Research in Psychoanalysis: An Area of Controversy

Research in psychoanalysis constitutes an area of controversy that recently has been the subject of intense debate among psychoanalysts. The 2002 Frankfurt International Conference on research in psychoanalysis, as well as the creation of two sub-committees of Research in the International Psychoanalytic Association (IPA), the first following an objective-quantitative research tradition, and the second subcommittee for Conceptual, Clinical, Epistemological and Historical Research, reveals the importance this topic has assumed recently in this area of study (Widlöcher, 2003).

The dialectical tension between practice and theory has been an implicit issue since the origins of psychoanalysis. By arguing that psychoanalysis is a metapsychology, Freud (as cited in Dreher, 2000) had in mind a theory based on, and with its roots in, clinical practice. As a matter of fact, it was in the course of the analytic work that the theory originated but it is also in practice that the theory must prove itself (Freud, 1927). In spite of the inseparability between cure and research in psychoanalysis, and the problem that research is an almost omnipresent topic in Freud's thought, since the analytic session is regarded as a privileged ground to test his concepts, this author (Freud, 1927) addresses this issue in a brief passage where he states,

In psychoanalysis there has existed from the very first an inseparable bond between cure and research. Knowledge brought therapeutic success. It was impossible to treat a patient without learning something new; it was impossible to gain fresh insight without perceiving its beneficent results. Our analytic procedure is the only one in which this precious conjunction is assured. It is only by carrying on our analytic pastoral work that we can deepen our dawning comprehension of the human mind. This prospect of scientific gain has been the proudest and happiest feature of analytic work. Are we to sacrifice it for the sake of any considerations of a practical sort? (p. 256)

According to Freud (1927), the analyst not only helps and cures but at the same time does research, which from a traditional point of view seems an uncommon association. The clarification of this relation between research and treatment remains the subject of discussion within the psychoanalytic community (Thomä & Kächele, 1985; Leuzinger-Bohleber & Bürgin, 2003; Dreher, 2000). It is interesting to note that Freud (1927) introduced the discussion of research in psychoanalysis in a paper where he discussed the academic background of analysts. Could non-medical practitioners be trained as psychoanalysts? What Freud (1927) was arguing in this section is that research in psychoanalysis has very specific characteristics, and to be an analyst involves the development of a research attitude towards clinical material and theoretical formulations. Previous medical studies could not assure or facilitate the development of such skills.

Dreher (2000) criticises Strachey's translation of the German word used by Freud (1927) in this paper “junktin” by bond since it introduces a slightly different meaning to Freud's (1927) original idea. Dreher (2000) argued that the word “junktin” “…draws our attention again to the fact that several things are held together which otherwise would fall apart” (p. 39), stressing the indispensable complementarity between both activities. According to Freud (1927), this conjunction between practice and research would take place at two different levels. On the one hand, research is developed on the basis of the single case. In the course of the treatment the analyst searches for the reasons and causes of psychic disorders in the patient's life, and how they change and evolve during the treatment, or why they do not.

According to a traditional experimental approach, that usually draws large samples from an even larger population, the use of single-case studies would create the problem of generalization since the results would have a validity restricted to the concrete individual case (Dreher, 2000). This critic reveals an underlying conception of science that aims to test law-like hypotheses through direct observation

in order to produce timeless and universal knowledge. As it will be argued in this paper, this is far from being the main objective of psychoanalytic research. Is the problem of generalization a relevant issue for psychoanalysis? Does psychoanalysis aim only to produce a theory that provides general and universal principles of psychic functioning? Moreover, it is important to develop a critical view towards a criterion of generalization attained by the induction from large samples. Why should validity be restricted in single case? As a matter of fact, each individual certainly has aspects that are universal and recurrent in the human species. In the same way, there are some features that are common to all psychoanalytic processes that seem to be forgotten in this classical view of the limitations of single-case method of research. As argued by Hampe (2003), “It is a mistake to believe that generalizing knowledge is always related to causal processes and aims at universal explanations, whereas individualizing research should aim at simply understanding …” (p. 49). Hampe (2003) illustrates this problem with an example from experimental physics. The exact determination of the brightness of a star or the speed with which a galaxy moves in space reveals the acquisition of individualised knowledge that may also produce, at a different level, generalised knowledge, for example the determination of the age of the universe. The same could be applied to psychoanalytic production of knowledge. For example, the research into the process of symbolic production in a single case may lead to a deeper knowledge of that specific case, but this individualised insight may have consequences for the general theory of symbolism. For example, Sayers' (2000) description of the case of Edward, who was a psychotic patient in analysis with Segal, shows that Segal's analysis of his communication, played an important role in the production of an individual insight in the understanding of that case, but it had also an important impact in the production of a general theory of the two modes of symbolisation because it enabled Segal to understand the primitive use of symbols in symbolic equations.

On the other hand, since research is inherent to the psychoanalytic method its use enables us to discover something new about general psychic functioning. As a matter of fact, it is impossible to treat a patient without discovering something new in reference to the single case but also regarding general psychoanalytic knowledge. In Freud's (1927) passage there is an implicit demand to establish as object of scientific research the determinants of the causes and changes of psychic disorder and also the factors responsible for the failure in analytic work. One of the obvious consequences of this thesis is the need for a constant redefinition of the relation between practice and research (Dreher, 2000). Nevertheless, in spite of this claim the present situation is still confused and the way in which research and clinical practice should be linked remains unclear since there is no agreement between different psychoanalytic schools as to what kind of science psychoanalysis is or what methods should be used in research.

The question of how to develop research in psychoanalysis is particularly complex and needs further reflection. Firstly, according to Fonagy (as cited by Leuzinger-Bohleber & Bürgin, 2003), the thesis defended by Freud (1927) of the inseparability between cure and research has raised some confusion since it has created the conviction that merely by using the analytic method, analysts are discovering something new which goes beyond investigation of a single case or even that they are developing research according to the experimental standards of research. The clarification of the scientific status of psychoanalysis is urgent because it has implications for methodology, namely in the choices of the methods that should be used to collect and treat data, and also in the definition of the criteria for scientific validation of knowledge.

Secondly, due to the specific nature of the analytic method, which is simultaneously a method of research into unconscious processes and contents and a technique of treatment, practitioners should reflect on the consequences and difficulties introduced by the use of the same

tool to investigate and to treat. Actually, with exactly the same instrument we collect data, indispensable for the construction of concepts, and test our theories. As argued by McAloon (1992), this involves the introduction of another dimension (research) to an already complex process (treatment). How can we be reassured that the research does not disturb the course of the treatment? The Heisenberg uncertainty principle (as cited in Meadow, 1996) points out that the act of observation already transforms the system under investigation. In this context, in what measure might this research attitude influence the therapeutic relationship? How might it affect phenomena like transference and countertransference? Furthermore, the use of the same tool to investigate/observe and to test the theory may create a certain propensity for auto-confirmation, particularly because there is a tendency to ignore data that are not collected by my instrument of observation. The use of a method based on words (like free-association, interpretation, etc.) creates the problem of non-verbal material and the recollection of data concerning non-verbal periods of development through free-association. In what measure does language interfere or transform non-verbal material?

In this paper, I intend to analyse the two main perspectives concerning the development of research in psychoanalysis. The Anglo-American approach proposes a method of investigation based on the traditional logical-rationalist model of science. In opposition, the European perspective defends a model of investigation closer to Freud's clinical and metapsychological view (Steiner, 2000). My main argument is that there is an urgent need for the clarification of the scientific status of psychoanalysis in order to define what model of research should be used in this domain. An effort should be made to reconsider Freud's (1927) claim of the conjunction between research and cure, and the use of Wittgenstein's (1953, 1975) notion of language-games might provide a possible framework to analyse the nature of communication between analyst and the patient and how theories and concepts gain their meaning in analytic practice.

What Kind of Methods in Psychoanalytic Research?

In the last decades, research has become a central issue within the psychoanalytic milieu. In most of Psychoanalytic Institutes the development of research projects is one of the final requirements to obtain the certificate as an analyst (McAloon, 1992). This policy reflects the recognition of the importance of stimulating a research attitude towards clinical material and analytic practice. Furthermore, since the 1940s and 1950s, the generalization of the use of psychoanalytic theory in other fields of humanities and science, involving the application of analytic knowledge to other settings and fields, has brought increasing pressure from the scientific community for psychoanalysis to develop research according to scientific criteria and experimental methods of investigation (McAllon, 1992). This tendency reflects the effort to submit psychoanalytic theory to scientific methods based on a traditional model of science. This has raised some confusion and has led research in psychoanalysis into a deadlock, since it is impossible to reduce psychoanalysis to the logical-rationalist approach of research. Recently, this topic2 has regained a renewed interest reflected in the creation of a Committee of Research in IPA, which followed a logical-rationalist approach to research. Green (2000) referred to the Ego-Psychology School and their effort to develop and institutionalise scientific research in psychoanalysis as follows,

I consider them, in my personal classification, as foreigners, alien to me within our multicultural community. An important sum of money has been given by the International Psychoanalytic Association (IPA) to

2. According to Green (2003), the recent debate on research in psychoanalysis has its origins in a conference held at the University College of London, chaired by Sandler, during which emerged a controversy between Green and Stern about clinical and observational psychoanalytic research (which was published in 2000 in an edition that included a previous controversy between Green and Wallerstein published in the Newsletter of the IPA). This controversy was fundamental for psychoanalytic research to regain a new space for debate within psychoanalytic institutions and meetings.

Page 39

Green's (2000) comment reflected the disagreement that this topic produced within psychoanalysis. More recently, a second sub-Committee for Conceptual, Clinical, Epistemological and Historical Research has been formed in order to develop the French and German tradition of research in psychoanalysis.

According to Steiner (2000), there are clearly two perspectives regarding the issue of how psychoanalysis should submit its knowledge to scientific validation. The Anglo-American perspective, close to the Ego-Psychology School, defends a compromise between psychoanalytic method and experimental research, meaning that psychoanalysis should be validated according to the traditional standards and criteria of science. On the contrary, the European approach, related to the British Independent Group, Kleinian and Post-Kleinian, Lacanian Schools and so forth, has a more orthodox position, since it recognizes the inadequacy of such methods to study psychoanalytic phenomena, and defends a model of research closer to Freud's clinical and metapsychological view.

Widlöcher (2003) seems to share this point of view and describes these two distinct positions as, on the one hand, the American position, which considers psychoanalysis as an object of research and submits psychoanalytic method and the empirical knowledge produced by it to verification and experimental confirmation, and on the other, the European perspective, which considers psychoanalysis as research, arguing that the psychoanalytic method is a tool of knowledge and discovery. Both perspectives reveals two divergent positions about what may be considered as evidence and therefore as empirical data indispensable for research – clinical data collected with the use of technical devices (e.g., video or tape transcripts) or whether a case study or an analysis reported by the analyst may also fulfil the

criteria of science. There are two fundamental questions in this area: what methods should be used in psychoanalytic research – psychoanalytic method or experimental ones? How do we gather analytic data and what is the role of the analyst in the transport of data from clinical setting or the use of technical equipment such as video, tape recording and others?

For the Ego-Psychology perspective, psychoanalysis should test its theories according to the traditional methods of research. In this context, psychoanalysis should be submitted to external criteria of validation and research activities should follow traditional scientific principles such as – a theory must have a direct empirical basis that allow predictability, the operationalisation of variables, experimentation and theoretical propositions must be verifiable and reliable (Couto, 1996). The Anglo-American perspective intends to develop research, trying to measure psychoanalytic processes and therapeutic outcomes objectively. In this context Wallerstein (1996), one of the pioneers of research in psychoanalysis in America, establishes as an aim,

…to go beyond the limitations of the familiar psychoanalytic single-case study, into formal and systematic research on the psychoanalytic process in a manner that we described as “at once meaningful and responsive to the subtlety and complexity of the phenomena and at the same time scientific in the best sense of that term… (p. 29)

The attempt to submit psychoanalytic theories to experimental research according to a logical-rationalist perspective of science, which involves the application of systematic methods and logically justified procedures, has brought psychoanalytic research to an impasse. French and German schools have claimed that to try to prove psychoanalytic knowledge according to this logical-rationalist model reveals an excessive adjustment to the classical paradigm of science that is inadequate for this area of study (Leuzinger-Bohleber & Bürgin, 2003). As ironically argued by Green (1996b):

The noble term “research” carries such an amount of prestige that it is to be expected that any reference to it might compel one to bow before it. Unfortunately, compared with the richness of the clinical experience of psychoanalysis, the findings of researchers look very meagre. Therefore, one has to be very cautious about the pretence to leadership of those who stand behind research. As some of the results of the requirements to make research possible end with an oversimplification of psychoanalytic knowledge, defended as a precondition of the path that would lead psychoanalysis to a science of proven facts, it is essential to preserve what is specific to the psychoanalytic endeavour (p. 21).

Since the 1940s and 1950s, there has been an increasing demand from the scientific community for psychoanalysis to develop research using “clean” devices and external observers, outside the clinical setting (McAloon, 1992). Closer analysis revealed the importance of preserving the psychoanalytic method in research as a way to guarantee that the material will emerge and can be clarified. As a matter of fact, the object of study of psychoanalysis – the intrapsychic material analysed from the vertex of the unconscious, can only emerge in a clinical setting. How can we be reassured that what we are investigating is the psychoanalytic object of study? Green (1996b) pointed out:

There is a neglect in most of the investigations of the specificity of what is intrapsychic and unconscious, and an underestimation of the parameters of the analytic situation related to the concept of setting, with the implicit idea that an observational procedure of interpersonal relationship can better account for the object of study of psychoanalysis than the speculation of psychoanalysts drawn from their therapeutic experience (p. 24).

This author (Green, 2000) emphasised the importance of the use of the analytic work and method within the analytic relationship to be sure that the phenomena occurs and that it can be explored and

interpreted in the analytic session. Green (2000) thinks that in analysis we can only work on the basis of the reconstruction – the analytic process works on the basis of psychic reality, meaning that we can only elaborate analytically those events after they were lived and after the psychic work has been done, involving all the distortions, unconscious processes, defences, personal projections, and so forth. As the Freudian (Freud, as cited in Green, 2000) concept of deferred action reveals, psychoanalysis does not work on the basis of a linear model of causality, which stresses the influence of the past event of childhood into the present. In this sense, the subject reorganizes, restructures subsequently, over and over again, the past events. It is this psychic re-elaboration of the event that gives significance to it and makes possible either the working through of the experience or, on the contrary, the development of a pathological effect. This argument is diametrically opposed to the simplistic common- sense view of, for example, the effects of a sexual abuse in children, or the experience of a traumatic event that would lead to direct pathological effect. After a traumatic event has been lived by the subject it is important to give space for the event to be worked through by the psychic because it will be reorganized in accordance with unconscious process, defences and so on. This event needs to be digested by the mind and the emotional structure of the subject. Those arguments implicitly reveal that psychoanalysis defends a different model of causality from the linear one proposed by logical rationalism.

A different conception of causality is revealed also by the principle of overdetermination in which psychoanalysts argue that a behaviour or psychic event may be an expression of different processes and contents in the same person or even in different persons and within the same person at different times. This is one of the most important arguments in favour of including the analytic method in research and clinical data obtained in the clinical setting, because the analytic technique makes it possible to clarify the relation between manifest content and psychological states of mind, using namely the clarification

of the material, the analysis of the transference and countertransference relationship, the analysis of dreams, free-association, and so forth. Therefore, only in the session, (i.e., within the analytic relationship) can those phenomena be understood. The use of experimental methods would lead to enormous difficulties since psychoanalysis deals with psychic and intersubjective phenomena that are not directly observable – it needs to be interpreted (Dreher, 2000). As stated by Dreher (2000), “In psychoanalysis, essential phenomena reveal themselves not by observation alone, but only through processes of understanding and interpretation – interpretation, that is, with the help of psychoanalytic concepts.” (p. 22). It is also recognised that the alternative use of observational methods is also limiting when compared with data collected through analytic methods (McAloon, 1992). It is a sort of interpreting of manifest content without the elucidation given by free-association.

Another area of controversy is how data from the analytic process should be collected. According to Dreher (2000) there are three different distinct categories of data involved in analytic research. The primary data are gathered directly in the analytic session and therefore form the basis of the analytic work. They are constituted by the verbal and non-verbal material that arises within the analytic session and are only accessible to the participants. One of the features of the primary data is its highly volatile character, and for research purposes, only the direct participants have privileged access to them. The secondary data are the ensemble of data or selection of data reported by the analyst from the analysis that can be gathered by the analyst through written or oral reports or with the help of technical devices or instruments such as videos, tape-recording, and so forth. Secondary data are the basic material for research activities. Finally, extra-clinical data are not gathered directly from the analytic situation, but are a result of the use of analytic methods outside the analytic situation. They may also be gathered by non-psychoanalytic methods but are interpreted in accordance with psychoanalytic theories.

From an experimental point of view, the use of the analyst in the process of collecting data and the transport from inside the therapeutic relationship to the exterior and out of the setting is undesirable because it produces biased and epistemologically contaminated data. These data are not objective or reliable. Moreover, according to the traditional paradigm of science, the recording of data made by the analyst is contaminated because they could have been selected according to the analyst's theoretical perspective, distorted consciously or unconsciously through countertransference or through the analyst's tendency to cover personal mistakes. Nevertheless, the use of technical devices, as pointed out by Dreher (2000), raises even more difficulties and poses two different methodological problems. Firstly, the technical transport of primary data is also subjected to systematic mistakes and distortions. What is lost or distorted by using this sort of instrument? For example while the video or audio recording capture the verbal dialogue between both participants they cannot register thoughts, feelings and what is communicated beyond words. Communication in the analytic session is viewed as a global phenomenon and cannot be reduced to one or few channels of communication. Secondly, to what extent might the introduction of the “technical third” distort and interfere in the analytic process? The introduction of technical devices changes completely the analytic relationship with consequences at the levels of transference and countertransference. It is the assurance of confidentiality that makes possible both the free-association of the patient and the freedom of the process that are indispensable conditions for the analytic work to take place. This transport of data also raises ethical problems. How does one select data and deal with the question of confidentiality of the process? Finally, the fact that data from the analytic session is collected through “clean” devices does not prevent the distortions introduced by the subjective beliefs of the researchers who will analyse those data (Dreher, 2000).

In this context, Dreher (2000) states that the current tendency is to rediscover the role of the analyst in research. In fact, the existence

of self-evaluation skills in the analyst is one of the preconditions for the development of analytic practice and it introduces a reassurance (through self-analysis, supervision of the cases, control-analysis, group discussion of cases, etc.) that the analyst is aware of how his personal aspects may intervene in the analytic process, indicating that there is a control inherent to the process. In Psychoanalytic Institutes the whole training structure is organized taking into consideration that personal analytic work – the training analysis – is a necessary condition for the future analyst, and it is considered to be a complement of the theoretical training. Free-floating attention is a central technique in the analytic function and it is regarded as the capacity to have access to unconscious communication, which requires the analyst to use his own unconscious as a receptor for the unconscious of the patient. In order to maintain this level of communication it is indispensable for the analyst to have easy access to his own unconscious, which explains the need for the candidate to have or to be under training analysis. Personal analysis would also provide an essential tool – the capacity to do self-analysis – which is fundamental to deal with countertransference in the analytic relationship. Moreover, the supervision with a senior analyst of a couple of cases is another precondition in the process of training and it has the function of guiding the trainee in the understanding and management of the case but also of helping him to become aware of his countertransference. Those requirements reveal an underlying conception of the therapeutic relationship as depending on the patient but also on the personality of the analyst and consequently, psychoanalytic institutions have developed control instruments to regulate those factors.

The Epistemological Status of Psychoanalysis

Up till now, the debate concerning the development of research in psychoanalysis has not produced any significant results. With regard to this topic the two most important perspectives have opposite positions. On the one hand, the Anglo-American perspective defends a

model of research close to the traditional standards of science and research, on the other, the European contribution (e.g., in Perron (2003) discussion of this subject) has been restricted to demonstrating at a practical and clinical level the impossibility of meeting such standards, arguing that this attitude undermines what is specific to psychoanalysis, since it involves a complete distortion of clinical data. As Green (1996a) states, “I regret to say that my feeling on the method adopted up till now has twisted the nature of the object.” (p. 37). From my point of view, this impasse is partly a result of a paradoxical position assumed by the European Schools in this debate. In spite of strongly arguing that traditional methods are not adequate, there is an underlying belief in the European perspective that since psychoanalysis does not follow those criteria it is a sort of ‘minor science’. Green (1996b) argued, “I have defended previously the idea that the ideal of psychoanalysis was to be part of science, but that the specificities of its practice and its mode of thinking were not compatible with the ordinary requirements of scientific evidence” (p. 22). Implicitly what this author states is that the ideal of psychoanalysis is to follow the standards of the exact sciences and to be submitted to systematic empirical evidence. As summarised by Leuzinger-Bohleber and Bürgin (2003), the actual state of art of research in psychoanalysis is as follows:

…on the one hand, there is a danger of over-adjustment to an understanding of research that is inadequate for psychoanalysts; on the other hand, there are the perils of refusing a scientific discourse by a self-absorbed retreat to an exclusive discussion within our psychoanalytic ivory tower (p. 7).

The arguments of both perspectives reflect a tendency to reduce science to a model of research based on experimental methods. This is perhaps the reason why the European Schools have not been able to produce an epistemological discourse that sustains the specificity

of analytic knowledge and therefore to present an alternative to the traditional model of research. Moreover, on the basis of the controversy between both perspectives, there also seems to be a different delimitation of the object of study since the European Schools emphasise unconscious processes and the Ego-Psychology School stresses the ego's adaptation to reality. From my point of view, this controversy needs to be analysed using a different framework and it requires a clarification of the epistemological position of both perspectives. Is it desirable to submit psychoanalysis to empirical evidence? Should psychoanalysis adopt a research perspective, which employs the methods of the traditional scientific approach? Moreover, is it useful for psychoanalysis to follow such rules/methods in research? There is an urgent need to discuss the scientific status of psychoanalysis in order to decide what model of research should be used to test its theories.

Recent developments in this debate seem to point towards a position that intends to clarify the scientific status of psychoanalysis that tends to stress the specific quality of psychoanalysis as a science of the unconscious that cannot be studied according to the dominant model of science (Leuzinger-Bohleber & Bürgin, 2003). In fact,

…modern psychoanalysis should neither feel obliged to adapt to the criteria of “science” in terms of a unified science characterised by the natural sciences, as some – mainly Anglo-American – therapy researchers claim, nor should psychoanalysts define their professions in the state of suspense between the arts and the sciences, as some German psychoanalysts would have it (e.g., ). (Leuzinger-Bohleber & Bürgin, 2003, p. 12)

Psychoanalytic research does not follow traditional scientific patterns and it is particularly difficult to submit psychoanalytic theories to external criteria of validation, namely the logical-rationalist ones. The American position has an underlying dogmatic and restrictive

view of science that limits it to a model of research based on traditional scientific methods, defending a universal and unique criterion of scientific validity. However, as happens in other areas of human sciences, most of the phenomena that make up the object of study of psychoanalysis such as the unconscious, transference, countertransference are not directly observable and therefore cannot to be submitted to direct observational evidence. They must be deduced, which involves processes of understanding and interpretation (Dreher, 2000). Moreover, how can psychoanalysis, as the ‘science of the unconscious’, be analysed according to a philosophy of consciousness based on rationality?

During the 20^th century, traditional patterns of research were questioned. Philosophy of science began to be critical of some of the traditional principle and since 1920s the logical-rationalist criterion has been criticized. A new conception of science and its procedures has emerged, particularly with the work of Popper and Kuhn. In Popper's (Popper, 1935) perspective, falsification as a criterion of scientific validation is more reliable than verification and proof. As argued by this author (Popper, as cited in Dreher, 2000) we must not exclude cosmic catastrophes when we state that the sun rises every morning because tomorrow it may not rise. It is not because we observe repeatedly a phenomenon in a single case that we are reassured that it will continually occur in the future. This is the reason why research should aim to refute scientific propositions. In this sense, Popper (1935) criticized the inductive conclusion based on experience and argues that science does not evolve in accordance with a linear and cumulative process of confirmed findings as defended by the traditional view. On the contrary, modification of theories and scientific progress results from an accumulation of falsification and Popper (1935) emphasises the process that makes possible for the old theory to be refuted and to be substituted by a new one. New theories would provide solutions to problems that the old theory failed to provide.

For Kuhn (1962), science is regarded as an historical and social process. He agrees with Popper when he recognizes that the evolution of scientific knowledge is not a cumulative process, it evolves by leaps and bounds. Paradigms limit and bind the scientist in his research, and Kuhn (1962) has a contextualist view of science where he identifies two different periods in scientific progress – normal science, which suppresses novelty and looks for the confirmation of the paradigm, and scientific revolution, when there is a shift in the paradigm and an emphasis on the refutation of knowledge. Kuhn's position is critical of the logical-rationalist argument of universal and timeless criteria of validity. He recognizes the importance of pluralism in scientific methods. Since science is a social and historical construct, a scientific method or a scientific knowledge is limited regarding its range of applicability and its fate is to be replaced over and over again each time a new paradigm emerges. Kuhn (1962) highlights the importance of social and historical influences on scientific change and construction of theories. In the development of psychoanalytic theories it is necessary to take into account the influence of internal aspects such as personal factors and also external ones. For example, the controversial discussions in the British Psychoanalytic Society between the Kleinians and Anna Freud's group in the 1940s constitute a paradigmatic example of the influences of internal factors in theoretical development. The concept of transference was at the origin of those discussions. Klein (as cited in Segal, 1972) argued that in the therapy of small children, the therapist was invested with transference phantasies from the beginning of the treatment. On the contrary, Anna Freud (as cited in Geissmann, C. & Geissmann, P., 1992) stated that transference reactions could only appear at the end of Oedipus complex. Till then the child could not invest other persons with fantasies, emotions and expectations because the model of relation with both parents was still under construction in direct relations with them. This discussion regarding the nature of transference assumed enormous proportions due to its implications on work

with children – a more pedagogical kind of intervention as supported by Anna Freud (as cited in Geissmann, C. & Geissmann, P., 1992) or the psychotherapeutic treatment of children with rules equivalent to the ones used in the analysis of adults as proposed by Klein . Implicit in this issue was also the discussion of Freud's heritage and both authors claimed that their theories were a direct continuation of the master's work. The extreme positions that both groups assumed in those discussions are a consequence of an emphasis on divergences and the impossibility to look for possible points of agreement, which was probably a result of certain aspects of both Klein's and Anna Freud's personalities.

In the same way, McAloon (1992) argued that the integration of psychoanalysis in the health system and the increasing importance of insurance companies in the financial support of psychotherapeutic treatment are other examples of the influence of external factors in the development of psychoanalysis. On the one hand, the influence of such factors has exerted a pressure for psychoanalysis to develop shorter treatments in order to reduce costs. On the other hand, it has also created the need to measure the efficiency of therapeutic intervention according to quantitative psychotherapy research model, which is the dominant model of research in health institutions, in order to identify the central aspects which produce changes.

More recently, regarding the methods admissible in science, Feyerabend (1975) assumes an anarchist “anything goes” position as a reaction against the logical-rationalist claim of a single and universal correct method. For this author, this procedure would create no inhibitions to new ideas and like Kuhn he defends a system of theoretical and methodological pluralism. The new hypotheses could even go against what is established but they should have a capacity of internal persuasion that allows the subject to understand the phenomenon from a new perspective but nevertheless allows the researcher to handle the problem on a practical or theoretical level. The acceptance of a certain scientific proposition is not determined solely by the

methodological rules followed in its production, and one can argue that there are no universal methodological rules. Moreover, there are no theories in full harmony with facts and each theory interprets facts in its own respective light. Rational-empiricist rules and their methodological procedures are not the only way to construct reality. Feyerabend (as cited in Couto, 1996) defends the need to accept the specificity of each area of knowledge claiming a position close to absolute relativism. A specific body of knowledge should be analysed in line with the rules, conventions and methodological practices adopted by the epistemological perspective which informs the paradigm and not according to external patterns of validation which belong to another viewpoint or perspective. This author defends a creative problem-solving strategy and he recognizes the importance of personal aspects of the scientist in the construction of theories. Frequently, the elaboration of a theory does not result from the application of logical and rational rules to concrete observed facts. Intuition and creativity plays a determinant role like for example, in Winnicott's (1971) theory of playing. The elaboration of the concept of playing in this author's work is mostly the result of the intuition of recognising the regularities in a multiplicity of phenomena and activities such as play activities in children, the cultural life of adults, elaboration of scientific theories, and so forth. In fact, Winnicott (1951, 1971) puts under the same denominator very distinctive activities since they all involve the same psychological process – the negotiation of the gap between internal and external reality. Winnicott's theory of transitional phenomena has some roots in his empirical observation of the small child's use of the first transitional object. In his effort to explain why the child develops such a strong link with a particular toy, he recognized that this use of the first not-me object had some similarities with the older child's attitude during play activities and also the adult's everyday experience in the cultural domain, since all those phenomena represent an attempt to manipulate objects of reality according to the subject's wishes and desires. Winnicott (, )

Wittgenstein's (1953, 1975) concept of language-games makes it possible to introduce a certain clarity into this discussion and to provide a different approach to some Psychoanalytic Schools' claim for pluralism of science. For this author (Wittgenstein, as cited in Monk, 1990), theories must be described mentioning the activities and the “way of life” of the “tribe” that uses it, meaning that a theory is not something fixed by timeless and rigid laws of logical form. It must refer to something that is always linked with a custom, a practice and a way of life, stressing the importance of historical, socially shared language. This author (Wittgenstein, 1969) makes an important distinction between propositions that involve a description of the world and propositions that constitute our conceptual framework. Contrary to the former, the latter do not describe a body of knowledge, but rather a way in which we understand the world. A framework cannot be justified or proven correct but it provides the limits within which justification and proof take place.

In psychoanalysis there is no agreement as to which framework should be used in research. The traditional model of science fails to recognize that the criteria for a correct or incorrect framework are

not provided by an idealized external objective world, but rather by ourselves, it is a result of a convention, a use with practical requirements. A convention of using rigid rules as defended by traditional science is not true and universal but simply more useful in some specific contexts. As a matter of fact, in psychoanalytic practice the aim is not precision, syntactic clarity and consistence as proposed by a logical-rationalist framework. In this sense, the American perspective defends the use in research of a framework that has nothing to do with the psychoanalytic “form of life”. In clinical practice, most of the time, the analyst does not aim to explain psychic phenomena. Interpretation does not result from the application of laws, and it does not provide a causal, mechanical explanation of psychic phenomena. With an interpretation the therapist intends to produce in the analysand a change of perspective in order to be reassured that a new meaning will emerge. In child psychotherapy, the interpretation of play does not aim to explain the causal structure that was at the basis of its formation but rather to enable the child to change the vertex under which he sees the content of his play phantasies so that now it make sense in a different way. The central point of an interpretation is to provide to the child the access to the vertex of the unconscious, to a new meaning that is produced in the context of the therapeutic relationship. Interpretation as a task of attribution of significance is not something that provides an explanation but an understanding that results from the communality of experiences between the child and the therapist in which emotions plays a determinant role (Neves, 2007). In the same way, the use of theories and concepts in clinical practice is subjected to a multitude of influences that go far beyond what is explicitly formulated (Dreher, 2000). There is a complex net of implicit influences, such as the personal aspects of the analyst (individual thoughts resulting from clinical experience, unconscious motives, values inherent to the analyst's personality, implicit theories shared for example by his/her training group, etc.) that influence the use of concepts in clinical practice.

In conclusion, the debate on research in psychoanalysis has only been recently analysed in accordance with the critical reflection undertaken by epistemology in the last century on scientific construction and progress. The effort to submit psychoanalytic theories to the model of exact science is mistaken and psychoanalysis should be able to denounce the inadequacy of logical-rationalist methods to test its theories. In this context, it is necessary for the psychoanalytic community to reach agreement concerning the definition of the scientific paradigm for psychoanalysis, which seems to have been recognised by some psychoanalytic movements with the creation of a second sub-Committee for epistemological reflection, among other subjects. Only by clarifying the model of science applicable to psychoanalysis is it possible to produce a scientific discourse that takes into consideration the specificity of its knowledge, which should be supported by internal criteria of validation focused on the practical use of its concepts and theories (Couto, 1996). What sense does it make to use a framework based on rationality, precision and logical rules if those criteria distort completely the object of study of psychoanalysis?

The application of Wittgenstein's (1953, 1975) methodology of language-games would clear up of some confusion in this domain and could provide the necessary ground for re-launching the Freudian debate on the conjunction between cure and research. Instead of regarding empirical research from a restrictive experimental point of view, it would be interesting to develop research with the purpose of understanding unconscious process and conflicts of the analysand, using concepts as tools and instruments for gaining knowledge and insights. Moreover, analysis of how concepts emerge in the psychoanalytic “form of life”, how they are submitted to constant transformation in clinical practice and how they are embodied with empirical content is an area which has gained increasing development (Dreher, 2000, 2003). Theories gain their meaning from a form of life, a language-game which, in the case of psychoanalysis, is quite different from that of describing and explaining the physical phenomena. Psychoanalysis

has an inventory of words and concepts specific to the use of its domain of action, therefore theories must look for justification in clinical practice and also in discourse since it is indispensable to analyse whether a specific concept is compatible with other concepts of the same conceptual field (Dreher, 2000, 2003). Moreover, this conjunction between practice and theory should complement conceptual investigation, which aims at the systematic analysis of the use of a theory or a concept, with empirical research.