I stand here in considerable trepidation. My task is to persuade you to do something which past experience may have shown you to be painful, and which most of you see no justification for doing. Why on earth should anyone want to grasp a nettle? Was it the Research Committee’s unconscious intent to underscore the futility of undertaking research in psychoanalysis by issuing this pointless imperative? Most of us would give nettles a wide berth and the same strategy has been widely adopted in relation to psychoanalytic research. My aim here will be not only to ask you to abandon this strategy, but more ambitiously to persuade you to embrace, or at least firmly grasp, this unattractive specimen of flora and enthusiastically go forth in the pursuit of psychoanalytic knowledge through research.

THE INTERNAL FUNCTIONS OF RESEARCH: THE FAULT-LINE

So, why is it that talking about research to psychoanalysts can feel like selling deep freezers to Eskimos? Paul Whittle (in press), the Cambridge psychologist, wrote a recent brilliant target article in Mark Solms’ new journal, Neuro-psychoanalysis, where he described ‘a fault line running down the middle of psychology’, a metaphor which we can easily extend to all disciplines involved in the study of the mind. There are two cultures, not so much opposed to one another but, rather like neighbours in a large apartment building, quite happy to walk past each other for years without even learning each others’ names.

Those in experimental psychology, cognitive neuroscience, neurobiology, human development and other sub-specialties of the ‘science of the mind’ have the benefit of a powerful, reasonably well-funded discipline, which has progressed particularly rapidly over the last quarter of a century, prides itself on a cumulative knowledge base, is strong enough both to generate a range of technologies and to interface with neighbouring disciplines, and is generally acclaimed as a relatively successful natural science. 

By contrast, psychoanalysis restricts itself to personal insight: that is the objective study of subjectivity. Whether understandings gleaned through psychoanalysis can be considered ‘true’ would depend on a generally accepted criterion for truth, which has eluded all students of the mind. In terms of personal experience, which may indeed be all that matters, most of us who have had substantial psychoanalytic experiences can readily testify to moments of genuine recognition which have extended our understanding of ourselves and that we would have little hesitation in labelling as ‘truths’.

For example, in the throes of intense transference during my analysis as an adolescent, I flew into a rage upon seeing my female analyst get into a car with the male psychologist who had carried out my original assessment. As she sensitively and subtly drew out my fantasies about the experience, the profundity and primitive and violent nature of the Oedipus complex suddenly became a reality in my mind. In trying to understand patients, I still frequently recreate this experience to help me appreciate the full strength of their emotional experiences in the transference.

Psychoanalysis provides a range of essential descriptions and constructs without which it would be far harder to talk about our selves or to understand our own and other people’s lives. No doubt, this is why so many psychoanalytic technical terms have found their way into everyday language (for example, ‘Don’t be so defensive’, or  ‘He is foul to her because he subconsciously fancies her’ etc.). 

According to some philosophical readings of Freud - Richard Wollheim (1995) and James Hopkins (1992) for example - a principal achievement of psychoanalysis was the extension of the everyday (so-called ‘folk’) psychology that people generally use to understand each other to an unconscious part of the mind. This would account for both its remarkable integration into our culture and (certainly in my view) for much of its therapeutic effect with certain groups of patients. But this is not our topic for tonight. Even if the roots of psychoanalytic constructs were to be found in the psychological understanding of the average person, there is no doubt that psychoanalysis has deepened and greatly elaborated these ideas to a point where a comprehensive model of the mind is now available to it.

The fault-line cannot be defined simply as the boundary between science and non-science, as many have tried to in the past. Whilst psychoanalysis is probably not a science, by most definitions of this term, the whole question of what is scientific is so fraught, so subjective and so ambiguous, that entering into it here would only obscure the debate. I have always maintained that psychoanalysis, because of its willingness to explore phenomena however painful, unpleasant and distressing they might be, possesses a quality of objectivity which few if any other approaches to human mental life possess. The issue is not whether psychoanalysis can be made into a science either by changing our definition of what science is or by changing the way we carry on our business. Far more important than a mere label, however prestigious, is whether the scope of research work undertaken by us can be meaningfully extended without destroying the precious understandings which numerous generations of psychoanalysts have achieved.

Yet, the gulf or fault-line to which Whittle refers, and where psychoanalytic research to a large measure falls, cannot be doubted. It is, to say the least, peculiar that subjects such as dreams, perception, thinking, the nature of mental representation, and human relationships can be written about so differently from these two perspectives. In fact, there are few in either community who ever read the papers of the other. If others, in either community, accidentally did so, they would share a reaction of shock at the profound misconceptions, puzzlement about the other’s purpose, derogation at the unnecessary jargon, but (above all) a deep conviction concerning the irrelevance and futility of the other’s enterprise. There is hardly any room for debate: for debate, you would need some common ground upon which a battle could be fought. Sadly there is only indifference.

BRIDGING THE GAP

Of course, there are those who attempt to bridge the gap. Stephen Frosh (1997a; 1997b), for example, perhaps one of the most wise of the ‘integrators’ and a most talented theoretician, has subtitled his book on psychoanalysis and psychology, Minding the Gap. He explores how important questions and ideas which have been fully taken up by psychoanalysis have been systematically ignored in so-called scientific studies of mental function. Other integrationists who have been less careful of the distinction are far more likely to be ignored or regarded as somehow ‘unsound’ or ‘inadequate’ by both sides. I would count myself as one of these unfortunates who has fallen into the gap. 

There are a few notable exceptions, and their work is in itself an interesting subject of epistemological study. For example, many of Joseph Sandler’s most important concepts (the representational world, role responsiveness, his writings on the nature of psychoanalytic concepts) have their origin in psychological theory. He and I discussed this on a number of occasions. I surmised from these discussions that his integrationist efforts were successful because, having travelled back and forth across the gap, he transplanted the concepts in toto, never intending for them to retain their links with their points of origin on the other side. Interestingly, the other side does the same. Cognitive behaviour therapy owes much to psychoanalysis. Not that you would be able to guess this from the writings of CBT clinicians.

TWO WAYS OF DOING BUSINESS

So what is the nature of the fault-line between the two approaches to the study of the mind? Whittle points out that, in psychoanalysis, communication - whether in writing or clinical discourse - occurs in terms of its impact upon the reader. As Adam Phillips puts it, paraphrasing Emerson, in psychoanalytic writing there is an attempt to return the reader to his own thoughts whatever their majesty, to evoke by provocation. According to this way of doing it, thoroughness is not inciting. No amount of ‘evidence’ or research will convince the un-amused that a joke is funny (1993, pxix).

In psychoanalysis, we accept that something has been understood when the discourse about it is inciting. Elusiveness and ambiguity are not only permissible, they may be critical in order accurately to depict the complexity of human experience.

By contrast, the other culture lives and communicates in the simplest and most matter of fact way possible. Whittle (in press) cites Wittgenstein’s preface to Tractatus: ‘What can be said at all can be said clearly, and what we cannot talk about we must pass over in silence’. No wonder, then, that subjective experience has largely eluded psychological disciplines other than psychoanalysis. No wonder, then, that psychoanalysts fear that the introduction of research methods from this barren world risks the destruction of the phenomena they cherish. Nietzsche talks of unpretentious truths that have been discovered by means of rigorous method and opposes it to metaphysics that blinds us and makes us happy. Nietzsche here distinguishes boring empirical fact from evocative narrative. Holding on to unpretentious truths demonstrates courage of a different sort to that shown by psychoanalytic investigations of the unconscious. It is a turning away from what is appealing. Whittle calls it cognitive asceticism. 

Cognitive asceticism is of little relevance to the clinician whose principal task is to create a narrative that fills the gaps in a person’s life. Theory has a heuristic value for the clinician. Theories support understanding. They are not bound by the minimalist principles that are the residues of positivism. They are adventurous, they dig deep. They are acts of imagination about how our minds function, that are judged principally according to how well they fit our own and our patients’ subjective experience. This is not to say that the theories are not true, rather that they are metaphoric approximations at a subjective level of certain types of deeply unconscious internal experience. 

There are examples of such theories in other early sciences. For example, the understanding of phonology through metaphor which European grammarians of the sixteenth century developed (distinguishing light and dark vowels, soft and hard consonants, moist and unmoist ones) has been shown to be far from arbitrary. The classification system has been demonstrated by modern phonetics (Fónagy, 1980; 1983) to be based on the actual functioning of the articulatory organs (mouth, tongue, vocal cords) as these sounds are pronounced. This is information of which the grammarians could only have had preconscious knowledge. By analogy, aspects of psychoanalytic theory may be thought of as attempts on the part of the theorist to grasp the nature of the mental processes and mechanisms of which they have no conscious knowledge and which are not available to introspection. We should not accept simplistic critiques of metaphoric thought in psychoanalysis. Science uses metaphor in the absence of detailed knowledge of the underlying process. Provided that metaphor is not confused with a full understanding, or to use Freud’s metaphor, the scaffolding is not mistaken for the building, heuristic considerations outweigh any disadvantages of their use.

Thus while there is wisdom and truth in our theories, they will not behave like theories in modern sciences. Psychoanalytic theories also impact on us at an unconscious level. The particular configuration of ideas fits with an inner experience. We are rich in theory and it is hard to imagine how this richness can ever be reduced either by research or by other methods without also compromising the quality of the fit between a psychoanalytic model of mind and subjective experience.

CAN YOU BE TOO RICH IN THEORY?

However, the very fecundity of clinically rooted concepts is beginning to threaten the clinical enterprise. Psychoanalysts emulating the founder of the discipline take special pride in discovery. This has led to an abundance of ideas in the field. What we do less well is to test these in meaningful ways. We are very vulnerable to charismatically presented new ideas which then come to be pooled in an eclectic purée of clinical strategies and techniques that create increasing problems in the transmission of psychoanalytic knowledge and skills. Sadly, it also makes for a built-in resistance against the systematisation of psychoanalytic knowledge, since those whose frame of reference depends on ambiguity and polymorphy can be threatened by scientifically based clinical reasoning.

Data is, of course, not the plural of anecdote. Psychoanalytic practice has limitations as a form of research. A physician practising internal medicine learns from clinical observations but is not under the illusion of being engaged in research. The physician’s work, we hope (if we are the patient, that is), is however influenced by the results of research, and his or her reasoning will have been disciplined by scientific training. The problem of using clinical experience as research is well known to be one of induction. Mostly, we tend to confirm our theory-based expectations from our patient’s material. Our memory for material is biased, even our perception is impure. We cannot be pre-Kantian objective observers. Our own discovery of the pervasiveness of countertransference denies us this possibility, even in principle.

RESEARCH ON MENTAL PROCESS

So what is the role of research in psychoanalysis? Research has a number of minor functions for us, which I will mention later, but one major one. The clinical aim of psychoanalysis and its firm grounding in the context of personal relationships inevitably pushes theory towards the understanding of mental contents, the key themes of feelings and ideas. The approach on the other side of the fault-line is concerned with mental processes, the way the mind functions, the machinery which gives rise to feelings and ideas. The representational world is the tune which the violins of mental processes generate. This is where psychoanalysts need alternative methods of inquiry. Our roots in folk psychology, our legitimate concern that our theories should hold meaning, not just for our patients and other psychoanalysts but also for the broader social world, handicap us in defining the mechanisms of the psychological world: mental processes. Our current ideas about how the mechanisms of the mind work are contaminated by our need to provide compelling explanations: compelling to our patients and compelling enough for ourselves to be able to guide our thinking in the cauldron of the clinical encounter. Most musicians are not well enough equipped to design a violin.

In a paper that I wrote, longer ago than I care to remember (Fonagy, 1982), I put forward the argument that research studies could not and should not be used to test psychoanalytic ideas. If a clinical observation cannot be replicated in the laboratory, there are a host of good reasons why this might be so. Rather, systematic observations could be used to inform us about the psychological processes underpinning clinical phenomena, which we currently use the metaphoric language of metapsychology to approximate. In the twenty years since writing the paper, I have actually attempted to do work along these lines, first in the area of neuropsychology and then in development. 

IS CLINICAL DATA THE ONLY DATA?

And here we come to perhaps the crux of the debate. To what extent should we allow psychoanalytic clinicians, with their permissive mental sets, to be sole arbiters of the psychoanalytic discourse? It has been powerfully argued, for example, by Andr¾ Green (in press) and Peter Wolff (1996), that free association and free floating attention are the hallmark of the prescribed method of psychoanalytic data gathering. Only facts gathered by these means are admissible to the psychoanalytic knowledge base. I firmly believe that their point of view should be resisted. The general adoption of this strategy, of course, implies that psychoanalysis exists in isolation from disciplines which do not use these methods, i.e. all other disciplines. Of course, like all such categorical statements, the assertion that theoretical change invariably originates in the consulting room is simply an idealisation and a fallacy. For a start, as Riccardo Steiner’s (1989, 1994) and George Makari’s (in press) work so clearly illustrate, history moves via its complex determinants, and psychoanalytic theory tracks along behind. As a Kuhnian (1962) ‘protoscience’, psychoanalysis is extensively organised by contemporary common sense and prevailing philosophical traditions (Makari, in press). It has not yet, and arguably it will never become immune to the social world in the Kuhnian sense of science (Makari & Shapiro, 1993; Makari, 1994). ‘Where have all the hysterics gone?’ asked Elaine Showalter (cited in Whittle, in press) rhetorically, and answered: ‘They have gone into discourse’.

The argument that psychoanalytic observations concerning human behaviour are in some sense incommensurate with any other form of observation is nonsense. The mind remains the mind whether it is on the couch or in the laboratory. To maintain otherwise is logically untenable, and risks denying psychoanalysis fields of observation that have historically proved to be of enormous value. It is inaccurate to state that Freud’s own observations were restricted to his consulting room. He was acutely aware of other domains of study - history, literature, anthropology as well as neuroscience - and drew on these at many points of contact. 

THE RISKS AND VIRTUES OF CONCEPTUAL INTEGRATION WITH RESEARCH IN OTHER DOMAINS

On the other hand, piecemeal integration from other domains runs a similar risk of inductivism as clinical observations. The temptation for all of us must be to identify those sets of findings from neighbouring fields which best fit our pre-conceptions. Conceptual integration, just as clinical work, is rarely truly without memory or desire (Bion, 1967).

Nevertheless, we have much to learn from other disciplines. It is arrogant to assume that we, as psychoanalysts, are superior to those in other disciplines (anthropology, cognitive neuroscience, philosophy of mind, developmental psychology, neurobiology and so on) in our commentary on the mind. Arnold Cooper noted that ‘it is inherent in the nature of science to be refreshed by discourse in other disciplines’ (Cooper, 1997, p9). The fear, expressed for instance by André Green, appears to be that fields adjacent to psychoanalysis have the potential to destroy the unique insights offered by clinical research. Whilst this may not be a dominant view in psychoanalysis, it is an influential view. Although most psychoanalysts welcome the insights which research from some related areas can bring, collaboration with neighbouring disciplines is patchy, unsystematic and usually focused on specific findings, discoveries or ideas which, not surprisingly, are already consistent with a particular author’s preconceptions (c.f. Wolff, 1996). 

Contrary to the suggestion that closer proximity to neighbouring sciences may destroy psychoanalysis, a strong case can be made (see for example, Kandel, 1998) that the rich insights from psychoanalysis can be strengthened by closer integration with biological psychology and psychiatry. Psychoanalysis, according to some of its most senior leaders (Cooper, 1997; Cooper, Kernberg, Schafer & Viederman, 1991; Michels, 1994; Olds & Cooper, 1997), will become extinct if we continue to isolate ourselves from important scientific advances in other fields. Systematic study could achieve a high level of integration and a great deal of increased sophistication in the way that psychoanalysts talk about remembering, imagining, speaking, thinking, dreaming and so on. What is required for integrative initiatives is a broader range of methods and an openness to and excitement about new ideas.   

Permit me to take an example from my work with Mary Target which has attempted to explore the utility of the concept of a ‘theory of mind’ for the understanding of early (self) development and adult borderline functioning. The theory of mind literature is a prototypical example of how observations of clinical phenomena (the absence or distortion of a capacity or function) may sometimes enlighten us about normal development. This is an epistemic strategy which psychoanalysis shares with the newer discipline of developmental psychopathology, bridging child psychiatry and psychology. The origin of the theory of mind concept is in philosophy of mind (Brentano, 1924; Dennett, 1978), adopted in comparative psychology (Premack & Woodruff, 1978), elaborated in experimental developmental psychology (Wimmer & Perner, 1983) then extended to the study of autism (Baron-Cohen, Leslie & Frith, 1985) and somewhat later to psychoanalytic approaches to borderline personality disorder (Fonagy & Higgitt, 1989). Through a combination of clinical (Fonagy, 1991) and experimental (Fonagy, Steele, & Steele, 1996) studies we have been able to demonstrate that some individuals with histories of maltreatment appear defensively to inhibit their capacity to think of others (or themselves in relation to others) as intentional beings, as having a mind, as being motivated not by external circumstances but by feelings, beliefs and desires. We have thus given some substance to many psychoanalytic clinical observations concerning the experiences of such patients on the couch, at the same time as being more specific about the kinds of mental processes that may be entailed by certain forms of splitting, attacks on linking and perhaps even dissociation. I believe this sequence speaks to the value of the interdisciplinary approach and of the cross-fertilisation of ideas.   

RESPECT FOR PSYCHOANALYTIC IDEAS AND THE UNCONSCIOUS

Whether psychoanalysts do or do not find much of value in other disciplines is a matter of taste and should not, in my view, be legislated upon. Nonetheless, progress in other disciplines concerned with the mind has been remarkable and the exclusion of such information is a high-risk strategy at a time when interdisciplinary collaboration is perceived as the driving force of knowledge acquisition. But are psychoanalytic researchers at risk of inadvertently destroying the theory they were trained in, the ideas they are committed to defending in neighbouring domains? André Green (Green & Stern, in press), for example, called infant researchers to task for  ‘.... trying to destroy psychoanalytic theory’. In my experience, these researchers are enormously respectful of psychoanalytic contributions. It is the motive force of their work to build bridges, develop ideas, and to protect psychoanalytic ideas in a cultural context which is massively hostile to it, rather than wishing to destroy them. 

Are psychoanalytic researchers trying to build a psychology alternative to the psychoanalytic, which is ‘simpler and easier to teach’? Surely this is a misconception. They observe and create models of mental processes based on this observation, governed by the principle of parsimony. They see their models as partial, representing singular aspects of behaviour. To take just one example, Gergely and Watson (1996) proposed an interesting model of how the infant develops an understanding of its own affect on the basis of the internalisation of his caregiver’s responses to his emotional expression. Their model, although highly innovative and firmly rooted in empirical observation, has attempted to enrich rather than displace earlier psychoanalytic formulations on mirroring. It has provided a further vector to account for the same mental processes described clinically by psychoanalysts such as Bion, Winnicott and others. 

A major, understandable source of anxiety for psychoanalysts associated with the interdisciplinary forays of research, is the risk of losing touch with unconscious determinants of thought, feeling and action. In particular, the observation of ‘behaviour’ (external reality) could undermine our concern with the internal world. For example, ‘observations of the actual past’ cannot (it is claimed) give us insight into the mental world, since representation of the unconscious past is distorted by projections, projective identifications and drives. The Sandlers’ theory of the past and present unconscious (Sandler & Sandler, 1987) teaches us that the past unconscious is inherently inaccessible. The essential point here is that autobiographical memories are always suspect, that experiences are aggregated into ways of thinking (templates) which unconsciously orient to later experiences. Interestingly, when - based on memory research - we (Fonagy, 1999; Fonagy & Target, 1997) offered a neuroscientific justification for psychoanalytic technique and the focus on transference interpretation, and argued that the recovery of memory was incidental to the process of psychic change, the predominant clinical concern appeared to be that the model underplayed the significance of real memories. It does seem hard to be heard across the gap. There is nothing inherent to research on mental processes that precludes full consideration of unconscious aspects of psychic life.

To enter into discourse, of course, does imply playing by the rules of the other side. But these can be helpful. For example, James Reason looked at slips of the tongue both from a cognitive and a dynamic perspective. Many slips, he claimed, could be accounted for in terms of dysfunctions of known cognitive mechanisms (such as the dominance of habitual over more rarely used language structures) and did not require explanation in dynamic terms. (This of course differs from Freud’s assumption that all slips were unconsciously determined.) Many remained, however, where the minor eruptions of unconscious processing were clear. For example, after extolling the virtues of Clinical Psychology at some length, a local counsellor formally inaugurated a new building for the department by saying, ‘I declare this Department of Cynical - er, I mean Clinical Psychology open’. Or the French radio announcer who, intending to exhort ‘le population immense du Cap’, found himself referring to ‘le copulation immense du Pape’. Setting our theories into the context of other disciplines can sharpen our perspective, legitimate our theories but also restrain us from the folly of over-extending them. 

A further anxiety about research concerns numbers. There is little to be said on this point beyond asserting that numbers have their place in our universe but they are not required for good psychoanalytic research. Joseph Sandler’s Annual Research Lecture in 1996 was entitled ‘Research without numbers’. He described a method that may be considered knowledge systematisation or epistemological/methodological research. This body of work (Sandler, 1962; Sandler, Dreher & Drews, 1991) builds on existing theory and aims at achieving theoretical refinement by asking questions such as, ‘How do we know?’ or, ‘How can we be more certain that we know?’. Part of Sandler’s methodology was to use the Socratic dialogue where, in a small group setting, through relentless questioning, he would attempt to elicit the implicit theoretical structures which clinicians pre-consciously constructed through clinical experience.

Recent years have seen an abundance of new techniques for doing research without numbers, so called qualitative research. The first Joseph Sandler IPA Research Conference, to take place this weekend, is appropriately focused on qualitative research in psychoanalysis. The approaches to data collection entailed in qualitative research are quite different from quantitative research in their epistemological roots. In many disciplines (such as psychology), they exist in active opposition to quantitative methods. There is no reason, however, why psychoanalysis cannot absorb both quantitative and qualitative methodologies. However, on a cautionary note, the effort and rigour demanded by qualitative methods are no less exacting than those required for quantitative techniques. As Marianne Leuzinger-Bohleber points out in her superb paper for this Saturday’s IPA Research Conference (4^th March, 2000), there are examples of good and bad in both. Incidentally, her work is a wonderful prototypical illustration of how qualitative and quantitative methods may be meaningfully combined to illuminate the types of long-term effects which may be observed in psychoanalytic treatments.

THE EXTERNAL FUNCTIONS OF RESEARCH

So far, I have focused on the role of research in expanding psychoanalysis, particularly in refining our models of psychological mechanisms. I have also argued that many of the dangers which the combination of psychoanalysis and research were thought to create are illusory. However, I have also tried to emphasise that research can never do away with the study of subjectivity through clinical psychoanalysis which must remain our key investigative tool, alongside other evolving methods for studying subjectivity. I have little time left to talk about the functions of research in relation to the world outside psychoanalysis. I see these as threefold: communication, theoretical and clinical validation. 

Concerning the communication function of research, I would simply say that research has the potential to translate our ideas into a language which is less dependent on the personal experience of psychoanalysis. Because when we undertake extra-clinical research we borrow research methods from other disciplines, we, as a by-product, facilitate communication with them. 

Concerning the validation of theory, this poses a formidable challenge. Even apparently easily operationalisable constructs such as defence mechanisms have rarely been formulated with the kind of exactness required by research studies. Extra-clinical investigations, however, may help to constrain theorising, for example through our growing knowledge of infants’ actual capacities. 

Validation of psychoanalysis as a therapy is a far more burning issue. The ‘outcomes’ enterprise is not marginally related to the validation of psychoanalytic theory. There is no more of a logical connection between the truth of psychoanalytic theory and the outcome of psychoanalysis than there is between a theory of headaches and the effectiveness of aspirin. As Freud (1937) pointed out, in ‘Analysis Terminable and Interminable’, knowing the truth about one’s life does not guarantee that one will live it any more successfully.

Yet, as mental health practitioners, our social responsibilities extend beyond those of scholarship. Society expects a mental health treatment to show marked reduction in the patient’s symptoms and conscious distress. Our patients probably share this expectation. One problem in outcome research has been a confounding between treatment technique and measures of outcome, favouring approaches such as cognitive therapy which directly focus on symptom change. However, the development within psychoanalysis of new, relevant measures, together with a trend towards assessing the impact of psychotherapies at a neurobiological level may serve to highlight the value of psychoanalysis. At least four neuroimagery studies have shown critical changes in brain activation patterns following psychotherapy (Baxter et al., 1992; Schwartz, Stoessel, Baxter, Martin & Phelps, 1996; Vinamäki, Kuikka, Tiihonen & Lehtonen, 1998), and Kandel (1998; 1999) has recently argued that psychotherapy may make neuroanatomical changes in the brain through altering gene expression.

Still, even with unfair yardsticks, psychoanalysis has been doing quite well. The Research Committee of the International Psychoanalytical Association, at the request of Dr Kernberg, has recently prepared a comprehensive review of North American and European psychoanalytic outcome studies (Fonagy et al., 1999). We concluded that existing studies did not unequivocally show that psychoanalysis was more effective than an alternative treatment or an active placebo. A range of methodological and design problems was identified. Nevertheless, the evidence is that psychoanalysis consistently helps patients with milder (neurotic) disorders and somewhat less consistently helps other, more severe groups. Longer intensive treatments tended to have better outcomes than shorter, non-intensive ones. The impact of psychoanalysis was apparent beyond symptomatology, in measures of work functioning and reductions in health care costs. 

The alternative to empirical validation is to step outside the constraints that society imposes on health practitioners. We could abandon the idea of offering treatment in favour of providing a particular form of intense subjective experience that deepens self-awareness. Thus we evade the ogre of ‘evidence based practice’ but exchange this for even more fickle and ambiguous social criteria. This is not an unreasonable approach, and it has significant currency in francophone countries. Arguably, the public no longer looks to psychoanalysis as a treatment for specific syndromes. But this would have an impact upon the way psychoanalysis is currently practised here. Our practice of psychoanalysis does not only depend upon patients, it also depends on attracting new recruits who are drawn primarily from the mental health professions. Opting out of outcome research would change the future face of psychoanalysis.

CONCLUSIONS

For all these reasons, (a) to elaborate our model of underlying mental processes, (b) to systematise our knowledge base, (c) to communicate with other scientists and mental health professionals and (d) to show that our treatment works, we look to research in psychoanalysis. But research is not for everyone. It is for those willing to live in no-man’s land, their motives regarded as suspect, sometimes treacherous to both sides, who may put in fourteen-hour days proving their loyalty to the separated parents, who can bear feeling incompetent in both their professions, and where necessary surviving on a thin diet of conviction... I could go on. It is my belief, however, that in time to come, the value of this enterprise will be recognised on both sides of the fault-line. And at that moment there will be rejoicing in both these lands. 

References