University of Leeds

Workshop Paper No 16
Prepared for Workshop Four
Methodologies for Researching Moral Agency
Friday 17 March 2000

Fiona Williams


Executive Summary

1. This paper attempts, first, to clarify those concepts, some of which we have been using in the Workshops, which seem to be important to the different areas of our research and, second, to chart how different concepts and ideas may relate to each other within the context of our research questions and aims. This is done rather tentatively, in the spirit of providing a paradigm of articulation between different parts, not a grand design. I have not tried to encompass everyone's work, but I have tried to provide a framework through which we can locate and link different ideas. I am also more concerned here with clarifying our methodological and analytical concepts, rather than our normative position(s). It is an exercise in finding commonalities and links through our differences.

2. Drawing on earlier work, the workshop papers and other work by the Cava research group, I suggest that the concepts we are using fall into clusters or constellations within four overlapping and interconnected fields of analysis. These are (i) the subject (ii) the social topography of enablement and constraint in the intimate/informal/ close /local/ context (iii) the wider discursive and institutional contexts and (iv) the dynamics of social change.

3 The paper then elaborates the use of the concepts in each field and the way they relate across the different fields of analysis. In relation to the first field - that of the subject - it is proposed that our research focuses upon people as moral, welfare and care subjects, as well as social subjects constituted through a spectrum of social relations. It then examines, with reference to empirical research examples, different aspects of the subject - subjectivity, identity, subject positioning and agency.

4. In the discussion of the second field of analysis, the paper debates, again with reference to empirical research examples, how far the sphere of the relational is a distinct field of analysis which inflects subjectivity, identity and agency with particular meanings.

5. The discussion of the third field of analysis, the wider discursive and institutional contexts, focuses largely upon the way in which we operationalise an understanding of complex and multiple social relations for our research.

6. Field 4 on the dynamics of social change discusses briefly the concepts to which our work refers (and generally critiques, such as individualisation). It also briefly refers to the issue of where cultural/ social change happens.

7. The conclusion starts to identify that which is common to all our approaches and
that which may be construed as distinctive about Cava research.

Three Concerns

In order to develop this chart, I draw on, and develop, earlier work on conceptualising the agency / structure relationship for rethinking research into poverty and social exclusion (Williams, 1996, 1998) and social policy more generally (Williams & Popay, 1999). Three concerns have been central to this work:

  • An emphasis upon the capacity of people to be creative, reflexive human agents, experiencing, negotiating and developing their own strategies for the management of their welfare and reconstituting the outcomes of welfare policies.
  • An attempt to apply to the framing and practice of research the theoretical insights drawn from more nuanced theoretical inquiries into social identification and differentiation which point to the multiplicity, complexity and variability of identities and social categories.
  • In pursuing questions of subjectivity, agency and identity, not to lose sight of processes of social formation, especially those constituted through power and inequality; to this end, to discover how researchers have developed and made use of 'mediating concepts' which link 'agency' to 'structure'.

The 'social policy' background to the emergence of these concerns is relevant here.


One impetus behind this rethinking was the inadequacy of the paradigm which had dominated social policy research. This provided analyses of policies in terms of measurable outcomes for social groups, and these social groups were categorised objectively, in terms of constructions by the Registrar-General, or by researchers, or policy-makers (that is, socio-economic/occupational groups, or groups in administratively-defined need - single parents, frail elderly, the poor, chronically ill etc.). In general, the people who inhabited these fixed categories were seen, and still are, as the passive beneficiaries (or not, as the case might be) of different policies, whose content might vary according to dominant political ideologies, or, in the political economy version, according to the compromises wrought between labour and capital. In this scenario, the concept of social divisions was generally limited to class, and the concept of social categories to income, occupation and/or an administrative definitions of social need. The concepts of agency, identity or personal experience were either non-existent, or they surfaced as collective class struggle of labour (or later, collectively organised women). In this framework the outcomes of policies were understood to be measurable in terms of access to benefits and services created to alleviate certain identifiable risks and ends (poverty, unemployment, housing, health and education). Occasionally less measurable terms, such as stigma, emerged to explain some of the mismatch between policy implementation and personal or cultural responses.

A number of social, political and intellectual changes forced a rethinking of this paradigm (see Williams, 1998, for full explanation). First, from the mid-70s, a whole range of political, economic and demographic forces challenged the social democratic welfare state. Amongst them were competing political discourses of welfare, so, for example, by the 1990s, New Right politics of "consumer sovereignty" and "individual choice" and "diversity of needs" competed with notions from the left and user/social movements of "user-control", "welfare citizenship" and the "diversity of social rights", and, from an emerging New Labour, with notions of "responsibilities with rights". However, in spite of their political differences and pedigrees, what all had in common was an emphasis upon the welfare citizen/consumer as, first, agent of her/his welfare destiny - whether through the market, through exercising social responsibilities, or through local, democratic forms, and second, as articulating her/his differential welfare needs. This helped shift the grounds of social inquiry. It also helps to explain why, in the early 1990s, the ESRC set up a new research programme on "The Management of Personal Welfare", subsequently directed by Jennie Popay and Ann Oakley(1) . Its brief included the development of new research on people's personal coping strategies which would identify them as 'creative human agents'. I was brought into the programme and charged with the task of reviewing the literature to see whether there was any research on stress and coping which provided an adequate conceptual framework capable of holding together these newer concerns with agency with the more traditional social policy concerns with the structures of poverty and inequality.

Needless to say, there was very little. On the one hand, the research on stress and coping was dominated by the methods of quantitative psychology in which individuals and their experiences were generally subsumed into aggregate measures of correlating variables through which stress was, or was not, held at bay. Gender, age, ethnicity, class, and so on were brought in simply as differentiating variables, rather than as intersecting modalities of difference which could render experience and action more meaningful, and through whose relations power is implicated and performed. On the other hand, the conceptualisation of agency in research influenced by economics was in the rigid shape of rational economic man driven by self-interest (see Duncan's papers for Workshops 1 & 3 (Duncan, 1999a and b)). Of greater concern was the fact that within the very broad area of welfare research there had been few attempts, with notable exceptions, to apply these newer theoretical developments to the agency / structure question. One of the most promising of these theoretical developments was Giddens' structuration thesis (Giddens, 1979, 1984), and the most notable exception was Janet Finch's Family Obligations and Social Change (1989), which took as its starting point the formulation by Giddens (1979) of the need to recognise the relationship between social structure, on the one hand, and the possibilities for independent human action on the other:

"Giddens' view of social structure is that it provides resources which people can use when they are interacting with others in daily life. In that process social structure is also reinforced and reconstituted. We need to see human beings acting purposefully in their use of such resources, although the outcome of their actions may have consequences which they did not intend. In the context of theory, Giddens rejects the view that patterns of action are straightforwardly imposed on people, including those theories which tell us that people internalise the norms of society and then produce appropriate action … People need to, and are able to, explain what they have done and why they have done it and this in itself forms part of the action. We all do this by drawing upon our understanding of how the social world works, by using the same shared knowledge of our society which we use to formulate our conduct" Finch (1989, p87-88, my emphasis)

(I will come back to this as this quotation, and Finch and Mason's subsequent work, begins to identify some 'mediating concepts').

Identifying/ Unpacking the Concepts

Bearing in mind the three concerns I outlined at the beginning - i.e. to develop (1) an emphasis upon creative human agency, (2) a more complex understanding of the welfare subject, and (3) mediating or middle-range concepts which connect 'agency' to 'structure' - I reviewed empirical policy-relevant research which was also attempting to deal with these concerns. From this I suggested that some 'exemplar' researchers were developing their frameworks across two or more levels of analysis. Drawing on this, as well as the workshop papers and other work by the Cava research group, I suggest that the concepts we are using fall into clusters or constellations within four overlapping and interconnected fields of analysis. I use the word 'field' deliberately to signify two meanings. The first is that of location, or boundaried space; the second is that of a sphere of operation, a region of influence or force (as in electric or gravitational fields). Thus, the four fields of analysis are, in brief:

1. The contemporary (welfare) subject, drawing upon concepts of subjectivity ( a sense of self), identity (a sense of belonging), subject position (one's positioning within socially constructed discourses, practices and categories) and agency (one's capacity to act). Related to agency are concepts of doing/acting such as reasoning, negotiating, and practices.

2. An intimate/informal/ close /local/ context, a social topography of enablement and constraint which the (welfare) subject inhabits, signified by risks, opportunities, relationships and resources; these are marked by variability in terms of both their distribution (usually interpreted quantitatively) and the differing and changing meanings attached to them (usually interpreted qualitatively). Later in the paper I discuss how far the intimate, or relational dimension, with its notions of informal rights, responsibilities, obligations, normative guidelines, gendered moral rationalities, and moral scripts, should be seen as a distinct field of analysis.

3. The discursive and institutional contexts (subnational/ national /international) of policy formation, implementation and delivery; extends to social, economic and cultural contexts. Key concepts: discourses and practices (welfare, legal, cultural, moral), social relations and formations, institutional arrangements (formal rights and responsibilities attached to citizenship, welfare reforms, forms of governance etc.).

4. The contextual dynamics of social, economic and political change
: all the posts, and associatedly, pluralisation, fragmentation, hybridisation, de-centering, globalisation, individualisation, reflexivity, unsettling, resettling, etc.

In our different papers and research we have different analytical starting points. For example Jennifer Mason usually starts from 'agency' and 'practices' in field 1, links it mostly to field 2, but brings in field 4. Carol Smart and and Bren Neale's work on divorce often starts with a mismatch between socio-legal discourses (3), explanations for change (4) and actual practices (1). Sasha Roseneil and Fiona Williams are often concerned with linking collective identities and actors (1), with social formations and social relations and cultural (S) or policy (F) discourses (3) and social change (4). Simon Duncan's and Alan Deacon's recent work starts with policies in field 3. For Simon Duncan, discourses, opportunities and risk in the local are extremely important in understanding agency and identity. Attempts to capture the dynamics between fields 3 and 4 are Fiona Williams Family-Nation-Work (see Workshop 1) and Simon Duncan's genderfare, (Workshop 1) and Sarah Irwin's moral economy (Workshop 1).

Although these fields of analysis are dynamically interrelated, the next sections explain them in sequence. The whole of the following discussion is represented in Figure 1.

Field of Analysis 1: The Subject

Whilst the focus of my original research was welfare subjects, the focus of CAVA's research concerns subjects whose fields are, at one and the same time, both broader and narrower than those inhabited by welfare subjects. They are narrower in the sense of having a rather more specific focus upon 'parenting and partnering practices'. But they are also broader in the sense that, to begin with, we are concerned with subjects who are constituted through more than welfare and law. We are interested in the ways in which moral subjects are constituted , how they exercise their moral agency, how they pursue moral practices, and how they create and construct moral subjectivities and moral identities and how all of this influences moral decisions. What is important about this approach is that it moves the moral subject out of the world of liberal philosophy in which 'he' has a given, or pre-social, moral identity into a sociological world in which he forms and reforms her/his moral identity, and engages in moral activity as a social practice. (It is, however, difficult to separate out this field of the moral subject as a discrete area. Moral discourses, practices and relations are implicated in welfare/ legal discourses of desert and dependency, etc., and they permeate practices of care and intimacy). Furthermore, in so far as we are concerned to research those who are caught up in parenting and partnering practices (and not simply those who are parents and partners themselves) then we are concerned with another kind of subject. Perhaps we could call this one the care subject (other options are the 'intimate' or 'family' subject; but I think 'care' is better). So, within the field of family, care and intimate relations we are concerned with how the care subject exercises creative agency, with the practices which follow, how discourses of care, intimacy, parenting and partnering constitute that care subject, and with the experiences and forms of identification that parenting and partnering practices engender.

We are also interested in people as social subjects. Again, the separation of different types of subject in reality does not occur, for all these different types of subject positions intersect. But in so far as we are interested in social and cultural change, then we are also concerned with how changing gender, class, ethnic, sexual, generational and employment relations are reconstituting people's experiences, identities and practices. In Sasha's paper (workshop 2) and mine (workshop 1) these changes in the subject are related not only to changing social relations but also to the plurality and the decentering of identities which challenge fixed boundaries between straight and gay, black and white, and so on. In relation to care subjects, this idea is also there in Bren Neale's, Carol Smart's and Amanda Wade's work in the shape of the challenge to the binary of blood and step parent.

In these ways, then, the notion of the subject is dependent upon a series of other concepts - namely, subjectivity, identity, subject position and agency. In their turn, these concepts are dependent upon others - for example, agency to practices, and subject position to discourses, and most of them to social relations. We can develop working definitions of these four dimensions of the subject as follows:

subjectivity - this refers to people's sense of self, usually interpreted (and constantly re-interpreted) through their experiences. Wendy Hollway's paper (Workshop 3) examined the development of one person's moral subjectivity with reference to conscious and unconscious elements in their life-history. She then related these to wider discourses which positioned Ron as a father, brother, criminal, and so on. What she emphasised was that while Ron's sense of himself was of his own making, it was still born of his relations with others, as well as of his positioning within his particular and a general social world (in other words her analysis of intersubjectivity crossed and connected the first three of the fields I marked out above)(2).

This connects to a point that Selma Sevenhuisjen makes:

"In contrast to an atomistic view of human nature, the ethics of care posits the image of a 'relational' self - a moral agent who is embedded in concrete relationships with other people and who acquires an individual moral identity through interactive patterns of behaviour, perceptions and interpretations" (Sevenhuijsen, 1998, p55). This raises the issue of identity:

identity - signifies a person's sense of belonging(s); it refers to the ways in which, and the sites through which, they attach themselves to the social world. Of course, these attachments are multiple, shifting and velcro-like - able to detach and reattach at different points and different times. Identity faces two ways: on the one side it looks to, indeed, shapes, the self; on the other, it looks to the discourses which place us as social beings. The first of these two faces is defined by Avtar Brah as follows:

"identity may be understood as that very process by which the multiplicity, contradiction and instability of subjectivity is signified as having coherence, continuity, stability; as having a core - a continually changing core, but the sense of a core, nonetheless - that at any given moment is enunciated as the 'I'" (Brah, 1996, p123-4, her emphasis)

For Stuart Hall identity is the pivotal link between subjectivity, subject position, practices and discourses, for it is both the meeting point between those discourses and practices which position us as social subjects and those processes which generate our subjectivities: "identities are thus points of temporary attachment to the subject positions which discursive practices construct for us." (Hall, 1996, p6). Furthermore, Hall places these concepts within wider social change - what he calls the "pluralization of social life" - which "expands the positionalities and identities available to ordinary people (at least in the industrialised world) in their everyday working, social, familial and sexual lives" (Hall, 1996, p234).

David Taylor, in an article on identity and social policy distinguishes these two faces of identity as ontological and categorical identity. Ontological identity derives from the attempt to impose a coherence upon fragmented experiences and concerns that which marks out one's unique and different self; categorical identity concerns an identification of oneself (or others) as belonging to a same social category (woman, cat lover, single parent etc.). It is also partly between these two that the discursive construction of subject positions are resisted / reproduced / and resignified. Is there any difference, then, between the concepts of ontological identity and subjectivity? Ontological identity signifies the process of creating coherence from experience whereas the span of subjectivity is much broader: its reach extends beyond conscious experience to unconscious interiority.

In terms of the moral subject, what would we mean by moral identity? Jennifer Mason's research on inheritance (Mason, 1999) is a good illustration of the way people create their moral identities through the practice of making wills. What it means to be seen as "a good mother / father" was often very important to her interviewees. Similarly, when grown children assess their mother's or father's parenting as "s/he did her/his best in the circumstances" they ascribe a positive moral identity to their parent based on experience whilst at the same time articulating some kind of general moral yardstick of "good enough" parenting. This idea of an identity expressed in relational terms is something I return to later.

subject position - refers to where and how one is positioned within the social world. As such, unlike subjectivity and identity which are precipitated subjectively (though constructed with reference to. the social), subject position is largely externally constituted - I say 'largely' because subject positions are continually being revised through changing identities (and these are themselves shaped by the sense of self continuously revised through the interpretation of experience). Subject positions are the positionalities that constitute us through discourses (3) and as subjects. As discussed above, these subject positions include the social categories associated with gender, race, ethnicity, class, sexuality, disability, age, generation and so on, but we can also use the concept more specifically to refer to moral, welfare and care subjects. The ways subject positions are ordered (as binaries - men/women, deserving/undeserving, good/bad, carer/dependant), as dominant/subordinate, as equal, same or different) constitute their social relations. Our research is particularly concerned with the interplay between changes in the ordering of gender, ethnic, sexual, intimate, parent, generational and friendship relations and constitutions of people as moral subjects and care subjects.

Recent empirical research in social policy on the relationship between experience / identity and subject position has usually examined the resistance of welfare subjects to dominant welfare discourses, (e.g. Simon Duncan's work on the New Deal and Labour's rationality mistake). One good example of this is John Baldock and Clare Ungerson's "Becoming Consumers of Community Care" (Baldock and Ungerson 1994). In this study they interviewed 32 stroke patients and their carers in order to explore how they cope with their sudden and often unexpected change of circumstances with reference to the mixed economy of care provision resulting from the 1990 NHS and Community Care Act. This legislation, amongst other things, instituted a move towards the encouragement of low-cost home care from the informal, voluntary and commercial sector, along with a right of users to assessment of their needs for care through a consultation process with a care manager. The implementation of the Act was couched firmly within the discourse of consumerism - that users will be able to articulate their needs and, with access to information, exercise choice within the mixed economy of care provision. Baldock and Ungerson found, however, that their respondents did not adapt in a straightforward way to a consumer identity. One reason for this was that their respondents felt extreme uncertainty in the face of their new medical condition:

"Organising social care, from the point of view of its users and consumers, is a largely unscripted process. That is to say that people have few ideas as to what can be done and at the same time and more importantly, they are very uncertain about what should be done" (1994, p44).

There was, to use the Finch-phrase, little sense of "the proper thing to do". Furthermore, their past experiences and how they understood these experiences (as users of public services or consumers in the market) failed to provide them with clear guidelines as to how to act. The 'normative guidelines' (my use, not theirs) drawn from the discourse of consumerism in the market did not fit well with the circumstances of requiring social care. And, whilst the role of service providers was to give people information in order to make choices, the researchers found that people did what they did, not on the basis of factual information, but on the basis of values and culture.

Part of the immediate problem for the stroke victim and their carer was of adapting to a new condition of 'social dependency'; it is this that was a key part of the 'unscripted' process. In addition, a consumer identity enabled people to think in terms of buying some things, say consumer durables, but not others, such as personal care. A third problematic element was that people brought to the situation a particular identity as a welfare user, accumulated from their experiences. The practices which flowed from these identities Baldock and Ungerson called "habits of the heart" (from de Tocqueville) and placed them along two axes which distinguished between four types of 'disposition towards the provision of care'. These were consumerism - where, at its extreme, one expects nothing from the state and subscribes to a view of individual self-sufficiency. It included those who, regardless sometimes of income, preferred the sense of control to be had from actively seeking and buying-in services through their own networks. In contrast, the second disposition, privatism, represented a different, more passive form of consumerism, in which people regarded dependency as a stigma, often refused offers of help and preferred to keep themselves to themselves, again valuing their autonomy. The third was welfarism, where people who drew on this believed in the welfare state and their right to use it. It often involved the active pursuit of their entitlements. Finally, clientalism generated a 'traditional' approach to the welfare state: "passive, accepting patient and grateful". None of the respondents was constituted simply in terms of the one disposition / discourse or another, but those who were able to participate more in the emerging landscape of social care were those who saw themselves more actively as either consumerists or welfarists.

I have elaborated this example because I think it demonstrates an interesting use of concepts which mediate between the field of identity, subject and agency and the concept of discourses in the third field of analysis - these include dispositions, scripts and habits of the heart. In addition, I have framed their work using concepts drawn from Finch and Mason - proper thing to do, normative guidelines and practices. These also mediate between identity, agency and discourse. In both studies, too, it is practices or habits which articulate with values, norms and 'oughts' and rewrite the guidelines or scripts. In a sense, by drawing out these, these researchers are spelling out what "agency" means.

agency refers to people's capacity to act. As such, and in order to research it, it has been tied to those concepts which allow for the detailing of actions, such as habits, practices and strategies. Agency runs through all three of the conceptual elements of the subject I have discussed so far. Experiences shape subjectivity, and subjectivity shapes the meaning of those experiences; identity informs our practices (often in ambiguous and contradictory ways: "I'm not a feminist, but…") and those practices signify and resignify our identities. The subject positions we inhabit frame our practices in particular ways, while our practices can reproduce or reconstitute the meanings of those subject positions and their ordering. Practices may be productive of power and they may challenge oppressive power. Furthermore, as Jennifer Mason's paper for Workshop 3 suggested, agency expresses itself in relational terms; individuals do not simply act as individuals within, or with reference to, relationships, but their very agency is constituted in and through their negotiations with others, their capacity to act, their reasoning, their actions and their practices. (Here, note, negotiation and reasoning are added to the conceptual toolkit of 'agency'.) I am not convinced that relationalism applies so thoroughly to all subjects in all positions, but it is a significant condition for the actions of the moral subject and the care/familial subject (A point discussed more fully later). In her discussion of the moral subject, Selma Sevenhuisjen ties some of these concepts together as follows:

"…the definition of identity and selfhood has shifted. Individuals are no longer seen as atomistic units with a pre-determined identity, who meet each other in the public sphere to create social ties. Identity, and with it the ability to engage in moral activity, is formed in specific cultural and historical situations, and thus it coincides with subjectivity, the ability to judge and to act. The self is not conceived as an entity, but as the protagonist in a biography which can contain all kinds of ambiguities and unexpected turns … The feminist ethics of care has more to gain from the idea of a processual self, a self which is continually in the process of being formed; moral identity is continually being developed and revised through this process. The construing of moral identities is thus, in this sense, inherently a social practice, something which we do and make within human relations and within specific social and political contexts, and the narrative conventions reflected in these." (Sevenhuisjen, 1998, p55-6)

In Cava we are concerned to research the moral agency of relationally constituted individuals, but we are also, in Strand 5, concerned with collective agency and the collective identities which generate and are generated by this, as well as the subject positions they challenge. Collective identity and collective agency cannot be read off from individual identity and agency, for the processes involved in their formation are different. To begin with, collective identification summons up commonalities across the experience of being different / particular (of being a woman, a mother, a racialised subject, etc.). In so doing, it freezes those collective identities in particular ways at historically and culturally specific times such that individual differences are temporarily erased. (At the same time, this process feeds back onto individual identity and practice: "I became a feminist and stood up to my husband"). This freezing generally takes place in the mobilisation of collective agency and in the articulation of needs or claims. The practices of claims-making and their challenges to discursive (and non-discursive - e.g. visual imagery) formations therefore become our focus in Strand 5.

But we are also concerned with a different form of collective agency and claims-making which is identified in Sarah Irwin's paper (Workshop 2 #1) and which distinguishes between explicit and implicit claims-making. She argues, in relation to changes in divisions of labour, family and employment, that implicit claims emerge through the very practices in which people engage, and these may not involve conscious reflection but be part of cultural embeddedness - what went before. This point helps us link the work of Strand 3 to Strand 5.

Although I have put agency in the first field of analysis as a dimension of the subject, it actually represents the closest point of entry - the stile - into the second field of analysis: the intimate/informal/ close /local/ context with its the social topography of enablement and constraint. Agency and practices take us into the realm of what's possible, into the practicalities and the realities of action, and is thus a more favoured habitat for sociologists as it is in listening to the narration of (or observing of) practices that they feel their ear is both closest to the ground, and closest to context / structure of fields 2 and 3. Concepts such as normative guidelines, moral scripts, obligations, responsibilities and rights, and the ways these are, related to social, cultural, legal, moral, policy discourses serve to mediate between the fields of analysis of the subject and those of context. Furthermore, in the research on moral agency, to recoin a phrase, practices mean principles, or at least, it is from people's moral practices or groups' claims that one can begin to (re)construct a normative order from which they are drawn and which they create (see for example Chapter 9 in Carol Smart and Bren Neale's Family Fragments (1998) (from practices) and my Good-Enough Principles for Welfare (from claims) (Williams, 1999) - a point pursued later.

1. Field of Analysis 2: the social topography of enablement and constraint in the intimate/informal/ close /local/ context

This refers to the immediate context in which people perform, negotiate and act. For our moral and care subject this is a field, crucially, of close relationships. What should be stressed however is the ambiguity of relationships - they enable and constrain, they provide opportunities and they generate risks (this counters the criticism that by emphasising the negotiational and the relational in parenting and partnering practices, one is in danger of suggesting an inevitable and positive correspondence between intimacy and relationships). The other key concept used in this field is that of resources. In Giddens' use the concept of 'resources' refers to the structural properties which individuals draw on in everyday life. Resources are both the medium and outcome of social life, and represent the interdependence of structure and agency. They are both enabling and constraining. In some research resources is the generic concept to cover not just 'things', such as material resources, but also personal resources such as coping strategies, relationships, and networks. Clearly, however, the implication from Jennifer's work is that relationships and the relational merit an analytical distinction apart from resources. Should we be treating the relational as an analytically distinct field of analysis? In order to discuss this I, first, describe a piece of research which subsumes relationships into a more generic understanding of resources drawn from both the personal and the local context, and second, apply the idea of a distinct relational field to some of the concepts discussed earlier.

How far do we treat the relational as analytically distinct?

(i) relationships as part of resources

A good example of empirical research which applies Giddens' generic concept of resources, and also relates it social relations, is Jonathan Gabe and Nicki Thorogood's study of tranquilliser use by working-class white and African-Caribbean women (Gabe and Thorogood, 1986).

Their data were drawn from two separate but related studies in Hackney: one on the use of benzodiazepines by middle-aged white women; the other on the relationships between health beliefs, health behaviour and the structural position of African-Caribbean women. They focus on data drawn from life history interviews with 45 white working-class women (of whom 32 used tranquillisers) and 15 working-class Caribbean-born women (of whom 9 used tranquillisers) all aged between 40 and 60. The resources which the women themselves said they used in the management of their daily lives were, besides benzodiazepines, housing, families, relationships and social support, leisure activities, cigarettes, alcohol and religion. Overall, they found that the white women used benzodiazepines more often and for longer than the Caribbean women and were more likely to see them as enabling. They explain this in a number of ways. First, in terms of the other resources available: for historical, cultural and economic reasons Caribbean women were more likely to have a full-time job, and they viewed this very positively; they were also more likely to have children living at home and find their daughters particularly supportive. In addition, Caribbean women were much more likely to go to Church and regard that as a very supportive resource. The authors suggest for Caribbean women their lesser use of the health services may be because they are experienced as racist or subordinating and that this may also reduce their use of drugs as a resource. The white women had far less access to such resources and/or those resources were not seen as sufficiently enabling to do without tranquillisers. They were less likely to have a full-time job, and more likely to have ambivalent feelings about their work. Also, long-term use was associated with being divorced and not having children at home. All in all, the access to, experience of and meaning of paid work, children, partners and leisure were very influential in differential patterns of tranquilliser use.

Gabe and Thorogood's research framework uses social categories not as methodological variables but in two different ways: first, as signifying subjective experience which is imbued with particular meanings, and, second, as being subject positions shaped by the social relations of class, gender, race and ethnicity which, in turn, find their structural expression through the institutions of the labour market, family, health services and of culture and religion. In addition, they see these subject positions not as discrete but as mutually interlocked and constitutive of subjective experience. In addition, the institutions of family, health service, labour market and religion are investigated in terms of the meanings attached to them by the subjects. In this way, we begin to see more clearly the process of the negotiation and the management of risks through differently available opportunities and resources. Thus, Caribbean women may attach particular negative meanings to a resource such as the health services in a way that white women do not, but their cultural identities and practices may better place them to view their work, family support and possible religious involvement more positively than some white women. The approach enables us to recognise both groups of women as creative agents operating within a complexly structured local context of risks and opportunities, in which their welfare outcomes are negotiated through their access to, and management of, available 'resources'.

This kind of approach, which identifies resources, risks and opportunities and social relations of the immediate context as the structuring forces shaping subjectivity, identity and agency, also creates possibilities to weave in quantitative analysis of risks, opportunities and resources (as Duncan's Workshop 4 paper on Localities does - I think!). It is clear too here that some of the concepts which operate in analytical field 3 also apply in the immediate / local context - for example, localised social relations especially of class, ethnicity and religion, or particular discursive practices (say, around gender and employment, as Simon's work on lone mothers does). In relation to the context of moral subjects and care subjects, not only will familial habits and conventions be important but also the nature and structure and practices of local care provision and welfare and legal services(4). These are all pathways into Field 3.

(ii) The relational as distinct

In the research example above, relationships are subsumed under a more generic concept of resources. However, Jennifer Mason's paper on relationalism (Workshop 3) mounts an important argument for drawing out the relational as a more distinctive area(5). Jennifer argues two things in her work on inheritance (Mason, 1999):

(i) 'ordinary people' see morality as central to their lives - in the sense that [they] are concerned with the morality of their actions … and their thoughts (1999, p2)

(ii) that (as argued earlier) moral agency is "not derived from abstract ethical principles which are either generally agreed upon or imposed by 'ethical experts' but have to be worked out in concrete situations & … most crucially in relationships with others ". (ibid.)

She goes on to argue, drawing upon her work on moving home that the decisions involved in this process imply a high degree of reflexivity (Mason, 2000, Workshop 3). However, the concepts of reflexivity available to us (Beck, Giddens, etc.) suggest that reflexivity is based upon the construction of individual selves and individual identities. By contrast, her research findings demonstrate that processes of reflecting, reasoning and deciding were accomplished in relational ways. As I said earlier, this is important because it begins to spell out what is meant by agency, but also to attribute to it a different meaning. In the work of Beck, Giddens, Bauman etc., agency (and therefore reflexivity) are understood in highly autonomous terms - pursuit of mastery, taking control, exercising autonomy, choice and so on, whereas in Jennifer Mason's work, these processes are understood as being accomplished through negotiation with others. (We could extend the argument to Giddens and New Labour's construction of a valid (positive) welfare subject as one who pursues their autonomy through paid work - see Williams,1999).

How does this argument for relational agency relate to the concepts of subjectivity and identity raised earlier? In Wendy Hollway's work, the notion of a relational subjectivity is central to her concept of moral intersubjectivity. She argues that one needs, in order to understand subjectivity, to have a theoretical perspective oriented towards understanding

(i) of the influence of dynamic unconscious processes in the construction of meaning and reality for people which involves the processing of past experiences through the unconscious.

(ii) that this, together with their life experiences, influences a person's capacity for moral agency.

She therefore combines the psychoanalytic with the social. The social includes the biographical and relational (not just situational and structural). The psychoanalytic depends upon the Kleinian idea that morality is born out of an infant's growing capacity to identify with others and to feel connected to their fate. In other words, moral activity develops relationally, in its conscious and unconscious dimensions.

What, then, of identity? Of course, both subjectivity and identity are relational in the sense that they reflect and refract social relations. But is there a distinct relational identity which is constituted in different ways from the ontological and categorical? Is there a (moral) identity which is created and revised through our close relationships with others through which we have a particular "sense of belonging"? One could imagine that one's moral reputation as "a good mother" in one's family could be achieved quite at variance to one's standing as a worker or a woman. Similarly, one could be a "good" professional, but a "bad" husband. What is interesting for us is where these social and relational identities create dilemmas - how does one enact good mothering with being a good colleague / employee; a "good" daughter with a "good" mother and a "good" lover etc.

What is important about spelling out these different modes and sites of identity (ontological, relational and categorical) and agency (individual, relational, collective) is that this way of thinking breaks down the dichotomy of public and private spheres which, I feel, has outrun its use. Here, instead, we have a much more complex set of interactions between dimensions of the subject and different contextual fields. In these terms, I think that when Finch and Mason in their work on inheritance (1994) spell out their levels of analysis between agency and structure in terms of 4 levels of analysis based upon a public/private dichotomy, they are actually doing their ground-breaking work a disservice. They propose 4 levels upon which family responsibilities can be analysed (1994, p98)

Level 1 public definition (a) social norms
Level 2 public definition (b) legal rules
Level 3 private definition (a) what I actually do
Level 4 private definition (b) what is appropriate in my family

Although they argue that the four levels in reality overlap and interrelate, they argue for analytical separation. However, the use of the public / private dichotomy suggests too great a fundamental separation in my view and limits the opportunity to think dynamically about the interrelations of different fields of analysis (for example, the place of the local meshes the private and public and disappears from view in this distinction). (I also suggest that Level 3 and Level 4 could be reinterpreted as relational moral agency and relational moral identity, respectively).

Even if we were to argue for the relational as a distinct sphere or field, it should not be seen as a separate field of analysis - it is important to keep in play the relationship between identities and practices constructed in relational settings and those constituted through one's subject positions. A good example of this sort of analysis can be found in Tracey Reynolds' empirical and qualitative study of Caribbean mothering practices (Reynolds, 1998, 1999). She demonstrates that mothering practices are informed by:

(i) An understanding of what it means to grow up black in a relatively hostile white world which involves teaching one's child how to negotiate this.

(ii) A "mythical" construction of a Caribbean female ethnic identity and associated practices which provide clear normative guidelines for mothering, employment and parenting practices.

(iii) That the "proper way to behave" that one teaches one's children is informed not simply by a view of "not letting the family down", but also "not letting the (black) community down".

In other words, notions of relational identity and agency are imbued with an understanding of a racialised and subordinate subject position and with ways of resisting or resignifying this. I would suggest, therefore, that the field of the relational can be seen as a distinct but not separate aspect of subjectivity, identity and agency, and as a distinct but not separate field of analysis.

Finch and Mason's identification of "the public" above is represented in this paper by the third field of analysis.

Field of Analysis 3: National / International / Subnational Discursive, Institutional and Relational Contexts

Our work on parenting and partnering practices will need to contextualise itself within the discourses, practices and institutional arrangements that help construct moral and care subjects. These will operate around health, welfare, social work, education and family law, as well as the media, political and sub-political and cultural groups. This is a very large field with many important debates but for reasons of space I discuss here just one of the key concepts which is central to our research questions and paradigm, and that is social relations. (The other key area which frames our research is that of welfare reforms and a central issue here for development is, I think, the notion of governance and associatedly, ideas such as ethico-politics(6)). Picking up from the discussion of 'the subject' earlier, the general social relations which underpin our research are those of gender, 'race'/ethnicity, class, sexuality, disability(?), age and generational relations. More specifically, we are interested in the way in which these reconstitute and intersect with moral, welfare and care relations.

To place our subjects within so many social relations allows us to insist upon the significance of this multiplicity, but it also demands that we clarify the specificitites of these different modalities, and also their particular salience for our research. I suggested, in a discussion of Sasha Roseneil's paper (workshop 2), that the art of juggling involved the following:

1. That we have to clarify the historical and cultural specificities of particular forms of social relations; that we cannot assume parallels between forms; nor can we, in the first instance, assume the privileging of any modality of difference over another.

2. Having established both their centrality (in different ways at different times) to the social formation, and their specificities, we need to consider their variability in social and cultural salience, and their particular salience to the issues we are researching.

3. However, one of the main forces that gives rise to their variable salience is the fact that they rarely operate in isolation from each other. As axes of difference they continually reconstitute each other, compounding or mollifying particular inflections of power. As individuals we experience this through the complex interweaving of our multiple identities and subject positions.

4. Furthermore, as Sasha'a paper pointed out in relation to sexuality, that which is historically interesting for the present is the unfixing of the binaries which have held social relations in place, and more than that, the construction of hybrid identities.

5. For our Cava research, the key thing is to be aware at what points in our studies of parenting and partnering practices are these different and intersecting social relations salient for the construction and living out of moral/care identities, subjectitivities, subject positions and agency. I would suggest that we are interested in (a) the differences that changing ethnic/ gender/ class etc. etc. relations bring to identities, subjectitivities, subject positions and agency (usually termed as 'diversity') but also (b) we are interested in the ordering of these relations as relations of power - for example, in relation to ethnicity and sexuality - how far does the recognition of the diversity that ethnic and sexual minorities bring still position those groups as racialised or sexualised 'others' (will our empirical work evince a picture of this - what sort of othering exists and how is it changing, are, for example, minority family practices still understood as pathological and how does this affect parenting and partnering practices?) and (c) do changing relations pose a challenge to the general cultural framing of what is understood as e.g. 'family' or 'nation', and how do we explore this? It's here that a proposition of hybridity is important - we could, for example, be interested in the ways in which the existence of black and white or straight and gay members of the same family network reconstitute the meaning of what that family is and does (and develop a notion of 'queer family practices').

This discussion above is predicated upon the importance of social and cultural change as a central issue in the research. It is also central to the different frameworks that members of the research group use in their work (e.g. Fiona William's Family-Nation-Work, SD's genderfare and SI's moral economy). It is also the area enclosed by the fourth field of analysis.

Field of Analysis 4: The Dynamics of Social Change

In Figure 1 I have listed the theorisations of change that appear in our different papers : notions derived from the sociological accounts of post-traditionalism, such as individualisation and reflexivity, emerge most frequently as the intellectual foil for our more relational, gendered and care-oriented accounts (FW, CS &BN, JM); also post-Fordism and globalization act as another more materially based foil (see FW). Ideas drawn from cultural theory, such as pluralization and hybridization are also evident (mainly SR but also FW). A critical engagement with all of these seems to be an important contribution of our work. But there is another aspect of social change which lies implicit in our work and that is : with whom or what do the significant processes of change lie in relation to morality? I raise this only as a point to think about. In a recent paper, Perri 6 distinguishes, in Durkheimian fashion, between three kinds of cultural change mechanism (6P,2000) . The first is climatic, where each individual is directly exposed to the same phenomenon (say, of exhaustion, disappointment, loyalty), the second is indirect and random, where a few individuals are affected by a phenomenon (such as stigma, persuasion, adaptation), but they then go on to influence others, but not deliberately. This he calls epidemiological. Third is apostolic where leaders or opinion formers (such as social movements) experience a phenomenon and then deliberately seek to influence others. Strand 3 seems to be dealing with the first two and Strand 5 with the last.


Figures 1 summarises this paper. It also identifies two other areas for discussion in the bottom left and right hand corners, both of which will be pursued in papers (by FW and CS) for Workshop 5. They concern the construction of principles and how far we can develop a relational understanding of morality and moral motivations. On the first point, I mentioned earlier that CS and BN in Family Fragments use people's moral practices around divorce as the basis from which to begin to construct principles, whilst FW uses welfare groups' claims to construct seven recognition principles for welfare in Good-Enough Principles for Welfare.

In a sense, this paper attempts to bring together our commonalties whilst acknowledging our differences, as researchers By way of finishing, then, I tentatively draw from this discussion a set of characteristics which hold us together and through which we can claim our research approach to be distinctive. These are not complete or fixed.

  • The excavation of the relational in concepts of agency, subjectivity, and identity, and in the theorisation of moral motivations.
  • A critique of the rational, unitary subject in favour of the reasoning, fluid, multiply- positioned and active subject
  • The connecting of relational moral agency to wider concepts of social relations, discourses, guidelines and principles, i.e. a more complex and detailed attempt to relate agency to structure /context and to social change.
  • The overall attempt to understand issues of morality and of social relations as dynamic and fluid, as being situational and subject to cultural and historic specificity, yet not as simply relative. That there are possibilities for drawing together commonalties in values which represent, in a particular time and place, a shared core.
  • A critical but constructive reading of post-traditional sociological approaches, and of New Labour and the Third Way, especially in terms of their partial-sightedness to issues of care, interdependence and equal moral worth in the morally self-regulating individual.

Figure 1


Discourse- see Footnote 3 and also Field of Analysis 3 on Figure 1

There seem to be two / three main critical questions asked about discourse: (1) what makes it any different from ideology? (2) Isn't it overdeterministic? And (3) isn't it ignoring of people's agency to construe them as subjects constituted through discourse?
Very briefly: Argument 1. The problem with Marxism and those theorists who used the historical materialism of Marxism as a base to examine culture and ideology is that it was very difficult to avoid, firstly, departing the material from the ideological and, secondly, privileging the material as the crucible in which the ideological was formed (and it never really could get to grips with sociobiology, the body and so on). Foucauldian theories which linked more clearly and complexly discourses to practices and subjects provided more scope for seeing the ideological and material as complexly intertwined (see Michelle Barratt's Politics of Truth, 1992). They also provided better conceptual tools for dealing with essentialist theories and ideas which rested on assumptions of pre-social beings. To be able to talk of people as welfare subjects and not as categories of needs-bearing individuals (lone parents, elderly, disabled etc.) makes for a much more reflexive approach to social policy.

Arguments 2 and 3: I would agree that much Foucauldian analysis is overdeterministic, simply replacing a crude economic / material determinism with a crude, top-heavy discursive constructionism. I would also agree that discourse does not encompass everything that structures our social world - there are non-discursive factors, too. However, we can attempt to correct this overdeterminism by developing the ideas of agency and identity. It is not insignificant that the areas of study where these concepts of agency and identity have been most developed are within feminist research, work on 'race' and ethnicity and on sexuality (i.e. in those areas where a notion of resistance resides). In other words, it is precisely through developing the concepts and ideas associated with 'creative human agency' and the associated capacity to resist, resignify, reconstitute and reproduce, that the best corrective to an overdetermining approach lies. So, on the one hand, 'discourse' is a term which signifies a more nuanced approach to the forces structuring people's lives, but. on the other, it cries out for an elaboration of 'lower order' concepts through which individuals perform their identities and act up/ out their agency. (Another related methodological issue I do not deal with is the significance of discourse in narrative)


Baldock. J. & Ungerson, C. (1994) Becoming Consumers of Community Care York, J.R.F.

Brah, A. (1996) Cartographies of Diaspora, London, Routledge

Duncan, 1999a 'Challenging Rational Action Theory' Cava Strands 1&2

Duncan, S (2000) Comparative research: The case of genderfare, Cava Strands 1&2

Duncan, S (2000) Localities Papers I, II, III. Cava Strands 1&2, University of Leeds.

Finch, J. (1989) Family Obligations and Social Change, Polity Press, Cambridge.

Finch J and J Mason (1994) 'Family Responsibilities and Inheritance in Great Britain' in M. Maclean and J. Kurckewski (eds. ) Families, Politics and the Law: Perspectives for East and West Europe, Clarendon Press, Oxford.

Gabe, J and N. Thorogood (1986) ' 'Tranquillisers as a Resource' in J. Gabe and P. Williams (eds) Tranquillisers: Social, Psychological and Clinical Perspectives, London, Tavistock.

Giddens, A.(1979) Central Problems in Social Theory, London, Macmillan.

Hall, S.(1996) 'Who needs 'identity'?', in S. Hall and P. du Gay (eds) Questions of Cultural Identity, London, Sage.

Hollway, W. Moral Intersubjectivity, Methodology and Empirical Research' Cava Strands 1&2, paper (Workshop 3)

Irwin, S. (2000) Conceptualising Social Change: Family, Work and the Changing Pattern of Social Reproduction Cava Strands 1&2,

Irwin, S.(2000) Patterns of Change in family and Structure and Household resourcing, Cava Strands 1&2,

Mason, J.(1999) 'Inheriting Money:kinship and practical ethics Centre for Research on feamily Kinship and Childhood, Working Paper 13

Mason, J (2000) 'Deciding where to live : towards a reflexive relationism' Cava Strands 1&2,

Mason, J (2000) 'Researching Morality' Cava Strands 1&2,

Perri 6 (2000) 'Shaping Citizens : responsibility and the politics of moral character' Paper for the seminar on 'Ethics Affect and citizenship' UEL, Birkbeck College, 17-18.2.2000

Reynolds, T. (1998) Afro-Caribbean Mothering: Re-constructing a 'New' Identity, PhD Thesis, South Bank University. Unpublished.

Rose, N. (1999) Powers of Freedom, Cambridge, CUP.

Sevenhuijsen, S. (1998) Citizenship and the Ethics of Care: feminist considerations about justice, morality and politics London, Routledge

Smart, C. & Neale, B. (1998) Family Fragments? Cambridge, Polity Press.

Taylor, D (1998) 'Social Identity and Socail Policy' in Journal of Social Policy, 27, 3 pp. 329-350

Williams, F. (1996) 'Feminism, postmodernism and the question of difference', in N. Parton (ed) Social Work, Social Theory and Social Change, London, Routledge.

Williams, F. (1999a) 'Understanding the Dynamics of social change, social relations and welfare settlements: towards an analytical framework' Cava Strands 1&2

Williams, F. (1999b) 'Good-Enough Principles for Welfare' Journal of Social Policy, 28,4,pp.667-687.

Williams, F. Popay , J. and Oakley, A. (1999) Welfare Research: A Critical Review London, UCL Press.

Williams, F. and J. Popay (1999) 'Balancing polarities: developing a new framework for welfare research', in F.Williams, J. Popay and A. Oakley op. cit.


(1) This was the last major ESRC Programme in Social Policy until CAVA, and was bitterly fought over as many felt that, after 11 years of Thatcherism, poverty and social polarisation were far more pressing issues for research. Back

(2) David Morgan commented, at the Family Centre ESRC Seminar, that it was not necessary to go to a psychoanalytic frame to develop the notion of intersubjectivity - one could revisit Goffman and the interactionists (3-3-2000). Back

(3) I clarify the use of the term discourse in appendix 1, since it is caught up in the ways I have been using identity, subject and subject position (and because its mention in the Workshops seems to invoke a sharp intake of breath). Back

(4) I think the notions of time / timetable / life course should come into this field of analysis, but I'm not sure how. Back

(5) In some ways, Jennifer's 'relational' is similar to the concept of the 'informal' developed in the 1970s by Abrams et al. Back

(6) For example, Nikolas Rose's concept of ethico-power (Rose, 1999) would seem to be an important idea for analysing current policies around parenting and partnering and the importance of our 'grass-roots' approach. Back