Methods in Dialogue. Hemingford Grey, Cambridgeshire, UK, May 18-20 2005, organised by Molly Andrews, Corinne Squire and Maria Tamboukou (Centre for Narrative Research, University of East London, UK), Phil Cohen (London East Research Institute, University of East London, UK)

Abstract: This paper comprises discussions from a residential symposium, "Methods in Dialogue", that took place near Cambridge, UK, in May 2005. The symposium concluded a series of seminars organised by the London East Research Institute and the Centre for Narrative Research at the University of East London and supported by the Economic and Social Research Council. Public support for social research increasingly depends on its ability to deliver scientifically valid and reliable studies to guide policy and practice. The theoretical foundations of social research, however, seem to be in a critical state. Evidence generated by both qualitative and quantitative methods is more and more seen to be conflicting, open to many interpretations. The aim of the event was to bring together qualitative researchers in the social sciences, many working in the field of narrative but also a number working with life history and auto/biography, discourse analysis, grounded theory methodology, visual methods and ethnography, to discuss the theoretical foundations of qualitative social research. The discussions addressed narrative itself as an index case for methodological debate; methodological considerations of objectivity and evidence, interpretation and context; appropriate levels of research focus and their interactions; the role of dialogue between disciplines; and the interaction between social science and the wider environment of which it is a part. Questions such as the following were discussed throughout the symposium: Who and what is social research for, and whose voices does it represent? What are social researchers' and participants' interpretative rights over their data and each other? How does thick description and the rich social interpretation it affords relate to the need for precise methods of explanation and generalisable conclusions? What special problems of research design or delivery arise when attempts are made to "empower" informants, to enable them to interrogate, and even co-construct the research story?

Key words: qualitative research, narrative research, methodology, objectivity, interdisciplinarity, transdisciplinarity, politics, practice, intersectionality, thick description

3.2 Margaret WETHERELL: Methods for studying multiple identities: Intersectionality, practice and troubled and untroubled subject positions'

4.2 Trisha GREENHALGH: Inter-disciplinarity in the study of rhetoric in healthcare policymaking

This report gives insight into the lively dialogue now taking place around qualitative social science methodologies. It presents discussions from a residential symposium, "Methods in Dialogue", that took place near Cambridge, UK, in May 2005. The symposium concluded a series of seminars organised by the London East Research Institute and the Centre for Narrative Research at the University of East London and supported by the Economic and Social Research Council. The aim of the event was to bring together leading qualitative researchers in the social sciences, many working in the field of narrative, to discuss the theoretical foundations of qualitative social research. [1]

Public support for social research increasingly depends on its ability to deliver scientifically valid and reliable studies to guide policy and practice. The theoretical foundations of social research, however, are in a critical state. Evidence generated by both qualitative and quantitative methods is more and more seen to be conflicting, open to many interpretations, and lacking in objectivity. The event brought into conversation researchers from a wide variety of disciplines: sociology, anthropology, psychology, political theory, linguistics, history, demography and cultural studies; and using a diversity of research approaches: critical theory, ethnography, ethnomethodology, phenomenology, grounded theory methodology, narratology, psychoanalysis, network analysis, action research, policy studies, visual methods, and social survey research. It also included participants working in a variety of research contexts: those working in government departments, local authorities, policy think tanks, research institutes and other public bodies, as well as from the voluntary sector, and from universities. Every discussion was prefaced by the presentation of a paper, and remarks by two discussants. For the purpose of this report, the discussions are presented chronologically. Speakers' and discussants' contributions are given in summary to contextualise the conversations that follow. [2]

The event was organised as a series of workshops, with the aim of creating a stimulating atmosphere where the "dialogic imagination" could take flight. The most novel aspect of this report is, therefore, the conversations, transcribed here in an edited form that has been approved by all the participants. These conversational condensations were produced by focusing on methodical and methodological considerations in general and narrative research in particular. The report thus does not present results from an academic gathering, but rather traces how the event's discourses unfolded through the interplay of individual contributions. It should be emphasised that the views expressed bear the marks of our own editing, selecting, condensing and consequently, "interpretation". In addition, most of the features particular to spoken language have not been transcribed, a sacrifice made to ensure easier reading. [3]

In an attempt to define more closely what "thick description" means, the discussion focused on three examples: a shell introduced by Marilyn STRATHERN, a little boy in a therapeutic situation described by Mike RUSTIN and a photograph of the South African National Women's Monument shown by Liz STANLEY. Thick description automatically includes a multiplicity of perspectives. The question is whether these perspectives are commensurable, whether there is a level behind the differences of the perspectives where they show complementarity. Can one take relationality as the answer to the problem of multi-perspectivity? Which interpretive framework is the "right" one to start the description from? Are we aiming at truth, or truths, and how can we address empirical and theoretical excesses? Can there be an objective perspective, or can we redefine objectivity? What kinds of closures do we have to face in our research and how can we stay aware of them? [4]

Social scientists create all kinds of objects of analysis in the course of their enquiries, but let me start with a found object. Or rather, it is a thing that we wish to make into an object of knowledge. You can both see what it is and not see what it is—you will register the colours, texture, know it is a shell, might have to be told the case is made of bark, just as one might interpret the contours of an ultrasound scan or the colours of a costume or figures on a graph. Yet probably you would say you don't "know" what this is, since clearly I have produced it in order to explain it, and you know there is more to come. John LOCKE similarly gestured towards objects of curiosity. In his Essay Concerning Human Understanding (1690), LOCKE conjures up the image of two cassowaries on display in St. James's Park, London. Cassowaries are large, flightless birds from Papua New Guinea and South East Asia, and to the Londoners of the time they seemed quite enigmatic, not fitting known categories, eluding immediate classification. But they are in his text for a purpose, and we know there is more to come. And we know that one effect of what is to come must be a better description of the items we have just encountered. In other words, the entity is not yet complete as an object of knowledge. [During this presentation, STRATHERN showed a pearl shell in its fabricated bark case, from Mount Hagan, Papua New Guinea, to all participants and allowed it to be passed around.] [5]

There is excess at the heart of the ethnographic enterprise, of which the aim is to study "everything". Any object can then be set in the context of a larger object (society, culture), and can help create the context for understanding other things. This excess bears on the question of "objectivity" in two possible ways. First, piling context upon context can produce the impression that all that is at issue is multiple perspectives, and multiplying perspectives leads one all over the place. However, it is much more usual in research to state a problem, and to decide in advance the nature of the evidence and the different perspectives needed to address it. This is objectivity of a kind: no one perspective dominates. Yet the more perspectives are piled up, the more they appear as mere points of view. Far from objectivity, we end up again with the suspicion of relativism. In the ethnographic context, relativism resides in analytical choice. Up to a point, one perspective is as good as another, though each will yield a different emphasis. [6]

A second possibility is that the ethnographic layering of contexts can offer precision of a kind. Objectivity can be redefined, usefully, as the recognition of the endurance of the characteristics of an object of knowledge, where the object is defined relationally, by a specification of coordinates. Looking at the pearl shell as a single isolated object, or even as an object of exchange, value and ranking, we cannot answer the question of whether we should think of it as a species of money. We would have to bring all those elements together to form one context and then cross it with another source of information—analysis of commodity exchange, the behaviour of markets, analogues of banking and credit systems—or with a further context that deals with the regulation of resource flows, the legalisation of currencies, and so on. If we were to find out that there are conditions under which it is apposite to call the pearl shell a currency that becomes its own context. Under this description, the objects of social research knowledge are "conglomerates". Can we, though, specify with some precision how the elements of the conglomerate relate to each other? [7]

Marilyn STRATHERN's shell is an object that "belongs" in many contexts and frames of reference. "Found objects" like this are a good way of illuminating latent frames, perspectives, and ways of seeing. Let me give you another example to show the necessity of different perspectives on one "object", this time taking a human "object". I heard about a very young child placed for adoption, who seemed quite uninterested in his adoptive parents' care of him. The parents and the child met with a psychotherapist and a social worker. The child careered around the room, knocking things over, screaming, apparently not relating to anyone, least of all his adoptive parents. The therapist had some toys. She started a game and the little boy became interested. He joined in. The therapist thought, "here is something to start from". More theoretically, "here is a capacity to symbolise". For a few moments, this apparently feral, unsocialised child, became a normal child, in both the therapist's and the mother's eyes. The parents' terror of diagnoses of brain damage, infantile psychosis, or autism was somewhat relieved. What was the child psychotherapist's method here? How did she think of her "object"? The child psychotherapist's method was to create the space for a relationship in which the child would feel "held in mind" and could find ways of expressing his feelings, desires, and terrors, so that these anxieties could be understood and given some symbolic expression. One can see how contentious the categorisation of something like this, a child and his family referred to a mental health clinic, can be. Many possibilities do have to be held in mind. In the instance mentioned, autism and brain damage were possibilities. Whatever confidence one may have in the diagnostic category of ADHD, there are child patients who may be helped by drug treatments, and certainly there are patients who benefit from Cognitive Behaviour Therapy. Given these differences of perspective, one might ask where "objectivity" lies in this situation. There is some. All the schools of mental health professionals work in the same setting, with the same manifest purpose. There will normally be a lot of agreement on what it means for a patient to get better. Since there is a lot of practical consensus about "ends", there can often be an acceptable "division of labour", and de facto agreements about what treatment modes, what perspectives, might work best for which kind of patient. But there are also deep-seated disagreements between professional perspectives, and there is no "objective" or "consensual" perspective on the horizon which is going to resolve all these differences. They depend at root on values, conceptions of how lives should be lived, even of how institutions and societies should be organised. Such differences can be creative. The most interesting kinds of human science have emerged from commitments to values. I don't think we should worry overmuch about the continued existence of such differences, the fact that we see "objects" through so many different lenses. Rather we should see the human sciences, like the parallel discourses of the humanities, as ways of elaborating and clarifying differences, of extending our understanding of possible worlds and possible ways of life (full paper version). [8]

My comments on Marilyn STRATHERN's paper are from a sociologist who lives abroad, who studies the past; particularly "race" and racism in South Africa. I shall add to the variety of methodological ideas she presented. There is plenty of so-called "thick description" and objectivity of a triangulated, indeed of a strangulated, kind in the history of sociology. I have drawn my (methodological, but also ethical and political) distance from in my South African research. Work of this kind is the mainstream of research on the "origins" of segregation and apartheid and it has repeatedly re-discovered more of the same: the "complete break with the past, external cause of racism" theory of the origins of apartheid in South Africa, and de facto (if not quite de jure) it has supported the "history" produced by Afrikaner nationalism of the South African past. It is, however, founded on implicit theoretical and methodological ideas about periodisation and causality which are, intellectually, ethically, politically, and in all other ways, suspect; it also has strong teleological features. By focusing on complex networks at work in South African political life during and in the long aftermath of the South African War (1899-1902), rather than "the state", I have tried hard not to fall into the trap of reifying "the state" or other social structures, to retain a strong sense of social and political life and its complexities, as well as of the construction of facts, histories and structures which persist—albeit not unchanged—over time. What I have done may well be problematic in other ways (all research is when reviewed by people with different methodological and intellectual agendas), but at least not in these. From my viewpoint, a sociology worth the candle must provide an analysis in which time and its passing is at the centre; which recognises the importance of social structure but also of changing social and political contexts and the events occurring in this and at the same time does not reduce persons acting in social life to simple—or indeed complex—reifications. Nothing else will do; and while bringing off the whole thing is extremely difficult, it still has to be attempted. Methodology is never just methodology; it always raises all the other -ologies and -isms; and it has to be taken seriously, first and last. A photograph called "Collected by the author ..." (shown by STANLEY and passed around) acts as a parallel, an awkward and indeed ironic parallel, to one of the "objects" referred to by Marilyn STRATHERN. This photograph is composed of a large array of objects gathered together on a table; these were purportedly collected by the author of a lengthy Boer woman's testimony about her "derring-do" activities during the 1899-1902 South African War and supposedly referred to in her circa 1940 published text. However, in fact, most of them are not referred to, although one in particular is—the cap badge of an African prison warder shot and killed by the author, Hendrina RABIE-VAN DER MERWE, when helping spring her husband Sarel from prison. A woman not to be crossed, then, and also one whose book was less a testimony of the past than a testament of the strongly Afrikaner nationalist present of 1940 in South Africa and its gospel of separation, binary racial hierarchy, an essential African inferiority, and apartheid. This photograph is, however, the cassowary of my commentary, in spite of my perhaps misleading earlier reference to its fetishistic qualities. And what is actually akin to Marilyn STRATHERN's pearl shell, a fetish object with excess built into every morsel, is what is shown in the other photograph (shown by STANLEY and passed around) of the Vrouemonument, the national women's monument. When people from Mount Hagen unwrap and look at the pearl shell they see a foetus and life; when people look at the Vrouemonument through the lens of "the history" and "the facts", the objective facts according to nationalism, they see death across the land. [9]

Mark FREEMAN: I would like to cast the conclusion of the first two papers into question. It seemed to me that Marilyn STRATHERN's paper argued for limited commensurability; she didn't go completely for the idea that these are irreducibly incommensurable perspectives. Finally, she said, there is the possibility of within-context trumping. Her perspective is a pragmatic one. Mike RUSTIN's commentary argued for an incommensurability based on the idea that there were ultimately disparate frames founded on ultimately disparate moral commitments and even perhaps ontologies. But I think about these in a different way. The discussion of non-human objects, such as shells, leans more to the incommensurability thesis than the commensurability one. If one asks "what is this shell?" there is no way to determine that, apart from the question one asks. Is it an object of exchange? Is it an aesthetic object? What is it? Well, one can only determine that relationally as a function of the question one brings to it. And so I wondered in that context whether it was possible actually to move in the direction of limited commensurability or whether there was a kind of basic irreducibility there. Mike RUSTIN moved in another direction when he said it was different when we are talking about the human realm. What is at stake in RUSTIN's account? The child. Whether the child is being constituted as an object of science, whether the child is being constituted as somebody with a thinking problem, or whether the child is constituted as somebody who is developing, who is encountering a challenge in his or her being and so forth—it seems to me that that is very important to try to come to terms with. I think RUSTIN is perfectly right to say: "Look, these people live in different universes and ultimately all we can do is battle it out and there is no possible way of ending that battle". But then I thought about the child and it seems to me that it is essential first and foremost to try to discern as best one can who that child is. And it seems to me that one can arrive very cautiously at some understanding that is appropriately complex, multifaceted, and does justice to the full measure of who that child is; it points one in the direction of particular kinds of understanding, not just any understanding. And so I understand RUSTIN more in the direction of commensurability than incommensurability. [10]

Marilyn STRATHERN: I wanted to get to a different point, which presupposes that you start with relationality. There is nothing that is not constituted by the relations it is enmeshed in. I did not use a human subject because you then muddle the conceptual relations with social relations. I would have said the question of "Who is the child?" is absolutely the wrong question. You have already singularised the child. [11]

Mark FREEMAN: I would say that singularising the child is certainly a very serious risk, and to the extent that one singularises in an objectifying way such that that child becomes a thing—sure, that is a problem. But to ask that question—who that still nameless child is—that question needs to be addressed in as multidimensional a way as possible and I don't know if that singularises. It certainly does particularise. [12]

Mike RUSTIN: The notion of these different perspectives battling it out does not mean that there cannot be a debate. Even if one thinks that there are different perspectives that remain in permanent conflict or difference, one might hold that there is one shared commitment to the norms of democratic discourse. I no longer think that there is any reason to believe that in the end everyone will finish up holding the same position. There plainly are different ways in which people can construct the meaning and value of life. It is a mistake to imagine that one's own certitudes are ever going to be everybody's certitudes. I think there is some sense in which differences of ontology and belief are part of what is there and are of value in themselves since they create the possibility of openness, complexity, change, variety and development. One thing that we have got out of the multiperspectival post-modern is the recognition and realisation that it not only won't be but shouldn't be closed down in that sense. [13]

Mark FREEMAN: I understand your reluctance to talk about the real child on some level, but it is also very problematic. It seems to me that there are some perspectives which are more "closed" ahead of time than others. For instance, there is an approach that presumes from the word go that this is a pharmacological problem—almost irrespective of how the child is. Now, there are also perspectives where one at least has a certain kind of openness to say "let's listen to this child and let's see whether it is possible that my perspective can be displaced, enlarged, transformed and so forth". But if the child is simply a prisoner of my perspective, then we have a problem here. Something presumably is going to be able to transform the way I see this kid. I am going to call that "the real child" which is going to be very important to deal with in a non-singularising way. [14]

Nira YUVAL-DAVIS: Beyond all the different perspectives, we have the construction of the child as a patient. And the common value of all the different practitioners is an attempt to cure the child or to make the child better and therefore there is a notion of complementarity, which is possible beyond the differences, beyond the value systems. And listening to the child is also a construction that is complementary to all these perspectives. To say "let's try relationality" is also just one perspective; this is not instead of perspectives. Very often we are deconstructing what others are thinking, but we are not deconstructing what we are doing at the same time by presenting an alternative kind of construction. I think in this sense we have to be much more reflexive. [15]

Srikant SARANGI: To what extent is description a premise for understanding? When understanding is shared, is there no need for description? Let me give the cases that we have been talking about: the shell. If it had a very symbolic ritual significance, then thick description is not necessary, because you need thick participation, not just thick description and thick interpretation. In the same way, categorising a child as ADHD: if you are trying to fit a token to a type, a category, then the role of description will be fairly selective, rather than thick. So I'm trying to connect these issues—but at the same time problematise the notion of thick description, because somewhere there is a sense that thick description is a "good thing". [16]

Matti HYVÄRINEN: Discussing thick description in terms of narrative theory, it's a nineteenth-century realistic novel in its tradition. The point of the story is not thick description anyway, it is evaluation, and that brings us back to interpretation. Therefore the vocabulary of "thick description" does not hit the point. [17]

Srikant SARANGI: My interest in thick description relates to another concept I have worked with, which is thick participation. Where does thick description come from? So, for instance, if I look at the shell, I can give a description, or multiple descriptions, of it. But if I have not participated in those activities that you have, then the kind of thick description I am going to offer is going to be categorically different from yours. It's not just a question of different contexts. Can one attempt thick description without thick participation and understanding? [18]

Liz STANLEY: Description is always already theorised because it's in language and there is no language outside of the social. [19]

Marilyn STRATHERN: Do we begin with description or do we end with it? Because one might take description as the primary activity, the initial laying out of what you think you are describing. But if in fact you take description as the end result, which subsumes analysis, interpretation, theory, you make it an explicit activity. [20]

Mike RUSTIN: Case material is very rich and powerful in its capacity to evoke individuals and their particular kinds of qualities. But it does not make a difference to the body of classifications of concepts. One describes an individual in relation to concepts and categories that are already there—not in a textbook, but in a kind of resource-bank of concepts and categories. But sometimes, somebody comes along and says, "well, with all the categories and concepts I have got available to me, here is a patient which does not meet any of those", and that is the point where change takes place, some evolution in the available body of categories and concepts. So there is both the use of a set of categories to understand a set of individuals, trying to register the particularity of the individuals as much as possible; and some evolution of the body of categories that is available to think about fresh individuals. And when we try to teach clinicians how to do clinical research, we are trying to get them to think about the differences between understanding the individual in relation to whatever is there already to understand the individual, and what it is to find out something new which is generalisable. Now, when we were doing our socio-biographical research we found the same kind of question: we were doing life history interviews whose objective was to produce a descriptive record because the subjects basically told us their life stories. We did not ask them any questions, we did not offer them any categories—or very few, perhaps too many but hardly any at all—one line, or one line of certain categories which however got some of them going in relation to our categories in ways that we did not quite intend, but mostly they told us for two hours what they thought. And then we had the problem: How do we classify this? What is it? And the problem we had was whether we could use these descriptions of these individuals, which are very vivid and interesting, to explain what their experience of unemployment or single-parenthood was like. Could we use these to generate changes, new knowledge, innovations, in the body of conceptual categories that we were drawing on to try to understand them? We had a set of concepts; we drew on whatever literature we could to find some resources to understand them, but our problem was could we say we learn anything of a generalisable kind that would then be relevant to any other kinds of subject? The conclusion I came to was that it was very difficult, because of our heterogeneous sample of subjects, there was not sufficient comparability. [21]

Jens BROCKMEIER: Description is always-already interpretation, because it is in language, and language is interpretation. I see this as the summary of what we are talking about, namely the transition from the modernist episteme to that which has happened afterwards. Marilyn STRATHERN's talk made it literally palpable that facts do not speak for themselves. This was the belief of the positivist and empiricist modernist, that you have to go to the facts and then you have the "truth". But if you take, for example, this shell, which Marilyn has shown us, you do not have "the truth"; you do not even know what you have unless you have an interpretative framework, a point of view on the world. Now so far this is mainstream thought. The difficulty now starts when we substitute the shell for the little boy, particularly because this boy has been presented and treated as a mute being. He does not speak and he does strange things. So what categories do we have now? What is our concept of the human subject? Let's call it the notion of the human. Does it do justice to the fact that this little boy might in a way speak, maybe under different circumstances and maybe in another different language. I would say, let's listen more carefully to that which he has to say in order to bring him out of this scheme and look at what kind of category we can apply to understand him as an agentic subject. [22]

Maria TAMBOUKOU: I have two platforms to think about these questions of objectivity, and although these platforms are different, they would both shatter this question of objectivity. One platform would be the HEIDEGGERian platform. If we take the HEIDEGGERian platform of being within the world, and of the subject/object divide not being relevant at all, everything we have been talking about now would not be relevant because we would need this embeddedness in the world, and this practice is used to enlighten our understanding of what is going on. My second platform would be a Spinozist-Deleuzian one. From this platform, my question would not be what this shell is, or what this child is, but rather what is the shell becoming? Or what is this child becoming? And I would take into my consideration the virtual very much as the other side of the actual. And I think Liz STANLEY talked very much about the virtual because she brought in history, and history is part of the virtual, of what constitutes not what is, but what becomes. I thought that our discussion today opened up these different spaces from where to think: we had a very good anthropological journey towards the shell, then we had a psychoanalytic approach towards the child, and then we had a socio-historical route into women's participation in South Africa. And that is the most crucial part of our session today, this opening up of the question of objectivity and this coming together into shattering the dominance of the importance of this question. For me, relativism vs. objectivity is not a question at all. [23]

Srikant SARANGI: But can you talk about what is becoming without talking about what is? Can you be specific about the two cases we have been studying today? [24]

Maria TAMBOUKOU: Mike RUSTIN came here very anxious saying "I want us to know what happened today, what happened to this child", so his concern was about the child's becoming. And Marilyn STRATHERN's shell kept changing, as we talked about it. [25]

Matti HYVÄRINEN: I want to repeat the point of Dominick LaCAPRA, who has written a lot on trauma. He still argues for objectivity since it is a reasonable and very important part of our research attitude, and his meaning is as a sort of a control of social transference, because he takes the psychoanalytic concept of transference into a larger context. His example in terms of trauma studies is always that when we study trauma and traumatic narrators, we should not put ourselves into the role of substitute victim. [26]

Corinne SQUIRE: Mike RUSTIN has an account of how therapists need to do certain kinds of child observations that provide a descriptive record—a record that would be unrecognisable as description to, for instance, an experimental child developmental psychologist who would do that description in very different kinds of ways. I find pragmatics of description very interesting. [27]

Marilyn STRATHERN: If we live in a world where the notion of history is powerful, you need to deal with it. And we actually live in a world where the notion of objectivity is powerful, and the question is then how do we deal with it? It is no good saying, "well, it is inconvenient and we will put it to one side and we will just be interested in other things". Do you people think I am not passionately concerned about the way we regard this shell? How does one convey the fact that these 21st century members of the world produce things that to our eyes are already pre-classified as somehow exotic? And I produced for you a sort of museum object in order to try to show that I can speak of that in the same breath as LOCKE, or feminist scholarship, or the interests of this roundtable. [28]

Margie WETHERELL: There seems to be real anxiety about getting beyond the intellectual's, the investigator's or the scholar's perspective, so we have "the real child" and "becoming", which is the scholar plus the world. I think a way of breaking the self-absorption of the analyst is to refocus attention on the participants' own orientations: their interpretations, their understandings of this flow of interaction. You have got this interesting case that seems to be outside interaction. But as Mike RUSTIN described it, the boy had orientations, he had actions, he had practices and a scream. So even in that hard case we can look to this as a source of information about participants' orientations. We can understand them because we are embedded in the same world that they are. [29]

Molly ANDREWS: I am thinking about ourselves: social science researchers generally, as being in this role of mediators between these things that we are trying to either understand or observe or describe—be they shells or children, or whatever—and actually both describing/interpreting what we see, and passing this on in some kind of fashion for some particular function. We are embedded and emotionally charged. I think we have to then question what does it take for us to be able to either listen to the child if there is an actual physical thing here about either screaming or silence—and these two things sitting side by side is quite dramatic of course—or also listening to this shell in some metaphoric sense? I agree with Srikant SARANGI when he says that "What are we becoming?" has to invite the question "well, what is this child, or shell, to us as these mediators?" This is where a lot of these important points almost intersect, in the complexity of this listening/mediator role and who we are in this kind of relational understanding. [30]

Lynn FROGGETT: The fact that objectivity won't go away has always been there; ever since the beginning of the profession, there has been a debate about how social workers should write their accounts. It is still going on, and there have been hundreds of attempted solutions to this problem, all in the name of preserving objectivity. And at the moment there is a great worry that on the one hand you have a sort of technocratic, managerialised, technical-rational objectivity, on the other hand you have people producing very emotive writing who try a thick emotional description, a sort of mise-en-scène of the encounter, and that raises huge anxieties often in the people who read those descriptions, despite the fact that they are exhilarating and deeply satisfying to read as sort of literary products. The difficulty seems to reside in the preconceptual, and it is something about what perceptual repertoires are brought to bear on the object that we are not quite grasping here. [31]

Mark FREEMAN: We have been talking about the preconceptual. I am not sure whether that may be the best term. We could talk more about the pretheoretical; even when one is watching, one is doing that within language, within the idioms of Western thinking, which it seems to me bring the interpretative dimension to bear on the objects in question. That is not a naïve process, it is certainly not a nonrelational process, but what we are asking people to do in that context is to keep their own theoretical lenses in abeyance as much as possible, recognising that probably that cannot ultimately be done. Here I would make that distinction: preconceptual versus pretheoretical. [32]

Lynn FROGGETT: The problem is that as soon as you say that there is not a preconceptual, because our lenses are incipiently conceptual, and rule out the possibility of a level of experience which is not yet formed in that way, I think you have huge problems in explaining dreaming, and in explaining art and in explaining children—and love. [33]

Jens BROCKMEIER: You said, "objectivity does not go away". We live in a world where objectivity is a very strong, powerful, emphatic term. We cannot ignore it, we must redefine it. I find it convincing that Marilyn STRATHERN tries to redefine objectivity by rejecting the traditional view that in order to become objective we have to leave out all that is messy, all that cannot be reduced to the neat and clear—objective—picture. A counter-term to objectivity is complexity. Complexity has a long tradition in the philosophy and theory of science. What has been called the "reduction of complexity" helps us to understand what is considered to be "objective", namely what can be defined, measured, controlled. Marilyn STRATHERN describes the important problem for the anthropologist of dealing with an excess of information and meanings and context and frames. It sounds very much like the problem that a poet or a writer has. So how uncomfortable would you, Marilyn, feel to be put in the company of writers and poets in dealing with the excess of information and world? [35]

Marilyn STRATHERN: In terms of the fabrication and manufacture of the narrative: very comfortable indeed. In terms of the content: very different because there's no free play. Everything needs to be demonstrably the product of experience, observation, and all the types of evidence. The reduction of complexity is not the way to go. With my little simple schema of co-ordinates you can arrive at specifying particular objects at the point where you can specify the combination of different perspectives. It is a way of doing it that was not complexity- reducing, that is the point. So that is more important than the solution: the formation of the question. [36]

Nira YUVAL-DAVIS: I wonder if the discussion can progress if instead of talking about objectivity we point to truth, because in terms of dialogical standpoint theory truth is the opposite of the reduction of complexity. We can only approach truth, we can never reach it, but we are nearer if we encompass a variety of gazes and a variety of standpoints and a variety of interpretations and approaches to the same object or subject. And I always found that working with this notion of trying to approximate truth is a much more productive notion than objectivity which is very stultifying and very hierarchical in terms of interpretation. Objectivity puts different perspectives necessarily in competition with each other in terms of their validity. [37]

Phil COHEN: What Marilyn STRATHERN said was that objectivity is what remains constant under transformation. Now, it seems to me that what we are struggling with is one version of reading, where you take what remains constant under transformation: concentrations of wealth and power, for example. So that version puts objectivity on the side of power: we do not want that kind of objectivity. You did offer us another way of reading that notion of what remains constant under transformation, and offer us a different way through. WITTGENSTEIN argues basically that a method of observation is one which struggles to let the subject or object be what it is for and in itself, but the paradox is that it produces the possibility of an alternative reading or interpretation. So that it is not something that involves reification. If we can work with the WITTGENSTEINian notion of observation, it seems to me that that then allows us to think about what remains constant under transformation as it travels across different contexts, but nevertheless remains something we can talk about in common. That offers us another model of objectivity, which means that we do not have to abandon objectivity. We don't have to say "It's their word, it's their language, it's their whatever"—we can try to work up our own meaning. [38]

Liz STANLEY: I have a problem with the notion of truth being used in the singular. If I understood, the point you made is that if you were to take many perspectives, many different views on something you would be moving towards the truth. Whereas I would say that if you were to do that you would be discovering many truths. [39]

Nira YUVAL-DAVIS: Many truths means that you have notions of truth, which are complete in themselves and multiple in an additive way. I think that truth is not like that. You can never really reach it, and it is not multiple. It contains in it the variety of perspectives and gazes in which different people approach it. Of course there are untruths, but in terms of truth it is never just one comprehensive, closed concept. I do not think that we can talk about truth in this way because then we lose the dynamic shifting contesting dimensions that Maria TAMBOUKOU was talking about, the becoming of the historicity of change, and in this sense all this has to be contained within the notion of truth. [40]

Liz STANLEY: One of the reasons why I chose to talk about what I did is that it provides a literal object lesson in what happens when there are not partial perspectives on truth, when there are processes at work by which some groups/sections successfully lay claim to the truth and there is no other truth possible. And what happens over time, is that the political origins of that, the processes of becoming, are forgotten as time passes, so it does become the truth, it becomes the facts, the history, what everybody knows. It has further consequences, too, producing very thin descriptions such as, "that is apartheid in South Africa" or "that is repression in the Soviet Union", or "that is fundamentalism in America under Bush". Also, we all work in universities, so for certain we have seen these kinds of closures over what is truth and what is bias happening. Such closures over the complexities of past "nows" seem to be a very powerful part of social life. Am I alone in the room in thinking well, there comes a point where you do not go search for a point of social origin, but make an intellectual choice about a problematic that you want to explore, such as, "how did that particular closure happen, by what means and processes?". I do not see how else it is possible to proceed without so tying what we do up in knots that we paralyse ourselves, which does not seem to me to be very interesting. [41]

Molly ANDREWS: How do we determine the boundaries around our research question? [42]

Brian ROBERTS: I did wonder why you chose the shell. Initially, why, given that it is important to you, and you allowed us to pass it round, which I think is very brave really, it having been broken once, but it did cross my mind. But what struck me was this ethnographic excess that you started with. And excess is a very loaded term. It seemed to worry you quite a lot, and I did wonder why we should be worried by excess, and I am going to side with the worry about reduction. But it did concern me: What is wrong with excess as an issue? There has been a lot of discussion of excess along a whole range of lines. To an extent we should have a notion of excess reported and invested in the process. How worried should we be about it? Excess in terms of context: One context on top of another. Excess of interpretation: As someone who does that I was quite aware of interpretation on top of interpretation. To what extent should we be participatory in approaching the knowledge around a particular object? At what point do we intervene in terms of participation? Are we more detached from it? And the last one is excess and reflexivity. It is often alleged against some of us involved in autobiography that we have more concern about reflexivity than in really doing things, or researching things. So there is a range of these things—emotion, participation, context, interpretation, reflexivity—and really the task before us is in this notion of what to rule in and what to rule out, where the boundaries lie, leaving aside objectivity. When we are looking at doing particular research, we think, "well, if I am taking that context or this context, or I am taking that kind of reflexivity, what are the possibilities and what are the closures?" That is how I would do it. Looking at those kinds of excesses or reductions and saying "What, within the choices we make among those five items, are the closures being made and what are the possibilities?" So I am not worried about excess particularly. I would be worried about possibilities and closure. I am not saying it is either bad or good, but what is closing down, what is possible, is more important. [43]

Jane ELLIOTT: Is there a problem with excess? The three papers point to different problems or non-problems with excess. In terms of the shell and the different conceptual frameworks there does not seem to be a problem with excess. What we want is more excesses of different interpretations of the history rather than just going down a certain narrow line. Where there does seem to be a problem with excess is with the child. What we are stuck with concerning the child is that the excess leaves us without a plan of action. It's when we start thinking about where do we go from here that the excess becomes problematic. [44]

Marilyn STRATHERN: I will just say two things in very short order: First on the idea of excess: I think that was working against the notion that method is reduction in advance; that is, you purify. And you can guarantee the purity of your data through the methods that you use. This characterises many social sciences and this is the short route of demonstrating objectivity. It is a form of defense of being open-ended, so that excess actually claims positive connotations. But one thing I wanted to end on was that when people look at these shells, they see a foetus: the shell is a child. [45]

Mike RUSTIN: What has been striking me listening to this discussion is how committed I am to the idea that knowledge and truth advance through strong research programs, that is to say through some idea of accumulation in which the ontology and the epistemology of the work are aligned; that is to say that the mode of finding out, and the categories, classifications and theories which arise, are aligned with one another. And I would give examples of such research programs as DARWINism, WEBERian sociology, MARXism, at one time, and psychoanalysis. One can think of a number of research programs where there has been a sense inside them that some further extension of knowledge is taking place without necessarily wanting to argue that this is the only kind of knowledge there is. What worries me about qualitative sociology, narrative methods and our own biographical methods is that I do not see emerging from those the frames of classification and theory, which allow accumulation to take place. So one has the risk of an endless set of particulars and perspectives with no development at all. What I am pointing to is an area which is somewhere between objectivity—because these perspectives have almost nothing in common with one another—and the idea of an infinity of perspectival choices. In either of those areas there are problems, but it is the middle space where there is a provisional order through which development takes place, which is where discovery and understanding take place. [46]

Liz STANLEY: While I share the interest and fascination with excess, it seems that research is always an exercise in closure and the trick is knowing the closures that you make and being able to name them. [47]

The discussion began from a distinction between approaches to social research on identity that emerge from social practices, and those that emerge from the theoretical position of intersectionality. Representatives of practice-based research presented themselves as less interested in the theoretical development of concepts such as identity, and more concerned with the working out of particular complex social problematics at micro, meso, and macro- levels. Advocates of intersectionality emphasised the necessity to analyse data at the same levels, but also to focus on intersectionality in relation to the multiple situational possibilities of identity, and in fields beyond "identity": those of social locations, and values. Problems of identity politics were discussed, that is, how the term identity has moved from indicating a politics of resistance into supporting a hegemonic politics. The important continuity of identity was for some dissipated in current overly general uses of the term. Another strand of the discussion concerned the language focus of most identity research; a focus that tends to neglect important aspects of identity construction such as social formations, broader discursive processes, and the unconscious. The political significances of a methodological focus on intersectionalities, versus one on practice, were presented in strongly differentiated ways. [48]

3.2 Margaret WETHERELL: Methods for studying multiple identities: Intersectionality, practice and troubled and untroubled subject positions'

I want to take up the issue of the multiple or plural nature of identity. How do we understand this and how can we build methods for investigating multiplicity? I contrast two different approaches—a more top-down theory of "intersectional identities" which has been very important in developing feminist and anti-racist politics, with a more grounded, inductive, bottom-up "practice" based approach, which is nevertheless still informed by larger, Foucauldian discursive analysis. I argue that in terms of building a research strategy for understanding lived identity and social action, the latter is more productive, particularly when the former takes a too-simple, additive approach to identity categories. It is helpful to make moments of practice the units of analysis, rather than large-scale identity categories per se. But macro, meso, and the micro levels of discursive functioning combine to order social action. In making this argument, my object of examination for us, my "shell", is the 2003 Big Brother exchange between Jade and Spencer on the location of Cambridge—just up the road from here—and East Anglia. While this exchange can be read in class terms, it also benefits from being investigated at the level of its intricate, moment-to-moment interactions, and in terms of the personal discursive orders of the people involved. [49]

I also argue that examining the ways in which multiple subject positions become troubled and untroubled in the flow of practice is a useful way of understanding how power configures the everyday multiplicities of identity. Mostly, they are untroubled, performed and understood complexly but unproblematically. It is when they become obviously troubled that multiple subject positions become obvious. [50]

I agree with Margie WETHERELL that it is necessary to take into account the three contextual forms of social order micro, meso and macro. Indeed people do not evolve in a vacuum and contextualising is necessary: their identity is being played out at the crossroads of these three levels. I also agree with the insistence on the importance of studying identity through the looking-glass of troubled positions. Major dislocations, such as uprooting or genocide, are heuristically particularly interesting because they offer opportunities to reveal aspects of deconstruction or reconstruction of identity, which might otherwise remain unnoticed. I don't really know to which camp I belong, in the dichotomy/opposition, which Margaret WETHERELL draws between the intersectionalist and the practice approach. As far as I understand the implications of each of these options, I would be inclined to argue that there is a complementarity rather than an opposition between the two perspectives. It is through the way a person acts in various situations or the discourses he/she holds within practices that identity can be examined. In other words, it is only from within the practices that it becomes possible to unravel how each of a number of axes of social signification on which people are located is being processed; how they are called upon, displayed or played down, come and go, depends on the practice. Is it really adequate to oppose intersectionalist and practice research strategies rather than to explore the intricacies of the two and in this manner to unravel the interplay of the three forms of social order, micro, meso and macro, in the construction of identity? Identity has to be theorised against the background of changing realities. Therefore rather than speaking of multiple identities, I prefer to speak of the changing realities of identity. Personal identity is shaped by, and closely connected to, broad socio-political movements and to the official discourses held about these movements. It is important to consider the dynamic construction and renegotiation of identity in the light of changing social realities. There are also changes in social reality, which open up new social options and possibilities and similarly make repositioning necessary. I have deliberately avoided speaking of identities, that is, adding a plural to identity. I basically disagree with the statement that "there is more than one "I" in identity". We may display different facets of our self; we have a variety of identity strategies and they may change over time, but I strongly believe that there is a "unified I" unless one falls in the multiple identity syndrome. But this is then a totally different story. [51]

I have no problem of looking at identity in terms of practices, in terms of experiential and performative aspects of identity. Identity is constructed, contested, authorised, in these practices, which are on the micro and the intersubjective levels. But intersectional analysis is not only about identity. Discussing all of the issues which relate to the syndrome of social relations under the notion of identity is very misleading. The fact that identity politics has been hegemonic for so long relates to the fact that often in identity discourse there is no differentiation between the analytical levels of positionings and social locations, identifications, and normative and political values. A level which is even more important to differentiate lies outside or between what Margie WETHERELL called "identity categories". For these categories are not identical, but are in different social locations. And a third level which has to be differentiated is the level of values, social and political. They also cannot be reduced to or automatically translated into identity politics. Margie WETHERELL said that intersectionality begins in the wrong place with identity categories or locations rather than practices, and with a narrow definition of identity, in terms of social categories. My approach and that of many others to the study of identity is precisely not about social categories. It is about narratives of self and others; but "social categories" are important in the sense of indicating vectors of social location and relations of power. If you try to differentiate between these two levels and to examine in particular social reality what are the vectors that are relevant in the hierarchy of power, in the dynamics of power, existing in a particular social reality? Isolating particular vectors of power in the foreground of social and political analyses is not redundant because it is important to interrelate the overall context and what is happening in minute everyday relationships so as to situate the narratives. [52]

The second point that I would like to raise relates to what Margie WETHERELL said about how the metaphor of intersections can pull the analyst back into additive thinking: gender plus ethnicity plus class plus sexuality. The issue is that we have to differentiate between the ontological bases of each dimension that are not reducible to those of another. There is a discourse about gender relations, about class relations, ability, sexuality and so on. And in this sense, we do talk about disparate discourses. But this is something very different than analysis of the concrete social realities in which they are mutually constituted. This is an important debate in the field of intersectionality and indeed the term "intersectionality" can be misleading in this way. But many of us reject the additive approach firmly without giving up the crucial insights that intersectionality analysis offers. [53]

Margie's third point was that intersectionality can encourage an overly deterministic perspective and neglect the self-determining and context-creating activity of social actors. This is exactly the opposite of the case. If we differentiate analytically between the three levels I have described, then this determinist approach is not possible. My basic answer to the question that Margie WETHERELL poses is exactly like Erika APFELBAUM's; that intersectionality and practice based are not two opposing approaches. They are complementary, and vitally complementary. [54]

Maria TAMBOUKOU: I read Margaret WETHERELL's paper quite differently from the discussants. The main point the paper was trying to make was to talk about the risk and danger of intersectionality becoming a stasis and a fixity. In this way, WETHERELL's suggestion of practice-based analysis looks at an ongoing destabilisation of the danger and risk of becoming fixed, without just bracketing the big differences and factors. I find it a very strong suggestion. My only concern here is how Margie WETHERELL has used the spatial metaphor of levels to talk about fixity. My other point is about how she looks at the analysis of trouble and power, and troubled positions. Foucauldian as I am, recently I have seen that power is not enough to account for what is happening in identity construction. Maybe desire in the Deleuzian notion of desire as production, not as lack, as in the psychoanalytic framework, could be used here to make sense of these practices. My final point is about all of us being aware that there are different philosophical underpinnings in how we understand a particular identity. When Erika APFELBAUM said, "I don't agree with the multiple I's", then we have different philosophies there. So it's clear for all of us to know the different platforms where we stand and talk about identity. [55]

Liz STANLEY: One thing that has been going through my mind is while intellectually I accept everything you say—and yes of course it makes absolutely perfect sense that we think of whole people and complexities. But as we in Britain have gone through an election period, a rather long and tedious election period, I was thinking of the way in which single—I am half Roma—racial categories have been used in the long run-up to this. I was also thinking of the conundrum that whilst academics, including academics who have some kind of political or ethical commitment, are moving in the direction of intersectional analysis, formal politics has moved the other way; there is the intractability of politics and the ways in which identity is fixed in the singular, and on particular persons. And added to this, those particular persons often find it much easier to mobilise around a single identity, something which is a puzzle in intersectional analysis. In life it's not a puzzle but it's something intersectional analysis has to grapple with. [56]

Margaret WETHERELL: When we are in the academic mode, we are concerned with machineries of representation around competing perspectives and getting the right epistemological positions sorted out. The process of politics is much more about choosing effective machineries. As I prepared this paper, the difference between an intersectional and practice-based approach was one I sometimes tried to construct; sometimes, I felt I was trying to describe it. That was a salient kind of activity as an academic. In the political arena we have to take whatever kind of intellectual tools will work, and sometimes intersectional analyses work best, they really add the best, the most effective kinds of politics. At other times that is not the case. [57]

Nira YUVAL-DAVIS: What is happening is that identity politics has moved from the politics of resistance into being a hegemonic politics because the hegemonic majority are feeling under threat in a way that in the past they did not. This is not just in terms of the risk society but of all the differences of identity, community, and globalisation. You have to fight against racism and sexism, the racism of the hegemonic majority and of our own communities. And you cannot chose or reduce or reify one identity over the other. Of course, it is more difficult to mobilise around this approach because it is less demagogic than identity politics is. It is very fascinating to see how New Labour is trying to find counter-ideological positions. They are trying to see how you can be a good Briton—a patriot in the same way as a Conservative—not by homogenising original Englishness or Britishness, but around diversity. What have they found? They have found two things with both defeating the whole purpose of the project. In (the White Paper) Secure Borders, Safe Haven, New Labour arguments about immigration control are exactly falling into line with the Conservative argument that if you control immigration then you can handle internal diversity and maintain social cohesion. Gordon BROWN has gone further than that. He says that we should not be ashamed of ourselves for wanting to have a cohesive identity. This is why, he says, we have to stop looking negatively at the British Empire. He says we have to change our whole attitude to the British Empire and to the whole of history. So they are trying to hold on to internal diversity of origin, but in a way the whole thing is falling apart. But it is all about identity politics, and I think this is where politically as well as analytically we have to work. [58]

Valerie HEY: What a lot of the discourse we are struggling with is about, is how to get to an identification politics that could mobilise around some very diverse and complex values. Identity or identifications are not just about those things we seek strategically, as in political mobilisations. Things get done to us, we do get fixed, we do get put in place because desire and power work like that. So we need to put power and desire back together. [59]

Alexandra GEORGAKOPOULOU: Within the practice-based approach, Margaret WETHERELL obscured some of the heterogeneity that exists. It is not one field, and maybe the opposition between practice-based and intersectional approaches also needs to be related to some internal dialogues. A pivotal part of any analysis that is bottom-up and practice-based is attention to language. There is a very strong emphasis on how people talk identities, what they are in talk, in discourse, in language. This is characteristic of the ways in which intersectionality works within practice-based approaches; we do talk about the interarticulation and coarticulation of our identities in these approaches. But then again, the emphasis is on articulation; in other words, it's about what people say about identity. We place so much emphasis on what is said, then what about the unsaid, or the unsayable? The way forward is an approach that will still look at the social and at discourse processes while at the same time bringing in the unconscious, as long as it is verbalised. [60]

Prue CHAMBERLAYNE: We started off with methodology, and presumably we are now discussing identity because it is important in qualitative methods to be clear about what theory of the subject we are using. I feel quite frustrated with the concept of identity because it is thin and static compared with concepts of interaction—Erika APFELBAUM was talking about interaction, and strategies which are much more action-based. And on the meso- level and the micro-level, I am thinking, "yes, but identity doesn't capture what you are talking about because you are talking about, for instance personal projects which might have a long dureé, so you are talking about something quite radical". In my language you are talking about the person or the self. There are richer concepts really. [61]

Margaret WETHERELL: I want to move away from identity. If you begin with practices your starting point is not identity and things that people display or carry around with them from situation to situation. It is the local ordering of activity. That has all kinds of methodological implications, because it affects the kinds of research questions that you ask, and it also makes a shift from sample-based work to site-based, institution-based, or problem-based work. So you are starting not by saying you are interviewing a sample of x kind of people but rather that you are interested in some efficacious site in social life, and you are trying to work up from that. Then identity appears within that, in the kind of positions that are available in our society. So it's about trying to not make identity such a dominant term. Part of my problem with intersectionality is that it takes us back all the time to identity. [62]

(Unidentified speaker): I was struggling with this juxtaposition between the two, intersectionality and practice, because it seems to me that by putting it in that way there is no practice involved in the intersectional work. We have been talking about the three levels: the meso, the macro, and the micro. There is an intersection between those three levels, but it is an analytic nicety to just keep them as three different levels so we can talk about them separately. But what I am particularly interested in, is what the differences are between those two paradigms—intersectionality and practice-based. If these two ways of looking at identities are to work at the participants' level, can we see whether they are doing intersectional work, or practice-based work? [63]

Margaret WETHERELL: The moves that you are making are exactly the ones that I was wanting to celebrate: That sort of sense of following the participants, trying to see what kind of work they are doing, and how the world has been constructed by them, so the focus turns to identity work, identity practice, rather than identity or identities. [64]

Mark FREEMAN: There is a tension between especially what Margie WETHERELL had said and what Erika APFELBAUM had to say, and I wonder whether it permits a kind of dialectical resolution in the following way: Erika APFELBAUM talked about displaying the idea that identity is displayed. The language of display is the language of unity. That is, someone is bringing a unitary—more or less unitary—self, which gets displayed in these different places. Margie WETHERELL, on the other hand, talked much more about practice, and it seems to me that that carries with it, sensibly enough, the idea of multiplicity. But as Erika would probably suggest, identity is also reconstructed through practice, and it seems also that there is a kind of specific meaning to the idea of identity, and it has to do with unity or continuity in multiplicity. Isn't that generally what we mean by identity, the idea of plural identity? I don't know if that really works, because identity presumes a kind of continuity, but you want to emphasise that it's reconstructed through practice, and hence displayed anew. [65]

Erika APFELBAUM: When you said "in multiplicity", it refers to identity in multiple situations. [66]

Margaret WETHERELL: Continuity has many sources. Continuity comes from practice, and from familiar, habitual routines. The point I was trying to make about trouble is that often you have intensely multiple positions that you can track through, but a complete lack of trouble. Does that make sense? And that's a kind of very interesting sort of phenomenon. [67]

Wendy PATTERSON: Narrative identity is what we do in the stories we tell about ourselves, or you could say identity is an accumulation of the stories one has about the self. But some identities are much easier to achieve than others and some ways of being are much easier to get to. So we do not have agency and free will in the construction of our identities through narrative, because events are prescribed to us and imposed upon us. Because what we do everyday is construct identities in talk or we do our identities, however, we do not have control always of them. I do not want to lose this idea of hierarchies and power relations. Intersectionality is a way of analysing imposed and prescribed identity. [68]

Erika APFELBAUM: I don't think that identity is about ascribed categories. Privately I still tend to want to dismiss altogether the word identity and replace it by something that clearly defines what it is we are talking about. We are talking of identity at multiple levels. [69]

Anneke SOOLS: I wanted to take up the idea of hierarchies. Intersectionality offers us the insight that the dominant and the marginalised are always constructed in relation to each other. [70]

Mike RUSTIN: I would like to raise a question about how one might think of these categories and concepts of identity being used in an actual research context. I have two points to make about it. One is that I think that the broad concepts of race, gender and class are much too high-level and generalised to touch and to get close to the kind of identity questions that bother most people most of the time: "Do I pass my exams?" "Am I any good as a daughter?" "Am I any good as a friend?" "Am I a reasonable kind of policeman?" and so on. All the time everybody is engaged in identity work, (involving our) norms and expectations of people. Of course, the broad heavy-duty sociological-political categories do come into it all somewhere. So the question is how do we try and raise enough categories to investigate the negotiation, development and choice of identity in everyday life? The second point is in relation to politics, which is that there are of course important, powerful specialists in the negotiation of identity in society out there. They are advertisers and marketers of consumer goods and of course they are politicians. And the difficulty of studying what they do is that the game that is played by people offering images and identities to potential consumers, or politicians in their not very competent way offering images and identities to the public, is a very quick-moving game, in which the takeup and response and the moving on happens like lightening. So how do you track that movement? How do you map it as it happens in the political process? Even more problematic, how do you do it in a way that might have some political consequences? These are very important topics because how people construct themselves—their values, their sense of who they are, how they live with other people—is the fundamental part of social life. But we need a fine-grained, in-tune analysis. [71]

Nira YUVAL-DAVIS: Identity is not just about narration of self. It is about narration of self and other, and about where the boundary between self and other—individually and collectively—is passed. And of course, the autonomy that you are given, in order to construct the boundary and where to construct it, differs a lot, and issues of forced identity are absolutely major. So this is one very important issue. The other issue arises from Mike RUSTIN's claim that the major categories of race, gender and so on are mostly irrelevant. I think they are mostly irrelevant only for people in hegemonic positions of power. Can a black person be a successful policeman? "Race" is relevant. You cannot say that the question about whether you can be a good policeman is completely detached from the issue of "race". And this goes for all issues of situated knowledge, though there isn't necessarily a congruity between the different power dimensions of your identity. Position is very important in these overall vectors of power and positioning, i.e., your location; then, how these factors reflect in a differential way your narration of self and other; and how this reflects what kind of ideology and political values you adopt explicitly or implicitly. These are all related and they need be. If we decide to take out of the equation all of the macro-social structures, social categories and vectors of power, and just concentrate on micro-power relations … Look, I think postmodernism has depoliticised so much of our work. The times are very dangerous now. We cannot afford for this to happen. The issue is—and this is going back to the objectivity versus truth debate—doing research is a political act, and we cannot just take issues of intersectionality out of the equation. They change quickly, but this is part of power relations and politics. [72]

In her paper Trisha GREENHALGH suggested a possible incommensurability between two disciplines, sociology, and computer-aided text analysis. The concept of "evidence" in research was discussed in relation to varying underlying criteria. Matti HYVÄRINEN traced two narratological and narrative-analytic histories of social research around personal stories, and suggested looking at the processes by which voices, registers, arguments, metaphors or "words" are included or excluded in research and decision-making. Karen HENWOOD debated the challenges researchers are confronted with when they wish or need to establish the relevance of their findings beyond the academy. It was pointed out that an incommensurability of perspectives provides meta-data from which new and richer conceptualisations can be drawn. In the discussion, the recourse to binary positions and foundation stories was disputed, as well as the missing politics within the academy, using narrative research as an example. Participants discussed the impact of qualitative/narrative research and the (im-) possibilities of increasing it. [73]

4.2 Trisha GREENHALGH: Inter-disciplinarity in the study of rhetoric in healthcare policymaking

This paper reports on an interdisciplinary research study conducted by medical sociologists, who seek to study the social processes of healthcare policymaking, and design scientists who seek to study arguments and to represent the use of rhetorical argumentation devices with computer graphics. By unpacking the micro-processes involved in the use of language, argumentation theory offers an opportunity to go beyond what sociologists have already demonstrated—that "researchers and policymakers inhabit different worlds". Making explicit the rhetoric of the argument, and using technology for information visualisation to present it in an accessible form to policymakers, has the potential to: a) draw out insights about how better to link the research world with the policymaking world and, b) provide a tool for improving policymakers' reflections on their work and thus their professional and team development. But we are discovering that the arguments of healthcare policymaking are remarkably hard to capture. We are beginning to learn lessons from what might be termed the incommensurability of our respective disciplines. [74]

Trisha GREENHALGH's paper documents in a compelling way how we cannot just resort to familiar, clear and safe methodologies and safe, privileged types of research material, if and when we try to investigate ongoing, politically relevant and utterly messy processes. Her key question is how the "evidence" for evidence-based policy is constituted. There might be, at the same time, another kind of question about how we understand "evidence". Quentin SKINNER has elegantly documented how the parties in a debate may either share the same concept or be pushed to innovatively reformulate, stretch or interpret concepts in order to get the upper hand in the debate. Trisha GREENHALGH's case clearly is one of those where aspects of the concept are tested, stretched, and fought for. SKINNER has outlined three separate aspects of conceptual change: firstly, changing criteria for the concept; secondly, a changing range of reference of the concept and, thirdly,, a changing appraisal of the concept. Coming back to evidence, we might be able to split the question. What are the criteria for evidence, and how do they vary between, say medicine, the court, and management? Are there any debates on the criteria and definitions of evidence? The second question is about the limits of applicability of evidence, and the process where they are negotiated. There may be an irony lurking within this appraisal: the broader, the more "humanistic" and open the range of reference that develops, the greater is the probability of a sinking appraisal. I wondered if there could be ways of connecting the two parts of Trisha's study more intimately together. One possibility might be an approach that is sometimes called "textual ethnography;" such as a study of the total paper-flow during the planning and decision-making process. You could ask, in BAKHTINian terms, about the administrative "intertextuality". Whose social voice, whose problems, whose complaints are textually included, foregrounded, and made present in the documents of final policy-making? Rhetoric and argumentation are too often conceptualised exclusively in terms of maximally effective persuasion and convincing. Could we instead make a heuristic distinction between a "closed" and "open" rhetorical culture? A vital part of "scientific" policy-making seems to involve the creation of a monological, formal, affect-free, and quasi-objective discourse. Interviewing people from different levels of administration gives a marvellous opportunity to: 1) hear different, informal stories, informal accounts of problems and decision-making and, 2) ponder the difference between the formal and informal registers of talk. In a good case, you may have two corpuses of rhetoric: 1) a confessional and informal corpus and, 2) a pool of formalised, scientific, objective, and neutral rhetoric of evidence. In such a case, it would be interesting to study the process by which voices, registers, arguments, metaphors, "words" indeed, are included or excluded, during the succession of planning memos, documents, and agendas. In the history of narrative study one can discern two parallel, narratological and narrative-analytic paths, the first subordinated to the second in contemporary social research. My argument here brings the two paths into relation (full paper version). [75]

There are risks associated with promoting the use of formalised, step-by-step guidelines for assessing research quality, if so doing fosters a culture of over-reliance upon protocols and audits. Trisha GREENHALGH's paper is a particularly good example of her activities as a boundary-crosser—not just between the social sciences and biomedicine, but also across the academic-policy divide. The study discussed in the paper is a particularly clear example of research that is working to also establish the relevance of its findings beyond the academy. Researchers face the demand of attaching further importance to their research by demonstrating its relevance to policy and practice. We are being encouraged to widen stakeholding in the research enterprise through practitioner-user involvement. Trisha GREENHALGH, in her study, sets out to involve health policy makers almost, although not quite, as co-workers in her action oriented research approach. One of her goals in so doing is to use the processes of practitioner involvement in the research and its products and outcomes in ways that can help them to reflect upon their work and improve teamwork and development. Trisha GREENHALGH's work sets out to embed health professionals in her study in the process of potentially becoming a new kind of practitioner. The openness of qualitative inquiry represents a key methodological issue for predominantly experimental disciplines to take on board, along with finding ways to study research objects in their full complexity and in situ, and to take into account the multiple perspectives of researcher and researched. Marilyn STRATHERN's suggestion about how to approach questions about objectivity and validity is also highly pertinent; that the layering of perspectives and contexts offers its own kind of precision. Marilyn STRATHERN's position is in favour of limited incommensurability. She poses an argument for assembling a range of irreducible frameworks to tackle the problems of relativism that occur when matters of perspective or frame are considered merely as different people's opinions and views. In discussion of Marilyn STRATHERN's paper and Mike RUSTIN's commentary, other voices attached importance to openness in the ways researchers can hold onto a particular frame or set of frames, and also to remaining open to the challenges posed by the object of study when viewed on its own terms. Trisha GREENHALGH's paper adds a further position to these arguments, that incommensurability of perspectives provides meta-data from which new and richer conceptualisations can be drawn. In making reference to her own area of research, she sees this point as also highlighting a key aspect of healthcare decision-making, while qualitative researchers, of course, view it as an aspect of reflexive research practice. [76]

Liz STANLEY: I would like to make three points. One is a general one about the discussions that we have been having, and the two others. Binary positions are put up and people are talking in terms of those binary positions and I find myself very uncomfortable with that for number of reasons. On the surface of the conversation, people are saying "complexity, complexity", but nonetheless, two positions keep being put up against each other. The second point is about foundation stories. Whenever I hear a foundation story I think: "Huh!" Because whenever people invoke foundations there is always another story. And of course there is another story from Cambridge in the 1940s, with philosophers of history also talking about narrative. What happened to feminism in these accounts? Since 1968, 1969, feminists have been writing about biography, autobiography, stories, and narratives, and this other foundation story gets removed. We should all be more cautious about foundation stories and I am not suggesting that in the long run Matti HYVÄRINEN's foundation story of narrative would not do that, but it is important to hang on to some of the extremely problematic reverberations of the way in which naming canonical sets of ideas removes from existence what are the more subtle and tenuous workings of such ideas; and to remember that, to name those kinds of positions is the end of a process rather than the beginning of one. I am much more interested in these more tenuous, more open ways in which ideas get bounced round. The third issue is to pick up on Matti HYVÄRINEN's point. Who says what the criteria are for judging what is good and what is crap qualitative research? Why do we have such general stories? What is there in going on in this, that might militate against the work that probably everybody in this room is interested in? What are the crosscutting things? So I want to have some politics in the academy, not just outside of it. [77]

Matti HYVÄRINEN: Where did the temptation and almost the norm or the trend to tell personally come from? In my account of "narrative", I go through the classics; no one says much about this history. But within this history there is also a nice way in which feminism even as a political movement became part of the academic change, the narrative "turn". [78]

Karen HENWOOD: I agree with Liz STANLEY's points about being wary about polarised arguments and about foundation stories. What I was trying to do was give a snapshot of contemporary conversations that have been part of the context of Trisha GREENHALGH's work. So I do not think the politics is not there. It is not spoken in a very loud voice, but I was trying to give a sense of the political context in which that paper was written. [79]

Liz STANLEY: It is when we are brought up against other academic formations that there's a temptation to slip into the terminology, the criteria—to use Matti's word—particularly when they are very powerful formations in academic or in political life. Is "narrative" on its way to becoming not complexities but something else, a single canonical story? And are these kinds of boundaries that we are talking about part of that process? [80]

Phil COHEN: Life histories can have many starting points, but life story historians tend to arbitrarily choose one. Maybe the issue is in trying to write the history of an idea, or of an intellectual movement—it can have many starting points, but the account you give of it almost by definition chooses one, and the one you chose is because of the kind of argument you want to make about it, and the way you position yourself within it. But I did not take from Matti HYVÄRINEN's presentation that one could not in principle hold a model in mind that there were numbers of possible starting points. The account you gave was rather focused by Anglo-American debates, and for example in France it's been a rather different story. It has had a different trajectory. You use Paul RICŒUR because Paul RICŒUR travelled to America, spent a long time in America. There are also those issues of translation. And then, finally, generational issues come into play, because partly what tends to happen when people give accounts of, if you like, the genealogy of the discipline, is that each generation tries to define itself in a way by rubbishing the patriarchs; the elders, they get written out of the script, or they get critiqued at best, at worst they are just not there any more, because this new generation needs to define its own place in history. So it begins with "us". It begins in 1980 or it begins in 1968 or whenever it is we are defining ourselves as this new Zeitgeist. [81]

Wendy PATTERSON: Narrative is ubiquitous, it's everywhere, suddenly everything is narrative. I get very concerned when I read a paper, which said it was a narrative, it was using narrative analysis, and I could not find any narrative anywhere in the paper; not even in the broadest sense. It is ok if people want to use the term, if they define what they mean by it. I know what I mean by narrative, so I get disturbed when I think people are using narrative discourse and claiming to be using narrative analysis when I can't see anything of what they are analysing as narrative. [82]

Brian ROBERTS: I am puzzled by the term "narrative turn". I have been puzzled by it for some time, because "turn" can mean various things. "Turn" can mean a glance, or it can mean a sharp change of direction, or it can mean a bend, or can mean something more revolutionary which is fundamental to the discipline in that you are changing the way in which you gather and interpret data. If you have a comparative view, then the term looks rather different. It is a hindsight term; something looking back and constructing. To me it is not a very useful term. It needs something to be defined within it. The history that Matti HYVÄRINEN gave, the two positions: there is sometimes much more of a commonality between them, an underlying structuralism. There is a certain lack of notion of time within both perspectives. One possibility about the combination of narratology and narrative analysis is to move beyond the reliance on story as such to a much broader view of the individual in performance, embodiment, and so forth. A much more aesthetic, open view of the way in which stories are constructed—a composition—is that people compose themselves along a range of registers. [83]

Corinne SQUIRE: I was wondering if the different heuristics of our dialogue that emerged have much in common. In Trisha GREENHALGH's work there is a notion of reasoned generalisation; that's the surface mode of that dialogue from the particular to the general in her work. But as Karen HENWOOD indicated there is also a very strong subtext of building research community from below in her work. I would not say it is a binary divide, because those things do inform each other. Then we have other models of this dialogue in Margie WETHERELL's paper and the interpretations that we developed of that, as allowing limited forms of generalisation where there were certain kinds of hermeneutically established commonalities. I am wondering how commensurable those different models of dialogue are. [84]

Karen HENWOOD: There are a lot of different ideas about why qualitative research is important, but they're not getting communicated very well. And that is been part of the initiative in the ESRC to provide more resources to qualitative research. It is to make more explicit these different ideas or models about what is important to research, so that more people can understand it. Scaling up for me means breadth, reach, making connections with more people to understand the relevance of the work. The more ideas we can proliferate about understanding the impact of qualitative work and the better we can communicate it, the happier we should all be, or I would be. [85]

Jens BROCKMEIER: I feel very sympathetic to Liz STANLEY's unease with the idea that there is an overarching definition that describes or prescribes: "This is narrative" or "This is methodology" or "This is what you're supposed to do". I see it more like Marilyn STRATHERN. Whether you like it or not, there is "objectivity" out there in the world, that is, there is a discourse of objectivity, and there are certain powerful definitions organising and regulating our work. How can we nevertheless try to find ways to develop what we could call qualitative or narrative approaches, or maybe approaches more adequate to understanding the phenomena, the fullness of life? Where and who is the person or institution who defines, for example, "What is narrative"? We all know that there are complicated hierarchical structures defining what a method is. What is the methodology that you are supposed to use if you want to get, for instance, a grant for something? In Britain now we even have a national centre for methodology that defines not just methods but methodologies. Let this centre work for 10, 15, 20 years, then very clearly you can even give an address for who defines the method someone has to work with in order to follow a career. This is not to say that we don't have the possibility of doing other things, but usually people who do other things are not people who participate in this kind of discourse. Mark FREEMAN and others have emphatically, perhaps romantically, referred to poets and writers and philosophers, and he could be right in emphasising that it might be only these "outsider" figures who are able to resist the kind of pressure in institutions to define, for instance, what "narrative" is. I believe that this is also what Matti HYVÄRINEN had in mind when he wanted not to define what narrative is but, rather, as a conceptual historian, to describe what regulates the discourse of and about narrative. Why is it that most social scientists and psychologists do not refer or use literary theory and knowledge, although for two thousand years there has been quite a repertoire of knowledge accumulated about language, and narrative in particular? Matti, as I understand him, wants to reflect on the conceptual consequences of this attitude about "scientificity" which excludes literary narrative thought and practice. So I would understand Matti's project as trying to understand why the "narrative turn" has possibly been forced to ignore so much knowledge about narrative. [86]

Anne CORDEN: I have always been interested in apparently how little interested we (social researchers) are in how much impact what we do and say actually has on people who have the real power and resources to do anything about the kind of things that we are interested in. One of the reasons for this may be because we have the anxiety that the impact is very little indeed. And I am not the least bit surprised about that, because I have had more than 25 years of working directly in policy making, and I have come to the conclusion that we do have very little impact. All the huge amount of policy-related research that we churn out, the concept of it being considered and argued about and acted on in terms of the value of its evidence or argument is not something that happens very much. Policy goes forward in terms of rhetoric and personal commitment and political idealism and energy and career-moving, and picking and choosing from what some researchers happen to pass forward at the right time. We have not looked very much at how our research findings are interpreted or whether they have any real impact on the people who have contributed to our thinking. We were talking about how we were doing the casting and not what it actually means to a person to identify themselves within the role as cast by us. I am trying to investigate the impact of the use of verbatim quotations in reporting applied social research. Increasingly it is thought to be good practice. I am increasingly skeptical about that. What are we doing when we do that? How is it received? Is it received as evidence or argument or does it reinforce our own underlying prejudices—things we have already decided and on which it is nice to see somebody else agrees with us? So we are doing something rather similar to Trisha and going to policy-makers and users of research and teasing out in depth what is the impact on them of reading the verbatim quotation of a research participant. Why do you read it like this? What impact does it make within the overall narrative of the research text? And at the same time as we are doing that, we are taking it back to the people who took part in the research, and we are saying "This is how we have cast you, this is what you have said, and this is how I have described you saying it. Now what do you think?" And that has been a very humbling and enriching experience and it is increasingly making me a bit uneasy. We don't challenge it enough. We are not open enough to be objective about what might be going on. [87]

Matti HYVÄRINEN: In terms of the "narrative turn" I try to put it in quotation marks every time I write because it is a term within the material. I am not using it analytically. "Now we have invented the "turn" and so, join us! And have a fantastic future!" There is a kind of romance in these "turns". When I talk about the criteria of the concepts I have no idea of normativity, but quite the contrary, I am just wanting to open it up. I am interested in studying the varieties, and I look for very personal and even idiosyncratic ways of defining narrative. If you have done something very funny in defining narrative, please send it to me! I'd love it. It is not really the purpose of the project to try to create a line of how we get here, to the point where we are wise and good persons. Finally, I fully share the criticism of foundation stories. It is one of the purposes of conceptual history in the first place to problematise all social foundational stories. But there already is a fairly established foundational story on the narrative turn, when it began, how it took shape, and there are all the separate strands of discussions that continue back as far as you want to go, so there is really no real moment of beginning, I'm sure. [88]

In his presentation on "Science and story" Mark FREEMAN argued that narrative inquiry can pave the way toward a more expansive, and adequate understanding of "science" than is generally found in the social sciences. He addressed the problem of fidelity to the phenomena and presented six paradoxes in which he incorporated key debates which had arisen during the symposium. In a response Jens BROCKMEIER suggested narrative imagination as central in probing and extending real and fictive scenarios of agency. Wendy PATTERSON pointed out the analogy between science and story in that they are both involved with theory-building and meaning-making. She emphasised the idea that writing up the research story may require poesis. The discussion dealt with the proposed paradoxes: the more useful, the more useless; the more art, the more science; and the more subjective, the more objective. It was stressed that science should be looked at as a continuum with no fixed boundaries; understood as only one of myriad possibilities to try to come to terms with the universe. In addition, competing notions of narrative were discussed and questioned. [89]

Drawing on an important essay by Martin HEIDEGGER entitled "Science and Reflection", I want to suggest that narrative inquiry can pave the way toward a more expansive, and adequate, understanding of "science" than is generally found in the social sciences. My takeoff point for the paper is also FREUD's discomforting realisation, early on, that his own case studies read like short stories and that they lacked "the serious stamp of science". Consoled by the fact that "the nature of the subject" was responsible for this and that more traditional scientific procedures "[led] nowhere", he would continue with such writing and, through it, continue to fashion and re-fashion psychoanalytic theory. FREUD therefore arrived at something of a paradox; even while by traditional standards his case studies seemed questionable in regard to their scientificity, it was precisely these studies that yielded the desired insight. Knowingly or not, FREUD abided by what is, arguably, the first and most fundamental responsibility of the scientific enterprise: fidelity to the phenomena. Given the clear and obvious value of case studies and narrative data more generally, it follows that the meaning of science, as customarily conceived, is problematically restrictive and that it ought to be reconceived in such a way as to include, rather than exclude, the kinds of pursuits that psychoanalysts and narrative psychologists more generally have found to be so central to their efforts to understand and explain the movement of human lives. "Science" as usually conceived can distort phenomena and be less faithful to reality than other approaches. Theory must do justice to the phenomena we are studying, be useful for this study. We might choose to, as Marilyn STRATHERN suggested, rearticulate objectivity itself; or we might choose to develop other, more useful, descriptions of what we are doing. [90]

[In his talk, Mark FREEMAN presented six paradoxes in which he wove together not only the issues raised in the written version of his paper, but where he also incorporated key debates which had arisen during the previous days. These paradoxes are summarised in the following paragraphs.] [91]

My argument here is sixfold, and as you will see, each aspect of the argument brings us to a paradoxical idea that I hope will serve, ultimately, to undo some of the binaries Liz referred to. [92]

First, I want to argue that much of the work that gets done in the social sciences, in its aim of being scientifically objective, actually ends up distorting the phenomena and in that sense ends up being less, rather than more, faithful to reality. I might as well offer one more disclaimer here. It also doesn't matter much to me whether we continue to use the word "objective". If we do, then it would be important, I think, to rethink it, to conceive it differently than it often has been conceived. It might also be desirable, though, to use different words, to articulate a different language, for better conveying what it means to be faithful to the phenomena—which, for many of us, concerns what I've called the "living, loving, suffering, dying human being": the Other. Hence the paradox I spelled out in the paper: Insofar as the human being is in fact of concern, and insofar as the first order of science—broadly conceived—is to be faithful to the phenomena, then much of social science hasn't been nearly scientific enough. [93]

The second aspect of my argument follows directly from the first—and that is that narrative inquiry, insofar as it aspires to be faithful to the phenomena, can open the way toward a more expansive, and adequate, conception of science itself than is often found in the social sciences. Let me offer a kind of corollary here as well. This one's somewhat paradoxical too: The more subjectively engaged we are with the phenomena—the more we are connected to the phenomena rather than disconnected, dispassionately, from them—the more likely it is that we'll be faithful to them. And so, the more subjective—which is not to say "private"—the more objective. [94]

This doesn't mean, of course, that we have to go home with the people we study or anything of that sort; it simply means that understanding, narrative understanding, is better achieved through engagement and connection than through disconnection and methodological distance. Along these lines, it might also be mentioned that a very rich area of thinking concerns precisely this relationship between the researcher and the researched—including the broadly transferencial issues involved therein. [95]

The third point I want to address concerns writing up the research story—specifically, to quote Phil, how we might "settle accounts with scientific objectivity ... without turning the researcher into just another unreliable postmodern narrator". Here, then, we might simply ask: What kind of research story, what kind of writing, best lends itself to the aim of being faithful to the phenomena? How are the phenomena to be depicted, "represented"? [96]

Well, in at least a portion of narrative inquiry, I suggest (I have no interest at all in turning this into some sort of universal recommendation), it makes good sense, I think, to move in the direction of a more poetic mode of writing—not for the sake of ornamentation or flourish or "romance" but for the sake of using language, using words, in such a way that they can carry the weight, and the depth, of the phenomena in question. On the first night, you may recall, Phil spoke about poetry as "putting truth back into memorable words". Carole followed up by speaking about "the way a poem can embody what it's about". The poem, she noted, isn't just "about" something. It's not only about content; it embodies its object, and it does so through form. And so, we have the idea of form serving content—the aim being (to quote Carole again), "devising a language that's appropriate" to the phenomena in question, a language that, through poiesis, can disclose the phenomena in a way that might not be possible through more "form-neutral" language. [97]

Now, one might argue in this context that this more poetic mode, oriented as it is to particularity, to the specific, can't approach the sort of generalising function that science is assumed to have. But Phil, you may recall, also spoke of the poetic as "taking a bit of the world and generalising"—or, to put it just a bit differently, revealing the general through the particular. (On some level, Margie's work speaks to this too.) In any event, this mode of writing, I emphasised, will be less oriented toward arguing, convincing, making a definitive case, than toward appealing, suggesting, opening, pointing toward the possible. So, this brings us to the third paradoxical statement: the more art, the more science. [98]

This is largely in service of the epistemological aim of increasing knowledge and understanding of the human realm, in its fullness, its excess—to refer back to Marilyn's talk. But it's also, I suggested, in service of the ethical aim of increasing empathy, sympathy, and compassion. Now, on the face of it, this too may seem like an utterly un-scientific way to go. And it might also seem like it can't possibly have anything to do with policy issues and the like, the world of practical affairs. But I actually don't think this is true. And the reason is that the more that narrative inquiry, narrative writing, can carry this ethical weight—the more it can evoke and can appeal to readers (including policy-makers and the like)—the more likely it is that people will care enough to want to do something on behalf of the people in question. This is a hope anyway. [99]

Now that I'm going out on a limb with all these weird paradoxes, let me offer another one in this context: the more useless this sort of writing is, the more useful it may be. This really is what I take MARCUSE's point to be in The Aesthetic Dimension. Art, he argued, didn't need to try to be politically useful, in the explicit way that MARXist aesthetics often urged. It didn't need to be instrumentalist, or directly oriented toward the political realm, in order to have value. It needed rather to be itself, to be use-less—its very value being a function of its own aesthetic purpose and meaning. FREUD's own "short stories" come to mind again in this context. [100]

One more suggestion, and paradox, that concerns the specificity of narrative as a concept (I offer this especially in view of some of Matti's comments this morning). In much of the narrative writing that gets done, including by some folks in narrative inquiry, there's frequently the idea that narratives are, in a sense, a step removed from reality. Reality, it's sometimes assumed, is what goes on now, moment to moment. And narrative, because it looks back upon experience, through the distanced perspective of the narrator, cannot help but "distort" reality in some way, or at least create a kind of illusory order that "life itself", moment-to-moment life, does not have. Now, there's undoubtedly a distinction to be made here: The reality of the moment-to-moment isn't quite the same as what we see, what we can see, in retrospect, through memory and narrative. But here, I would suggest that narrative, rather than necessarily distorting or falsifying reality, can actually point the way toward a deeper, more capacious view of reality than the one that's often presumed. We might even say in this context, paradoxically, that narrative is more about the real, the humanly real, than reality itself—at least as it's often conceived. And here, to Matti especially, I would say that the narrative turn is in fact an extremely important one—not only because it points to new methods, methodologies, and so on, but because it can allow us to rethink the very idea of reality itself. [101]

This brings me to the sixth and final moment of the argument—one that I hope really moves beyond what might be seen as the binary between science and story. On some level, I want to suggest, narrative inquiry is where science and story, indeed science and art, meet. It therefore has the capacity not only to lessen the distance between the two terms but to "unconceal", as HEIDEGGER might put it, their latent meanings and hidden potentialities. Hence the somewhat paradoxical idea of poetic science (full text version [German]). [102]

I want to examine meaning and meaning construction as forms of human agency. Drawing on notions of meaning, agency, and subjectivity by Jerome BRUNER and Klaus HOLZKAMP, my argument emphasises the human potential to act, choose, and imagine as integral to the human condition. Against the backdrop of this discussion, I am particularly interested in the meaning-making potential of language, particularly, of two forms of language use. One is the discourse of agency, because it brings to the fore the constructive dimension of language. The other is narrative, because it is the most complex and comprehensive construction site of human imagination. I would suggest that narrative imagination plays a central role in probing and extending real and fictive scenarios of agency. [103]

Thinking about the objects of analysis—the pearl shell, the monument—it occurs to me that they are traces of human activity, rather than self-contained objects. In this way we can think of a narrative of personal experience as a trace of a process of living, remembering and telling—always partial, always subjective. However I think we can also draw an analogy between science and story in that they are both involved with theory-building and meaning-making and both use prediction as the basis for action. Personal narratives are not only about past experience. When we are planning for the future, we tell future stories of how things will turn out. We act on the basis of such plans and expectations and, very importantly, we evaluate what actually happens in the light of what we expected to happen—much as with testing hypotheses. A standard scientific approach can distort the phenomenon and is therefore less faithful to reality. The more subjectively engaged we are, the more faithful we will be and, paradoxically, more objectively truthful. Writing up the research story may require poesis in order to reveal the general through the particular so, paradoxically, the more art the more science. The more "useless"—that is, in the sense of resisting being used in an instrumental way—the more useful it might be. [104]

Maria TAMBOUKOU: I am feeling very unsettled. And the reason is all these male philosophical names that were thrown around, and not even one name of a feminist philosopher. So I will try to follow a feminist philosophical line where these questions have already been discussed a lot. I suggest Adriana CAVARERO and her work on narrative and the self. She offers narratives as an alternative project to philosophy, and bases her argument on a notion of the familiarity of the body, so she talks very much about the embodiment of memory. She calls it the unreflective structure of memory, and describes how it creates the basis for the self to become situated as narratible and relational. CAVARERO draws on the philosophy of Hannah ARENDT, who is HEIDEGGERian, and that's the link to our point. She highlights mortality as opposed to death, which is the point of the main philosophical line, and she talks about the birth of new concepts and ideas and her interest is in this birth, and mortality, which she relates to embodiment. There is a whole line of relevant feminist thinking, and Liz STANLEY is part of it. She has talked about feminist methodology and how the self of the researcher is totally embedded in the research practice. I wanted to remind us that there is a whole lot of history of feminist philosophy. [105]

Chrysanthi NIGIANNI: I was struck by the three paradoxes Mark FREEMAN mentioned: The more useful, the more useless; the more art, the more science; and the more subjective, the more objective. I had this feeling of moving into the binary. The question is, why not go beyond this binary division? [106]

Mark FREEMAN: In some ways I tried to use those two terms in each of the paradoxes in my own mind—and perhaps did not do this successfully—in order to undo the binary, rather than to create another one. [107]

Chrysanthi NIGIANNI: But the movement was the reverse? From science to art, from uselessness to usefulness. [108]

Mark FREEMAN: A reversal of the ordinary sense in which we think of those terms, for sure. In other words, ordinarily we think of "subjective" as something that prevents us from acquiring the kind of objectivity that's necessary in order to do science. But I would rather rethink that term entirely. Not in a binary way but in a much more expansive one. To talk about the subjective or the engaged is not to talk about the purely private, necessarily. It is to talk about the relation. It is to talk about one's own engagement. And so you are right to suggest that the way I set it up in each case had a kind of duality, but the aim was precisely to undo those. [109]

Phil COHEN: Obviously what we have been doing has been creating a very hard and fast line between science and non-science. There is much more continuity between what is science, or scientific, and what is not. There is also another development known as the new sociology of knowledge. There is no one poesis of science, but a number of different poetics of science, or sciences. The poetics of a particular science are about a particular way of doing the observation that remains faithful to the objects specific to that science. So there would be a specific poetics of physics, which is faithful to the physical world; there would be a poetics of sociology, which is faithful to the social as an object; there would be a poetics of psychoanalysis that was faithful to the unconscious as an object. So these would be different and they would not all conform to the same sort of model. It would seem implausible for that to be the case. Some of the implications of this afternoon have been to say that we have suffered under the heel of hard science for so long; for so long social science was supposed to conform to that. Now the boot is on the other foot in a way, and we are saying that we can draw on the models from poetry, literature etc. [110]

Mark FREEMAN: You are referring to the poetics of particular disciplinary pursuits, and I do not think that that is the direction I was trying to articulate. The other concern that you have seems to be one of trespassing by carelessly importing language into a domain. That seems to be not an argument so much against trespassing, as much as using imprecise language. I am using the idea of the poetic in the broad sense, addressing the idea of what it is that poets seem to want to do, which is to articulate a language that is somehow appropriate to the phenomena in their fullness or in their excess. I do not know if it's a stretch to move into that direction, although admittedly it is something that one has to do carefully. [111]

Matti HYVÄRINEN: There was a historical movement against narrative; structuralist poetics was one of the strongest anti-narrative movements. Only recently has the narrative way of thinking got some air. Now narrative can give a fuller and more complex account of experience. [112]

Mark FREEMAN: There's a lot of work that somehow presumes that, whether by using the language of distortion or embellishment, narratives somehow falsify reality. And that of course presumes that reality is only the sort of sensuous stuff that we have in front of us. But it seems to me that narrative can not only distort or deform but often reveal things that could not be revealed in the flux of immediate experience, in the flux of the now. And so it seems that there are important regions of truth that can be revealed and articulated through narrative in a way that they cannot otherwise be. I hate to relegate all of that to the realm of the purely fictive or the false or whatever, and I think Jens BROCKMEIER's idea of imagination is very important here. Narrative involves imaginative labour. [113]

Lynn FROGGETT: There is an anxiety which is very pervasive in social science that if we go to the arts, we are abandoning ourselves to a state of merger, to fusion, and to the loss of analytic distinction, the loss of a critical consciousness. The problem with a poem is that it doesn't invite the critical consciousness with which we feel all so comfortable. And yet, the whole argument that has been put forward in our discussion does not necessarily mean an abandonment of the critical consciousness to the aesthetic, but perhaps a little bit more unpacking of what is involved in the artistic experience, for instance the visual arts, which can incidentally capture some of those areas of the sensuous perhaps even better than narrative and the linguistic arts. Whether it is the visual or the musical or the linguistic arts, the aesthetic experience is not only an experience of fusion and merger, which is sometimes assumed. It is more likely to be an oscillation between self-abandonment and empathic identification. This is a necessary part of a compassionate imagination: the ability to stand back, distinguish oneself and appraise. I have come to this view through my most recent research project. I started with the problem of how does an organisation become a narratively disposed organisation? Particularly if the people in that organisation are people for whom words are not a natural choice of self-expression, and do not always tell terribly good stories at first sight. And in that case it became clearer watching them do art work that you have this alternation of states of mind in aesthetic experience, which does also include a moment of precise critical appraisal, because you produce something in the world which then has an existence separate from the self. Looked at like that, there is more possibility of a bridge between the social sciences and the arts than seems to be there at first sight, and some feminist philosophers can help us get to that position as well. [114]

Nira YUVAL-DAVIS: Max WEBER said that good sociology is about Verstehen, so the debate is not just art versus science but it also involves empathy, engagement and closeness. As Donna HARAWAY said, "there is no view from nowhere". You have to embody, you have to have a standpoint when you do science as well as when you do art. Those binaries and dichotomies, which were supposed to be cancelled, are (often) reinforced in discourse. Feminist theorists have transcended them a long time ago. Embodiment and the fact that knowledge and imagination are always situated is absolutely critical both in doing social research and in doing art. Sometimes stylistic construction, an obscure but poetic style, comes to replace some very thick contents. It is more beautiful to read poetry and also social science written as poetic if it fulfils what we want it to fulfil from its variety of other resources. We should not fall into the fetish of style. We should look (also) at the substance of what is being discussed. [115]

Corinne SQUIRE: We would really be in trouble if autoethnography was judged as literature or art rather than as part of the research process that it is in the service of. Similarly if you look at DERRIDA's more imaginative writing in the service of philosophy, it is not literature in the literary sense. We seem also to have taken on the HEIDEGGERian real as this kind of immanent "that which will be revealed" which leads to this fidelity ethic. I would rather take a slightly more distanced approach to this, and talk about attention to the real. We have been paying attention to the real of those objects that drop out of or exceed inquiry, or narrative. Doing that does not necessarily lead us in the direction of compassion, which is a dangerous thing to immediately leap to. [116]

Jens BROCKMEIER: There are many theorists who say that there is no such thing as "science". "Science" is an invention of positivist philosophers of science. If you look, as Richard RORTY did, at subjects like material meteorology, theoretical physics, geology, chemical biology, earth sciences, ethology, anthropology, psychology, social sciences without even differentiating these disciplines, we find so many different methodological and non-methodological approaches to what we call the world and our being in it, that it is bizarre to believe this all falls under one single category. So we always have to be aware that when we talk about "science" we mainly talk about a positivist idea or an empiricist abstraction, an idea that started to be formulated around 1870-1890 and that had its peak maybe in the 1940s-50s, the heyday of the modernist episteme. Since then, at least in the philosophy of science, there is no "science" anymore. There are a myriad of different approaches and techniques, of practices and theoretical reflective modes with which we try to come to terms with our life in this universe. [117]

Alexandra GEORGAKOPOULOU: I would like to bring a paradox to the table as well. On one hand we have got so many widely-ranging and even competing notions and definitions of narrative. It is a methodology, it is an epistemology, it is an alternative antipositivist inquiry, it is a communicative mode that has complex interrelationships with time, life, history, past, present and future. In other necks of woods, it's a genre, an archetypal one. It's a social practice, which we didn't say anything about here in this session. We have got all this which one could look at positively or negatively. Is it a sign of richness? Has it made the concept completely loose? Could we treat it as an open-ended totally sensitising concept? But if we do that, what are the implications? So that's one part of the equation. The paradox is that with all these wildly competing notions of narrative, on the other hand we have what I call narrative hegemony. On one hand we have got all those different levels, all that work that narrative does for us, and on the other hand we have a very well established orthodoxy of what actually constitutes social science narrative data, when in fact we have got a wide array of narrative data, cropping up in all sorts of places: verbal, visual, multimodal, etc. Conversations alone cough up all sorts of little and bigger stories that we could look at. It's not a question; it's just a paradox. The more wide-open it is, the more confined it is when it comes to being data. [118]

Molly ANDREWS: I would like to go back to the idea of compassion as being something like "feeling with", because it is at the heart of a lot of things, including methodology. I think of Robert Jay LIFTON's wonderful study of Nazi doctors. He says that his goal was to understand these worlds as these people frame them, and to understand not to condone. It's an incredibly important distinction. It is this question of how you project yourself into a situation that you have, experientially, no or limited access into. That continues to be for us a critical challenge. [119]

Mark FREEMAN: Iris MURDOCH talks about objects of attention, and she thinks about not only other people, but about art and nature. She says the more one is capable of truly attending to otherness, whether in human form or non-human form, the more likely it is that you will be unable to treat a person as a thing. For her, attention is an in-road into compassion. [120]

Srikant SARANGI: What struck me talking about fidelity and attention and compassion—to the phenomenon, to what we study—is the notion of the responsibility of us as researchers. The issue of informed consent comes out in various ways; first when you collect your data and people give you the permission to record. It is very interesting when professionals then come back worried about your interpretation rather than the consent to the data collection, and about what control they have on how you are going to interpret the data. The phenomenon of case notes (and concern about them), is not about the subject, but about fellow professionals and how the case has to be handled in an accountable and responsible way. So how we interpret data, and who owns the data and their interpretation, is as important as what we collect. [121]

Mark FREEMAN: I do not know that being responsible necessarily means that the interpretation that we arrive at as researchers should be able to be "owned" by the people we study. In some cases that is certainly true, but oftentimes we go well beyond what they intend to say, or perhaps their own understandings, and then one has to deal with that responsibility very carefully. [122]

Srikant SARANGI: But then this raises the whole question of what you might call an interpretative licence, like poetic licence. If we do not restrain our interpretative licence, why should we talk about making our work relevant to whoever it is? [123]

Suzanne BUNKERS: My comment is one that is in line with a lot of what some of you have already discussed: What is the difference between the experience and the representation of the experience in the phenomenon of the narrative? And that is something that is key to theory and key to practice in the field of life-writing and narrative from a humanities perspective. So it is not really a question of "Is the narrative a fictive narrative in the sense of being false?" It is much more a sense of a shaped, selected representation of experience that cannot be the experience, but that can give another person an insight into what an experience might be, via the route of narrative. I was curious when you were taking about compassion, because I was thinking about that more in a language of empathy. Are they meant to be seen as synonymous? Are they somewhat different? Is compassion more along the line of sympathy than of empathy? [124]

Wendy PATTERSON: The point Alexandra GEORGAKOPOULOU made about the paradox between all the different things writing can be and the restriction of how narrative appears as data is really interesting. What I have tried to do with the questions of "What is a narrative?" and "What is not a narrative?" staying with that concern about data, is to move away from an idea that narrative is made up of a connected series of events, and to say that it is about experience. An experiential definition of narrative includes any narratable experience, any talk of personal experience as a narrative. For example the whole of an interview where a person is talking about their experience is narrative. But that is only one sort of narrative. There has been a move to increase the range of talk and text data, without ending up in the situation where anything anybody says, anytime, is narrative. [125]

Jens BROCKMEIER: You, Alexandra GEORGAKOPOULOU, have made a nice distinction between narrative being used as a methodology, and narrative as a subject of inquiry. In the latter, narrative is not being used in order to investigate something else, but is a state or mode we have to understand in itself, and that we probably have not really understood yet. Historically we are just beginning to find out what is going on in those processes that we label narrative. This is an interesting distinction, helpful for example to describe my own work. I try to find out what "narrative" is actually all about. I focus on the individual, how he or she makes sense of the world and herself, and how narrative is essential in this business. I believe that without narrative he or she won't find this out. The strange thing of human existence is that it is not just an existence in the real, it is also an existence in the possible. There's a permanent surplus of meaning, of possible meaning. There is this interpretative excess that somehow we cannot understand, which actually constitutes the classic hermeneutic circle. We are never outside of this narrative circle, this project of understanding. Only when we believe that we can look at it from the outside, can the empiricist and positivist illusion emerge. In reality, we are already-always narrative beings. [126]

Mark FREEMAN: Jens, what you had to say before in terms of the great many things that are subsumed under the rubric of science, I think is right on one level. And yet my guess is that many of us in this room still experience a monolithic dimension to the scientific project as it is often conceived in the social sciences. I also want to question the notion that any talk of experience is narrative. There is a lot of talk of experience that does not qualify as narrative. And while I am not going to try to define narrative at this point, I would say that it seems to involve not only talk about experience, but also reflective talk about experience that may not be oriented only to the past, but that may be oriented also to the future; and that may be oriented to narrative sequence but is not necessarily so. Finally, narrative has to do with issues of meaning and significance from the reflective standpoint. [127]

Phil COHEN: What we have been trying to do this last couple of days, was to get a clearing in the dense forest of ideas about dialogue and narrative and a time and a place where people from different necks of the woods could come together and discuss, worry about and maybe discover some common issues. Now of course, there is a risk in that. We have deliberately brought you together from different spaces. There is a risk that you can simply create space for competing monologues or a dialogue of the deaf. What we were trying to do was create a space for a play of difference. What Margie WETHERELL did in her presentation was to bring out that play of difference, and work through it in a very productive way. Although I did notice that identity was used—talk about counting—208 times in twelve different ways. So what's been happening here was the dialogic imagination that you talked about. It does help to have a place and time like this to enable such discussions to unfold, and minds meet and things change because of those encounters. We are not only talking about our conversations with one another, but the conversations we have with those people we call our informants. Looking back at the whole range of contributions we have had, they seem to have fallen into three broad groups. There have been contributions about we could call it the narrative turn in inverted commas, and what that makes possible in terms of dialogue between disciplines. There have been a cluster of contributions about the uses and possibly abuses of psychoanalysis, as ways of unsettling the sociological reasoning that we have all grown up with. Finally, there have been many contributions about the interaction between social science and the wider environment of which it is part. [128]

Exploring questions regarding the larger project of making what social scientists do more meaningful, more useful, and more accountable can be seen as a priority of this event. Problematising such understanding is often partial and pursued at the margins of research projects when time and resources allow. Theoretical debates about problems of method can become dissociated from practical questions of research design, delivery and dissemination. This residential event has explored how these issues play out when research is conducted across different methods, modalities, and disciplines. The form of presentation chosen here, aimed to transmit the lively discussions as vividly as possible. The discussions addressed narrative itself as an index case for methodological debate; methodological considerations of objectivity and evidence, interpretation and context; appropriate levels of research focus and their interactions; the role of dialogue between disciplines; and the interaction between social science and the wider environment of which it is a part. Throughout the discussion questions such as the following were asked: Who and what is social research for, and whose voices does it represent? What are social researchers' and participants' interpretative rights over their data and each other? How does thick description and the rich social interpretation it affords, relate to the need for precise methods of explanation and generalisable conclusions? What special problems of research design or delivery arise when attempts are made to "empower" informants, to enable them to interrogate, and even co-construct the research story? This report does not provide conclusions but traces the discursive dialogue. The report ends here, but the conversation is alive and ongoing. [129]

1) We are grateful to the ESRC for its support for this event, and for the series of seminars that preceded it. <back>

Paul Sebastian RUPPEL is student of psychology at Freie Universität Berlin. He is co-editor of the German "Journal für Psychologie". He is especially interested in qualitative research methods and narrative and postmodern approaches.

Martin DEGE is PhD candidate at Clark University, Worcester. He is co-editor of the German "Journal für Psychologie" and founding member of the Clark University Radical Psychology Group. In his current work he is focussing on the story, history, and philosophy of psychology as well as the institutional processes that constitute the field of psychology.

Molly ANDREWS is co-director of the Centre for Narrative Research, University of East London, author of Shaping History: Narratives of Political Change (Cambridge University Press, 2007) and co-editor (with SQUIRE and TAMBOUKOU) of Doing Narrative Research (Sage, 2008).

Centre for Narrative Research, School of Social Sciences, Media and Cultural Studies
University of East London
4-6 University Way, London E16 2RD, UK

Corinne SQUIRE is co-director of the Centre for Narrative Research, University of East London, author of HIV in South Africa: Talking about the Big Thing (Routledge, 2007) and co-editor (with ANDREWS and TAMBOUKOU) of Doing Narrative Research (Sage, 2008).

Centre for Narrative Research, School of Social Sciences, Media and Cultural Studies
University of East London
4-6 University Way, London E16 2RD, UK

Ruppel, Paul Sebastian; Dege, Martin; Andrews, Molly & Squire, Corinne (2007). Tackling Problems of Qualitative Social Research: A Conversation. Conference Essay: Methods in Dialogue [129 paragraphs]. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 9(1), Art. 41, http://nbn-resolving.de/urn:nbn:de:0114-fqs0801416.

Forum Qualitative Sozialforschung / Forum: Qualitative Social Research (FQS)

ISSN 1438-5627

Creative Commons Attribution 4.0 International License