Medical Hypotheses 75 (2010) 594599

Contents lists available at ScienceDirect

Medical Hypotheses
journal homepage: www.elsevier.com/locate/mehy

Psychoanalysis on the couch: Can neuroscience provide the answers?

Andrea Mechelli *
Department of Psychosis Studies, Institute of Psychiatry, Kings College London, De Crespigny Park, London SE5 8AF, UK

a r t i c l e

i n f o

s u m m a r y
Over a century after Freuds attempt to establish psychoanalysis as a natural science, there is renewed interest in the integration of psychoanalytic and neuroscientic ndings within a single theoretical and experimental framework. However, it is important that any intellectual exchange is not motivated only by declining condence in psychoanalytic theory and practice or awareness of the rising fortunes of the brain sciences. The present paper considers three possible ways in which psychoanalysis and neuroscience might be integrated. These include the investigation of the neurological organisation of psychoanalytically dened phenomena; the evaluation of psychoanalytic theories based on their neurobiological evidence; and the use of neuroimaging techniques to assess the progress and outcome of psychoanalytic treatment. The author argues that these exercises are unlikely to provide psychoanalysis with the unlimited opportunities for overcoming its uncertainties and doubts that some have anticipated. For instance, the argument that mapping psychoanalytically dened phenomena in the brain may provide biological validity to these phenomena should be considered an expression of logical confusion; the evaluation of psychoanalytic theories based on their biological evidence is critically dependent on speculative interpretation of what the theories predict at neuronal level; and the supposedly objective evaluation of the progress and outcome of psychoanalytic treatment on the basis of neurobiological data relies on the subjective reports of the patient and analyst. In light of this conclusion, there are a number of outstanding questions which remain to be addressed, including whether psychoanalysis should adhere to scientic canons and whether this would necessarily require an experimental methodology. 2010 Elsevier Ltd. All rights reserved.

Article history: Received 8 May 2010 Accepted 22 July 2010

We must recollect that all of our provisional ideas in psychology will presumably one day be based on an organic substructure. Sigmund Freud [1].

Introduction In the 1890s, Sigmund Freud produced several drafts of the Project for a Scientic Psychology [2] in which he attempted to translate psychological concepts into the language of neuroanatomy and neurophysiology. The work was motivated by the assumption that a psychological theory deserving any consideration must be grounded in the natural sciences. Psychological processes would become perspicuous and free from contradiction [2] once characterised in terms of quantitative states of speciable material particles. Throughout the Project, Freud distinguished between mechanical and biological explanations of psychological phenomena. Mechanical explanations referred to the fact that cer* Address: Department of Psychosis Studies, Institute of Psychiatry, Kings College London, P.O. BOX 67, De Crespigny Park, London SE5 8AF, UK. Tel.: +44 (0) 20 7848 0833; fax: +44 (0) 20 7848 0976. E-mail address: a.mechelli@kcl.ac.uk 0306-9877/$ - see front matter 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.mehy.2010.07.042

tain phenomena were determined directly by contemporary physical events; for instance, consciousness could be described in terms of simultaneous mechanical events occurring in specic types of brain cells. Biological explanations referred to the fact that some phenomena were determined genetically by virtue of their survival value; for instance, psychical attention had developed in the course of psychical evolution because any alternative had led to the generation of greater unpleasure. Freud discussed both mechanical and biological aspects of a number of psychological phenomena in the context of normal function as well as psychopathology. However, this exercise was doomed by the lack of an established, rigorous method for integrating psychological and neuroscientic knowledge. Furthermore, the paucity of the neuroscientic knowledge available at the time meant that Freud had to rely on imaginings, transpositions and guesses [3, p. 120]. As a result, he never published the Project and even came to describe it as a kind of aberration [3, p. 134]. Most importantly, Freud failed to demonstrate that mechanical or biological explanations would indeed be capable of making psychological concepts perspicuous and free from contradiction. For instance, it is hard to see how exactly a characterisation of the ego in terms of a complex of neurones which hold fast to their cathexis was supposed to shed light

A. Mechelli / Medical Hypotheses 75 (2010) 594599

595

on this concept and inform clinical practice. Later Freud abandoned the idea of biology as a scientic model for psychoanalysis in favour of a descriptive model based on verbal reports of subjective experiences: in the end I may have to content myself with the clinical explanation of the neuroses [3, p. 137]. However, throughout his life he remained aware that all subjective experiences also exist in the form of physiological mechanisms and that future developments in brain sciences might shape our understanding of psychological processes. Over a century later, neuroscience has witnessed more signicant developments than Freud might have imagined. In the 1960s, new treatments in the form of psychopharmacological drugs began to be available for patients with serious psychological conditions such as depression and manic-depressive illness. More recently, the introduction of brain imaging techniques such as positron emission tomography and functional magnetic resonance imaging allowed the measurement of brain activity associated with different mental processes in vivo. These developments have captured the public imagination and have established neuroscience as one of the most fertile disciplines of our time. In contrast, over the past 50 years psychoanalysis has been repeatedly called into question by the dominant scientic paradigms of the time. First, logical positivism charged that psychoanalytic hypotheses could not be veried or conrmed due to the vagueness of their theoretical terms [4]. Later, Popper suggested that psychoanalytic hypotheses could not be falsied on the basis of empirical evidence [5]. More recently, Grnbaum has argued that alternative psychoanalytic theories cannot be compared because it is not possible to manipulate variables of interest as required by the scientic paradigm of eliminative inductivism [6,7]. In addition, new developments in psychoanalytic theory and practice have failed to capture the public imagination and a growing number of patients are turning to faster, less costly treatments. The psychiatrist and neuroscientist Eric Kandel argues that the method of free association and interpretation that was proposed by Freud at the beginning of the 20th century has long exhausted its investigative power [8,9]. Furthermore, the achievements of psychoanalysis over the past 50 years have been hindered by a lack of a questioning tradition based on rigorous experiments designed to falsify, support or explore the insights produced by clinical practice. However, a discipline of psychoanalysis in intellectual decline could benet greatly from the major developments in brain sciences. As Eric Kandel put it: the future of psychoanalysis, if it is to have a future, is in the context of an empirical psychology, abetted by imaging techniques, neuroanatomical methods, and human genetics [8, p. 468]. This opinion is shared by a growing number of psychoanalysts who are becoming more open to the brain sciences as well as several neuroscientists who are beginning to engage with psychoanalytic ideas [1016]. Over a century after Freuds rst attempt to establish psychoanalysis as a natural science, there is therefore renewed interest in the integration of psychoanalytic and neuroscientic ndings within a single theoretical and experimental framework. This has resulted in the foundation of a new International Neuro-Psychoanalysis Society in July 2000 with the aim of promoting inter-disciplinary work between the two elds. This goal is based on the assumption that psychoanalysis and neuroscience ultimately share the same scientic endeavour, namely to make the complications of mental functioning intelligible by dissecting the function and assigning its different constituents to different component parts of the apparatus [17]. However it is important that any intellectual exchange with neuroscience is not motivated by declining condence in psychoanalytic theory and practice or awareness of the rising fortunes of the brain sciences. Rather, one has to consider whether psycho-

analysis and neuroscience indeed share the same scientic endeavour and how, if at all, psychoanalysis could benet from the integration with neuroscience. There are at least three possible ways in which psychoanalysis and neuroscience could be integrated within the same theoretical and experimental framework. First, it has been suggested that mapping psychoanalytically dened phenomena in the brain may provide biological validity to these phenomena [9,18]. Second, it has been argued that the ndings of the brain sciences could be used to test psychoanalytic theories [8,9,11,12,1820]. Third, it has been proposed that the neuroscientic method could be used to measure the progress and outcome of psychotherapy [8,9,11,18]. These possible ways in which psychoanalysis and neuroscience could be integrated will now be considered in turn.

Neuroscience as a means of mapping psychoanalytically dened phenomena in the brain It has been suggested that it is possible to nd the neurological correlates of some traditional psychoanalytic concepts and thereby set them on a rm, organic foundation [18, p. 104]. Here the expression psychoanalytic concepts refers to psychoanalytically dened phenomena or, in other words, phenomena which have been described by and are dependent on psychoanalytic theory (e.g. repression). In recent years, a number of investigations have attempted to identify the neural correlates of psychoanalytically dened phenomena. Most of these studies have used the clinico-anatomical method which involves looking at the different ways in which a psychological function of interest breaks down following focal lesions in different regions of the brain. Using this approach, Mark Solms explored the neurological organisation of dreaming in 361 patients with localized brain damage [21]. The research revealed that lesions to different areas of the brain were associated with distinct patterns of disturbances, such as absence of conscious experience of dreaming, lack of any visual imagery, recurrence of stereotyped nightmares and increase in the frequency of dreaming; in addition to these dreaming-related disturbances, different lesions resulted in distinct patterns of psychological decits that shed light on the elementary function of each area. The author argued that this investigation provides us with a rst foothold on the anatomical and physiological representation of some crucial psychoanalytic concepts [22]. The clinico-anatomical method has also been used to explore the neurological organisation of unconscious processes in 35 brain-damaged patients undergoing psychoanalytic therapy [23]. The investigation revealed that lesions to different areas of the brain were reliably associated with distinct unconscious processes. For instance, a bilateral lesion in the ventromesial quadrant of the frontal lobes resulted in exemption from mutual contradiction, primary processes (mobility of cathexis), timelessness and replacement of external by psychical reality which was interpreted as a breakdown of the unconscious, inhibitory functions of the ego. In addition to the clinico-anatomical method, the neural correlates of psychoanalytically dened phenomena can be investigated using alternative approaches such as brain imaging techniques and psychopharmacological manipulations. In the case of brain imaging techniques, brain responses during different experimental conditions are measured and compared [24]; in the case of psychopharmacological manipulations, the impact of psychoactive drugs on conscious and unconscious mental processes is investigated. There are a number of methodological limitations with each of these approaches that need to be considered when planning an investigation or interpreting the ndings. For instance, it is extremely rare to nd any two patients with exactly the same brain

596

A. Mechelli / Medical Hypotheses 75 (2010) 594599

lesion and yet the clinico-anatomical method relies on multiple observations. Even when two patients with the same lesion are available, there is no established method for determining whether the resulting unconscious mechanisms are indeed the same. In addition, psychoanalytic ideas such as repression and hysteria may be difcult to manipulate experimentally and evaluate in the laboratory as required by brain imaging techniques and psychopharmacological manipulations. From the point of view of the psychoanalyst, however, the issue of whether it is currently possible to identify the neural correlates of psychoanalytically dened phenomena is somewhat secondary; a more fundamental question is whether this exercise can provide any contribution to psychoanalytic theory and practice. In other words, if the current methodological limitations were overcome, would mapping psychoanalytically dened phenomena in the brain provide biological support for these phenomena? It may be tempting to assume that the identication of the neural correlates of psychoanalytically dened phenomena would provide biological support for these phenomena (indeed, this is how the nding would be likely to be reported by the media and understood by the public). However, the biological evidence afforded by this exercise would be critically dependent on the assumption that the phenomena have been identied and dened correctly. In other words, mapping psychoanalytically dened phenomena onto the brain is not a source of empirical justication, because the assumption that the phenomena under investigation are valid hypothetical constructs underlies the whole effort. For instance, let us consider the phenomenon of repression as the process by which certain thoughts or memories are expelled from consciousness and conned to the unconscious [25]. One could map this phenomenon in the brain by looking at the effects of brain damage on the mechanism of repression as revealed by psychoanalytic treatment of neurological patients. However, the biological evidence afforded by this exercise would rely on the assumption that indeed there is a process by which certain thoughts or memories are expelled from consciousness and conned to the unconscious. A cognitive psychologist who did not share this assumption would come to a completely different conclusion about the effects of brain damage in the same neurological patients. Thus, the argument that mapping psychoanalytically dened phenomena in the brain may provide biological validity to these phenomena should be considered an expression of logical confusion. This is not to say that mapping psychoanalytically dened phenomena in the brain should not be pursued. For instance, this exercise may be extremely valuable from the point of view of the cognitive neuroscientist who is interested in the neural correlates of conscious and unconscious mental processes. Indeed, the perspective of psychoanalysis may provide the brain sciences with a comprehensive conceptual framework which may inform and stimulate new research. However, mapping psychoanalytically dened phenomena in the brain bears little relevance from the point of view of the psychoanalyst who is concerned with the theory and practice of psychoanalysis.

Neuroscience as a means of testing psychoanalytic theories It has been proposed that the neuroscientic method should be used to test psychoanalytic theories in terms of their biological plausibility [8,9,11,12,16,1820,26]. For instance, Solms and Turnbull [18] argue that psychoanalysis has proved hopelessly inadequate for the purpose of deciding between competing possibilities, which has resulted in a gradual proliferation of incompatible theories and practices. In contrast, the conclusions produced by the brain sciences over the past few decades have been remark-

ably reliable. This discrepancy, they suggest, is due to the different observational perspective of the two disciplines: while the subjective point of view of psychoanalysis generates data of a eeting and fugitive nature, the objective perspective of neuroscience leaves little room for ambiguity and speculation. However, linking the metapsychological conclusions of psychoanalysis with the equivalent conclusions of neuroscience [..] provides psychoanalysis with unlimited opportunities for overcoming its uncertainties and doubts [18, p. 305]. This exercise comprises of two main steps: rst, one has to establish the neural correlates of the components that make up the psychoanalytic theory under investigation by using the clinicalanatomical method, brain imaging techniques or psychopharmacological manipulations. These components include phenomena which are widely agreed to exist, such as sleeping and dreaming, as well as phenomena which have been described by and are dependent on psychoanalytic theory, such as censorship and regression. Once these phenomena have been localized, one can ask whether they relate to each other in the physical reality of the brain in the way that is predicted by the theory under investigation. For instance, in order to test the wish-fullment theory of dreams, one has to identify the brain areas that are implicated in the various components of the theory such as sleeping, dreaming, censorship, regression. Once the neural correlates of these concepts or phenomena have been established, it is possible to ask whether the relevant brain areas interact in the way that is predicted by the theory. If all aspects are supported by biological evidence, the theory should be accepted; in contrast, if some aspects are not consistent with the available neuroscientic data, the theory should be revised accordingly; nally, if most of the aspects are not biologically plausible, the theory should be rejected altogether. The idea that neurobiological data could be used to make inferences about how the mind works can be found in many of Sigmund Freuds comments on the possible integration of psychoanalysis and biology. For instance: Biology is truly a land of unlimited possibilities. We may expect it to give us the most surprising information and we cannot guess what answers it will return in a few dozen years to the questions we have put to it. They may be of a kind which will blow away the whole of our articial structure of hypotheses [27]. However, it should be noted that Freud himself never attempted to make inferences about the workings of the mind on the basis of purely neurobiological data; indeed, one of his greatest contributions was the discovery that the mind could be better understood by looking at apparently trivial phenomena such as dreams, associations and delusions. Furthermore, when Freud borrowed concepts such as instinct and discharge from biology, he re-worked them signicantly within the psychoanalytic framework. This leads us to the question of whether it is actually possible to make inferences about psychoanalytic theories on the basis of neurobiological data. A rst consideration is that one has to assume that psychological and neurobiological domains are isomorphic [28]. In other words, for every pair of individual entities (whatever these entities may be) in the psychological domain which are in some relation to each other (whatever this relation is), there must be a corresponding pair of individual entities in the neurobiological domain (whatever these different entities may be) which are in some relation to each other (whatever this different relation is); the relation in one domain can be very different from the relation in the other domain but must be of the same kind (e.g. binary and asymmetric). Isomorphism needs to be assumed when testing psychoanalytic theories on the basis of neurobiological data using the two-step approach described above. Whilst it may be reasonable to assume that the psychological and neurobiological domains are indeed isomorphic, a second consideration needs to be made which is more problematic. Alterna-

A. Mechelli / Medical Hypotheses 75 (2010) 594599

597

tive psychoanalytic theories can be compared on the basis of their biological evidence only if they make different predictions in the neurobiological domain. However, psychoanalytic theories per se do not make any clear-cut predictions about the underlying quantitative changes in the brain. Conversely, neurobiological data per se do not provide any self-evident information about the nature or quality of the associated psychological processes. In order to bridge the gap between the two domains, one has to make a number of assumptions and interpretations. First it is required that the neural correlates of the psychoanalytically dened phenomena that make up a psychoanalytic theory of interest have been identied successfully; as argued above, this relies on the assumption that these phenomena have been identied and dened correctly. Then one has to assume that the psychoanalytic theory predicts a specic neurobiological mechanism in the physical reality of the brain; however, in identifying what that mechanism might be, one cannot avoid subjective interpretation. For instance, does the wish-fullment theory of dreams predict that during dreaming there will be an increase, a decrease or no change in communication between the brain regions that are implicated in censorship and regression respectively? There is nothing in the theory itself to indicate which of these should be predicted. Finally one has to interpret the neurobiological data as consistent or inconsistent with the assumed predictions of the theory. To summarise, any outcome of the test will rely on the correct identication and denition of the phenomena that make up the theory in question and on speculative interpretation of what the theory predicts at neuronal level. While this exercise may as well provide valuable insights, it is important to appreciate that the use of neurobiological data does not allow a completely objective, interpretation-free assessment of psychoanalytic theories after all. This is well demonstrated by two recent investigations of the unconscious by Schore [12] and Kaplan-Solms and Solms [23]. Both investigations used neuroscientic data to make inferences about the Freudian unconscious but their conclusions are strikingly different. Schore [12] reviews a signicant body of neuroscientic evidence including clinico-anatomical and brain imaging studies to conclude that the right hemisphere acts as the neurobiological substrate of Freuds dynamic unconscious. The author goes on to propose that the the current expanding body of knowledge of the right hemisphere suggests a major alteration in the conceptualization of the Freudian unconscious [12, p. 269]. Schore argues that the mechanism of repression requires left-hemispheric inhibition of right-hemispheric sexual and aggressive wishes. However, such inter-hemispheric inhibition cannot take place until the age of four when the left hemisphere is thought to become dominant. This, he suggests, means that the formation of the unconscious in the early years of life must rely on mechanisms other than repression. As Schore himself puts it: the current ndings of neuroscientic imaging research that study the brain in real time suggest that the adaptive operations of the dynamic unconscious in everyday life are best described not in terms of ongoing repressive functions but rather in terms of nonconsciously mediated processes [12, p. 247]. In contrast, Kaplan-Solms and Solms [23] used the clinico-anatomical method to investigate ve patients with damage to the perisylvian area of the right hemisphere and who were undergoing psychoanalytic therapy. The authors argue that if the Freuds conscious and unconscious were localized in the left and right hemisphere respectively, then left-hemisphere patients should have gross ego-decits and be at the mercy of primary processes whereas right-hemisphere patients should be more realistic and rational than average. However, the investigation revealed that right-hemisphere patients displayed greater ego-decits than did their left-hemisphere counterparts, a nding which was interpreted as inconsistent with the idea that the right brain acts as

the neurobiological substrate of Freuds dynamic unconscious. This conclusion clearly does not support the major alteration in the conceptualization of the Freudian unconscious proposed by Schore [12]. While the investigations by Schore [12] and Kaplan-Solms and Solms [23] may have provided some valuable insights, they also illustrate that use of the neuroscientic method may not necessarily make psychological concepts perspicuous and free from contradiction.

Neuroscience as a means of testing the outcome of psychotherapy There is growing evidence that the acquisition of new skills may cause alterations in gene expression which in turn may produce long-term changes in brain structure and function [8]. These neuroplastic changes may also be the result of downstream effects on protein biosynthesis or the regulation of rho GTPases which are not necessarily caused by alterations in gene expression. For instance, learning to play a string instrument is associated with increased representation of the ngers in the premotor and motor areas of the brain that are implicated in skilful hand movement [29]. Likewise, learning to navigate within a complex environment is associated with structural increases in the hippocampus, a region of the brain that is critical for navigation [30]. These plastic changes have been observed not only in the developing brain but also in adult life albeit to a lesser extent [31]. Furthermore, it appears that regular training over a period as short as three months is sufcient for neuroanatomical changes to take place [32,33]. Insofar as psychotherapy produces long-term changes in conscious and unconscious attitude and behaviour, this is likely to occur through alterations in gene expression, protein biosynthesis or the regulation of rho GTPases which in turn may result in longterm changes in brain structure and function. In other words, psychotherapy could be seen as a form of micro-plumbing which initially changes the functional connections among neurons, and then later converts these functional changes into changes in the actual structure of the cerebral cortex itself [34, p. 68]. It has therefore been suggested that one should be able to measure the progress and outcome of psychotherapy in terms of cortical reorganisation in the brain [8,9,11,18,32,33,35]. This idea is supported by a number of recent studies that have reported changes in brain function following various forms of psychotherapy (see [36] for review). There are two possible ways in which psychoanalytic treatment might affect the brain [9, pp. 518519]. One possibility is that therapy might induce signicant changes in those areas of the brain that showed structural or functional abnormalities before therapy; in other words, psychoanalytic treatment might normalize the structure and function of the brain. An alternative possibility is that therapy might lead to compensatory changes in areas of the brain which were not impaired before therapy. It is not difcult to see why some psychoanalytic institutions might be open to this approach: it might promote a shift in the public perception of psychoanalysis from speculative discourse of little scientic value to evidence-based treatment of the mind and the brain. However, it is important to carefully consider the value of this exercise in relation to the aim of psychoanalytic treatment. If we accept Freuds notion that the aim of psychoanalytic treatment is to bring about a mitigation of neurotic symptoms and a decrease in the propensity to symptom-formation, then one has to wonder whether neuroanatomical changes may provide any clue as to whether successful treatment has taken place. Let us assume that signicant increases in gray matter density were observed in the amygdala of patients who undertook psychoanalytic treatment compared to control subjects. The amygdala is a region of the brain that plays a critical role in emotional regulation and has been

598

A. Mechelli / Medical Hypotheses 75 (2010) 594599

implicated in a number of psychiatric conditions. It would be tempting to conclude that the observed changes indicate that the patients have become more aware of their unconscious complexes and no longer show the same propensity to respond to frustrating experiences with symptom-formation. However, there are several other possible interpretations. For instance, the observed changes might reect the fact that patients have learnt to repress their emotions even further or have developed new unconscious complexes. It might also be that the increases in the amygdala simply reect growing frustration or disappointment in the patients after hours of fruitless, expensive therapy. The critical point here is that neuroanatomical changes are of a quantitative rather than qualitative nature and therefore can provide little information about the nature or quality of the analytic experience. In other words, the observed quantitative changes do not have any obvious meaning per se and cannot provide biological evidence of successful treatment unless such evidence is added as an interpretation. Thus, it appears that the use of neuroscientic data may not allow a truly objective, interpretation-free measurement of the progress and outcome of psychotherapy after all. Ultimately, one has to rely on the subjective reports of the patient and analyst to interpret the quantitative changes that have taken place in the brain.

The conclusion that neurobiology may be unable to shed light on the questions and doubts of psychoanalysis is by no means novel; indeed, over the past few decades several authors have argued that neurobiology is irrelevant to psychoanalysis e.g. [28,38,39]. However these authors did not systematically consider the possible ways in which psychoanalysis and the brain sciences could be integrated to map psychoanalytically dened phenomena in the brain, test psychoanalytic theories or evaluate the outcome of psychoanalytic therapy. In light of the present conclusion, there are a number of outstanding questions which remain to be addressed. For instance, should psychoanalysis adhere to scientic canons? Does adherence to scientic canons necessarily require an experimental methodology? How might the progress and outcome of psychoanalytic therapy be assessed? These questions may be troublesome but cannot be avoided in the context of a rational, critical evaluation of psychoanalytic theory and practice. Whether fortunately or unfortunately, psychoanalysts will not be able to rely on neuroscientists for the answers. Conicts of interest statement None declared. References

Conclusion Over the past 50 years, the scientic paradigms of positivism, falsicationism and eliminative inductivism have called psychoanalysis into question for relying on theoretical hypotheses which cannot be conrmed [4], falsied [5] or compared experimentally [6,7]. In recent years, it has been argued that although psychoanalysis has historically been scientic in its aim, it has rarely been scientic in its methods [9, p. 506]. Some psychoanalysts have reacted to these criticisms by rejecting the view of psychoanalysis as a science; others have implied that adherence to scientic canons does not necessarily require the experimental manipulation of independent variables [37]; more recently, it has been argued a discipline of psychoanalysis in intellectual decline could benet greatly from the major developments in brain sciences [18,10,14]. Here I have considered three possible ways in which psychoanalysis and neuroscience could be integrated. First, I have discussed the investigation of the neurological organisation of psychoanalytically dened phenomena; then I have explored the potential of the brain sciences for testing psychoanalytic hypotheses; nally I have considered the use of the neuroscientic method to measure the progress and outcome of psychoanalytic treatment. These integrative exercises may provide valuable information from the point of view of the cognitive neuroscientist who is concerned with the neural correlates of conscious and unconscious mental processes. Indeed, the insights of the psychoanalytic perspective can inform the quest for a deeper understanding of the neurobiological basis of mental processes. However, I have argued that the use of the neuroscientic method is unlikely to provide psychoanalysis with unlimited opportunities for overcoming its uncertainties and doubts [18, p. 305]. For instance, the mapping of psychoanalytically dened phenomena in the brain does not necessarily provide biological support to these phenomena; in addition, the comparison of psychoanalytic theories based on their biological evidence depends on speculative interpretation of how the theory should be implemented in the neuronal domain and what the observed quantitative changes in the brain mean; nally, the supposedly objective evaluation of the progress and outcome of psychotherapy on the basis of neurobiological data relies on the subjective reports of the patient and analyst.

[1] Freud S. On narcissism: an introduction, vol. SE 14. Vintage, The Hogart Press; 1914. [2] Freud S. Project for a scientic psychology, vol. SE 1. Vintage, The Hogart Press; 1950 [1895]. [3] Freud S. The origins of psycho-analysis. New York: Basic Books; 1954. [4] Nagel E. Methodological issues in psychoanalytic theory. In: Hook S, editor. Psychoanalysis, scientic method and philosophy. New York: Grove Press; 1959. p. 3856. [5] Popper K. Conjectures and refutations. London: Routledge & Kegan Paul; 1963. [6] Grnbaum A. Logical foundations of psychoanalytic theory. In: Essler W, Putnam H, editors. Festschrift for Wolfgang Stegmller. Boston: D. Reidel; 1982a. [7] Grnbaum A. Can psychoanalytic theory be cogently tested on the couch? Parts I & II. In: Psychoanalysis and contemporary thought, vol. 5; 1982b. p. 155255 [311436]. [8] Kandel ER. A new intellectual framework for psychiatry. Am J Psychiat 1998;155:45769. [9] Kandel ER. Biology and the future of psychoanalysis: a new intellectual framework for psychiatry revised. Am J Psychiat 1999;156:50524. [10] Corrigall J, Wilkinson H, editors. Revolutionary connections: psychotherapy and neuroscience. London: Karnac Books; 2003. [11] Green V. Emotional development biological and clinical approaches towards an integration. In: Green V, editor. Emotional development in psychoanalysis, attachment theory and neuroscience. Creating connections. Brunner-Routledge; 2003. [12] Schore AN. Affect regulation and the repair of the self. WW Norton & Company; 2003. [13] Westen D, Gabbard OG. Experimental psychology and psychoanalysis: what we learn from a century of misunderstanding. Neuro-Psychoanalysis 2002;1(2):23345. [14] Balick A. Shifting paradigmspsychotherapy, neuroscience, and a changing world: the seventh professional conference of the UK council for psychotherapy. Int J Psychother 2001;6(3):28995. [15] Damasio A. The feeling of what happens. London: Heinemann; 1999. [16] Panksepp J. Emotions as viewed by psychoanalysis and neuroscience: an exercise in consilience. Neuro-Psychoanalysis 1999;1(1):1539. [17] Freud S. The interpretation of dreams, vols. SE 4 & 5. Vintage, The Hogart Press; 1900. [18] Solms M, Turnbull O. The brain and the inner world: an introduction to the neuroscience of subjective experience. London: Karnac Books; 2002. [19] Hobson J, McCarley R. The brain as a dream state generator. Am J Psychiat 1977;134:133548. [20] Rubinstein B. On the possibility of a strictly clinical psychoanalytic theory. In: Gill M, Holzman P, editors. Psychology versus metapsychology. New York: International Universities Press; 1976. p. 22964. [21] Solms M. The neuropsychology of dreams. Mahwah, NJ: Lawrence Earlbaum Associates; 1997. [22] Solms M. Preliminaries for an integration of psychoanalysis and neuroscience. In: Winer JA, editor. The annual of psychoanalysis, vol. XXVIII; 2000. p. 179 200. [23] Kaplan-Solms K, Solms M. Clinical studies in neuro-psychoanalysis: introduction to a depth neuropsychology. London: Karnac Books; 2000.

A. Mechelli / Medical Hypotheses 75 (2010) 594599 [24] Shevrin H, Ghannam JH, Libet B. A neural correlate of consciousness related to repression. Conscious Cogn 2002;11(2):33441. [25] Freud S. Repression, vol. SE 14. Vintage, The Hogart Press; 1915. [26] McCarley R, Hobson J. The neurobiological origins of psychoanalytic dream theory. Am J Psychiat 1977;134:121121. [27] Freud S. Beyond the pleasure principle, vol. SE 18. Vintage, The Hogart Press; 1920. [28] Edelman M. Hypothesis and evidence in psychoanalysis. The University of Chicago Press; 1984. p. 10920. [29] Ebert T, Panter C, Wienbruch C, Hoke M, Rockstrom B, Taub E. Increased use of the left hand in string players associated with increased cortical representation of the ngers. Science 1995;20:213. [30] Maguire EA, Gadian DG, Johnsrude IS, Good CD, Ashburner J, Frackowiak RS, et al. Navigation-related structural change in the hippocampi of taxi drivers. Proc Natl Acad Sci USA 2000;97(8):4398403. [31] Mechelli A, Crinon JT, Noppeney U, ODoherty J, Ashburner J, Frackowiak RS, et al. Neurolinguistics: structural plasticity in the bilingual brain. Nature 2005;431(7010):757. [32] Draganski B, Gaser C, Busch V, Schuierer G, Bogdahn U, May A. Neuroplasticity: changes in grey matter induced by training. Nature 2004;427(6972):3112.

599

[33] Balbernie R. Circuits and circumstances: the neurobiological consequences of early relationship experiences and how they shape later behaviour. J Child Psychother 2001;27(3):23755. [34] Vaughan SC. The talking cure: the science behind psychotherapy. New York: G.P. Putnams Sons; 1997. [35] Pally Regina. The mindbrain relationship. Int J Psychoanal Key Papers Ser Ed, 2000. [36] Etkin A, Pittenger C, Polan HJ, Kandel ER. Toward a neurobiology of psychotherapy: basic science and clinical applications. Neuropsych Clin Neurosci 2005;17(2):14558. [37] Lacan J. crits. A selection, trans. Alan Sheridan London: Tavistock/Routledge; 1977. [38] Reiser M. Changing theoretical concepts in psychosomatic medicine. In: Reiser M, Arieti S, editors. American handbook of psychiatry, vol. IV. 2nd ed. New York: Basic Books; 1975. p. 477500. [39] Blass RB, Carmeli Z. The case against neuropsychoanalysis: on fallacies underlying psychoanalysis latest scientic trend and its negative impact on psychoanalytic discourse. Int J Psychoanal 2007;88(1):1940.