Global transformations of the state,governance and teachers’ labour: Putting Bernstein’s conceptual grammar to work

Introduction

This paper engages with Basil Bernstein’s (1990, 2001; Bernstein and Solomon, 1999) rich conceptual grammar in order to generate a sociological account of the outcomes for teachers’ work, identity and social class as a result of strategic shifts in governance to the global scale. In doing so it builds on earlier published work (Robertson, 2012, 2013; Sorensen and Robertson, 2017) on the global governance of teachers, where we drew upon aspects of Bernstein’s rich corpus of work on pedagogic governance to examine transformations in teachers’ work over time. However, the lines of argument in our previous papers were suggestive rather than systematic, and in need of fleshing out and filling in.

This Special Issue of the European Education Research Journal, on regenerating the sociology of Bernstein through a focus on international policies and local affects, has provided us with the opportunity to meet that need. Our main purpose here will be on developing a two-way conversation between Bernstein’s conceptual grammar and how best to theorise the nature of the social regulation of teachers at play in the Organisation for Economic Cooperation and Development’s (OECD) Teaching and Learning International Survey (TALIS). The OECD is, of course, an intergovernmental organisation which operates globally and which has also been engaged in international policy framing aimed at student learning via its Programme for International Student Assessment (PISA). Bernstein’s theoretical work, and that of his students, was largely derived from a detailed analysis of education in national education settings (see, for example, Sadovnik, 1995; Moore and Muller, 2002; Frandji and Vitale, 2011). The questions we seek to address in this paper are how useful Bernstein’s theoretical tools are for examining global policies, and what new insights this engagement with contemporary social life generates for Bernstein’s theories. As such we also hope to put Bernstein’s theories to work in ways that are true to his own project of generating theoretical tools having greater levels of explanatory power as they encounter new phenomena (Bernstein and Solomon, 1999; Bernstein, 2000).

We begin our paper with some general remarks on Bernstein and his conceptual grammar, before turning to describe the emergence of the OECD’s TALIS as an international policy.We draw substantively on a particular family of Bernstein’s conceptual tools aimed at understanding pedagogic governance, and argue that the OECD can be understood as a global actor that has come to dominate the field of symbolic control over what counts as ‘the good teacher’ and ‘quality education’. In this context, TALIS itself can be understood as a pedagogical device acting as a symbolic regulator of consciousness.

We reflect on the prominence of constructivism in TALIS as a preferred pedagogy and the eschewing of disciplinary knowledge as the basis of expertise, and ask: what does this mean for teacher knowledge, consciousness, identity, the division of labour, and the social base (Bernstein, 1996: 45)? Understanding the significance of this ideological shift in the discourse contained in TALIS as a pedagogic device is particularly relevant to examining whether, as Bernstein observed in one of his last texts, a new kind of pedagogic identity is emerging from a market-oriented ideology shaping many societies around the globe. We conclude by reflecting on our account in two ways. First, what are the challenges facing the OECD, as an intergovernmental institution with a global horizon of action, in realising its imagined teacher for the 21st century when it attempts to govern over a number of scales reaching down into the classroom? Second, what does our analysis mean for Bernstein’s theories, which have been characterised by a focus on the state as regulator, by a methodological nationalist lens on the state’s education sectors (see Robertson and Dale, 2009), and assumptions of the penetrability of boundaries (Archer, 1995)? All of these elements are at issue in offering a robust account of international policies and local effects in education and how the pedagogical work some policies are engaged in actually happens.

Encountering Bernstein

Bernstein had a long and enduring interest in the production of knowledge, the structuring of intellectual fields (Moore, 2004: 120), and education ‘…as a crucial instrument of symbolic control through which horizontal solidarities of nationalism and identity were constructed, and vertical cleavages of class were reproduced’ (Bernstein, 2001: 22). However, it is his determination to develop a conceptual grammar that links transmission, agencies and the social base to the production of consciousness, dispositions and desire which helps us to establish the link between the ‘what’, ‘who’ and ‘how’ of knowledge acquisition and processes of cultural reproduction.

To make more visible the epistemic gains to be had using Bernstein’s conceptual grammar in analysing discursive and material shifts in governance, it is helpful to look at alternative accounts of the global governance of teachers’ work using theorists interested in the relationship between discourses, knowledge and power. Foucault’s (1988) work is perhaps an obvious starting point in that he focuses attention on the discursive positioning of the subject – in our case the teacher – by globally-located agencies such as the OECD. Discourses, of course, are socially-produced and are a particular kind of social practice, and would go some way toward helping us build an account that is sociological. Much of the work on the OECD’s governance tools, mostly centred on PISA, has tended towards this kind of analysis (e.g. Grek, 2009; Ozga, 2009). However these accounts tend to focus attention on numbers themselves as a means of governing. Whilst insightful, we feel this kind of theoretical account has a limited vocabulary to talk about power, the range of actors on the terrain who might mediate processes of transmission, and how new kinds of identities might be produced as a result.

A second approach might be to map out the different actors who coalesce around the OECD and the creation of TALIS, much as Ball did when he mapped out the different actors and their connections in what he called ‘network governance’ (Ball, 2007; 2012). Here, it is now possible to describe different agencies and actors on the education landscape – from financiers to accounting firms, global education providers and private companies all attached to, or replacing, public institutions. However, what we do not see in this approach are which networks matter (Goodwin, 2009), who gets to make what kinds of decisions over what aspects of teachers’ work and identity, quite how power and control are exercised, how this set of rules of the game is established, what kinds of specialised agents these actors are in relation to each other, and how the network itself – as some kind of transmission device – works pedagogically.

The major issue regarding these first two approaches, both broadly Foucauldian, is, as Bernstein puts it, that

…there is no substantive analysis of the complex of agencies, agents and social relations through which power, knowledge and discourse are brought into play as regulative devices. What is absent from Foucault is a theory of transmission, its agencies, and its social base. Nor is there any analysis, or study, of different modalities of control. (Bernstein, 2001: 23)

Bernstein was insistent on the need for theory to evolve as it sought to generate greater powers of description and explanation and as descriptions of the empirical world enrich theories. This explicit move – of putting Bernstein’s theories to work on making sense of education governance, and in a globalising world – is important in that in many respects it is also a very different world to the one that Bernstein sought to describe and theorise. These transformations have included (but are not limited to) the rescaling of key aspects of the governance of education to the global level (Robertson et al., 2012), horizontal shifts in governance to include many new (including for-profit) actors (Ball, 2007, 2012), and pressure on teachers to use their knowledge and skills to ensure higher student learning outcomes and competitive knowledge economies (Robertson, 2005). Understanding and explaining these developments using Bernstein’s sociological tools demands movement backward and forward between description, our empirical observations of the contemporary world, and re-description/theory development, in order to understand the mechanisms of cultural production and transformation.

We will therefore be arguing that, ontologically, epistemologically, methodologically and politically, Bernstein’s tools have the necessary rigour and purpose, yet flexibility, with which to study the social worlds of teachers and international agencies, and abstract the structuring rules. Moreover, this conversation potentially allows us to bring something back to Bernstein’s conceptual grammar. Ontologically he is asking about how our social worlds, dispositions, identities and desires are constituted, as well as how they are disturbed, challenged and transformed, and through what mechanisms with what outcomes. Epistemologically he guides us to theory-informed description that then moves to levels of abstraction, to determine the links between the real (or social base) and the structuring mechanisms at work, and the relations within and between. Methodologically, he asks us to be attentive to emergences, or ‘announcements’; in other words, to refuse to regard categories as fixed and frozen, and the social world as unchanging. How does the voice of the category called ‘good teacher’ alter as a result of the role of the OECD in shaping teacher policy? In addition, if theory should do political work in some way – or should ‘disturb’ (Bernstein and Solomon, 1999: 275) – it can also help make visible those processes and social relations, such as power and control, privilege and marginality, to help us challenge and change the transmission processes that leads to these unequal outcomes.

Bernstein’s conceptual grammar on pedagogic governing

Here we introduce a family of concepts from Bernstein in order to explore shifts in the global pedagogic governing of teachers’ work: these include the ‘field of symbolic control’, the pedagogic device, classification and framing, and pedagogic identities.

Bernstein (1990: 134–135) developed the concept of the field of symbolic control to refer to those agents and agencies that specialise in discursive codes which they dominate and thus act as ‘regulators of consciousness’ (Bernstein, 2001: 22). Such control can be ‘…normalising or disturbing, reproductive or productive; consensual or conflicting’ (Bernstein, 2001: 23). In his early works, Bernstein distinguished between agents who are functionally similar but operate in different fields of activity; for example, psychologists who operate in the economic/private field compared with those who operate in the symbolic/public/government field. Similar occupational functions but in different fields, he suggests, may well produce ‘…different interests, different dispositions, different motivations and different ideologies’ (Bernstein, 2001: 25). Given the extent to which important functions in the education sector in countries like England and the United States – from school inspection to textbook production and school management – have been reallocated to the economic/private field, or some combination of public/private, what does this blurring or collapsing of the boundary between the public and the private mean for symbolic agents? Following Bernstein, we might expect that combinations of symbolic agents in the economic field, with their dispositions, motivations, interests and ideologies different to those of the state/symbolic field, are in turn constitutive of different social and class factions. What, however, are the implications of rescaling important aspects of national state powers to global agencies, such as the OECD, as we see with education policies, for how we understand the field of symbolic control?

Bernstein also goes on to describe different categories of agents of symbolic control (Bernstein, 2001: 25–26) engaged in different kinds of work: from (i) regulators (e.g. religious/legal), to (ii) repairers (e.g. medicine, psychiatry, social work), (iii) reproducers (e.g. school system/teachers), (iv) diffusers–recontextualisers (e.g. mass and specialised media, including arts), (v) shapers (e.g. universities and cognate agencies, research centres), and (vi) executors (e.g. administrators in the civil service). He noted that from (ii) downward, ‘…all are regulated directly or indirectly by the State and closely subject to its policies’ (Bernstein, 2001: 26). Again, there is considerable evidence that the hold of the national state over these various specialist activities has changed dramatically as a result also of the vertical and horizontal rescaling of the functional division of the labour of education (Robertson and Dale, 2006; Ball, 2007; DiMartino and Scott, 2012).

If important aspects of the work of the reproducers, diffusers–recontextualisers, shapers and executors, are not regulated by the national state but, rather, through new combinations of agents operating at the global level as a new nodal point of power, what does this mean for new modes of social control/governance, class interests and projects? Concretely, what does the close alignment between the OECD and transnational capital (see Woodward, 2009) mean for the nature of the symbolic control over codes and the production of interests, dispositions and identities? To what extent do these different categories of symbolic agents continue as a result of transformations in governance, or are new factions visible?

Bernstein (2000) also introduced the idea of the ‘pedagogic device’ which, we will argue, is particularly powerful for understanding TALIS. By pedagogic device Bernstein means the relay itself, as well as what is relayed. The relay itself has rules that regulate what can be relayed, and these rules are ideological (Bernstein, 2000: 28) in that they carry within them ideas about how the world could, and should, work. As a relay itself, TALIS uses hierarchy (from best country to worst on a particular question, such as teachers feeling prepared to teach) with limited capacity to carry complex information on teachers and their local contexts. It also represents education systems as national because it is the national (and not sub-national) level that usually engages with and funds the OECD as an intergovernmental organisation. However, not all education systems are organised and governed nationally, with the result that this adds further spatial scales – or levels of governing – to be penetrated by the relay in order to have effects. In terms of what is relayed, the specific areas of focus of the TALIS survey carry ideas about what and how teachers should teach, and how they might be helped to overcome their limitations through a repair function offered by the OECD.

Bernstein describes three rules and associated fields of pedagogic governance – (1) distributive, (2) recontextualising, and (3) evaluative – which, in combination, make up the pedagogic device (Bernstein, 2000: 37). In essence the pedagogic device ‘…acts as a symbolic regulator of consciousness, on the one hand, and the production, reproduction and transformation of culture, on the other’ (Bernstein, 2000: 38). These rules are organised hierarchically, in that the first frames the second, which in turn frames the third, linking power at the moment of production to transmission and then to the moment of practice.

By using distributive rules and its associated field, Bernstein is describing the rules and processes involved in the formation of pedagogic agencies, agents and discourses responsible for determining and encoding what constitutes valid knowledge. In our case, the questions then are, what is it that 21st century teachers should know and do; and, if this is different from how teachers have tended to act in the past, what space is opened – and filled – by a new set of ideas or ideologies concerning what counts as the good teacher? Bernstein insisted the recontextualising rules in the pedagogic device are connected upward to the distributive rules. Thus he is describing the pedagogic rules of recontextualisation which involve governing particular agents using particular functions (by whom/for whom, and so on), and this in turn gives rise to recontextualising fields. However, if – as we have argued above – the state and its governing functions are being transformed vertically and horizontally, do the recontextualising rules create new recontextualising fields, enabling new sets of actors to be active in recontextualising processes?

Finally, by ‘evaluative’ Bernstein means processes of acquisition where ‘evaluation condenses the meaning of the whole device’ (Bernstein, 2000: 36), this condensing resulting in a moment of acquisition (Bernstein, 2001: 37) or practice. Bernstein’s pedagogic device is thus useful for linking together those creation, transmission and acquisition processes necessary to give rise to pedagogic identities – in this case, the possibilities (or not) for the OECD of using TALIS as the device for disrupting an old teacher identity and inserting a new one.

We also draw on our earlier engagement with Bernstein’s concepts of ‘classification’ and ‘framing’ (see Robertson, 2012) on understanding the dynamics of power and control over the global governance of teachers’ work over time. Classification refers to principles that establish the social division of labour – such as teacher, teaching assistant, social worker, head teacher, school consultant, and so on (Bernstein, 2000: 6). Dominant power relations establish and maintain the boundaries that give rise to these divisions. Strong classifications have strong insulation between categories; weak classification means the insulation is broken and one or both categories are in danger of losing their identities or having it redefined. Framing is concerned with who controls what – or the forms of realisation of the discourse; that is, the voice of the category and the projected message.

Where framing is strong, the transmitter has explicit control over selection, sequence, pacing, criteria and the social base. Where framing is weak, the acquirer has more control over the communication and its social base. (Bernstein, 2000: 13)

For example, weak framing of the ‘good teacher’ by the international agencies up until the emergence of TALIS gives the acquirer (in this case the teacher) more control over the discourse, its rules of realisation, and therefore practices and forms of consciousness. This raises the question of the extent to which TALIS, as a global pedagogical device, suggests a stronger set of rules for framing, in turn placing limits on teacher’s autonomy in local settings over pedagogy, curriculum and forms of assessment.

In concluding this section we dwell on an observation made by Bernstein in one of his final texts regarding the rise of neoliberalism and market-based logics in education. Bernstein was aware of major changes taking place in what might be accepted as official knowledge in national education systems, though in his view these developments were still fragmentary and waiting announcement. In a prescient chapter in Pedagogy, Symbolic Control and Identity (Bernstein, 2000) he sketched out figures described as ‘barely recognised, but which might be; “…in the future”’ (Bernstein, 2000: 65). These figures represent distinct groups in the official arena as resources for different modalities of reform; these positions differ in their bias and thus the pedagogic identities they project. His ‘De-Centred’ (market) identity is contrasted with three other possibilities regarding discourse and identity: ‘De-Centred Therapeutic’ (professional); ‘Retrospective’ (old conservative); and ‘Prospective’ (neo-conservative). Reminding us that we were not, at that time, likely to recognise this neoliberal De-Centred Market (DCM) identity (Bernstein, 2000: 66–69), he nevertheless pointed to the various ‘resources’ which might contribute to its emergence, including pedagogical practices contingent on the market such as efficiency, effectiveness and accountability. He wrote:

We have a culture and context to facilitate the survival of the fittest as judged by market demands. The focus is on the short term rather than the long term, on the extrinsic rather than the intrinsic, on the exploration of vocational applications rather than applications of vocational knowledge. The transmission here views knowledge as money. And like money, it should flow easily to where demand calls. There must be no impediment to this flow. Personal commitments, inner dedications, not only are not encouraged, but also are regarded as equivalent to monopolies in the market, and all such monopolies should be dissolved. The DCM position constructs an outwardly responsive identity rather than one driven by inner dedication. Contract replaces covenant. (Bernstein, 2000: 69; emphasis added)

What is quite remarkable about this sketch of a ‘barely visible’ identity is the extent to which this would appear to have become a dominant identity in many education landscapes. Important questions emerge for us, however. Are the OECD and TALIS respectively a mobilising agency and specialised device generating this new identity with its distinct dispositions, forms of consciousness and desires oriented toward producing a knowledge economy rather than a knowledge society? – and, if so, what are the implications for teachers if covenant is replaced by contract?

Putting Bernstein into conversation with the OECD’s TALIS programme

In this section we put Bernstein’s theoretical tools into conversation with the OECD and TALIS. We will suggest that the OECD is not only a nodal agent of symbolic control that has acquired for itself state-like functions at the global level, but also one seeking to use programmes like TALIS strategically as a pedagogical device to transform teachers’ contemporary identities toward being a part of, and producing, competitive knowledge economies. We draw on empirical work, some of which has been reported more generally (see Robertson, 2012; Sorensen and Robertson, 2017; Sorensen, 2017), to develop our analysis below.

Teachers and the field of symbolic control

The OECD has, since around the year 2000, strategised and been increasingly successful in acquiring for itself a pivotal role in the field of symbolic control over national education systems through its expert advice and comparative country data. It has done this in three ways. First, it has limited the influence of previously dominant global agencies (UNESCO and ILO) by offering an alternative set of discursive codes concerning quality education and the good teacher. Second, it has had ceded to itself important power and authority (especially legitimation) from national governments, enabling it to take on functions once the exclusive powers of the national Official State. (Note that in many cases education systems are not national but, rather, sub-national so that the impetus of ceding power to the OECD could well have come from the economic bureaus of national states.) Third, it has bought in a new set of more globally-oriented actors to flank its governing activity in the field of symbolic control – from the European Commission to transnational education companies, global academics and global consultancies. In combination these developments have also begun to expand the circuits of pedagogic power in an increasingly globalised education world (Singh, 2017).

In relation to an alternative set of codes regarding what might count as the good teacher, the OECD drew on its indicator work on knowledge economies, especially what and how to measure them (Robertson, 2009), to provide a frame for the good student (measured through PISA) and the good teacher (TALIS); and, in turn, linking these pedagogical projects to the realization of competitive knowledge and service economies (Robertson, 2005). PISA also enabled the OECD (2014c) to demonstrate its power through the visible effects on sub-national and national governments’ education, and thus to be enrolled as a model of pedagogic governance of other populations, such as teachers.

Within the OECD, the TALIS Board of Participating Countries (from January 2016 called the TALIS Governing Board, as a result of the status of the programme in the OECD being raised) stands out as the most important body for multilateral decision-making on TALIS. The TALIS Governing Board meets approximately a dozen times in each round of TALIS. Major issues for the first two rounds of TALIS were to be approved by the OECD Education Policy Committee and, in some cases, the OECD Council. However, it is in the TALIS Governing Board where decisions concerning policy objectives for the survey are established, and where the standards for data collection and reporting are determined (Sorensen and Robertson, 2017). The TALIS Governing Board is thus at the very core of the field of symbolic control because of its important executive position, akin to Bernstein’s ‘executors’ identified above, but this time with a global orientation. The TALIS Governing Board consists of government representatives from each participating political entity, and from a number of organizations. In accordance with the distinctive historical relationship and formal status given to the European Commission (EC) by the OECD (as another global actor), the EC is represented by policy officers from the EC Directorate General for Education and Culture (DG–EAC). The Business and Industry Advisory Committee (BIAC) and the Trade Union Advisory Committee to the OECD (TUAC, the OECD formal mechanism for social dialogue with trade unions) have also participated in the negotiations (Sorensen and Robertson, 2017; Sorensen, 2017).

It can be argued that the EC plays an important flanking role for the OECD in the field of symbolic control. This is because the EC also seeks to guide national governments across Europe toward the development of competitive knowledge-based economies; but, because of subsidiarity rules, the EC has limited or no authority to govern the education systems of its member states. The separate needs of these global or regional actors for legitimacy and a political mandate to govern national states are revealed in our interviews with personnel from the OECD and DG–EAC. Interviewees emphasised that their organisations were, first of all, ‘bottom-up’ intergovernmental fora for cooperation. By funding the OECD’s first round of TALIS in 2008, DG–EAC was able to get the OECD to collect data on European teachers, giving it, DG–EAC, access to quite detailed accounts of teacher’ work and how it might move the teachers’ Member States in the direction outlined in the TALIS pedagogical device.

Education International (EI), the global federation of teachers unions, has – in its capacity as the primary organisation working for teachers’ interests in TALIS – also been provided with a broad mandate by its affiliates to negotiate in the TALIS Governing Board. ² Our fieldwork revealed that EI had taken part in TALIS Governing Board meetings since they were initiated in 2006, and was granted permanent observer status in 2009. Attaining this status, EI has been consulted on draft chapters and enjoyed enhanced opportunities for submitting comments and ideas. However, EI does not have the opportunity to decide on the encodings of what a good teacher is, only to react and comment on them. The TUAC representatives participating at TALIS Governing Board meetings report to a sub-group of EI affiliate member organisations set up by EI through the TUAC Education, Training and Employment Policy Working Group. Accordingly, EI affiliates are encouraged to mobilise support for TALIS amongst their members (Education International, 2012), and thus can be seen to diffuse and recontextualise TALIS in their different national contexts. This is a less powerful role because they are some distance from the work of producing the rules and encoding them in the TALIS survey.

The OECD has also used its authoritative power to mobilise a range of agents and agencies, some from the recontextualising fields in national education systems, to come together for a series of high visibility events. The most notable of these is the series of annual ‘International Summits on the Teaching Profession’, the first of which was convened in New York in 2011 by the United States Department of Education, the OECD and EI. Thereafter a Summit has been held each year, with the seventh taking place in March 2017. ³ The aim is to promote learning amongst ministers, union leaders, government officials, teaching professionals and education experts on teacher policy (Gomendio, 2017). The link can now be seen here between the distributional and the recontextualising rules and fields at work.

These new symbolic agents are also visible in the events that have surrounded the launch of TALIS results. A global live webinar ‘Education Fast Forward 10: Better Teaching for Better Learning’, hosted by edu-tech firm Promethean Planet, connected studios in Atlanta, Boston, London, Paris, Tokyo and elsewhere to provide for a debate on TALIS results and teacher quality (Education Fast Forward, 2014). The debate took place between what was labeled ‘global education experts’, dominated by consultancies and foundations based in North America and England – including the Bill and Melinda Gates Foundation, TSL Education (now TES Global), Innovation Unit, and Promethean. The OECD and UNESCO were represented, together with renowned academics, including Michael Fullan and Pasi Sahlberg, who are simultaneously entrepreneurs and legitimators of the TALIS process and product (Education Fast Forward, 2014). In this sense, private consultants, consultancy firms, corporate philanthropists and entrepreneurial academics act as diffusing and shaping symbolic agents. In national settings there are also new kinds of agents involved in the recontextualising field. For example, in England, the Education Policy Institute, an independent evidence-based research body with specialisms in accountability and inspection, school performance and leadership, interprets and discusses TALIS data and results and in doing so has the potential to influence the evaluative rules and thus practices (see Sellen, 2016).

These globally- and nationally-oriented private sector actors are located in Bernstein’s economic field as cultural agents. As a result, they have different orientations, values, dispositions and identities to those in the public symbolic field. In terms of policy preferences, strategies, capacities and horizons of action (Ball, 2012; Robertson and Verger, 2012) they share the feature that their activities, whether for-profit business or venture philanthropy, produce policy recommendations, on the basis of TALIS data, which feed into new products and services. In other words, they help to profile and hence ‘sell’ TALIS whilst also selling their own expertise and services at the annual Education Summits (OECD, 2014d, 2015).

Conclusions

There is a great deal to be gained from engaging with Bernstein on international policies and local contexts. In our view there are genuine insights to be gleaned from the ‘who’, ‘what’ ‘how’, and ‘so what’ of contemporary forms of governing in a globalising education world when putting Bernstein’s conceptual to work ‘in the world’. We hope that in our endeavours we have stayed true to the spirit of intellectual enquiry that Bernstein championed. Of one thing we are sure: were Bernstein to enter the research field now, his accounts of the world of education would be very different from that of the world which gave rise to descriptions that breathed life into his conceptual grammar. In Jessop’s (1999) terms, the Keynesian National Welfare State has given ground to a Schumpeterian Post-National Workfare Regime, with huge implications for the world of education and the pedagogical work that is then set in motion to produce new social orders, identities, relations and base. For this reason we have tried to address the contemporary world of education and its social role in making pedagogical identities with many questions, to which we have only the traces of answers.

Our focus specifically on teachers as workers in the 21st century is particularly revealing, because they are the symbolic agents whose work it is to make the new 21st century citizen and worker. In focusing on the role of the OECD and devices such as TALIS we are able to show how the very work of governing is now different. So, what tentative conclusions do we draw from the TALIS case?

First, we find that the OECD has emerged as a dominant actor in the field of symbolic control, encoding a very specific understanding of the ‘good teacher’ and developing TALIS as a means for controlling both the distributional rules as well as some of the key repair functions, through the ways it suggests what is to be done.

Second, the OECD has expanded the global governance of teachers through partner agreements, event cooperation and contracting out to scores of flanking agents, from the European Commission to global commercial firms, and entrepreneurial scholars, whose alignments are much closer to the economic sphere and particularly that of global capital. However, this alignment also brings risks, especially with regard to the rules and fields for recontextualisation. It is still thinkable that education can be organised differently, that there are actors out there who believe that the dominance of global symbolic agents located in the borderlands of the economic and public/state fields is illegitimate and thus without authority, and that teachers should have autonomy over their work. This dissident voice, at least viewed from the point of the OECD, is also a voice that will need to be governed, now and into the future. Perhaps the time will come when the unthinkable becomes thinkable, and when the OECD will have won the day as a regulator of consciousness; but there is a great deal of distance to go with TALIS before this is the case.

Third, viewing TALIS as a pedagogical device is valuable in that it enables us to see the what and the how of the relay, and that we are able to link form, framing and ideational content, and the work that the different recontextualising and evaluative fields need to do across a greater space – vertically across scales and horizontally across countries. However, this multiplicity of spaces and scales also creates new challenges in that each has its own recontextualising rules and fields, as well as the points of entry of TALIS into national settings which do not articulate with the organization of the sector, and generates new frictions and places new limits on the fidelity of the idea of the pedagogic identity for teachers that is being relayed. This then compels us to enquire about the questions with which Bernstein did not engage to a sufficient extent, which in Archer’s terms (1995: 212–223) include the wider structure of education systems (in our case a very multi-scalar structure with a distinct power geometry), questions about the ease of permeability, of competing resources for identities which now operate across scales, and of differences across systems. All of these issues are at play in the OECD’s TALIS programme, demanding some new insights into how power is projected over space, the multi-scalar nature of symbolic agents and fields and how these are navigated and governed, what makes one system more permeable than another, and what the incremental relocation of the symbolic work to the economic field means more generally for governing.

In raising these questions we do not intend to underplay the significance of what Bernstein’s conceptual tools have helped reveal. It is possible to trace out and see the developing shape of an emerging governing project aimed at teachers for 21st century capitalist economies and societies. However, there are political questions now thrown to the fore regarding the prominence of a market identity and what this might mean for knowledge itself. Will the decentred market identity replace inner dedication and ‘covenant’ with ‘contract’; or what might a more hybrid identity begin to look like if market-mediated therapeutic identity emerges out of the arena of resources? In which ways do such hybrid identities emerge as distinctive and tension-laden outcomes – potentially constituting divided-self pedagogic positions (Bernstein, 2000: 77–78) – of multi-scalar governance, and how do they play out in specific fields of recontextualisation? It is too early to tell.

Indeed, these are questions that must be explored empirically, and not claimed as a truth in a game of academic guesswork. Will the OECD continue to play a central role, as a global nodal point, in the face of tightening national boundaries, populist political projects and a growing sense that neoliberalism has delivered to the ruling class but not the middle and working classes – of which teachers are an important class faction? The only thing we can be certain of for the moment is that the social world is not fixed and static. It is full of struggles over meanings and what becomes thinkable: ‘…in the process of controlling the unthinkable it makes the possibility of the unthinkable available’ (Bernstein, 2001: 38). By attempting to think of teachers in a particular way, perhaps what has also really happened is that teachers’ taken-for-granted worlds – shaped by specific and very different social and material contexts – are now suddenly revealed to them as thinkable.