The meaning of dreams in the psychotic state: Theoretical considerations and clinical

Paola Capozzi and Franco De Masi

The authors consider that the Freudian theory of dreams is not directly applicable to psychotic and borderline patients with their constantly varying states of mental integration. Because these patients’ dreams lack associations, the usual psychoanalytic approach cannot be used to ascertain their meaning. After reviewing the literature on the specific quality of dreams in the psychotic state, the authors point out that such dreams have nothing to do with the metaphorical language of the dream work but instead express the concreteness of the hallucinatory construction. For this reason, a dream’s meaning may fail to be understood by the patient even if it seems clear to an observer. Yet the analyst’s reception of a ‘psychotic dream’ is a unique and essential source of valuable information on the manner of construction of the delusional system, allowing analytic work on the psychotic nucleus. In the authors’ view, such dreams may help the analyst and the patient-while still lucid-to acquire insight, thus affording a stable foundation for emergence from psychosis. The paper includes some case histories, in one of which a psychotic female patient is enabled by work on dreams to reconstruct a psychotic episode and thereby to ward off an imminent fresh lapse into psychosis.


The Schizophrenia Bulletin of the US National Institute of Mental Health includes a section, ‘First Person Account’, which is open to patients and their families. In some of these articles, patients diagnosed as psychotic discuss the way they live and their problems.1 One of these descriptions seems to illustrate particularly well the state of mental imprisonment in which some patients live even when they have succeeded in recovering a part of their social capacity.

The patient writes2: ‘Over the years since then I have had many “mad” thoughts, but my main delusions, of grandeur (mediumistic phenomena) and persecution, have remained constant. These delusions are supported by such vivid hallucinations that I cannot usually distinguish them from reality, except by saying to myself that none of this is real, that what the people I love and trust tell me is true, that I am unusual only because I am schizophrenic, and that everything I think about the supernatural is due to an illness’.

This statement helps us to understand how the relationship between the healthy and psychotic parts of the personality resembles a precarious compromise that can very quickly break down. The delusion is always present as a ‘second system’ and can be activated at any time.

After all, the psychotic condition cannot be seen as something static, in balance with something else (from this point of view, the terms ‘psychotic part’ and ‘neurotic part’ used first by Katan and later by Bion may be misleading), but is in fact an entity endowed with a dangerous power, for the psychotic world can always colonise and engulf the healthy one.

To preserve their psychic equilibrium, some patients eventually decide not to experience their own emotions. As another patient puts it: ‘Flat emotions are an obvious compromise to reduce psychotic symptoms. As a result, one feels emotions at a lower level than a normal person would.’

The kind of improvement to which this patient aspires (which can also be secured by spontaneous remissions or by treatment with drugs only) is anti-analytic in that it runs counter to the purpose of therapy, which is to recover the unity and completeness of the subject’s self. However, we know that any attempt to recover emotions and vitality entails the risk of relapse and that, when the analytic process facilitates integration and a closer approach to the patient’s emotions, a fresh crisis becomes possible. Attempts at improvement and integration will therefore always be risky.

It is perhaps for this reason that psychoanalysis has traditionally been deemed ineffective, if not positively contraindicated, for psychosis.

An analyst engaged in the treatment of psychosis must, of course, internalise a mental setting appropriate for the psychotic patient and be capable of understanding the specific condition of that type of mental functioning. Keeping the emotions ‘flat’ is a defence deployed by the patient because even pleasurable experiences may be felt too intensely to be tolerable. Again, as the patient improves, mental activity or affective cathexis may be feared on account of the risk that they might be transformed into omnipotent devices potentially beyond mental control.

The state of inhibition that often characterises the therapy of these patients for long periods may also result from underlying confusional states which make it impossible for the patient to distinguish the meaning of his or her emotions, so that vital aggressive aspects are confused with destructive explosions (De Masi, 1997). This emotional paralysis often conceals the activity of a confusing and terrifying superego.3 It is therefore essential for the analytic therapy to create the conditions whereby the psychotic nucleus can be confronted, with a view to reducing its power.

Until the underlying delusion has been understood and transformed, the psychotic experience continues to bring about a state of intimidation that conditions the patient, who sees himself as a person whose liberty is restricted-a prisoner who can feel relatively safe provided that he does not overstep certain limits and move on to dangerous ground.

As soon as an improvement is discernible and the patient’s collaboration can be relied upon, the conditions predisposing the patient to the specific delusion must be explored. In this sense, only the reconstruction of the psychotic state and a profound understanding of the reasons for it can reduce the risk of its recurrence.

One of the complex and paradoxical aspects of the analytic therapy of psychotic patients is that, for long periods, the ‘psychotic part’ is absent, invisible or impossible to grasp in the clinical material, and that, when it emerges, it does so suddenly and unexpectedly, after the transformation into ‘psychosis’ has taken place.

This paper postulates that exploration of the dreams of certain psychotic patients may prove useful for understanding their psychosis and contribute to the analytic work of modifying the delusional state.

We begin by investigating the structure and meaning of dreams in the psychotic state. These considerations are important because they suggest a possible analytic approach to psychotic dreams where the preconditions required for full application of the analytic method are lacking. As we know, one of the difficulties encountered in analysis of the dreams of psychotic patients is the absence of associations. Since these patients are unable to make significant connections to their dreams, they bring them into their sessions in such a way that their meaning cannot be analysed or explored in depth.

To demonstrate how dreams may prove useful for understanding the psychotic state and, more especially, how they can provide the analyst with valuable information that cannot be obtained from other material, some clinical examples are given below. A case treated by one of the authors is discussed at greater length.


Freud (1911) notes that, at the time of onset of Senatspräsident Schreber’s second psychotic episode, he dreamed more than once that his old nervous disorder had come back. Once, in the early hours of the morning, while he was in a state between sleeping and waking, the idea occurred to him that it must be very nice to be a woman submitting to the act of copulation. Schreber later had the delusion that he had been transformed into a woman and was being fertilised by divine rays. In Freud’s view, such patients are struggling with the same material as will feature in their psychosis. Subsequent authors (e.g. Katan, 1969) have observed that the psychosis may begin with a dream and that psychotic patients do not distinguish between dreams and psychotic symptoms.

On the relationship between dreams and psychosis, it is worth rereading the paper by Mack (1969), which reports on an American Psychoanalytic Association panel held in Boston in 1968. Among the participants, Frosch (1969) wonders whether there is anything specific about the mental content or affective quality of a psychotic dream. He concludes that a definitive answer cannot be given, because some dreams exhibit nothing specific, whereas others include a persecutory content in which the psychosis is expressed. Frosch considers an important and characteristic aspect of psychotic patients to be their inability to distinguish dreams from reality: their dreams are so vivid that they are confused with the events of the day. Postulating that dreams may be an attempt to forestall the illness or to control the conflict, Frosch enquires as to the further significance of dreams that herald psychosis. Dreams of disintegration of the world might reflect the disintegration of the ego or, alternatively, they could represent a higher level of psychic development in which the ego reacts to the drives directed against the external world and the self.

Finally, in the view not only of Frosch but also of Katan, dreams sometimes not only plunge the dreamer into psychosis but actually become part of the delusion, so that they are incorporated into the psychotic behaviour (the dream then being the construction of a delusional reality). Frosch thinks that in these cases ego shortcomings allow the emergence of pathological phenomena in the waking state, or else excessive instinctual pressure is present. As a result of the ego deficiencies, the dream may be accepted as reality.

Atkins (1969) also thinks that psychosis may be anticipated by certain dreams, which in this case will be of two kinds: those expressing the regressive and ego-disorganising potential and those directed towards defence and delusional restitution. In the latter type, the manifest content displays a coherent defensive construction similar to delusion, erected in an attempt to counter the threat of disintegration.

The relationship between dreams and psychosis is complex even on the simple phenomenological level.

Certain authors (for example, Racamier, 1976) have outlined some clinical characteristics of the dreams of psychotic patients. The first relates not so much to the dream as to the activity of dreaming, which, in acute psychotic phases, becomes confused with external reality. In psychotic dreams, especially in the acute psychotic phase, there is often no distinction between the waking state and that of sleep, and between hallucinations and dreams; nor is there any distinction between delusions, hallucinations and nocturnal events that might be called dreams, to which they are often psychically equivalent. In these acute phases, dreams may be experienced as real events.

The second characteristic concerns the ‘psychotic dreams’ occurring not infrequently in the therapy of borderline patients or in the intervals between psychotic episodes. These dreams feature violence, sexual sadism, sudden aggressive explosions and monstrous nightmare figures, which sometimes worry the patient but more often alarm the analyst.

Finally, there are aspecific dreams – i.e. ones indistinguishable from those of neurotics.


Analytic theory investigates the meaning of dreams by linking them closely to the dreamer’s mental functioning.

For Freud (1900), the dream work consists in fulfilling, albeit in disguised form, an unconscious wish that would otherwise disturb the dreamer and interrupt his or her sleep. Dreams also serve to reduce and modulate the tension of the drives and to express instinctual wishes in manifest form through the distortion of the censorship (Freud, 1922). Dreams enact a conflict of which the patient is not conscious, but their interpretation allows the conflict to be resolved by making it conscious. The interpretation of dreams thus constitutes the royal road to a knowledge of the unconscious and to the understanding and subsequent transformation of psychoneurotic disturbances. Translation of the manifest content into the underlying latent thoughts is made possible by a knowledge of the mechanisms of the dream work, such as condensation and displacement. This work transforms the latent dream thoughts, which are unacceptable to the ego, into the manifest content. The only exception to this process are traumatic dreams, which are a repetition of the traumatic experience in which the function of fulfilment has failed (Freud, 1920). They are an attempt to master the traumatic event, but, like the repetition compulsion, are one of the phenomena that led Freud to his theory of the death instinct.

Freud (1940) considers psychosis to be analogous to dreams because it exhibits the prevalence of the instinctual wish and of the primary process, the lack of the notion of time, the absence of contradiction and the hallucinatory fulfilment of the wish. On this basis, Freud suggests an interpretation of the latent content of Schreber’s psychotic delusion, according to which the persecutory state conceals his repressed homosexuality.

Modern hypotheses on the relationship between dream mechanisms and the mental functioning of psychotic states have not confirmed this equivalence, but, in their enrichment and supplementation of the Freudian model, throw light on the modalities of psychotic thought.

There are indeed mental states that abolish the capacity to represent wishes and emotions, and which therefore interfere with the function of dreaming.4 It follows that the meaning of dreams differs according to the patient’s mental state. This clinical problem arises particularly with patients such as borderlines and psychotics, whose state of integration constantly varies. It may therefore be postulated that the function, meaning and use of dreams differ according to the mental state and structure of the analysand, and that different mental states give rise to ‘different’ dreams with diametrically opposed meanings.

Today, however, we see dreams very differently from our predecessors. Dreams for us are a particular manifestation of the symbolic function, a first step in the process of thought. Whereas Freud considered dreams to be the hallucinatory fulfilment of a repressed infantile wish, and repression to be the engine of the psychical apparatus, Klein held that it was not the repressed that gave life to dreams but instead the dynamic between internal objects, which were formed by splitting and by projective and introjective identification.

Further intuitions about dreams and the process of dreaming were contributed by Bion, who saw a dream as an activity that creates thoughts rather than a limited functional regression revealing a repressed content. Riolo (1983) points out that, whereas a dream for Freud was the distorted representation of an already existing unconscious meaning, Bion saw dreams as attempts to generate a new meaning on the basis of the perception of an emotional experience. According to Bion, in order for the perceptions of an emotional experience to be used for the production of a dream, they must take on the characteristics of thought – that is, they must be transformed by the alpha function. If the alpha function is damaged, the beta elements will remain unmodified and emotions will be experienced as sensible objects, so that the patient cannot dream. Both dreams and the unconscious are products of the differentiation effected by the alpha function. For beta elements are non-thoughts that cannot become a part either of consciousness or of the unconscious; they can be neither known nor repressed. Dreams constitute a barrier, or alpha screen, which creates consciousness and continuously differentiates it from the unconscious. Dreams are the royal road to the unconscious and constitute evidence of the functioning of the alpha screen and the contact barrier. However, if the contact barrier is destroyed and replaced by the beta screen, it is no longer possible to distinguish conscious from unconscious, to repress, to remember, to think, and hence also to dream. In Bion’s view, the destruction of the alpha function makes it impossible to store experience: patients feel that they contain not the visual images of things, but the things themselves. Conversely, things are perceived in the same way as ‘thoughts’ and ‘ideas’ are seen by non-psychotics: the patient expects them to behave as if they were visual images in his or her mind. The alpha dream work allows a neurotic subject to experience reality through unconscious thought, which is always active in these subjects; as Bion put it, ‘the dream-work we know is only a small aspect of dreaming proper-dreaming proper being a continuous process belonging to waking life’ (1992). Unlike psychotics, neurotics understand the experience of reality through alpha elements, which also constitute the foundation of dreams. Hence a neurotic dream does not correspond to the psychotic’s omnipotent fantasy, but is the opposite of it. In psychotic states, ‘dreams’ remain attempts at construction (from chaos to the formulation of meanings), but are doomed to fail and remain unused because the function of thought is lacking.

Many analysts have noted the particularity of the function of dreaming in psychosis, but have interpreted it in different ways. Bion (1967) had already pointed out that dreams in the psychotic state perform a discharge function. Green (1977), too, suggests that the purpose of dreams in psychosis is not so much to express the fulfilment of a wish as to evacuate-to rid the psychical apparatus of an excess of painful stimuli. Segal (1991) asks what happens to dreaming and dreams when the ego is temporarily or permanently unable to carry out the task of producing a normal or neurotic dream. In order for dreams to perform their function of elaboration and insight, the benign interaction of projection and introjection must accompany the constructive container-contained relationship. In this author’s view, the classical theory of the function of dreams takes it for granted that the ego is capable of repression and that the subject possesses the capacity for symbolisation, which in turn calls for repression. If, however, excessive projective identification is present, concrete thought-confusion between the object and its symbolic representation-will result. Only the perception of separateness and the working through of mourning can allow the development of the capacity for symbolisation. Segal draws attention to the technical problem of analysing dreams in which the dream work is defective: it is totally useless, if not counter-productive, to analyse the dream content as if it referred symbolically to something else. A convincing and extreme example is that of a sado-masochistic patient who had succeeded, during the course of his analysis, in giving up perverse fantasying in the waking state, but who for a while used night dreams to enter the perverse dimension that was closed off while he was awake. Dreams for this patient constituted not ‘dreaming’ but ‘acting out’. As soon as the vigilance of conscious life gave way to the onset of sleep, he began to perform actions in dreams, whereby he again achieved a similar perverse satisfaction to that of the former sado-masochistic situation of waking life. Instead of actively summoning up fantasies with his eyes open during the daytime, he ‘dreamed’ them at night.

In such cases it is superfluous to analyse dreams, but the operation of dreaming is seen to throw important light on the dreamer’s mental functioning.

Segal, too, refers to Bion’s model and his concepts of alpha and beta elements, as well as that of the mother capable of containing the child’s projective identifications. Discussing two of her borderline patients, she mentions predictive dreams-that is, ones that foreshadow action, in which what has been dreamed is to be acted out. The action of the dream is totally literal and is performed down to the slightest detail. Rather than substituting for acting out, these dreams contain every detail of what is to be acted out. Predictive dreams would seem to operate in the same way as what Bion (1963) called a definitory hypothesis. They differ from evacuative dreams, in which certain internal perceptions of the patient are successfully got rid of, in that they appear to remain in the psyche as a bad object that must be disposed of by acting out the dream.

In Bion’s terms, psychosis is not a dream but the opposite of a dream. Not only do psychosis and dreams not coincide, but the dreams that appear in the psychotic state are not, in fact, dreams, even if they appear to be so, but instead concrete facts or acts of evacuation of parts of the self. These patients’ dreams cannot be genuine dreams because they are not the result of a process of symbolisation and because there is no internal mental space to contain them. These dreams thus no longer conform to the rules applicable to the unconscious laid down by classical metapsychology: they are not subject to censorship and their manifest content does not refer to a latent content. What may be regarded as the psychotic unconscious is not, in fact, such, because there has been no dream work (which is the precondition for representation, symbolisation, thinkability and the unconscious). In this context, dreams may be fragmentary and meaningless constructions, but they may at the same time constitute the beginnings of the construction of an internal space and a potential communication.

The general validity of the traditional Freudian view of dreams is not denied, but the conditions for its complete application are lacking in psychotic patients.

A useful approach in our terms is that of Meltzer (1983), who investigates how dreams succeed in creating symbolic forms capable of representing emotional experience and hence truth. Exploring the path followed by dreams in the attempt to construct the meaning of an emotional experience, he shows how these experiments sometimes succeed in solving the problem and sometimes fail. In the latter case, the dream process becomes psychopathological, resulting in the formation of hallucinations or delusions.

Like hallucinations and delusions, lies are a distorted representation of emotional experience. The psychotic experience is underlain by a conscious distortion of truth, due to attacks on the organs of perception that prevent the use of consciousness.

Combining the above considerations, Blechner (1983) writes that dreams in psychosis not only perform an evacuative function but sometimes also have a constructive meaning. They often evince a capacity to think, whose expression may be suppressed and removed from consciousness. Patients may be reluctant to work on the themes expressed manifestly in their dreams because the method of association would interfere with their intention to communicate with the analyst directly. The authors suggest that the situation in these cases resembles that of a psychotic patient who occasionally has a rational thought. It is then inappropriate to stress or interpret the parts that continue to function. Blechner thinks that in the case of borderline dreams the unconscious cannot be regarded as an entity with specific contents and processes (e.g. repression), as in classical theory. Borderline dreams express a quality of psychic experience that lacks consciousness (and here the term ‘unconscious’ is used mainly adjectivally). In the author’s view, as a borderline patient improves, the dreams, whose content tended to be very concrete in the early phases of the analysis, become less sharply delineated and more complex, exhibiting a greater capacity for symbolisation.


In a recent contribution, Blechner (2000) notes that a quick perusal of the psychoanalytic literature reveals examples of dreams incorporating psychotic material. Referring to a patient of Silvano Arieti (1963) who had emerged from a delusional persecutory state, he notes that a dream had heralded progress in her psychoanalytic therapy. Resnik (1996) also draws attention to the role of dreams in borderline pathologies, where they often constitute attempts at reconstruction.


A young male psychotic patient, A, in analysis during a period of remission, emphasised in every possible way how difficult he found it to communicate his ‘dreams’, as if they were among his most intimate, personal and private possessions, which he was afraid the analyst might destroy or annihilate.

A once reported that he had dreamed of a church with a tennis court. This was at a time when, feeling better, he had taken up this sport again, having greatly enjoyed it during his teens. The patient added nothing to this statement, and the analyst also felt unable to say anything about this dream.

In subsequent sessions he went on talking about his pleasure in playing tennis, and said that he had entered for a tournament organised by his club. Eventually, however, it became clear that he was totally and delusionally identified with the then world champion, John McEnroe. An instance of delusional behaviour acted out in his parish church later revealed that he also saw himself as a reincarnation of Jesus Christ. Only then did it become possible to understand the communicative value of his ‘dream’, in which he had concretely specified two aspects of the content of a grandiose delusion, sport (the tennis court/McEnroe) and mysticism (the church/Christ).

As we were later able to understand, the ‘dream’ represented the psychotic solution which the patient wished to keep to himself in case the analyst destroyed it and prevented his flight into delusion. It then occurred to the analyst that, although he had improved (or perhaps precisely because of the improvement), A had become desperately aware of the severity of the disturbance and of the impossibility of a genuine reconstruction of the self, so that he was again impelled towards a delusional solution. As with other patients, for A too the moment of improvement and clarity corresponded to a precarious equilibrium that could easily break down and plunge him back into psychosis.

All this pain is, of course, not normally verbalised by the patient; it may be intuited only by a psychoanalyst experienced in working with psychotic patients, who is aware, partly through his experience of other cases, of the risks accompanying attempts at integration. We contend that in such cases patients dream of something that is going to happen, which is only indirectly connected with the analytic relationship but concerns their possible future mental functioning-for example, the development of a new psychotic state, the themes of which are prefigured. In other words, we postulate the existence of ‘predictive’ dreams that anticipate the onset of the psychotic state and specify the basic elements that will be contained in the delusional construction. According to our hypothesis, such dreams arise when the psychotic crisis is not yet manifest.

Patient A, who brought the dream of the tennis court and the church with great difficulty and without associations, produced his ‘hallucination’ in a session, but it was not possible to work on it because the analyst stubbornly held fast to the idea that the dream referred symbolically to something else.

While emerging from the megalomanic state indicated in the ‘dream’, the same patient subsequently brought another dream apparently representing the psychotic transformation and persecutory guilt.

A dreamed that he had been invited to the home of an analyst couple who allowed him to get into their marriage bed. There he became turned on as he performed a series of twisting and turning movements, ending up in a blissful state of pleasure. Then the figures of his parents appeared, faded and were far away in the distance. Eventually he found himself surrounded by deformed basset hounds with their heads turned back, which complained that he was responsible for their deformity.

On this occasion A said that he felt that the little deformed animals in the dream had to do with himself and were like little selves of his own. In the dream the patient had visualised the state of masturbatory pleasure that had underlain the onset of the delusion of omnipotence, but he was now conscious of the destructive quality of this process, which was responsible for causing mental damage and for weakening the analyst and the parents. The entire process of psychotic transformation, which had come about ‘in silence’, could be visualised only retrospectively, once the patient had become capable of transcribing it in the dream images.

Since an unconscious distortion of the truth lies at the root of the psychotic disturbance, understanding dreams in psychosis is always an awkward and complex matter; it is difficult to make use of the dream content even when such dreams seem to be attempts at communication.

Because the distortion affects the perceptual apparatus, cancelling out the sense of truth and shutting down the organs of consciousness, Meltzer (1983) tends to equate lies and delusions. However, whereas in the case of lies the alteration of truth is conscious and the attempt to distort the truth is intended to confuse the other, in psychosis the first person to be confused is the patient himself, who shows his agreement with the liar.

Yet if the psychopathology of the process of dreaming coincides with the distortion of thought and the formation of hallucinations or delusions by mental self-poisoning, valuable traces or signs of the origins of this poisoning might perhaps be found in the dreams of psychotic patients. One of the hypotheses advanced in this paper is that ‘dreams’ occurring during the course of a psychotic process may have a meaning: they may be the patient’s way of describing mental states specific to the psychosis, of allowing the dreamer and the analyst to understand and anticipate the nature of the psychotic poisoning or, conversely-albeit with more difficulty-of describing a possible way of emerging from it. Sometimes, in the optimum case, they may constitute an attempt to make delusional contents assimilable-that is to say, to dream the psychosis in order to act it out again in the potential space of the analytic relationship.

We now present a dream in which a patient in a persecutory psychotic state repeats the persecutory psychosis by representing it. The patient, B, is in a dark tunnel like an underground railway or subway and is trying to get back into the open by climbing up a ladder. The ladder is very long and weird, as if in a painting by Miró. While climbing in the dream, B meets a man with a notice saying: ‘These last five minutes will be the most terrible of your life’, and the persecutory prediction immediately comes true. As he climbs, the patient feels a spray of liquid on his shoulders; he is then surrounded by a group of threatening men, who grab him by the shoulders and violently plunge him down a steep slope.

B had begun the session by talking about Marie Cardinal’s book The Words to Say It; this had comforted him, because he had read that at a certain point in her analysis the woman patient described had emerged from the nightmare of hallucinations and begun to dream.

What distinguished the experience of the dream brought in the session from the ordinary experiences of persecution afflicting the patient virtually every day? For the sake of brevity, we shall disregard all the implications of the delusion in terms of B’s history and personality structure, as well as the possible transference meanings, and concentrate on the new element introduced by the dream.

In the dream the persecutory experience, which seemingly repeats the persecution of waking life, takes on a new meaning precisely because it is dreamed, and the patient who constructs the dream becomes the bearer of intentionality. The intentionality of the dream is to convince the patient (and the analyst) of the ineluctability of persecution: the delusional part delivers him up, in a state of terror and total isolation, into the hands of his enemies. No one comes to his aid nor, in the dream, does the patient ask for help.

However, if, as B had indicated in advance by his reference to Marie Cardinal, a dream is a communication, a space shared with the analyst, then the delusion no longer takes place in total isolation. By interpreting the dream as propaganda of the delusional part (the part that wants to convince the patient that no escape is possible, plunging him back down into the dark, and describing its power), the analyst can show the patient that it is a part of himself that creates the delusion and wants to believe in it.

Enumerating other episodes when the delusional intuition appeared out of the blue, B said that a ‘click’ would sound in his mind and that when this happened there was no escape. If the click did not occur, there would be no terror and he would not be a prey to delusion. The analyst replied that the click was facilitated by a mental configuration that pushed forward and conquered his mind when he isolated himself in fantasy and believed himself to be alone against the world. When he isolated himself, he lost all protection and perception of friendship. He would then enter into the mode of the ‘enemies’, who would subject him to incredible force and confront him with ever more refined and sophisticated persecution systems, which only his own fantasy (represented in the dream by the weird ladder) could create-for example, dusts, smells, television cameras, and the like. The process had begun a year earlier when he himself had developed a competitive and megalomanic attitude.

During the psychotic episode, B had attributed his failure not so much to the megalomania, which had not stood up to reality and the consequent frustration, as to the ‘enemies’ who were still persecuting him today. The weird, surrealistic ladder was connected with fear, because it alluded to the danger of being completely captured by the psychosis and of discovering too late (the ‘last five minutes’) that he had wasted his life in the dark tunnel of delusion.

A few months later, during a period of distinct improvement, the same patient brought a very long dream in which, at the Olympic Games, a male Italian athlete was taking part in a difficult event and he felt impelled to help him. A group of female athletes were performing movements in which they flew up into the air and defied the laws of gravity. He watched them in admiration but knew that he would not be able to imitate them. Later, a group of athletes were engaged in an event in which he too would have liked to take part.

Notwithstanding the many grandiose elements in the dream (Olympic Games, astonishing movements and events with heroic winners), B himself-associating with his own pain when his national football team eventually won a difficult match-manifested the good atmosphere that not only appeared in the dream but which he himself experienced when he recovered the area of relationship, co-operative playing and possible identification with coevals. He had indeed experienced a similar atmosphere of well-being, which he had been able to contrast with that created by the persecutory solution.

It should be noted that, for B, the two dreams were not only dreams but also configurations representing two possible mental dispositions of the waking state that conditioned the perception of reality.

We now present some excerpts from the analysis of a case that bears out the hypothesis that certain dreams ‘predict’ psychosis and occur before the psychotic crisis actually becomes manifest.

The patient, C, was 23 years old and had, while still at high school, already suffered from fits of anxiety and periods of isolation, anorexia and obsessional dieting; she had been referred for analysis by a psychiatrist who had treated her when she was hospitalised for several weeks after a bizarre attempt to ingest objects, which had been thwarted by her mother in time to stop her suffocating herself.

During her few weeks in hospital, the patient had had threatening auditory hallucinations and delusional ideas of guilt. The diagnosis on discharge had been uncertain-somewhere between borderline pathology and incipient schizophrenia.

Having felt empty and abulic at the beginning of her analysis, C said that, before her hospitalisation, she had felt herself to be very intelligent, lucid and creative, and had immersed herself in reading the great authors, in particular Virginia Woolf. In a crescendo of mental excitement, she had conceived the idea of being Lucifer, the bearer of light, God’s chosen angel. Later she had felt demoniacal, the bearer of all the world’s evil, and had understood that the only way to ward off the imminent end of the world was to do away with herself. She had then suddenly lost her mental lucidity. The elated state of the psychotic episode, although it had aroused anxiety, had also seemed to constitute a special, extraordinary condition, and was remembered not without a sense of longing.

This patient’s therapy can be divided into two parts. The first was characterised by psychotic-type anxieties, while in the second she lapsed back into psychosis.

In the first year of her treatment, C had not presented any psychotic symptoms, the material having instead revealed depressive and loss-of-identity anxieties. Given the seriousness of the case, the analysis seemed to be going well, and transference interpretations, especially of her fear that the analyst might be unable to understand or support her, enabled her temporarily to overcome the various attacks of emptiness and anxiety.

Although the analysis was going well, the impression might even then have been gained that things were going ‘too’ well, and it might have been felt that, as regards improvement, C was being in part very obliging towards her analyst, almost as if she were conforming to the latter’s therapeutic expectations.

At any rate, notwithstanding her belief that she had devoted a great deal of attention to the material, the analyst was surprised by a fresh psychotic episode that began after the patient had consented to her father’s insistent demand for her to go away with him. This had involved her missing a week’s sessions.

During the trip C had fallen ill again, suffering from the delusion of being sexually excited by her father, who was seen as the Devil. The trip was broken off and the patient returned to the psychiatric clinic.

The virulence with which the psychotic symptoms resurfaced showed that all the analytic work thought to have been accomplished had been swept away in an instant. The analyst wondered how she could have failed to recognise the persistence of the psychotic situation on such a massive, unmodified scale. The analysis had manifestly proceeded up to that point in an excessively reassuring climate inconsistent with in-depth penetration into the psychotic nucleus, which had continued to exist and to exert silent power over the patient.

Upon the resumption of her analysis, C said that she had experienced the trip as an incestuous event. It had seemed to her indecent for her father and herself to be there among a group of couples. People had taken them for a husband and wife on honeymoon. The room in which they had slept, with its double bed, had appeared to her to be an enormous red vagina, suggesting ‘sinful’ thoughts towards her father. During the psychotic episode she had been terrified of the analysis and of her analyst; she had thought that the analyst would be furious because she had broken certain laws or destroyed her (the analyst’s) brain. Already in the previous months, the patient had every so often fleetingly mentioned that she felt she was entering into the analyst’s thoughts.

The separation from the analysis had constituted a dramatic discontinuity, in which the previous psychotic episode had been repeated. In the new delusional episode, the patient seemed, by the incestuous flight with her father, to have acted out an attack on the (female) mother-analyst, who was transformed into a damaged and vengeful figure. For this reason the patient had felt guilty towards the analyst and threatened by her, representing as she did the psychotic, vengeful superego.

At the height of this destructive state, C had felt diabolical (as in the first episode, in which she had been Lucifer), in league with the figure of the father, who also represented a part of herself-the part that urged her to fling herself into a manic, sexualised reality.

Confronting the fresh psychotic episode, the analyst realised that she had arrived on the scene ‘after the event’, when the psychic catastrophe had already taken place. Because she had been unable to oppose the father’s wish (which, moreover, had sounded to the patient like a diktat), the analyst must have been placed in the position of a weak and vulnerable mother swept aside by the paternal sexualisation. The psychotic episode induced the analyst to review the prior analytic material; in the light of the events described above, her attention focused on a series of dreams that seemed particularly meaningful, in which the patient had mentally anticipated and virtually predicted the psychotic episode that was to ensue shortly afterwards. These dreams had remained unanalysed partly because they were so terrifying that the patient had been unable to produce any associations to them, and partly owing to their intrinsically ‘alien’ quality. With hindsight, it was easier to understand these dreams as a communication of the psychotic mental state to which both analyst and patient had closed their eyes.

Two dreams that had preceded the crisis and whose meaning could have been recognised referred to the erotisation of the feared father figure, as subsequently manifested in the delusional explosion:

‘We are skiing in the mountains, and are to go to our rooms in the evening. I hope to end up in the room of a boy I like, but instead I find myself going into the room of Father X, an Italian teacher. He makes to seize hold of me, I run away, he follows me, and the house turns into a castle. Wherever I go, fires break out …’

‘My family are at home having an orgy; they seem to be drunk, and there are people jumping on the beds. My family are amused to see me turn up in this awkward situation. I act furious, and try to get my creams and make-up. But I can’t do it, because I am somehow hanging up, my hands tied, as if in a porno movie.’

In these dreams C had described how, by transforming her parents into perverse objects, she was able to create an orgiastic state of mind and consequently trap herself in a sexualised world. Her acquiescence (‘I am somehow hanging up, my hands tied’) lay at the root of her entrapment in psychosis and her total loss of the perception of human relations.

Further clarification of the complexity of the psychotic anxieties and persecutory guilt that followed the annihilation of C’s healthy self was afforded by another dream, which she brought in terror a few months after the second psychotic episode:

‘My sister and I are prostitutes in a brothel. Each of us has a speciality; one of us goes to bed with murderers. We are expecting Robert De Niro from Cape Fear. Everyone is terrified because he is a savage murderer. We wait; an Indian prostitute ties herself to the foot of the bed, saying that she must suffer the death penalty because she neglected her child in order to be a prostitute. A general silence descends and I burst out crying to the depths of my being in enormous sadness.’

This dream is particularly touching because it betrays the punishment that threatens the patient when she has no choice but to submit to the power of the psychotic part, which is represented as invasive, terrible and terrifying (Robert De Niro and Cape Fear). The patient feels not only guilty but also bewildered and horribly depressed at the devastation in progress (‘crying to the depths of her being in enormous sadness’).

Given all the possible limitations due to the complexity assumed by this treatment, C’s dreams became a particularly valuable medium for the understanding of altered mental states that could not otherwise have been revealed and elaborated. The reception of these dreams helped in the construction of an analytic relationship that laid the foundations for preventing further psychotic implosions, the analyst now being more capable of seeing the prevalent expression of the psychosis through the dreams. It was after all always possible for C to lapse into psychosis, and the analysis of her dreams confirmed that the patient’s communication had to do with this problem, which it was important to confront.

Some dreams had already described the mechanisms of transformation of psychic reality, of the control exercised by the psychotic part and of the drugged mental condition, all of which were central to the creation of the psychotic state. Here are some examples:

‘I am at a Salvador Dalí exhibition. There is an empty anteroom with columns similar to the giraffe-like necks of Dalí’s women. I ask Dalí to go with me into a room because I am afraid. He accompanies me into another room and gives me a piece of a sculpture of the Virgin Mary to eat; there is a notice that says: “Now it is too late”. I ask him to take me home because I have not got my car keys. The car turns into an escalator, going up; it is like an elastic band that you have to pull with your hands. It is very steep; I laugh like a mad woman and fall back down.’

‘I have to go and see Di Pietro, a university lecturer, who sets me an exercise on surrealism. In the waiting room I am very worried because I know nothing about surrealism; I would rather have written about something else.’

‘I am in my father’s village. My father is at home, copying some of my notes with his left hand. He thinks they are shorthand symbols, but in fact they are only scribbles …’

In connection with these dreams, C told me of her attraction to Dalí’s paintings and surrealism, to the automatic writing that ‘liberates the unconscious’, and to mental logorrhoea. She herself said that in the past she had had periods of writing mania, in which she would scribble away for days on end. In some of these periods her mind would create continuous links, where she would think constantly of everything with great lucidity, in a state of uninterrupted intense mental activity accompanied by a great sense of exaltation. The scribbles in the dream were the exalted and surreal production of the psychosis, which transformed reality and presented itself as superior; the father who copied them suggested that the patient thought her father and the analyst should copy and admire them and take inspiration from her. The madness idealised in this way excited the father, whereas a part of the patient knew that the scribbles were only faeces.5

This was a particularly important dream because it led to the reconstruction, with the patient’s participation, of her entry into the psychotic state, which had been responsible for her first hospitalisation. In this situation the patient had felt particularly lucid and creative and had had the idea of being Virginia Woolf. The state of exaltation, initially perceived as a positive, uplifting mental openness, had subsequently been experienced as destructive. The patient had then felt herself to be Lucifer, the transgressive, negative bearer of light. This biphasic sequence is characteristic of the psychotic mental state. The patient is first excited and then plunged into persecutory, self-destructive guilt.

Later dreams described the dominion of the psychotic part over the rest of the personality, exercised through threats-‘an evil devil joined our group and told us that we must cut the veins out of dolls drawn on a piece of paper with scissors. I absolutely did not want to, but was afraid of rebelling’-or erotised allurements -‘I can see my sex organs from a distance and they look very small. Then they open and I can see the tubes and an alien, who seems to be winking. I feel somewhat surprised, a sort of “Aha”.’

The bizarre transformation of objects-in this case, in particular, of the parent figures-is also evident:

‘I am flying … Then I am on the telephone to my mother, but I do not tell her that my father is there because I am confused by the racket he is making. Instead I tell her that an ornament has got broken. It is a glass cock in the form of a human being, dressed like a character in a Manzoni novel. I am very fond of this object, in the way that one is fond of dolls. I put it in a glass crusher that vaporises it and what come out are substances of different colours that recombine in the form of a doll that looks like me and flies …’

There is also seduction with the promise of omnipotent mental states:

‘F is running with his head only and is much faster than I am. I say to myself: just look at that, it’s marvellous!’

Finally, consciousness appears, with the beginnings of repudiation of the psychotic part, represented as a destructive gang of delinquents:

‘Two boys know how to fly, and I do too, but not so well. They summon me into a room to show me that they have destroyed the carpet, which is like the skin of the house. I call my aunt, who is so upset that she gets drunk; she tells me I am mad to let these delinquents into the house-and meanwhile they have also wrecked the bodywork of the car.’

Together with the other factors favouring change, this type of analytic work on dreams contributed to the success of the analysis, because it enabled the analyst to make the patient receptive to the danger and to confront the silent, omnipotent allurements of the psychosis. The analyst, having understood the psychotic transformation described in the dream and communicated it to the patient, was able to recognise and internalise it at a time when she was still able to use her mind. This method also made it possible to reconstruct and understand the two previous psychotic episodes.

The analytic work was pursued in two complementary directions, involving, on the one hand, the reconstruction of the psychotic episode so as to make it understandable at a time when it was threatening to reappear and, on the other, confrontation of the attempt at omnipotent reparation whereby the patient sought to abolish the catastrophic sense of guilt that followed the psychotic episode. The therapeutic work concentrated on these points with a view to breaking the vicious circle between irreparable damage and manic reparation (Salvador Dalí) and to sweeping away the patient’s collusion with the psychotic part.


Owing to the nature of the psychotic process, the dreams of psychotic patients present very complex problems of reception to both members of the analytic couple. On the formal level, these dreams possess all the characteristics described by Racamier. They may be very short, like visual flashes, or extremely long and full of transformations, a symbolic interpretation of which is not feasible. The patients always have great difficulty in associating to them, sometimes displaying a sense of alienness, surprise or alarm at their terrifying and incomprehensible content.

The dreams reported here, on which we have concentrated, are connected with the psychosis in so far as they ‘describe’ it, so that they offer valuable material for the therapeutic process.

Quinodoz (1999) recently wrote about dreams that ‘turn over a page’-dreams that are full of anxiety and whose primitive content frightens the dreamer. In the authors’ view, notwithstanding their regressive aspect these dreams may be seen in dynamic terms as signs of progress. For despite their primitive content, they may constitute attempts at integration and reunification of parts of the patient. Although Quinodoz concentrated on the problem of dreams in neurosis, he considers that similar processes take place in the dreams of more seriously ill patients who attempt to put together the scattered parts of their personality (Quinodoz, 2000). In these cases, too, the anxiety appears to be connected with the attempt to represent and reintegrate parts of the self. Dreams featuring psychotic parts might be attempts by the patient to reunify split-off, scattered or not readily integrable experiences which, however, nevertheless manage to achieve representation. Although we agree with Quinodoz on the progressive, integrative meaning of these dreams, it is difficult to say how far the anxiety accompanying ‘psychotic dreams’ results from the difficulty of integration and how far it is due to perception of the threat from the psychotic part that terrifies the patient.

Grotstein (1983) states that dreams perform a fundamentally important function in allowing psychic life to be observed. For this to happen, however, an unseen observer must be preserved to annotate the plot and verify and validate its truths and messages. Psychosis, by altering the structure and coherence of the mind, represents the failure of this harmony. The dreamer who dreams a dream and the one who understands it may together be deemed an integrated unit whereby the existence of a stable sense of identity is rendered possible. The dreamer who understands the dream is a representation of the internalized maternal container, which perceives the narrative urgency and modifies the story so as to allow a solution. However, if this object is felt to be damaged, dreams are experienced as malevolent oracles, resulting in the psychotic experience. The realm of -K, or hallucinatory transformations, is established.

One of the difficulties presented by the understanding of dreams in the psychotic state is precisely the absence, in the patient, of the dreamer who understands the dream; as a result, a psychotic patient’s dream becomes indecipherable. The dream no longer represents a dialogue between the dreamer and someone who understands the dream, but sometimes instead only the product of a sensory experience that bears witness to the collapse of the mind.

Meltzer (1973) described the processes of mental sexualisation that occur in certain psychotic states, perversions and drug addictions, when patients succeed in creating an excited masturbatory mental state likely to distance them from reality. The entry into such a mental state is clearly experienced and anticipated in some of the dreams we have described. The point we wish to emphasise is that these patients (like the female patient we described) are not conscious of this process. The transformation is therefore so invasive and overwhelming that the patient is prevented from asking for help or drawing attention to the danger.

Since working with the psychotic part of the personality is the analyst’s purpose in the treatment of a psychotic or borderline patient, such ‘dreams’ are important because they afford access to the area of the delusion. Psychotic ‘dreams’, which remain totally lacking in meaning for the dreamer, must first be understood by the analyst as a delusional transformation occurring in the here and now.

Patient A, who overcame powerful resistances before eventually revealing that he had dreamed of a tennis court in a church, said nothing either about the content of the dream (there were no associations) or about its quality. It was only with hindsight, when the content had been acted out, that we were able to say that it had described a delusion several days before it invaded consciousness. What had been ‘dreamed’ had then been translated completely into reality. The patient described the delusional transformation as if it were a dream, thereby deceiving his analyst.

In our material, a ‘dream’ is the same as a hallucinatory construction; in the specific case, the ‘dream’ and its content refer not to the metaphorical discourse of the dream work but to the concreteness of the hallucinatory construction. They constitute the delusional reality in concrete form, and that is why they cannot be understood by the patient even if they are represented so vividly. For this reason, even if the delusional reality is clearly ‘represented’, the term ‘representation’ is inappropriate because it presupposes a capacity for symbolisation. Our point is that, whereas some authors treat delusions as if they were dream material, it is in fact necessary to treat this dream material as if it were a delusion. The importance of the clinical material reported here lies not only in the fact that dreams anticipate the content of the psychosis but also in that they precisely describe the functioning of the psychotic part and its exciting, confusing power to destroy the patient’s mental health.

Meltzer (1973) drew attention to some important aspects of this situation. For example, he wrote that the (destructive) psychotic part of the patient’s personality presents itself as promising protection from pain by arousing sensuality and working on the patient’s vanity, while concealing the sadistic and brutal facets.

As we know, in the analysis of borderline patients, failure to recognise the elements of the analytic communication connected with the action of the psychotic part increases the risk of delusional breakdown. For instance, underestimating the messages appearing in dreams or other material of an initial erotic transference may encourage the development of a violent delusional erotic transference. An ‘absent’ analyst colludes with the patient’s impotent, subjugated self.

We have presented two convincing clinical examples of this situation. The failure to take account of patient A’s ‘dream’ of the church and tennis court, leaving it uncomprehended and uninterpreted, may have caused the patient to perceive the analyst’s mental absence as potentiating the delusional situation. After all, in his next dream the patient described how he had been allowed to get into the marriage bed of an analytic couple and do whatever he wanted.

The failure to discern the erotisation in the dreams of patient C may have broken down some defence against the advance of the erotic delusion. In this last case the analyst had allowed the patient to go away without understanding the situation, abandoning her to the power of the sexualised internal father and the subsequent action of the destructive superego. In view of this situation, the analysis of C, who was fascinated by Salvador Dalí, concentrated for a long time on understanding the dreams in which this situation was clearly represented; the analyst sought to make the patient see how she was acting out the psychotic part, how it transformed reality, and how it seduced her mind, aligning itself with the level of functioning present at that particular time. The concentration on the communication of the ‘dream’ made it possible to show C how the psychosis (Salvador Dalí) was seducing her vanity, convincing her that she was special and idealizing the world of surreal transformations.

In other dreams, the seduction tended to follow the path of sensuality (the alien in the masturbatory vagina).

The leitmotif of the fascination of the ‘surreal’ is also present in many other patients, appearing for example in the themes and dreams of patient B, whose persecutory delusion we described earlier. The painter appearing in the dream reported is Miró.

Our experience suggests that there are only a small number of themes that describe the entry into psychosis. The most common are those describing the fascination of the surreal and the weird, sexualisation and perverse excitement or the drugged mental experience. These themes are described clearly and without disguise in the dreams.

It is never easy for an analyst to be capable of discerning the communicative meaning of a dream, so that its use becomes a highly complex operation. Reception by the analyst may help the patient to escape from the psychotic trap, whereas, if the analyst does not ‘understand the dream’, the patient feels even more impelled in the direction of psychotic colonisation. Indeed, one of the clinical difficulties in the cases presented here is precisely the docility with which the patient is trapped in the psychotic world. Steiner (1982) considers in this connection that the psychotic part gains control by persuading the rest of the personality to enter into a ‘perverse’ relationship. In this process, it is important to pay close attention to the patient’s propensity to act as a subordinated, consenting object and not to present any resistance to the psychotic process.

Another reason why the analyst is deprived of dreams and associations is that psychotic patients’ ‘dreams’ contain the delusional nucleus, which purports to be the ‘sole exciting and seductive reality’, a way of escaping or fleeing from the analysis and the analyst, who, in the patient’s eyes, is liable to destroy these ‘dreams’. Whether dreams betray the patient’s submission to the seductive power of the psychosis or bear witness to his/her resistance to the delusion, they will always be clinically important, because they may sometimes be the only communication that makes psychosis representable, describing its self-destructive and exciting nature. It is certainly helpful if this work can be done on dreams rather than on instances of psychotic acting out, considering that all consciousness of the catastrophic and destructive character of what is happening is lost in a psychotic crisis.

We should like to conclude by recalling the statement made at the beginning concerning the difficulty of using the associative method where the dreamer’s mental state is near-psychotic. Such difficulties make the therapy of psychotics particularly arduous.

In the classical model of the dynamic unconscious, unconscious ideas maintain their pathogenic power because they are kept out of consciousness: free associations serve to overcome the resistances and, by virtue of the patient’s becoming conscious of the unconscious content, to modify the psychopathological structure. The ‘dream work’ whereby the construction of the dream becomes possible presupposes that the personality structure is whole and that the unconscious function is clearly differentiated from consciousness-although at the same time possessing permeability in both directions (conscious-unconscious); it entails a capacity to use symbolic language, calls for a sense of personal individuality and integrity and, finally, demands the potential to perceive one’s history and infantile past. All this is remote from the psychotic state. The ‘dream work’ and associative capacity are, as we would now say, the prerogative of a neurotic personality’s self.

To understand the patient’s communication, an analyst makes use of conscious and unconscious thought operations, working in a mental condition resembling that of an intuitive scientist seeking to clarify a problem.6 Psychoanalytic interpretation is nothing but the birth of a thought that arises spontaneously and unconsciously, presenting itself as an intuition endowed with the character of obviousness and truth. Since the process is unconscious, it is only after the event that the route and the associative chain underlying the birth of the relevant idea becomes possible.

Psychotics, too, think themselves intuitive scientists, and they too seek new meanings, but their intuition (or rather illumination), which gives them the certainty of understanding reality, has a special character. The delusional intuition, in which the boundary between truth and falsification is lost, may be accompanied by an ineffable feeling of pleasure at the triumph of the psychotic part, which has subverted the organising function of thought and affects.

Psychosis alters the communicative function of the unconscious, the system able to formulate intuitive thoughts (De Masi, 2000). By blinding the unconscious, such modifications paralyse and distort its self-reflective and creative function and thereby make dreams inaccessible to the analytic method of free associations. In the psychotic state a particular quality of thought, nourished by visual combinations, combinations of assonances or linguistic identities into which the delusional meaning is constantly introduced, takes the place of the ordinary capacity to evaluate the subject’s own mental processes through the links (free associations) that lead to the perception of clarity and psychic truth. Psychotic constructions are fuelled by ‘revelation’, which acts by manipulating thought and transforming it into anti-thought, an already saturated proposition which claims that facts adapt to the ‘revelation’ and not the other way round.

Thought in the psychotic state is essentially sensory thought that ‘sees’ and ‘feels’. Psychotic ‘revelation’ is like a film that is projected into and captures the patient’s mind. Here the spectator ‘falls’ into the film shown, and becomes the protagonist of a concrete series of events that are not seen as a construction of his/her own.

While direct access to the ‘unconscious’ is possible with these patients, they are at the same time unable to make appropriate use of their own conscious perceptions because the conscious ego is in thrall to the psychotic part. The psychotic perceptions display unparalleled tenacity and adopt the guise of realities that do not readily lend themselves to transformation. Hence the importance of working on ‘dreams’, which allow a modicum of representation of the psychosis, albeit at a distance. For this reason, although the content of such a dream may be understood by the analyst, it remains totally meaningless to the dreamer, who has lost the capacity for insight.

The analyst must concentrate for long periods on the specific mode of functioning of a patient battling with psychosis, and must establish a complex and thoroughly well-structured setting tailored primarily to the nature of the ongoing process.

These considerations will help us to understand why the analysis of such patients may remain seemingly precarious for long periods but may subsequently begin to flow more easily. From this point of view, dreams in psychosis, which refer not to the search for a symbolic content but to the constitution of the psychotic state, may prove to be a means of communication essential for working usefully on the psychotic nucleus and the disposition to delusion. In dreams, the psychotic reality can be visualised and compared with the healthy part, which has not been overwhelmed by the invasion of the psychotic state and continues in part to ‘see’.


We have sought to show how it is possible, by working on certain dreams, to influence the action of the psychotic part, which, as Freud suspected in connection with the death instinct, ‘operates in silence’. Only the psychoanalytic method and the particular form of psychoanalytic listening can help to make the psychotic process manifest. Since, in a psychotic crisis, the capacity for symbolisation is totally lost and the analytic space collapses, this work can be done on dreams and not on the active psychosis.

In other words, if the main purpose of analytic therapy in the treatment of psychotic patients is to try to modify their internal balance so that they do not slide into psychosis, appropriate ways must be found of confronting the psychotic part before it unleashes all its devastating force.

If delusion is a specific, omnipotent ‘dream’ that transforms psychic reality, the psychotic ‘dream’, which also represents the pressure of the delusion that is colonising the ego (which is inclined to submit to it), may become one of the available means of highlighting what is occurring in silence, with a view to warding off the catastrophe and helping the patient to escape the threatening and seductive grip of the delusion. Such an approach to psychotic ‘dreams’ may, by making patients receptive to the danger and putting them in a better position to confront the silent and omnipotent means of seduction deployed by the psychosis, help to direct the analytic process into more favourable channels.

Translations of summary

Les auteurs pensent que la théorie freudienne des rêves n’est pas directement applicable aux patients psychotiques et borderline dont les états d’intégration mentale varient constamment. Du fait du manque d’association  qu’ont ces patients, l’approche psychanalytique ne peut être utilisée pour établir leurs significations. Après avoir examiné la littérature se rapportant à la qualité spécifique des rêves dans les états psychotiques, les auteurs montrent que de tels rêves n’ont rien à voir avec langage métaphorique du rêve, mais expriment plutôt  l’état concret de la construction de l’hallucination. De ce fait, le patient peut ne pas réussir à  comprendre le sens du rêve même si celui ci parait clair à un observateur. Cependant, la façon dont un analyste réceptionne un ” rêve psychotique ”    représente une source d’information essentielle et unique sur la manière dont se construit  le système de l’allucination  ,   permetant ainsi d’effectuer un travail analytique sur le noyau psychotique. Les auteurs pensent que de tels rêves permettent au patient et à l’analyste – étant encore lucides- de gagner de l’insight ,et donc créent une fondation stable pour sortir de la psychose. L’article comprends quelques cas cliniques dont un avec une patiente psychotique qui a la possibilité, en travaillant sur les rêves, de reconstruire un épisode psychotique et ce faisant de se protéger d’une chutte imminent dans la psychose.

Nach Meinung der Autoren lässt sich die Freudsche Theorie der Träume nicht direkt auf psychotische und Borderline Patienten mit ihren ständig wechselnden Zuständen seelischer Integration anwenden. Da bei den Träumen dieser Patienten die Assoziationen fehlen, kann die übliche psychoanalytische Zugangsweise zur Erfassung ihrer Bedeutung nicht angewandt werden. Nach einer Übersicht über die Literatur zur besonderen Qualität von Träumen im psychotischen Zustand, weisen die Autoren daraufhin, dass solche Träume nichts mit der metaphorischen Sprache der Traumarbeit zu tun haben, sondern die Konkretheit halluzinatorischer Gebilde ausdrücken. Aus diesem Grund kann die Bedeutung eines Traums von einem Patienten nicht verstanden werden, selbst wenn sie für einen Beobachter klar erscheint. Doch ist das Aufnehmen eines “psychotischen Traums” durch den Analytiker eine einzigartige und essentielle Quelle wertvoller Information über die Art und Weise, wie das Wahnsystem aufgebaut ist, die analytische Arbeit am psychotischen Kern ermöglicht. Nach Meinung der Autoren können solche Träume dem Analytiker und dem Patienten, solange er bei klarem Verstand ist, helfen, Einsicht zu gewinnen und auf diese Weise ein stabiles Fundament für das Auftauchen aus der Psychose zu schaffen. Der Artikel beinhaltet einige Fallgeschichten; in einer von ihnen wird eine psychotische Patientin durch Arbeit an Träumen in die Lage versetzt, eine psychotische Episode zu rekonstruieren und dadurch einen drohenden Rückfall in die Psychose abzuwehren.

Los autores consideran que la teoría freudiana de los sueños no se puede aplicar, directamente, a los pacientes psicótico y fronterizos. A causa  de la constante variación de sus estados de integración mental  y de  su dificultad para  asociar en relación con los sueños, el enfoque psicoanalítico corriente  no puede ser empleado para descubrir el significado de éstos. Después de revisar la bibliografía la cualidad específica de los sueños en el estado psicótico, los autores subrayan que dichos sueños no tienen nada que ver con el lenguaje metafórico de la elaboración onírica y que, más bien, expresan lo concreto de la construcción alucinatoria. Por esta razón, es difícil que el significado sea entendido por el paciente aunque resulte claro  para un observador. Sin embargo, la comprensión del “sueño psicótico” por parte del analista es una fuente única y esencial para una información valiosa sobre el modo en que se organiza el sistema alucinatorio y que permite el trabajo analítico con los núcleos psicóticos. En opinión de los autores, tales sueños pueden ayudar al analista y al paciente – mientras éste está lúcido – a lograr “insights” , lo que da lugar a bases sólidas para poder salir de la psicosis. El artículo incluye algunas historias clínicas, en una de las cuales  una paciente psicótica , a través de trabajar los sueños, puede reconstruir un episodio psicótico y, de este modo, protegerse de otra inminente recaída en la psicosis.


Arieti, S. (1963). The Psychotherapy of Schizophrenia in Theory and Practice. Psychiatric Research Report No. 17, American Psychiatric Association.

Atkins, N. (1969). In: Mack, 1969.

Bertrando, P. (1999). Vivere la schizofrenia. Turin: Bollati Boringhieri.

Bion, W. (1962). Learning from Experience. London: Heinemann.

– (1963). Elements of Psycho-Analysis. London: Heinemann.

–(1967). Second Thoughts. London: Heinemann.

–(1992). Cogitations. London: Karnac.

Blechner, M. J. (1983). Changes in the dreams of borderline patients. Contemp. Psychoanal., 19: 485-498.

–(2000). ‘Dreams that turn over a page …’ Letter to the editor. Int. J. Psychoanal., 81: 174.

De Masi, F. (1997). Intimidation at the helm: superego and hallucinations in the analytic treatment of a psychosis. Int. J. Psychoanal., 78: 561-575.

–(2000). The unconscious and psychosis: some considerations on the psychoanalytic theory of psychosis. Int. J. Psychoanal., 81: 1-20.

Freud, S. (1900). The Interpretation of Dreams. S.E. 4-5.

– (1911). Psychoanalytic notes on an autobiographical account of a case of paranoia. S.E. 12.

– (1920). Beyond the Pleasure Principle. S.E. 18.

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

– (1940). An Outline of Psycho-Analysis. S.E. 23.

Frosch, J. (1969). In: Mack, 1969.

Green, A. (1977). The borderline concept. In: Borderline Personality Disorders. Ed. P. Hartocollis. New York: Int. Univ. Press.

Grotstein, J. (1983). Do I Dare Disturb the Universe? London: Karnac.

Katan, M. (1969). In: Mack, 1969.

Mack, J. E. (1969). Dreams and psychosis. J. Amer. Psychoanal. Assn., 17: 206-221.

Meltzer, D. (1973). Sexual States of Mind. Perthshire: Clunie Press.

– (1983). Dream Life. Perthshire: Clunie Press.

Quinodoz, J.-M. (1999). ‘Dreams that turn over a page': integration dreams with paradoxical regressive content. Int. J. Psychoanal., 80: 225-238.

– (2000). Response (to M. J. Blechner). Int. J. Psychoanal., 81: 175.

Racamier, P.-C. (1976). Rêve et psychose: rêve ou psychose. Rev. franç. psychanal., 40: 173-193.

Resnik, S. (1996). In: Panel report: Psychic reality in borderline conditions. Int. J. Psychoanal., 77: 43-47.

Riolo, F. (1983). Sogno e teoria della conoscenza in psicoanalisi. Riv. Psicoanal., 29.

Rosenfeld, H. (1978). Notes on the psychopathology and psychoanalytic treatment of some borderline patients. Int. J. Psychoanal., 59: 215-221.

Segal, H. (1991). Dream, Phantasy and Art. London: Routledge.

Steiner, J. (1982). Perverse relationships between parts of the self. Int. J. Psychoanal., 63: 241-251.

Winnicott, D. (1971). Playing and Reality. London: Tavistock.

Translated by Philip Slotkin MA Cantab. MITI

2 [Translator’s note: These extracts are here retranslated from the Italian.]

3 This problem arises with particular intensity in patients whose own living childhood world of passion met with crushing responses from the adults around them.

4 The fact that psychotic mechanisms can interfere with the capacity to dream is noted by Winnicott (1971), who distinguishes the withdrawal into fantasy from the capacity to dream. Discussing a female patient of his who spent her days engaged in continuous fantasying, he stresses the pathological character of the withdrawal into fantasy, which absorbs energy but does not contribute either to dreaming or to real life. Whereas fantasying interferes with action and with life in the real world, it interferes much more with dreaming and with internal psychic reality.

5 It is interesting to note that the figure of Di Pietro (the famous Italian anti-corruption judge) in the dream represents an exciting and persecutory superego, of a kind that, as also pointed out by Rosenfeld (1978), confuses the patient by first exciting her and then making her feel guilty and persecuting her.

6 Poincaré, Hadamard and even Einstein all drew attention in their writings to the importance of unconscious work in scientific intuition.

Copyright © 2001 The International Journal  of Psychoanalysis