Phillipson (2010) as agents of ‘linguistic imperialism’ (see also Pennycook 1994) and which Block et al. (2012) collectively refer to as the ‘ELT industry’. As outlined above, CEFR has seen increasing use by the ELT industry in the production and marketing of tests and textbooks, and it is in part through these products that the conceptual nexus between CEFR and English has been reinforced. Indeed, the close association between the ELT industry and CEFR has had practical implications for how the framework has been implemented at the local level in Thailand and Malaysia. In the former, the British Council has run workshops in support of the rollout of the framework since 2014. In the case of the latter, the implementation of CEFR has had an even more pronounced effect, with locally-developed materials being replaced in 2017 by CEFR-aligned global textbooks—produced by MacMillan and Cambridge, the latter of whom had previously conducted an influential CEFR-based assessment of the Malaysian educational system (Cambridge English 2013).

While CEFR may thus be seen purely as an instrument of imperialism, a conclusion potentially reinforced by the way in which the policies examined in this research juxtapose the framework to other instruments of transnational governance like PISA, such a purely structural view only tells part of the story. In particular, it ignores the way in which CEFR also perpetuates local nationalist agendas by solidifying the position of English as a second language and, by extension, pushing codes other than the national language (Thai or BM) further down the pecking order. As argued by Kirkpatrick (2017), it is this rearrangement of the linguistic hierarchy in favour of the national language and to the detriment of local languages that is a key consequence of the growing focus on English in South East Asia. This indicates that while the transnational focus of neoliberalism means that it is often pitted in opposition with nationalism, being seen as endangering a community’s homogeneity or indigeneity, the agendas of both ideologies may also co-exist and develop synergies.

The ability of CEFR to facilitate such a synergy in the Thai and Malaysian contexts may be seen as a logical extension of the framework’s previous trajectory. While CEFR may have been produced under the auspices of the Council of Europe, thus representing the humanist agenda of that organization, its relatively rapid adoption by the European Union meant that it was soon embedded into a much more complex discourse, one in which the cultural agenda in language policy is often marginalized in the face of neoliberal arguments focussed on promoting multilingualism as a means of economic mobility (Krzyzanowski and Wodak 2011). At the same time, the ‘European’ legacy of CEFR is also rooted in local concerns over the influx of migrants, with the motivation underlying the development of its earliest precursor in the 1970s being directly linked to nationalist calls for the linguistic integration of migrants (Tabouret-Keller 1991). It is notable that the framework has until today also continued to mediate nationalist agendas in the European context, being regularly used as a means of gatekeeping in citizenship testing, for example (Extra et al. 2009).

In conclusion, what does this trajectory tell us about the nature of CEFR and other globalized language policies? Perhaps their defining characteristic is that they are either by nature able to mediate multiple agendas at different scales or are sufficiently open-ended that actors are able to twist them according to their needs. In this respect, globalized language policies come close to what Holland terms foundational documents, texts which ‘are essential in defining larger religious, political and social agendas’ (2014, p. 386). By using the example of Quotations from Chairman Mao Tsetung, also known as The Little Red Book, Holland shows how the broad social impact of such documents is, while intertwined with their contents, also heavily dependent upon the ways in which particular fragments are recontextualised. This study suggests that a similar approach may be taken when examining globalized language policies, with the contents of the text playing as crucial a role in determining their meaningfulness as the ways in which the text (or fragments of it) are recontextualised and incorporated into new texts and discourses.