Epistemic governance: an approach to the politics of policy-making Open Access

Although it is feasible to think that actors’ ability to influence others’ behavior is based on structural sources of power such as money or military force (see, e.g. Mann, 2012b), from the actors’ perspective it all boils down to their conceptions about the facts and about themselves as actors: what they are able or forced to do, and what their desires and obligations are. Therefore it is apparent that – whether they are conscious of it or not – actors who aim to be influential in politics attempt to affect others’ views of reality. Reminding us that government and mentalities are entwined and constitutive of each other, Foucault's neologism of governmentality captures well the idea of government that guides the comportment of others by acting upon their hopes, desires, or milieu (Foucault, 1991; Inda, 2005). According to Foucault this kind of governance became increasingly crucial for the political elite when the sovereign power of monarchy was gradually replaced by a constellation in which the art of government consists in managing public opinion and the support of several factions of society (Foucault, 2007, 2008). The same has been said about governance on a global scale: national states adopt global standards and policy models not because they are coerced to do so but primarily because governments are convinced that it is good for them, and hence global governance works particularly through knowledge production and consultancy (Alasuutari, 2011a; Buduru & Pal, 2010; Radcliffe, 2010).

While there is growing interest in considering politics and governance as a playing field in which actors act upon each other's conceptions, aspirations, and motivations, there isn't a satisfactory analytic of what we call epistemic governance as a set of processes describing how this happens. Several research traditions touch on governance and policy-making from this perspective, but the crux of the matter needs to be better unpacked, which is the task of this article.

When thinking about relevant previous scholarship, we find framing theory captures part of the problematic by focusing on how ‘opinions can be arbitrarily manipulated by how issues are framed’ (Chong & Druckman, 2007, p. 104) so that ‘different meanings compete for support’ (Fiss & Hirsch, 2005, p. 30). Building on the seminal work by Goffman (1974), this tradition emphasizes the use of ‘schemata of interpretation’ by individuals to understand and act in the world around them. Framing theory has become particularly popular in research on social movements (Benford & Snow, 2000; Kubal, 1998; Snow & Benford, 1992) to explain how collective action is organized in one way rather than another through the contested ‘core framing tasks’ of diagnosis, prognosis, and mobilization. This research tradition also stresses the effect of the mass media on public opinion (Entman, 1993; Scheufele, 2006). Framing is itself a widely employed metaphor in governance, but these two common theoretical applications capture its salient features for our purposes here. However, in all these uses the otherwise helpful metaphor of framing is constricting in that it treats frames as independent from the issues themselves and from the actors involved. Such ‘light’ frames assume that an actor is more or less independent of the frames she uses, choosing them at will much as one does a pair of spectacles. Hence framing misses the point that meaning construction is a much deeper question. To continue using that concept, the contesting frames used in a particular issue are embedded in larger institutional contexts and practices, entwined with and reflected in actors’ tacit presuppositions of the world, which also constitute actors themselves (cf. Glaeser, 2011). Consequently governance that acts upon people's understandings not only consists in daily political framing contests of meaning; epistemic governance speaks to and evokes actors’ deep-seated values and beliefs, and we argue that success in epistemic governance is based on those paradigmatic assumptions.¹

The importance of the institutional context in understanding power and governance that always functions through actors’ epistemic premises has also been addressed in some other approaches to global governance. For instance the epistemic communities approach has stressed the role of scientists and policy experts whom policy-makers in different states turn to for advice: when a network of professionals with recognized expertise and an authoritative claim to policy-relevant knowledge within a particular domain share a set of normative and principled beliefs, they influence the coordination of state policies (Dunlop, 2009; Haas, 1992a, 1992b; Miller & Fox, 2001). Likewise, attending to the global shaping of national policies, policy diffusion research (Dolowitz & Marsh, 1996, 2000; Simmons & Elkins, 2004) shows that the same global models are enacted in different national states, drawing on and enhancing a global epistemic community. The neoinstitutionalist world society theory tradition makes a similar point by maintaining that in fact national states function as a kind of epistemic community, sharing norms and knowledge that comprise a rationalized world culture (Meyer, Boli, Thomas, & Ramirez, 1997). Consequently, states voluntarily enact the same global models. In a similar manner, building on Gramsci's famous idea of hegemony as ‘intellectual and moral leadership’ (Gramsci, 1971, p. 57) critical policy studies examine domination by way of inclusion and exclusion in political agendas (Howarth, 2010).

On the other hand, following in the footsteps of Foucault, the governmentality research tradition (Dean, 1999; Dean & Hindess, 1998; Rose, 1996, 1999a, 1999b, 2007), the closest to addressing our question, has studied the effects of public policies on people's mentalities, which shape individuals as willing and docile subjects. Scholarship in this tradition has analyzed the rationales and effects of public policies on individuals’ mentalities (see e.g. Dean, 1998; Rose, 2000) to show that the self-regulating capacities of subjects, created by different indirect ‘technologies of government,’ are important for governing in liberal democratic societies, even extending to controlling populations beyond the state's purview (Rose & Miller, 1992).

Previous scholarship gives us a good general picture of the complexities of governance in current global conditions. Yet it has its shortcomings when thinking about the politics of policy-making in more detail. While scholars generally agree that a simple conception of power as a property held by one actor does not work in the global system, each school of thought seems to have its own subtext that claims who, or which more or less abstract actor, pulls the strings behind the scenes. For example the epistemic communities approach stresses the influence of scientists and experts, and in much of governmentality scholarship power is attributed to neoliberalism and its veiled technologies of government, whereas framing theory ascribes key agency to the media. Similarly, policy diffusion research emphasizes knowledge innovators while critical policy studies stress the nodes of capital. In world society theory the power-holder is more abstract, since power is ascribed to ‘world culture,’ with ‘world models’ as its embodiments and international nongovernmental organizations (INGOs) as its agents (Boli & Thomas, 1999) but, symptomatically, world society theory has been criticized for ignoring the question of power (Beckfield, 2008). These theories all make valid points about actors or factors that impact on social change on a global scale, but by hurrying to point out power-holders (or by ignoring agency) much existing scholarship brushes aside the question about the particular ways in which actors influence the comportment of others by working on their beliefs and aspirations. Even most governmentality studies focus on how particular policies change society and human subjects within it, but tend to overlook how decision-makers are convinced in the first place that particular policies must be implemented, especially in a global context. The question of whether there is any analytical unity in the processes of such persuasion work has not been addressed.

To get a better grasp of the actual means by which actors in policy-making convince each other about the best policies and practices, more in-depth research and consequent theory-building are thus needed. The article at hand contributes to filling this gap by focusing on this aspect of power and politics, which we call epistemic governance. We suggest that there is, indeed, analytical unity in the strategies of epistemic governance and that is what we will unpack in this paper. Based on evidence from our research and theorizing, we find that this conceptualization of strategies of epistemic work functions across a range of different cases. In that sense it is a useful analytical and methodological tool for those who want to study politics and governance from the viewpoint of knowledge claims and rhetoric.

The remainder of the article is organized in the following fashion. In the next section we open up the concept of epistemic governance. After that we discuss what we mean by the three aspects or objects of epistemic work. In the final section, by way of conclusion, we discuss some implications of epistemic governance and outline ideas for future research.

By proposing an epistemic perspective on governance, we do not aim to present yet another competing theory on a par with previous theories of social power. Rather, our point is to say that policy-making is always premised on actors’ understanding of the world and the situation at hand. Actors themselves are, likewise, constituted by such understandings. From that viewpoint, similarities in decisions reached by governments in different countries can often be viewed as indicating likenesses in their epistemic assumptions, in other words about what can or must be done and what is virtuous or ethically acceptable (although there may be other reasons for such isomorphism as well).

The way we use the term, epistemic governance does not refer to a particular form or resource of power, for instance ‘soft power’ (Nye, 2004) in contrast with coercion, and neither do we argue that the contemporary world is witnessing the ‘end of power’ as we used to know it (Naím, 2013) although trying to influence common views has probably become an important aspect of governance in the present-day world (Hajer, 2009). Rather, the epistemic governance analytic depicts an approach to studying governance, however rough, violent, and easily discernible or ‘dislocational’ (Foucault, 1977) and subtle it is. This means that we approach governance as more or less unself-conscious ways by which actors work on people's conceptions of reality. This entails strategies that affect people's wishes and aspirations, but a threat or use of military force and economic constraints are also means to affect people's conceptions of the situation and hence make them adopt a particular line of action. As Castells (2011, p. 779) puts it, ‘Violence and the threat of violence always combine with the construction of meaning in the production and reproduction of power relationships in all domains of social life.’ In this sense, apart from the use of sheer force with no other objective or effect than removing an obstacle – itself an extreme case – most social relations can be viewed from the perspective of epistemic governance. The epistemic aspect is better visible when governance relies less on overt military force (when objectives are clearly and coercively spelled out) and more on opinions and sentiments of the general public. Recent scholarship has correctly emphasized the use of subtle techniques of governance in democratic nation-states (see, e.g., Castells, 2009, 2011; Foucault, 2007; Hajer, 2009; Mann, 1986, 1993, 2012a; Rose & Miller, 1992), and even authoritarian regimes are increasingly subject to global public opinion in the compression of global political processes. However, this scholarship has not paid much attention to what this convincing consists in and how it mobilizes popular notions.

Our approach to epistemic governance thus makes a contribution to theories of power and governance by identifying a processual layer of the political process that most theories ignore. In other words, ‘epistemic governance’ does not indicate an alternative to such theories but it does push theorizing to a thicker description of how actors are socially embedded and how they employ (even if implicitly) that position in influencing others. Here, we follow Foucault's famous distinction between analytics and theory in being more interested in how power operates in a specific, historical domain rather than in an abstract definition of what it is (e.g. Deacon, 2002; Foucault, 1980, p. 82).

So how does epistemic governance work? To illustrate it, consider a politician arguing for a reform, say a new law. In any given country, the politician would provide sources of authority aimed at convincing the citizens that the current state of affairs is unsatisfactory and that the proposed measures to improve the situation will be effective. Depending on the case and the country in question, the politician would appeal to reliable sources of evidence, experts’ or other authorities’ views, to what she assumes are widely acknowledged values and principles, and based on all that she would argue that passing the law will be in the best interest of the nation. The opponents would naturally challenge the politician's case by shaking different elements of the line of argument (see e.g. Alasuutari, 2013; Kurz, Augoustinos, & Crabb, 2010; Perelman, 1968; van der Valk, 2003).

This kind of discourse about political issues, the pros and cons of the policies being debated, is part and parcel of epistemic governance in the sense that to pass a new law policy-makers need to win sufficient support for it within their constituency and among stakeholders. Such support is sought for by what we call epistemic work as a reference to the particular techniques used by actors engaged in affecting views and hegemonic definitions of the situation at hand.

By talking about work and techniques we do not necessarily denote self-consciousness. An actor in the context here, any agent engaged in attempts to effect or comment on policy-making, does not always or even often try to work on epistemic premises. Rather, we suggest that in most cases policy discourse is aimed at concrete issues and targets, and the actors promoting a particular policy believe that the epistemic premises on which it is grounded are well-accepted and hence solid grounds for the proposal. Indeed, epistemic governance is most effective when its mechanisms and premises are left unnoticed.

What are these premises? When considering the above example, most existing research would pinpoint the way in which arguments are grounded on facts. If, however, we pay more careful attention to such policy discourses, we can notice that epistemic work does not consist only in ontological claims. Speakers also appeal to commonly shared values, and they address their audience as a community with shared interests, such as the nation. Indeed, based on existing empirical research of national policy-making, we find that there is analytical unity in the techniques by which policy-makers generally get convinced of, and in turn try to convince others of, policy solutions. Epistemic work can be targeted on three different aspects of the social world: what is the environment, who are the actors, and what is virtuous or acceptable.

In actual practice these three objects of epistemic work – ontology of the environment, actors and identifications, and norms and ideals – appear in combination so that a single argument by proponents, opponents, or analysts of a particular policy or political event deals with more than one aspect simultaneously. In fact, we propose a strong claim here: there is no epistemic work that does not entail all three objects. This is obviously the case with ontology: it is inconceivable to think of epistemic work on actors or norms that is not embedded in some claims about reality. But it is also true of the other two objects of epistemic work. To argue what needs to be done on the basis of the current state of affairs always includes a normative element, an idea about good or decent. And to say anything about what must be done implies actors and what they identify with.

As to how actors employ these three dimensions of epistemic work, we suggest a distinction can be made between paradigmatic and practical dimensions. By the paradigmatic dimension we refer to basic ontological premises. The fact that we think of ‘society’ or ‘economy’ as objects of governance indicates an initial grasp, or assumption, of what these categories mean. This ontological aspect of what Charles Taylor (2004) calls the ‘background understanding,’ is a pre-conceptual basis on which our further understanding of facts and measurements about a particular policy is founded. Particular actions make sense only in relation to this broader, deeper grasp on our reality, just as that grasp can only be realized in specific actions. Of course, in our account the paradigmatic–practical distinction is an analytical one rather than an ontological ‘fact.’ Another way of putting this is to relate to sensemaking: the paradigmatic dimension is akin to institutional conditions within which the routinized practices make sense, while the conditions themselves can only make sense in relation to the specific practices that institute and reproduce them.

Affecting the shared view of what is a truthful and accurate picture of the situation at hand is an essential element of epistemic work in most cases. If, for instance, a politician is able to show that her nation is not functioning well in a policy area, it is a strong argument for a reform or change of government. In such a case the government may either try to prove that reform measures are already underway or present evidence that challenges the bleak picture of the situation, in either case reinforcing the ontology of the environment. For instance, Zeev Rosenhek (2013) has pointed out how diagnoses and explanations of the ongoing global financial crisis by the US Federal Reserve and the European Central Bank built their epistemic authority and reinforced ideas of the nature of the global financial system, which led to particular policy prescriptions.

Considering ontology of the social world as the object of epistemic work, we propose that there are two aspects to, or dimensions in, it: paradigmatic and practical. Arguments used to justify or criticize a policy proposal relying on ontological claims often entail elements along both these dimensions. To take an example of paradigmatic premises used in political argumentation, one of the ways to justify a reform in a national context is to resort to the idea of modernization as a natural process that steers national trajectories in the right direction. This understanding of societies following a law-governed trajectory is coupled with a Darwinian evolutionary idea, according to which the direction of social change is determined by the acts with which states and other actors adapt to changing external conditions (Alasuutari, 2011b). Presuming an a-cultural view of societies ‘evolving’ through a universal developmental process, this view relies heavily on the authority of science and Enlightenment notions of progress (Taylor, 1999). This cultural model implies that a reform is tangential (on this concept, see Qadir, 2011) with development by referring to it as an act that ‘modernizes’ the area in question. As a culturally ingrained background understanding it feeds off and reinforces how, in a number of tacit ways, our popular concepts promote reforms and newness that are presented as better than the old or existing state of affairs. Thus, for instance, social change is routinely called development, differences between different countries are placed within an evolutionary continuum by talking about developing and developed countries, and existing practices may be deemed as ‘outdated.’ It is also common to refer to different countries as ‘leaders’ or ‘laggards’ in their policies, thus implying that the direction of change is already known and that different countries and cultures can be placed in a quasi-temporal order regarding their developmental stage.

Actors may, of course, take part in political argumentation by challenging such paradigmatic premises of the social world and appealing to other ones. But considering the tacit character of background understanding it is probably more common that the basic premises are not questioned and instead, participants are engaged in epistemic work in the practical dimension while reproducing the epistemic regime. That is, they produce knowledge based on the established paradigmatic foundation and utilize existing facts and measurements of society. In other words actors typically try to appeal to authoritative accounts of reality, which is why science and research-based institutions play a significant role. It is not only that science has an authoritative position in definitions of reality. In addition, knowledge production that takes place in an institution that is independent from direct government or other political control is considered more credible than other knowledge claims, facts, and figures. For the same reason, government research institutes or intergovernmental organizations such as the Organisation for Economic Co-operation and Development (OECD) pay much explicit attention to ascertaining that their knowledge production is unbiased by a particular political agenda and that the facts they provide are based on transparent scientific procedures. This does not of course mean that ‘evidence-based policies’ justified by such knowledge claims are removed from the sphere of politics; it is rather that political actors try to justify their views by science both in defining a problem and in ending up with what can be called policy-based evidence (Black, 2001; Hughes, 2007; Naughton, 2005). The way in which the OECD's Programme for International Student Assessment (PISA) is used to justify reforms in national contexts is a case in point (Pons, 2012; Rautalin & Alasuutari, 2009; Takayama, 2010).

Influencing actors’ conceptions of reality and utilizing them in policy-making does not mean that participants need to accept the definition of a situation that outlines political decisions. What is at stake here is the belief of political actors (leaders and popular activists or commentators) that a particular definition of a situation is accepted broadly enough to be used when influencing others. Whether or not individuals believe in assumptions about the ontology of the environment, what is crucial here is their action based on the belief that those assumptions are socially established: they know that this is the way people think of things. Affecting the view of policy-makers is important, but the public view creates resources or constraints for political actors (Adut, 2012; Hajer, 2009) particularly in liberal democracies, which is why the media and the public sphere are important in epistemic work.

Another key aspect of epistemic governance is to work upon people's understandings of themselves and others as actors: who they are, what community they belong to, and what other actors there are in the social world. Both national and international politics and governance are commonly seen as activities in which different entities or stakeholder groups seek their own interests, and in which future policies depend on the relative strengths of the entities in question. But such a view ignores the identity politics by which actors try to influence each other's identifications with and definitions of the groups in question. When a collective actor such as a nation-state or a political party is well established, it appears to be a given entity, but such groups’ existence is conditioned on people's identification with and commitment to them, which entails epistemic work focused on actors and identifications.

In this case, too, we can identify a paradigmatic dimension of background understanding on the one hand and a practical dimension of knowledge production or other practices based on the paradigmatic face, on the other. The paradigmatic dimension concerns established conceptions about actors and agency in the social world and in contemporary societies, particularly assumptions concerning the constellation of forces involved in politics and policy-making. These background assumptions entail ideas about agency, such as an individual as homo economicus rationally pursuing her individual or group interests, and concepts that are used in conceiving of groups and stakeholders such as gender, class, nation, ethnicity, or religion. The paradigmatic aspect of this epistemic work may invoke a particular discourse by which to describe the situation, whereas work in the practical dimension will build arguments based on the tacit assumptions about actors and their identifications.

In addition to such ontological claims, epistemic work upon actors and identifications focuses on making people identify with a proposed group or strengthening their association with it, thus affecting people's activity. Such epistemic work on actors and their identifications is not restricted to language and speech. It may also include ritualistic aspects, which function through individuals’ bodily experiences and through the emotions they arouse. That is, creating a team spirit and identification with the group in question is strengthened by means of rituals, comprising sacred symbols and highly formalized modes of talk and behavior, often coupled with music (Alasuutari, 2004, pp. 93–104; Bell, 1997; Bloch, 1989).

Strengthening and appealing to the nation and national interest is a prime example of epistemic work on actors and identifications. It is so prevalent in everyday life and national political discourse that it easily goes unnoticed as a self-evident framework within which to view the world and to debate policy-making. Deemed banal nationalism by Michael Billig (1995), its core aspect is identification with the nation, which is something that nation-states reproduce by invented traditions (Hobsbawm & Ranger, 1983) or commemorative rituals (Spillman, 1997). According to Billig, banal nationalism is also and most successfully reproduced through mundane ‘flaggings’ of the nation in routine everyday communication that policy-makers then draw on when promoting their views. This kind of rhetoric about the nation as an imagined community (Anderson, 1991) with shared interests is endemic throughout the world, for instance in environmental policy discourse. Subscribing to ‘ecological modernization’ theory and rhetoric (Spaargaren & Mol, 1992; York & Rosa, 2003), all countries and economic areas such as the European Union seem to aim at becoming world leaders in clean energy technology, thus benefiting from ‘green growth.’ A similar identification of actors with an even more imagined community is evident in the way conceptions of the world as divided into competing civilizational blocs shaped the effect of the Danish cartoon controversy of 2005 (Eide, Kunelius, & Phillips, 2008; Lindekilde, Mouritsen, & Zapata-Barrero, 2009). Informing national politics as well as international diplomacy, the event was used by actors to discursively construct a new institutionalized player, the ummah or ‘Muslim world,’ to which actors related themselves in various ways (Powers, 2008; Saunders, 2008).

Such examples illustrate well how constructing and appealing to actors’ identities functions as a tool of politics and governance. There is, of course, a long-noted political strategy of labeling certain actors as illegitimate (communist, capitalist, terrorist, etc.). However, we wish to draw attention here to, so to speak, the ‘positive’ aspect of appealing to actor identifications in order to motivate legitimate action. So, speakers may appeal to a seemingly self-evident community such as the nation, within which members are supposed to have the same interests. Building and strengthening such patriotism are in themselves powerful means of epistemic governance. On the other hand appealing to the shared interest is a strategy by which different parties and stakeholder groups advance their own interests while also defining themselves as collective actors in such a way that a large number of citizens supports and identifies with them. The aim is to present partial group interests as the common interest, and hence to be a major player in epistemic governance. The way in which the new terminology of ‘war on terror’ was domesticated across the world as ‘our’ issue, relevant to the survival of each nation, almost independently, is another good example of this (Qadir & Alasuutari, 2013).

Epistemic work is, lastly, focused on people's norms and ideals and the way they are applied to particular cases. If and when an actor is able to convince others about the right thing to do by appealing to a general ideal or principle and what it obliges or allows people to do in a given situation, it is a persuasive way to steer their behavior. This is similar to the point Foucault was approaching in his later work regarding the construction of ethics and technologies of the self (Foucault, 1980). Commonly held values and principles such as rationality, equality, freedom, and fraternity or loyalty toward fellow citizens, are also part of world culture, which is one further reason for global isomorphism, although naturally norms diffuse not in isolation but as part of ontological and epistemological packages. On top of such general values, actors engaged in this form of epistemic work create ever-new ideals such as ‘child-centeredness’ or ‘sustainability,’ which often quickly circulate across the world when they become popular and actors justify their political goals by applying them.

Again, we can identify a paradigmatic aspect of background understanding of norms and ideals. Any particular notion of what is good or desirable in a given situation presupposes a broader sense of what ‘good’ and ‘desirable’ mean, and a sense that actions can be classified as such. This typically includes a notion of morality based on natural rights, in which people have mutual rights and obligations in respect of one another, as well as a sense of a moral hierarchy of those rights and obligations and also whether they apply reciprocally or vary by social position. Furthermore, as Charles Taylor (2004) points out about the moral background, this paradigmatic dimension comprises a grasp on what makes particular ideals realizable, and not just utopic. Rhetoric used to convince others of virtuous and desirable actions reinforces this paradigmatic background understanding, which is materialized in practical arguments for or against a particular action. In this second, practical dimension of argumentation, too, assumptions about particular norms and ideals may be explicit or implicit.

Norms and ideals used in justifying a policy often derive their rhetorical force from their apparent self-evidence, hence setting the agenda for political struggles. To take an example, the noble objective of providing all children with the opportunity to fulfill their potential is prevalently used in promoting early childhood education policies, despite huge variations in actual policy objectives (Alasuutari & Alasuutari, 2012; Mahon, 2010). Epistemic work focusing on norms and ideals may also utilize understandings about sacred, highly emotional, or religious principles, of which the Danish cartoon controversy is a good example. Many Western participants in that conflict articulated the protests as an outrage against freedom of speech as a highly valued principle, whereas most Muslims interpreted the publication of the cartoons as baiting and blasphemous.

Epistemic work drawing on norms and ideals is not detached, however, from claims on rationality and scientific evidence. For instance universal principles such as the right to freedom of speech or other human rights are often considered as self-evident starting points and functional requirements of social and economic development in any society, backed up not only by United Nations’ declarations but also by scientific authorities. The same is true of new ideals such as individually tailored education or support. In particular, the catchwords that ‘fly’ and spread throughout the world are not considered as norms honored only by a group of believers but rather are viewed as upshots of rational, evidence-based development toward a better world. They appear to be self-evident values, which is why they function so well in promoting reforms articulated with them.

Appealing to commonly valued ideals and articulating a reform with them is an effective way to promote the reform in question but it is not, of course, the only mode of epistemic work upon norms and ideals. Policy-makers may for instance point out that the current state of affairs is in contradiction with an international treaty or declaration, which is why a reform is needed to avoid sanctions or to maintain the international reputation of the country in question.

We have tried to show in this paper that governance works by way of influencing other actors’ beliefs and views of the way in which the social world is organized, the role and subject positions of actors, and the nature of ethical or valued goals of social change. In other words, epistemic work revolves around constructions of what the world is, who we are, and what is good or desirable. When anybody attempts to affect policy change, even by way of commentary, they resort to challenging or influencing these three conceptions. As we claimed earlier, there is no epistemic work that does not operate on all these three objects. Moreover, we distinguished between paradigmatic and practical dimensions in each object of epistemic work, indicating a distinction between elements of broader, more amorphous background understanding and elements of narrower, more focused actions in which that understanding is realized.

This approach to epistemic governance, again, does not posit a new theory of power but rather pushes existing theories to include a thicker description of the politics of policy-making, a processual layer of epistemic work that most theories tend to brush aside. For example, from an epistemic governance perspective we would argue that even rational actors depicted by realist theories will attempt to influence the definition of the situation, what actors should identify with, and what is virtuous or acceptable. That is, we urge realist actor-centered theories to acknowledge contexts, webs of meaning, and epistemic assumptions. At the same time we argue that structural theories need to acknowledge interpersonal meaning construction and actors’ agency when describing the processes of policy reform.

We see that the ways actors attempt to convince each other about a policy are based on globally shared epistemic assumptions about the social world. But this does not mean that world leaders are unthinking conformists hypocritically enacting foreign policy models that they know won't fit their local context. Rather, world cultural ideas and ideals form the premises that actors appeal to in various combinations when articulating their goals and aspirations. In that sense, world culture forms the global field of politics and policy-making and its various national and local subfields within which actors struggle for their views, benefits, and influence.

This consideration of epistemic governance is also not yet another refined conspiracy theory about someone or some people ‘high up’ or behind the scenes pulling strings. Rather, our purpose has been to indicate how conceptions of the social world are utilized as tools of governance in the normal course of events, not necessarily with the aim of oppressing or controlling another group but as part of routine decision-making. This means that we consider epistemic governance not as something ‘done’ by particular actors but rather as something that is analyzable in processes of decision-making.

When analyzing power from this viewpoint we need to bear in mind that affecting or maintaining hegemonic views of society is part of epistemic governance. As Foucault famously pointed out, power is particularly successful when it is able to hide its own mechanisms, and so the simple view of power as external coercion, as a pure limit set on freedom, is part of its acceptability (Foucault, 1980, p. 86). Hence, decisions are often successfully justified by picturing a ‘must’ and a loss of sovereignty around the corner if the proposed measures are not taken. Similarly, the conclusion that a particular policy is a necessity is an outcome of a negotiation concerning the situation at hand. Affecting actors’ wishes and aspirations is perhaps most effective because the decisions that ensue appear to be untouched by relations of dominance. In other words, epistemic governance depicts not a particular form of power but rather an ever-present aspect of governance: the point is that strategies of power work on people's conceptions of the situation. We have not examined in any detail here the kinds of notions of society that actors work on, and how they relate social scientific conceptions and institutions to popular imageries, but we believe this to be an important part of epistemic work and worth further analysis. This raises an important question about the phenomenological context of the social psychology underlying epistemic governance, including the place of constitutive Othering and exclusion, which, again, we believe is worth examining.

Also in line with Foucault's broad conception of power, it is clear that epistemic governance does not necessarily work from a single center, and is not necessarily controlled by an identifiable, elite group. Rather, actors involved in contesting definitions of situation and appropriate solutions to evolving problems end up synchronizing social change on a global scale through knowledge production, circulation, and opinion formation by institutions such as science, its cultural mediators, and the media. The micrological channels that help construct and naturalize such epistemes include the mass media, educational institutions, and transnational professional associations. However, much more empirical research is needed to better understand how specific institutions, and constellations of institutions, contribute to epistemic governance in a particular case.

Contrary to some approaches, we therefore argue that the actors who make decisions are important. Likewise, we acknowledge that some actors have access to more resources by which to influence others, and such resources may well be classified in different ways, for instance as ideological, economic, military, or political (Hall & Schroeder, 2006; Mann, 1986). Yet whether actors use, say, science, money, or tanks as their consultants, the objective in utilizing those resources is to convince others of what they want, or what they should or must do in a given situation, and such convincing is epistemic in nature in the sense that its target is people's conceptions of reality, of themselves and others as actors and of norms and ideals. And even though actors and their resources make a difference, that is not to say that actors are a pre-given reality, and hence un-constituted. Rather, we emphasize that although actors and actorhood may be constituted institutionally, there are no inevitable mechanisms or trends of social change that all societies must follow and that we as social scientists can discern as hiding ‘behind’ evident social change. Change is a result of actors making decisions, or attempting to influence society, by way of the three objects of epistemic work. Of course, the way people behave in a particular situation and what repertoire of rhetoric gets used can only be said after close scrutiny.

Crucial to such investigations, and this way of viewing governance as epistemic work, is special attention to the use of particular language to create discourses that reflect the construction of epistemes wherein the view of the world and desirable policy actions are shaped. Understanding the discourse that generates and sustains such epistemic work becomes crucial to the analytical process of understanding how governance actually works across the world. There is need for much closer examination of particular terms and how they are used in motivating or opposing a specific prescription. However, rather than attempting to attach some metaphysical ‘meaning’ to such terms, we call for paying closer attention to the grammar of their use. It is by such examinations of epistemic work that we can move forward from more or less structuralist explanations of sources and forms of power to reveal the strategies of power at play in specific attempts to influence change in the contemporary world.

The authors are grateful for inputs from, and numerous discussions with, Risto Heiskala, Anne Kovalainen, Risto Kunelius, Seppo Poutanen, and George Thomas. We also gratefully acknowledge comments from scholars in the Tampere research group on Cultural and Political Sociology, whose empirical research this article is based on and cites.