The Concept of Truth in Psychoanalysis

Charles Hanly

[Paper presented on June 15, 1991, at the Annual Meeting of the Rapaport-Klein Study Group]


Psychoanalysis is passing through a difficult period in its history (Wallerstein, 1988). It is still unclear whether there will emerge further splintering, further dilution, or a gradual re-unification and re-integration (Rangell, 1988). This paper seeks to explore a core issue in our current differences: the concept of truth in psychoanalysis.

Philosophers have advocated two different theories of truth: correspondence and coherence. The correspondence theory states that truth consists of the degree of correspondence between an object and its description. It assumes that under normal conditions the human mind is able to gain knowledge of objects by means of observation and its experimental refinement. This observational knowledge can then be used to test beliefs and theories. The correspondence theory is implied with oblique eloquence in Galileo's 'eppur si muove' (see Drake, 1978, pp. 356–7). Neither his official recantation of his astronomical discoveries, nor the majestic coherence of Ptolemaic astronomy, nor its obvious agreement with experience, nor the consensus of generations of scholars, could alter the fact that Galileo's observations of the moon, planets, and sun had enabled him to describe much more accurately what was actually happening in nature. This same view of truth and of science has been held by the great seminal scientists: Harvey, Newton, Darwin, Einstein, and Freud, and by scientists generally. The school of thought in philosophy with which the correspondence concept of truth is associated is realism: critics of correspondence would say naive realism; advocates would say critical realism.

The coherence theory of truth adopts the view that of the question: 'What objects does the world consist of? only makes sense within a theory or description… Truth… is some sort of (idealized) rational acceptability - some sort of ideal coherence of beliefs with each other and with our experiences as those experiences are themselves represented in our belief system - and not correspondence with mind-independent or discourse-independent "states of affairs"' (Putnam, 1981, pp. 47–9). Thus there may be more than one true description of the world. The correspondence theory allows for only one. In effect the coherence theory abandons objects as they actually are as the ground of truth for objects as they are constructed or constituted by the belief and theory investments that govern their observation and the way in which they are experienced by observers. The mind must, as a matter of psychological and epistemological inevitability, subject the objects which it seeks to know to the conditions under which it is able to know them. The original form of this idea is traceable to Kant (1781) although Kant was a scientific realist. Among its modern adherents have been Bradley (1897), Merleau-Ponty (1945), Sartre (1943), Ricoeur (1970), Habermas (1971), and the philosophers of science Kuhn (1970), Feyerabend (1965), and Putnam (1981). The school of thought, in philosophy, to which the coherence theory belongs is idealism.

Two further ideas tend to accompany the correspondence idea of truth: one epistemological, the other ontological. The epistemological premise is that objects are able to cause our senses to form more or less correct observations of them as they actually are. These observations can be, or can be made to be, sufficiently independent of theories held by observers concerning their objects that theories can be objectively tested. The ontological premise is that anyone's thoughts, and actions of any kind, are caused. Minds are part of nature.

Similarly, two further ideas tend to be associated with the coherence idea of truth. Epistemologically, it is assumed that our ways of thinking and perceiving unavoidably condition what we observe. Objects are unable to exert an independent influence upon our senses such as would enable us and oblige us to correct our theory-laden ideas of them. Facts are theory-bound, never theory-independent. Objects are amorphous and unintelligible in themselves. They have no means by which to define themselves. They must wait upon the definitions inherent in the theories we invent to try to understand them. The ontological idea is that human beings are unique in nature on account of a consciousness which supports the capacity for voluntary actions of a special kind - actions which are motivated by reasons rather than by causes. Minds constitute nature.

The idea of truth as coherence, of the intrinsic indefiniteness of persons as objects of knowledge and of voluntarism, are logically interconnected in the following way: if a person's actions are motivated by reasons which are neither causes nor caused, if a person freely chooses his motives, then his actions become at once immunized against the influence of his past and unpredictable. Voluntarism is a source of an intrinsic indefiniteness of the human mind which allows it always to slip away from any description that would seek to correspond with some fixed and determinate nature. The link between voluntarism and the indeterminacy of psychic life has been nicely stated as follows:

It has often been said that one's past determines one's present and future. Let it be underlined that one's present and future - how he commits himself to existence at the moment - also determines his past (May, 1958, p. 88).

Present choice determines the meaning of the past and the motives of actions. Psychic life ceases to be sufficiently determinate to be a suitable object for descriptions whose truth resides in their correspondence with an objective state of affairs.

Bound to this view is the hermeneutic, phenomenological, existential, and idealist idea that self-consciousness involves the capacity for self-transcendence. Self-transcendence allows for the abrogation of causality, the transformation of motives as causes into sui generis reasons at the disposal of consciousness. Thus Habermas (1971) claims that, when a neurotic conflict is resolved, self-reflection has actually 'dissolved' or 'overcome' the causal connexion between the symptom or inhibition and the repressed drive demands. Where psychic determinism was, the uncaused choice shall be.

This constellation of ideas has found its way into contemporary psychoanalytic theorizing, where it has been pressed into service in a number of ways.

Sometimes there is an appeal to the coherence theory of truth as a means of defending a theory against criticism. Goldberg (1976), (1988) has used the coherence theory to defend self-psychology against its critics. The philosophical idea that observations are theory-bound is used to explain differences in clinical observation: when two individuals with roughly similar neurophysiological equipment view the same thing or event and each see it differently, it is not necessarily true that one is incompetent or even wrong; rather it may be that they each observe with a different theory (Goldberg, 1976, p. 67).

This idea also agrees with Putnam (1981) that there may be more than one true theory about the same thing because the observations that confirm theories are contaminated by the very theoretical concepts they confirm. As Putnam states it, 'the very (experiential) inputs upon which our knowledge is based are conceptually contaminated…' (p. 54). However three difficulties arise when the coherence theory is used in this way. One: it is a double-edged sword. The observations of classical analysts cannot falsify self-psychology, but neither can the observations of self-psychology falsify classical - or any other - theory (Hanly, 1983). Two: one and the same patient can have a neurosis caused by a failure to resolve oedipal conflicts when he is treated by a classical analyst, and a failure to find a suitable object for idealization when he is treated by a self-psychologist (Kohut, 1971). The Oedipus complex is both a cause and not a cause - only a symptom. This consequence defies an elementary principle of logic, the principle of identity: nothing can be both p and not-p. Three: a scientific theory must be falsifiable in principle. Goldberg's use of coherence implies that neither self-psychology nor classical theory are scientific, since neither has a domain of observations that could ever falsify it. This use of the coherence theory results in theoretical solipsism and truth by conversion.

Goldberg (1988) has implicitly addressed these difficulties by considering the conditions for theory testing in psychoanalysis. Goldberg suggests a pragmatic test although, unlike scientific pragmatic testing, this form of testing requires 'commitment' over time to the theory being tested (p. 27). This requirement opens the door to the claim by adherents of a theory that falsifying observations are the result of a lack of such 'commitment'. It is of interest for our argument, however, that in the end Goldberg indirectly appeals to correspondence. If, for example, we are to be able to carry out the injunction to 'remain alert to the effects of our observations' (p. 110) we must be able to make observations of objects that are not subject to these effects. This crucial issue will be further considered below.

Spence (1982a), (1982b), despite some ambiguity, comes down in favour of coherence as the criterion of truth in psychoanalysis when he claims that 'the analyst functions more as a pattern-maker than a pattern-finder…' and goes on to refer to analyses as 'artistic masterpieces'. This account agrees with the notion that a present intention or perception interprets the past - that there is no discrete, specific, particular past which continues to be what it was; there are only the diverse perspectives on the past brought about by the intentions inherent in current projects, moods, affects, attitudes, and theories. The idea that the meaning of a person's past, as well as its influence upon his current life, is determined by present choices is very different from the idea that contemporary affective experiences activate chains of associated memories leading back to infantile precursors. The latter idea assumes that memories thus reactivated have inherent meaning that remains the same even if its conscious recall does not; the former idea assumes that memories are a kind of opaque mass than can be redesigned and informed with meaning by present intentions and investments: that is, by volitions conceived as uncaused causes.

These ideas belong with those of Habermas (1971) and Ricoeur (1974), (1981). Habermas (1971) advanced the view that psychoanalytic 'self-reflection' is able to suspend or transcend psychic causality.

Although some self-psychologists may not agree with his position, Kohut (1959) rested his theorizing on the coherence theory premise of indeterminacy:

What we experience as freedom of choice, as decision, and the like, is an expression of the fact that the I-experience and a core of activities emanating from it cannot at present be divided into further components… They are, therefore, beyond the law of motivation, i.e., beyond the law of psychic determinism (p. 232).

Psychic causation becomes a product of self-disintegration along with the Oedipus complex. The cohesive self rises above the bounds of causality. Kohut (1977), in explicit agreement with Habermas (1971), introduced the idea of the mutuality of observer and observed in order to claim on behalf of self-psychology a more fundamental knowledge of human nature than that of psychoanalysis. Kohut's concept of empathy disallows the degree of epistemic independence of subject and object required by correspondence. Goldberg (1988) has extensively elaborated the epistemological implications of the self-psychological version of empathy. Ricoeur (1974) asserted, in line with Habermas, that 'there are no "facts" nor any observation of "facts" in psychoanalysis but rather the interpretation of a narrated history' (p. 186). And although Ricoeur (1981) intended to abandon his earlier (1970) view that reasons and motives are sui generis, he failed to do so in so far as he continued to conceive of the relation between unconscious wishes and dreams, neurotic symptoms, or parapraxes as one of referring, denoting, signifying: that is, as an a-causal, semantic relation. Ricoeur (1981) appealed to coherence in a particularly naive and unsatisfactory form: 'a good psychoanalytic explanation must be coherent with the theory or, if one prefers, it must conform to Freud's psychoanalytic system' (p. 271). Such a dictum makes nonsense of the clinical testing of his ideas which Freud advocated.

The relationship of the concept of causality to the problem of truth in psychoanalysis is also raised by Schafer (1976), (1978) and Klein (1976). Schafter (1978) asserts that 'thinking historically, we do not say an agent is causally motivated to perform some action … we say that this agent did that and perhaps gave or could have given these reasons for doing so…' (p. 56). Like Klein, Schafer wants to deny that reasons, intentions, purposes, wishes, motives in general, are causes. Schafer and Klein erroneously identify causality with fatalism as did Sartre (1943). Schafer and Klein are also committed to a coherence theory of truth.

There are more ways than one to understand reality (Schafter, 1976). Reality is not, as Freud usually assumed, a definite thing to be arrived at or a fixed and known criterion of objectivity (Schafter, 1978, p. 66).

Once motives become reasons rather than causes they acquire a wonderfully amorphous, open-textured nature which allows them to be 'correctly' construed in a variety of ways. Interpretation is an expansion and complication of the context of an action. Different theories expand and complicate the context differently. Narrative coherence becomes the operative criterion of truth. There are as many true understandings as there are coherent, comprehensive, unified narratives about the motivating reasons.

Schafer (1978) states the ontological basis for this relativism: 'the concept of action requires us to regard each action as inherently spontaneous, as starting from itself' (pp. 48–9). This astonishing assertion certainly provides the basis for a 'free' construal of the reasons for an action. But if actions are not the outcome of past and present events but really are free creations, as Kant (1785) believed of morally willed actions - actions which have consequences but no antecedents - then they do not have a history at all. Schafer's position is the same as the existentialist view of Sartre (1943) despite Schafer's (1976) disavowal, and it is fundamentally at odds with the view of Ryle (1949), upon whose ideas Schafter otherwise has relied for his attack on Freud's metapsychology. Schafer's concept of action provides no justification for widening the context of an action to earlier events and actions (conscious or unconscious); it provides a justification for a life history that is no more than a phenomenological chronology. Anything more, if, indeed, every action, as Schafer claims, spontaneously starts from itself, would be sheer invention: an invention for which the only possible criterion of truth would be coherence. This idea provides unlimited opportunity for Spence's pattern-making.

A number of considerations lend credence to the coherence criterion of truth for psychoanalysis. Freud appears to have espoused coherence. Freud often (1895, pp. 194–5), (and 1909, pp. 165–9, for example) testified to his awareness of the complex, seemingly arbitrary, fragmentary, subtle, evasive mass of material produced by associations. Is not this material typically so ambiguous, so rife with uncertainties, that the best we can achieve is a coherent account with the possibility of other no less coherent accounts being constructed? In apparent support of Goldberg, Freud (1915) pointed out that even at the earliest stages of description a new science already applies concepts that are not drawn altogether from the field of observation to which the descriptions apply (p. 117). Freud (1927) remarked that 'a number of very remarkable, disconnected facts are brought together … into a consistent whole' (p. 23) by his 'Totem and taboo' hypothesis. Given the multiple variables at work in the clinical situation, which in this regard only reflects the human situation; given the complex, shifting nature of transference; given the difficulty of sorting out what the patient has innocently suffered at the hands of others, what he has provoked, what he has only fantasized, and with what he has been complicit … is it not heuristically judicious to adopt a concept of truth that refuses to lay claim to an objectivity that is not attainable?

Moreover, does not an analysis bring about changes in the meaning that events in the patients' pasts have for them in the present and future? - for example, a woman who was unable adequately to enter into, let alone resolve, the conflicts of the oedipal stage but who manages to do so in the transference comes, in the course of this experience, to remember her father as a sexually exciting object when, during childhood, she had experienced him only as an indifferent, zombie-like figure who was silent and remote. Is not this routine clinical experience evidence for Spence's 'pattern-making', for Ricoeur's 'narrated history', for Habermas' 'self-reflection', for the determination of the past by investments in the present and the concept of truth in psychoanalysis as coherent narration?

Moreover, the current state of psychoanalytic theory lends plausibility to the idea of coherence. There is no unified theory. There are only divergent, often mutually inconsistent theories supported by clinical observations. Does not this state of affairs cohere rather well with the coherence theory? Perhaps there are as many true life histories as there are theories that can give a consistent account of them?

Attractive as these possibilities are in certain respects, to these questions I believe the answer is No. The description above of the uncertainty of associations is tendentious and incomplete. Even when a patient is filling the hour with reports of manifest dream contents to the exclusion of associations, the details of the material are clear and determinate. There is nothing indefinite or illusive about it. Of course, it is unintelligible and uninterpretable in the absence of associations; but this fact has itself an obvious interpretation. The patient is anxiously clinging to the manifest dream content. This interpretation, properly timed and expressed and linked to the transference, will begin a process of change that will enable the patient to begin associating to his dreams. These associations will then also be determinate and discrete. If they are incomplete - as they are likely to be - it will be because further resistances are at work. If they become vague and uncertain it is for the same reason. Vagueness and uncertainty are themselves determinate states of affairs that have an explanation. They are not characteristic qualities of mental contents and states as such. The same is true of fantasies, memories, character traits, etc. Pattern-making by the analyst is not required so long as resistances and defences are interpreted in such a way as to allow the intrinsic forces at work in the psychic life of the patient to make themselves known. These forces will determine the pattern as they will determine the transference. The forces in question are the drives, their vicissitudes and their derivatives. The ideas of pattern-making, of theory-bound observation, and the like are rationalizations for countertransferential resistance to the threats posed by the drives, that is, by the instinctual unconscious. It is for this reason that psychoanalytic adherents of coherence have to find some way to banish them conceptually. Psychoanalytic theories that repudiate the drives are also likely to employ coherence as a concept of truth. Freud (1900), (1923) was certainly aware of the complexity of dreams and the extent to which they are representative of all mental phenomena; however Freud (Dora, 1905) also believed that the obscurities of a dream can be cleared up, that each manifest element can be traced along the paths of displacement and condensation from whence it came and that the meaning of the dream is to be found in the unconscious wishes of the dreamer. We are not always able to find the meaning, but it is there to be found, independently of any pattern-making activity on the part of the analyst. The task of interpretation as Freud conceived it is to make the interpretation correspond with the operative unconscious wishes of the dreamer - wishes that have a definite nature of their own. (For an opposing view see Viderman, 1970, 1972).

It is also true that Freud appreciated the extent to which any inquiry has to be guided by preliminary ideas. In this respect Freud's grasp of epistemology was more realistic and empirical than that of Bacon (1620), the great founder of modern empiricism. But Freud also thought that these preliminary ideas can and must be continually criticized and made to reflect the facts of observation more accurately. From the need to have a theory that will enable us to make predictions about what we will observe in order to make systematic observations, it does not follow that these predictions must govern what we will find. The preliminary ideas Harvey had concerning the circulation of the blood did not add or subtract anything from his crucial measurement of the amount of blood ejected by the heart in a single pulse. Hawking's (1988) mathematical derivation which proves, on current thermodynamic and quantum assumptions, that black holes emit particles does not affect the observations that will now have to be made on cosmic radiation to test the empirical truth of this derivation. Freud's prediction of the incidence of infantile seduction required by his seduction theory did not in some subtle way influence the number of such occurrences or Freud's ability to estimate them. Adequately formulated scientific theories or common-sense beliefs yield predictions and give rise to expectations that can be tested by observing what actually happens. These observations have meaning in their own right, independently of the theories or beliefs we have about their objects.

Freud (1927) defended science against those philosophers who assert that all knowledge claims, whether religious or scientific, are ultimately equal because they are equally subjective. Our observations, it was argued then as now, are inevitably conditioned by what we believe and how we observe. As we have shown above, various psychoanalysts are now advocates of variants of this idea of truth. Freud offered three arguments (biological, methodological, and epistemological) on behalf of scientific realism. Mental activities have developed in order to explore the world; it is likely that they have acquired a structure that facilitates that exploration (the biological argument). These mental activities are a part of the world; they can, themselves, be investigated in order to discover their degree of facility, its causes, and methods of improvement (the methodological argument). Scientific knowledge, because of its methods, is determined not only by our mental activities and their structures but, primarily, by the objects observed (the epistemological argument). I have elsewhere (Hanly, 1983), (1988) set out supporting evidence from cultural history for Freud's biological argument. His methodological and epistemological arguments are supported by evidence from the history of science. Science has been able to identify and to take into account the influence of our sensory apparatus upon our experience of nature. Copernicus and Galileo discovered the influence of the earth's daily rotation upon our observation of stars and planets and, by correcting for it, they were able to construct a genuinely objective description of the solar system. Einstein's invention of relativity theory enabled the human mind to realize that the self-evident rectilinearity of space is only a consequence of the organization of our sensory apparatus. Psychoanalysis has made its own contribution to this process of observational 'correction' for the field of human phenomena. These examples indicate that the expediency of the adaptation of the human sensory apparatus and thought activity to reality brought about by environmental pressure has been sufficient to allow it to proceed beyond the requirements of survival.

Freud employed a coherence criterion within the framework of his realist epistemology. Nowhere is the use of coherence more evident than in his effort to prove the objective reality of the Wolf Man's primal scene (1918). But even though his interpretation makes coherent sense of the details of the Wolf Man's infantile history and its connexion with both his infantile and adult neuroses, Freud does not claim that he had succeeded in proving that the primal scene was an occurrence rather than a fantasy. A crucial fact concerning the Gruska scene (the boy's urination) could only be established inferentially. Similarly, Freud only claimed that his hypothesis in 'Totem and taboo' was more plausible than existing theories and that it probably contained some measure of truth. He did not claim either that its coherence made it true or that such coherence constituted a limit beyond which knowledge could not reach. Freud used coherence as a necessary but not a sufficient criterion of truth. He took correspondence to be necessary and sufficient. Freud used coherence as a formal, logical criterion and correspondence as a material, epistemological criterion. Correspondence is built into the foundations of psychoanalysis. It is part of the meaning of the reality principle.

Yet Freud may have been in error. And, in any case, it would be to argue fallaciously from authority to treat as evidence what Freud believed rather than to weight the force of the arguments on which his view was based. Here are two additional arguments, one drawn from mathematics and physical science and one drawn from psychoanalysis.

Euclidean geometry is a mathematical system which is complete and completely coherent. For this reason it escapes Gödel's theorem. Yet it turns out that it does not describe the space of the universe. Physics has been able to identify facts which show this to be the case. The axioms of Euclidean geometry were believed to be self-evidently true. Plato, Descartes, and Leibniz considered them to be innate to the mind. Kant considered them to be a priori conditions of experience. Yet despite this self-evidence and coherence, and despite the fact that we actually do observe the world in a Euclidean fashion, these axioms have been shown by relativity physics to be approximations suitable only to regions of space smaller than the solar system.

Psychoanalysis is familiar, in the psychoses, with systems of belief, observation, and behavior that are remarkable both for their coherence and for their detachment from reality. The following bit of case history is representative. A psychotic student in a university seminar experienced growing agitation when certain topics were under discussion while the sounds of shuffling feet were accompanied by the sounds of a streetcar passing under the windows. This agitation cohered perfectly with his belief that such conjunctions of sounds signalled the approach of evil forces at work in his Manichean cosmos. Coherent as well with his beliefs were the ritual precautions he undertook to oppose the advance of those forces. Nowhere are the shortcomings of coherence as a sufficient criterion of truth more forcefully demonstrated than in our own field. Just as an argument may be valid and yet have a false conclusion, so a system of beliefs or a narrative may be coherent but false. The concept of coherence is not sufficient to bridge the gap between ideas and objects.

Moreover, psychoanalytic findings offer a defence against one of the philosophical criticisms of the correspondence theory. Putnam (1981) has argued that realism requires, in addition to the observer and the observed, a third party to whom reference can be made in order to compare the observer's perception with the object independently of the observer. How else could we form a judgment about its correspondence? Since the human observer is not in a position to form this judgement, realism is flawed with the hidden theological assumption of a God whose perception of objects is the ideal against which human perceptions can be measured.

But psychoanalysis has shown that this third-party observer is none other than the human observer himself. During the period of the pre-oedipal anaclitic bond, children, while having their own perceptions of things, use their idealized parents to carry out the very function assigned to God by Putnam's argument. Parental perceptions are taken to be authoritative - the standard against which children are able to measure their own experience. Certain ego regressed patients continue to have to establish a relation with an idealized figure because they cannot trust the evidence of their own eyes unless vouched for and authenticated by an authority. Even Harvey had to attribute his discovery of the truth about the circulatory system to Galen, in whose works no such idea is to be found. Psychoanalysis (Freud, 1923) ; (Waelder, 1934) ; (Hanly, 1979) has shown that the identifications involved in the resolution of the Oedipus complex normally bring into play a capacity for critical self-awareness that includes the perception of objects. An individual acquires the ability sufficiently to objectify his own perceptions and beliefs to enable him to consider to what extent they correspond with and are adequate to the object and to what extent they are not. This capacity forms the psychological ground for critical common sense and for scientific realism. The reality principle requires neither the alleged olympianism of correspondence nor the demiurge of coherence.

In psychoanalysis, through sympathetic identification clarified by countertransference awareness, this same self-critical capacity can facilitate our search after an understanding of our patients in their terms rather than our own. The complement of this self-critical receptive observation on the part of the analyst is the struggle for self-honesty in the analysand. The view that an analysis consists of a mutual construction by analyst and analysand of the analysand's life fails to do justice to this struggle. There are analysands who have been able to use the analytic process to discover more about themselves, to recover more of their past and find ways to reconcile themselves with it, than their analysts could comprehend. Fortunately for our profession and for our patients the process that the analyst facilitates can yield for the analysand a degree of resolving self-knowledge and improved functioning that exceeds those of his analyst. There is a common human nature, although to be sure not in the form of an Aristotelian essence, that exists in nature, that awaits our better understanding. It is embodied in the lives lived by individuals. These individual lives are part of nature. They are there to be known, however difficult that may be. The self-honesty of an analysand in his realization that he feared his father because he wanted to murder him, or of an analysand in her realization that her frigidity and the pleasure she took in rape fantasies was caused by her wish to use intercourse to castrate the man, implies that these realizations correspond with real wishes that continue to influence the individual's life. Neither the pain of those realizations nor their beneficial effects can be accounted for by any other assumption. In the end, each person has only his own life to live, however shared with others. At the core of the being of each person there is a solitude in which he is related to himself. Truth resides in this solitude to the extent that one can remember one's own past as it actually was. The ground of genuine analytic work in the analyst is his attitude of respect for this solitude.


A philosophical controversy concerning the nature of truth has begun to play an important part in psychoanalytic theorizing. The two major philosophical notions, the coherence and the correspondence theories and their use in psychoanalytic theory making, are examined. It is argued that although coherence is part of the criteria of truth, correspondence is the more essential and fundamental criterion. It was in this way that Freud used these concepts in creating psychoanalysis. Psychoanalytic discoveries concerning the psychogenesis of objectivity in perception and thought support the correspondence theory of truth and provide, in addition, an answer to the third party critique of correspondence. The correspondence theory as a basic attitude of mind is a necessary element in the respect for the patient upon which psychoanalytic therapy depends.


BACON, F. (1620) Novum organum. In Philosophers from Bacon to Mill ed. E. A. Burtt. New York: The Modern Library, 1939

BRADLEY, F. H. (1897) Appearance and Reality London: Swan & Sonnenschein.

DRAKE, S. (1978) Galileo at Work: His Scientific Biography Chicago: Univ. Chicago Press.

FEYERABEND, P. (1965) Problems of empiricism. In Beyond the Edge of Certainty: Essays in Contemporary Science and Philosophy ed. R. G. Colodny.

FREUD, S. (1895) Studies in hysteria S.E. 2

FREUD, S. (1900) The interpretation of dreams S.E. 4 & 5

FREUD, S. (1905) Fragment of an analysis of a case of hysteria S.E. 7

FREUD, S. (1909) Notes upon a case of obsessional neurosis S.E. 10

FREUD, S. (1912–13) Totem and taboo S.E. 13

FREUD, S. (1915) Instincts and their vicissitudes S.E. 14

FREUD, S. (1918) From the history of an infantile neurosis S.E. 17

FREUD, S. (1923) Remarks on the theory and practice of dream-interpretation S.E. 19

FREUD, S. (1927) The future of an illusion S.E. 21

GOLDBERG, A. (1976) A discussion of the paper by C. Hanly and J. Masson, 'A critical examination of the new narcissism' Int. J. Psychoanal. 57 67–70

GOLDBERG, A. (1988) A Fresh Look at Psychoanalysis: The View from Self-Psychology Hillsdale: The Analytic Press.

HABERMAS, J. (1971) Knowledge and Human Interests trans. J. J. Shapiro. Boston: Beacon Press.

HANLY, C. (1979) Existentialism and Psychoanalysis New York: Int. Univ. Press.

HANLY, C. (1983) A problem of theory testing Int. Rev. Psychoanal. 10 393–405

HANLY, C. (1985) Logical and conceptual problems of existential psychiatry J. Nervous & Mental Disease 173 263–81

HANLY, C. (1986) Review of The Assault on Truth: Freud's Suppression of the Seduction Theory and In the Freud Archives. Int. J. Psychoanal. 67 517–19

HANLY, C. (1988) From animism to rationalism In Philosophy, History and Social Action: Essays in Honour of Lewis Feuer ed. S. Hook, W. L. O'Neill & R. O'Toole. Dordrecht: Kluwer Academic Pub., pp. 221–34

HAWKING, S. (1988) A Brief History of Time New York: Bantam Books.

KANT, I. (1781) Immanuel Kant's Critique of Pure Reason trans. N. K. Smith. London: Macmillan, 1950

KANT, I. (1785) Groundwork of the Metaphysic of Morals trans. H. J. Paton. London: Hutchinson, 1947

KLEIN, G.S. (1976) Psychoanalytic Theory New York: Int. Univ. Press.

KOHUT, H. (1959) Introspection, empathy, and psychoanalysis: An examination of the relationship between mode of observation and theory In The Search for the Self Volume 1 ed. P. H. Ornstein. New York: Int. Univ. Press, 1978 pp. 205–32

KOHUT, H. (1971) The Analysis of the Self New York: Int. Univ. Press.

KOHUT, H. (1977) The Restoration of the Self New York: Int. Univ. Press.

KUHN, T. S. (1970) The Structure of Scientific Revolutions Second edition, enlarged. Chicago: Univ. Chicago Press.

MAY, R. (1958) Contributions of existential psychotherapy In Existence: A New Dimension in Psychiatry and Psychology ed. R. May. New York: Simon & Schuster, 1967 pp. 37–91

MERLEAU-PONTY, M. (1945) Phenomenology of Perception trans. C. Smith. London: Routledge & Kegan Pual, 1962

PUTNAM, H. (1981) Reason, Truth and History Cambridge: Cambridge Univ. Press.

RANGELL, L. (1988) The future of psychoanalysis: The scientific crossroads Psychoanal. Q. 57 313–40

RICOEUR, P. (1970) Freud and Philosophy New Haven: Yale Univ. Press.

RICOEUR, P. (1974) The Conflict of Interpretations ed. D. Ihde. Evanston: Northwestern Univ. Press.

RICOEUR, P. (1981) Hermeneutics and the Human Sciences trans. J. B. Thompson. New York: Cambridge Univ. Press.

RYLE, G. (1949) The Concept of Mind London: Hutchinson's Univ. Library.

SARTRE, J.-P. (1943) Being and Nothingness trans. H. E. Barnes. New York: Philosophical Library, 1956

SCHAFER, R. (1976) A New Language for Psychoanalysis New York: Yale Univ. Press.

SCHAFER, R. (1978) Language and Insight New Haven: Yale Univ. Press.

SPENCE, D. (1982a) Narrative Truth and Historical Truth: Meaning and Interpretation in Psychoanalysis New York: W. W. Norton.

SPENCE, D. (1982b) Narrative truth and theoretical truth Psychoanal. Q. 51 43–69

VIDERMAN, S. (1970) La construction de l'espace analytique Paris: Denoel.

VIDERMAN, S. (1972) Comme un miroir, obscurément… Nouv. Rev. Psychoanal. 5 131–152

WAELDER, R. (1934) The problem of freedom in psychoanalysis and the problem of reality testing In Psychoanalysis: Observation, Theory, Application New York: Int. Univ. Press, 1976 pp. 101–20

WALLERSTEIN, R. S. (1988) One psychoanalysis or many? Int. J. Psychoanal. 69 5–21

[Note: A version of this paper was presented at the 36th International Psychoanalytic Congress, Rome, July 1989, and appeared in The International Journal of Psychoanalysis, 1990, 71: 375-383. It appeared also as the first chapter of Charles Hanly's book The Problem of Truth in Applied Psychoanalysis, New York: Guilford Press, 1991. We thank the The International Journal of Psychoanalysis for the permission. Copyright © Institute of Psychoanalysis.]

Web Editor: Paolo Migone. For suggestions or corrections, please e-mail to: