Repository logo

English medium instruction policies as a form of symbolic control: Making visible the changing faces of language and content in higher education



Change log


Chang, Sin-Yi 


With the growing trend of internationalization in higher education, the rising status of English has drawn much research attention in recent years, as manifested in the promotion of English medium instruction (EMI) policies in many non-Anglophone contexts (Dearden, 2014; Macaro, 2018; Pecorari & Malmström, 2018). While a sizable body of EMI literature can be identified, the lack of systematic analysis between the macro and the micro has led to a number of problems, including domain loss, detrimental effects on subject learning, in addition to other social concerns of equality and access. As the language of instruction plays an important role in shaping what is re-contextualized and reproduced, there is an urgent need to examine the changing processes of organizing, transmitting, and acquiring knowledge in EMI contexts. In other words, this means to carefully trace what is intended, enacted, and learned when the medium of instruction transitions from one language to another.

To do so, in this study I position EMI as a language policy with educational consequences. Specifically, I take a social realist approach by integrating Spolsky’s (2004) theory of language policy and Bernstein’s (2000) conceptualization of the pedagogic device. The central questions that I ask in this study are: How does EMI policy influence curriculum-planning, teaching, and learning within the higher education pedagogic device? What are the changing faces of language and content under such policy influence? Methodologically, I frame my research as a qualitative case study to investigate the educational outcomes of EMI at one university in Taiwan. In particular, the case is not the university itself, but the institutional policy that promotes EMI. The participants include 17 university lecturers and 37 students across disciplinary communities, as well as 6 policy-makers and administrators with varying levels of authority. Data collection methods encompass curricular document analysis, classroom observations, semi-structured interviews, and focus group interviews. This amounts to approximately 75 hours of audio data and field notes obtained over a period of eight months.

The findings of this study are divided into three parts that look into what EMI means in the curriculum, classroom, and individual learning experiences of the students. I first explore the different roles of English on each level, and then examine how the diversifying roles of English have altered ways of organizing, transmitting, and acquiring knowledge respectively. By theorizing EMI policy as a form of symbolic control, I attempt to understand how the conceptualization of knowledge is redefined in EMI contexts. As an overview, the findings showed that internal epistemic rules and external market forces rivaled for the control of the pedagogic device, with the latter occupying the higher ground. This generated a conflicting picture: on the one hand, EMI was strongly desired for a range of material and/or symbolic benefits (language-related); on the other hand, it also struggled to live up to its name when the teaching and learning of disciplinary knowledge faced increasing risks (content-related).

At this crossroad, the solution to the dilemma may not simply be to reverse the EMI trend – as this is unrealistic and against the wishes of many who are involved – but to ensure the quality and success of EMI by retaking control of the device: to bring knowledge issues back to the center of thinking about EMI, putting education as priority instead of external regulations that may or may not be educational. In an effort to resolve “the great ambivalence toward EMI” (Tsou & Kao, 2017, p. v), I introduce a new language-content model – a dynamic version of the language-content continuum that is increasingly referenced in EMI literature (Airey, 2016; Lyster & Ballinger, 2011; Macaro, 2018; Met, 1998) – to help with curriculum-planning, teaching, and learning. The model emphasizes different degrees of integration based on what is controlled and how much control is taken. Indeed, content and language are naturally blurred in reality in many ways, but I argue that understanding both the intersections and internal fences between the two is necessary before considering the possibility of meaningful integration, especially when issues of proficiency are of major concern.





Liu, Yongcan


English medium instruction, Language policy, Higher education


Doctor of Philosophy (PhD)

Awarding Institution

University of Cambridge