1

The royal road

In the New Introductory Lectures Freud says of his theory of dreams:

It occupies a special place in the history of psycho-analysis and marks the

turning point; it was with it that analysis took the step from being a
psychotherapeutic procedure to being a depth-psychology.
(Freud, 1933:7)

This is not surprising. Freud’s studies of neuroses revealed to him the significance and
psychic meaning of symptoms. It is the study of dreams—a universal phenomenon—
which opened up the understanding of the universal domain of dream thought and dream
language which goes far beyond the understanding of the actual night dream. Freud came
to see the analysis of dreams as the royal road to the unconscious.
Unlike many of his other theories, Freud altered but little his theory of dreams, first
fully formulated in The Interpretation of Dreams (1900), the book which, according to
Jones, he considered his most important work.
Freud regards dreams as guardians of sleep. As we know, sleep can be disturbed by
external stimuli such as a loud noise. To protect sleep, the sleeper can produce a dream in
which the noise is taken up by the dream and, as it were, explained away. These are rare
occasions. More regularly our sleep is disturbed by internal stimuli. Unfulfilled desires
and wishes, unresolved conflicts, give rise to inner tensions which could trouble our
sleep. In sleep our relation to reality is temporarily suspended. Repression is partly
relaxed and regression occurs so that archaic unconscious wishes strive for expression.
Motility and action are suspended and repressed desires seek expression ‘in a harmless
hallucinatory experience’ (Freud 1933:16). Common speech recognizes the wish-
fulfilling aspect of dreams by using the same term ‘dream’ both for the day-dream (a
wish-fulfilment fantasy) and the dream we have whilst asleep. But there is a fundamental
difference between the two. The day-dream expresses conscious wishes— organized,
rationalized, acceptable to our waking consciousness. In the night dream, on the contrary,
it is precisely those wishes which have been repressed and which trouble our psychic life
that seek fulfilment. Freud thought at that time that dreams are predominately (though
contrary to the common belief never exclusively) of a sexual nature. St Augustine, in his
confessions, complains bitterly that God should permit that he be troubled by sexual
dreams. He says that it would be so easy for God to arrange it differently. Had St
Augustine been familiar with psychoanalysis, he would know that the task is not so
simple. One cannot overestimate the importance of repressed sexuality in dreams, though
originally Freud may have underestimated the equal importance of repressed aggression.
Wishes which are powerful enough and dynamic enough and yet repressed enough to call
for an expression in the dream rather than in reality invariably have their roots in infantile
 The royal road 3

conflicts repressed in childhood but continually active in the unconscious. ‘Dreaming is a

piece of infantile mental life that has been superseded’ (Freud 1900). Only infantile
wishes have the power to mobilize forces which produce the dream:

These wishes in our unconscious, ever on the alert and so to say immortal,
remain one of the legendary Titans, weighed down since primeval ages by
the massive bulk of the mountains which were once hurled upon them by
the victorious gods and which are still shaken from time to time by the
convulsion of their limbs.
(Freud 1900:553)

The dream is usually linked with some event which happened in daytime. Freud called
this event a ‘day residue’. Such an event may be important enough to make it
understandable that it should influence the dream. But whether important or trivial, the
day residue is an event which in some way is in the patient’s mind connected with and
represents some deeper unconscious conflict. In some way the day residue which triggers
off the dream is similar to an event which could have triggered off the onset of a neurosis
or of a particular neurotic symptom. Non-fulfilment of deep-seated wishes gives rise to
inner tensions. Their fulfilment, however, would give rise to anxiety and guilt. It is not
for nothing that these wishes were repressed in the first place. Freud’s basic work on
dreams precedes his concept of the superego. He calls the repressing agency that forbids
the fulfilment of wishes unacceptable to consciousness, the censor, or the censorship, and
he describes the conflict as between the unconscious wishes striving for expression and
wish-fulfilment in the dream and the censorship which forbids such fulfilment. The ego
does not disappear in sleep. It must protect itself both from the tension arising out of
unfulfilled desires and the anxiety and guilt that would accompany their fulfilment. Freud
sees dreams as the result of a compromise between the repressed and the repressing
forces—a way of bypassing the dream censorship.
A dream is produced by what Freud calls dream-work. The dreamwork converts the
latent dream thoughts unacceptable to the ego even in the state of sleep into the
apparently innocuous manifest dream content. The dream-work is Freud’s first
description of a wider concept which is, I think, fundamental to the understanding of
psycho-analysis, that is, psychic work.
Psychic dream-work aims at fulfilling the unacceptable and conflicting wishes by
disguising them, and it evolves a particular mode of expression—the dream language.
This is constructed by such mechanisms as condensation, displacement, indirect
representation of various kinds, and symbolism. Those mechanisms Freud calls
sometimes ‘agents’, sometimes ‘Werkmeisters’ (foremen, or masters), again conveying
the psychic powers that create a dream.
Displacement can be of two kinds. One is the displacement of psychic values. The
manifest dream may put emphasis on a dramatic and apparently important situation, but it
is some insignificant detail that contains the most important latent dream thought. For
instance, a patient, through a concatenation of circumstances, had a glimpse into a room
in my house and he saw a print which he thought was of Venice. That night he had a long
dream in which he was walking with a girl in a place which reminded him of Venice. That
part of the dream led to a lot of associations, coming without much resistance, having to
 Dream, phantasy and art 4

do with his past flirtations with girls, with fantasies about me and my holiday, and a
fantasy of meeting me on holiday. But there was a detail in the dream to which he offered
no spontaneous associations and which, in the dream, had no apparent emotional
significance, compared with the richly evocative scene with the girl. There was
somewhere in the background of his walk in the dream a concrete structure at the seaside.
I asked him what this detail made him think of. Here the associations were much less
pleasant. He said that he once saw at the Venice Lido remnants of German military
installations. This in turn led him to associations about the German concentration camps
and the extermination of Jews. That took him back to the glimpse he had of the room in
which he saw the print and some books and he said that he had the thought that he ‘was
trapped in a Jewish household’. He connects Jewishness with intellectual and artistic
interests; and his feelings about Jewishness are ambivalent. There was a great deal in him
of unconscious anti-Semitism, which consciously is rather repugnant to him. It is the
insignificant detail in the dream which contained all his repressed hostility and cruel
unconscious wishes, his, unacceptable to him, anti-Semitism, stimulated by the thought
of my family life and the holiday he imagined me having. But in the dream it is displaced
and condensed in a little detail and is ignored in associations; and there is a displacement
of the importance to the more innocuous parts of the dream. He unconsciously wished
that my husband and I would perish in a German concentration camp.
Another kind of displacement is the displacement of feelings or phantasies belonging
to one situation on to another. A patient dreamed of an angry quarrel with a man towards
whom he had no antagonism, but in the background there was the figure of another man
loosely connected by a similarity of name with the first one. Towards that man he had
many hostile thoughts, the expression of which would give rise to guilt, as he was much
beholden to that man. An incomplete displacement of that kind is shown in the following
dream.
A man dreamed that he saw a little chicken being quartered, and he heard the
desperate crying of a baby or a small child. After a while he realized that the sound was
coming not from the chicken but from a small child who was nearby. In this dream, the
phantasied attack he had wished to make on his baby brother is displaced on to the
chicken, but the displacement is not quite successful. It is a little boy who cries and the
dreamer woke up in anguish.
This kind of displacement can also be seen as indirect representation: one man
represents the other; the chicken represents a brother.
Condensation is an invariable feature of dreams. However short the dream, the latent
thoughts that it contains range widely, and many thoughts and wishes, often contradictory
ones, are contained in the dream as a whole and in the various elements of the dream.
That is one of the reasons why it is difficult to report fully on the analysis of a dream, and
indeed a dream can never be analysed fully in one session. In the next session the patient
brings new associations and new dreams long before the analysis of the first one can be
exhausted—if indeed it can ever be.
An interesting example of condensation has been shown to me in a repetition dream
dreamed by Patient O, who suffered from a gastric ulcer. He has had this dream, close to
a nightmare, on and off ever since he can remember. As a very small child he remembers
waking up from his dream in a panic. In the dream he is completely tied to a chair in a
 The royal road 5

half-lying position. From all sides he is threatened by some elongated animals with
crocodile mouths.
In the course of his analysis the dream first occurred in the context of castration fears
of having his penis bitten off or chopped off as a punishment for masturbating. He is
being tied to a chair to immobilize his hands. It appeared again in the context of a
phantasy of myself being pregnant, and anxiety about attacking the inside of my body
and the babies therein. The unformed elongated shapes with crocodile mouths
represented the dangerous babies inside mother. The dream kept recurring in various
contexts. One day I was struck by his description of himself in the chair as being in a way
bandaged to it, and that he himself was the elongated shape. I asked if he had ever been
swaddled, and he told me that he was completely swaddled up to the age of three or four
months. He also told me that he apparently suffered severely from abdominal colic (as it
was diagnosed) at that time. It seemed to me then that the animals attacking him, the
huge, angry, hungry mouths, were a projection of his own bodily perception of himself
immobilized, hungry, and with a perception of his hungry mouth as enormous. Probably
swaddling intensified the process of a violent projective identification of his perception of
himself, as he was deprived of any kind of motor discharge by his musculature.
From the time of our work on that level, the dream stopped recurring. As the dream
was formed it seems to have condensed his experiences at many different levels. In this
condensation it also shows how the earliest primitive phantasies coloured, and found
expression in, later phantasies and anxieties. My understanding of this dream derives, of
course, not only from Freud’s concept of condensation and displacement, but also from
my own experience and from later theoretical developments. For instance, I used the
concept of projective identification to understand the way unconscious phantasies were
expressed in the dream, and I saw the condensation in this dream as an evolution from
very primitive oral and concretely psychosomatic phantasy to a later, more symbolic
level.
A more complex condensation is illustrated by a tiny fragment of a much longer
dream. In the fragment the patient saw the analyst accompanied by a little hairy boy or
man who looked rather ridiculous and jumped around the analyst in a very subservient
way. His associations were to another analysand of mine who has a nice crop of dark hair
and is not very tall. The patient had some reason to feel jealous of that man, who was
professionally ahead of him. He had met the man the previous day and on that occasion
felt rather contemptuous of him. This is the day residue. He remembered that he thought
that my husband looked rather like a gorilla and his own father had a very hairy chest. He
often meets in my street a long-haired adolescent whom he describes as ‘rather a hood’
and whom he thinks is my son. He met that boy also in the vicinity of the Tavistock
Clinic and wondered if he was treated there. He thought I must be a very bad mother,
neglecting my children so that they needed treatment. The subservient attitude he linked
with himself, always coming punctually to the sessions and feeling abjectly dependent,
which is a feeling he hates.
But the litde man in the dream did not look quite human. The patient had recently seen
a film about a werewolf. The figure in the dream could well be a werewolf.
So up to that point one could say that the figure in the dream represents a rival—my
other analysand, my husband, and my son. They are all condensed into one figure. Past
and present are equally condensed. My husband and son and his father and brother are all
 Dream, phantasy and art 6

represented by one figure. But there are many other ideas represented in that fragment of
a dream. The father and my husband, standing for him, are derided by being made small
and ridiculous. There is also the fear of the rival, thus attacked—werewolves and gorillas
are dangerous—but the fear is counteracted by his being made small and ridiculous.
There is also the idea of my cruelty and badness, as accounting for the bad psychological
state, not only of my supposed son, but of himself as my son. They both suffer from my
neglect and I am blamed for their neuroses. Towards the end of the session there was a
further association which revealed an even more painful repressed thought.
Whilst speaking of werewolves, he said, ‘According to the legend, you become a
werewolf if you are bitten by one. I suppose the wolf is at the door now.’ He was
referring to an impending holiday. So there is another layer to the dream. When the
analyst, the feeding mother, goes away, he dreads hunger, felt like a biting wolf- the wolf
is at the door. This bite of the wolf makes him into a werewolf. It mobilizes his oral greed
and aggression, to which is added an extreme jealousy of those he thinks will stay with
me—husband and son. In the dream he deals with his werewolf-like feelings by
projecting them into his rivals, thus achieving the double aim of getting rid of pain and
guilty feelings in himself and of attacking his rivals and making them bad. The resulting
persecutory fear of his rivals now turned into werewolves is dealt with in a manic fashion
by making the werewolf small and ridiculous. (He also projected into them his own
smallness and the hated feeling of dependence, seen as subservient.) So one can see how
one fragment of a whole long dream can condense and express a most complex psychical
process.
What is the essence of what Freud so beautifully, I think, calls ‘the dream thought’? I
think Freud originally had in mind simply the repressed wish, disguised in the dream. But
wishes are contradictory and complex, and I think the dream thought is more than a
simple wish. It is itself a complex organization of wishes and defences. The dream
thought of my patient’s dream could be verbalized thus: ‘When she goes away I am bitten
by hunger. She is a bad biting figure inside my tummy. I feel full of greedy and biting
feelings. This is intolerable. I shall put it into the rival who is with her. But that makes me
frightened of the rival. I shall try to diminish and ridicule him’, and so on. Condensation
itself is not accidental. The dream thought, as I see it, is an expression of unconscious
phantasy and our dream world is always with us.
In my understanding of condensation I may possibly differ from Freud. I do not think
he sees condensation necessarily as a connected ‘story’. He sees it more as various
strands arising possibly from different impulses and trends of thought, converging
together and being expressed in one condensed element.
Apart from condensation and displacement, there are other methods of transforming
the latent dream thoughts; for instance, by indirect representation. There are many ways
of achieving it: by similarity, the possession of a common attribute, using a part for a
whole, by opposite, by verbal connection, and many, many others. Those representations,
when understood, are sometimes very amusing—wit and humour, as Freud has shown,
having similar features to the dream-work. As part of a long dream a patient dreamed of a
column of soldiers marching eight abreast. Rather perplexed by that part of the dream, I
asked her what she thought. She answered immediately: ‘Ate a breast, of course. What
else could it mean?’
 The royal road 7

A more complex example of representation by the opposite, or reversal, is shown in

the dream of Patient B, with a manic character structure. The patient’s mother died in his
early adolescence and he avoided the mourning by schizoid and manic defences which
deeply affected his character structure. The day preceding the dream, we were talking
about a quite serious car accident he had had over the break and his preference for big
and powerful cars over small cars, which his wife prefers, and in which he feels too
vulnerable. He would really prefer to travel in armoured cars or tanks, he feels so
vulnerable.
The next day I accidentally collected him from the waiting room a couple of minutes
early. He said he was very pleased and felt warmly welcomed. After a time he reluctantly
admitted that his first thought was very anxious: he thought I might have left the session
with the previous patient early because I was ill, and it immediately reminded him of an
operation I had had several years previously, and of his mother’s unsuccessful breast-
cancer operation.
He then told me a dream. He was in a kind of lab. There were some chemical benches.
Near him was a younger man, Bob. He slipped a little box into Bob’s locker. Then a
beautiful young woman brings the prize of £500, probably a winning in a raffle—‘when
your number comes up you win’. She approaches him and Bob, and up to the last minute
they do not know which of them is the winner and they both feel teased. Then she gives
the prize to Bob. He does not feel jealous; he feels generous, very much aware of how
rich he is and how poor Bob is. Bob is not only poor; he is also unworldly and naïve. He
would like, with the prize money, to buy a bottle of whisky for his wife, and he turns to the
patient to ask him if he can do that, and how one proceeds to do it. The patient helpfully
instructs him.
The patient’s first batch of associations started with telling me that Bob in the dream
reminded him both of his brother and his eldest son (his attitude to Bob reflects very
much his attitude to his younger brother) both as a child and as a young man, since his
brother was idealistic and became a poor parson, while the patient made mints of money.
The dream reminded him of all the ways in which he looked after his brother as a small
child, particularly after his mother’s death, but also later in life, when the patient
administered the family estate. The willingness with which he gave the prize made him
think of his elder son giving a family party now and he, the patient, in such a situation
feeling that he is giving way as paterfamilias.
The sum of £500 he did not link with anything. I made the comment that everybody in
the dream was younger than himself, but he reminded me that a young woman could well
stand for his mother since his mother died young. And this led him to another batch of
associations. He remembered that his daughter, about whom he is always troubled, is now
approaching the age at which his mother died, and it is also the time of the year close to
the time his mother died. This immediately drew my attention to the importance of
reversal in the dream. What he was concerned with was not his lucky number but his
unlucky number. His mother’s number was up prematurely, it was unlucky for him, and
he has been recently preoccupied with his vulnerability and his fear of death, stimulated
by the car accident. Also his son’s growing up could be felt by him as his number coming
up.
Once alerted to the importance of the reversal, we could see that every element in the
dream is reversed. The woman who gives the prize is older than he (his mother), not
 Dream, phantasy and art 8

younger; he feels not rich but poverty-stricken, and this refers to two situations: first that
his brother, of whom he is acutely jealous, was born and got the prize from mother (the
little box) and later, when his mother died, he became an object of special care and
attention from his father. In the dream the patient also feels generous and benevolent to
his brother, which he would wish to be, but in psychic reality he resented bitterly both the
early care his brother got from mother and the extra attention he got from father and the
family after mother’s death, whilst my patient’s needs were completely ignored and he
was expected to look after his brother. He also at the time blamed his brother for her
death, since some people attributed her breast cancer to his little brother having bitten the
breast when a baby.
After some bringing together of the underlying experience of the dream, represented in
such a reversed fashion, he remembered that he did in fact inherit a little box from his
mother but he did not give it to his brother. We then got an association to the £500. He
had a bill for £500 for the urgent repairs to his car that had been smashed in the accident.
He also remembered that for more than a week he forgot to pay my bill. So the dream has
to do with urgent anxieties and needs of repair, and in contrast to the apparent generosity
of the dream, a remaining reluctance to pay my bill, or the one for the repair of the car.
So every element in the dream is reversed: the lucky number is the bad number; the
benevolence and generosity replace rage and meanness. The jolly raffle is in fact a re-
living of his mother’s death and his fear for himself. Even the bottle of whisky in the
dream is a kind of reversal. I made a comment at some point that Bob wanting to get a
bottle of whisky for his wife, after getting the prize from the woman, could represent
Bob’s wanting to use the goodness he got from the breast in giving good sexuality to his
wife. However, the patient immediately corrected me, since his association to cheap
whisky was to a drunkard who committed incest with his daughter and committed
suicide—very bad sexuality.
Right at the end of the session, he suddenly realized that the chemical lab with
benches in his dream reminded him that, as his mother was dying of cancer, he used to
spend hours in a little chemistry laboratory which he rigged up for himself cut off from
the rest of the family and from his own feelings.
In this dream I think we can see how complex dream language is. One could say that
in the dream everything is represented by its opposite, by the reverse, but at the same
time that way of representing it changes a deeply traumatic situation into a wish-fulfilling
one. One could look at the reversals of the dream as a disguise: an effect of dream
language is to disguise a distressing experience, but at the same time in the dream
language is hidden an unconscious wish to change the painful reality, internal and
external, into a glorious raffle.
Each dreamer has a favourite style in his dream language and the style itself often
reveals their personality. ‘Le style c’est l’homme’, said Buffon. It applies to dreams as
well as to art. The very style of the dream, like the style of a personality, reflects the
broad combination of object relationships, anxieties, and defences that moulds one’s
personality.
All the various methods of representing an idea in the dream constitute dream-work.
Freud excludes symbols from dream-work. He considered symbols as universal and
deriving from the ancient past. He says:
 The royal road 9

Things that are symbolically connected today were probably united in

prehistoric times by conceptual and linguistic identity. They are, one
could say, given, not achieved by psychic work implied in other methods
of indirect representation.
(Freud 1900:352)

This view of symbolism has been questioned implicitly by Melanie Klein and explicitly
by myself, and I shall return to it later.
The process of analysing a dream is doing the dream-work in reverse. The associations
to the dream expand again what had been condensed, rectify the displacement, decipher
the indirect representation. But the associations to the dream are not, as some therapists
think, in themselves the latent content. They are only a path leading towards latent
content, because repression continues to operate and to manifest itself as resistance.
Indeed, it is from the clinical experience of resistance that Freud deduced the mechanism
of repression (Heimann 1950)—a theoretical concept. The analysis of a dream proceeds
against resistance. Trends of association break off or acquire a defensive character, or
else the patient resists seeing the significance which may be apparent to the analyst. The
interpretation of the analyst has to demonstrate the resistance and indicate the latent
content. Where the patient’s own work falters, the analyst’s interpretation provides the
missing link. The psychic work of deciphering the dream-work is essential in the analysis
of dreams. This is done jointly by the patient and the analyst. It must be remembered that
Freud did not say that dreams are the royal road to the unconscious. He said that the
understanding of dreams is the royal road to the unconscious, and this understanding has
to be reached by psychic work.
There is a further factor which conceals dream thought and which operates after
waking, and that is the distortion in the actual remembering of the dream whilst awake
which Freud calls secondary elaboration. As we recollect, so we distort the dream.
Sometimes in the session this more conscious distortion can be corrected and a more
genuine remembrance of the dream emerges. This secondary elaboration, according to
Freud, is a continuation of the repression of the latent dream thoughts. But he also
describes how intolerable to our waking mind are illogicality, chaos, and disorder. He
quotes Havelock Ellis:

Sleeping consciousness we may even imagine as saying to itself in effect:

‘Here comes our master, Waking Consciousness, who attaches such
mighty importance to reason and logic and so forth. Quick! gather things
up, put them in order—any order will do—before he enters to take
possession/
(Freud 1900:501)

Whether this need to rationalize and make sense, make a story, is not the same as
resistance against latent unconscious dream thoughts I am not certain. It is also
questionable what ‘remembering’ a dream is. After analysis of the secondary elaboration,
one may recover a dream nearer to what was remembered immediately on waking. But in
the course of the session new elements of the dream may appear. What is remembered
may be altered as the dream reveals new aspects and deeper levels. The remembered
 Dream, phantasy and art 10

dream has its roots, in my view, in an unconscious phantasy the full depth and extent of
which can never be remembered.
Freud considered three kinds of dream. The first is the undisguised wish-fulfilment
dream, characteristic of children. He reports a dream of a little girl in which she gorges
on strawberries and of a little boy deprived of a meat dish dreaming of a ‘roast that got
itself eaten’. However, since we have learned to psychoanalyse children, I think we are
much more doubtful about the innocence of such dreams. The second kind are dreams
involving dream-work and disguised fulfilment of unconscious wishes, and it is to this
kind of dream that the bulk of Freud’s work is directed. The third kind are those dreams
which seem to run counter to the wish-fulfilment theory: namely, anxiety dreams and
punishment dreams. His comments before 1920 on those dreams are, first in relation to
anxiety dreams, that the dreamer makes an attempt at fulfilling his wishes in a dream
language but that this attempt is not necessarily successful. When an undisguised ego-
dystonic wish breaks through, the ego will respond by anxiety. As to the punishment
dreams, he reminds us that dreams are the outcome of various compromises between the
censor and the instinctual desires, and in the punishment dreams it is the censor that has
the upper hand.
Unlike his other theories, Freud never much altered his theory of dreams to bring it in
line with the vast development of his theory of psychic life, particularly after 1920. Since
Beyond the Pleasure Principle (1920) he saw the basic psychic conflict as that between
the life and the death instinct. This was his final instinct theory. He then evolved the
structural theory of mind—described in terms of the ego, the superego, and the id (Freud
1923). He had revised his theory of anxiety and repression (Freud 1926). Starting with
the idea that anxiety was due to repression, he had come to realize that, on the contrary, it
is anxiety which causes repression. This discovery was linked with his view that it is the
ego, not the superego (the old ‘censor’), that is responsible for repression as well as for a
variety of other defence mechanisms. In the New Introductory Lectures (1933), in the
chapter on ‘The Revision of the Theory of Dreams’ he did bring the theory partly up to
date. For instance, he did bring in the superego in the place of the censor, and speaks of
the dream having to reconcile the claims of the id and the superego. Since this was
written after The Ego and the Id it would carry the implication that the superego contains
the death instinct. But Freud does not explicitly bring the concept of the life and death
instincts into the revision of his theory of dreams. He gives particular attention among
anxiety dreams to the repetition in dreams of traumatic events which have led to a
traumatic neurosis. And, like other anxiety dreams in which the wish-fulfilment role of
dreams had failed, he explains the recurrent traumatic anxiety dream as being like other
anxiety dreams—an example of the dream-work having failed in its wish-fulfilling task.
However, by that time, in other publications (Beyond the Pleasure Principle, for
instance) he speaks of those dreams as one of the important phenomena which made him
develop his concept of the death instinct. But in his revision he does not seem to expand
the idea that the work of the dream is not only to reconcile the forbidden wish and the
superego or the ego, but also to find a compromise or resolution for contradictory
unconscious wishes, and the basic conflict between the life and death instincts.
In the dream of Patient B, I emphasize his wish-fulfilment in converting the traumatic
death of his mother into a happy, glorious event. However, that was not all there was to
the dream. Why did he have that particular dream at that time? The context was that
 The royal road 11

certain events had stirred in his transference a particular jealousy of a brother figure and
brought about death wishes towards me and that representative of his brother. His dream
is a solution to his conflict between those death wishes and his reparative wishes in a
wishful phantasy of bringing his mother back to life, helping his brother and giving him a
prize, and so on, but without losing his sense of his great superiority to the latter.
Freud also evolved the concept of working through, but he did not explicitly apply it
to the dream work as one of the ways of working through a conflict. He speaks of the
dream as a harmless hallucination, a’harmless dream-psychosis’ (Freud 1933:16). He also
speaks of the dream as similar to psychosis but happening entirely in sleep. This seems to
me in some way questionable, since the kind of psychic work elaborating a conflict, akin,
I think, to a working through, which happens in the dream, is precisely what is lacking in
psychosis. Freud says that anxiety dreams happen when the dream-work fails. Can we
today say more about this failure of the ego to perform the dreamwork?
One is always struck by how formidable is the task of the ego in creating a dream. It
has to carry out repression adequately, but not excessively. It has to be capable of the
psychic work involved in dreamwork, and I shall suggest that this includes the formation
of symbols. What happens when the ego is too damaged to fulfil those functions? I think
some of the later work done on the development of the ego and its pathology has thrown
more light on these problems. I shall therefore return to dreams after discussing the
concepts of unconscious phantasy and symbolism, and the differences between psychotic
and nonpsychotic functioning of the ego.
 2
Phantasy

Freud’s discovery of unconscious thoughts underlying hysterical symptoms can be seen

as equivalent to the discovery of unconscious phantasy. He showed how conflicts,
instincts, and defences are expressed and contained in an unconscious phantasy, such as,
for instance the phantasy of fellatio in Dora’s globus hystericus. Originally he thought of
phantasies mainly as defences against memory: ‘Phantasies are psychical façades
constructed in order to bar the way to these memories’ (Freud 1950 [1897]: Letter to
Fliess). Having abandoned his seduction theory in favour of the view that seduction
scenes which were apparently remembered or recovered from repression and believed to
be fact were most frequently a child’s wish-fulfilment phantasy, he gave phantasy a
major place: ‘Hysterical symptoms are not attached to actual memories but to phantasies
erected on the basis of memories’ (Freud 1900:491).
In a letter to Fliess (21 September 1897), Freud says that his discovery made him
realize that phantasies are as real and important as any external reality. After that he
sometimes spoke of phantasy as the psychic reality (Freud 1914c: 18; 1925a: 34).
Unlike his theory of dreams, Freud never worked out in full a theory of unconscious
phantasy. He never devoted a book, or even a paper, wholly to that subject, despite the
importance of the concept in his work.
One could say that generally for Freud phantasy is pretty close to day-dreaming. It is a
wish-fulfilling idea which comes into play when external reality is frustrating. Basically,
a phantasy consists of an unconscious wish worked on by the capacity for logical thought
so as to give rise to a disguised expression and imaginary fulfilment of the instinctual
wish. Phantasies remain subordinate to the pleasure principle, but they are formed by the
‘secondary process’; that is, by the normal rational logic that is characteristic of the
Systems Preconscious and Conscious. Hence Freud sees phantasy as a rather late
phenomenon, appearing only when the reality principle and the capacity for logical
thinking have been firmly established. In the case of conscious fantasies, day-dreams, the
fantasy is known not to be true. When the wish-fulfilling fantasy is unacceptable to
consciousness it is repressed and becomes unconscious phantasy. In his clearest statement
about this, in the ‘Formulations on the two principles of mental functioning’ he says:

With the introduction of the reality principle one mode of thought-activity

was split off; it was kept free from reality-testing and remained
subordinated to the pleasure principle alone. This activity is phantasying,
which begins already in the children’s play, and later, continued as day-
dreaming, abandons dependence on real objects.
(Freud 1911:222)
 Phantasy 13

If fantasies are found to be unacceptable to consciousness so that they are repressed into
the System Unconscious, they are subject to the ‘primary process’, supremacy of the
pleasure principle, no sense of time or temporal causation, and all the other features that
Freud considered to be characteristic of the System Unconscious. Once they have been
repressed into the System Unconscious, phantasies are no longer known to be untrue so
that they become indistinguishable from memories. In the System Unconscious
phantasies ‘proliferate in the dark’, as he put it.
Freud, however, seems uneasy about the exact relation between instinct and phantasy.
On the one hand, phantasies were subordinate to the pleasure-pain principle, which
implies they derive from instinct. On the other hand, he frequently states that at some
point instincts attach themselves to phantasy. ‘Every desire takes before long the form of
picturing its own fulfilment’ (Freud 1916–17:372).
Freud’s uncertainty about the exact relation between instinct and phantasy, and his
prevalent view that they are a kind of repressed day-dream, may well be at the root of his
hesitation about the role of phantasies in dreams.
In Chapter 6 of Interpretation of Dreams he writes:

[I] cannot completely escape a consideration of phantasies in this

connection since they often make their way complete into dreams and
since still more often clear glimpses of them can be seen behind the
dream.
(Freud 1900:493)

He says that phantasies that make their way into the dream are thereafter treated in
general like any other portion of the latent material, though the phantasy often remains
recognizable as an entity in the dream. In describing those phantasies, he assumes that
they must have undergone repression to appear in the dream. (At this point he seems to
mean that they were conscious fantasies later repressed.) But he also speaks of primary
phantasies which have never been conscious (which he directly connects with instincts).

They strike me as being, as it were, more fluent, more connected, and at

the same time more fleeting than other parts of the same dream. These, I
know, are unconscious phantasies which find their way into the fabric of
the dream, but I have never succeeded in pinning down a phantasy of this
kind.
(Freud 1900:493)
He asks:
Whence comes the need for these phantasies and the material for them?
There can be no doubt that their sources lie in the instincts; but it has still
to be explained why the same phantasies with the same content are created
on every occasion.
(Freud 1916–17:370)

His answer is that they repeat ‘what were once real occurrences in the primaeval times of
the human family’ (Freud 1916–17:371).
 Dream, phantasy and art 14

It is in keeping with Freud’s notion of phantasy being a relatively late phenomenon in

mental life (‘It begins with children’s play’) that he sees phantasies as highly organized
and referring mainly to whole objects. More primitive phantasies do not enter much into
his description. More primitive functioning on a pre-verbal, even pre-visual,
psychosomatic level is not included in his concept of phantasy. Hence he does not
connect early hallucinatory wish-fulfilment with phantasy, but only with memory.
Mostly Freud refers to unconscious phantasy as connected with pathology, but he is
also very aware that it is only a ‘short step’ between phantasy resulting in a symptom or
in artistic creativity; and in all his papers on works of art he views unconscious phantasy
also as a source of sublimation. But seeing it as a kind of repressed day-dream, he cannot
quite account for unconscious phantasy as a source of creative art and gets into
difficulties in some of his theorizing about art. I shall return to this subject later.
It is as though Freud had opened a door to a fascinating, rich, mysterious world, but
did not quite take the full measure of his own discovery and the connections between that
and his other major discoveries, such as the dream-work and the dream language. It is the
psychoanalysis of children which revealed the ubiquity and the dynamic power of
unconscious phantasy, and Klein gave this concept its full weight. From the beginning of
her work with children, she was struck by the extent to which the child’s life is
dominated by unconscious phantasy. In one of her earliest papers, ‘The role of the school
in libidinal development’ (1923), she describes how all the child’s activities, not only in
play but also in work, contain an elaboration of unconscious sexual phantasies. At that
time she was still mainly concerned with phantasies leading to neurotic symptoms like
inhibitions in learning, such as her description of how a little girl could not learn
grammar because parsing represented for her eating rabbit and that in turn was connected
with unconscious cannibalistic phantasies; or the little boy who could not do division
sums because it represented cutting his mother into bits; and so on. But in addition, one
gets the impression of a rich unconscious phantasy life underlying the child’s relation to
school and all his activity: not just a pathological formation interfering with activities. In
the consulting-room she observed the child’s expression of his unconscious phantasies
and realized the degree to which they could distort perception and dominate the child’s
life. And the younger the child, the more dominant was its unconscious phantasy life.
Most of Freud’s statements give the impression that he thought of unconscious
phantasies as if they were like islands in the sea of mental life. Reading Klein’s work
with children, one gets a glimpse of an internal phantasy world like a vast continent under
the sea, the islands being its conscious, external, observable manifestations.
When it became apparent that her views on phantasy extended Freud’s concept and in
some way differed from his, Susan Isaacs addressed herself to this new use of the concept
in her paper ‘The nature and function of phantasy’ (1948). As Freud had said, phantasies,
like dreams, are wish fulfilments, but while he considered that they form relatively late in
development, in Klein’s and Isaacs’ view they are active from the start. In the omnipotent
mind of the infant and child desires become wish-fulfilling phantasies, and such
phantasies are viewed by Klein and Isaacs as a direct expression of instincts and
impulses. Unlike Freud, who considers, in some of his writings, that at some point
instincts and phantasies get linked, Klein boldly assumes that instincts from the
beginning of life give rise to phantasy. It is not at some point that ‘every desire takes
 Phantasy 15

before long the form of picturing its own fulfilment’ (Freud 1916–17:372), but that such
picturing is inherent in the process at all times.
Freud wavered between two definitions of instinct. In some papers he called it ‘the
psychical representative of the stimuli originating within the organism and reaching the
mind’ and ‘the concept on the frontier between the somatic and the mental…the
psychical representative of organic forces’ (Strachey, in Freud 1915a:112). In his later
papers, however, he more often speaks of instinct itself as having a psychical
representative. In The unconscious’ he says:

An instinct can never become an object of consciousness—only the idea

that represents the instinct can. Even in the unconscious, moreover, an
instinct cannot be represented otherwise than by an idea [which he also
calls ‘psychic representative’].
(Freud 1915b:177)

An instinct, according to Freud, has a source, an aim, and an object. The ‘idea’, which is
the psychical representative of the instinct, in Klein’s view includes a phantasy of a
drive-fulfilling object.
The hallucinatory wish-fulfilment postulated by Freud would be seen by Klein and
Isaacs as part of a primitive phantasy. Freud wondered, ‘but it has still to be explained
why the same phantasies with the same content are created on every occasion’ (Freud
1916–17:370). Taking the Klein and Isaacs view, the reason those phantasies are
common to all humans is not that they reproduce some real events in the prehistoric past
but that we share a common instinctual endowment and ways of dealing with it. All
phantasy activity belongs to the ego, and some of the differences in views on phantasy
mainly hinge on different views of the early ego. In Klein’s view, from the beginning of
life there is sufficient ego to experience anxiety, to form some object relationships in
reality and phantasy, and to use primitive defences. Unlike Freud, she does not hold the
view that phantasies can only be formed when the infant or child has developed a
capacity for logical thought.
The view that phantasy is operative from the beginning, at the most primitive stages of
development, implies that this phantasy is to begin with physical: the hallucinated breast
is not to begin with a visual experience, but a bodily one. Early experiences, such as
hunger or satisfaction, are experienced and interpreted by the infant in terms of object-
relationship phantasies. Susan Isaacs assumes that behind every phantasy of introjection
there is an earlier one of concrete incorporation. Satisfaction is experienced as containing
a need-fulfilling object; hunger as a persecution. Our language reflects this. We speak of
being ‘gnawed by hunger’; or ‘the wolf being at the door’. My favourite is the French:
hunger being described as eating an enraged cow (‘manger de la vache enragée’). Such
primitive psychosomatic phantasies evolve with growth and reality-testing, but they
remain at the core of our personality and can still play a dynamic part in later
development. Physical experiences are interpreted as phantasy object relationships,
giving them emotional meaning. A baby in pain may feel itself as being hated. But also,
conversely, the phantasies are so close to the somatic that they affect physical
functioning. It is well known that an emotionally upset baby often develops digestive and
 Dream, phantasy and art 16

other physical symptoms. This may persist into adulthood, as in my Patient O [Chapter
1], whose unconscious phantasies resulted in a gastric ulcer.
After 1920, when Freud introduced the duality of the life and death instincts, he did
not revise his views of the early hallucinatory wishfulfilment; nor his views on primary
narcissism. It is Klein who used the concept of the death instinct most fully in describing
the functioning of the early ego. Thus in her view the early phantasy wishfulfilments
would relate not only to hallucinated or phantasied satisfaction of the libidinal wishes by
an ideal object but would also relate to all those wishes springing from the death instinct.
What Freud described as the deflection of the death instinct by the organism is seen by
her as a projection by the ego of the death instinct into an object, giving rise to phantasies
of an object which is destroyed and destructive (Segal 1964).
One can see here that from the start, in her view, phantasies have both a wish-fulfilling
and a defensive aspect, since splitting and projection, which are mechanisms of defence,
are also expressed through phantasies, just as are the impulses.
Phantasies of course are linked with defences. The very fact of phantasying is a
defence against painful realities. Freud’s earliest view was that phantasies were defences
against memory, but it soon became apparent to him that they could be used as defences
against any painful reality. However, though he recognizes that phantasies have a
defensive function, he does not connect defence mechanisms with phantasies. Susan
Isaacs firmly links the concept of phantasy with that of defences. According to her, the
concept of defences is an abstract and generalized description of what the individual does
actively in the particular content of his phantasies. Underlying the defences that we
observe and describe in an abstract way is a detailed phantasy of their implementation.
I shall illustrate this with material from a little girl, aged two and a half, who suffered
from severe sleeping and digestive difficulties. She presented what could be described as
a splitting of the object into a good and bad figure in the following way.
She had missed her last session before the holidays because she had become ill with
diarrhoea. In her first session after the holidays she became very much concerned with
and frightened by a shapeless shadow on the wall, which she named ‘the lady on the
wall’. She started the session saying that she did not bring a cardigan because though it
was cold outside it was so nice and warm in the room. And she smiled warmly at the
analyst. But when she turned to the wall and saw the shadow she started screaming The
lady on the wall!—it’s nasty, it’s bad! I don’t like her! It bites! Put her into the dustbin!’
She felt confronted by two objects in the room: the flesh-and-blood analyst felt as giving
warmth and protection; and the lady on the wall full of threat. The analyst interpreted that
the ‘lady on the wall’ was the bad analyst who was absent in the break; and the child
relaxed and started looking at the wall in a questioning way. The analyst reminded her
that the slight discoloration on the wall was due to the little girl having herself splashed it
sometime before the holiday and reminded her of her diarrhoea, which deprived her of
the session; and she linked it with her anger at the analyst and her wish to splash her with
urine and fill her with faeces. This she believed turned the analyst in her mind into a bad
lady like the lady on the wall. The child then relaxed more and started playing freely. She
poured some water into a cup, called it ‘milk for the baby’ and produced a complicated
game in which the baby was first refusing to take the milk, saying ‘it’s dirty’, then
shaking it and drinking it with great satisfaction.
 Phantasy 17

This child, who had great feeding and sleeping difficulties as well as digestive
troubles, was, I think, expressing positive and negative impulses as well as defences such
as splitting and projection. In her anger at the break she fantasied that she put bad parts of
herself (biting urine, faeces) into the analyst’s breast, which then turned it into a very bad
object. But she also loved and needed, and wished to keep in her mind, the good analyst.
So she split off this badness from her image of the analyst and projected it on to the
biting, dirty ‘lady on the wall’. What in abstract terms was splitting, idealization, and
projection was in the child’s experience a vivid phantasy resulting in a near-hallucination.
The following material from an adult patient shows similarly the unconscious
phantasies underlying the defence mechanism of splitting.
Here are two dreams described by a patient during the hour preceding my holiday.1 In
the first dream, the patient was in a dark room which contained two human figures
standing close to one another as well as other less well-defined people. The two figures
were exactly alike, except that one of them looked drab and dark, while the other was
illuminated. The patient was sure that she alone could see the illuminated figure—it was
invisible to other people in the dream.
The patient made extensive use of the mechanisms of splitting, denial, and
idealization. She had had the opportunity that same week to see me in a room full of
people, a situation unusual for her, and her association with the dream was that the two
figures represented myself. One was the person whom everybody could see in the
crowded room, but the other one was ‘her analyst’, her special possession. She felt that
she was not going to mind the holiday any more than she had minded or had been jealous
about seeing me with other people, because she had this special relation to me which was
permanently hers alone. In the first dream it is clear that she deals with her jealousy,
stirred both by seeing me with other people and by the analytic holiday, through splitting
and idealization; she has got the illuminated, idealized analyst whom no one can take
away from her.
In the second dream the patient dreamed that there was a small girl sitting on the floor
cutting out paper with a pair of scissors. She was keeping the cut-out piece to herself; the
floor was covered with discarded bits of paper which other children were busy collecting.
The second dream is another fuller version of the first: it shows how this splitting and
idealization were in fact felt by her. The splitting is expressed in the cutting. She is the
little girl who has cut out of her analyst the cut-out figure which, like the illuminated
figure in the first dream, is her analyst’s good part. The people who could only see a drab
analyst-figure are represented in the second dream by the children who have nothing but
the discarded bits. The splitting seen in the first dream is clearly experienced in the
second dream as an attack, an actual cutting of her analyst into an ideal part and a
worthless part; and what is represented in the first dream as idealization is experienced in
the second dream as taking and retaining for herself the best cut-out bits of her analyst.
The second dream makes it clear that, for this patient, the processes of splitting and
idealization were felt as a very aggressive, greedy, and guilty activity.
Phantasies can defend against realities, but they also defend against other phantasies.
For instance, phantasies of an idealized object and an idealized self are both a wish-
fulfilling phantasy and a defence against underlying terror.
A patient for a long time tenaciously maintained a very idealized and rigid relationship
to me. But it was very brittle. Any disappointment, and particularly any hint of a possible
 Dream, phantasy and art 18

narcissistic injury, would lead to the phantasy collapsing, and she would experience
blackness and terror. Gradually we got in touch with the most horrifying phantasies of
herself disintegrating, disappearing, and being subject to endless persecution. It became
apparent that she had to maintain this idealization of me because of her conviction that
neither she nor I would ever be able to stand the intensity of her hatred, terror, and
despair.
Thus phantasies are a meeting-ground between impulses and defences, as in Freud’s
description of the dream as a compromise formation between impulses and defences.
Klein, though, sees unconscious phantasy not so much as a compromise but as an
expression of both impulses and defences.
‘Unconscious phantasies form the operative link between instincts and mechanisms’
(Isaacs 1948:112).
However, there is a further important link to be made, and this is one between
unconscious phantasy and the structure of the personality. If one thinks of Freud’s
structural model, the superego is a structure resulting from a phantasy of having
introjected an object, and the nature of that object taken inside in phantasy is determined
not only by the reality of the object but also by the projections of the child’s impulses.
And such a projection done in phantasy results in the creation of a phantasy internal
object.
If phantasies exist from the beginning of life, early phantasies of projection and
introjection would lay a base for the personality structure. Rosenfeld, in his paper on ‘The
psycho-analysis of the superego conflict in an acute schizophrenic patient’ (1952) shows
how in a session the patient demonstrated that he had a phantasy of three painful internal
objects which could be called an early superego. They were the lupus, a brown cow, and
a yellow cow. As was clear from this and the preceding sessions, the lupus was a punitive
superego directed at the patient’s oral sadism; the yellow cow was the superego for his
urinary, envious, and jealous attacks and the brown cow in Rosenfeld’s view was a
destroyed breast changed into faeces. (I think it was, probably, similarly to the urinary
yellow cow, also a breast filled with faeces by projection.)
The phantasies underlying a much more evolved superego are illustrated by the dream
of a neurotic patient.2 Presented by him very early in his analysis, the dream illustrates
the relation between unconscious phantasy, reality, mechanisms of defence, and ego
structure. The patient was an officer in the Polish navy. He had a very rigid personality.
He often felt his superior officers as overbearing, but he also had a very strict inner sense
of honour and duty by which he felt oppressed. He was afraid of being too strict with the
sailors, with whose rebelliousness he often felt identified. They also presented him with
great homosexual temptations and guilt, since as his own subordinates he considered
them as the most forbidden objects. He dreamed of a pyramid. At the bottom of this
pyramid there was a crowd of rough sailors, bearing a heavy gold book on their heads.
On this book stood a naval officer of the same rank as himself, and on his shoulders an
admiral. The admiral, he said, seemed in his own way to exercise as great a pressure
from above and to be as awe-inspiring as the crowd of sailors who formed the base of the
pyramid and pressed up from below.
I was of course rather suspicious of the dream as bookish. But I had gathered that the
patient knew very little about psychoanalysis. Later on I found out that all he knew was
that Freud and psychoanalysis had something to do with sex. In his associations he
 Phantasy 19

identified the admiral with his father, who in reality was rather rigid and oppressive, and
the sailors with his own rebellious and sexual feelings. He also said that the admiral was
just as strong and frightening in the dream as the sailors. He tried to keep to the ‘golden
mean’ and felt squashed between the two.
This made it clear that the severity of his superego was also due to his own
projections. We can see here the interrelation between phantasy and external reality, the
reality of the father being altered by projections. His defence mechanism of repression is
represented in the phantasy by the combined pressure of the admiral-superego and the
officer-ego trying to keep the instincts, represented by the sailors, under. We can also see
the operation of projection and introjection, in that his father is altered by projections
since his power is the same as that of the sailors and he is introjected to form the
superego.
The dream was such a clear pictorial representation of his mental structure,
corresponding so exactly to Freud’s diagram of the id, the ego and the superego, that it
struck me of course as being rather pat and too easily accessible. We later discovered that
the structure of his personality, as represented in the dream, and quite near consciousness,
was rigidly maintained to protect him from much more primitive anxieties and objects.
This may be why it was so easily represented and presented to me. Rosenfeld (1962)
describes how an idealized, oppressive, and over-rigid superego can be internalized at the
beginning of latency as a defence against underlying paranoia and depressive anxieties.
One can look at unconscious phantasy and the structures evolving out of it as
determining the basic structure and character of the personality, as the matrix of our
mental structure and life.
Klein’s view of the early object relationships and primitive phantasies affected her
views on narcissism. Freud himself wavered in his view about the earliness of object
relationships. While his main line of thought, till 1920, postulates early phases of auto-
erotism and primary narcissism, in other places he describes auto-erotism as a turning
away from the mother’s breast towards one’s own body. He also states in his paper ‘On
narcissism’ (1914b) that he has to admit that he never found any clinical evidence of the
existence of primary narcissism.
Freud sees instincts as ‘merely soldered to objects’ (1905a, 1911). In Klein’s view
instincts are object-seeking and phantasies are about the relation between the self and the
object. Even if on the surface a phantasy is apparently wholly narcissistic, on analysis it
always manifests itself as containing an unconscious object relationship (Rosenfeld
1964a). For instance, the phantasy may be of oneself having incorporated, and become
identified with, an ideal breast, or some later version of an ideal object, and this is
accompanied by a projective phantasy of the distressed, angry or envious infant self being
in others. Another common phantasy is of being inside, and having taken possession of,
such an ideal object. This is illustrated by a very narcissistic patient who would either go
into a dreamy, withdrawn state or act out in a manic way.3 He would spend hours
exercising in order to pile up muscles on his chest and shoulders. His dreams and
associations would reveal that those muscles represented to him breasts. He had
conscious sexual phantasies of getting into women’s bottoms. They occurred frequently
while he was exercising. There were more unconscious phantasies of getting inside his
own bottom, felt as even preferable to women’s bottoms. These states were invariably
accompanied by feelings of persecution. He was forever fighting with intruders at work
 Dream, phantasy and art 20

and at home. His dreams revealed the underlying unconscious phantasies. He often
dreamed of fortresses that were under siege. In one dream, a desert island he lived on
was surrounded by cannibals in boats who were going to land. He was narcissistically
inside of and identified with me as mother (island), his depressed, intrusive, cannibalistic
child self was projected outside. His awareness of his own destructiveness was minimal,
and persecution was mainly dealt with by denial.
A typical dream: He dreamed of a girl mime of his acquaintance. He saw her in the
dream, miming a very gaudy figure. She was in a room with a funny ceiling, round,
concave, and also gaudy. The ceiling he associated both to a womb, because of its colour
and gaudiness, and to the breast. He immediately recognized that the mime girl was
himself since in the previous session he had spoken of his imitating me. So he phantasied
he was both inside a womb/breast and being it. His very way of associating was partly an
acting-in. He was interpreting the dream to himself, imitating me, and at the same time
complying with what he thought I would say, so that he would be ‘in my good books’.
The personality grows, matures, and develops. Growth and evolution of an individual
are due not only to physiological growth and the maturation of the perceptual
apparatus—memory, and so on—but also to accumulated experience and learning from
reality. Learning from reality in turn is connected with the evolution and changes in
phantasy life. Phantasies evolve. There is a constant struggle between the infant’s
omnipotent phantasies and the encounter of realities, good and bad.
The little girl who in the session could express her phantasies of a biting, spitting,
dirtying ‘lady on the wall’ did not during the break have phantasies of a visual kind; nor
dreams, as far as we know. She suffered from stomach pains and night fears which she
could not describe. When she came into the consulting-room the phantasy became visual,
and it was mistaken for a perception, a misperception close to a hallucination. In the
course of development phantasies evolve from part-object relationships to whole-object
relationships; from the predominance of most primitive oral, urinary, and anal drives to
more genital ones.
But the evolution is not in the content of phantasy only. There is a simultaneous
evolution from a concrete primitive perception to a differentiation between phantasy and
reality perceptions. This evolution depends crucially on a gradual overcoming of the most
primitive omnipotence through reality-testing, leading to increasingly realistic
perceptions of one’s self in the world. There is a constant struggle between the infant’s
omnipotent phantasies and the encounter of realities, good and bad. In the earliest and
most primitive forms of feeling and thought, which Klein describes as the paranoid-
schizoid position, the infant resorts to projective identification, omnipotently to alter the
reality. However, from the beginning there is some reality-testing, and the mental life of
the growing individual will be crucially affected by his capacity to recognize and tolerate
the discrepancy between his omnipotent phantasy and expectation, and the reality that he
encounters. This in turn will be affected both by the nature of the external experience
being tolerable enough and the capacity of the infant to tolerate such discrepancies as he
encounters. An angry baby with a phantasy of a persecuting breast may, to begin with,
turn away from the returning feeding object, misperceiving it as bad. Some infants will
never accept the returning object and they develop feeding difficulties. Others may, after
an initial refusal, recognize the reality of the returning good object and accept the feed.
And the mother’s own capacity to tolerate patiently the initial refusal may affect the
 Phantasy 21

outcome. Expecting an ideal breast, which would return one to the imagined bliss of
intra-uterine existence, the infant has to tolerate the breast that falls short of such an ideal.
But, even though reality-testing has to start from the beginning of life, the real
battleground for the development of a mature relation to reality lies in the move from the
paranoid schizoid to the depressive position. The depressive position has been described
by Klein (1935, 1940) as the state of mind appearing in the infant when he starts relating
to mother as a ‘whole object’. Previous to that, the infant is in the paranoid-schizoid
position (Klein 1946, 1952). This is characterized by a wholly egocentric ‘part-object’
relation, in which the infant perceives the object only in terms of his experiences split
into good and bad objects, attributed to a good or bad breast respectively. Splitting,
idealization, projective identification, and fragmentation predominate. In projective
identification the infant projects not only his impulses but also parts of himself into the
object, thus confusing the internal and external worlds. With a gradual withdrawal of
projective identifications, together with a change in content and intensity of projection, a
truer perception is established of mother as a separate person with her own continuity and
characteristics, good and bad, and of oneself as having contradictory impulses, loving and
hating, towards that person. This allows for the differentiation between oneself and the
object, awareness of guilt, and fear of loss. In the earlier state of mind one could hate and
wish to annihilate the bad object, and love, idealize, and keep the good one. When mother
as a whole is felt to be omnipotently destroyed in hatred the needed and loved one is
destroyed as well. New impulses appear—the wish to restore and regain the lost object—
reparation. The implication of this shift is enormous and still being worked out (Spillius
et al. 1989).
I have been particularly concerned with the implication of that shift on reality-testing
and symbolism. There is a reciprocal relation between reality-testing and the depressive
position: some reality-testing is necessary to initiate the changes from the paranoid-
schizoid to the depressive position. Through tolerating discrepancies between his ideal
and persecutory expectations and the encounter with the real object, the infant can
gradually integrate the good and the bad, and perceive a whole object alternately or
simultaneously as good and bad, loved and hated. And this kind of reality-testing paves
the way to the recognition of his mother as an actual person with human characteristics
rather than a part-object wholly related to himself. Once this change begins to happen, the
change of orientation brought about in the child’s mind pushes him to ever-increasing
recognition of reality. An important part of reparative impulses in relation to the object is
the withdrawal of projective identifications and allowing the object a separate existence.
If this is achieved, omnipotence is diminished and the acceptance of the differentiation
between the self and the object leads to the differentiation between the phantasy, which is
a product of the self, and the reality, which is outside. And the objects that are
internalized in phantasy and become such internal structures as the superego, become
increasingly realistic. The nature of the phantasy changes from ideal and persecutory
part-object relationships to whole objects such as the parental couple, siblings, and
family, and the internal conflict worked out in phantasy is concerned more with
reparation of damage done to those entities. It is also in the depressive position only that
gradually repression replaces the more primitive defences of splitting, idealization, and
projection. The infant becomes more separate and differentiated from his object and
capable of feeling guilty about his impulses and phantasies. He therefore represses them.
 Dream, phantasy and art 22

And it is when repression functions that repressed impulses and phantasies can give rise
to sublimation.
To give but one example, Patient A, a schizoid woman whose very disturbed mother
had committed suicide when A was adolescent, had the following dream. She dreamed of
two balconies, one bathed in sunshine; one black and in gloom. Her association was to a
party in the castle of a foreign analyst in a foreign country which she knew her analyst
attended; and the black balcony she associated to a joke that Dr S used to throw his
patients down the cliff from his balcony. That patient tried to adhere to a rigid
idealization of her analyst, to defend against a very persecutory situation. In her life she
was rigid, relatively unproductive, and quite paranoid. She experienced her whole family
as destroyed, fragmented, and persecutory. Many years after that dream, she had one in
which she was doing a jigsaw puzzle, and when it was finished it represented a house in a
garden and a complete family on the lawn. It was very hard work but she was delighted.
Her associations were manifold. First, to her changed view of her family, having
recaptured some memories of the time when it was happier and her mother was more
accessible, to a jigsaw puzzle which she had accidentally knocked off, and to the mental
activity of putting things together in the jigsaw puzzle. She said, ‘It was like thinking—
putting those together.’ At the time of that dream she was writing a book, totally
unconnected with her family or anything to do with it, but obviously her mental activity
was felt to restore something in her mind felt as a home and a family. The change is not
only in the content of her unconscious phantasy life; it is also a change in relation to her
mental life, in that she now recognizes that her phantasies are a mental activity.
Higher mental activity, like thinking, is an interplay between phantasy and reality. We
do not approach reality with a mind that is blank. We approach reality with expectations
based on our preconscious or unconscious phantasies and experience reality, not only in
infancy but throughout our lives, as a constant implementing and testing of our phantasies
against reality. What is reality-testing? You can only test a hypothesis. Unconscious
phantasies are like a series of hypotheses which can be tested by reality (Segal 1978).
In ‘Negation’, a paper in which Freud gives to phantasy a more fundamental role than
in many other papers, he says:

The first and immediate aim therefore of reality-testing is not to find an

object in real perception, which corresponds to the one presented, but to
re-find such an object to convince oneself that it is still there.
(Freud 1925b:237)

The richness, scope, and correctness of our mental activity is linked with our relation to
unconscious phantasy. If unconscious phantasies are split off or too severely repressed,
our conscious life is impoverished and restricted. On the other hand, if our reality-testing
is impeded and the unconscious phantasies affect perception and dictate behaviour
uncorrected by reality-testing, our mental life may appear to be rich but it is delusional
(Segal 1978).
I think that the main thrust of Freud’s thinking is that phantasy is not a primary
activity. It has the same roots as, and is comparable to dreams, symptoms, parapraxes,
and art; it does not underlie dreams, symptoms, thought, and art. For Klein, on the
contrary, unconscious phantasy is a core primary activity, an original expression of both
 Phantasy 23

impulses and defences, and it is continually interacting with perception, modifying it but
also modified by it. With maturation and increasing experience, the phantasies become
more complex, with more differentiated sensory components and motives, and elaborated
in various ways. Hence, for Klein unconscious phantasies underlie dreams, symptoms,
perception, thought and creativity. They do not intrude into a dream; they are ‘such stuff
that dreams are made on’.
Unlike Freud’s view that unconscious phantasy occasionally intrudes into dreaming,
the view of phantasy that I present is that dreaming is but one of the expressions of
unconscious phantasy. It has been noted by Freud that phantasies have the same
unconscious content, and the same mechanism, the same formation, as dreams. One could
say that they use the same language. When Freud discovered dream language and the
dream-work, he discovered a dream world and a dream language which are with us
whether asleep or awake—the world and the language of unconscious phantasy.

Notes
1 Patient described before in the Introduction to the Work of Melanie Klein (Segal 1964) at page
18.
2 Patient described before in the Introduction to the Work of Melanie Klein (Segal 1964) at page
20.
3 Patient described before in Models of the Mind (Rothstein 1985) at page 41.
 3
Symbolism

It is not possible even to approach the topics of phantasy and dreams without the concept
of unconscious symbolism. The discovery of symbolism and phantasy were
interdependent and interrelated. Freud had discovered that hysterical symptoms had a
meaning. They had a meaning in that they were a symbolic expression of underlying
repressed phantasies. It is through symbolism that unconscious phantasy is expressed,
whether in symptoms, in dreams, or in ordinary human relationships and endeavour.
From the beginning, psycho-analytic work has been largely concerned with
understanding the symbolic meaning of the patient’s communication. At the beginning of
his work, Freud used the concept of symbolism in this broad sense. He does so
throughout his work on unconscious symbolism of hysterical and obsessional symptoms,
and to begin with in his work on dreams; he speaks of symbol-formation (Freud
1900:349–52). In the course of writing The Interpretation of Dreams, however, he
introduces the term ‘indirect representation’, and begins to differentiate what he calls
‘symbols’ from other forms of indirect representation. In the second volume of The
Interpretation of Dreams he says:

When I had become familiar with the abundant use made of symbolism
for representing material in dreams the question was bound to arise
whether many of these symbols do not occur with a permanently fixed
meaning. Like the ‘grammalogues’ in shorthand… I felt tempted to draw
up a new dream book on the dictating principle.

And later:

We must restrict ourselves here to remarking that representation by a

symbol is among the indirect methods of representation, but that all kinds
of indications warn us against lumping it in with other forms of indirect
representation without being able to form any clear conceptual picture of
their distinguishing features.
(Freud 1900:351)

He eventually comes to emphasize the following characteristics: that symbols are almost
entirely universal; that they may vary in certain respects in different cultures; that they
are given and not formed; and that they derive from an archaic past. In the dream he
describes symbols as ‘mute elements’, in that patients have no associations to them and
the interpretation has to be done uniquely by the analyst. Jones, in 1916, in his paper ‘The
theory of symbolism’, based on Freud’s work, starts by differentiating between conscious
symbolism (which he unfortunately describes as metaphor) and unconscious symbolism:
 Symbolism 25

1 A symbol represents what has been repressed from consciousness, and the whole
process of symbolization is carried on unconsciously (Jones 1916:97).
2 All symbols represent ideas of ‘the self and of immediate blood relations and of the
phenomena of birth, life and death’ (Jones 1916:102).
3 ‘A symbol has a constant meaning.’ Many symbols can be used to represent the same
repressed idea, but a given symbol has a constant meaning which is universal (Jones
1916:97).
4 Symbolism arises as the result of intrapsychic conflict between the ‘repressing
tendencies and the repressed’. Further: ‘Only what is repressed is symbolized; only
what is repressed needs to be symbolized’ (Jones 1916:115–16).
He further distinguishes between sublimation and symbolization. ‘Symbols’, he says,
‘arise when the affect investing the symbolized idea has not, as far as the symbol is
concerned, proved capable of that modification in quality which is denoted by the term
sublimation’ (Jones 1916:139).
Summarizing Jones’s points, one might say that when a desire has to be given up
because of conflict and repressed, it may express itself in a symbolical way, and the
object of the desire which had to be given up can be replaced by a symbol.
This I think narrows the view of symbols even further, in that he considers that
symbols are not part of sublimation. He is also more definite in the statement that the
symbol has only one meaning, whilst Freud allows more flexibility (Freud 1900:353):
‘They frequently have more than one or even several meanings, and, as with Chinese
script, correct interpretation can only be arrived at on each occasion from the context.’
Some of Jones’s statements are very fundamental and up till today not controversial;
namely, that symbolism at depth represents the ‘immediate blood relations and the
phenomena of birth, life and death’. His statement that symbolism arises as the result of
intrapsychic conflict, that it represents what had been repressed, and that the whole
process of symbolization is carried on unconsciously is also universally accepted. In
other ways, however, later work has thrown doubt on some of his statements. For
instance, does a symbol really have fixed meaning? For instance, one can see how many
meanings may be packed into one symbol. Also, are there symbols that are given and do
not necessitate psychic work and are therefore not part also of dream work? And are
those symbols really any more mute than other forms of indirect representation? And,
more importantly, do symbols really only appear in dreams or symptoms, and have no
part in sublimated activities? Certainly, both Freud and Jones freely used the concept of
unconscious symbolism in explaining works of art. And yet there is no doubt that there
must be some fundamental difference between symbolism giving rise to symptoms or
being expressed in a work of art. Freud had noticed that unconscious phantasy underlies
both, but is it not also expressed in both in a symbolic way?
I think a new approach to the problem of symbolism came in Melanie Klein’s work
with children. She uses the concept of symbolism as it was originally used by Freud. She
understands the child’s play in the consulting-room as a symbolic expression of
underlying unconscious conflicts, desires, and phantasies. Her 1923 paper, The role of the
school in libidinal development’, shows how she saw the symbolic expression of
unconscious phantasy not only in the playroom, not only in the symptoms, but also in the
child’s everyday activities. For instance, she notices that to many children the school
building symbolizes the mother’s body and the teacher a father or a penis inside it. And
 Dream, phantasy and art 26

in the consulting-room the child expresses his unconscious phantasy life in play that
symbolizes it, and gives access to the analyst’s understanding of the phantasy through its
symbolic meaning.
She gradually developed a particular interest in the intellectual inhibitions of children,
and attached great importance to the role of symbolism in intellectual development. In
her paper ‘A contribution to the theory of intellectual inhibitions’ (1931), Klein equates
intellectual inhibitions with an inhibition of the symbolic function, though she does not
specifically spell it out. In her 1923 paper Klein still puts emphasis on libidinal trends and
castration fears, strictly following Freud’s ideas. As her work developed, however, she
paid increasing attention to aggression and the resulting anxiety and guilt. She saw
anxiety and guilt as one of the prime movers to symbol-formation. The child’s
epistemophilic instinct with libidinal and aggressive components gives rise to wishes and
phantasies of exploring the mother’s body. Anxiety and guilt due to the aggressive
components lead to a displacement of the epistemophilic urge to other objects, thus
endowing the world with symbolic meaning. But if the anxiety is excessive the whole
process is inhibited.
In one of her most seminal papers, ‘The importance of symbol-formation in the
development of the ego’ (1930), she addresses herself specifically to an inhibition of
symbol-formation and its catastrophic effect on the whole development of the ego. She
described an autistic little boy of four, Dick, who could not talk or play; he showed no
affection or anxiety and took no interest in his surroundings apart from door-handles,
stations, and trains, which seemed to fascinate him. His analysis revealed that the child
was terrified of his aggression towards his mother’s body, and of her body which he felt
had turned bad because of his attacks on it; because of the strength of his anxieties he had
erected powerful defences against his phantasies about her. There resulted a paralysis of
his phantasy life and of symbol formation. He had not endowed the world around him
with any symbolic meaning and therefore took no interest in it. Melanie Klein came to
the conclusion that if symbolization does not occur, the whole development of the ego is
arrested. In this paper she concludes that excessive anxiety in relation to the mother’s
body and the onset of guilt bring about a paralysis of symbol-formation.
On re-reading that paper I was struck that what she describes in Dick’s relation to his
object is not only aggression but also a massive projective identification. She describes
here a process which she conceptualized much later on. Dick in his phantasy attacks his
mother’s body by projecting into her what he feels to be his bad urine, faeces, and bad
penis, representing also parts of himself. As a consequence, he sees her body as filled by
bad and dangerous fragments of himself, and feels himself empty. The analysis of Dick
must have been one of the important sources of her later formulation of the process of
projective identification.
In my analysis of my first psychotic patient, Edward (Segal 1950), I was struck from
the beginning by the nature of his concrete thinking—for instance, that words could be
the same to him as objects. I had great difficulty in interpreting to him, because my words
were experienced as things or as actions. For instance, interpreting to him a fear of being
castrated would be experienced by him as my actually castrating him. In the same way, if
he was angry with me in one session he would hallucinate my face as black with anger
next day and shout for help to protect him from this ‘black Indian’. I had at the same time
a patient described in my paper (Segal 1952) ‘A psychoanalytic contribution to
 Symbolism 27

aesthetics’, who was a writer and frequently inhibited in writing when she started to
experience words as broken-up bits of things. A border-line patient I had later on often
could not read because she thought words were jumping out of the page and actually
biting her eyes.
In 1957, in my paper ‘Notes on symbol formation’, I tried to tackle theoretically the
problems of symbolic functioning I was struggling with in my clinical work. To help my
understanding I used Klein’s theoretical framework of the paranoid-schizoid and the
depressive positions. At the beginning of that paper I described two patients, a hospital
psychotic who since his illness had stopped playing the violin, and who, when asked
why, answered with violence, ‘Do you expect me to masturbate in public?’; whilst an
analytical patient I had at the same time had dreams about playing the violin, also
representing masturbation and associated phantasies which in no way interfered with his
sublimation in playing the violin.
I came gradually to the conclusion that one could differentiate between two kinds of
symbol-formation and symbolic function. In one, which I have called symbolic equation,
and which underlies schizophrenic concrete thinking, the symbol is so equated with the
object symbolized that the two are felt to be identical. A violin is a penis; playing the
violin is masturbating and therefore not to be done in public. In the second case, that of
true symbolism or symbolic representation, the symbol represents the object but is not
entirely equated with it. To the patient who dreamed of the violin, the violin represented
the penis, but was also differentiated from it, so that it could both embody unconscious
masturbation phantasies and yet be sufficiently differentiated to be used as a violin as
well, to make music which could represent intercourse but not be equated to intercourse.
The move from symbolic equation to symbolic representation is beautifully described by
Claudine Geissman (1990). She describes the evolution of the use the child made of litde
stones or marbles in the course of six years’ analysis. This psychotic litde girl started her
analysis in the setting of a day hospital when she was eight. She could speak only a few
words and could apparently understand the speech of others only a little. She could not
play, and her main activities were to tear, break, cut objects, and kick and hit other
children or adults. Any object she was interested in she would immediately put into her
mouth or throw out. Often she put in her mouth litde round stones, smooth and brown,
and if she could get hold of them, small brown marbles. She also used them as missiles
for hitting objects, human or inanimate. If she lost or mislaid one she would get into
states of uncontrollable violence towards others and herself—tearing her hair and
scratching her forehead.
She started her analytic treatment eagerly, and it appeared in her analysis that what
interested her, and what made her follow the analyst to the treatment room, was the fact
that the dress the analyst wore had some circles forming part of a geometric pattern in the
fabric. Geissman traces the fate of the stones and marbles in the child’s analysis. To
begin with, the child used them in the way described, sucking them or spitting them out.
In the course of the first few months she discovered in the lavatory that the lavatory chain
was made of a succession of little metal balls. And she started to play with it. The main
game was trying to put the lavatory chain as far as possible into her own throat and then
to pull it out. Sometimes she would stroke the balls with her fingers. Dr Geissman
thought it was an important move in the transference in that she related to something
provided by the setting. The analyst understood the child’s various activities in the toilet,
 Dream, phantasy and art 28

as well as the game with the chain, as meaning that the child was treating the toilet and
the chain concretely as the mother’s body. She was trying to swallow the balls of the
chain as a concrete equivalent of parts of the analyst/mother’s body.
The first attempt which the child made at a symbolic representation was also
connected with the marbles. She asked the analyst to draw for her a large grey marble.
This was also her first complicated sentence, in that she could put together two qualities,
large and grey. She crossed out the drawing with great satisfaction, but it was followed
by a violent motor discharge, agitating her whole body.
In the next few sessions she continued to ask the analyst to draw the marbles, but she
herself coloured them and kept comparing them with the marbles she would take out of
her mouth. Gradually she started to draw them herself. She also asked the analyst to draw
a variety of objects and to name them. Her vocabulary increased considerably during this
time.
This seemed a first step in her greater ability to represent things symbolically.
However, her symbols quickly became concretized, in that she would then proceed to wet
the drawings and stuff them into her mouth. This process was linked with a variety of
hypochondriacal complaints: fear of vomiting, of expelling poisonous gases, and so on.
But the phantasies that she expressed in the stones and marbles and in the drawing of
them became gradually clearer, showing, for instance, a clear split between blue marbles
(the analyst has blue eyes), which represented her love and an ideal breast; and the red
and the black marbles, into which she projected violence, hatred, or despair. The
drawings were still so concretely felt as part of herself and containing her feelings and
thoughts that it was impossible for her to leave the room without taking them with her,
which the analyst allowed since it seemed that if she left the room without them she
appeared to feel completely empty of any thought, feeling, or capacity for movement.
Gradually she was able to shift her interest to other, similar objects—for instance, the
necklace of pearls which the analyst wore, and eventually little rubber balloons that she
blew up and then tore. At the end of one session she collected the debris of these balloons
and asked the analyst to draw a whole balloon with the debris inside. This the analyst
thought was the first time she tried to bring a whole, undamaged object together with the
one that was broken or bad. And by making the good one contain the debris of the other
she was making a step towards integration and reparation.
In the months that followed, the patient brought jacks, and added to them her litde
stones and marbles. And she started playing games with the analyst, using these objects
as proper toys. This development in her way of using the litde marbles was paralleled in
the rest of her development. Now in the sixth year of her analysis, the girl can speak,
read, write, relate to the other children, and play.
These very concrete methods of symbolization appear not only in the schizophrenic.
They often underlie inhibitions, as in my writer patient.
Anthropological examples are sometimes very explicit. James Mooney, in ‘The ghost-
dance religion and the Sioux outbreak of 1890’ (Annual Report of the Bureau of
American Ethnology, XIV (2) (Washington, 1896); 721, 724), reports:

An Indian prophet, Smohalla, chief of the Wanapum tribe, refused to till

the ground. He held that it was a sin to mutilate and tear up the earth,
mother of all. He said: ‘You ask me to plow the ground! Shall I take a
 Symbolism 29

knife and tear my mother’s bosom? Then when I die she will not take me
to her bosom to rest. You ask me to dig for stone? Shall I dig under her
skin for her bones? Then when I die, I cannot enter her body to be born
again. You ask me to cut grass and make hay and sell it, and be rich like
white men! But how dare I cut off my mother’s hair?’

The Indian chief expresses here poetically what was the unconscious dilemma in my
writer patient as well as the dilemma of many pastoral people forced into agriculture.
Concrete symbolization is conspicuously at the root of pathological mourning. If the
dead person is felt as a concrete dead body, or as faeces, inside oneself, then normal
mourning is not possible. It is only if the dead person can be felt as symbolically
introjected and the internal object is symbolic of the lost person, that internal reparation,
necessary to overcome mourning, can be achieved. An actually dead person cannot be
brought back to life; nor can faeces be changed back into milk. It is only if the dead
person is symbolically represented in the mind that the symbolic internal reparation can
be done. Patient A, whom I quoted in the previous chapter, and who dreamed of putting
together a jigsaw puzzle, said in her associations that it was a mental activity. Previously,
among other difficulties, she had suffered from hypochondriacal and psychosomatic
difficulties connected with a concrete phantasy of a fragmented dead mother and family
attacking her body. It is only when the nature of her symbolism changed, as expressed in
her dream of the jigsaw puzzle, that reparative internal work could be done, leading to a
restoration of an internal family and the possibility of acceptance of the death of her
mother and mourning. In her case the mourning had to involve the recognition of the real
illness and associated badness of her mother, so different from the mother she wished to
have, and only very seldom experienced. (Often such a recognition is even more painful
than an actual loss.)
I have come to the conclusion that the two modes of symbolism pertain respectively to
the paranoid-schizoid and the depressive position. Klein links symbolism with projection
and identification. She says that she agrees with Ferenczi that symbolism starts with
projection of parts of the infant’s own body into the object. And yet her work on
symbolism is mainly centred on introjection and reprojection. The child introjects and
symbolizes the mother’s body, and it is this internal mother that is then displaced on to
the external world. Klein at that time had not yet worked out the actual interplay of
projection and introjection. I think that in her later work her concept of projective
identification threw a new light on the whole problem of symbolism. It struck me in my
work that concrete symbolism prevailed when projective identification was in
ascendance. This also seems logical. Symbolism is a tripartite relationship: the symbol,
the object it symbolizes, and the person for whom the symbol is the symbol of the object.
In the absence of a person there can be no symbol. That tripartite relationship does not
hold when projective identification is in ascendance. The relevant part of the ego is
identified with the object: there is not sufficient differentiation between the ego and the
object itself, boundaries are lost, part of the ego is confused with the object, and the
symbol which is a creation of the ego is confused with what is symbolized. It is only with
the advent of the depressive position, the experience of separateness, separation and loss,
that symbolic representation comes into play.
 Dream, phantasy and art 30

I would like to illustrate it with two extracts of material from a patient at about a two-
year interval.
A neurotic young man, Patient C, is able much of the time to function on a depressive
level. He can communicate in a symbolic way and has numerous sublimations. These
achievements are, however, insecure, and at moments of stress he tends to use massive
projective identification accompanied by a regression to concrete levels of functioning.
Sometimes, for instance, he has near hallucinatory states of mind. He came to one session
very perturbed because on waking up he had a hallucinatory experience. It differed from
hallucination only in so far as he clung desperately to the belief that it must be the
product of his own mind. When he woke up he felt his head was solid and he saw a
motor-cycle riding into his head. The rider had a kind of mask on, which made his head
look like a gorilla. He felt terrified and thought his head would explode. Then he looked
at his own index finger, and got frightened because his finger looked like a gorilla. He
only emerged from a state of acute anxiety when he made himself remember the previous
session in which he was disturbed by a very intrusive noise of motor-cycles outside the
consulting-room windows. He thought the motor cycles were connected with my son. He
associated the gorilla to a psychotic boy who was described in a paper as looking like a
gorilla. The finger was associated to anal masturbation, which he had spoken of a few
days previously. His anal masturbation was always associated with violent projective
identification into the anus of the analyst/mother, as described by Meltzer (1966). We
understood that the motor-cycles outside the window represented his own intrusive self,
identified with his finger and penis, projected into an external object, the motor-cycle of
my son identified with it, and intruding into him. The same patient about two years later
showed clearly in a dream how he felt this process of concretization had occurred.
One day he told me that, as he was going past my consulting-room door to the
waiting-room, he became very anxious because the thought occurred to him that there
was no guard at the door and nothing to stop him from getting into the consulting-room
and interfering with the session of my other patient. Then he added, ‘Come to think of it,
there is nothing to stop me doing what I want on the couch. For instance, if I wanted to, I
could lie upside down.’ Then he giggled, and became embarrassed as he realized that
upside down in the bed is the position he was in during some love play with his girlfriend
the night before. So, apparently the situation was as follows: There is no guard at the
door, no husband. He could have intercourse with me as his girlfriend and have our
positions upside down; that is, with him dominating me—apparently, a plain Oedipal
situation. He went on to tell me a dream. He said: ‘I had a dream in which I was
explaining to M (the girlfriend) about my hallucinations. I was telling her, “Look, I
dream up a car and there it is.” And the car appeared.’ He got into the front seat. But
there was no partition between front and back—no pole to lean against. He started
falling backwards, feeling an utmost panic. And he woke up with severe anxiety.
My understanding of his associations preceding the telling of the dream, and the
dream, was as follows. The pole is a phallic symbol. But also, I am of Polish origin, and
he knew that my husband’s name is Paul. In the absence of the pole, the father, or the
penis in the vagina, there is nothing to stop him not just from having intercourse with his
mother on a genital level, there is nothing to stop him from unrestrained projective
identification with her, leading to the loss of boundaries, confusion, and panic. In this
 Symbolism 31

dream the father’s penis is absent, but in other dreams or hallucinations, the persecutory
penis would return, as in the hallucinations of the motor-cyclist.
What he seems to explain to his girlfriend in the dream is that when he projects
himself into his mother, what used to be his thoughts, ‘what he dreamed up’, is felt by
him as a reality in the external world. But the whole process, instead of hallucination at
that point gave rise to a dream. It was subjected to dream-work which converted it into a
meaningful internal and external communication. The dream-work has somewhat failed,
since the patient woke up in a panic. Nevertheless, since the time of that dream his
hallucination disappeared completely. I do not mean, of course, that that dream cured him
like a magic wand. It represented the integration and assimilation of the insights acquired
in our analytic work.
The aim of primitive projective identification is to deny a psychic internal reality by
getting rid of a part of oneself and simultaneously possessing and controlling the object.
A gradual change happens with the formation and use of symbols in the depressive
position. Projective identifications are gradually withdrawn and the separateness of the
subject from the object becomes more firmly maintained. With that comes a greater
awareness of one’s own psychic reality and the difference between internal and external.
In such a situation the function of symbolism gradually acquires another meaning.
Symbols are needed to overcome the loss of the object which has been experienced and
accepted and to protect the object from one’s aggressiveness. A symbol is like a
precipitate of the mourning for the object. The relation between the capacity to symbolize
and that of mourning is shown clearly by the same patient another couple of years after
his hallucinations ceased.
After considerable change and improvement, the patient was getting married. Before
his wedding, for which he was missing a few days of his analysis, he showed
considerable ambivalence to myself representing the father. When he came back from his
honeymoon he said that he had never been so moved in his life as he was at the actual
wedding ceremony. He decided to marry in church in deference to his dead father, though
he himself was not religious. He asked for his father’s favourite hymn—‘The Lord is my
shepherd’—to be sung at the wedding ceremony. He said that he had never in his life
been so happy and so unhappy at one and the same time. He did not know if at that
moment he was regaining his father or losing him. He was so aware of his father’s
presence in his thoughts, and so acutely aware of his real absence from the wedding. He
then told me a dream he had had the night before in which a fisherman was to take him
out to teach him to fish. The fisherman’s hands were bandaged because they had cuts on
them. The patient was afraid that he would be too hurt to go fishing. But the fisherman
assured him that he could still keep his promise.
In the last session before the break I had occasion to point out to him that he was very
cutting to me. I was the fisherman/father with the bandaged hand. The symbol is formed
by the subject in the process of working through mourning. It represents the object but it
is a creation of the subject and therefore it can be freely used. This is unlike a concrete
symbolic equation, which is never felt sufficiently separate from the object for the subject
to be able to use it freely. Also, in that the symbol is not equated with the object, the
proper characteristics and functions of a substitute used symbolically are fully recognized
and acknowledged. So that my neurotic patient, unlike the psychotic one, could recognize
 Dream, phantasy and art 32

the violin for what it was, however much he needed it also as a symbol for an
unachievable relationship.
Artists in particular, when successful, combine an enormous capacity for symbolic use
of the material to express their unconscious phantasies with a most acute sense of the real
characteristics of the material they use. Failing that second capacity they could not have
used it effectively to convey the symbolic meaning they wish to embody.
I have summarized the differences between the two modes of functioning as follows
(Segal 1957:57):

I should like at this point to summarize what I mean by the terms

‘symbolic equation’ and ‘symbol’ respectively, and the conditions under
which they arise. In the symbolic equation, the symbol-substitute is felt to
be the original object. The substitute’s own properties are not recognized
or admitted. The symbolic equation is used to deny the absence of the
ideal object, or to control a persecuting one. It belongs to the earliest
stages of development. The symbol proper, available for sublimation and
furthering the develop-ment of the ego, is felt to represent the object; its
own characteristics are recognized, respected, and used. It arises when
depressive feelings predominate over paranoid-schizoid ones, when
separation from the object, ambivalence, guilt, and loss can be
experienced and tolerated. The symbol is used not to deny but to
overcome loss. When the mechanism of projective identification is used
as a defence against depressive anxieties, symbols already formed and
functioning as symbols may revert to symbolic equations.
Symbol formation governs the capacity to communicate, since all
communication is made by means of symbols. When schizoid
disturbances in object relations occur, the capacity to communicate is
similarly disturbed: first because the differentiation between the subject
and the object is blurred, secondly because the means of communication
are lacking since symbols are felt in a concrete fashion and are therefore
unavailable for purposes of communication. One of the ever-recurring
difficulties in the analysis of psychotic patients is this difficulty of
communication. Words, for instance, whether the analyst’s or the
patient’s, are felt to be objects or actions, and cannot be easily used for
purposes of communication.
Symbols are needed not only in communication with the external
world, but also in internal communication. Indeed, it could be asked what
is meant when we speak of people being well in touch with their
unconscious. It is not that they have consciously primitive phantasies, like
those which become evident in their analyses, but merely that they have
some awareness of their own impulses and feelings. However, I think that
we mean more than this; we mean that they have actual communication
with their unconscious phantasies. And this, like any other form of
communication, can only be done with the help of symbols. So that in
people who are ‘well in touch with themselves’ there is a constant free
symbol-formation, whereby they can be consciously aware and in control
 Symbolism 33

of symbolic expressions of the underlying primitive phantasies. The

difficulty of dealing with schizophrenic and schizoid patients lies not only
in that they cannot communicate with us, but also with themselves. Any
part of their ego may be split off from any other part with no
communication available between them.
The capacity to communicate with oneself by using symbols is, I think,
the basis of verbal thinking—which is the capacity to communicate with
oneself by means of words. Not all internal communication is verbal
thinking, but all verbal thinking is an internal communication by means of
symbols—words.
(Segal 1957:58)

I have presented two types of symbol-formation in a very extreme way. There is a long
transition between the one and the other mode, and I do not think I have ever seen a
patient the whole of whose function would be on a concrete level or whose concrete
symbols would ever be completely concrete; only predominantly so. Nor do I think the
symbolism of the depressive position is ever free of concrete elements. Any art, in
particular, does embody concrete symbolic elements that give a work of art its immediate
‘punch’; it has a concrete impact on our experience provided it is included in an
otherwise more evolved type of symbolism, without which it would be no more than a
meaningless bombardment. One of the great achievements of the depressive position is
the capacity of the individual to integrate and to contain more primitive aspects of his
experience, including the primitive symbolic equations.
Verbalization is a particular and highly evolved form of symbolism and I should like
to show in some child material supervised by me the appearance of a verbalization in
connection with reaching depressive anxieties. The eight-year-old boy was subjected to
many separations due to the divorce of his parents. For the first six months of his analysis
he hardly spoke. It was reported by his mother that at home he tended to be withdrawn
and taciturn. He could work at school and play but his play was solitary. He played freely
in the consulting-room and responded to interpretations by a change of expression or
play. Occasionally, he briefly answered a question but never spoke spontaneously or
freely. After some sessions before the ones I shall report, he would mutter a hardly
audible ‘thank-you’ at the end of the session. I shall report briefly two sessions and more
fully a third session in the six months of his analysis in which he started communicating
and associating very freely in the context of mobilized depressive feelings.
On the Wednesday preceding a holiday he attached an aeroplane to a string and
whirled it over his head. The analyst* interpreted that with the approaching holiday he
may have thought she would fly away on a plane and he wanted to keep her and control
her movements. He responded by saying spontaneously that his grandparents were
arriving that day on a plane and went on associating very freely to their visit, which
excited him. It appeared he was very fond of them and felt cherished by them. He also
admitted some jealousy of his father, who, in contrast to himself, had a proper couple of
parents.
On the Thursday he was very despondent. He tried to place strings on two pieces of
furniture, but they kept falling between the two. He tried very unsuccessfully to attach the
strings to the furniture. The
 Dream, phantasy and art 34

* Mrs C.Duthy.

analyst interpreted his despondency about people coming and going and of himself
falling between the two parents. She said he felt it unsafe to attach himself to anyone.
Towards the end of the session he cut a string into five bits and then made a knot between
the two last ones. She interpreted that the five pieces of string were five sessions and that
he did want to make a link between today and tomorrow and was less afraid of feeling
attached to her. Throughout this session he did not talk, but at the end he was less
despondent and smiled at her.
I shall report Friday’s session more fully. When he came to the consulting-room he
quickly went to his box and took out some long pieces of string which he had been
cutting off a ball of string over the last few days. He took out his trains and his cars. Up
till then he had always kept his coat on in the session. So the analyst commented on the
fact that he was not wearing his coat and that she noticed that his mother was not with
him in the waiting-room. In response to her comment he said that his mother had to go
off, and at once busied himself with the pieces of string, placing them on the floor and
around the room, hooking them over the furniture and hanging bits here and there. She
said that things appeared to be very precariously tied together. Nothing was properly
knotted and everything was in danger of slipping. She pointed out to him that it showed
something about his relationship with her—how he liked to make only a loose connection
with what she said, to speak to her would be to strengthen the tie with her which might
make things difficult when they parted at the end of the session. At this he paused in his
play and, turning to her, said: ‘Our au-pair, Sophie, is going on the 10th.’ She said, ‘Oh,
and how do you feel about that?’ He replied: ‘Well, it’s OK because another one comes
and replaces the old one—so I don’t notice.’ He paused a moment thoughtfully, and
added: ‘We’ve had lots—ten’, and she agreed with him that it was indeed a lot for a boy
of eight.
Now he set about playing again, making more loose connections with his pieces of
string, though tying one piece a little more firmly across two low cupboards in the room.
She said that it was best, in his view, to continue not to tie things down as it was simpler
to make replacements, but she noted that the new piece of string appeared more firmly
fixed and suggested that their talking together had put them on a firmer footing. As she
spoke he was throwing some pieces of string over the line that he had just strung out,
with a competent and business-like gesture. She said that he was now showing her how
Sophie had looked after him, washing his clothes for him and hanging them up on the
line. As he continued with his washing-line, flinging up the clothes to dry, she said that
he would like so much to replace her quickly so that he did not have to feel sad at her
going. He appeared quite oblivious of what she was saying, but then lost his balance
slightly and bumped his head on the wall, adjusting himself quickly so that she would not
notice his loss of composure and the pain of the bump. She showed him how the thought
of losing Sophie was confusing and painful. He said Sophie looked after him very well.
He then began to play around with the trucks, putting them in a little pile and placing
all the strings in a heap on top of them. He then lifted the trucks from under the strings
and put them on the top of a shelf. Having done this, he swept all the trucks off the shelf
and on to the floor. She said that she thought he was showing her that he was fishing with
a net, that having caught the fish he was throwing them back into the water. He was very
 Symbolism 35

pleased and excited at her having understood, and immediately began to tell her how, on
holiday in Italy, he had been fishing with his father. When they had caught the fish he
had had to throw them back into the water. He explained that it would have been too
expensive to keep them because they had to pay for the fish they caught and retained.
Curious to see if the fish were alive when they were pushed back in, she asked if the fish
swam away. He replied, ‘Oh, no, they were half dead.’ She interpreted his feeling ‘half
dead’ when he thought she was throwing him away on her holiday. After a time he said
that a few of them may be still alive and swam away if they were only half dead. She also
interpreted that he felt it was too expensive and painful to think about Sophie and the
analyst going away and that made him also throw away his feelings and the words that
expressed them. But that he was also pleased that their talking together brought his
feelings to life. At that he became a little anxious and she interpreted that he was
immediately fearful of being sad at the loss of somebody.
There was a long quiet moment, and the child sat thoughtfully for a while. He began to
play in a way that she had seen him do recently quite often, whizzing his cars over
hillocks that he had made in the rug that he was sitting on. She ventured that he had been
thinking about his holiday in Italy because that was where Sophie might come from, and
he at once replied that she was Swiss but that Switzerland bordered on Italy. She then
realized that he had been preoccupied with Sophie’s leaving for quite some time, and she
suggested that the whizzing over hillocks game represented skiing. First he objected,
saying he had never been skiing and did not think about it, but afterwards, when the
analyst suggested that he may have watched ‘Skiing on Sunday’ on TV, he said, yes, he
very often watched it with Sophie, who loved skiing. He told her warmly that she was
very nice and they talked a great deal about Switzerland together. The analyst interpreted
that he was imagining that she would take him with her skiing.
As he spoke he began to gather up his cars and then his mood changed as he began to
run them against the wall of the room violently. She said he was very angry with Sophie
for leaving without him and with her too, as she would be leaving soon for the weekend
without taking him with her. The weekend seemed to him like a cold, snowy place. He
was particularly angry, as the fact that he had spoken to her tied a dangerous knot
between them and it would not be so easy to feel that she was a new au-pair—like a new
analyst on Monday. The patient stopped what he was doing abruptly and turned his back
to the analyst. Suddenly she heard the alarm on his watch and she asked why he, such a
punctilious boy, had set his alarm to go off three minutes before the end of the session, as
it was then, and not at the end as usual. He replied that it had gone off by mistake, that he
was setting it for ten minutes’ time, the time he would reach home. She said that he was
anxious about having been angry with Sophie and wanted to see her urgently to make
sure that she was OK. She also pointed out that crossing from the consulting-room to his
home was frightening, just as crossing from Friday to Monday left a dangerous gap.
What to me is striking and moving is how the little boy started associating freely and
meaningfully when he was enabled to face his depression—anger and sadness at the
partings. I think he was re-living the emergence of speech as part of working through the
depressive position.
I think it is also significant that virtually his first free association (in the Wednesday
session he spoke of the arrival of his grandparents who, unlike his real parents, are a
properly married couple) on the Friday was to the fishing expedition with his father. I
 Dream, phantasy and art 36

think it is significant that, though the actual material is about loss of the girl representing
mother and the mother-analyst, it happens in the presence of the father and the analyst-
father in the session. In the dream of my Patient C, who dreamed about falling into a car,
the father’s presence is needed to stop uncontrollable projective identification in relation
to mother.
In the material of the litde boy, the presence of the married couple and the supportive
father-analyst seem to facilitate the problems of mourning in relation to his mother. It is
an important aspect of the depressive position that the recognition of mother as a separate
person includes the recognition of father as her partner, rather than as a part-object seen
as her possession or as an object confused with her, as in the phantasy of the combined
parents.
It has been recognized, particularly by Lacan, that father’s penis to begin with, and
then father, play a fundamental role at the beginning of language. His explanation,
however, is very different from mine. I think one of the important functions of the father
is that of an object seen as stopping a stream of mutual projective identifications between
child and mother. When the depressive processes are initiated it enables the child to
recognize the father as a separate object and that object in turn becomes a necessary
factor in the further elaboration of the depressive position.
The changes in the nature of symbolism affect an evolution in one’s experience of
internal objects. I think Patient C’s experience at his wedding shows this very beautifully.
He is aware simultaneously of his loss of his father as a real external object and his
regaining of a father as an internal object, but this internal object is not concrete as in
pathological mourning; he is very aware that he regained him, as he said, in his thoughts.
There is also an acknowledgement of his aggressive castrating wishes towards his father,
but no projection of the aggression. The father remains helpful and supportive of his
sexuality.
A similar awareness is shown by Patient A and her association to the jigsaw puzzle as
being a mental process. When internal objects are felt in that way they can be projected
into the external world, as in the case of my Patient C, into the helpful analyst as father,
or, in the case of Patient A and her jigsaw puzzle, into writing a book that will bring
together various elements symbolizing the bringing together of fragments of her mother
and her family. It also allows for greater diversity because various aspects of the internal
object can be symbolized by other events or figures in the external world without
insistence on the object or situation being in all respects a substitute for the lost object.
Money-Kyrle (1965), in his paper ‘Success and failure in mental maturation’,
emphasizes that with integration there is an increasing differentiation of the various
aspects of the object, coupled with a capacity for generalization. Freud is reputed to have
said that every man marries his mother, and yet some marriages are successful and
fruitful; others are pathological because the wife is the mother. I think Freud was right
that every man marries his mother, but the sanity or otherwise of such a procedure
depends on the type and degree of symbolization. The wife may symbolize and contain
some aspects of the mother; or she may be felt to be the mother, in which case the
marriage carries all the prohibitions and conflicts of the relation to the mother.
In talking about phantasy and symbolism I want to emphasize the evolution not only
of the content of phantasy but also of the level of symbolic function. Money-Kyrle, in
‘Cognitive development’, puts it succinctly thus:
 Symbolism 37

To fit such observations the theory of conceptual development has to be

extended to include not only growth in the number and scope of concepts
but also the growth of each single concept through at least three stages: a
stage of concrete representation, which strictly speaking is not
representation at all, since no distinction is made between the
representation and the object or situation represented; a stage of
ideographic representation, as in dreams; and a final stage of conscious
and predominantly verbal thought.
(Money-Kyrle 1968:422)

Once the step has been taken between concrete and depressive symbolism, the basis is
also laid for further abstraction, including verbalization.
However, it does not follow that the capacity for purely abstract thought is of necessity
a sign of mental health. It may be the result of splitting, in which abstract thought is
completely devoid of emotional meaning. Indeed, with schizophrenics very often there is
a simultaneity between crude concrete symbolization and complete abstractions, devoid
of emotional, and sometimes even intellectual, meaning.
Once a higher mode of mental functioning has been achieved, it is of course not
achieved once and for all. There is always a potential for regression. The more extensive
the split-off area of unresolved psychotic conflicts, the more dramatic such a regression
can be. In the typical schizophrenic breakdown in adolescence or early adulthood, the
pressure of adolescent conflicts leads to a regression, to massive use of projective
identification and an associated breakdown of the symbolic function. For instance, the
capacity for speech, which has been acquired—and this I think implies necessarily that
some depressive functioning has been achieved—gets disturbed and words themselves
are treated as concrete objects.
A schizophrenic adolescent described by me (Segal 1957) when in a good phase of her
illness wrote fairy-tales; in a bad phase, some of the fairy-tales ‘came to life’ and the
figures she invented persecuted her. But in a less gross way those regressions occur
momentarily in all of us. Patient C, even after his marriage and all it signifies, still
frequently resorted to projective identifications of the kind described, though never to the
degree of again being hallucinated.
What about symbolism in dreams? Money-Kyrle links dreams with the second stage
of development, which he calls ‘ideographs’. It is arguable that for the imagery of dreams
to be formed at all a depressive level of functioning must have been achieved.
Nevertheless, in some dreams, or some elements in some dreams, a regression occurs to
concrete symbolization with all its consequences for the nature and function of the dream.
 4
Mental space and elements of symbolism

In the previous chapter I presented my ideas on the development from concrete symbol-
formation to the formation of symbols in the depressive position. My assumption was that
concrete thinking and symbolization are a regression to the paranoid-schizoid position.
Following Klein, I thought in terms of excessive protective identification as leading to
pathology. Since then, more work has been done on the actual pathology of the paranoid-
schizoid position, and the difference between more normal forms of protective
identification, even before the depressive position, and forms which are pathological. In
his paper on the ‘Differentiation of the psychotic from the non-psychotic personalities’,
Bion (1957) distinguishes between normal and psychotic forms of projective
identification. In normal development the projection is of parts of the personality or
internal objects which are split off but not unduly fragmented; projections of that kind
can be gradually withdrawn in the depressive position and re-owned by the self. The
more pathological forms are characterized by a great hatred of reality and therefore also
of the perceiving part of the ego, including its perceptual apparatus. The individual
splinters this hateful part of the ego into minute fragments and projects it into the object
with great violence, which in turn splits the object in a similar manner. The result is that
the subject feels surrounded by ‘bizarre objects’. Such objects are minute fragments of
the subject’s personality embodied in minute fragments of the object and imbued with
extreme hostility. The existence of those bizarre objects in the mind can easily be
detected in the psychotic, but they may also exist in a split-off part of the mind in non-
psychotic people, particularly in the severe neurotic. In the case of more normal splitting
and projective identification, projections are gradually withdrawn and integration is
possible. Bizarre objects cannot easily be withdrawn, and the establishment of the
depressive position is severely impeded. Also, unlike the ideal and the bad breast, and the
good and bad parts of the self, which can be integrated, bizarre objects cannot be
integrated; they can only be agglomerated. (Such agglomerations are, I think, at the root
of crowd phobia.)
Bion extended his considerations about the fate of projective identification to a theory
about the formation of the mental apparatus based on the interplay between the
‘contained and the container’. This interplay can take on a benign or a malignant
character. At the dawn of life every infant tries to deal with his pains and needs by
projecting them into an object. Projective identification is then in ascendance. What
Freud describes as the ‘motor discharge’ in a normal infant I see as an expression of
violent projective identifications.
I had a patient who dealt with being hungry by defecation, and had an elaborate theory
to account for why he thought such a solution would work. But the persistent evacuation
of need, pain, or hatred into the object, and then identification of the object with such
projection, leads to the creation of a bad and fragmented object. This in turn leads either
 Mental space and elements of symbolism 39

to a reintrojection of such an object, resulting in an increase of fragmentation of the self,

or to attempts at blocking all introjection. We have always assumed that a good
experience can modify the perception of the object and the self. The question is, what is
the nature of this good experience? According to Bion, the good experience for the infant
is that the containing object modifies in some way the part that had been projected into it.
He describes how sojourn in the breast seems to ameliorate the projected parts.

Melanie Klein has described an aspect of projective identification

concerned with the modification of infantile fears; the infant projects a
part of its psyche, namely its bad feelings into a good breast. Thence in
due course they are removed and re-introjected. During their sojourn in
the good breast they are felt to have been modified in such a way that the
object that is re-introjected has become tolerable to the infant’s psyche.
(Bion 1952:90)

This idea, that the projected part is modified by the action of the container, had been
foreshadowed in Strachey’s (1934) ‘imitative interpretation’. Strachey says that the
severity of the superego projected into the analyst is modified by the analyst’s
understanding so that its severity is diminished, and it can be re-introjected in a more
benign form. However, Bion goes very much further in these considerations. According
to him, at the first primitive stages of development, the infant is filled with raw
perceptions, objects, and emotions. In Elements of Psychoanalysis he describes it thus:

The infant suffers pangs of hunger and feels it’s dying; racked by guilt
and anxiety and impelled by greed, it messes itself and cries. The mother
picks it up, feeds it and comforts it and eventually the infant sleeps.
Reforming the model to represent the feelings of the infant, we have the
following version: the infant, filled with painful lumps of faeces, guilt,
fears of impending death, chunks of greed, meanness and urine, evacuates
these bad objects into the breast that is not there. As it does so, the good
object turns the no-breast (mouth) into a breast, the faeces and urine into
milk, the fears of impending death and anxiety into validity and
confidence, the greed and meanness into feelings of love and generosity
and the infant sucks its bad property, now translated into goodness, back
again.
(Bion 1963:31)

He calls those raw primitive elements ‘beta elements’. Beta elements are raw, concretely
felt experiences which can only be dealt with by expulsion. They are like a very primitive
form of what I call ‘concrete symbolic equation’. When those beta elements are projected
into the breast they are modified by the mother’s understanding and converted into what
Bion calls ‘alpha elements’. If the beta elements are felt to be concrete things that can
only be ejected, the alpha elements on the contrary lend themselves to storage in memory,
understanding, symbolization, and further development. They are the elements which can
function in the symbolic way which characterizes the depressive position. If the
interchange between the infant and the breast is good, then the infant not only reintrojects
 Dream, phantasy and art 40

its own projections made the more bearable, but he also introjects the container-breast
and its capacity to perform the alpha function; the mother’s capacity to bear anxiety that
is projected into her by the infant is crucial in this interplay. When the infant introjects
the breast as a container that can perform what Bion calls the alpha function of
converting the beta elements into alpha ones, it is a container which can bear anxiety
sufficiently not to eject the beta elements as an immediate discharge of discomfort. An
identification with a good container capable of performing the alpha function is the basis
of a healthy mental apparatus.
This may seem to be very abstruse. Nevertheless, once alerted to it one can observe it
in sessions, and its clinical relevance is enormous. I would like to show this in some
fairly simple clinical material. I presented this material in another context1 but I quote it
here to illustrate the transformation of beta into alpha elements.
The mother of Patient D became pregnant when the patient was only four months old.
Following the birth of the next sibling, the little girl turned her face to the wall and her
back to her mother, and for quite a long period she did not relate to her mother at all.
Later in childhood she had phases of mutism, though not of very long duration. The
patient is not psychotic, but the existence of a split-off psychotic core in many ways
interfered with her development.
A fortnight before the dream I shall report, she accidentally met me outside the
consulting room in conversation with a young woman linked in her mind with her sister.
During that fortnight she was persecuted; severely disturbed and her way of associating
was completely different from her usual manner. The material she brought was
fragmented, incoherent, sometimes nonsensical and it was thrown at me in a hostile,
provocative, and disruptive way, making it almost impossible for me to think. Her
communication was verbal as well as non-verbal, but the words were used as missiles,
and the experience for the analyst was not of receiving a relevant communication but of
being under real bombardment—the kind of behaviour described by Bion as a’screen of
beta elements’. I did, however, manage to make some contact with her, getting in touch
with her distress and the violence provoked in her by the encounter and linking it with her
preverbal experience of her mother’s pregnancy.
After a fortnight she came one day in a quite different mood, and reported the
following dream. She dreamed about being in a session, and that she brought me a
complaint. ‘I mean I didn’t talk about it, I actually brought it to the session. The
complaint was I was shedding from inside my body all sorts of bits and pieces, little
animals, maybe rabbits, bizarre fragments that could be more faecal. I thought (in the
dream session) that the rabbits could be babies, but no, they were too bizarre. I was
terribly anxious and I felt I was falling to pieces. You started explaining what was going
on, but you conveyed that it could not be put into words. You painted for me a
background and some figures. I wondered which was more important, the background or
the figures. The figures became my parents. When that became clear I threw the bits at
them and they—the bits—became dots. I wondered if it is an attack and I thought, yes, it
probably is. Then you say: but the dots are also tears. I feel tremendously moved and not
anxious any more and this is the first clear communication in words within the dream.’
I shall not report her associations, though they were very relevant and enabled me to
make a number of connections, because I want to concentrate on the ‘container’ and
‘contained’ aspect of the dream. I think the experience she showed me in the fortnight
 Mental space and elements of symbolism 41

was of an infant who feels deprived of normal projective identification because she
perceived her mother as already full with the next infant, and blocking it off. In fear and
hatred she tries to project beta elements into me, standing for the pregnant mother, but
she experienced me as blocking them; they rebound unmodified and she feels persecuted
by what I say, feeling it as a return of hostile beta-element fragments. But gradually, as
she finds that I neither collapse under the bombardment nor retaliate, nor block her out,
she begins to feel that she is understood; and her experience alters, as she shows in the
dream. The concrete complaints, bits falling out of her, become thoughts and feelings that
could be put into words. But first the words must be supplied by me. Later on it also
appeared that my painting in the dream was in a frame and the frame represented the
psychoanalytic framework. The experience described in the dream is the benign interplay
between the container and the contained, including a shift from beta to alpha elements,
and it is conjoined with the move from the paranoid-schizoid to the depressive position.
The psychoanalytic setting, with its regularity of time and place, the supporting couch,
and so on, is one of the factors in this containment. But the crucial factor is the analyst’s
understanding. It is when the patient feels understood that he feels that what he projected
into the analyst’s mind can be processed by that mind. He can then feel mentally
contained. When, for external or internal reasons, this benign interchange does not
happen its place is taken by a relation between the container and the contained which is
mutually destructive or denuding.
I remember an experience right at the beginning of my psycho-analytic work which
for some reason remained very vivid in my mind. A colleague working with an elderly
psychotic woman asked me to see her on a Sunday to hold the fort between her sessions.
She was obsessed with a childhood experience, and endlessly and repeatedly told me
about it. A cruel gardener caught a horrible rat in a frightful trap full of teeth. She also
hated her analyst, Dr R, and never wanted to see him again. I tried to talk to her about her
feeling trapped in her analysis, and she said yes, Dr R was the cruel gardener with the
trap. But when I said that she felt then like a rat in a trap she immediately countered that
Dr R was the rat. Whichever way I tried to approach it I could not establish contact with
her. Today I would have understood, theoretically at least, that she was telling me that the
only relation between the container and the contained can be mutual cruelty and
destruction. I might have said that the only way she could visualize Dr R or myself and
her coming together was in mutually attacking one another.
We frequently see patients, even non-psychotic patients, who experience any coming
together as mutually destructive. I think the appearance of bizarre objects described by
Bion belongs to this general category of relations between the container and the
contained. In this case the container itself is split and, instead of the infant or patient
experiencing that fragmented projections are contained by a container capable of standing
them and the anxiety they produce, and bringing them together, the split bits of the
container become part of persecution. The claustrophobic and agoraphobic anxiety
frequently present in borderline patients can often be related to a bad relation between the
container and the contained. The patient longs to be inside a room or a place representing
mother, but that container is felt to have the projected characteristics of greed, envy or
cruelty, and the claustrophobic anxiety forces the patient out, only to be confronted by
agoraphobic anxiety of falling into a void—uncontained and disintegrating.
 Dream, phantasy and art 42

Bad relations between the container and the contained may be due to deficiencies in
the maternal response, or they may lie more in the child himself. The failure in the
mother may lie in her incapacity to tolerate the infant’s projections. She may respond
with hostility, by falling to pieces herself, or by blocking any projections. Sometimes the
experience of a mother blocking projections or falling into pieces is part of the same
pattern.
My Patient E was hard and brittle. At times she seemed so defended and hard on
others that nothing could reach her; at other times she would collapse and feel that she
was disintegrating under apparently minor provocations. She also suffered from great
mental blocks. She often thought of herself as mentally deficient. She came to a session
very distressed because at work she had been mildly rebuked for her inattention. Her boss
had told her she was so preoccupied that at times it seemed as though she had no room in
her mind for anything to do with her work. She then added, in a very distressed way, that
she was the same with her children. While she was looking after her toddler he climbed
on a ladder and fell, and it was as though she was not there, as though, ‘in the words of
my boss’, she said, she had ‘no room for him in her mind’. She also cannot spend much
time playing with them or she feels they crowd her in. We have often seen with her how
persecuted she feels by any demand when she is withdrawn. In the next session she was
complaining bitterly about the lack of parking near my consulting-room. She was also
very preoccupied with my recently having lost weight. She thought I had a heart
condition. It appeared later that she was frightened of telling me anything painful or
shocking because I might have a heart attack. This contrasted with her usual view that I
was completely invulnerable. She told me she valued this view of me as invulnerable, and
it was the only thing that made her analysis possible, since she did not feel me to be as
fragile as her mother was. But she reluctantly agreed with my comment that this total
invulnerability meant to her that talking to me was like beating her head against a wall, as
though there was no parking room in my mind for taking in and feeling anything about
her concerns. She had to maintain the idea of my invulnerability, however, because the
alternative was that if she reached my heart I would collapse, and it might kill me. It
seemed to me that much of what she was talking about in previous sessions—her own
alterations between states of extreme hardness and total persecuted collapse and
disintegration, was an identification with a mother experienced as narcissistic, unable to
stand demands, criticisms, or aggression, and at the same time extremely vulnerable. It
was guilt in particular, my patient thought, that her mother could not stand, and failure as
an analyst with her is what she thinks I cannot stand, and yet this is what she relentlessly
inflicts on me. Should I not be invulnerable, I would undoubtedly collapse and die.
In the next session she came very early and rang the bell wrongly (patients are
supposed to ring my bell twice). When she came into the consulting-room she felt
anxious about being so intrusive, being the first patient in the morning, and having rung
the bell incorrectly, but not nearly as anxious as she used to be in the past. And she said,
with great satisfaction, that coming early she found plenty of room to park. It seemed to
me that the analysis of the previous days had given her some feeling that I could give
parking space to her intrusions without either rejecting her or collapsing. (She could in
fact be extremely intrusive.) E, in her material, shows her own experience as a child, but
she also identifies with her mother and shows the state of mind of a mother incapable of
tolerating the infant’s projections.
 Mental space and elements of symbolism 43

In any situation when the child’s projections are not well responded to, the child feels
the return of his projections as made worse rather than better. In most cases both external
and internal factors contribute to faulty development. On the part of the child, the
interference comes from excessive envy. The child cannot tolerate his dependence on the
maternal containment, and destroys it in his mind, with results at first sight not very
different from those of an actual maternal failure. On many occasions, when E felt that I
could tolerate and understand the various states of mind she could provoke in me, and
interpret that to her in a way that made her feel contained, there was an initial great relief
and improvement, promptly followed by envious attacks, at times quite annihilating the
work done.
The clinical and theoretical implications of the understanding of this phenomenon are
very wide, but I want to concentrate on that aspect of the concept of the relation between
the container and the contained and the beta and alpha elements and the alpha function
which are directly relevant to my theme of phantasy, symbolism, and dreams. The beta
elements as described by Bion seem very close to what I called concrete symbolic
equations. One could think that those equations are formed by beta elements. On the
other hand, and I incline to that view, one could consider that the concrete equation is a
transitional stage between the beta and the alpha elements. I think that, however concrete
they are, they have certain qualities lacking in beta elements. For instance, they are not
always minutely fragmented and are therefore more recognizable; their significance is
more easily detected, and they have some symbolic meaning. The alpha elements,
according to Bion, are elements of dream thought, myth, and symbolism, and I think that
the alpha function is closely related to the symbolic function. Bion describes beta
elements as ‘saturated’ through; alpha elements, on the other hand are ‘unsaturated’.
They are open to various ‘realizations’, in Bion’s terms, or reality-testing and therefore
open to many and varied transformations (hence there may be many symbols for one
object or quality and, conversely a symbol may have many meanings). They lend
themselves also to generalization, abstraction, and differentiation, a theme developed by
Money-Kyrle (1968), in his paper ‘On cognitive development’.
I have related the move from concrete symbolization to symbol proper to the move
between the paranoid-schizoid and depressive positions. Bion considers the question of
whether the move from beta to alpha precedes the depressive position, or whether it is the
result of the shift. On the one hand, one could assume that there must be an alpha-
function type of mental apparatus to be able to bear the depressive position, On the other
hand, one could argue that the alpha elements could not be formed outside the depressive
position. Bion speaks of a benign interchange in normal projective identification.
However, one must take into account that the identification part of projective
identification would preclude the kind of benign interchange that he describes. If the
container becomes completely identified with the projected part, it disappears. A
complete identification between the container and the contained probably occurs because
of envy and the inability to tolerate the dependence on the container, and complete
identification precludes the experience of feeling contained. This was very frequent with
Patient E. After a good experience with me she could for a time become a perfect tolerant
mother, but all link with me would disappear, and soon the identification would collapse.
It seems to me, and I think this is a view to which Bion inclines, that the move from beta
to alpha and from paranoid schizoid to depressive are conjoined phenomena which are
 Dream, phantasy and art 44

interdependent. Some depressive dependence must be acknowledged for the experience

of being contained to be admitted at all, but the intemalization of that experience in turn
facilitates the toleration of depressive anxiety. Thus the formation of alpha elements, and
eventually proper symbols, are part and parcel of evolution in the depressive position.
Bion approaches separately what he calls the formation of the mental apparatus, which
he links with the container-contained relationship and that of the occurrence of thoughts
which have to be dealt with by that apparatus. He relates thought, as I have related
symbolism, to the capacity to recognize and experience absence. To begin with, the lack
of the breast is felt as a bad breast inside (hence my patient’s conviction that he could get
rid of hunger by defecating). It is only when the infant can recognize the absence of the
object that he can either symbolize or think. Bion (1970) described it succinctly: ‘No
breast—therefore a thought’. However, to be capable of having such an experience there
must be a part of the mind that can contain the anxiety of missing an object, a ‘no-breast’.
The container-contained relationship must have given rise to the creation of this part of
the mental apparatus, without which even the thought ‘no-breast’ could not be formed.
According to Bion, the container-contained relationship determines the mental
apparatus. He does not seem to link it with the concept of mental space, even though the
container is essentially a spatial concept. I find it useful to think of it also as a way in
which the mental space is formed and our experience of our minds, as Wollheim (1969)
has pointed out, is always spatially tinged.
Using the words ‘mental space’ sometimes leads to confusion with Winnicott’s
concept of ‘potential space’ (Winnicott 1971), sometimes referred to as ‘transitional
space’. They are very different. For Winnicott the relevant space is the space between
mother and child, which, if not intruded upon by the mother, is the space in which
transitional phenomena develop and which becomes the cultural space. Bion’s container
or space is the result of interaction of projective and introjective identifications, and it is
not just a neutral space but an active container with the capacity to perform the alpha
function. This space is not between mother and child; it is an internal mental space
formed by the introjection of a breast capable of containing the infant’s projective
identifications and giving them meaning.
Bion’s concept of the container and the contained is strictly a two-body relationship,
relating to the earliest relationship between the infant and the breast. However, if we
think of the beginnings of ‘no breast therefore a thought’ or of the beginnings of
symbolic thinking, and the shift between beta and alpha being a conjoint phenomenon
with the depressive position, we have to ask ourselves about the status of the third object.
It is implicit in the depressive position that the perception of a mother as a whole person
implies the beginning of a recognition of her having a whole, separate life unrelated to
the infant, primarily a relationship to the father. In fact, in his later work, Bion does refer
to the place of the third object in relation to the relationship to the container and the
contained.
In ‘Attention and interpretation’, Bion (1970) says that when there is a good
relationship between container and contained it gives rise to a third object in a way that
two objects share a third to the advantage of all three. In contrast, a bad relationship
between container and contained gives rise to a third which is destructive to all three. I
think that in the creation of such a bad object splitting also plays a part. In order to
preserve the good relationship with the primary object, the breast, the infant splits off the
 Mental space and elements of symbolism 45

bad relationship and projects it on to a third. I have described (Segal 1964:57) a ‘third
area’ due to this kind of splitting. In the chapter on the psychopathology of the paranoid-
schizoid position, I speak of a borderline patient who said, ‘Here is my head on the
pillow, and there you are in your armchair, but between the top of my head and you there
is nothing but horrible bloody mish-mash.’ The other patient I describe in the chapter was
a hebephrenic schizophrenic. In both those patients there was a third area split off from
both patient and analyst containing bad fragments. The father, or more primitively the
father’s penis, easily becomes the ideal recipient of such projections.
My patient C, whom I described in Chapter 3 (pages 39–40), who suffered from
transitory hallucinations, one day hallucinated a motor-cyclist driving in his head. He
maintained a good relationship with me as mother by splitting off his hostile intrusive self
and the bad part of me into a third object. He felt that as a child he could never be at
peace alone with mother because of an intrusive elder sibling, and according to his
mother his breastfeeding was also disturbed by his father’s demands on her. The
hallucination itself was due to a splitting process in which the bad, mutually intrusive
relationship between himself and me was split off into a hallucination.
The appearance of a third object is the beginning of the realization of the existence of
the father in the Oedipal triangle. R Britton made an interesting extension of Bion’s
concept of the container and the contained to include the relationship with the father:

The acknowledgement by the child of the parents’ relationship with each

other unites his psychic world, limiting it to one world shared with his two
parents, in which different object relationships can exist. The closure of
the Oedipal triangle by the recognition of the link joining the parents
provides a limiting boundary for the internal world. It creates what I call
‘triangular space’, i.e., a space bounded by the three persons of the
Oedipus situation and of their potential relationships.
(Britton 1989:86)

Britton, like myself, sees the container as being also related to mental space. This new
space has new characteristics. It contains three kinds of possible relationships, the
vertices of the triangle: the relationship between the mother and the child; that between
the father and the child and that between the parents from which the child is excluded.
Each of these relationships can be seen as mutually beneficial, the third—excluded—not
being necessarily a hostile entity, unlike the father and sibling of my Patient C, or another
patient, who dreamed that a man (obviously related to his father) comes into the kitchen
to steal all the food. This extension of the mental space is crucial for the perception of a
variety of relationships, not mutually exclusive or necessarily hostile to one another. The
third can become an objective and/or benevolent observer. This observing part is a
necessary feature in mental life for the existence of insight, benevolent curiosity; it is the
basis of a constructive, epistemophilic attitude.
Two dreams of a patient in an advanced stage of her analysis illustrate some features
of this triangular space. The first dream follows an event in the transference. I asked the
patient, K, if she could come ten minutes earlier for her session. She gave me apparently
good reasons for not being able to change her time. As I had unavoidably to leave my
consulting room at the earlier time I had to tell her that I was sorry, but I would have to
 Dream, phantasy and art 46

cut her session ten minutes short. It soon transpired that in such a case she could think of
ways of coming ten minutes early. The next day she had the following dream. She was
walking down a lovely road bordered by leafy trees. But she came to the end of the road.
There was no way to go on. She retraced her steps, and on the way she saw a clearing,
and in that clearing a couple were having very vigorous sexual intercourse. The
intercourse was not only physical. The man seemed to have been telling the woman how
passionately he loved her. She observed this scene with great interest. Her first
association was to tell me that there was nothing voyeuristic or exhibitionistic about the
scene. The couple were making love in a clearing in a forest. They were concerned with
one another, not exhibiting. She had no voyeuristic excitement. She did not seek to see
them, unlike in some other dreams, and she was just interested, not excited. She
associated the end of the road to my firmly telling her that I wished to end the session ten
minutes early.
I think that the leafy road which she enjoyed so much represents her phantasy of being
inside me controlling me, a very common psychic stance in her. The frustration about the
end of the session shows her that this has come to an end. She sometimes uses the
expression ‘No way!’. ‘No way’ can she maintain the phantasy of being inside and
controlling me. When she gives up that phantasy she is confronted by another space
which opens a clearing in which the parental intercourse happens. And in the dream it is
unspoilt by voyeuristic projections.
Further work of course led to the recognition of repressed feelings of jealousy and
rage, and she had to make room in her mind for those feelings as well. But it is important
that she could be also an objective and appreciative observer.
Some time later she had a similar dream, bringing another aspect of the new space and
a new way of relating. In the dream she was in the kitchen with me, washing up. We were
talking about marriage, and I was telling her that I married a Pole because Poles are so
passionate. In that dream her relation to me in terms of feeding and washing up is not
spoilt, and is even enriched, by her knowledge of me as part of the couple. Similarly, her
relation to me as mother does not disturb her own appreciation of father.
A patient of mine, F, is tormented by recurring panics. Though intelligent, when in a
state of panic, whether conscious or unconscious, she loses her concentration and all
ability to think. In one of the sessions she described acute anxieties about a hole in a wall.
Some building work had to be done in her house; she could not imagine it being done
without a big hole in the wall. She thought her children could fall through the hole, or
herself, and anyway the thought of this big black hole put her in a panic. It was fairly
clear in the session that she had no idea in her mind that the builder might know what he
is doing. (The transference implications were obvious.) When this was interpreted she
told me that she had another panic which assailed her on the way to the session: she had
lost her diary which had confidential information about herself and the work in her office,
involving other people, and the loss of the diary could be catastrophic. It slowly emerged
that the loss of the diary had to do with her having read a poster of a lecture I was giving,
jointly with a man. She did not intend to attend this event, but she also wanted to forget
the date, not to be troubled by it. The indiscretion referred to was what she felt was the
indecency of my exhibiting myself on a platform with the man.
The next session was very difficult, and analytic work was hardly possible because of
the patient’s determination to break any links. Two themes seem to have emerged: one of
 Mental space and elements of symbolism 47

an internal tormentor. It eventually led to a phantasy of a hand cruelly squeezing out a

breast that was filled with a boil. The second theme referred to sexuality. She felt that I
and others were trying to impose on her a view of sexual intercourse as something
pleasurable and non-destructive, which was a denial, and idealization, since ‘I thought
everybody knew that sex was always sado-masochistic’. Another view she had of sex was
as of two people coming together in intercourse in cruelty to the third person in front of
whom they were exhibiting, so as to inflict on this third person unbearable feelings of
exclusion, inferiority, and jealousy. Some links could be established—for instance, the
parents-myself cruelly exhibiting to the child and the resulting destruction of her
relationship to the breast; but the patient remained on the whole disconnected from me,
angry, and very anxious.
In the third session, as she was speaking of the hole, some music from the
neighbouring flat could just be heard in the consulting-room. She drew my attention to it,
and said it made her think of people dancing. She did not feel disturbed or persecuted by
it. Eventually we could establish that the terrible hole in her mind was the missing space
in which she could perceive the parents, represented by me, dancing together. Later in the
session, she started speaking about her difficulties in writing a report she had to do at
work, but in a much quieter and understanding way. This patient had an idealizing
relationship to the breast, but this relationship was very fragile, only to be maintained by
a splitting off of ambivalence. The appearance of a third object turns everything into
torment. Not only is she tormented by phantasies of sexuality infused with cruelty, but
the original relation to the breast is destroyed by this intrusion and also turns into
torment. When this intrusion happens her mind becomes itself disjointed and fragmented,
and she becomes filled with intense paranoid and hypochondriacal phantasies. At those
points her capacity for understanding and symbolization gets lost. My talking to others is
experienced as my forcing her to witness an actual sexual intercourse. The bad feelings
about me become a tormented and tormenting, poisonous breast inside her, experienced
in her physical symptoms. She can get rid of the resulting persecution by making a hole
in her mind instead of a space accommodating the dancing parents.
The fear experienced by the patient, F, of the hole in the wall is similar to the fear in
the dream of Patient D described earlier, in which she brings me her complaints as bits
and pieces falling out of her body, among them something that could be children. She
brings her complaints as bits and pieces falling out of a hole in her body. What she
represented at the beginning of her dream is very similar to Patient D’s fear about the
hole in the wall through which the children would fall out. Though later in the dream D
shows how the containment and understanding provided by the analytic situation led to a
transformation of beta into alpha elements. Patient F shows the change in the third
session. Another patient, less severely disturbed, exemplified clearly the effect on
thinking of a difficulty in establishing the triangle.
She started the session by telling me that she had two very tiny fragments of a dream.
In one, she saw me surrounded by middle-aged, stupid, altogether despicable men. Of the
second fragment she could remember only that it had something to do with African land
and African people. The first dream seemed to both of us pretty obvious, with an
impending long weekend. But the hardly remembered, fragment of the dream brought
surprisingly rich associations. To begin with, she expressed again her horror of racial
prejudices from which she cannot free herself, and which she detests in herself. That
 Dream, phantasy and art 48

seemed to provide a link between the dream of men with whom I may spend my weekend
and the one about the Africans. But her other associations were more unexpected and
illuminating. The patient is a form teacher, and she started speaking about a child’s
difficulty in learning any grammar, particularly foreign grammar. She thought that Africa
might represent what this girl feels about foreign grammar as totally exotic and
incomprehensible. This child, she said, is quite clever, but very disjointed. She seems
unable to make certain connections, and this seems particularly obvious in her total
inability to grasp the rules of grammar. After all, grammar, with the sort of patterns it
describes, should come more naturally. Then she laughed, and said, ‘Maybe to her
grammar is so foreign and exotic—just like parents in intercourse must appear to the
child—beyond reach, incomprehensible, exotic, foreign.’ This patient is often
preoccupied and disturbed by very primitive fantasies of the primal scene. In this session
she seems to feel that there are certain natural patterns of interrelationships, as in
grammar (shades of Chomsky!), and that this includes an intuitive awareness of parental
intercourse.
In the first dream that intercourse is attacked and derided. She has a prejudice against
it, like a prejudice against Africans. And her associations to it suggest that she is aware
how her thinking is dislocated by her attack on those natural patterns of relationships.
I think that the triangular space is also like the space for a new baby. It is not
‘saturated’ by the mutual projections between the mother and child. As a new,
unsaturated, mental space it allows for the birth of new thoughts: two can come together
like parents in intercourse to produce a third new thought.
The various considerations I have brought in about the relationship between the
container and the contained, and extension of that concept to include the mental space
which accommodates the parents together, and the potential new baby, seem to be far
removed from my general theme of dreams, phantasy, and symbolism. I think, however,
that they are crucial in the understanding of the mental processes involved in the
formation of phantasy and dream thoughts and the malfunctions of the mental apparatus
which account for its pathology.

Notes
1 Patient described in ‘Early infantile development as reflected in the psychoanalytical process:
steps in integration’ (Segal 1982).
 5
The dream and the ego

In previous chapters I put forward the idea that a dream is a way of expressing and
elaborating an unconscious phantasy. What Freud called a ‘compromise’ takes the form
of a wish-fulfilling phantasy, satisfying contradictory wishes and defences. That phantasy
is expressed and further worked through in the dream. A dream performs many functions.
It provides a phantasy expression of an unconscious conflict and seeks a phantasy
solution—wish-fulfilment. Dream-work is part of the elaboration of unconscious conflict.
It provides also an intrapsychic communication between the unconscious and the
conscious. When we remember a dream we retain communication with symbolic
expressions of the unconscious. In the analytical process this internal communication
becomes also a means of communication with the analyst.
In Chapter 1 I commented on how great is the task of the ego in accomplishing
psychic work—dream-work—involved in dreaming. And I raised the question of what
happens to dreaming and dreams when the ego is temporarily or permanently not capable
of carrying out the tasks involved in producing a neurotic or a normal dream. In the acute
psychotic there is often no distinction between a hallucination, a dream, and a real event.
I remember sessions early in the treatment of an acute schizophrenic, when he would give
me a jumbled account of his night in which it was impossible to tell what really
happened, what he hallucinated whilst awake, or what he hallucinated whilst asleep. Bion
describes a patient who was terrified because he had dreamed of his analyst and therefore
concluded that he must have devoured him, and was amazed to find his analyst alive in
the real external world. In such cases, the patient is unable to differentiate between the
psychic event which is a dream and actual events in the external world.
In the less acute cases in our daily practice, we find patients whose dreams
occasionally or habitually do not fulfil the dream function as described by Freud. The
dreams of these patients serve the function not of elaborating and symbolizing latent
dream thoughts, but the function of getting rid of psychic content (Bion 1958). In the
analysis they serve predominantly not for communication but for acting-in. That is, such
dreams function not as symbolic communication but as symbolic equations, or beta
elements which are expelled. Dreaming is thus felt as an expulsion and is sometimes
equated with actual urination or defecation. In the session the patient projects into the
analyst with various consequences. Though the aim of such projections is not usually
intended as a communication, if the analyst can understand them they become a
communication as well. Mr H1 dreamed profusely. He had numerous notebooks in which
he would write down his dreams and associations. He would report them to me
sometimes on the following day, but often weeks or months later he would produce the
notebook and start reading from it. Mr H was a past master at getting rid of his psychic
life. For instance, after the session he would spend hours in the lavatory defecating and
carrying out what he called his ‘post-analysis’, which was his way of getting rid of any
 Dream, phantasy and art 50

feeling or insight he might have experienced in the session. His dreams served a similar
function. Whatever touched him could give rise to a dream which would be defecated
into his little notebook and in that way disposed of. In his case it was often the more
insightful part of himself that was thus disposed of. This was particularly marked after
the death of his mother, when he had several dreams expressing his mourning. The
dreams were duly recorded and forgotten, reported to me weeks later, whilst in his
conscious life he seemed untouched by mourning. In An Outline of Psycho-analysis,
describing the splitting in the ego in psychosis, Freud (1940:202) speaks of a psychotic
patient in whom the split-off, more sane part of the ego was functioning whilst dreaming.
Borderline patients like Mr H differ from those psychotics who cannot differentiate at
all between dream, hallucination, and reality; they know they have dreamed, they speak
of their dreams, but in spite of that their dreams are psychically equivalent to concrete
events, and they have similar consequences. Miss G2 would start a session complaining
that my room smelled of gas. We would find subsequently that she had dreamed of gas
balloons exploding. If, in her dream, I or a character standing for me persecuted her, she
would upbraid me for it, as though the character in the dream was actually myself. She
had a similar waking phenomenon. She had numerous fantasies which she called
‘fantasies’, but she lived them as though they were hallucinations. For instance, she
walked very awkwardly because she had a phantasy faecal penis stuck in her vagina or in
her anus. What she called a fantasy was nearer to a psychosomatic hallucination.
When patients use dreaming for evacuating mental contents, the process can take
many forms. Sometimes the dreaming itself is felt to accomplish the evacuation, or the
recording of it, as in the case of Mr H’s notebooks. Very often it is dreaming and telling
the dream to the analyst that accomplishes the evacuation. The telling of the dream may
be devised to rouse feelings in the analyst and thereby accomplish a projective
identification. Mr H sometimes told me dreams of torture or terror which shook me,
whilst he seemed to have no feelings about them whatsoever. Also, some of the dreams
about his mother’s death could evoke in me a feeling of being very moved, which
contrasted with a total lack of feeling in himself. The evacuation is into an object, and in
the analytic situation the object is the analyst. The aim is twofold: one, to split off and get
rid of certain psychic contents; and two, to affect the object. For Mr H, the getting rid of
content was usually the primary aim and the way it affected his object often secondary.
In Mr M, on the contrary, the aim of affecting the object was the more prominent. At
the beginning of his analysis he used to flood me with dreams and numerous associations.
When I pointed out to him his inability to listen to the interpretations and the way his
numerous associations, with which he interspersed the telling of the dream, were in fact
obscuring the meaning, rather than helping me to clarify the dream, he was very shocked
and surprised. He said he thought that was how one analysed dreams—‘Isn’t that how
Freud analysed his Irma dream?’ It soon became apparent that he was being Freud,
analysing his own dreams with me as his dazzled and mesmerized audience. After some
analysis of this meaning of his way of dreaming and recounting his dreams, he produced
a few shorter dreams with more coherent and pertinent associations. Those dreams, and
the associations to them, dealt with and illuminated the psychic function of his dreaming
and recounting the dreams.
He dreamed that he was injecting anti-rabies serum into a big football. It was
important that not a drop should touch him. He was also carelessly putting something
 The dream and the ego 51

into a woman’s purse. He associated to his promiscuity and seduction of women—an

activity which he wished to see as reparative (the serum)—but which was, in fact,
destructive and projective. He had a conviction that when he approached a woman, he
had the power to implant in her what he called ‘the need for M’, thus projecting his own
infantile need into the woman. Once so infected, only intercourse with M (the serum)
could save them. But the injection by his penis was re-infecting them—getting them more
addicted to him. The serum in the dream is destructive and not a drop must touch him.
This, of course, is also a reference to analysis. He injects his dreams into me, but not a
drop of an interpretation must touch him.
The next day the situation became even clearer. He washed his hair just before the
session and literally soaked my pillow. He had a dream of piercing a balloon and making
it burst. ‘It popped like a cherry.’ He associated this dream with the previous one of
injecting the football. It had become clear that for him his profuse dreams and his way of
telling them to me were like a stream of sexualized urine, meant to dazzle, seduce,
confuse me, and make my mind burst. He was motivated both by his need to get rid of his
own infantile needs and by envious rivalry with me. He was Freud in the sessions. (Note
also the theme of injection continuing the identification with the Irma dream.)
Sometime later he brought four dreams. I shall not give them in full, since the telling
of dreams and the associations took nearly the whole session. The first dream referred to
masturbation; the second to an injunction that one must not fuck one’s daughter, the
fourth one referred to Othello’s killing of Desdemona. But it is the third dream which is
relevant to my theme. He dreamed that he was putting tiny fragments of something into
two oval shapes which became warped. He was also trying to bring them together. He
interrupted himself to associate that the shape reminded him of a grapefruit, since later in
the dream he was buying a grapefruit from a woman and was annoyed that the price had
been raised. (His fee had been raised shortly before this.) He thought the small fragments
were his disintegrated dreams and that he hoped that putting them in the oval shapes and
bringing the shapes together meant he was integrating them. He used an expression
typical of him—‘I want it to mean that’—then he continued the dream. He was running
in a transparent warped passage. There was a supervisor at the entrance who might not
let him into the place he was running to. Then came the grapefruit part. Possibly he had
to buy a grapefruit from a woman to get in.
I think that in this dream it is the warped grapefruit and the warped corridor that are
the essential clues. The other three dreams contained fragments of a fragmented and
projected Oedipal situation—for instance, projecting his Oedipal feelings into his
daughter. In the third dream he shows how putting his dreams into me—the oval shapes,
the breasts, the grapefruit he has to buy—he is both projecting a fragmented Oedipal
situation into me and warping my judgement. His ‘I want it to mean’ is also his wanting
me to interpret the dream’s meaning to be what he wants it to be, so that I would let him
‘into the place’. Dreaming and telling me dreams were experienced by him as an actual
urinary intercourse in which he both projected parts of himself into me and wished to
affect my mind—burst it, as in the previous dreams—or warp it. His conviction that he
could affect my mind that way led to his tendency to experience his analysis as a folie à
deux, reproducing his experience with his mother, who erotized her relationship with him
and idolized him in a virtually delusional way. Dreams were, of course, only one of the
manifestations of this process, but a very central one.
 Dream, phantasy and art 52

I also think that the two oval shapes were also his own lungs, since the re-
internalization of the destroyed breasts was, I think, at the root of the asthma from which
he occasionally suffered.
When people experience dreams as concrete events or objects, a stool, anal gas, or a
stream of urine, which are expelled into an object, their reality perceptions are
unavoidably affected. I have described how Miss G would complain that my room
smelled of gas, and only later we would find out that she had a dream in which a gas
balloon exploded. When she had a quarrel in a dream, I was perceived as a quarrelsome
person. The dream obviously spilled into her perception of reality. An example of such
misperceptions was reported to me by a colleague*, in a patient I was supervising.
The patient started the session by saying that he was very worried because he thought
his analyst had shaved his beard, but he could not be sure. Later in the session he said that
in a certain light he could see his own face reflected in the lamp overhead. Other
associations dealt with his intrusiveness into the analyst’s mind and his fear of getting
confused with the analyst. At that point the analyst drew his attention to the fact that if he
could see his own reflection in the lamp he could also see the reflection of the analyst’s
face. The patient was amazed. He looked again and said he had never noticed it before.
He only saw his own face. But he still could not see whether the analyst had shaved his
beard. The patient’s projective identification, seeing only his face where he should have
seen both, prompted the analyst to ask whether the patient himself had thought of shaving
his beard. ‘No, he hadn’t,’ but he suddenly remembered that he had dreamed the previous
night that he looked into the mirror and that his own beard was shaved. It seems that his
dream was completely expelled into the analyst with the effect that he perceived the
analyst’s face the way his own face was in the dream.
This case shows a striking sensory misperception. My patients—Mr H, Mr M, Miss
G—showed similar occasional misperceptions. But the distortion of the perception on the
emotional level, like Mr M’s experience of the folie à deux, is invariably present. In
addition, all the dreams of the kind I have reported are also to a large extent acted out in
the session: Mr M’s rapid confusing and confused way of reporting dreams, his wetting
the pillow, Miss G’s quarrels with me, and so on.
In certain patients, however, or in some patients at some times, the
*
Dr R.Britton.

acting-in is the most prominent feature. The dream has to be acted out in detail in the
session. For instance, Mr H had numerous dreams in which he would be late for a meal or
a meeting, a precise number of minutes—anything from two to forty-five—and then he
would come late to the session by exactly the same number of minutes.
A patient whose analysis I supervised† showed a similar mechanism. In a recent
session he took half an hour on procrastinating, empty associations. Then he produced a
dream in which he arrived at a zebra crossing at 8.20. There was a policeman and an
incident. He did not want to see the incident and walked away. The analyst glanced at her
watch and it was precisely 8.20. Then the patient let his associations wander away from
the dream. As in the dream, he wandered away from the incident he did not want to see.
The session as such was a re-enactment of the dream.
 The dream and the ego 53

Miss G had a number of dreams of paranoid quarrels with thinly veiled parental
figures. When this happened she would start a kind of quarrel with me in the session.
Whether I was silent or whatever I said would be interpreted by her as quarrelling. I felt
like a puppet in someone else’s nightmare. Eventually we would find that she had had a
dream and was acting it in the session. In these cases I have a feeling that the dream
content has to be evacuated by a very precise re-enactment.
I have called this kind of dream a ‘predictive dream’, because it seems to predict
future happenings as they are almost automatically acted out. They function like a
‘definitory hypothesis’, described by Bion (1963). Such dreams are often also acted out
outside the session. The more usual analytic experience is that when a patient brings a
dream and we deal with it in the session he is less compelled to act out. It is not so with
predictive dreams. These seem to be more like blueprints for future acting out, both
inside and outside the session. This was particularly striking with Mr H. If he brought a
dream indicating a wish for a homosexual acting out as a reaction to the weekend break,
no analysis of the dream would prevent his carrying out the action almost precisely as
planned in the dream. It could of course be the effect of the analysis being incorrect or
insufficient. But I also had the impression that no analytic understanding could contain
the powerful urge to get rid of the dream content by acting it out precisely. Of course all
dreams are to a certain extent acted in and out. Or rather, the phantasy expressed in the
dream is to a certain extent acted in and out in what Rosenfeld (1964b) described as
‘normal acting out’. However, the dreams I am speaking about in this chapter are
†
Analyst, Dr Piontelli.

predominantly used for acting in and out, and in the case of predictive dreams this is
particularly compulsive and often uncontainable in the session.
Those dreams in which the dream-work has partially or completely failed are
characterized by interrelated phenomena. One, which is not always present, is a certain
crudity in symbolization. For instance, Miss G would regularly dream of houses where
there was a restaurant on the first floor, a gas boiler or a furnace in the basement at the
back, and scenes of sexual violence or murder in the basement at the front. Or she would
dream of two wasps in sexual intercourse which consisted of urinating or defecating into
one another, and so on. Mr H, who hunted guardsmen for sado-masochistic sex, would
dream of a guardsman pursuing him with a red-hot poker. It is as though the barest
minimum of effort went into symbolization. Other features are constant: the experience
of the dream as a concrete event, the use of the dream for the expulsion of psychic
content, and the associated blurring of boundaries between internal and external worlds.
The burst balloon of Miss G is not felt as an inner mental event. It invades her perception
of my room. It is equally clear in the case reported by Dr Britton: the dream is not even
remembered, it becomes part of the analyst’s face.
Freud speaks of the dream as a harmless psychosis and also as of a hallucination
happening during sleep. I do not think that this applies to most dreams. I think that
hallucination is a pathological process linked with pathological projective identification.
The kind of dreams that I am describing here are indeed like hallucinations happening
during sleep. But in the normal dream, where the process of repression, symbolization,
and working-through is achieved by the dream-work, the dreams, even though they are
 Dream, phantasy and art 54

pictorial representations of a phantasy, do not have the psychic function of a

hallucination.
The benign interplay between projection and introjection, which occurs when there is
a constructive relationship between container and contained, is at the very basis of the
mental functioning, including dreaming. The failure of a good relation to an internal
container leads to the concretization of mental events.3
An example comes from an unusually gifted and able man who has a constant struggle
with psychotic parts of his personality.4 We ended a Friday session with the patient
expressing enormous relief and telling me that everything in that session had a good
resonance in him. On the following Monday, he came to his session very disturbed. He
said he had had a very good afternoon’s work on Friday and Saturday morning, but he
had had a dream on Saturday which had worried him very much. In the first part of the
dream, he was with Mrs Small. She was in bed and he was either teaching or treating her.
There was also a little girl (here he became rather evasive), well, maybe a young girl.
She was very pleasant with him, maybe a little sexy. And then quite suddenly someone
removed a food trolley and a big cello from the room. He woke up frightened. He said it
was not the first part of the dream that frightened him but the second. He felt it had
something to do with a loss of internal structure. On Sunday he could still work, but he
felt his work lacked depth and resonance, and he felt something was going very wrong. In
the middle of Sunday night he woke up with a dream, but he could not hold on to it and
instead became aware of a backache low in his back—maybe the small of his back.
He said that the Mrs Small part of the dream did not disturb him because he could
quickly see through it. In the past, Mrs Small, whom he does not think much of,
represented a belittling of Mrs Klein (klein = small). He understood that and supposed
she represented me changed into a patient and also into a sexy little girl. He supposed it
was an envious attack, because on Friday he felt so helped by me. He then had some
associations to the cello—his niece had one, his admiration for Casals, and a few
others—which led me to suggest tentatively that it seemed to be a very bisexual
instrument. That interpretation fell rather flat. What struck him more, he said, was that it
is one of the biggest musical instruments around. He then said that I had a very deep
voice, and that another thing that frightened him was that when he woke up from the
dream he could not remember what we were talking about in the Friday session.
It seems to me that the whole situation, which in the first night was represented by the
dream, in the second night happened concretely. By changing me into Mrs Small, he had
lost me as the internalized organ with deep resonance. The cello represented the mother
with deep resonance, the mother who could contain the patient’s projections and give a
good resonance; with the loss of this organ there was an immediate concretization of the
situation. On Saturday night, he belittled me, as is shown by his changing me into Mrs
Small in his dream. This led to the loss of the cello—‘one of the biggest musical
instruments around’—and the disappearance of the food trolley. He woke up anxious.
The function of the dream to contain and elaborate anxiety began to fail. The next night,
instead of a dream, he had a pain in the small of his back, a concretization of Mrs Small.
Hypochondriasis, much lessened now, had at one time been a leading psychotic-
flavoured symptom. The attack on the containing functions of the analyst, represented as
the organ with the resonance, resulted in the patient’s losing his own resonance (his depth
of understanding) and his memory (he could not remember the session). When this
 The dream and the ego 55

happened, he could only experience concrete physical symptoms. The belittled analyst,
who in the dream was represented by Mrs Small, became a concrete pain in the small of
his back.
There is a technical problem in analysing those dreams in which the dream-work is
defective. It is useless in those cases to try to interpret only the content of the dreams. In
recent years some analysts have expressed a certain pessimism about the usefulness of
dreams in analysis. I think this is the effect of trying to analyse the content of the dream
in a classical way in relation to dreams which may have to be tackled differently. In the
case of Mr H, analysing the content of his dreams would make him very happy and
excited, and have no therapeutic effect. Miss G would be likely to feel infinitely
persecuted, taking every interpretation as a forcible pushing down her throat of her
projective identifications. Mr M would experience it as our joint sexual game, with
marked anti-therapeutic effect. In fact, in the beginning of his analysis with me he would
get very manic and sexually excited if I started interpreting the content of his dreams.
Some patients come flooding the analyst with dreams and confusion. In such a case the
first thing to interpret is the flooding and the effect it is supposed to have on the analyst.
Generally, in these dreams, which are primarily used for acting out in the analytic session
it is this function of the dreams that has to be interpreted first of all. Only gradually, and
where it connects with this function, can one address the actual content of the dream. The
analyst’s capacity to contain the projective identifications, understand, and eventually
interpret them, provides a container which restores the mental space and helps to restore
the symbolic function. The process described in the dream of the lost cello can be
reversed. Such a restoration is shown by another patient.
After a long holiday, he lay on the couch silent and seemed very weighed down. After
a long silence, he spoke inaudibly, saying the break was awful. When he got up from the
couch before the holiday he felt as if he was glued to it. He felt almost paralysed. In the
holiday he could not dream. He woke up feeling weighed down by stones. In the daytime
he was like a zombie. From my own counter-transference experience of the heaviness of
his silence, I had no doubt that those were not idle complaints, and he was projecting into
me his experience. Going on my past knowledge of this patient, I interpreted to him that,
unable to face the separation, he glued himself to the couch and left most of himself
inside me, glued to me and confused with me, so he either felt weighed down by a
conglomerate of him and me, or when he cut himself off from this dreadful experience he
became a zombie. He then remembered that right at the beginning of the holiday he had
one nightmare in which a huge animal, a cross between a dinosaur and a rhino, burst
into a shed, bits of the shed sticking to his skin. He woke up in fright and had not
dreamed since. He associated the dinosaur to something very archaic and the rhino to
nosiness, intrusion, and aggression. In the next session he said with relief that dreams had
returned and he could breathe again and write again. He dreamed of a person picking up
a kitten. He thought the kitten would be very dirty, but in fact it was not. In another
dream there was a bag of letters, like letters for Scrabble, only the bag was much bigger.
He was beginning to sort them out. Another dream (which I will not report here) dealt
with separation. It seems that the experience of the previous session converted his
perception of himself as a dinosaur/rhino bursting into the shed, which he breaks, and
with which he gets confused (bits stuck to the skin) into that of a kitten which was not as
dirty—meaning, I think, dangerous—as he thought. The kitten is picked up by a person,
 Dream, phantasy and art 56

so I have also become more human in his perception. The jumble of letters—fragments
projected into the bag—can then be sorted out to form words. (In Bion’s terms, one could
say that his beta elements got converted into alpha elements.) When this happens he can
begin to work through the experience of separation and recover his capacity to dream.
Freud considered the understanding of dreams to be the royal road to the unconscious.
As I suggested, following only the content of the dream has its limitations. If we analyse
not the dream but the dreamer, and take into account the form of the dream, the way it is
recounted, and the function it performs in the session, our understanding is very much
enriched and we can see how the dream’s function throws an important light on the
functioning of the ego.

Notes
1 Mr H has been described in my 1972 paper, ‘A delusional system as a defence against the re-
emergence of a catastrophic situation’, and in ‘The function of dreams’ (1981).
2 Miss G has been described in the same paper (Segal 1981).
3 This patient also is described in ‘The function of dreams’ (Segal 1981).
4 This patient’s dream is also described in ‘The function of dreams’ (Segal 1981).