Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

This book is based on the critical analysis of publicly available policy documents relating to international students. I understand policy as discourse, drawing on a Foucauldian understanding of discourse. The approach to analysis has been informed by Carol Bacchi’s “what is the problem represented to be” framework.

Bacchi ’s framework is grounded in discourse theory, heavily informed by Foucault . Discourse theory posits that through both language and knowledge, discourses represent, structure and imagine the world, changing it in line with particular ideologies (Fairclough 2003). Rules and regularities develop which create a code of knowledge about a subject (Thomas 2013), often around the people, theories, systems and techniques for defining and acquiring knowledge (Rose and Miller 2008). They define and “police the boundaries” of acceptable statements and debate (Devos 2003, p. 157). Discourses are dynamic, applied, interactive social processes of production and reproduction of knowledge and reality (Fairclough 1989). These affect what it is possible to say about an object, and consequently, discourses can be understood “as practices that systematically form the objects of which they speak” (Foucault 1972, p. 54). The material object, its “ground”, is therefore not the focus of analysis because the discourse in question actually creates the reality it talks about.

A discourse is a collection of formations, practices and events within which “a group of statements…constituted its object” (Foucault 1972, pp. 35–36). Discourse comprises both the language and knowledge about an object and, therefore, establishes the logic and rules for that which is possible to be said (Foucault 1965, 1972). Discourse goes beyond describing reality: by “enabl(ing) and constrain(ing) the imagination and social practices” (Sidhu 2006, loc944), it helps to create and constitute reality. There may be multiple or even contradictory understandings of an object, for discourses are not homogenous, but make meaning in social contexts (Bacchi 2000; Iverson 2007).

Therefore, the notional object is problematic. What is called, for example, “madness” may in different eras refer to substantially different understandings, and further, the same term may be used in different discourses (juridical, religious, etc.) in the same era with different points of reference (Foucault 1965). Miller and Rose (1990, p. 5) emphasise discourse as technologies of thought. Knowledge of an object requires inculcation in particular procedures and techniques, such as statistics, experimentation, and so on, such that objects are talked about in particular ways which enforce power. This limits the potential for objects to be talked about or known differently. Discourse shapes epistemology, by determining socially which objects are appropriate focuses of activity, and how they can be known. As social practices, discourses are culturally conditioned tools for thought, which make it impossible to escape the “web of beliefs” (Moscovici 2000) without accessing shared knowledges or to speak from outside a discourse (Foucault 1972). But these discursive representations are real, not illusory, inasmuch as they are a shared system of social practices. In other words, the discursive representation has a reality distinct from but not independent of the object it purports to represent.

Problematising

A discursive approach to policy investigates how certain topics become policy objects (Foucault 1972), for which problematisation is a useful concept. Problematisation explores how something becomes “an object of concern, an element for reflection, and a material for stylization” (Foucault 1988, p. 24). Discursive formations such as policy are understood as social practices characteristic of particular times and places, so discourses and social representations can be deprived of their common-sense status (Foucault 1982; Fairclough 1989; Filippakou 2011), because they have been different in other times and places. The goal is to look for the “rules by which a particular statement has been made” (Foucault 1972, p. 30), identify how it excludes other statements of possibility and “examine the interplay of (the) appearances (of concepts) and dispersion” (Foucault 1972, p. 37). Problematisation begins from the premise that the nature and content of discourses, and of policy discourses in particular, could be different, that their form and substance are not inevitable or natural. Nor, however, are they arbitrary, for their nature and content reveal power dynamics. Instead, they are determined by normative frameworks.

The construction of a “problem ” is a particular characteristic of policy discourses (Bacchi 2009), wherein it becomes a real object. Governmentality and policy studies often focus on problems for problematisation (Rose and Miller 2008; Bacchi 2009). Government involves the creation of problems, and their solutions, which confers legitimacy on the ruler; the agent who identifies or names the problem positions themselves as having the power to solve it (Saarinen 2008b). Bacchi (2009) argues that the problematisations embedded in the policies reveal the mode of governance . Problems, as represented in policy, rely on particular assumptions of knowledge and reality, which can be challenged and contested: “what starts out as claim comes to be transformed into a matter of fact” (Rose and Miller 2008, loc1473). The way that problems and solutions are framed and represented is indicative of the logics of governance. For instance, Foucault distinguishes between sovereign and disciplinary power, where the former uses pomp and ceremony to rule, and the latter uses techniques of surveillance and discursive normalisation (Foucault 1977). Typically, modern modes of governance are hybrid, employing both sovereign and disciplinary modes (Bacchi 2009). In policy, it is primarily the disciplinary mode which is of relevance.

Hence Bacchi ’s (2009) “what is the problem represented to be ” (WPR ) approach (Fig. 4.1) provides a framework of questions to structure an analysis of discursive problematisations, beginning with “what is the problem represented to be”. Often the problem may be implicit and may be read back from the solution presented, to explain why certain things are thought and how these representations are created (Webb 2014). Problematisations are often plural or nested within a single policy and may be contradictory (Bacchi 2009; Webb 2014). In Bacchi’s framework, there is a double problematisation : firstly, the policy constructs the problem; and secondly, the analyst problematises the problem representation.

Fig. 4.1
figure 1

“What is the problem represented to be” (WPR ) framework (Bacchi 2009)

Full size image

Applied to the rationales for engagement in international education—economic , educational, sociocultural and political—Bacchi ’s approach supposes that the presentation of arguments in favour of a particular course of action can be read as solution, from which underlying problems can be inferred. Thus, the need to gain income through international recruitment implies a lack of money as a problem. The need to gain political influence through the agency of international alumni implies a lack of such influence, and so on.

Assuming

Dominant discourses can naturalise certain ideological assumptions as common sense (Fairclough 1989) because they limit and shape what can be imagined (Sidhu 2006). This leads certain discourses to become hegemonic, reducing the usual plurality of contradictory discursive alternatives (Foucault 1972). In van Dijk’s (1996, p. 85) words: “dominant groups or institutions may influence the structures of text and talk in such a way that as a result, the knowledge, attitudes, norms, values and ideologies of recipients are—more or less indirectly—affected in the interests of the dominant group”. This suggests that, while the plurality of discourses is important, studying dominant discourses, such as political and state discourses, is more likely to reveal imbalances of power .

Thus, Bacchi ’s Question 2 asks what premises or assumptions are required to accept this problem representation, what is taken for granted or common sense (e.g. Spanger 2011; Stevenson 2013; Lancaster and Ritter 2014). This explores the logic of the discourse, its judgements, reasoning, and necessary precursors (Foucault 1965). These premises are the background knowledge or beliefs that the reader must have to make meaning of the text (Saarinen 2008b; Loutzenheiser 2015), revealing the underpinning discursive structures which are shaped by governmentality (Bacchi 2009).

Public discourses such as mass media and policy texts often address an ideal subject or reader, forcing readers into a particular position or sharing assumptions to understand the text (Fairclough 1989; Saarinen 2008a). Implicitness can be used to create ideological common ground between the text producer and the reader (Bacchi 2009), reducing the space for disagreement or competing voices and reflecting existing power structures. The reader is incorporated at least temporarily into the discourse community of the policy text, for if they do not share in those assumptions, the text loses coherence. This is not to say that disagreement and rupture are impossible, but rather to highlight that the most powerful effects of policy discourses are likely to be those least spoken about for this reason—that disputing essential presuppositions causes the texts and actions to lose meaning. In particular, these may be found in specific understandings of social representations, relationships and narratives and can be operationalised by looking for keywords. Bacchi (2009) also identifies key “binaries”, oppositional dichotomies that underpin problematisations such “licit/illicit drugs”. These tend to simplify complex relationships or gradations by reducing them to binary categories which reveal the operation of conceptual logics.

In essence, this question operationalises Foucault ’s archaeological approach (1972), creating a window on discourse rules which determine what can be said. This involves exposing the metaphorical layers of concepts which have shaped how an object has come to be viewed, and what the conditions are that make the emergence of a policy problem possible (Gale 2010).

Silencing and Excluding

Access to powerful, dominant discourses is limited, such that those with political power can define the discourses of the state (van Dijk 1996). This means that the “idealized schemata” created will necessarily include certain dimensions and exclude others, along lines which sustain the interests of the dominant group (Foucault 1972; Rose and Miller 2008). Power is not unilaterally exercised, however (Sidhu 2006). Instead, it is deployed through a series of routine micro-practices, in a heterogeneous range of institutional contexts. Where contributions to the discourse are made by less powerful participants, they are shaped by more powerful participants (Fairclough 1989). Indeed, such contributions can only be made in adherence to the rules of the discourse. This means that the content, relations between concepts and subject positions will be primarily defined by the most powerful participants. Silences are created in these exclusions: “The manifest discourse, therefore, is really no more than the repressive presence of what it does not say; and this ‘not-said’ is a hollow that undermines from within all that is said” (Foucault 1972, p. 28).

Bacchi ’s Question 4 asks what is left unproblematic, not discussed or could be thought about differently (Bacchi 2009). Exposing silences shows what and who is marginalised in the process of policy creation and text production, and alternative ways of knowing (Tikly 2004; Taylor 2004; Spanger 2011; Stevenson 2013). Silences may be issues not discussed, often to do with inequalities or power relations, and particular subjects or indeed different discourses (Spanger 2011). Powerful discourses will tend to silence the discourses, assertions or representations of the less powerful (Lombardo and Meier 2009).

It is this focus on the power relations implicit within discourse which generates the critical potential of a discourse approach (Fairclough 1989; Foucault 1965). There is some disagreement about whether such criticism is intended to produce real-world changes. Foucault’s (1972) position is that it opens space for alternative ways of thinking and speaking, that it “seeks difference and complexity” (Webb 2014, p. 369), and given that discourses are real, that this constitutes real-world change. In this sense, policy activism can consist of reworking and reinterpreting texts strategically (Taylor 2004). In the idiom of policy and policy research, this lack of an “answer” may appear inadequate. However, it is consistent with the philosophical assumptions of a discursive approach wherein any attempt to provide a definitive account of policy, or how it should be, would necessarily be specious and partial. Instead, the creative critical potential of this approach is to throw the familiar practices and assumptions of policy into question, to open discursive spaces to alternative representations, and to reveal shared understandings and social practices as a precursor to developing ethical alternatives (Tikly 2004).

Creating People

Discourse constitutes the object (Foucault 1972), and for people, this means establishing subject positions , such that they can only take meaningful action within these positions. Categorical subject positions might include “victim” or “criminal”, “husband” or “wife”, “worker” or “manager” (Spanger 2011; Widding 2011; Svender et al. 2012). As Ball (1993, p. 14) puts it, “we are the subjectivities, the voices, the knowledge, the power relations that a discourse constructs and allows”. Because identities are embedded in discourses, people speak from social categories. Indeed, Hacking (1999) argues that the act of naming a group of people creates an identity for that group, which people come to fit. He suggests that developing a category (he gives homosexual as an example) causes bureaucrats and academics to recognise people who fit that category, where previously they would have described them differently (for example, as deviant). This “making up people” then defines what is possible for people to do, say and act out in their lives, because description allows action; if we can describe a thing or a person, it can be done. If it cannot be described, it cannot be conceived. Mass media can affect people powerfully because it creates through narrative formulae for how to live, the “habits of conduct” (Rose and Miller 2008, loc3098), which are internalised as standards, aspirations and habits. Policy discourses do the same, constructing images of political and social subjects embedded in power relations. This is Bacchi ’s (2009) Question 5: “what are the effects of this representation?” This question emphasises the discursive creation of subjects (Tikly 2004). Indeed, Loutzenheiser (2015, p. 107) revises this question to focus entirely on subjects: “who is the subject implied (to be)?”

Social categorisations represent objects and people conventionally, establishing a model or an ideal type for people to fit into (Fairclough 2003), which can marginalise them (Rose and Miller 2008; Van Leeuwen 1996). They set people in opposition to each other, or divide their own consciousness, an effect known as “dividing practices” (Foucault 1982). Many are “problem categories”, such as non-participants (Stevenson 2013), resistant or in deficit (Bacchi 2009). Such discursive representations discipline people by limiting possibilities for action and identity (Moscovici 2000).

They can also be discursively marginalised in other ways. People can be described generically or specifically or aggregated as statistics, a key mechanism by which people are rendered calculable and governable (Rose and Miller 2008). Counting reifies people, turning them from agents into objects. Further, people can either be active agents in grammatical terms or passive (Fairclough 2003). This is not necessarily intentional manipulation, indeed it is usually not, but it can have material and discursive consequences (Bacchi 2009), and tends to reinforce existing structures of power .

However, this discipline is never total, and there are always possibilities for agency and struggle (Foucault 1982; Moscovici 2000). While limiting in some ways, categorisation also paradoxically empowers us to act as social agents, by defining social practices we can perform, as well as those we cannot (Fairclough 1989; van Leeuwen 1996). A single individual may occupy multiple subject positions (Loutzenheiser 2015) and opt in or out of certain positions. People interact with these narratives and discourses creatively, recombining them innovatively, and overtly resisting them (Foucault 1982; Fairclough 1989; Rose and Miller 2008). “Individuals and groups, far from being passive receptors, think for themselves, produce and ceaselessly communicate their own specific representations and solutions to the questions they set themselves” (Moscovici 2000, p. 30). Because power in a Foucauldian sense is productive and diffuse, it can rest with the individual, offering them the agency to create positive outcomes (Sidhu 2006). Discourses therefore both empower and disempower individual agents, sometimes simultaneously.

Self-subjectification is a particularly powerful dimension to this process (Bragg 2007), meaning the acts of individuals to create themselves as social subjects (Foucault 1982), sometimes by conforming to the idealised expectation of the category or opposing them. Self-subjectification refers to:

those intentional and voluntary actions by which men (sic) not only set themselves rules of conduct but also seek to transform themselves, to change themselves in their singular being, and to make their life into an oeuvre that carries certain aesthetic values and meets certain stylistic criteria. (Foucault 1988, pp. 10–11)

In other words, people choose how to live their lives and how to define themselves, and do so in reference to particular values and norms. This process makes individuals responsible for their “choices”, discursively amplifying certain behaviours and minimising others (Rose and Miller 2008). In viewing oneself as a project, everybody is an administrator or regulator of their conduct and lives (Marginson 1997). Responsibility for who we are then falls on of the individual. Therefore, every moral decision is an instance of self-subjectification because it refers to the “unified moral conduct”, the broader social system of rules (Foucault 1988, p. 28).

Discourse is a particularly effective tool of power because it operates on cognitive and linguistic levels below consciousness (van Dijk 1996), enabling consent to be manufactured through definitions and limiting possibilities (Fairclough 1989). These indirect techniques of control mean that discourses can be internalised, affecting individual thought, action and self-identification (Lukes 2005). Subordinate or marginalised groups or individuals can train themselves out of desires and beliefs which fall outside the discursive norm for their group or identification. This has particular relevance in a neoliberal capitalist context, where the professionalisation and training of subjects for superior performance in the workplace becomes a “personal development ” project (Rose and Miller 2008). In this model, homo economicus is rational, making decisions as an individual, not in a social context, where the skills and knowledges they acquire are commercialised, comprising their “human capital ” (Marginson 1997). The augmentation of this capital entails “becoming an entrepreneur of oneself” (Tikly 2003, p. 164). However, this project of self-work means adopting and internalising the values and behavioural norms of free-market capitalism.

Although the structural focus of a Foucauldian approach may appear negatively deterministic, it is also critically productive and radical. It shows how power works through discourse in social representations, thus undermining power and opening up space for alternative representations and discourses, offering a tool for agency and resistance. Therefore, the discursive effects which may close off different options for agents and the power of dividing practices are significant and worthy of study (Bacchi 2009), which is the focus of this book. I acknowledge and value students’ agency, but it is not the object of study. Instead, I focus on the structures of policy discourses in the critical interests of exposing their effects.

Methods

The documents having been identified and selected on the basis described at the beginning of Chap. 3, the qualitative analysis was conducted using software to facilitate an inductive thematic coding . Relationships between these themes were organised around rationales for or against recruiting international students. These rationales constituted the starting point for Bacchi ’s (2009, 2012) “what is the problem represented to be” (WPR ) analysis. This framework was then applied to identify how international students are discursively represented in policy.

Documents were coded inductively, based on emerging themes (Braun and Clark 2006). In essence, inductive or open coding involves assigning codes as they emerge from the data, driven by data (Gibbs 2008). A software programme for qualitative data analysis, NVivo , facilitated the analysis, which enabled a large volume of documents to be included in the corpus: 125 documents in total were included, with over 3000 pages.

In the first stage of coding , after a preliminary reading for familiarisation, extracts were coded if they related to the research questions (Ashwin et al. 2015). Sentences relevant to the research questions were established by examining regularities, disjunctures and attention to discourse (Bazeley and Jackson 2013). The unit of coding varied according to the document, either line-by-line or a whole paragraph, depending on the document (Strauss and Corbin 1990; Saldana 2009). Line-by-line coding occurred in the thematically dense areas, which were also simultaneously coded where a particular sentence incorporated several themes. This lent a sense of the richness of the data (Bazeley and Jackson 2013). It was more common to code by sentences, but in some texts, the same code was applied to several sequential sentences or whole paragraphs where the subject of the text warranted.

Open coding began from the most central policy documents and proceeded to texts originating from quasi-independent or independent public bodies. I worked in chronological order to gain a sense of changes over time, enabling constant comparison (Gibbs 2008). I coded “close to the documents”, using open coding and labelling the sections of text with language and terminology derived from the documents themselves (Strauss and Corbin 1990; Bazeley and Jackson 2013). Automatic coding was occasionally used to identify relevant key phrases, such as “the international student experience ” and “the best and the brightest”, which were used frequently verbatim. Reports were then checked to remove irrelevant results. It was not used otherwise, in order to preserve the integrity of an inductive approach. Inductive coding meant that when new codes were added, previously coded documents needed to be reviewed for relevance to these codes. I used text searches to find these. This enabled the constant comparison that Gibbs (2008) argues is essential to good qualitative analysis; it maintained consistency and ensured that the same codes were used throughout the study.

“Second order” or “axial coding ” involved reviewing each code individually for consistency. It also established relationships and connections between codes expressed through a hierarchy, reviewing redundant nodes, and renaming codes for consistency (Gibbs 2008). During this review, codes were merged when very similar meanings or patterns of coding were apparent, and “hierarchies” of codes were developed (Saldana 2009; Bazeley and Jackson 2013). Compound queries were used to check the comprehensiveness of coding. Internal consistency of codes was reviewed by exporting the reports and checking them manually. With a single coder, Bazeley and Jackson (2013) suggest the transparency of the analytical process must replace second coders as a measure of reliability. To this end, a selection of code reports is included in Appendix 3. A selection of code reports was reviewed by an external researcher.

Codes were categorised as rationales for or against recruiting international students; and descriptions of international students. This corresponds to grounded theory’s “selective coding ” (Gibbs 2008), where codes are deleted, retained and regrouped. These were recorded in NVivo using either “issues” or “enhance” or “benefits to indicate rationales against and for recruitment, respectively. During the qualitative analysis, I compiled a list of the codes which related to each rationale.

The quality of a qualitative analysis relies on the transparency of the analytical process, reflexivity, evidence of its grounding in the data (Gibbs 2008), its internal coherence, as well as whether the book proves convincing (Kuckartz 2014). To ground the conclusions in the data, the prevalence of particular themes is captured in citations. To provide a rich description of the entire corpus (Braun and Clark 2006; Gibbs 2008), quotes are used frequently and contextualised where possible.

Once the codes were organised into rationales for or against recruiting students, Bacchi ’s “what is the problem represented to be” (WPR ) framework of 6 questions was applied to each rationale, as illustrated in Fig. 4.1.

For example, Chap. 5 explains the argument that international students should be recruited to enhance the UK’s diplomatic influence through soft power , which can be generated through the goodwill created when students study in a foreign country. This rationale incorporates an implicit understanding of a problem, a proposed solution and a number of assumptions. Because there are several rationales with overlapping and competing discourses, Bacchi ’s framework is applied to each separately. Bacchi demonstrates the power of her questions in an integrated narrative, rather than a rigid march through each question individually, particularly as smaller problematisations may nest within larger ones. This is consistent with Foucault ’s earlier work (1965, 1977), in which he identifies key concepts within the broad research study and within each explores the historical development , ruptures and silences . Literature is used to highlight assumptions and offer alternatives.

Conclusion

The approach used in this study can be characterised as broadly Foucauldian, adopting a problematisation analysis. It is document-based, but understands policy as discourse, such that texts are a momentary crystallisation of discourses used in social events and networks . A wide range of documents have been used, from a range of actors and genres, which helps to illustrate the fluidity of the domain of international student policy. These documents have been approached using open qualitative analysis, supported by NVivo software. After initial coding , a hierarchy of key themes was developed, which were further analysed using Bacchi ’s framework of questions.

Policy discourses were found to relate almost entirely to the recruitment of international students. The capacity of the UK HE sector to attract international students is explicitly valued: “Higher education has a fundamental value in itself and our universities are, in many ways, world-class: in research; in attracting international students; and in contributing to the economy ” (BIS 2011, p. 7). This may seem an obvious, indeed a “natural” point (Foucault 1972), but there are other possibilities for policy: it could relate to discrimination, racism and insecurity experienced by students (Marginson et al. 2010); it could speak to inclusive practices and multiculturalism; it could seek to engage students as temporary citizens. That UK national policy discourse does not speak to these issues is not natural or inevitable; it is the culmination of a policy process of selection, prioritisation and discursive formation which have made certain statements “unsayable”, naturalising particular assumptions (Fairclough 1989) and instituting silences .

This rest of the book critically analyses these assumptions and is organised around four key rationales relating to international student recruitment :

  • International students increase the UK’s global influence ;

  • International students increase the quality of UK higher education and its reputation ;

  • International students increase income to the UK;

  • International students are less desirable when they are seen as migrants.

Within these rationales, international students are represented in a range of different ways:

  • As ambassadors ;

  • As educational resources ;

  • In cultural deficit ;

  • As financial resources;

  • As migrants.

The following chapters present each rationale, giving the findings of the textual analysis, and then explaining the WPR analysis.