Field theory,role theory and role conflict: Reappropriating insights from the past

Abstract

This paper puts Bourdieu’s influential field theory into dialogue with the classic sociological tradition of role theory. I argue that while role theory per se is certainly problematic, it nonetheless bore the virtue of conceptualising in clear terms something amply borne out by empirical research but underemphasised in Bourdieu’s account of the social world: that people are situated in more than one social structure, with differing salience to them, and that this can have profound effects on their experience, their wellbeing and their practice. Following Bourdieu’s own prompts, I thus appropriate and recast selected insights from role theory to elaborate field theory so that it might articulate how experience and practice are structured by multiple fields, introducing the notions of field-set, meta-habitus, lifeworld horizon, illusio space and greedy fields.

This paper endeavours to put Pierre Bourdieu’s influential field theory into dialogue with the old sociological tradition of role theory. The immediate motivation for this is simple: role theory, I will argue, conceptualised in clear terms something abundantly apparent in swathes of empirical research but underemphasised in Bourdieu’s model of the social world. This is the fact that people are usually simultaneously oriented towards multiple bundles of social relations, involving different expectations and satisfactions, often with profound consequences for their inner life and their activity in the world. Role theory is, for sure, otherwise problematic from a Bourdieusian point of view, but it nonetheless contains several distinctions and observations that need to be incorporated into the body of ideas developed by the Frenchman. The wider context, which I will return to in the conclusion, is that the theory and research programme initiated by Bourdieu need to be put into continual constructive conversation with other perspectives if they are to thrive, and that this conversation must occur not only with new and emerging perspectives across disciplines but with older and long-established perspectives, even those, like role theory, finding rather less favour nowadays.

I begin by giving a basic outline of field theory’s key concepts and how they appear to side-line the everyday experiences, struggles and issues – amply documented in social inquiry – that people face by virtue of being concurrently positioned within multiple social structures. Bourdieu did, in fact, recognise the possibility and even necessity of developing field theory to account for multiple positioning, and even coined a term for it, but seemed ambivalent about it and, ultimately, took it no further himself. Taking up the task, in Bourdieusian spirit, I turn to role theory for assistance. After a brief overview of its historical trajectory and core assumptions, I dwell on the nuanced ways in which it conceptualised, under the banner of ‘role conflict’, how people’s thoughts and practices are inexorably contextualised by orientation towards myriad ends that they value to different degrees. Although role theory per se is, for reasons I will make clear, flawed, these conceptualisations can nonetheless help guide elaboration of field theory to better render and research the effects of multiple field membership. I thus introduce several new concepts, namely, the ‘field-set’, the ‘meta-habitus’, the ‘lifeworld horizon’, the ‘illusio space’ and ‘greedy fields’.

Capital, field, habitus

The scope of Bourdieu’s influence on contemporary sociology is immense. Studies of class, culture, education, race/ethnicity and more routinely deploy his concepts and precepts and find his specific theses useful for making sense of present-day phenomena (see Atkinson, 2020). This is partly because field theory was developed through and for Bourdieu’s own empirical research, partly because it was expressly designed to overcome limitations of existing perspectives and partly because it rests on a set of tightly integrated yet flexible concepts. The starting point is Bourdieu’s relational conception of social structures as so many spaces of difference, domination and struggle over varying forms of capital. The latter act as resources and powers, and sources of authority and legitimacy (or ‘symbolic power’), but ultimately offer major forms of recognition in a social order, that is, forms of value, worth and purpose (Bourdieu, 1990a, 2000a). Given the arbitrary and transmissible nature of capitals, however, and that they tend to be perceived as something other than they are – that is, as natural and/or inherently valuable or superior – they are, argued Bourdieu, more accurately described as principles of misrecognition.

The major space of difference in a social order is the ‘social space’, which serves as Bourdieu’s (1984) conception of the class structure, and in contemporary capitalist societies this is defined by the distribution of economic capital (money, wealth), cultural capital (mastery of symbolic systems) and social capital (connections and memberships) and manifests in different (judgements of) cultural tastes and lifestyles (Bourdieu, 1984). There are also, however, numerous ‘fields’ with varying levels and forms of autonomy from the social space – fields of cultural production (art, literature, science, etc), religion, politics, business (the economic field) and state activity (the bureaucratic field), for example (see e.g. Bourdieu, 1993, 1996a, 2005). They have relative autonomy insofar as class positions and capitals are commonly translated into the specific logics of the fields yet played off the existence of other forms of capital exclusive to the field – artistic capital, scientific capital and so on.

Each field has its own dominant and dominated players and its own internal struggles and movements as people endeavour, with varying success, to acquire or maintain capital, which is always, in practical terms, an effort to find meaning and purpose through recognition from others. Members are united by a set of common assumptions, beliefs and knowledges related to the field (doxa), but aspects of this can and do, as part of specific struggles, become openly contested, defended and subverted (Bourdieu, 1977). Members of a field also share a belief that the game is worth playing (illusio), or that the capitals and forms of meaning associated with a field are valuable and worth striving for. Immersion within a field, but also inhabiting a specific position within the field, generates a ‘feel for the game’ – a sense of what is likely or possible within the field, including what is likely, possible and thus desirable regarding one’s own place and practice within the field (see esp. Bourdieu, 1977, 1990b, 2000a). Phenomenologically, the feel for the game manifests in the contents of perception being fringed with – to use Husserl’s term – a ‘horizon’ of meanings and possibilities, and the resultant dispositions or schemes of perception and evaluation, which Bourdieu called habitus, are the generative base for specific strategies to accumulate or conserve capital but also routine presentations of self and judgements and classifications of others. ¹ Of course, the feel for the game need not always be acute – people can misread situations, especially in times of rapid change (Bourdieu, 2000a) – nor is habitus always well integrated. An unusual trajectory – upwards mobility in the social space, for example – can generate dispositions forged in quite different contexts and at odds with one another, and thus a ‘cleft habitus’ (Bourdieu, 2004).

Bourdieu later expanded his concept of field beyond the class structure and what might be thought of as the ‘professions’ (art, religion, politics, law etc), or what he called the ‘field of power’, to encompass specific organisations too, not only in the sense that, for instance, a firm or a higher education institution may be positioned within a field, but also to articulate that an organisation can itself operate as a field, with its own internal forms of capital/authority, doxa, illusio and so on (Bourdieu, 2005) – an idea that some have since considered a highly fruitful means of synthesising and surpassing numerous strands of organisational sociology (esp. Emirbayer and Johnson, 2008). Similarly, Bourdieu (1996b, 1998) also broached the possibility that familial or intimate relations can operate as a field, structured by the distribution of economic, cultural, physical and affective forms of capital, which again bears potential to integrate varied family sociologies and branches of psychology while proving fertile in empirical inquiry (Atkinson, 2016). This is not to say that intimate or organisational relations always operate as fields; like any other field they may instead, under certain conditions, take the form of a homogenous ‘corps’ or rigid ‘apparatus’, though these relational bundles still, I would argue, implicate illusio and habitus (Atkinson, 2020).

An unfulfilled promise

Bourdieu’s own research nearly always focussed on the structure, transformations and effects of a specific field and its relation to the evolving field of power and social space. Even when he trained his sights on the trajectory and practices of specific individuals of consequence for a field, such as Heidegger (Bourdieu, 1991a) or Manet (Bourdieu, 2017), his emphasis was on the manner in which their class capitals were converted into philosophical or artistic capital and class dispositions ported into the field – philosophy or art – and played off the space of possibilities there. Everything about Manet’s life, from his family relations to his friendship networks, was either translated directly into the logic of the artistic field – as the source of capitals or class dispositions – or stripped away as irrelevant. This was bound to Bourdieu’s methodological precept that social science should proceed by putting aside ‘concrete’ or ‘empirical individuals’ and focus on ‘constructed’ or ‘epistemic individuals’, or simply ‘agents’, that is, individuals as defined only by their capitals and habitus in relation to a specific field (Bourdieu, 1988b: 21ff; cf. Bourdieu, 2000b: 303n8).

However, Bourdieu’s later broadening and multiplying of the applications of field-thinking to cover structures of varying scale, including organisations and families, ineluctably raises a major corollary at odds with the logic of epistemic individuals: people – and not just those in the field of power – are situated in more than one field at a time, with their different possibilities and doxa and their competing sources of (mis)recognition, and this is likely to have a significant bearing on what they do in any one field. Several commentators have flagged this tension (e.g. Atkinson, 2016; Burawoy, 2019; Decoteau, 2016; Hadas, 2021; Schmitz et al., 2017), the most tenacious and emphatic among them being Lahire (2011, 2015). Bourdieu’s model of practice is generally hermetically sealed within a single field, he says, whereby positions engender dispositions which engender practices. ² Even the cleft habitus is seemingly defined by movement within a single field. Yet it might well be that what one can and wants to do in any one field is inexorably affected by one’s positions and practices in other fields.

Think, for example, of the lawyer torn between the possibilities and expectations within their firm (an organisational field) and the possibilities and desirability of specific stances and strategies within the legal field more broadly that run counter to them, or – closer to home – the academic likewise torn between commitment to position and practice within their university and the academic and/or disciplinary field. More ubiquitous, though, is all that is commonly put under the label of ‘work-life balance’ or ‘work-family articulation’ (see e.g. Crompton, 2006; Morgan, 2011), which evidently implicates the relationship between – and difficulty negotiating exigencies of – the family field and employment-related fields (organisational and/or professional) and its classed and gendered dimensions in the wake of feminisation of workforces after the Second World War (McNay, 1999). This is hardly a marginal sociological theme, minor element of social experience or small socio-political issue. It is, instead, the spur for extensive intellectual inquiry and policy wrangling across the world, a source of profound concern for many people and, ultimately, the background context in which the (un)desirability and (im)possibility of specific practices are formed. Bourdieu (2001: 107) too recognised the tensions and sacrifices made by women in juggling ‘success’ within paid work and the family, though he did not articulate it in terms of fields.

It would be wrong, however, to claim that Bourdieu never considered the theme of multiple field membership. In fact, although it usually only warranted a passing mention, he even coined a term to articulate it – he just never did anything meaningful with it afterwards. This is the notion of the ‘social surface’, which Bourdieu (2000b: 302) defined as a

rigorous description of the personality designated by the proper name, that is, the collections of positions simultaneously occupied at a given moment of time by a biological individual socially instituted, acting as support to a collection of attributes suitable for allowing him [sic] to intervene as an efficient agent in different fields.

The term thus encompasses occupancy of multiple positions but also a set of attributes facilitating cross-field agency. What those attributes are is a mystery, and the term ‘social surface’ is certainly open to interpretation given its fleeting mention. ³ Its primary function is evidently to denote capacities to operate in and affect more than one field, almost casting it in a positive light (because it implies greater agency/power), and there is nothing on possible strains or conflicts between these positions for a person. Moreover, Bourdieu himself then tracked back to his methodological distinction between epistemic and empirical individuals, admitting that focusing on the former ‘gives rise to a number of problems normally ignored, especially in statistical treatment’, specifically that it ‘allow[s] certain properties of the field to escape’ (Bourdieu, 2000b: 303n8), including the likelihood already highlighted that what someone does in one field is affected in part by their position, properties and expectations in other fields. There is ambivalence or ambiguity here, then: he presented the epistemic reduction as a necessary if regrettable element of rigorous social science yet conceded that ‘progress in this theory’ – theory oriented around fields – will ‘have to stem from the invention of categories and operations able to reconcile the theory with properties provisionally excluded’ (Bourdieu, 1988b: 23). The latter sentiment is correct, since it expands the range of sociologically explicable experiences, making it necessary to invent or develop those categories and operations, though in the process the very division between constructed and concrete individuals may need to be rethought.

Looking for solutions: Role theory and role conflict

Bourdieu’s own intellectual practice made clear that conceptual development is aided by critical dialogue with other perspectives both present and past within and beyond sociology. Hence the wide range of thinkers and theories alluded to, accommodated or rebuffed throughout his corpus. In this instance, I would suggest, Bourdieu’s toolkit can be expanded with the help of insights from – among other perspectives – an old body of thought seemingly passed its peak: role theory. This is an amorphous intellectual tradition with multiple progenitors and developers, among them Georg Simmel, Florian Znaniecki, Robert Park and Ralph Linton, though at its core, obviously, is the notion of ‘social role’. Probably the most well-known elaborations and uses of the latter term are to be found in the two dominant – and onetime seemingly opposed – schools of thought in mid-20th Century American sociology: functionalism and symbolic interactionism. Both operated with an understanding of roles as, at base, discrete sets of expectations of behaviour in specific situations. For Parsons (1951/1991) these expectations were understood with the vocabulary of ‘norms’ and ‘values’ – a sense of what is and ought to be done in a situation – which were said to be ingrained during early socialisation and to operate via internalised sanctions (Parsons and Bales, 1956). Parsons also stressed that roles were always relational – they bore on how people were expected to relate to one another in a specific situation – and for that reason came in oppositional couples, such as husband/wife, doctor/patient and so on. These oppositions were, moreover, defined according to their relationship with the so-called ‘pattern variables’, that is, the degree to which they embodied contrasting notions of universalism/particularism, achievement/ascription, neutrality/affectivity, specificity/diffuseness and self-/collective-orientation. The resulting meshwork of binary roles organising interaction across myriad contexts constituted the structure of the social system, though at a more substantive level they were organised into so many more or less bounded ‘institutions’ and ‘collectivities’ – the family, organisations, professional or ethnic groups, political parties and so on (Parsons, 1951/1991: 36ff). Their progressive differentiation from one another was, following Durkheim, held to be the hallmark of industrialised society, and the overarching assumption, of course, was that these roles contributed in some way to the functional integration of the social whole.

Like Linton, Parsons (1951/1991: 25) distinguished roles, as expectations of behaviour, from ‘status’, which was taken to denote an individual’s place or position in society. This might itself entail several roles across situations of interaction. Merton (1968) – who famously coined the phrase ‘role model’ – offered a characteristically precise vocabulary here. A ‘role set’ was all the roles one occupied within a certain status position. For example, a medical student – that is their status – occupies different roles, with different expectations and requirements, vis-à-vis patients, nurses, senior medics and so on (Merton, 1968: 423). A ‘status set’, on the other hand, was the collection of different statuses an individual occupies with all their associated roles. A medical student is also a wife, a mother, a member of a political party and so on. Parsons (1951/1991) also made room for prestige as a ranking of general esteem attached to status positions in society and principle of stratification. This was based on a highly skewed reading of Max Weber’s writings on Stände, divested of any connection to the unequal and arbitrary distribution of life chances key to the German thinker’s writings and cast instead as wholly functional. ⁴

The interactionist version of role theory was initially built against functionalism. Its starting point was Mead’s (1934) Hegel-inspired analysis of ‘taking the role of the other’, that is, the projection of oneself into the place of another – whether a concrete other or a generalised other (the ‘community’ or ‘society’) – as the basis of our sense of self and conduct. Blumer (1969/1991) doggedly translated this into a critique of functionalism by insisting that if there are such things as social roles and associated expectations these operate not through entrenched norms and values, such that the causal explanation of activity can be located at the level of collective ideas, institutions or pattern variables, but via the specific meanings, acquired and continually modified through interaction with others and reflexive interpretation, that symbols have for individuals. Goffman (1959/1971), exploiting the theatrical meaning of the term role, then famously extended the interactionist concern with taking the role of the other to examine the way in which people ‘play’ roles in order to give off certain impressions, developing along the way the affiliated vocabulary of performance, dress, settings, teams, levels of commitment and so on. Stryker (1980) later saw in the notion of social role a vehicle for building a ‘social structural version’ of symbolic interactionism capable of examining the degree to which people identify with the institutionalised roles they play and the consequences of that. Both Goffman and Stryker understood that people play or inhabit multiple roles/statuses – judge, mother, customer, etc.– and Goffman (1955), at least, showed interest in how symbols of prestige manifest in interaction too.

Although some asserted the positive benefits of ‘role accumulation’ (Sieber, 1974; cf. Coser, 1991) – reminiscent of Bourdieu’s treatment of multiple field membership – role theorists were long concerned with the widespread phenomenon or role conflict (or sometimes strain), functionalists because it bore on system maintenance and stability and interactionists because it was a focal point of experience and negotiation in everyday life. Often the interest was in competing expectations or demands arising from the same role, but in other instances research focussed on contending expectations and demands arising from different roles – or statuses – occupied by an individual, that is, the mental strains and tensions that come with juggling disparate ‘personas’ and/or obligations (Goode, 1960; Hughes, 1984). Two modes of inter-role/status conflict that attracted particular attention were precisely those already used as examples of field multiplicity: ‘occupational role conflicts’ and so-called ‘sex role conflicts’ (Stryker and Macke, 1978: 75–79). The first of these referred to the friction generated by discrepancy between the demands or expectations flowing from an individual’s employing organisation (a company, a hospital, a university, etc.) and those associated with their membership of a broader profession (law, medicine, science, etc.), a typical theme being dissatisfaction induced by organisational pressure to contravene professional standards or otherwise privilege organisational goals (e.g. profit) over professional ones (e.g. knowledge production). The second covered what is now commonly referred to as work-life (im)balance for women: conflict between family responsibilities and employment demands. Culminating in Hochschild’s (1989, 2001) famous analysis of the ‘second shift’ and ‘time binds’ faced by women balancing paid work and motherhood, scholarship consistently revealed, and still reveals, the sacrifices and compromises made by women – giving up work or going part time, choosing flexible jobs, foregoing or limiting children, abandoning any concept of free time – and the mental strains and diminished sense of self they can generate. How men experienced familial and occupational demands garnered much less attention since the male breadwinner model was still taken for granted in role theory’s heyday and perhaps, therefore, no real role conflict was assumed to exist, at least on any wide scale. Indeed, labelling the theme one of ‘sex role conflicts’ rather than simply ‘family/employment role conflicts’ seemingly, in line with the general logic of functionalism, essentialised or at least normalised male/female roles as they existed in the mid-20th Century. Coser’s (1974) notion of ‘greedy institutions’ – workplaces or other organisations (including the family) demanding all or most of an individual’s attention and interest – did, however, have the virtue of implying some degree of imbalance between roles for men and women alike (see also Coser, 1975).

Role conflict as a notion was premised – whether implicitly or explicitly – on the idea that the clashing roles are of differing salience to the individual (Hughes, 1984; McCall and Simmons, 1966). In part that salience was said to be anchored in the situation – certain stimuli prompt an individual to foreground one role and background another, as when someone receives a work call or email in need of response while otherwise spending time with family. There is also, however, the issue of how important roles and their associated identities are for the individual compared to others. A person’s roles, in other words, are hierarchically arranged, with some being more highly valued or taking priority more often. Stryker (1968) claimed this was measurable in terms of the likelihood of the identity and behaviour of an associated role being invoked across unrelated situations – a politician, for example, neglects family responsibilities to further their career or is constantly thinking about politics even when with family. ‘Identity’ here was understood in Mead’s sense as the ‘me’ reflexively grasped by the individual in the situation, with its constellation of associated wants and desires. When my child cries, it is me as ‘father’ and the corresponding motivations and wishes that is brought to mind and prompted to act. The importance of a role identity, moreover, was defined in terms of degree of ‘commitment’ to it, which was itself understood to be modulated by the intensity and extent of interactional or network connections, and therefore the ‘cost’ of not performing (cf. Becker, 1960; Bielby, 1992). As for why those networks are entered into in the first place, Stryker (1968) stressed choice flowing from an individual’s prior identity or sense of self. In this much, Stryker effectively joined with Parsons and Bales’ (1956) functionalist account of ego socialisation, which held that a person’s job choice, though constrained by adjustment to expectations and judgements in the light of educational performance (‘achievement’), was essentially ‘up to him’ (p. 129). The pronoun is telling: the main aim of Parsons and Bales’ analysis was to explain why boys end up prioritising occupational roles and girls end up prioritising family roles (the answer being familial and cultural role models).

Removing the conceptual chaff

Biddle (1979, 1986) valiantly tried to wrestle role theory from attachment to any specific overarching theory and present it as a generalised, if not axiomatic, element of sociological thought in need of integration and organisation. By the mid-1980s, however, its star had started to wane. This was largely because of the collapse of what Giddens (1984: xv) called the ‘orthodox consensus’ in sociology – the dominance of functionalism (with interactionism as a minor heterodoxy) – as more radical perspectives, spurred by tumultuous times in American society, questioned the absence or scotomisation of class, domination and struggle in mainstream American sociology, including its concept of role (e.g. Gouldner, 1979; Mills, 1959/1970; Stanley and Wise, 1983). With the simultaneous growth, diversification and internationalisation of sociology, alternative perspectives – ethnomethodology, (post-)structuralism, neo-Marxism, structuration, feminism, neo-pragmatism and so on – with either no place for the term social role or a critical attitude towards it proliferated. Sporadic efforts have been made to reinvent the notion by integrating it with cognitive sociology (Danna-Lynch, 2010), neo-pragmatism (Smyth, 2021) or neuroscience (Abrutyn and Lizardo, 2023), but it was the various languages of ‘positions’, ‘practices’, ‘subjects’ and ‘discourses’ that would come to dominate sociological understandings of what people do and why.

Bourdieu himself was embedded in the French intellectual field of the 1960s and 1970s and never had much time in his written oeuvre for functionalism or role theory beyond summary dismissals of Parsons-style sociology as objectivist and theoreticist (e.g. Bourdieu, 1988a, 1991b), and though he had a lot of respect for Goffman’s analysis of impression management (see Bourdieu, 1982) he seemed to claim to accommodate it without much reference to roles (e.g. Bourdieu, 1987: 5). ⁵ He did, however, frequently criticise interactionist logic more generally (see e.g. Bourdieu, 1991c: 64–65; Bourdieu, 1996c: 204–205; Bourdieu and Wacquant, 1992: 256–257), and, perhaps beholden to the imperatives of teaching, he did address Parsons and ‘the Parsonian tradition’ more thoroughly in his lectures and workshops (e.g. Bourdieu, 2014: 73ff; Bourdieu and Wacquant, 1992: 242–243), so we can extend on these remarks to summarise how field theory differs from and advances over role theory (see also Krais, 2006).

The entry point has to be role theory’s handling of power, domination and struggle. In short, symbolic interactionists typically ignored them while functionalists euphemised them. This applied not just to class – where occupational roles were considered complementary, in Durkheimian fashion, economic differences largely overlooked and prestige rankings viewed as motivationally necessary (e.g. Davis and Moore, 1945; Parsons, 1951/1991) – or gender, where domestic roles were considered normal and functional (see Stanley and Wise, 1983). Nor does it apply only to the social system as a whole, where power was narrowly and innocuously cast as facility to achieve one’s goals, and regulate behaviour, via institutionalised rights. It also applies to the manifold professions, organisations and intimate relations – the supposed institutions and collectives – vying for people’s attention. The associated roles – physicist, professor, mother, etc. – were homogenised, that is, assumed to have the same typical set of expectations or requirements attached to them, when in fact the expectations and assumptions that structure the role holder’s practice, and their evaluation of their ‘role identity’, depend on their perception of their place within a discrete system of fine difference and competing forms of value. A lawyer/professor/mother is always a lawyer/professor/mother of greater or lesser standing backed by myriad resources within a specific set of relations between lawyers/professors/family members, with a specific set of possibilities for action and interests open to them as a result. They may just use the authority delivered by those resources, moreover, to contest and impose specific definitions of the ‘role’ as the legitimate one to be adhered to by all, whose form and degree of ‘fit’ is then judged accordingly. At best classical role theorists acknowledged the differentiation of senior/junior roles and its part in structuring appropriate behaviour (e.g. Coser, 1966), or leader/led roles as modes of ensuring collective interests are met (Parsons, 1951/1991), but not only is this too simplistic in its binarism and uni-dimensionality, and still laced with assumptions of functionality in Parsons’ case, but it went unconnected to the issues of commitment and salience and the differences in practice they may generate.

The concepts of social space/field, symbolic power and feel for the game address precisely these problems. There are, for sure, some points of connection with role theory: like functionalism – because of shared debts to Weber and Durkheim – Bourdieu was interested in the progressive differentiation of social orders in the modern period (Bourdieu, 2014: 75); like Mead – because of a shared debt to Hegel – Bourdieu understood that seeing ourselves through the eyes of others is fundamental to our social existence; and like both functionalists and interactionists, Bourdieu put a premium on expectations in guiding our behaviour. Yet differentiation is not of homogenous and functionally complementary spheres, institutions, professions, sub-systems or collectives but of fields of struggle which, if anything, mark a differentiation of the division of labour of domination (Bourdieu and Wacquant, 1993). Perceiving ourselves through the eyes of (concrete or generalised) others, moreover, is always contextualised by our and their (imagined) possession of capital; the sense of the forthcoming and ‘what is (to be) done’ are, contra role theory, individuated according to our specific position in the field without losing entirely the shared element encapsulated by doxa; and the very meaning of ‘roles’ and what they (should) entail become stakes within the struggle for capital. ⁶ These same principles apply to the class structure – where the stakes include, but are not reducible to, economic capital ⁷ – and the various fields – professions, organisations, families – with relative autonomy from it.

The pathway into social roles, moreover, was always underspecified in role theory. The functionalist model of ego socialisation applied only to boys and girls in general (and is now outdated), and when it comes to why one work-related role is committed to over another, both Parsons and Bales and Stryker defaulted to voluntarism, which is a foundational principle of both symbolic interactionism and functionalism (no matter what Parsons later built on top of his ‘voluntaristic theory of action’). This is despite both strands of role theory acknowledging the efficacy of expectations and anticipations in shaping choices and practices, specifically regarding roles as identified by symbols and signs but also in relation to educational performance in Parsons and Bales’ formulation. The conceptualisation of family as a field offers a solution by specifying the process by which individuals come to value the things – the capitals and fields – they do: it is rooted in the earliest struggles within the familial field for recognition from caregivers, who are not mere embodiments of universal roles but themselves equipped with varying levels and balances of capital and thus distinct dispositions regarding the achievable and the praiseworthy (Atkinson, 2016; Bourdieu, 2000a: 164–167).

Still, for all the limitations and outdated assumptions of role theory per se, the notion of role conflict, and its related conceptual distinctions and vocabulary, articulates in precise terms the observable tendency only hinted at by Bourdieu for people to be embedded in different sets of structural relations – professional, familial, organisational, etc. – with competing demands for time, attention and interest and significant consequences for mental life, social conduct and wellbeing. This tendency is all the more prominent in the wake of changing patterns of employment and their effects on family life, or, more specifically, the tensions and strains faced mainly by women as they juggle newfound expectations to enter and progress in the workforce and old expectations that they will shoulder the majority of housework and childcare. How, then, to integrate the residual insights of role theory to fulfil Bourdieu’s call for conceptual development of his system?

Conceptualising the phenomenology of inter-field conflict

The ‘social surface’ offers a starting point for capturing multi-field membership and conflict but needs to be elaborated and refashioned. Reconceiving it as a ‘meta-habitus’ is the first step. Let us say an individual has a habitus in relation to a specific field – a sense of the possible, likely and desirable in relation to capital accumulation and position-taking there. This is a specification of Bourdieu’s own malleable use of the term habitus, which occasionally seemed synonymous with the social surface but which was otherwise said – perhaps most definitively in Bourdieu and Wacquant (1992: 16–23, 124–140) – to be inherently defined by its relation to ‘a’ or ‘the’ field (and was obviously considered different from the social surface, otherwise he would not have coined the latter term). In phenomenological terms, we might say this field-specific habitus depends on a ‘field horizon’: the set of meanings and expectations co-given with a percept that relate specifically to the structure, doxa and dynamic of the field, such as an artist’s ‘feel for the game’ of art, or an employee’s ‘feel for the game’ of the organisation they are a part of.

That same person is, however, a member of multiple fields – which, after Merton, we might usefully call their ‘field set’ – meaning they also have a sense of how what goes on or will happen in one field will affect their possibilities and practice within another field. A family obligation or attention-sapping tension means they will not be able to finish the artwork when or how they would have wanted, or complete that report for the management team of the firm, or an upcoming exhibition or report deadline means they cannot spend time with their children or attend (as promised) a specific family event. This feel for the multiple games they are playing feeds into their decisions and practices – whether to attend to this or that first, to give up this or that, with more or less clearly anticipated consequences, and so on. Strategies may operate across fields, and strategies – and capitals – within one field may even be sacrificed, compromised or timed according to another.

Underpinning the feel for the games one is playing is what might be called the ‘lifeworld horizon’, since it denotes the meanings and forthcomings co-given with a percept relating to the myriad other fields structuring the everyday lifeworld. ⁸ I take this to be the core of the notion of social surface, or meta-habitus, comprised of not only all the individual’s habitus in relation to specific fields but the practical mastery of the relations between them. Although it may often seem impossible and superfluous to disentangle them in analysing specific experiences and practices, the distinction between field-horizon/field-specific habitus and lifeworld horizon/meta-habitus is nonetheless useful for capturing the practical sense of ‘if’ – ‘if I did not have this going on (in that field) I would/could do that (in the other field)’ – and thus the sense of being thwarted or constrained in what one does. Of course, the relation between fields in an individual lifeworld may also be one of facility – ‘because of this going on (in this field) I am able to do that (in that field)’ – which is perhaps what Bourdieu was more interested in. ⁹

The lifeworld horizon can be further specified by accommodating role theory’s dual sense of ‘salience’. On the one hand there is the situational element of trans-field experience. Some situations – or more generally, perception of some person, thing or event – ‘call forth’ or at least foreground the expectations and dispositions relating to a specific field, like the crying infant while home-working invoking familial dispositions and expectations (thoughts related to other fields may recede into the margin or periphery of consciousness rather than out of it altogether). In most cases, though, they will be pertinent to more than one field simultaneously and be perceived or felt as such (‘I must go to her’ and ‘I’ll never finish this report!’), and the resultant practices themselves may bear on strategies, capital and positions across different fields. Some fields, however, will be ‘greedier’ than others insofar as they – or other players within it holding symbolic power, to be more precise – demand more time and attention for effective participation. Their effects – phone calls, emails, or just ‘taking up headspace’ – may be experienced as encroachments or invasions with deleterious impact on participation and position in other fields. Practices adding up to strategies in a specific field, for example, may be delayed, modified or sacrificed altogether to face ‘what must be done’ in another, ‘greedy’ field.

On the other hand there is degree of ‘commitment’. This theme makes plain that illusio is never singular and uniform in strength but plural and variable in force, thus structuring priorities across situations and what one tends to think about. Adapting the language of Schutz (1970), we might say that some field effects become salient or relevant in consciousness because they are imposed by outside events, including the actions of others and their knock-on effects, while others are intrinsically relevant in the sense that we value them such that our minds incline towards them (whether it be problem-solving, projecting, fantasising or worrying). This may itself be a source of tension of conflict, as with the politician neglecting family duties because of their differing illusio for each game and visibly thinking about politics while with their family – to the annoyance of family members and with effects on their position and perceived value within the familial field – or the parent resenting having to work and be away from their child(ren). Or it may be a source of succour, as negative experience, a failed strategy or domination in one field is compensated by throwing oneself more fully into another (Atkinson, 2022a, 2022b).

Breaking with the interactionist vision propounded by Stryker and others, however, but also the Freudian model of desire associated with Parsons, if strength of illusio across different fields should be thought of in relational terms it is in the structural rather than ego-alter/self-other sense of the phrase, that is, as organised around an intra-psychic polarisation of dominant and dominated desires and commitments but with varying distances between field illusiones specifying and weighting their evaluation and salience across situations. ¹⁰ Someone may value family and their professional career above their employing organisation or the overt trappings of class (money, ‘cultivation’, etc,), making those field desires dominant, but still value family to a greater extent than their career (or structural proximity generates tension between them). This polarisation is crosscut by a dimension of time, however, capturing the waxing and waning, or genesis and disappearance, of illusiones over the life course – trajectories dialectically related to experiences within fields – and with it the sense of what one used to care about, what was once more important relative to other things, and even what one might care about in the future (e.g. anticipating parenthood). Every individual, then, can be characterised by an illusio space, which, together with the lifeworld horizon, offers a means of articulating intra-personal relations between social structural relations and, more pertinently, a conceptual tool for explaining how and why people navigate and have navigated the play of structural forces across their lives and in their contemporary lifeworld. Although individual, there are also likely to be discernible patterns and tendencies in the structure of illusio spaces according to, for example, gender, age and class, opening up possibilities for exploring the social distribution of saliences and inter-field clashes on a broad scale too.

Field-set, meta-habitus, lifeworld horizon and illusio space are categories capable of rendering concrete or empirical individuals in conceptual terms. Bourdieu’s constructed/concrete division thus begins to dissolve, since individuating existence across multiple fields has been subjected to sociological construction. There is no escaping construction, in fact, as Bourdieu knew, and thus no such thing as a pure ‘empirical individual’ – people (like anything else) are always perceived and evaluated, by anyone, through schemas of acquired symbol-sign associations, some of them subjected to the trials of sociological reason and some of them not. Let us replace the questionable epistemological divide, then, with an operational one: that between field analysis and lifeworld analysis. The former focuses on the major structural relations of a specific field, isolating the capitals and habitus in play there in order to understand strategies and field transformations in broad terms – Bourdieu’s usual modus operandi – but with open awareness that extra-field factors are being ‘bracketed out’ (to use an apt term from Giddens, 1984: 285). The latter, however, takes as its object of construction and analysis the play of fields within individual lifeworlds and social surfaces so as to explain a specific biography, event, pattern or regularity more fully, which may involve bracketing out the detailed structure of specific fields. It also envelopes the major substantive contribution, of continued if not increasing prominence in sociological research, of the jaded tradition of role theory: multiplicity of social positioning, commitments and expectations and the mental strains and practical outcomes it generates.

Conclusion: The past informs the future

Bourdieu’s conceptual toolkit continues to find application and test in empirical research as well as frequent discussion across theory journals and books decades after it was first presented. There are many points of contention, and countless claims that it misses or underplays this or that aspect of social life – change, resistance, agency, materialism and so on have been the usual bugbears – sparking manifold efforts to modify or tinker with the basics. This is desirable: if the conceptual toolkit, with all its philosophical underpinnings, and the vigorous research programme it has generated are to survive, progress and thrive into the future, it must be put into continual dialogue with contrary positions and arguments and take on their lessons. That obviously includes perspectives that have emerged or flourished since it was first formulated, including globalisation theory, the new materialism or postcolonialism, as well as perspectives from beyond sociology, including cognitive psychology, neuroscience or geography. Yet it also includes continual dialogue with the past – with the ‘classics’ old (Marx, Weber, Durkheim, Simmel) and new (Du Bois, Khaldun, Beauvoir, etc.), the pre-classics (Hegel, Smith, etc.) the ancients and so on, but also with older traditions from the 20th Century. The backward-facing dialogue must be discerning, carefully identifying advances and oversights, but it is nonetheless vital if we are to avoid losing relevant insights from the past or misguidedly reinventing the wheel – and if we are to be faithful to the collective practice and history of social science, which requires antagonism and reasoned mutual agreement to improve (Atkinson, 2020; Bourdieu, 2004).

Such is the broader context for the specific tasks pursued in the present paper. Engaging with role theory and plumbing its insights have allowed us to address an identified gap in Bourdieu’s conceptual toolkit, namely, multiple positioning, its effects on practice and the strains it induces. For sure, I have argued that the notion of role and the conception of social structures that went with it were problematic and should be replaced by the notions of habitus and field, and their coterie of allied concepts, but nonetheless a backwards trawl over some of the themes that role theory was used to make sense of has served several purposes. First, it has emphasised that the gap in Bourdieu’s toolkit is no small oversight but something that has informed and still informs rich bodies of research, thus pressing the need to remedy it. Second, and relatedly, it has revealed multiplicity to be an old issue, not a new one, and that there have been efforts to theorise it in the past which must be considered so as not to pretend to be the first. Third, those previous efforts have disclosed sub-themes – especially related to salience and commitment – necessitating careful conceptual elaboration to accommodate them. Thus I suggested the notions of meta-habitus, field-set, lifeworld horizon, illusio space and greedy fields. And fourth, it has forced us to convert a questionable epistemic boundary in Bourdieu’s thought – the confinement to field analysis – into a distinction between two complementary modes of sociological analysis – field analysis and lifeworld analysis. These moves expand both the explanatory range of Bourdieu’s relational sociology and its field of possible research objects – markers, I would hold, of a progressive shift in any conceptual system.