Inclusion/exclusion: Educational closure and social differentiation in world society

Abstract

Niklas Luhmann’s late concept of inclusion/exclusion in world society compensates for the lack of a systems theory of social differentiation, but ultimately remains incomplete. Moreover, it is remarkable that Luhmann apparently never linked the inclusion/exclusion concept to his late theory of the education system. The following considerations seek this missing link, assuming that both functional and social differentiation constitute a double-sided form of societal structuring generated by the polycontextural operation of social closure mechanisms. The term ‘educational closure’ refers to the autopoietic form of inclusion/exclusion processed by an education system internally coupling operational and social closure. The article argues that organised education worldwide operates as an observing system autopoietically generating categorical differences, as well as inequalities between persons. While grouping, sorting, and tracking pupils, the education system seeks to maintain the recursion of educational communication, but also generates unequal social differentiation as a far-reaching side effect. Finally, reconstructing Weber’s theory of social closure as an earlier inclusion/exclusion concept may provide a theoretical basis for further empirical research into the social consequences of educational inclusion/exclusion in a world society.

Starting point: The paradox of education

Understanding Luhmann is a challenge even for sociologists socialised in the traditional discourses of their own discipline. Especially the replacement of ‘familiar’ sociological terms like action, intention, norms, domination, inequality, group, or individual by imported and thus unfamiliar concepts such as system, autopoiesis, observer, or even communication complicates the reception of Luhmann’s theory of society. Education scientists in particular have to cope with Luhmann’s methodological ‘anti-humanism’ (Luhmann, 2013: 12). Considered as psychic systems, human individuals develop in operational closure against their societal environment, inaccessible to communication – and that includes ‘educational communication’ (Kade, 2004). Education, in other words, is impossible; paradoxically, however, its function is to transform every human individual into a person capable of meeting the communicative demands of the functionally differentiated society.

Luhmann’s (2002) systems theory of education focuses on how educational communication constitutes and sustains itself under these conditions, that is to say, how education enables its own autopoietical continuation. Defined as an operationally closed system, only educational communication communicates, not ‘human beings’. Thus, an autopoietic educational system can only emerge if its communicative operations are decoupled from the cognitive operations of the psychic systems, which is tantamount to creating a constitutive difference between environment (psychic systems) and system (education). If the occurrence of intended learning effects were a prerequisite for the continuation of systems operations, no education system would exist. Due to the intention to control ‘not intentionalizable’ cognitive operations, the education system thus paradoxically constitutes itself on the basis of this ‘structural deficit’ (Luhmann, 2004a: 96–97). Unlike functional systems such as economy or politics, educational communication apparently misses a success medium (e.g. money or power) and a corresponding binary code but compensates for this by organising educational communication as a specific interaction form (Vanderstraeten, 2002).

Since there is no educational technology available to make learning outcomes more certain, the education system invents ‘replacement technologies’ (Luhmann and Schorr, 1982). Grouping, tracking, and selecting individuals turns the paradox of education into a durable problematisation of the cognitive and behavioural state of its students. Nevertheless, education needs to be comprehended as a global phenomenon, because every single learning interaction constituted through an organised education system has to cope with the same paradox: trying to structurally change operationally closed psychic systems through operationally closed communication.

The following considerations intend to explore the extent to which inclusion/exclusion can be theorised as a replacement technology used by the educational system to enable operational closure. However, the construction of inclusion/exclusion as a specific educational operation that differs, for example, from economic and political inclusion/exclusion requires theoretical modifications to Luhmann’s approach to a systems theory of society. This calls for preliminary remarks on the relevant theoretical basics, in particular those our argument modifies. Given the circular nature of Luhmann’s theorising, the following sections set out from two points: first, the distinction between societal and social forms of differentiation and, second, the concept of observation as an operation processed by communication systems. The first addresses a blind spot in Luhmann’s theory of society while the second could prove a useful concept in tackling this issue.

Sociology without groups: Societal and social differentiation

According to Luhmann, society is neither observable as a given object nor tangible as a distinct entity. No epistemological path leads beyond society since epistemology itself is a historical and thus societal invention. Contrary to any ontological notion, society exists only through and during the elementary operation of communication in the medium of sense. ¹ Following Luhmann’s ‘operational constructivism’ (Luhmann, 2009: 14), society emerges as an interpretation-based reality and social systems operate as interpretative or sensemaking systems. A world based on interpretation constantly needs to be reinterpreted since future is an inexhaustible source of uncertainty, providing an endemically contingent world of (un-)predictable possibilities. Societal systems seek to control future by drawing boundaries of contingency: not everything is possible and the impossible on the one side enhances the likelihood of successful communication on the other. While communication systems continuously seek to reduce world contingency through generating system-specific boundaries of sense (‘Sinngrenzen’, Luhmann, 1994: 95), they continuously regenerate world contingency.

Societal systems therefore need to be understood as self-referentially constituted interpretation contexts framing the modes of making sense of the world. Functional systems such as the economy and education facilitate world interpretation as such by separating economic communication from educational communication, thus dynamically stabilising their continuation by reducing the complexity of subsequent communications. This ‘outdifferentiation’ (Luhmann, 2013: 2) of functional systems historically invented by modern society not only changes the basic form of generating societal order but extremely increases societal complexity through constructing an ever-changing environment that constantly requires reinterpretation. Functional differentiation therefore implies a plurality of sensemaking systems as contexts that make society a polycontextural (Luhmann, 2013: 46) reality. Thus, societal events always emerge synchronically in a multitude of interpretative contexts: becoming a pupil is always an educational, legal, political, economic, scientific, etc., reality. Contrary to the tradition of sociological thinking, systems theory serves no anthropology. According to Luhmann, neither do human individuals inhabit societal systems nor are they the ground for systems differentiation. The autopoiesis concept simply refers to the fact that one cannot think anybody else’s thoughts: Ego’s consciousness and Alter’s consciousness operate self-referentially; they neither share nor exchange any elements. Thus, without communication, only self-referential behaviour would be possible but, not social operation. But human individuals need to be taken into account as psychic systems structurally coupled with the societal system, since communication presupposes cognition and articulation capacity on the part of individuals – and nothing more.

The idea of society being constituted as a union of distinguishable social groups in terms of ascriptive categories such as nation, ethnicity, religion, class, gender, milieu, is part of the sociological tradition. But this tradition still leads to the empirically misleading culturalisation or ethnicisation of society. Such categories might have been adequate for describing societal segmentation or stratification. However, they lose their analytical capacity when modern society abandons estatist and group membership for functional differentiation as the primary principle for generating order. The notion of society being structured in terms of ascriptive categories methodologically reproduces a ‘groupism’ (Brubaker, 2004: 11) inverting explanans and explanandum: it is the societal process of ‘group-making’ (Brubaker, 2004: 13) that needs to be explained. In line with this epistemological doubt, Luhmann suggests that modern, functionally differentiated society operates with indifference towards ascriptive categories; categorically grouping human beings no longer has a ‘social function’ (Luhmann, 2013: 107). Even opportunities for interaction lose their segmented and stratified exclusiveness – the modern school system, in particular, organises interaction across social strata and formally attains social mobility. The operation of interaction systems is sensitive to the individualised person, but these systems emerge independently of stratification or caste membership.

In other words, Luhmann discovered the structural de-coupling of societal and social differentiation as the fundamental and constitutive invention of modernity; but he ignores the fact that, in this case, indifference is not equivalent to irrelevance. Hence, the epistemological advantage of the systems theory of society is its conception as a sociology without groups. Therefore, in what follows we assume that the relation between functional differentiation and social differentiation needs to be considered as a co-evolutionary (or ‘double-sided’) form of societal structuring. Luhmann decided to focus on societal (functional) differentiation, leaving the other side, social differentiation driven by ascriptive categories, as an ‘unmarked space’. To compensate for this theoretical blind spot, we shall use the concept of observation, arguing that societal systems address and differentiate persons in order to stabilise their autopoietical continuation and thus perform inclusion and exclusion as social differentiation. The basic question this discussion raises is whether the social closure concept is compatible with a theory of functional differentiation (as Brubaker suggests) and, vice versa, whether the systems theory of society is amenable to a closure concept.

Observing systems

Societal systems are capable of observing their environment and themselves; they function as observing systems. The concept of observation is one of the hardest elements in Luhmann’s epistemological approach to accept, since he shifts the capability to observe ‘world’ from subject to society (Emmerich and Huber, 2013). Consistent with theorising communication as the elementary operation constituting society itself (communication is society), he redefines self-reflection as a specific mode of distinguishing and relating self-reference and other-reference exclusively processed by social systems. In contrast to the individual, societal systems are, of course, incapable of perceiving; they use communication self-referentially as medium and object of observation instead. Since the consciousness of psychic systems (human individuals) operates autopoietically, communication provides the only possibility for generating a shared world. This is also the case for social systems: while the elementary operation of a system emerges only as autopoietic communication, observation generates a difference between self- and other-reference by indicating the ‘world outside’ for the system. Hence, observing systems have to be considered as performing systems creating an external world through internally processed observation and under self-determined conditions. And this includes observed individuals becoming part of the system’s external world as persons.

Observation is conceptualised as a double operation processing distinction and indication at the same time: while distinction evokes a binary difference, indication marks one side of this difference in order to refer to ‘something’ that is not the observing system but environment. Therefore, the system needs to redundantly reproduce the difference of self-reference and other-reference as an internal prerequisite for observation (Luhmann, 2013: 49). Replacing cognition by communication, Luhmann continues the second order cybernetics approach (von Foerster, 1979) distinguishing between first- and second-order observation. First-order observations ‘only observe what they observe and do not observe how they observe’ and thus cannot observe the ‘what’ as depending on the ‘how’. While observing, the system cannot treat itself as an observer: ‘Phenomenology is practiced as ontology’ (Luhmann, 2013: 50). Whilst first-order observation constitutes ‘something’ for the observing system, second-order observation observes an observing system asking about the distinctions this observed system uses while indicating ‘something’. In the case of self-observation, the observing system observes how it observes and thus needs to self-referentially treat itself as both an observing system and an observed system. ²

What, finally, is this theorisation of ‘observing systems’ empirically good for? First, assuming that societal systems are capable of self-referentially constituting a relevant environment also implies they are capable of observing human individuals as persons. Second, if societal systems observe persons by indicating one side of a distinction (e.g. customers and non-customers, pupils and non-pupils), they differentiate individuals on the basis of autological group-categories that primarily make sense for the observing system. Third, despite Luhmann’s systems theory of society obviously overlooking the problem of social differentiation, these two implications may provide starting points for theoretically compensating for this blind spot. The empirical question is now how societal systems observe persons – and to what end. The following section argues that Max Weber’s concept of social closure helps identify the generative mechanism ‘doing difference’ within societal systems.

Social closure revisited: The making of social differences

Available concepts of societal structuring concentrate mainly on explaining the re-production of vertical forms of social differentiation, especially economic, political, and educational inequalities. Generally, these mechanisms have been conceptualised as matching forms of either domination or hierarchy (e.g. capitalism, patriarchy, and colonialism). In contrast, ascriptive categories indicating social difference (e.g. race, ethnicity, religion, gender) are not ‘intrinsically linked to inequality [. . .]. The relation between difference and inequality is contingent, not necessary; it is empirical, not conceptual’ (Brubaker, 2015: 11). However, domination and exploitation may causally explain the reproduction of class relations, and unequal allocation may explain the stabilisation of social strata; but neither concepts of exploitation nor theories of domination consistently explain the societal (historical) emergence of ethnic, cultural, religious, etc., differences. Unable to theoretically construct a societal mechanism generating horizontal forms of social difference, these concepts have inevitably to assume that social groups emerge pre-societally.

In the light of eroding estate-based German society after World War I, Max Weber (2005) sought to develop a sociology grounded in the difference between open (society) and closed (community) social relations, that is, forms of societal structuring. The concept of social closure thus theorises the mechanism that transforms open relations into closed forms of societal order. ‘Openness’ describes a principle of societal (associative) cohesion including all individuals regardless of their estate or class membership. Closeness, in contrast, derives from exclusive social (communal) cohesion preferring individuals as members of families, estates, classes, ethnicities, and/or social groups in general. Weber’s thoughts on open and closed forms of social cohesion provide the basis for subsequent systems theories of functional differentiation. But systems theorists such as Parsons and the ‘early’ Luhmann did not share Weber’s sensitivity to the mechanism of ‘social closure’; they also overlooked the relevance of the mechanism for a systems theory of social differentiation. However, Luhmann’s late theory of inclusion/exclusion implicitly reacts to this deficit.

Hence, reintegrating social closure promises to partially complete the systems theory of society. However, it requires a close re-reading of Weber’s concept of social closure and the reconstruction of Luhmann’s concept of inclusion/exclusion. Finally, constructing a consistent theoretical connection between social closure and operational closure is the challenge to be met.

One frequent economic determinant is the competition for a livelihood – offices, clients and other remunerative opportunities. When the number of competitors increases in relation to the profit span, the participants become interested in curbing competition. Usually one group of competitors takes some externally identifiable characteristic of another group of (actual or potential) competitors – race, language, religion, local or social origin, descent, residence, etc. – as a pretext for attempting their exclusion. It does not matter which characteristic is chosen in the individual case: whatever suggests itself most easily is seized upon (Weber, 1978: 341–342).

Comparing the German original ³ with the English version (Weber, 1978), mistranslations are apparent that could lead to theoretical misconceptions: ⁴

• – First, Weber speaks of ‘one’ and ‘another’ part of competitors instead of ‘one’ and ‘another’ group of competitors. Weber focuses on individuals involved in economic competition and does not refer to pre-existing racial, religious, local, etc., groups as a prerequisite for social closure.

• – Second, the German phrase ‘äußerlich feststellbares Merkmal’ (translated as ‘externally identifiable characteristic’) means ‘visible attribute’ and emphasises the ascriptive ⁵ character of indicating social categories to individuals.

• – Third and most misleading: Weber identifies the exclusion of individuals from competition (‘Ausschluss vom Mitbewerb’) as a major effect of social closure, but the translation omits ‘from competition’, which suggests that social closure evokes the total exclusion of one group.

Unfortunately, the English translation suggests that social closure takes place between pre-existing social groups, which runs contrary to Weber’s intention. At least, the concept may oscillate between ‘closing the market to competitors and closing the group to outsiders’ (Cardona, 2013: 1). Nevertheless, the closure concept is obviously used to explain how economically irrelevant imaginary categories (‘visible attributes’) turn into relevant operators for societal structuring.

As far as this interpretation is concerned, social closure should be conceptualised as a mechanism of differentiation constituting both economically included and excluded persons and/or groups. Moreover, Weber emphasises the arbitrary character of categories chosen to generate exclusion: since any distinction – race, language, religion, etc. – will do for exclusion, no prior linearity or causality between ascribed attributes (difference) and societal structuring (inequality) can be assumed. In other words, the generation of social differentiation by social closure is co-evolutionary but a-causal in relation to societal differentiation. Openness thus needs to be understood as a prerequisite for closeness and vice versa: otherwise, no such mechanism as social closure would, theoretically, make sense. Finally, Weber’s concept of social closure makes a pitch for theorising social differentiation beyond methodological groupism, since it seems to be connectable to a theory of observing systems: insofar that observation performs a difference for the system, observed individuals become differentiated by the system.

Neglecting this constructivist ⁶ line in Weber’s conception of social closure, subsequent approaches to social closure theory by Parkin (1979) or Murphy (1988) tend to espouse the idea of pre-constituted groups intentionally using closing strategies in order to succeed in societal conflicts. Focusing on the reproduction of societal inequalities through education, Bourdieu and Passeron (1977) suggest that school systems serve as institutionalised arenas in which social milieus fight a class struggle by cultural means. Yet, all these contributions integrate an important aspect into closure theory: the concrete role public education plays within the process of social structuring in modern society. But despite this insight, they refrain from paying attention to the autological operations of organised education.

In contrast, Tilly’s (1999) and Brubaker’s (2015) instructive contributions to social closure theory seek to analyse the crucial role of organisations (Tilly) and to evolve the social-constructivist implications of Weber’s initial concept (Brubaker). According to Tilly, organisations such as firms, governments, and schools generate durability in societal inequality by matching social differences (ascriptive categories) with hierarchically structured positions inside the organisation. Hence, organisations match interior categories (hierarchy) with exterior categories (race, ethnicity, religion, gender), and thus internally reinforce the environmental inequalities they utilise in order to internally facilitate exploitation (Tilly, 1999: 75–76). Tilly identifies exploitation, opportunity hoarding, emulation, and adaptation as the four ‘causal mechanisms’ (Tilly, 1999: 10) ensuring the operational continuation of organisational inequality-making. Tilly’s argument is instructive insofar as it precisely reformulates Weber’s insight that closure functions by turning operationally irrelevant ‘visible attributes’ into relevant operators for societal structuring. By matching interior and exterior categories, organisations generate the causal link between social difference and societal inequality. Furthermore, this ‘organizational view of inequality-producing mechanisms’ (Tilly, 1999: 9) is able to epistemologically avoid methodological groupism, since it explicitly identifies the ascriptive categories doing the crucial work of social closure by generating unequal groups of human individuals.

Critically discussing Tilly’s categorical conception of durable inequality, Brubaker (2015: 43–44) suggests distinguishing between ‘strong’ and ‘weak’ meanings of the term ‘categorical’ inequalities and contrasting ‘categorical and gradational forms of inequality’ with respect to a ‘strong’ concept. Particularly with respect to complex societal allocation mechanisms, Brubaker queries a causal linkage between organisational positions and categories of difference. As Brubaker (2015: 44) states, the modern principle of functional differentiation structurally decouples ascriptive categories and inequalities (e.g. with respect to the law); but in a global perspective, different forms of tightly coupled relations between categories of person (ethnicity, sex, caste) and unequal access to resources, public health, legal rights, or education co-exist.

Brubaker’s critique is instructive, since it sensitises the analysis of social closure to the complexity and dynamic of societal structuring. First and foremost, he – implicitly – outlines the theoretical necessity to distinguish between societal differentiation (inequality) and social differentiation (ascriptive categories) as well as – explicitly – insisting on the empirical importance of assuming an acausal relationship between the two forms of societal structuring. However, Brubaker overlooks that Weber developed a theoretical concept capable of describing an operation of social differentiation based on the ascription of social group categories to individuals (social closure). Tilly’s organisational approach may ‘overextend’ categorical exclusion with respect to the concept’s explanatory force, but gives an instructive answer to the remaining conceptual question: Who does the grouping? With regard to the problem of social differentiation, Tilly points to the role modern organisations play in structuring society; as far as the concept of observing systems is concerned, the answer to the question ‘who’ societally observes human individuals as grouped is: organisational systems.

Social closure theory offers two instructive countervailing insights: first, Weber’s approach opens the way towards a ‘non-groupistic’ and non-ontological theory of social differentiation in modern society; second, group-making as such needs to be understood as an internally produced resource stabilising the operational continuation of organisational decision-making. The following reflections on Luhmann’s inclusion/exclusion concept accordingly seek a connection between a constructivist revision of social closure theory and a theory of observing systems. Thus, theorising organisations as autopoietic systems (Luhmann, 2006) is a prerequisite for turning the inclusion/exclusion approach into an operational concept of social closure.

Inclusion/exclusion: Social closure and autopoiesis

Within the architecture of Luhmann’s theory of society, the inclusion/exclusion concept is to be found only in outline. Few theorists attempt to enhance the concept, although it promises to provide a more profound analysis of the ‘collateral damage’ functional differentiation causes with respect to diverging living conditions (Nassehi, 2004; Stichweh, 2009). Despite its ambivalences and inconsistencies, criticised by both opponents and proponents of Luhmann’s systems theory approach, discussion on the analytical value of the inclusion/exclusion concept has continued among sociologists and education scientists, in particular (Stichweh and Windolf, 2009; Farzin, 2006; Emmerich and Hormel, 2013). Emphasising inclusion as the evolutionary force of societal cohesion, the theory of functional differentiation disregards the ‘other side’ of modernity: nationalist and racist politics dividing society into members and non-members, legitimised and non-legitimised persons, and granting categorically differentiated access to rights, citizenship, economy, or education. ⁷

Luhmann (1995) recognised this blind spot of functionalism after visiting Brazil, where he ‘observed’ that the favelas were a mere habitat for individuals excluded from functional system performance. ‘Exclusion’ apparently reduces the person to an individual body, irrelevant for functional systems (Luhmann, 1995: 262). But the terms inclusion and exclusion do not describe two distinct socio-ontological states of ‘being’; they indicate the two sides of a specific modern form of differentiation relating individuals to societal systems:

The predominant relation is no longer hierarchical, but one of inclusion and exclusion [orig.]; and this relates not to stratification but to functional differentiation. [. . .] Functional systems presuppose the inclusion of every human being, but, in fact, they exclude persons that do not meet their requirements (Luhmann, 2015: 42–43).

On closer inspection, conceptualising inclusion/exclusion proves difficult to adapt to the architecture of systems theory, particularly with respect to the theory’s concept of differentiation.

Inclusion is to be understood as a form whose inner side (inclusion) is marked as the opportunity for the social consideration of persons and whose outer side remains unmarked. There is therefore inclusion only if exclusion is possible (Luhmann, 2013: 18).

This formalistic definition of inclusion/exclusion is apparently guided by the idea of boundary-making and replaces the binary distinction system/environment. Luhmann insists that inclusion/exclusion operates with reference to the level of society and constitutes a ‘primary differentiation of society’ (Luhmann, 1995: 261) – although systems theory intends only functional differentiation for the primary form of societal structuring. This is where the system level of organisation comes into play.

Functional systems treat inclusion, that is to say, access for all, as normal. The opposite applies to organizations: they exclude everyone except highly selectively chosen members. As such, this difference is functionally important. For only with the aid of internally formed organizations can functional systems regulate their own openness and treat persons differently even though all have the same access. The difference in modes of system formation thus makes it possible to practice both inclusion and exclusion at the same time (Luhmann, 2013: 151–152).

Luhmann’s argument has a number of implications. First, inclusion is the mode in which functional systems consider ‘persons’, whereas organisations process exclusion by marking/unmarking persons, thus enabling functional systems to not consider an individual as a person. Second, processing exclusion on the organisational level is also ‘functionally important’ and – somehow – related to the autopoiesis of a functional system. Third, and consequently to the latter, Luhmann’s statement that functional systems ‘regulate their own openness and treat persons differently’ stringently implies that they also regulate their closeness to persons at the same time and by the same operation; organisation-based inclusion/exclusion selectively opens and closes access to societal systems. Finally, inclusion/exclusion needs to be analysed at the different levels of functional, organisational, and interactional systems.

One key, previously disregarded, aspect prevents this construction from theoretically scotching autopoiesis, since it provides a bridge to the theory of observing systems: while defining inclusion/exclusion as a form generating a marked and an unmarked side, Luhmann adapts the operational logic of distinction/indication (observation) to the concept of inclusion and exclusion; observation and in-/exclusion seem to operate homologously. It is crucial to mention the function observation performs regarding the continuation (dynamic stabilisation) of operational closure. In line with the autopoiesis paradigm, observation supports operational closure in a certain way:

This is the case even, indeed especially, where such operations are observations or operations whose autopoiesis requires self-observation [. . .]. Observing systems, too, have no contact with the environment at the operational level. All observation of the environment must be carried out in the system as an internal activity with the aid of the system’s own distinctions (for which there is no correspondence in the environment) (Luhmann, 2012: 49).

If societal systems are able to ‘consider persons’, the only plausible internal operation enabling them to do so is observation: human individuals become a relevant environment for a system while being selectively observed as persons and non-persons by the system. The remaining question is: To what end?

Gathering the loose ends of inclusion/exclusion and observation theory and combining them with a constructivist redefinition of social closure, a systems theory of social differentiation can be (re-)constructed. The inclusion/exclusion concept describes a discriminatory treatment of individuals – the consideration/non-consideration of persons in societal systems – apparently relevant for the operational closure of such systems; inclusion/exclusion operates as an observation structurally coupling social and operational closure. Whereas operational closure constitutes a social system (society, organisation, interaction), social closure constitutes persons as relevant environments for systems and through systems. Both forms of closure generate a difference, but only the former elicits systems formation; the latter generates social differences as a resource for the system’s dynamic self-reproduction.

Taking into account the polycontexturality (Luhmann, 2012: 46) of societal reality resulting from functional differentiation, inclusion/exclusion also needs to be seen as a polycontextural phenomenon: there is no single and encompassing logic of social closure (Emmerich and Hormel, 2013: 76–79). In this sense, economic, political, and of course educational inclusion/exclusion operate independently, without causal linkage. The economy excludes individuals where they have no money available, while individuals without citizenship status remain politically excluded (‘sans-papiers’); but political exclusion does not determine economic exclusion. In contrast, public education determines further societal inclusion/exclusion by function; unequal exit examinations generated by internal selection procedures predetermine unequal allocation to organisational positions and legitimise exclusion from career fields. Inside the education system the differentiation of persons is guided by the opportunities for inclusion/exclusion the given organisational and interactional structures provide: special needs education, for example, generates an organisationally constituted status of ‘inclusive exclusion’ (Stichweh, 2007).

Hence, Luhmann’s notions of inclusion/exclusion seem to reformulate Weber’s idea of social closure under the epistemological conditions of operational closure – except for one significant detail: Weber’s closure concept would suggest that systems ‘regulate their own openness by (!) treating persons differently’, implying that societal systems process social differentiation through ‘group-making’. The ‘excluded’ of the favelas or any other comparable peripheral and marginalised urban or rural habitat may be marked as societally irrelevant for functional systems, as Luhmann assumed; but they become societally visibilised as the excluded from these systems. Locatable and socially marked (or named), these persons were made a group without any internal cohesive principle; socially ‘outdifferentiated’, they emerge as an in-/visibilised ascriptive category.

Educational closure and organised education

Organised education processes a legitimised form of social closure. Flanking credentialism, meritocracy ideologically enables an individualistic rule of school-based closure (Parkin, 1979) since industrialisation demands manpower, knowledge, and competence instead of birth right or caste membership. With respect to Luhmann’s analysis of the education system, what follows concentrates on how schools function as autopoietic systems and how social closure relates to operational closure. The term educational closure precisely articulates the system-specific autologic of processing this relationship.

The education system’s ability to observe persons is based on the ability to selectively ascribe categories to individuals; but the distinctions that enable the system to indicate individuals make sense only for the indicating system. Sorting customers into good and bad persons would not help a grocery store boost sales; but sorting pupils into high and low performers or fast and slow learners helps teachers construct classroom order and professional certainties. Thus, performing the communicational consideration/non-consideration of persons and granting access/no access to the system’s communication finally depends on the operational autologic and on the structural variables of each system. Particularly with respect to people-processing organisations such as schools, the autopoietical relevance of individuals seems intuitively evident. School systems therefore need to be considered as a form of organised education not only capable of observing persons as environment; in fact, they depend operationally on the ability to differentiate individuals by interior categories (the system’s own distinctions) – and for their own purposes. In the following, we therefore speak of educational closure, since organised education (school systems) is the societal operator of this highly specialised form of education-based social closure (Emmerich, 2016b).

Taking the recursiveness of elementary operations and observations into account, organising instruction-controlled learning generates an operational paradox, since educational communication cannot reach – let alone change – the operations of its psychic environment (Luhmann, 2004b: 117). From a functional point of view, the reference problem of public education everywhere is the same: the unlikely intentionalisation of unintentionalisable socialisation (Luhmann, 2004a: 94). But since the education system lacks a symbolically generalised success medium such as money or power as well as a binary code, educational communication operationally regenerates an ‘endemic uncertainty’ regarding its intended effects. Hence, educational autopoiesis apparently requires organisation and selection compensating for this structural deficit. ‘Whereas other functional systems outdifferentiate on the basis of binary codes [. . .] the outdifferentiation of the education system itself depends on organisation’ (Luhmann, 2004c: 185; own translation).

Finally, Luhmann assumed that the education system features two different media (the primary medium child and the secondary medium lifecourse/career) as well as two different binary codes (the primary code impartable/non-impartable and the secondary code better/worse) (Luhmann, 2002). With regard to these apparent theoretical inconsistencies, Vanderstraeten (2004) convincingly proposed to theorise the education as a system constituted through ‘organized interaction’. Hence, taking this operational doubling – interaction and organisation – into account, also the doubling of medium and code becomes plausible: ‘impartable/non-impartable’ codes educational communications on the interaction level while ‘better/worse’ codes communications on the organisational level. But assuming a primary and secondary relation of media and codes basically implies a constitutive operational relation between them; taking this insight into account, only the internal coupling of both media and both codes would enable the recursion of educational communication, that is to say, the emergence of education as a societal system (Emmerich and Hormel, 2013).

As far as educational communication as a societal system is concerned, operational closure requires persons capable of observing the interaction system as a relevant environment, as well; otherwise, no interactive communication would occur. Hence, the education system expects human individuals to have a linguistic and cognitive command of the common school language as a prerequisite for the system’s own autopoiesis; otherwise, individuals would interrupt operational closure – and they do (Fuchs, 2002). Apparently, this paradoxical situation double-binds the system’s operations and constantly provides for the structural disablement of educational communication. At this stage, educational closure comes into play in ‘dissolving’ the structural paradox: observing individuals as disabled enables the education system to exclude their contributions to classroom interaction (educational communication). Closing educational communication to ‘persons that do not meet its requirements’ needs to be considered as a genuine form of social differentiation processed by selective observation guided by ascriptive categories.

Hence, the education system achieves operational closure through social closure: in reproducing the capacity to differentiate considerable and non-considerable individuals it reduces the complexity of educational communication while insisting on this capacity as an operational requirement. The inclusion/exclusion mechanism prevents the ongoing system from collapsing. ‘Learning disabilities’ or other special educational needs classifications apparently function as interior categories devised within the education system, whereas diversity indicators such as race, class, gender, ethnicity, sexual preferences, etc., do their crucial work as exterior categories. Thus, differentiating persons in terms of externally ascribed categories continues the recursion of educational communication in cases when no ‘interior’ educational category can guide observation. For the education system, the ascription of ‘learning disabilities’ (or any other diversity category) is the solution, not the problem.

The organisation-based management of time, space, and persons enables a school-internal micro-economy that categorises, groups, sorts, and tracks pupils, as well as continuous decision-making to meet internal selection demands. But legitimisable decision premises – high or low educational achievements – need to be produced at the level of classroom interaction: operationally decoupled from the systems organisation level, the ascription of individualised achievement remains an abstruse professional task. Attaining operational closure at the interaction level requires cognitive and social operations to be decoupled: if educational communication were to depend operationally on successful learning, neither autopoiesis nor dynamic stability would be possible. In the case of education, system formation emerges only because education operationally neglects its psychic environment but refers to individuals as persons in the mode of observation. What, therefore, needs to be differentially observed is the participation of pupils in educational communication (interaction or reading/writing) in order, first, to find clues for further educational support activities and, second, to establish the premises for making decisions on grouping and selection. Thus, educational inequalities result from inclusion, not from exclusion; without being considered as a pupil, no unequal certificates could be allocated providing unequal access to subsequent systems such as working organisations.

The global educandus and the national educator: Educational closure in world society

Functional systems generate boundaries while reproducing the difference between system and environment, but they do not produce borders (Luhmann, 1982). Symbolic generalised media such as money, power, and law reduce world complexity because they operate only within their boundaries of sense. Borders have to do with the state, that is, the territorialised organisation of the political system. Ideas like transnationalism, internationalism, or even globalisation only transcend the political and economic function of borders. Luhmann’s concept of world society thus implies a critique of the common methodological nationalism in sociological thinking incapable of appropriately theorising the dynamics of modernity. But taking inclusion/exclusion as a ‘primary differentiation of society’ into account, the inclusive order of a functionally differentiated world society turns into a highly exclusive spatial order at the level of organisation.

Transnational education policies such as the OECD Programme for International Student Assessment (PISA) aiming to improve learning outcomes in a global perspective primarily observe the education inequalities produced by national systems on the basis of classical statistical social indicators (age, sex, class, race/ethnicity) construed as ascriptive categories. This methodological groupism guides globalised survey research, suggesting educational inequality grounded in social difference. From an educational closure perspective, these survey data need to be reinterpreted as indicators for grouping effects grounded in the system’s internal ‘grammar’ of inclusion/exclusion (Emmerich, 2016a). From a world society point of view, a global educandus represents the paradox of education affecting every single educational communication, whereas school systems as national educators organise and thus vary forms of educational inclusion/exclusion.

Organisations such as the welfare state, firms, and particularly school systems generate social differentiation through the broad spatialisation of communication, linking modes of segmentation and stratification while seeking to control their environment as a resource. Creating spaces and placing persons to give them a locatable social address enables organisations to control the future: spatialisation and location provide permanence and certainty. The unequal dispersion of schools on the earth’s surface; the centre/periphery structure channelling access to low and high education; the segregation of schools and school districts: all these phenomena described by educational research result from the spatialising nature of organising world society. People processing organisations such as school systems depend strongly on the constant communicational availability of human individuals if they are to stabilise themselves dynamically: operational closure requires addressable persons and inclusion guaranties access to this resource. National school systems solve these resource problems through compulsory schooling. Once included as pupils, individuals participate in educational communication and thus become observable and distinguishable for the education system with respect to operational requirements. Individuals incompatible with cognitive and behavioural norms structurally disable the flow of educational communication, that is, the operational closure of the system. But these pupils only represent the general paradox of education: changing cognitive structures by communication remains an unlikely effect.

Paradoxically, the educational icon of the ‘dis-abled’ pupil only articulates the ‘technological deficit’ pertaining to every national school system – worldwide. But the strategies for coping with the problem of blocked recursion at the operational level may differ from one national school system to the next, particularly with respect to basic organisational variables. Each national system seems to define the boundary between inclusion and exclusion differently depending on a historical ‘memory function’: whereas special educational needs (SEN) education in the United States has a broad psychological tradition generating a system of highly differentiated individual support measures, the institutionalisation of SEN education in Germany has been driven by a specialised profession and its strategy to implement a separate SEN school system (Powell, 2010). Both national systems process educational closure but generate different forms of exclusion and inequality. With regard to transglobal mobility and forced migration, educational closure generates a variety of structural disparities affecting newly arrived migrant pupils. Within the faltering German secondary school system, which provides unequally valued exit certification, the ascribed category ‘migration background’ statistically indicates significantly lower opportunities (odd ratios) to succeed in higher education compared with non-migrant pupils (Autorengruppe Bildungsberichterstattung, 2016), whereas in Canada no such educational closure effects can be statistically observed (PISA, 2015).

Compulsory education entirely includes every human individual at a certain age, and this raises the question of how and to what extent inclusion/exclusion operates as the modus observandi of national education systems. Allocating pupils to different tracks or school types providing unequal education opportunities requires decision premises capable of legitimising the allocation. Ascriptive categories such as ‘special educational needs’ or ‘migration background’ help solve the operational problem of indicating students for differentiation and allocation. According to Tilly (1999), even schools observe individuals as internal and external environment, ascribing unequal categories to individuals and thus constituting unequal social groups in order to solve their endemic problems. Schools thus operate as ‘sorting machines’ (Domina et al., 2017), matching interior (educational) and exterior (ascriptive) categories to allow their internal differentiation processes (grouping and tracking individuals) to continue (Emmerich and Hormel, 2013).

Finally, these theoretical considerations point to an empirical desideratum. Comparing national school systems with regard to educational inequality, empirical educational research commonly addresses the question of whether one system produces more or less inequality than another among the same social groups (class, ethnicity, gender). But in the light of the above reflections, the questions changes: To what extent does educational closure generates the variance in inequalities to be observed between and within national education systems? What intersecting ascriptive categories do educational organisations use to generate social differentiation? The functionalist analysis of inclusion/exclusion we have been discussing can avoid the methodological ‘groupism’ pervading international assessment and common sociological survey design. The proposed educational closure perspective accordingly rejects the inversion of explanans and explanandum: why and how education systems perform inclusion/exclusion and thus generate social differentiation in a world society horizon still needs to be explained empirically.