Psychoanalysis in social psychological research

I'm increasingly impressed with how radical a thinker Freud was, now that I appreciate what it takes to excavate anxiety-provoking experience for the purposes of understanding. Freud identified with 'others' – notably women, homosexuals and 'mad' people – without the discursive support that we have today. However, the main reason for my continued passionate engagement with psychoanalysis is that its insights appear inexhaustible. It continues to challenge and advance my psychological understanding even after all this time.

I am a social psychologist and have no clinical training. I have learned what I know of psychoanalysis from reading, from two short spells of non-intensive psychoanalytic psychotherapy, and from using its insights to illuminate the actions and relationships around me. I was not encouraged in this as a student in psychology; on the contrary in my first undergraduate year, the unconscious was dismissed as impossible to disprove (and therefore unworthy of scientific attention). I persisted because I experienced the difference between knowledge that left me feeling uninterested and psychoanalytic ideas that excited me. Now I know that psychology and psychoanalysis parted company over what constituted acceptable scientific method.  

I was in my late twenties when psychoanalysis became contentious within second wave feminism; a time when the personal was political and I was asking how gendered identity changes in relation to changes in social practices. I still think that social scientists writing about social change need to use psychoanalysis to save them from social deterministic accounts. For example, the idea of unconscious identifications within (and beyond) families helps to theorise inter-generational transmission and the cultural continuities which ensure that identities are not simply the product of current social conditions.

Post-structuralist theory revived questions of how subjectivity should be theorised, which psychology's notion of the individual had tended to take for granted. It featured a systematic critique of the individual of Enlightenment thought – unitary, rational and masculine – which was also the individual of scientific psychology. Critical psychology, in which I was involved from the early 1980s, was centrally concerned with what to put in place of this individual subject. Freud's 'decentring' of the subject through the concept of the unconscious provided us with a key and a combination of psychoanalytic and post-structuralist concepts led to a group of us 'changing the subject' of psychology.

One of the ideas that I developed in my co-edited book Changing the Subject took the idea of subject positions in discourses and applied it to how people talked about sexuality. Why did people take up some discursive positions and not others? The power and gendering of certain discourses was only a partial answer, and we looked to Freudian ideas for the rest. I used the term 'investment' to open up questions of motivation, affect and unconscious conflict. The discourses had been widely taken up in feminist qualitative social science, but without this discussion of investment the 'discursive turn' seemed distinctly deterministic.

Psychoanalysis was itself undergoing huge changes in the late seventies and eighties, taking a 'relational' turn coming from three directions: feminist, British object relations and US relational psychoanalysis. 'Drives' and 'instincts' were associated with unchangeable 'nature' and therefore with women's oppression. Psychoanalysis was recast by many academics (those who engaged in it at all) in terms of the formative influence of real relationships. The mother-daughter relationship became a central object of intellectual inquiry (how could daughters grow up not to reproduce their mothers' oppression?) and Oedipal dynamics were often rejected (with the influential exception of Juliet Mitchell's 1974 book Psychoanalysis and Feminism).

Mothering became an influential topic where feminism and psychoanalysis meet and I, like many other women, was drawn to it when I became a mother myself. The wonder of how motherhood changes women, not just the initial event but over and over again, has fed into several of my academic projects, including my latest book The Capacity to Care and a current research project, with Ann Phoenix, on the identity processes involved in becoming mothers. Both are examples of using psychoanalysis to inform questions that have been addressed in psychology and social science. Part of such projects is to make psychoanalytic ideas accessible to those outside the psychoanalytic community. I worry that my work will be too simple for the psychoanalytically trained and still off-putting for social science students and academics.

So far I have talked about ontology – how psychoanalysis has influenced the theory of my work. But the way psychoanalysis approaches knowledge of subjectivity has radical implications for methodology and methods across the social sciences, particularly in psychology. In the psychoanalytic tradition, knowing and thinking is understood more affectively and relationally than cognitively. I now turn to how the use of psychoanalytic principles has transformed my approach to epistemology and method.  

Qualitative empirical research has had disappointingly descriptive tendencies in the last two decades, fostered in part by worries about the potentially oppressive power dynamics involved in making interpretations of participants' statements without their approval. Psychoanalysis tells us that we (researchers and researched) are not transparent to ourselves, that our conscious and unconscious mind may well be in conflict, that anxiety and its related defences are part of the human condition. In my book with Tony Jefferson, Doing qualitative research differently, I developed an interview method based on the principle of free association, in order to produce data that did not reproduce the rational intentional and rather coherent subject which is the qualitative and psychometric researcher's stock in trade.

However, it is not straightforward to apply to empirical research either Freud's concept of free association or the whole method of psychoanalytic interpretation, developed as they were in a clinical setting. For example, how do you know if an interpretation has validity if it is not made within the analyst–patient encounter? There, the analyst can judge the value of the interpretation based on the responses of the patient. In interpretation of research data, we have needed to devise different devices, such as group interpretation methods, clinical-style supervision to get insight into our countertransferences, and the triangulation of different methods of data (or data from different parts of a participant's account).

The Freudian concepts of transference and counter-transference and the rich subsequent history of these concepts arising from clinical experience have much to contribute to the methodological debate about researcher reflexivity. How should reflexivity be understood now that much of social science and qualitative psychology have been convinced that there is no 'view from nowhere' on which researchers can rely in their pursuit of knowledge; now that the history of science and social science has been revealed to be shot through with prejudiced knowledges about subordinated groups? As researchers we try to minimise class, 'race' and gender differences (and others) between researchers and participants, or at least guard against their effects. Psychoanalysis can remind researchers of all the idiosyncratic ways in which unconsciously researchers will project their own issues onto participants, both in the face-to-face relationship and in data analysis. It also offers ways to help bring these to awareness so that they can serve as a resource and are less likely to compromise the research. The concept of transference (the unconscious projection of our feelings and meanings onto the other person) is an example. It enables researchers to become more aware of the difference between what belongs to the participant and what belongs to the researcher. Various methods (as above)
can be used to support this awareness.

If objectivity is rejected, are we left with relativism? I have used psychoanalytic concepts of objectivity (for example, Winnicott's notion of subjective objects and omnipotence as an intrapsychic obstacle to accepting reality) to try and develop a post-positivist concept of objectivity. The fact that psychoanalysis has been through its own critical turn means that such debates are now to be found in a clinical context.

There is still some way to go in incorporating psychoanalytic ideas into research methods. Even my Free Association Narrative Interview did not go far enough beyond the 'individual' research subject who presents through the spoken word as self-directed, consistent, rational and largely autonomous. It was hardly adequate for inquiring into embodied subjectivity either (something else that psychoanalysis has been good at). So my current empirical project on maternal subjectivity involves using the psychoanalytic method of (infant and child) observation. Because the method was formed to understand babies' development, trained observers tend to be very good at 'reading' bodily expressiveness and identifying unconscious intersubjectivity in first-time mothers too. Careful description precedes interpretation and the group seminar acts as a check on the observers' use of their own subjective response within the research setting.
Finally, social science methodology has been based on the idea of the representative sample. Psychoanalytic knowledge is based firstly on the single case. Increasingly my ontological and epistemological choices have led me towards in-depth analysis of relatively few cases. This too is part of my ongoing agenda in drawing on psychoanalytic principles as a psychologist and methodologist.

- Wendy Hollway is Professor of Psychology and co-director of the Research Centre for Citizenship, Identities and Governance at the Open University.
E-mail: [email protected].