Special issue - Dialoguing across divisions

Tony Manstead and Margie Wetherell introduce a special issue on the state of UK social psychology, with lessons for all.

UK social psychology is highly distinctive – nowhere else in the world is the field of academic social psychology so clearly divided into different camps. Major differences in approach and perspective have been a source of division and strife. The most obvious one is between those who advocate a broadly positivist and quantitative approach and those who adopt a broadly qualitative approach. Like any crude classification, this does not do justice to the varieties of social psychology that are practised either side of (and in some rare cases across) this division. But it does capture the primary dividing line.
Mostly, these days, it is no longer a case of actively warring factions. This peaceful coexistence is welcome and has many advantages, but it has some problems too. We believe that more can be done to create a working environment based on understanding, mutual respect and some common goals. There is a new dialogue emerging, which works across these old divisions, and this special issue describes such a dialogue. We are all social psychologists working in British universities who have come together to talk to each other about the state of UK social psychology. Here we explore new and constructive forms of engagements and set out our conclusions.

Then and now

But first, what do we mean when we talk about warring factions 'then' and peaceful coexistence 'now' in social psychology. To make sense of this we need to sketch the recent history.

As any social psychologist will tell you, where there are two or more groups there is potential for conflict. This conflict is greater when one group begins in a position of dominance and defines the playing field in terms of access to journals, acceptance for conference papers and jobs. It is probably true to say that at the beginning of the 1980s quantitative social psychology was in that position. This position was challenged, however, by an initially small but rapidly growing group of social psychologists who were developing new qualitative approaches. These approaches have led to the variety of cultural, critical, psychosocial, narrative and discursive psychologies evident today.

The conflict was fierce at times as the 'upstarts' (as seen from a quantitative viewpoint) struggled with 'the establishment' (as seen from a qualitative viewpoint). Voices were raised, conferences were full of challenging but fascinating methodological and theoretical disputes, and there was competition for power and influence. This phase of conflict was followed by a period of stand-off, starting in the mid-1990s and lasting until now. Quantitative and qualitative social psychologists largely went their own ways. They have their own journals, books and academic networks, and more often than not they work in different institutions. Although they may be physically present at the same conferences, in reality there are often two mini-conferences operating in parallel, each appealing to its own constituency.

What, one might be forgiven for asking, is the problem with this state of affairs?

A state of benign indifference between disputing groups sounds acceptable, if not particularly fruitful. Why does it matter? Why should we try to do better? There are several answers to these questions:
l    The division between the quantitative and qualitative approaches has become so ingrained that there are many social psychologists in the UK who are simply unaware of what is happening in the other camp. Far from there being any creative complementarity between the two approaches, there are in effect two parallel social psychologies.
l    Then there are practical issues to do with the evaluation of research quality. If researchers in each camp lack understanding of the other camp's research methods, how can they arrive at informed assessments of the quality of that research? This is an issue that arises in reviewing papers, research grants and RAE submissions.
l    Being viewed by the world outside as internally divided is unhelpful to the identity of social psychology and is probably damaging in terms of academic reputation and influence.
l    There are also consequences for undergraduate and postgraduate training. How desirable is it that students are being trained in one set of methods without being trained in the other? Whichever side one is trained in, the result is 'partial' in both senses of the word. The danger is that the parallel universes of quantitative and qualitative social psychology will be sustained in future generations.

This special issue is about how to move on from this impasse. It is about recognising and naming the current state of affairs, working towards a rapprochement and formulating a more secure and healthy UK social psychology.

The articles that follow are one outcome of a series of seminars titled 'Dialoguing Across Divisions in UK Social Psychology', organised by Wendy Hollway and Tony Manstead with funding from the BPS. Selection of a small group of participants from different corners of social psychology combined with careful attention to group process (facilitated by the group analyst Sheila Ernst) and multiple meetings within a short space of time led to a very exciting set of exchanges. These exchanges clarified similarities and differences, and the first two substantive articles in this special issue pick up that theme.

In the opening article three strong programmatic statements illustrate some of the divisions in social psychology. Next, Stephen Reicher and Stephanie Taylor put these similarities and differences in a broader context, and discuss what is behind them. Then Alex Haslam and Brian Parkinson develop one analysis of how we, as social psychologists, might understand our own situation and deal with these divisions. Our seminar series led to greatly increased mutual respect and understanding among the participants and produced a set of suggestions for 'where next' – the special issue ends with this manifesto. We hope that much more will flow from these beginnings.

Lessons for all

If you are not a social psychologist, please don't stop reading now. The main issue – divisions within psychology and how to understand and work with them – is relevant to all. As psychology grows in popularity in all arenas it is vital that our diversity is seen as a strength, not something pulling us apart from within.

- Tony Manstead is Professor of Psychology at Cardiff University. E-mail: [email protected].
- Margie Wetherell is Professor of Psychology at the Open University. E-mail: [email protected].

Weblinks

Social Psychology Network online database: www.socialpsychology.org
Papers on Social Representations – online journal: www.psr.jku.at
Discourse analysis information: www.lboro.ac.uk/
research/mmethods/resources/links/da.html
ESRC Identities and Social Action Programme: www.identities.org.uk

Three views on hate

Russell Spears, Wendy Hollway and Derek Edwards tackle one topic from several angles to identify core differences between camps. 

On 24 May 2004 The Guardian published a story about an Iraqi family, a mother and her children. The woman's husband, and children's father, had died mysteriously in detention during the American/British invasion. The newspaper headline quoted the woman's response – 'I will always hate you people'. Mrs Izmerly's response, extreme emotions, unequal power and national, global and group conflict are part of the territory social psychology covers. We want to use the story as a way of exploring what social psychology offers and to probe the differences between three core approaches – the experimental, the psychoanalytic and the discursive. Each author speaks from their particular perspectives within each of these broad approaches.

Experimental social psychology (Russell Spears)
Where do you begin to devise an experiment to gain insight into this headline? This reminded me of a conference I recently attended called 'Why neighbours kill', an interdisciplinary meeting of political scientists, sociologists and social psychologists. After diverse talks on genocide in Rwanda, the Balkans, and Cambodia, by people who had actually been there and talked to the survivors, I was starting to feel a bit uneasy that my little package of 2 x 2 designs (with no killing in sight) might come over as ever-so-slightly trite. This was not helped by a well-known 'experimental' social psychologist (who shall remain nameless) who boldly announced before my talk that he would not insult the audience by presenting experiments on such a grave topic.
So what is the use of experiments on such extreme emotions? First, however real and high-impact these examples are, they remain what psychologists often disparagingly call 'anecdotal'. There is a serious point here: empirical evidence is the lifeblood of psychology, and experiments provide the control to assess causal relations and patterns among variables that may not be apparent to the naked eye. Although few would doubt the evidence of exclaimed hatred from the headline, this is only the first step to understanding it.

I will try to explain why I think an experimental approach can be useful, and perhaps even necessary, to answer key questions about feelings of hatred. Let's take the research I talked about at the conference, conducted with Colin Leach, on the topic of Schadenfreude: pleasure felt at another's failure or downfall. Although not the same as hate, Schadenfreude is closely related and can be fuelled by hatred. It can explain pleasure in an enemy's demise, and the failure to stop, or even the tendency to participate in, some of the extreme acts associated with conflict.

However, Schadenfreude is rarely openly expressed like hatred. Nietzsche noted that Schadenfreude is an opportunistic emotion that relies on a third party for the rival's demise, and this can make it less legitimate, for example, than direct victory over the rival. Open gloating in such circumstances can also be dangerous if the rival retains power (hence Arafat's concern at the open gloating of some Palestinians about 9/11). So how do we detect Schadenfreude if it is not legitimate to express it? Experimental techniques can help us.
In our research we have employed a 'bogus pipeline' technique in which we attach a sensor to participants and tell them that this can detect their true emotions (rather like a polygraph). Although this is not actually connected to anything, we find that people are more honest and show more Schadenfreude towards a hated rival as a result. More generally, experiments allow us to detect causal relations (e.g. that threats to identity can incite Schadenfreude) and meaningful patterns among variables (e.g. that the pain caused by threats to identity predicts Schadenfreude).

Or course we cannot reproduce in the lab the conditions that foster the kind of hatred that motivates some people to become suicide bombers (nor would we want to for obvious ethical reasons!). However, we can model some of the proposed processes and test implications of theories. In other research we have tried to show that the disempowering conditions of stable low status can be associated with more aggressive forms of discrimination. This reflects a 'nothing to lose' strategy of the hopeless and helpless, epitomised by our headline.

The point about experiments is that they offer depth of explanation, enabling us to dig beneath the surface, and to investigate the psychological processes that are not always visible or accessible in direct accounts. Experiments are particularly good at getting at the parts that other methods cannot reach: they are useful in telling us things that people either don't want to reveal, or can't. This may be because they are ashamed to admit to them (e.g. malicious emotions like Schadenfreude) or are not even aware
of them (e.g. unconscious thoughts and desires, or patterns of behaviour that are only apparent at the group level, through the lens of the experimental panopticon). As with all methods, these need to be treated with interpretative care, but they provide explanation that goes beyond conscious accounts and surface appearances.

The claim that experimentalism is the only show in town is a dangerous one, however. While some advocates of the experimental approach are wont to claim scientific superiority, this is not inherent in the method itself, and experiments (or quantitative psychology more generally) can be seen as complementary to other approaches. To confine oneself to experiments is surely partial and unhealthy (a bit like a dietary fad). Returning to the conference example, it would have been foolish to see myself in a contest with political scientists about the causes of genocide in Rwanda. To admit only experimental evidence here would be absurd. But to deny their utility when they can sometimes 'tell us more than we know' is equally ludicrous. Rather, we had different parts of the jigsaw, relating to different levels of explanation, and the choice of methods was contingent on this analysis. In the spirit of methodological pluralism (or unholy alliances), experiments could complement discursive approaches (by digging beneath the discourse) and supplement psychodynamic approaches (by uncovering unconscious processes).

Psychoanalysis (Wendy Hollway)
I stare at a headline: 'I will always hate you people' and monitor my feelings, grabbed by the picture of Mrs Izmerly and her three children. It conjures in me a knot of knowledge, belief and feelings about the Iraq war. I am furious with Bush (and Blair) and regularly have a tussle between my hate and my better judgement when it comes to how I feel about American people in general. I feel guilty when I read in the main text that the daughter accuses all British citizens, as well as American citizens, of being complicit in this war because we live in a democracy. There are some powerful group constructions going on even in the six words of the headline and I do not – cannot – stand outside of them. I am British, in this instance to my shame. I start here because I believe that social psychologists should reflect on their own subjective responses to any issue on which they conduct an inquiry in order to clarify where their commitments might lie.

It is not only my meaning frames that will shape the analysis of this headline however. If the journalist did not choose the words, he certainly framed the story, and it is always within a frame that meaning is achieved. I know immediately that this family of an Iraqi man were expressing their hatred of what the coalition was doing in Iraq. After reading the article, I could see that this case was framed in terms of the terrible effects that such treatment has on the reputation, acceptance and, ultimately, purpose of the coalition's presence in Iraq. The theme of hatred in the main heading was mirrored in the final sentence quoting one daughter: 'I won't allow myself to rest until I have got revenge for him.' Meaning and interpretation are co-productions; in this case the interviewees, the journalist (perhaps editors) and me.

Psychoanalysis is one of the few theoretical perspectives in social psychology that does not shy away from hate, understands it and takes it seriously. There is no established psychoanalytic method in social psychology because psychoanalysis is a clinical method not a research method. I have used psychoanalytic principles to inform my understanding of individuals (ontology) and of how research can come to know them (epistemology). There is no single accepted way of doing this, but it informs my mode of interviewing (see box).

If I wanted to do research on this topic, how would I go about it? What research question would I choose of the many possibilities? Each is made possible by the salient ideas in a given approach. Social identity theory, for example, will be interested in the construction of group identity in the 'we', or 'I' and 'you people'. A social psychoanalytic approach takes account of the feelings and investments that are involved in such constructions – the inner psychic contributions to meaning – and recognises that these are negotiated in relational, discursive and wider social contexts. So here, hate is a dynamic kindled in Iraqi-Anglo-American relations in the current context of the actions of coalition forces in Iraq. Of course this is a gloss on a situation with a very complex history in respect of relations between Islam and Judaeo-Christianity, but it informs understanding of the affective loading on a category like 'you people'.

In summary, the research question is not an innocent, neutral tool but an intervention that already carries a payload of meaning that will shape the knowledge produced from the research.

Until the research question is clarified, decisions about design are premature. It affects what we take as the unit of analysis. It could be the headline, the whole media text, the existing interviews from which the story was constructed, or new interviews with this family or other families specifically set up for research purposes.
Should it be based on a single-case interview? This question raises the issue of how extrapolation ('generalisability') can proceed from one or few cases. On the other hand, how many interviews is enough? What is the justification for needing a number that is amenable to statistical analysis? Is it that one case does not provide 'proof'? In the case of this extract, the phenomenon of Iraqi hate comes as no surprise. If the question is how widespread it is, the design goes in a survey-based, quantitative direction. If it is what makes some Iraqis hate and not others, the design must be comparative.

My purpose is to understand more deeply what it means to 'hate you people', how it has come about and its likely effects. Such questions require qualitative methods because only these can understand experience and meaning. One case will provide the depth and can be extrapolated using theoretical understandings of hate and group relations as long as this is restrained by careful contextualisation.

This makes it appropriate, in principle, to conduct one or more in-depth interviews to establish the specificity and detail of this woman's hate in the context of her life history and especially of the treatment of her husband. In practice, there might be political and cultural barriers. The analysis would then take what I call a psychosocial direction, which means I would look at how Mrs Izmerly's account of her experience (never separable from emotions) draws on actual events and makes something unique of them in her inner world. In terms of generalising from single-case data, the approach is one of theoretical extrapolation.

This method provides data concerning the complexity of a person's meanings and their relation to specific experiences. Like experimental social psychology, it goes beyond 'conscious appearances and surface accounts', although incorporating a complexity and attention to particularity which that cannot achieve. Unlike the discursive approach, its focus is the person who speaks, rather than the text, which in my view is a central location for emotion.

Discursive social psychology (Derek Edwards)
My first reaction to the headline, and the story beneath it, is that of an ordinary reader. It is a powerfully evocative report. But rather than exploring my emotions, or developing my stance on Iraq, finding people to interview, or devising experiments on how emotions are caused, I start to get interested in the report itself. This is not a pursuit of deep, underlying significances, but rather, of how specific words, descriptions and accounts are assembled and put to work.

A common objection to discursive psychology (DP) is that it only analyses discourse, when there are other, more important things to do. We are turning away from the events themselves, whether in the world or in the psyche – in this case death, politics and hatred. Yet to take an immediate interest in those matters is also to turn away from the actual object presented for analysis, the newspaper report, which is also real. Discourse is both real and important. If it were not for discourse there would be no politics, no war in Iraq, no understanding of what is happening there, nobody to quote, nothing to say. So there is no immediate requirement to use the report as a point of departure and do some other study instead. Our immediate focus is on reports themselves, how they formulate the nature of events, how they provide for causal explanations, invoke psychological states (see box), and build implications for politics and policy.

Out of such an analysis may come further questions, and the need for more materials. But those materials will probably be more discourse rather than an experiment, survey, or even a depth interview designed to probe the psyches of the participants. In DP there is a preference for collecting discourse as we find it, rather than doing research interviews. The reason is a basic conception (and observation) of how discourse works. Everyday discourse deals not only with its obvious topics, but also with the conditions of its production.

It is always situated, indexical, sequentially relevant, always of and for its context, and always doing something. Research interviews are the basis of a great deal of qualitative research, especially where the aim is to discover how people think on some chosen topic. But interviews inevitably usurp the circumstances in which people ordinarily say things in and for living their lives.

Apart from uses of emotion words, and other items from the commonsense psychological thesaurus, DP examines how psychological business is generally handled and managed when people talk together. One specific topic, again starting with our newspaper headline, might be how direct quotation works ('I will always hate you people'). What are the general characteristics and uses of direct quotation? What does it do? In what discourse contexts, at what junctures, and in the performance of what kinds of activities,
do people actually produce quotes of what other people say? Does it have regular characteristics, types, functions and occasions? In fact there is already plenty of work on this (by writers ranging from Goffman and Bakhtin, to the detailed conversation analyses of Elizabeth Holt and Robin Wooffitt), and it turns out to be more interesting and systematic that we might imagine. Rather than formulating, as independent variables, a range of theoretically generated types of quotations and measuring their effects, we find order in the ways that quotations are actually used.

From DP's perspective the tendency of psychologists to turn the world into causal factors and variables, and get them under laboratory control, often seems inappropriate. Experimentation conforms to (some) canons of scientific research but it also, especially in social psychology, seems based on an assumption that the everyday world of social activities is actually a complex mix of causal factors and variables, rather like the relationship between a physics experiment and the world outside the laboratory. In DP the social world is already orderly and intelligible precisely because people make it so – that's how it works. Experimental social psychology is also grounded in an understanding of how social life works, but not in systematic observation and analysis of it. Further, DP is not a preliminary, 'natural history' phase of research that will eventually generate experiments. Rather, our methods are adequate and appropriate to the phenomena. The orderliness we find is that of actions oriented to norms, rather than effects stemming from causes. It seems weird to treat discourse and social interaction in the same way as one would
a chemical reaction.

- Russell Spears is Professor of Psychology at Cardiff University. E-mail: [email protected].
- Wendy Hollway is Professor of Psychology at the Open University. E-mail: [email protected].
- Derek Edwards is Professor of Psychology at Loughborough University.
E-mail: [email protected].

Box: A psychoanalytic perspective
The free association narrative interview method reaches beyond structured interviewing, which is dominant in qualitative research and runs the risk of constraining interviewees within the assumptions provided by the questions. This method can elicit deeply felt and difficult emotions, possibly conflictual, as well as taken-for-granted issues like identity and identifications, so it fits well with this subject.
To guard against eliciting generalisations (like the headline) or common discourses, questions are open but specific, eliciting a narrative grounded in actual events. Main questions could include 'Can you tell me about your life under Saddam's regime?' and 'Can you tell me about your husband's life?' The simplicity of these questions belies the complexity and richness they give rise to in the context of an attentive and respectful relationship with the participants.
The resulting narratives are developed by follow-up questions following the ordering and wording of the interviewee, based on the principle that the researcher should elicit participants' experiences, meanings and free associations, imposing as little as is possible of their own. Because meaning is achieved in the context of the wider whole (the gestalt principle), the material is not broken up (as is common) for analysis. Analysis of data involves, among many other things, noticing signs of the affect and potential conflict interviewees show in their narratives.

Box: examining everyday emotion
Discursive psychology (DP) examines, among other things, how people deploy commonsense psychological ideas. Rather than taking those ideas out of context and finding that they amount to a messy, contradictory and inaccurate theory of mind, we explore how people actually put them to use in their everyday lives, when accounting for actions and events.
In a relevant study of emotion discourse (Edwards, 1999), talk from relationship counselling was analysed along with newspaper reports in the aftermath of Princess Diana's death. A Sun editorial said: 'In the depths of his grief, Diana's brother is entitled to be bitter about her death.' Analysis focuses on how this formulation selects emotion rather than, say, judgement as his reaction (her brother, Earl Spencer, had produced a heavy rebuke of the role of the press in her death), and names that emotion grief rather than, say, anger. Conceptually, grief's object would be Diana's death, whereas the object of anger would be (in this context) the activity of the press. Similarly, bitter evokes a disposition within Spencer, and perhaps a motive for producing emotive criticisms, whereas anger directs attention to its object and cause – the press and their paparazzi. Very briefly, these kinds of observations reveal a range of functional uses of emotion terms including, in the case of the Sun's editorial, how to deflect attention from a criticism of their own journalistic practices onto the psychological state of the critic. Everyday emotion talk turns out to be very precise when examined inside the real-life practices where it is used, and for which it is surely designed.
Edwards, D. (1999). Emotion discourse. Culture and Psychology, 5(3), 271–291.

Discuss and debate

 Are the three methods described here complementary or incompatible? How should researchers define the limits of applicability of each method?
 Where does emotion reside – in individual persons and their experiences, or in the cultural practices of language and social interaction?
Do these approaches encourage or rule out certain kinds of question as relevant, legitimate, interesting?
Which (if any) of these methods would provide most solace or utility to Mrs Izmerly? And is that an appropriate criterion for choosing one approach rather than another?
Have your say on these or other issues this article raises. Write to our Letters page on [email protected] or at the Leicester address – 500 words or less, please. Or you can contribute to our online forum on this or any other topic – go to www.thepsychologist.org.uk and follow the links.

Download PDF for whole feature.