The cultural construction of global social policy: Theorizing formations and transformations

Introduction

As a field of study institutionalizes, claims for stronger theoretical underpinnings, conceptual clarifications and distinct analytical approaches emerge as important issues for defining the basis and boundaries of research. This has not only clearly been the case for the fields of sociology or history but for each of the (sub)fields that situate themselves within – and increasingly across – established disciplines. Different contributions sometimes counteract each other, but at other times they multiply their efforts and analytical resources to enhance the reflexivity of a field of study. The contributions in this special issue might differ with regard to specific suggestions of the how, but they all agree in the need to ‘theorize’ global social policy. This attempt is a reflection of a problem observed as the field emerges and grows: How to provide for conceptual tools that allow for situating particular types of studies within a broader context? Which analytical heuristics are able to transcend a specific case and prove to be meaningful for comparisons as well as for the further institutionalization of the field of global social policy?

In this article, I argue that a combination of world society and discourse theory is fruitful for the endeavour to theorize global social policy. ¹ The argument is developed in four steps: In the first part, the core term of global social policy is under scrutiny: What is meant by ‘global’ in ‘global social policy’? Here, the first conceptual similarities – zones of contact – between world society theory and global social policy research will be pointed out. The second section elaborates on the core concepts, assumptions and arguments of world society theory and points out the parallels between world society and discourse theory. The third section presents a discourse theoretical world society perspective on global social policy and discusses how to study and analyse the formation and transformation of global social policy. In the fourth part, this cultural and constructivist approach to global social policy is exemplified with regard to the institutionalization of early childhood as a global issue. The article concludes by summarizing the theoretical and analytical consequences of these suggestions for global social policy research.

Definitions: How to define global social policy? Or: Some remarks on the ‘global’ in ‘global social policy’

One part of the definition of global social policy (GSP) that seems to be troubling in every attempt to theoretically ground it is the term ‘global’. ² Like ‘globalization’, this term is not only used frequently and in a variety of contexts (Guillén, 2001) but it remains – to some extent due to the expansion and explosion of its use – opaque and problematic (Selchow, 2007; Stäheli, 2003a). The definition of the term ‘global’ is necessary in order to reflect on what the field of GSP encompasses: What is meant by global social policy? How is social policy as ‘global’ differentiated from national (and from international and supranational) social policy? Here, I distinguish four different definitions of ‘global’: global as inter-, trans- and supranational; global as a distinct and emerging level of social reality; global as universal; and global as worldwide.

A first approach to defining ‘global’ refers to processes and structures that are distinct from and transcend national borders, including international processes between nation-states and international organizations; supranational, referring to the transfer of sovereignty to higher levels; and transnational, to processes below or beside state–state relations, involving non-state actors such as civil society, NGOs and TNCs. From this perspective, global social policy encompasses international, supranational, transnational and regional social policy. The challenge that comes with this definition of global is to not merely replace these terms or to use them interchangeably. One option here is to define ‘global’ as an overarching conceptual term for referring to all forms of social policy that are not (or no longer) bound to national borders, encompassing inter-, supra- and transnational processes, conceptually linking them and studying their relationships.

Another suggestion is put forward by a second approach of defining the global: to distinguish global as a distinct and emergent level of social reality. ³ In this definition, the global is often positioned in relation to the ‘local’. It is the conceptualization of the global–local relationship that is challenging for this definition of the global. Especially a ‘realistic’ understanding of the global – based on the assumption of different levels of social reality that have empirical reference – ontologizes the global–local distinction and struggles with the question of which level has more explanatory value or quality: Is the global influenced (or constituted) by the local, or is the local influenced (or determined) by the global? Different scholars have criticized this ‘ontologization’ of the global and the juxtaposition of the global and the local (Robertson, 2001; Selchow, 2007; Stäheli, 2003a). Whereas it is possible to argue that the global is always localized (Collier and Ong, 2005: 4) and that the construction of global policy models takes place in very specific settings, it is questionable if the dimensions and features of ‘global’ processes can solely be explained by reducing them to ‘local’ interactions and structures. A definition of the global as a distinct and emergent level can be analytically fruitful if the global–local distinction clearly refers to different levels of analysis. ⁴ Based on an analytical (and not ontological) distinction, global and local are regarded as complementary perspectives or modes of describing social reality, focusing on processes on a level of either social (inter)action or institutionalized practices and conventions.

A third suggestion refers to global social policy as social policy that is built on universal norms and values. Here, as in the first definition, terminological replacement is involved, in this case, the use of the term ‘global’ interchangeably with the term ‘universal’. Comparing the current sociological interest in ‘globalization’ to classical sociology, which was interested in social universals and temporal differences between modern and pre-modern times, Therborn (2000: 149) argues that the ‘substitution of the global for the universal’ has been accompanied by ‘a substitution of space for time’. The problem with such a substitution is that the conceptual distinction between global and universal is blurred. The term ‘global’ brings into play a (re)territorialized notion of the social, which is not bound to (Western) sociology’s search for ‘social universals’, but imbued with an interest in the territorial dimension of sociality. Such a conceptualization of the social as spatial ‘implies at least three crucial differences with classical universalism, [namely] variability, connectivity, and inter-communication’ (Therborn, 2000: 149), thus an awareness of differences, of relations and of the need for cooperation. In using the terms global and universal interchangeably, these differences become indistinct. Stichweh (2003) thus proposes to differentiate more strongly between the two concepts. He defines ‘universal’ as a social category related to claims of universal validity that are true for everyone, and ‘global’ as a spatial category referring to worldwide institutionalization or diffusion.

A fourth approach points out that national social policy itself is not only increasingly bound to global spheres of influence, but is, moreover, in itself an expression of the global: nation-state formation itself is the far side of globalization (Meyer et al., 1997). In this understanding, the spread of national social policies worldwide can be analysed as global social policy. This use of ‘global’ as ‘worldwide’ is indeed another common strand of defining the global. Here, global is used to refer to policies, institutions, practices, or norms that can be found in a number of places – mostly countries – around the world (e.g. Schofer, 2003). However, the vague formulation ‘a number of countries’ indicates already the limitations of the definition of global as worldwide: How many countries need to have a certain social policy in place to argue that this policy is global? Does ‘worldwide’ require a defined number to be truly global, and is there a (numerical) tipping point from which a policy (pension, health, child care) can be said to have diffused globally? Mostly, the use of the term global as worldwide thus merely indicates a tendency; the tendency of diffusing, travelling, institutionalizing in different countries and regions around the world, without explicit reference to numbers of social units as defining elements.

Taking these four definitions into account: How has the ‘global’ been defined in global social policy so far? Deacon (2007: 1) proposes a two-fold definition of global social policy: ‘first, it is the social policy prescriptions for national social policy being articulated by international organisations; second, it is the emerging supranational social policies and mechanisms of global redistribution, global social regulation and global social rights’ (the three global social Rs). In his earlier work, Deacon refers to GSP as the ‘supranationalisation or globalisation of social policy’ and the ‘socialisation of global politics’ (1997: 3). While the first dimension refers to the three global social Rs, the second brings attention to the changing content of international cooperation. In both cases, the term global is used in two regards. First, ‘global’ indicates processes and structures that are distinct from national ones and refers to a distinct and emerging level of global social governance (which is however often equated with supranational and international). Second, the definition of global as ‘worldwide’ clearly comes into play here, for example when Deacon (2010) states that GSP analyses how the policy views and recommendations formulated on the global level (by international actors) shape national social policies. The argument involved here – but not made explicit – is that social policies are not just spreading worldwide across countries, but that the global (worldwide or cross-national) diffusion is related to the emergence of a global (supranational) level of governance. It is specifically this two-fold definition where parallels between global social policy research and world society theory exist.

In the understanding of world society theory, ‘globalization is a dual process: global institutionalization of world society and diffusion of world societal models to social units embedded in the global environment’ (Drori, 2008: 453). These two dimensions of globalization are regarded as complementary: ‘similar issues – such as environmentalism, health, or rights – are institutionalized cross-nationally (as national ministries) and globally (as international organizations)’ (Drori, 2008: 453). We find here a conceptually elaborated version of the ‘global’ that is implied in global social policy research as well. While the notion of cross-national influences on social policies is already commonly accepted in political science and sociology, the idea of the emergence of a global level (world culture and world polity) has been strongly put forward by world society theories. These theories argue that the consolidation of a global level influences the worldwide diffusion of policies and structures and vice versa: ‘globalization is a co-constitutive dual process: there are reinforcing relations between the cross-national and the global’ (Drori, 2008: 454). Globalization in this understanding thus encompasses more than interrelations, interconnection, exchange, or worldwide transactions. Besides this ‘horizontal’ dimension of globalization processes, which emphasizes linkages and flows, a vertical dimension focuses on the emergence of new ‘global’ forms, of ‘global’ structures of ordering and the constitution of an ‘awareness of the world as a whole’ (Robertson, 2001). In such an understanding, global social policy would refer to the worldwide diffusion and to the global-level institutionalization of social policy, and indeed GSP research has referred to global social policy as transcending national borders and as emerging on a global level. As this co-constitutive duality is what is implied in definitions of global social policy as well, it seems productive to ask how world society theory is developing its arguments. Contrary to Deacon’s (2010) critique on world society theory, this article argues for the fruitfulness of linking global social policy research to world society theory.

Theories: How to conceptualize global social policy? Or: Why world society theory would work

Deacon’s (2010) reading of world society theory centres on a critique of ‘too little’ and ‘too much’: what is missing in world society theory, according to Deacon, are actors and conflicts; overemphasized are homogeneity and the unity of the world society. In the following section, core concepts, assumptions and arguments of world society theory will be laid out briefly in order to discuss if – despite Deacon’s critique – world society theory can be theoretically and heuristically fruitful for global social policy studies.

Central for neo-institutionalist world society theory is a conception of institutions based in the phenomenological thinking of Berger and Luckmann (1967), which has been influential for the development of this theory. World society theory defines institutions as ‘whole edited models of the world and effective activity in it, culturally (Czarniawska and Sevón, eds. 1996); and whole arrangements of organizations and roles and relations, structurally’ (Meyer, 2009: 41). These two perspectives on institutions, the cultural and the structural sides, are referred to in world society theory by two concepts: world culture and world polity.

The concept of world culture that constructs and legitimizes models of the individual, organizations and the nation-state as ‘actors’ – and of ‘the world as enactment of culture’ (Meyer et al., 1997: 151) – is one of the central concepts of world society theory. By diffusing the general principles of individualism, universalism, rationality and notions of social progress condensed in institutions globally, world culture is seen to lead to unexpected homogenization of structures and behaviour of these constructed entities (Meyer et al., 1997: 152ff.). World culture is generally used as an independent variable explaining worldwide isomorphism through diffusion; however, the ‘development of the world polity or culture itself is … recently emerging as a more prominent topic of inquiry’ (Schneiberg and Clemens, 2006: 197; see also Lechner, 2009; Lechner and Boli, 2005). One strong conceptual and analytical approach to the analysis of world culture can be found in Lechner and Boli (2005). With the two concepts of ‘cultural universalism’ (referring to Roland Robertson) and ‘world polity’, Lechner and Boli (2005) present two modalities that allow for recognizing world culture. Referring to Berger and Luckmann, they define ‘culture’ as collectively shared symbolic rules that operate through language and are embedded in cultural products (objects, techniques), but also in organizations and individuals via principles, norms and social behaviour (Lechner and Boli, 2005: 16). Culture itself constructs social identities by assigning reality as well as meaning and legitimacy to them, i.e. ‘it defines the ontological value of actors and action’ (Meyer et al., 1987: 22). With reference to ‘cultural universalism’, Lechner and Boli (2005: 21) define those concepts as world cultural that are intended and perceived in form and content as universally valid, appropriate, or useful. The two logics that fundamentally infuse world culture are rationalization (involving systematization, standardization, scientization) and actorhood (Drori, 2008: 461). Besides ‘cultural universalism’, the second notion for the definition of world culture is the linkage to ‘world polity’. World polity is conceived as a social unit operating and being integrated on a global level through governance and authority structures which refer to the world as a whole (Lechner and Boli, 2005: 20). Especially international organizations, science, professions and the nation-state are structures of world polity; these structures are ‘materializations’ of world cultural principles. Drori (2008: 457) summarizes the features of modern world polity: ‘it is (a) expansive, (b) heterogeneous, in both organizational forms and substantive issues, (c) dynamic, and (d) loosely organized and highly decentralized’. With its ‘straightforward focus on culture as a fundamental social factor’ (Drori, 2008: 462), world society theory defines globalization as global institutionalization and cross-national diffusion, and these processes of globalization are part of the emergence and diffusion of world culture and world polity in the world society – including contradicting and conflicting models. Linking global social policy to the concept of world culture thus emphasizes that global social policy can be understood not only as social policy spreading worldwide but moreover as social policy linked to institutions and structures operating within a global scope and to content presented and perceived as universal. More importantly, however, a combination of the concepts of world polity and cultural universalism with thoughts on the social constitution of reality can shed light on the question of how something is constituted as global social policy. It is this interest in construction processes – in the formation and transformation of global social policy – that requires a different analytical approach, which will be laid out in the following section.

This constructivist basis also implies that – contrary to Deacon’s (2010) first critique – world society theory does not assume a united or unified system. Although the astonishing isomorphism over time between the structures of different social units has been pointed out, world society scholars take a constructivist stance: global influences upon national policy are analysed as worldwide diffusion processes, but world culture is not a point into which everything – even the contradictory – merges but rather a point of resonance, constructed as well as referred to in specific ways. Indeed, world society scholars (Meyer, 2009: 53, 58; Ramirez, 2010) have referred to Anderson’s (1991) ‘imagined community’ to emphasize that world society or polity is an imagined community, not a united or unified system. World society theory thus theoretically proposes a phenomenological understanding of the world as horizon (Krücken and Drori, 2009: 16). Analytically, however, this understanding needs to translate more clearly into correspondent research strategies, challenging unidirectional and top-down diffusion analyses to include analyses of world cultural construction processes. In analysing such construction processes, ‘acknowledging heterogeneity challenges conventional images of causality and pushes institutional analysis away from strong forms of structural determinism to a much greater emphasis on agency, conflict, contingency, and process’ (Schneiberg and Clemens, 2006: 214ff.). ⁵

The lack of attention to actors is Deacon’s (2010) second main point of critique. One of the main differences to global social policy studies is indeed the use of ‘culture’, not of ‘actor’ as explanatory concept, as ‘defining dimension of society’ (Krücken and Drori, 2009: 16). From the perspective of world society theory, not only social action, but also actorhood itself is culturally constructed and constituted. Thus actors derive their purpose, rationality and sovereignty from the institutional context (Meyer, 2009: 41; Meyer and Jepperson, 2000). Cultural constructions ‘work by defining actors at the front and centre of the social stage’ (Meyer, 2009: 57), including interests and goals that actors pursue. The most powerful actors today – to which high levels of agency are ascribed – are individuals, organizations and nation-states, and it is a historically recent phenomenon that these actors are seen as imbued with agency. Actors – individuals as well as organizations and nation-states – are thus not regarded as ontological entities, but social entities that are built on ‘models of proper actorhood’ (Meyer, 2009: 47) which are produced in socio-historical processes. World society scholars differ from each other with regard to how much agency they conceptually ascribe to actors. ⁶ From a Goffmanian (1959) perspective, individuals, organization and nation-states are seen as ‘agents’ on a social stage, acting according to a script that they perform. This perspective ascribes a higher level of agency to actors than the concepts of ‘carriers’ and ‘markers’. As ‘carriers’, individuals, organizations and states are conceptualized as conveyors of ideas that channel scripts and deliver them. Even less agency is assumed at the other end of the conceptual spectrum: ‘markers’ or ‘signifiers’ are regarded as sites where practice is expressed. In this understanding, individuals, organizations and states are signals for global processes, for something broader than themselves. Change is reflected in these entities, not constructed by them. Despite the different levels of agency ascribed to actors, in all three conceptualizations, is the cultural construction of actorhood is clearly one of the core assumptions of world society research. The processes of change that world society research analyses are not attributed to (individual) actors but to larger cultural processes, logics and mechanisms. Rather than assuming agency, world society research analyses how agency is produced in socio-historical processes, how differences and ‘power’ are constructed and how they relate to processes of change. Here, world society theory’s (de)constructivist perspective on social reality is discernible, and indeed the parallels to and epistemological commonalities with Foucault are increasingly pointed out by world society scholars (cf. Meyer, 2009: 43).

In the following, I argue that a discourse theoretical perspective on globalization processes, as it is slowly emerging in world society research, is analytically and methodologically inspiring for global social policy research. ⁷ Such a perspective allows for analysing how global social policy itself has been and is constituted and constructed, how it could be established and in which forms, how certain ideas are included, institutions are emerging, instruments are being developed and actor positions are created and assigned that call for the expression of ‘agency’. This (not only) historical perspective asks for the conditions for the formation and transformation of elements of social policy that become ‘global’ and provide (new) frameworks for orienting national social policy and the work of ‘global actors’ (IGOs, INGOs, experts, consultants, epistemic communities). The basis for these reflections is the assumption of a constitutive character of social reality, leading to an interest of how forms of knowledge (including ‘truth’ and ‘reality’) are produced in historical processes (Landwehr, 2008). I argue that it makes sense to apply the notion of the constitutive character of social reality as well as the interest in processes by which knowledge is produced to GSP research in order to analyse the formation and transformation of global social policy.

Analytical strategies: How to study global social policy? Or: Why Foucault has to enter the ‘global’

The attempt to introduce a Foucauldian perspective of world society research to GSP is based on two parallel assumptions of world society theory and discourse theory: both assume a culturally constituted and constructed character of social reality, and for both the assumption of a cultural construction of actors is essential. In this section, I will first approach global social policy research from a discourse theoretical perspective; second, link this perspective back to world society theory; and third, discuss analytical heuristics for a discourse theoretical world society perspective on global social policy.

Applying Foucault to GSP analysis and interpretation means to approach global social policy as a field of knowledge and a system of rules of formation, linked to institutional and technical forms as well as different practices (Foucault, 1972, 2002 [1966]). Foucault’s discourse analysis aims at reconstructing the discursive formations that govern the appearance and dispersion of statements at a specific moment in history (Foucault, 1972). In his Archaeology of Knowledge, Foucault elaborates on his conceptualization of discourse ‘as the group of statements that belong to a single system of formation’ (1972: 107). Foucault focuses on the formation of knowledge within scientific disciplines using examples from his earlier work on The Order of Things (2002 [1966]) as the main source. However, asking himself during the course of his arguments whether ‘archaeological’ (discursive) analyses only apply to epistemological figures and science, he refers to the realm of political ‘knowledge’ (knowledge used in a broad sense, referring also to behaviour and practices). An archaeology of political knowledge analyses in which way language and behaviour are permeated – and thus allowed and restricted – by discursive practice. Discourses are treated not as symbols ‘but as practices that systematically form the objects of which they speak’ (Foucault, 1972: 49). Historically specific rules of formation are constitutive for the formation of ‘objects, modes of statement, concepts, thematic choices’ (1972: 38); they constitute the way that something can be referred to as an object, which forms statements can take, which concepts (terms) can be used and which strategies can be employed. In order to analyse the formation and transformation of ‘political knowledge’, the question of which episteme ‘exists’ is linked to their relationship with modes of decision and behaviour, cooperation and conflict, struggles and tactics (Foucault, 1972: 194). Focusing on the analysis of discourses in his Archaeology, Foucault formulates here an early interest for the relationship between knowledge and power, which he will more distinctively elaborate in his later work. The relationship between power and knowledge is important when analysing political processes, as they are mutually constitutive, thus enabling and restricting the emergence of ideas and their institutionalization. In later publications, the question of how ‘knowledge’ is formed is complemented by the question of their effects, especially on how individuals are produced as ‘subjects’ first of all. As has been pointed out, this question is of great interest for world society theory, as it assumes that individuals (as well as organizations and the nation-state) are culturally constituted, in Foucault’s terms: discursively produced by orders and rules of knowledge.

In discourse theory, the analysis of the formation and transformation of ‘subjects’ as well as of ‘political knowledge’ seems to be thought of however only in relation to the knowledge and practices of a specific society or social group. In order to be able to provide heuristics for analysing the formation and transformation of global social policy, discourse analysis has to adopt such a ‘global perspective’. As both theories share claims about the cultural constitution and construction of actors and social reality, world society theory can provide such a global perspective for discourse analysis.

Which analytical heuristics are provided by world society and discourse theory that would allow for analysing the constitution and construction of global social policy? World society research rests on the important claim that world culture as a ‘virtual sender’ is decisive for the institutionalization and diffusion of models. The strategies common in world culture research require data for longitudinal and large-N analyses, and this renders the ‘cultural’ operationalization of ‘world culture’ difficult. Thus, the influence of world culture has mostly been operationalized through political-structural variables such as linkages to international organizations via membership or the ratification of conventions for which longitudinal and large-N data are available. In recent years, however, correlation, regression and event history analyses have been increasingly complemented by analyses of discourses and discursive construction processes (cf. Berkovitch [1999] on women’s rights; Drori [2006] on governance), and new conceptual as well as analytical tools have been developed that focus on the ‘dynamics of meaning’ (Zilber, 2008), such as the concept of ‘theorization’ (Strang and Meyer, 1993), the suggestion of Czarniawska and Joerges (1996) to use the concepts of ‘translation’ and ‘editing’ instead of ‘diffusion’, and the discourse analytical perspective on institutionalization as proposed by Phillips et al. (2004) (see Zilber, 2008: 161–163). The concept of ‘theorization’, for example, is one of the strong world society theory concepts for analysing the constitutive conditions for institutionalization and diffusion processes. Referring with ‘theorization’ to the development of abstract and rationalized categories and models, Strang and Meyer (1993) argue that social entities – individuals, organizations and states – are constructed in the world polity as modern actors in theoretically complex, standardized and comparable forms. As theorization is especially linked to scientists, experts and consultants, theorized forms gain legitimacy. Both their complex construction and the legitimacy acquired by their relationship to rational science leads to increased diffusion and in the consequence to increasing isomorphism of structures. Theorization in this understanding is one ‘cultural condition for diffusion’ (Strang and Meyer, 1993), and an interesting point of entry for the question of relevant conditions for the institutionalization of cultural conceptions. When analysing the constitution and construction of cultural forms, not only broader context variables need to be taken into account but also the specific semantic processes and mechanisms which enhance the likelihood of concepts and issues to be communicated, to find ‘addresses’ in different geographical and cultural contexts, and to stimulate further communication.

The analytical heuristics proposed by Foucault allow for analysing such processes: How are certain issues constituted as global social policy issues? How are the conditions for communicating those issues changing? Here, the world society emphasis on constitutive conditions is strongly linked to the construction of meaning, and it translates into analyses of how social concerns and policy scripts for dealing with them are produced and positioned as global and legitimate. From such a discourse analytical world society perspective, global social policy is constituted and constructed by processes in which knowledge is produced and legitimized: historically specific rules – embedded in language, practices and objects – allow for the formation of certain statements which can be given from specific institutional positions and which have effects on the formation and transformation of global social policy. In Foucauldian understanding, ‘discourse’ not only refers to the content of statements, but to systems of formation and the logics of language that govern the appearance of statements. Analysing global social policy thus requires focusing on the formation of modes of statements, objects and concepts, and on the positions of enunciations that increase the legitimacy and success of specific models, categories and policies. ⁸ A combination of world society theory with discourse analysis implies that ‘macro’ and ‘micro’ perspectives are linked – to take broader cultural processes such as universalization and theorization into account and to analyse the discursive practices and logics that constitute them. In the next section, I will use the case of early childhood to exemplify how the formation of global social policy can be approached from such a discourse theoretical world society perspective.

‘Making the case for EC’. Or: Analysing the constitution and construction of early childhood as a global social policy issue

The institutionalization of early childhood as a global issue has gained momentum since the 1990s and even accelerated further since the turn of the 21st century (Sugawara, 2008). Two major events mark the culmination of this process. In 2010, the First UN World Conference on Early Childhood Care and Education (ECCE) took place in Moscow (UNESCO, 2011), and the UN General Assembly adopted its first resolution specifically dedicated to the early years (UN GA, 2010). The constitution of early childhood as a global issue is an interesting case for global social policy: the ‘discovery of early childhood’ (May, 1997) is a historically recent phenomenon, and it has long been constituted as a ‘local’ issue, deeply embedded in specific familial and communal settings (LeVine and New, 2008). How could such a historically recent and locally embedded issue institutionalize as a global social concern? In the following, I outline how the institutionalization of early childhood as a global issue can be approached from a discourse analytical world society perspective. With reference to Foucault’s terminology (1972, 2002 [1966]), I suggest analysing conditions for the formation of statements, specifically authoring positions and modes of statements; the formation of objects and concepts; and the formation of strategies. ⁹

The first strand of a discourse and world society theoretical analysis focuses on analysing the conditions for the formation of statements: Where, when and in which mode are statements produced? Two different angles can be used: the analysis of positions of enunciation and the analysis of modes of statements. Analyses of positions of enunciation focus on the institutional and authoring positions from which statements are made, such as international organizations, practice or epistemic communities, scientific disciplines, as well as conferences, workshops, or expert meetings. How do these institutional contexts differ? Who enters into them? Which positions of enunciation do they provide with regard to statements that can be issued?

Especially since the early 1990s, the number of international associations working on early childhood has expanded to a high degree. With this expansion, the sheer number of communications on the early years increased – not only in the form of documents, but also in the form of organizational decisions, of structural relations ranging from more permanent networks such as the Consultative Group for Early Childhood Care and Development to temporary linkages such as the OECD/UNESCO Early Childhood Policy Review Project (UNESCO, 2005), and in new forms of communicating on early childhood, such as the First UN World Conference on ECCE in 2010 (UNESCO, 2011). An overview on the emerging structures for dealing with early childhood globally is important; but the interest of a discourse analytical approach goes beyond organizational histories to identifying and distinguishing authoring positions from which statements can be made in specific ways. Although such positions are sometimes strongly related to one actor, commonly different assemblages of organizations, epistemic communities and individuals represent and mark specific positions of enunciation in different historical periods. ¹⁰

The incorporation of early childhood in the League of Nations is an example for a then newly emerging authoring position – one that positions early childhood as an ‘international concern’ and demands ‘international action’ for young children. With the foundation of the League of Nations in 1919, the ad hoc cooperation between nation-states was transformed into stable and formalized relationships, and within this new structure, issues were discovered as legitimate international concerns (Berkovitch, 1999: 15). The early years were constituted as an issue of international cooperation – at that time however only with regard to health, protection and welfare of young children, not their education or psychosocial development. ¹¹

Analysing the conditions for the formation of statements also requires taking the different modes of statements such as statistics, declarations, flagship publications of international organizations and monitoring mechanisms into account. How are they related to each other and to the different positions of enunciation? By which individual logic are they characterized?

One interesting example is the development and use of indicators that measure the well-being of children worldwide (Ben-Arieh, 2008) and define childhood norms with regard to international policy initiatives such as the Education for All (EFA) initiative, the Millennium Development Goals (MDGs), the UN World Summits for Children and international monitoring instruments such as UNICEF’s Multiple Indicator Cluster Survey (MICS). The core international indicators for early childhood are education and care indicators. Their changes over time – for example the continuity of child survival indicators, the increase of early education indicators, the expansion of well-being indicators beyond health indicators, and especially the recent attempts to develop an index for measuring the implementation of child rights in early childhood – tell a story of the perspectives from which early childhood has been taken into account, and thus about key issue that are regarded as universally relevant in relation to the early years.

Analysing different modes of statements from a world society perspective requires, moreover, taking a closer look at the logics involved that enhance the likelihood of an issue to be addressed across communicative contexts, and at the effects that different modes have on the institutionalization of a ‘global issue’. The construction of indicators, for example, is based on selectivity. Dimensions of early childhood that cannot be quantified and compared across countries will be more difficult to include into international monitoring mechanisms. Especially the logic of quantifications produces objectivity, factuality and legitimacy (Heintz, 2007) and positions indicators as relevant and valid, as the ‘universal language of number’ cannot be easily rejected. Their numerical representation – in form of statistics, rankings, world graphs, tables and classifications – establishes global horizons of observation and comparison that list and rank countries according to their performance on agreed goals, and universalize the meaning and importance of what is measured (Tag, 2009). Indicators thus generate universalizing effects for constituting early childhood as a global issue.

The second strand for analysing the institutionalization of early childhood as a global issue focuses on the formation of objects and concepts. Again, two different sets of questions can be asked. A first perspective asks how statements concerning a specific issue appear. Which terms are used to refer to the early years and what are their implications? In which relation to other semantics do they appear, such as health, education, or development concepts? Which roles and responsibilities are constituted, for example for the state, parents, or professionals? Here, analyses of rhetoric, of arguments and metaphors, but also narratology (see Czarniawska, 2008) provide useful analytical tools. Different arguments for the importance of focusing on the issue of early childhood, for example, constitute different perspectives, and each of them involves different sets of problem definitions, (policy) solutions and taken-for-granted assumptions. ¹² An interesting starting point for the analysis of the formation of objects and concepts are the terms used by different international organizations with which they refer to the early years in their publications and at international conferences: ECCE (Early Childhood Care and Education) in UNESCO, ECD (Early Child Development) in UNICEF, ECCD (Early Childhood Care for Development) promoted by the Consultative Group for Early Childhood Care and Development, or ECEC (Early Childhood Education and Care), the acronym used by the OECD. Each of the terms indicates different perspectives on and priorities for the early years, ranging for example from a focus on early education within UNESCO to a holistic child development perspective in UNICEF.

Here again, the world society interest in constitutive conditions that allow concepts to stimulate further communication and to become taken-for-granted requires taking a closer look at the logics by which objects and concepts such as ‘early childhood’ are constituted as global issues. In 2008, for example, UNICEF, UNESCO and Save the Children jointly organized an international conference for early childhood policy in Asia for governmental representatives from 15 countries, early childhood experts and staff from international organizations. During the months of preparation, the working title for the conference was ‘Making the Case for ECCE/ECD/ECCD’. In the end, however, the simple reference to ‘EC’ was chosen, as it constituted early childhood as a consensual and shared interest on which the agencies could work together, and which would allow the countries to communicate more easily on what was positioned as a common goal. In this sense, ‘EC’ was used as a consent formula to which all agencies and country representatives could refer. Such a consent formula can be interpreted as semantics that allow communication to continue, and it can be compared to some extent with the position of an ‘empty signifier’ (Laclau, 1996) that ‘fixes meaning by providing an horizon for possible meanings’ (Stäheli, 2003b: 8). The reference to ‘EC’ and to specific assumptions in relation to early childhood, such as the formative and foundational effects of early childhood experiences for future life, were based in cultural universalizations, especially as the authority structures involved were international organization representatives and scientific experts. However, the reference to and repetition of the consent formula ‘EC’ does not mean that a homogeneous conception of early childhood was established during the conference. Rather, the reference to an abbreviation and the emphasis on common, consensual knowledge (‘As we all know…’) left empty spaces for the creation (supposition) of meaning: countries and agencies could refer to one common set of terms without addressing different assumptions about the early years. These open spaces for meaning-making in turn fostered referencing the semantics of ‘EC’ and contributed to positioning early childhood as a global issue. ¹³

A third and related research strand focuses on the formation of strategies in the institutionalization of early childhood as a global issue: How are statements legitimized? To which theories do they refer, to which legitimized positions, and how are (established) discourses employed? Which audiences are addressed?

The history of childhood has mostly been described as a story of rationalization – the theorization of childhood in the evolving childhood sciences such as pedagogic, development psychology and child medicine (Turmel, 2008); the measurement of development, including the definition of standards and developmental stages (Morss, 1990); and the evolution of professional roles for child rearing and education. Indeed, the scientific interest in early childhood has expanded massively during the last century: international journals on early childhood have been founded and study programmes for early childhood education, child psychology and infant health have been created. The evolving childhood sciences provide a basis for legitimizing interest in early childhood. In documents constituting early childhood as a global issue, not only are certain arguments repeated over and over again, but also the legitimizing basis on which they rest. Next to reference to ‘scientific evidence’ in general, especially the High/Scope Perry Preschool study and Head Start are cited repeatedly in publications of international organizations, as well as the economist James Heckmann (e.g. OECD, 2006: 36, 250; UNESCO, 2006: 113; UNICEF, 2001: 51; World Bank, 2002: 6; World Bank, 2007: 67–79). The increasing reference to ‘science’ and scientific ‘findings’, ‘evidence’, ‘outcomes’, ‘effects’, ‘studies’, ‘research’, ‘experts’ and ‘data’ provides a basis for demanding investments in the early years. The rationalized legitimacy of scientific evidence is one powerful cultural resource in the institutionalization of early childhood as a global issue.

For a discourse and world society theoretical analysis of one global social policy issue – early childhood – I propose to analyse the rules of formation of statements with regard to their modes and positions of enunciations, the formation of objects and concepts and the formation of strategies. These three strands are of course linked, and thus discourse analytical world society research also pays attention to their relations, regularities and irregularities: How are positions of enunciation and modes of statements linked with statements and strategies? Which patterns of repetition and concentration, of dispersion and complexity can be reconstructed, and which disruptions and oppositions?

Such a research approach reconstructs the conditions of existence, the conditions for the emergence of an object and of an issue in the way it appears: Why does it appear in this way, at this time and place? Thus, the chronological ordering of events on the level of statements as well as the relationship to broader societal discourses and their changes is important. The current interest in early childhood, for example, links to Enlightenment and Romanticist ideas on individuality, education, development and progress that provide historical legacies for the current conceptions of early childhood as a specific and individually, as well as collectively important period in human life. Also, changes in the constitution of women from mothers of the nation to female labour force and to equal citizens and bearers of rights (Berkovitch, 1999) are relevant for the constitution of early childhood, and, in more recent years, the change from a neoliberal paradigm in global social policy to a social investment perspective (Jenson, 2010a, 2010b) which positions children at the centre for investing in human capital.

From a world society perspective, world polity transformations such as the emergence of the UN system, other international associations and nation-state formation are central for the constitution of issues as global. Longitudinal studies can provide important information on which of those changes can be related to the emergence of early childhood as a global social policy issue and identify causal influences on its constitution as a national policy issue (Sugawara, 2008). Moreover, however, it is important to take the constructivist stance seriously and not to ontologize the ‘globality’ of issues. Although world society theory assumes a distinct and emergent global level, this ‘level’ is understood as an analytical level. This perspective opens the question of how EC is positioned as global – explicitly as well as due to different horizons of reference: Who is the audience to which certain statements are directed (Heintz and Werron, 2011)? Is the world horizon that is constituted by the semantics of ‘early childhood’ one of the ‘world as a whole’ (Robertson, 2001)? Are regions or are the Global North and the Global South addressed differently? To what degree is the concept of ‘early childhood’ universalized? Such a perspective can point out how global horizons of communication, of observation and of comparison institutionalize for specific issues.

Whereas world society theory emphasizes the broader cultural processes involved in the global institutionalization of issues such as universalization and theorization, the discourse analytical perspective allows for analysing the logics and discursive practices involved. Analysing such processes means to shed light on the constitutive conditions for early childhood to institutionalize as a global issue, and thus the formation and transformation of a recent global social policy field.

Consequences. Or: And what does all this mean for global social policy research?

The main argument of this article was that a dialogue of world society theory and discourse analysis can enrich the theoretical basis for global social policy research. With reference to world society theory, the meaning of the term ‘global’ in global social policy has been discussed, and a discourse analytical world society perspective has been proposed for analysing the formation and transformation of global social policy, of social policy models that are positioned as global, imbued with universalizations and embedded in world polity structures. Choosing early childhood as an example, the article reflected on how to analyse the institutionalization of global social policy issues and their constitution as ‘global’ social concerns. The three analytical strands can serve as a starting point for discussing new analytical approaches to global social policy formation and transformation. The initial suggestions here present a broad framework that has to be used innovatively; still, this generalized perspective allows for relating different studies with each other and with a specifically ‘theorized’ version of global social policy research. Debate is not only needed on how to further specify these initial suggestions but also on how they could be transferred to questions regarding the construction of other social groups (women, workers, ethnic minorities) and the individual. The question of the emergence of global social policy needs to be linked to the construction of the human being as social actor, ontologized with certain rights and needs, and the construction of the state (and other organizations) responsible for meeting those needs. Those actor constructions are increasingly universalized; along with the development of human rights, world responsibilities are constructed as problems of the world of a whole to be met by the world as a whole, even transcending the ‘international’ community by including actors outside and transcending national entities. Monitoring, reporting, statistics, evaluations, programming and world conferences are mechanisms and instruments by which these constructions are stabilized and transformed, spanning a global horizon of communication, observation and comparison in which global social policy is constituted, enabled and restricted at the same time.

For global social policy studies, the discourse theoretical reading of world society theory means a specific approach that differs in many regards from the ‘traditional’ conceptualization of the field and object of study: an approach that, first, is focused on cultural explanations, not on actor orientation, and thus able to take the construction of actors itself into consideration; that, second, is based in constructivist, not in realist epistemology; and that, third, is analytical in the first stance, not normative (without denying or propagating normativity, as an individual choice of each researcher). Perhaps more debate is needed if global social policy as a field of research wants to move in this direction.