ESRC RESEARCH GROUP ON CARE, VALUES AND THE FUTURE OF WELFARE

University of Leeds

Workshop Paper No 14
Prepared for Workshop Four
Methodology for researching moral agencies
Friday 17 March 2000

Jennifer Mason

RESEARCHING MORALITY

Introduction

I have changed the title and focus of this paper (my original brief was 'Exploring moral obligations in research interviews') for two reasons, and I hope you will forgive this indulgence. First, I have done this because I want to ask a wider range of methodological questions about researching morality than that raised by a discussion of research interviews alone. Second, because I do not want to tie the discussion into the concept of obligation, since this is only one specific way of conceptualising morality and again I want to think a bit more widely. I do also have some reservations about the concept of obligation itself - in our work on kinship and responsibilities, Janet Finch and I moved away from that concept towards the more active notion of negotiated commitments because that seemed better to express the active (although not unconstrained) ways in which responsibilities to kin were worked out by our interviewees (Finch and Mason, 1993).

The paper remains a methodological one, however. I shall use it to raise questions about how we might go about researching morality empirically. The ideas I set out, however, are emphatically not to be seen as a blueprint or as a conceptual framework. Instead, I am presenting a range of possibilities, which is designed to leave doors open, not to close them. In the spirit of grounded theory, or at least my version of it, I think it is vital that our methodologies are conceptually and theoretically open so that we do not close off or fail to see or listen to the fullest range of significant possibilities our data might be able to raise. But I also feel strongly that we do not, cannot and should not approach empirical research through a theoretical or conceptual vacuum (Mason, 1996a). This paper is an attempt first to identify some conceptual doors onto morality (whilst acknowledging there will almost certainly be others I have missed), and then to leave them open!

I shall do this by asking some questions which problematise the ontology and epistemology of morality but in rather practical ways. These are questions about where morality is located, and therefore where and how should we look for it in our research?

1. Locating morality - methodological issues

The question here is really about where we might spot or glimpse morality, methodologically speaking, and below I suggest some possibilities. I want to emphasise again here that these are not meant to constitute a conceptual framework. Other than recognising that the different 'locations' I discuss are overlapping and connected, I want to make no other claims about how or whether they fit together into one picture or model. I think we need to leave those questions open at this stage, whilst at the same time making sure that our methodologies allow us to generate data which will get us closer to answering them. In other words, I want to advocate a data driven approach to explanation and theorising, whilst recognising that we cannot be (should not pretend be) theoretically neutral in our empirical work. Thinking broadly about 'locations' of morality, and sharing and discussing these, is my attempt to keep these various balls in the air.

I shall not therefore pretend that my list of 'locations' has no theoretical orientation. There is one, which underpins my whole discussion, and I think this is an orientation, which we share in CAVA. This sees morality as situated not abstract, as practised, as fluid and of course as contested. If you simply expect moral views to be held decontextually by individuals, and acted upon consistently in every situation, then researching morality is straightforward: you simply ask people what is right and what is wrong in a large survey, using a representative sample, and you have your answers. Unsurprisingly I favour an approach to researching morality which does not expect it to be static, unchanging and consensual (Mason, forthcoming). Attitude studies, which try to seek people's 'real' or 'true' moral views (untainted by circumstance and context) do carry that kind of expectation, and are flawed as a consequence. We have plenty of evidence that morality does not take this static form (Smart and Neale, 1999; Finch and Mason, 1993, 2000; Duncan and Edwards, 1999). But how do we investigate morality, in its complexity and its different 'locations'? Here are some ideas, 'location' by 'location'.

i) Moral practices

Many of us have argued that morality - and especially moral agency - is located in practices. Practices are not separate from reasoning and reflexivity, and are therefore not to be confused with a concept of 'behaviour' which sees it as separate from 'attitudes'.

The methodological implications are that we want to find out about people's experiences and their practices, so that we can see how they activate morality in their own lives, rather than asking them about what is right and wrong in the abstract. The most commonly used method to do this has been qualitative interviews which focus on people's experiences in selected areas. It is difficult simply to 'observe' practices (because they are not just behaviour, and also because the contexts in which they are 'done' are not fully available to a researcher), but there may be possibilities here (see especially collective practices below). It is difficult to investigate practices in quantitative surveys, because of the impossibility of exploring context through structured means. It may be possible to explore practices, or at least elements of them, through documentary research, and certainly some elements of institutional practices can be investigated through these means. In our study of Inheritance, Janet Finch and I and our colleagues examined a sample of 800 probated wills to try to understand something about will-making practices. Using SPSSX we were able to produce an overview of patterns of bequeathing in our sample (limited, of course, by the nature of the sample, including that only around one-third of people make wills, not all of these are probated, etc). However, we supplemented this broad but not too deep and not at all processual picture of will-making practice with interviews with will-making professionals, and with qualitative interviews with people about their experience of inheritance more generally (Finch et al, 1996; Finch and Mason, 2000).

ii) Moral thinking, moral reasoning, moral feeling, and non-verbal expressions of morality.

I have written about reasoning, thinking and feeling elsewhere where I argued that we should be interested in the sentient activities and the active sensibilities involved in relationships between kin, and that we should use ideas like this to avoid the unhelpful dichotomy of care as labour or care as a state of feeling (Mason, 1996b). As ever, looking back I am not sure I would conceptualise the issues in the same way now, and debates about these issues have certainly moved on rapidly since I wrote that chapter. But I think there is a domain in which morality might operate which is to do with thinking, feeling, reasoning (not necessarily in rational ways), emotion, mood, imagination, fears, sentiment, memory. These have often tended (rather simplistically) to be seen as 'inside' individuals in the sense that they are not articulated verbally, and they become something different if and when they are articulated verbally. But they are not necessarily individualist, or necessarily 'inside' in the sense of being separate from practices; they can be collective as in the example of family moral 'secrets' which 'everybody' knows and perpetuates, but no-one articulates.

This location - which is unexpressed in words - may be rather significant in the study of morality (Mason, forthcoming). It has had the least empirical attention, from sociologists at least. Perhaps the closest sociology has got is to recognise that we can discern 'taken-for-granted' but unspoken assumptions in people's practices and in their words; that we can listen to what people do not say as well as to what they do say; and that we can make inferences about how people feel on the basis of what they say or how they behave. Or (as I discuss in Mason, 1996) that we can study emotional labour, and the social construction of emotion, neither of which quite get the point. The 'discursive turn’ in sociology, and the decentring of the creative subject, rules this location out of the frame altogether, and our understandings of morality may be impoverished as a consequence.

Researching the unspoken is inherently problematic, and we may need to find ways of encouraging non-verbal forms of communication to explore dimensions which people find difficult to express in words (Mason, forthcoming). These kinds of methods are probably the most advanced in sociology in research with children, where play and drawing in particular are now commonly used, although we have some way to go in developing our expertise in analysing the products and working out what they represent or what they tell us. Children, in particular, (if they are computer literate) may respond to the use of interactive technology, computer graphics and games and so on, to express the non-verbal. But why just use such methods with children? It is somewhat patronising to assume that they are unable to use language to express themselves if we are simultaneously assuming that adults can do this. It is better to think through what dimensions of social experience we can explore in these ways - with both children and adults.

Those of us who analyse kin relationships have used various ways of mapping and charting relationships (such genograms/family trees, concentric circles), and discussions with people about why they draw it or picture it in that way can tell us quite a lot about morality (for example, who should be included, who is excluded because they have 'sinned'). The point is that neither the picture/chart or whatever, nor the account, tells the whole story.

Other ideas for exploring the non-verbal include:

• observing non-verbal elements in, for example, qualitative interviews

• we can ask people to talk about feelings, memories and the like (contextually, not in the abstract), although we should recognise that expressing them verbally turns them into something else.

• We can sometimes discern whether someone is used to thinking in a particular way by what they say, and how emotionally engaged they are with a particular issue.

• We can use personal photographs, objects and possessions as starting points for discussion, as well as for analysis in their own right (eg interesting work on analysing 'family photographs'; eg 'spatial' household studies, looking at where technology is located in households, looking at use of space, where certain things get done; eg objects and possessions - in our study of inheritance, we looked at (literally, very often) objects people had inherited - what kinds of objects they were, where they kept them, discussed them whilst holding them, looking at them, admiring them etc. It emerged that it is these kinds of inheritance, and the symbolic practices associated with them, with which people get most morally and emotionally engaged).

• We can get a sense, very often, of what matters to people in conversations with them, and in observing their practices, their contexts, and this may be quite important for understanding morality.

iii) Moral identities and reputations

Morality may be 'located' in or practices through moral identities and reputations - a good mother, a bad father, a wayward child, a good son, a caring person. These are likely, of course, to be shifting, multiple, situated, individual, relational and collective (Williams and Popay, 1999).

Identities may be unspoken, spoken, practised (individually or collectively or relationally), and so on, and the key methodological issue (other than those discussed in (i) and (ii) above, is probably to determine the likely and appropriate range of contexts in which they might be expressed, or in relation to which they might be constructed. Simply talking to individuals (eg in qualitative interviews) is unlikely to be enough to get at collective and relational expressions/forms of identity. For example, in our Family Obligations research, Janet Finch and I learned a lot about people's family 'reputations' by talking to their kin (Finch and Mason, 1993). But see also collective practices, below.

iv) Morality in relationships

Morality is likely to operate in relationships between people. This includes how people think, feel and reason about others and themselves in their daily lives, in different contexts as well as how they interact with or in relation to others (Mason, 2000).

We can use a range of methods as already discussed to focus on relationships, and we can get the perspectives of more than one party to a relationship. We can 'read' our data for relationships rather than just individuals (see Mauthner and Doucet, 1999, for an interesting discussion of reading data for relationships).

v) Collective moral practices, morality in collective settings and localities

I am referring here to settings where morality gets done or practised collectively, and which may indeed be organised around a particular moral agenda or framework. These might range from playgroups, NCT classes, self-help groups, social movements; to ritualistic or ceremonial collectivities of families and friends at for example birthdays, weddings, funerals, Christmas, Ramadan, Diwali and other festivals; to 'informal' collectivities such as children at play, families/others out shopping or travelling, on holiday, watching TV, cooking, eating. Some collective practices may not be spatially bounded, and on the other hand localities may incorporate a range of elements not all of which are to do with collective practices.

If collective practices happen in spatially and temporally bounded locations, then researchers might be able to visit them to research them. However, many may be private, or our gaze may be very partial, or selective in ways, which we cannot control or would not choose. Researchers might be able to engineer them, using focus groups for example which are good for exploring collective reasoning and wisdom, and for seeing what points a group of people can reach; or getting people to interview each other. However, both of these strategies may have problems and limitations for some purposes, in which case researchers might fall back on asking for people's accounts about and experiences of collective practices. Some elements of morality in these 'locations' may be documented.

vi) Moral biographies

Biographies do not exactly constitute a distinct location, but they evoke a sense of time and narrative in and through which moralities are likely to be situated and constructed. Biographies, of course, do not have to be individual. Morality may emerge biographically, or be biographical, or form part of a biographical narrative, so our methods need to allow for this possibility by ensuring they include an historical, development or biographical gaze (in interviews, for example, getting personal biographies and asking questions like 'how did that come about?), and explore the connections between past and present on a range of levels.

vii) Collective/shared understandings/public norms

One example of this is the old idea of moral rules and 'public' norms. Of course much of CAVA is premised on the idea that such things do not exist, or at least do not operate in this way (instead, moral agency, moral practices, moral discourses). In our Family Obligations study, Janet Finch and I included a large survey designed to find out whether there was a 'normative consensus' about the proper thing to do for relatives (Finch and Mason, 1993). Hypothetical situations were given, and the respondents were asked to say what the people in those circumstances should do. The survey recognised that morality might vary situationally (ie by varying the situations in the questions) whilst also trying to see if such a thing as a societal or 'public' normative level exists in the sense that people will say what is right and what is wrong. Then, to see if there is agreement or not in the normative content.

In some ways, this is highly problematic, since there is not much point in knowing whether people are prepared to say 'a daughter should do x for a mother in these particular circumstances', if we don't know what that means in practice, whether and how people use this, how this relates to their own practices and moral reasoning in its context. (Of course our study explored moral and family practices in detail in its qualitative interviews, and we did some work in relating these to what emerged from the survey). But it is rather interesting that people will answer these kinds of questions, and make moral judgements, and I suspect rather enjoy doing so. The problem is to work out what this means - it is certainly an aspect of morality, although I wouldn't now wish to call is 'public' morality because of the artificial public/private separation that implies. For our purposes in that study, the main point was to challenge the orthodoxy that there are moral rules, or that there is a normative consensus (there wasn't!), and we wanted to be able to make broad based general claims about this, hence the quantitative survey and the representative sample. In other words, we wanted to be able to say, in a general way, whether or not there are shared understandings at this level.

Another way to explore shared understandings is to look at what people think they share more qualitatively, and of course more contextually (see Discourses and narratives, below).

viii) Moral discourses, composite moral narratives

This is a gross oversimplification, but it seems to me that the term discourse is commonly used in one of two broad ways. The first concerns 'formal' discourses - moral regulation, expert discourses, welfare and legal discourses, and oppositional discourses, which all emerge through or are constituted in a variety of practices, but which have some kind of formal (often textual) expression. These are traceable in texts, policy documents, policy debate, institutional practices, and the like.

The second is less formal, and is about discourses in people's everyday practices and reasoning. This second is not the same as collective/shared understandings because, although people may operate as though there were a common understanding (eg of what is a 'good mother'), what is actually happening is that people are actively engaging with what they think that might be, constituting it in and through their practices (not just following the agreed upon rules, nor indeed just 'drawing on' a discourse). This is the kind of morality, which many of us investigate, primarily but not exclusively through qualitative interviews. Other methods above may also apply.

The first way of conceptualising discourse is a top-down route. The second is supposedly 'bottom up', although there is a danger that it too becomes 'top down' in studies where people are simply seen as 'drawing on' discourses, and where everything has to be slotted into the discursive framework (ie everything is a product of/constituted through discourse). I do not like this top down version of the bottom up route because it closes doors (just as seeing morality as static moral views does, or seeing it as rules which are followed). I think we need to be able to look for discourses without assuming that is all we are looking for.

In an attempt at a more truly bottom-up version of the bottom-up route, Janet Finch and I have developed the concept of 'composite (moral) narratives'. These are narratives about the moralities of specific types of situation - in this case associated with inheritance and kinship - which we constructed from putting together people's individual accounts, experiences, hopes and fears and so on. Essentially, the narratives are a more general expression of 'what mattered' individually to the people in our study, and they tell us a great deal about how people make connections between their own experiences, fears and aspirations on the one hand, and what they perceive to be more general practices and assumptions on the other (Finch and Mason, 2000; Mason, forthcoming).

Four general points about 'locations'

1. Underlying all of the examples of locations of morality, which I have given, is a concern with appropriate contexts, and these need to be thought through carefully for each of the empirical studies in CAVA. We might see these as localities, we might see them as the substantive concerns and issues we focus on in interviews for example (motherhood, divorce, etc), or areas for policy analysis. We might see them as issues for sampling (where we will wish to move beyond the framework provided by individual characteristics/variables for sampling, as this is simply not applicable in many of the examples above).

2. I have insisted that these locations are not part of a conceptual framework which we 'know in advance'. Instead we must explore, from the vantage point of each location, its connections with other locations and contexts.

3. Also underlying the examples is a fluctuating distinction between where morality is 'located', and where/how you can find out about it, which needs to be examined in each of the empirical studies. (eg can words tell us about thoughts, and so on?)

4. All the usual rigours of doing good research apply. Particularly useful are the principles of analytic induction where we make sure that we not only look where we expect to find what we are looking for, but also where we look for 'negative instances' and 'counter examples'. Also we should think broadly about method - different methods for getting at different versions of morality and so on (Mason, 1996). This is of course particularly important for strengthening the explanatory power of qualitative research.

Conclusion

Overall, the point I am making is that we cannot 'research morality' in the abstract - only moral practices, moral agency and so on in specific contexts (Mason, forthcoming). But how can we be sure that what we see there is morality? Is there a danger that our interest in morality closes our eyes to other elements in the way partnering and parenting are done and constructed? Perhaps pragmatism, instrumentalism, need and serendipity are just as important? I have a number of concluding points to make here:

• First, of course we should not assume that parenting and partnering are always framed as moral issues. We need to look for these other things too, and to be aware that in interviews, for example, people may be keen to frame their narratives in ways which cast them in a good (moral) light.

• Second, however, we should examine places where people, documents, practices, policies and so on treat issues as though they are moral issues. I have proposed that the concept of 'what matters' (and what doesn't) - to people, in policy, in practice, and so on - is a broader way to approach this than the more narrow 'proper thing to do' (Mason, forthcoming).

• We need to explore (rather than assume we understand) points of connection with what is treated as moral in different 'locations', eg where an interviewee talks about issues which we know have points of connection with more formal moral 'discourses'.

References

Duncan, S. and Edwards, R. (1999) Lone Mothers, Paid Work and Gendered Moral Rationalities, Basingstoke: Macmillan.

Finch, J. and Mason, J. (2000) Passing On: Kinship and Inheritance in England, London: Routledge.

Finch, J. and Mason, J. (1993) Negotiating Family Responsibilities, London: Routledge

Finch, J., Mason, J., Masson, J., Hayes, L., and Wallis, L. (1996) Wills, Inheritance and Families, Oxford: Clarendon Press.

Mason, J. (1996a) Qualitative Researching, London: Sage

Mason, J. (1996b) 'Gender, Care and Sensibility in Family and Kin Relationships', in J. Holland and L. Adkins, (eds) Sex, Sensibility and the Gendered Body, Basingstoke: Macmillan.

Mason, J. (2000) Deciding Where to Live: Relational Reasoning and Narratives of the Self, Centre for Research on Family, Kinship and Childhood, University of Leeds, Working Paper No. 19.

Mason, J. (forthcoming) 'Qualitative Interviews: asking, listening and interpreting as theoretical projects' in T. May (ed) Companion to Qualitative Research, London: Sage.

Mauthner, N. and Doucet, A. (1998) ‘Reflections on a voice-centred relational method: analysing maternal and domestic voices’ in J. Ribbens and R. Edwards (eds)

Feminist Dilemmas in Qualitative Research: Public Knowledge and Private Lives, Sage, London.

Smart, C. and Neale, B. (1999) Family Fragments, Cambridge: Polity.

Williams, F. and Popay, J. (1999) 'Balancing Polarities: Developing a New Framework for Welfare Research' in F. Williams, J. Popay and A. Oakley Welfare Research: A Critical Review, London: UCL Press.

BACK