493

and symptom formation. From all that we know of the structure
of the simpler neuroses of daily life it is highly improbable that a
neurosis can occur only through the objective fact of peril without
participation of the deeper unconscious strata of the psychic
apparatus. In the unconscious, moreover, there is nothing at hand
which can give content to our conception of the destruction of life.
Castration can be visualized through the daily experience of parting
with the intestinal contents and through the loss of the maternal
breast through weaning; but anything like death has never been
experienced, or if so, like a swoon, has left no trace behind which
could be demonstrated. Therefore I hold fast to the supposition
that death-fear is to be comprehended as an analogue of castration
fear and that the situation upon which the ego reacts is the abandon-
ment by the protecting super-ego–the power of fate–whereby the
assurance against all danger has an end. Besides, there comes into
consideration that in the experiences which lead to traumatic neurosis
external protection from irritation is broken through and over-large
quantities of excitation appear in the psychic apparatus; so that the
second possibility is present that not only does fear fail to signalize
the affect but the situation is newly generated from the economic
conditions.

Through the last remark that the ego through its regularly repeated
loss of objects is prepared for castration, we have obtained a new
conception of fear. Having regarded it hitherto as the affect signal
of danger, it now appears to us often in connection with castration
danger as the reaction to a loss, a separation. While many
things may speak against this conclusion, as is at once apparent, a
very noteworthy agreement must strike us. The first fear experience
of mankind is birth, and this signifies objectively separation from
the mother, which may be likened to castration of the mother (accord-
ing to the equation child equals penis). Now it would be very
gratifying if fear should be repeated at every subsequent separation
as a symbol of separation; but unfortunately another valuation of
this agreement stands in the way, for subjectively birth is not experi-
enced as a separation from the mother, for to the narcistic fetus the
mother is quite unknown as a separate object. Another consideration
will be that the affect reaction to a separation is known and that we
feel pain and grief but not anxiety or fear. Indeed, we recall that
we have not understood why grief should be so painful.

VIII

It is time to reflect. We seek obviously for insight into the nature
of fear, for an “either–or” which will separate truth from error.

494

This, however, is difficult, for fear itself is not easily comprehended.
Hitherto we have attained nothing but contradictions between which
no choice is possible. I now propose something different. Without
bias we will assemble everything which may be said of fear and at the
same time give up the expectation of a close synthesis.

Fear or anxiety in the first place is something felt and we call it an
affective state, although we do not know what an affect is. As a feel-
ing it obviously is of the painful type but this does not exhaust its
quality. Not every painful sensation is fear, for there are others
(tensions, pain, grief), so that as a painful quality fear must have
something special. We may ask: Is it possible to comprehend differ-
ences between these various painful affects? From the sensation of
fear we can always deduce something. Its pain has a special note and
while we cannot prove this it is probable and would not be striking.
From this peculiar quality so difficult to isolate we note the fear of
certain somatic sensations which we refer to different organs. We are
not interested here in the physiology of fear and will be satisfied to
isolate individual representatives of this sensation and the most
frequent and evident have to do with the respiratory organs and
the heart. They are to us proofs that the motor innervation (dis-
charge processes) take part in the totality of fear. Hence the
analysis of the fear stage gives us (1) a specific pain character;
(2) a discharge reaction; and (3) perception of the same. Points
2 and 3 at once give us a distinction between fear on the one hand
and grief and pain.

In the two latter there is no motor expression; where they are at
hand they separate themselves distinctly not as constituents of the
whole but as consequents or reactions. Fear is then a special state
of pain with discharge reactions along definite paths. According to
our general views we shall believe that fear is based on increased
excitation which on the one hand gives it its character of pain and on
the other relieves itself through discharge. But this purely physio-
logical comprehension will hardly suffice. We are tempted to assume
the existence of a historical factor which unites the sensations with
the innervations. In other words, the state of fear is the reproduc-
tion of an experience which contains the conditions of an increase of
irritation and leads off along certain paths whereby the pain of fear
receives its specific character. As such a visualized experience birth
in mankind presents itself, and therefore we are inclined to see in
a fear state a reproduction of the birth trauma.

At the same time we have not maintained that fear is an exception
in this respect as compared with other affective states. We believe
that other affects are also reproductions of old experiences which

495

are important to life and perhaps preindividual, and we adduce them
as general, typical, congenital hysterical attacks in comparison with
late and individually acquired attacks of hysterical neurosis, the
genesis and importance of which have through analysis become dis-
tinct. Naturally it would be very desirable to be able to carry this
notion out for a series of other affects in a demonstrable fashion
but at present we are far removed from this prospect.

Reference of fear back to the birth trauma must defend itself
against objections which are close at hand. Fear is probably a reac-
tion experienced by all animals although only the mammals experi-
ence labor and it is questionable whether the latter is a trauma with
them. There is therefore fear without the prototype of birth. But
this objection is outside the limits between biology and psychology.
Just because fear has a biologically indispensable function to fulfil
as a reaction to the condition of danger it may be arranged in differ-
ent fashions for different creatures. We do not therefore know
whether in the creatures far removed from man there is the same
content in sensations and innervations as in mankind. This does
not, however, prevent fear from being a prototype of the birth
process in man. If this be the structure and origin of fear, a further
question is, What is its function? Under what circumstances is it
reproduced? The answer seems near at hand and convincing. Fear is
a reaction to a state of danger and is reproduced regularly when
such a state reappears.

There is something else to note. The innervation of the original
fear state was probably characterized by feeling and was purposeful,
quite as with the muscle actions of the first hysterical attack. If one
would explain the hysterical attack one need only to seek the situation
in which the movements concerned were part of an authorized action.
So probably during labor the direction of the innervation upon the
respiratory organs prepares the activity of the lungs, the acceleration
of the heart-beat works against the poisoning of the blood. This
efficiency naturally fails in the later reproduction of the fear state
as an affect, as it is also missed in the repeated hysterical seizure.
If, then, the individual engages in a new situation of danger it can
readily become unsuitable to answer by a fear state, the reaction to
a former danger, in place of that of the now adequate reaction. The
utility again appears when the danger situation is recognized as
approaching and as signalized by the outbreak of fear. The fear can
then be disposed of through suitable measures. Hence two possi-
bilities separate themselves in the appearance of fear: one, unsuit-
able, in a new situation of danger, and the other suitable for signalling
and preventing danger.

496

But what is a danger? In the act of birth there exists an objective
danger to the preservation of life; we know what it imports in real-
ity, but psychologically it tells us nothing. The danger from
birth has as yet no psychic content. We certainly may not presup-
pose that the fetus has any kind of knowledge that it is in danger of
annihilation. The fetus can do no more than perceive a wholesale
disturbance in the economy of its narcistic libido. Large quantities
of excitations penetrate to it and generate a new series of unpleasant
situations; many organs are forced to heightened fixations constitut-
ing a prelude to the soon to begin object-investments. What among
these find evaluation as a characteristic of a danger situation?

Unfortunately we know far too little of the psychic comprehension
of the new-born to answer this question directly. I cannot even
concede the utility of the above depiction. It is easy to say that the
new-born will repeat the fear affect in all situations which recall to it
the event of birth. However, the decisive point remains, what is
recalled, and how.

Hardly anything remains but to study the occasions in which the
nursling or older child shows itself to be prepared for fear. In his
book on the birth trauma, Rank^* makes an energetic effort to demon-
strate the relations of the earliest phobias of the child to the impres-
sion of the event of birth, but I do not regard this as fortunate.

One may make two objections to this view: First, that it rests on
a presupposition that the infant has definite sense impressions, espe-
cially those of a visual nature, received during birth and renewed in
memory of the birth trauma, thus producing the fear reaction. This
assumption is fully unproved and devoid of probability, and it is not
credible that the child at birth has anything beyond tactile and general
sensibility. If it later shows fear of small animals as they vanish
into holes or emerge from the same, Rank explains this reaction
through the perception of an analogy which, however, could not
become striking to a child. Second, Rank, in the evaluation of this
later fear situation according to the requirement of the recollection
of the happy intrauterine life or of its traumatic interruption, leaves
the door open to any kind of arbitrary interpretation. Individual
cases of child fear are directly opposed to the application of Rank’s
principle. If the infant is brought into darkness and solitude we
should expect that it would accept this restoration of the intrauterine
situation as gratification. If the fact that it reacts to this very
situation with fear is referred back to the recollection of the disturb-

^*Otto Rank. Birth Trauma and Its Importance for Psychoanalysis.
Psychoanalytic Bibliothek, XV, 1921.

497

ance of its happiness by birth one can no longer ignore the forced
character of the interpretation.

I must draw the conclusion that the earliest child phobias do not
permit of a direct transference to the impression made by the act of
birth, and thus far they have withdrawn themselves from explanation.
A certain predisposition to fear is unmistakable in the nursling. It
is not at its worst directly after birth, to decrease slowly, but appears
during psychic evolution and persists through a certain period of
childhood. When such early phobias extend beyond this period they
awake the suspicion of a neurotic disturbance, although their relation-
ship to the later distinct neuroses of childhood is obscure.

Only a few cases of the expression of child fear are intelligible to
us and we must limit ourselves to them. When the child is alone, in
the dark, and when it finds a stranger in place of its mother, all three
states reduce themselves to one condition, absence of the loved and
longed for person. But from here on the way to understanding fear
and to reconciling the contradictions which seem associated, is free.

The memory picture of the person longed for is certainly intense,
probably at first hallucinatorily fixed. But this has no consequence
and now it appears as if this longing changes to fear. There is
exactly the impression as if this fear is the expression of perplexity,
as if the undeveloped creature knew nothing better than to start with
this yearning fixation. The fear appears as a reaction to missing the
object and there is forced on us the analogy that castration fear has
also the content of separation from a highly prized object and that
the original fear (the primary fear of birth) arises from separation
from the mother.

The next consideration leads to the emphasis of the loss of the
object. If the nursling longs to behold the mother it is only because
it knows from experience that she satisfies all its requirements with-
out delay. The situation which it evaluates as fear and against which
it would be assured is one of unsatisfaction, the increase in the
requirement tension, against which it is powerless. I mean that from
this point of view on, everything is arranged; the situation of unsat-
isfaction attains a painful height from the mass of stimulus and does
not find itself overmastered by psychic employment or discharge.

For the nursling this must be analogous to the experience of birth,
the repetition of the danger situation. Common to both is the
economic disturbance through the increase of the mass of stimulation
demanding discharge.

This factor is the essential nucleus of the danger. In both cases
the fear reaction appears, having shown itself already in the nursling
as useful, since the direction of the discharge upon the respiratory

498

and vocal musculature summons the mother, just as formerly the lung
activity did away with internal irritation. From its birth the child
does not need to maintain more than this characterization of danger.
With the experience that an external object comprehended through
perception can make an end to the dangerous warning situation, the
content of danger now displaces itself from the economic situation
to its condition, the loss of the object. To miss the mother is now
the danger at the appearance of which the child gives the fear signal,
even before the economic situation has appeared. This transforma-
tion signifies the first great advance in the impulse toward self-
preservation and at the same time it includes the transition from the
automatically unwilled new origin of fear to its intentional reproduc-
tion as the signal of danger.

In both respects, as an automatic phenomenon and as a rescuing
signal, fear shows itself as the product of the psychic helplessness of
the nursling which is the self-evident counterpart of its biological
helplessness. The striking coincidence that birth fear as well as
nursling fear recognize the condition of separation from the mother
needs no psychological explanation. It is explained biologically
simply enough from the fact that the mother has soothed all the
demands of the child before birth and partly continues to do so after
delivery. Intrauterine life and earliest childhood represent far more
of a continuum than the striking censure of the act of birth allows
us to believe. The psychic mother-object replaces to the child the
biological fetal situation. We must therefore not forget that in intra-
uterine life the mother was no object and that at that time there was
no object.

It is easily seen that in this connection there is no room for reacting
off the birth trauma and that another function of fear as a signal for
avoiding danger is not to be found. The fear condition of the loss
of the object carries us still further. The next transformation of
fear, the castration fear which develops in the phallic phase, is a fear
of separation bound to the same condition. The danger here is
separation from the genitals. A thought train from Ferenczi which
seems fully justified allows us to recognize distinctly the line of the
connection with the former content of the danger situation. The
high narcistic valuation of the penis can rest on this, that the posses-
sion of this organ contains the danger of a reunion with the mother
(mother substitute) in the act of coitus. Deprivation of this mem-
ber is as good as a fresh separation from the mother, thus again
indicating an unpleasurable requirement tension (as in birth) to be
delivered helpless. This requirement, the increase of which is
feared, is, however, only a specialized one, that of the genital libido,

499

and is no longer an optional one as in the nursing period. I add here
that the fancy of return within the mother's womb is the substitute
coitus of the impotent (inhibited by the castration threat). In
Ferenczi's sense one may say that the individual who will allow him-
self to represent the return into the womb with his genital organ,
replaces this organ regressively by his entire person.

The progress in the development of the child, the increase of its
independence, the acute separation of his psychic apparatus in several
instances, the appearance of new needs cannot remain without influ-
ence on the content of the danger situation. We have followed this
transformation from the loss of the mother object to castration and
see the next step caused by the might of the super-ego. With the
depersonalization of the parental instance, from which one fears the
castration, the danger becomes more definite. Castration fear devel-
ops into conscience fear and social fear. It is now no longer easy to
say what is feared. The formula of separation, exclusion from the
herd, applies only to that later portion of the super-ego which has
developed from leaning upon social prototypes, and not to the nucleus
of the super-ego, which corresponds to an introjected parental
instance. More generally expressed, it is the anger, the punishment
of the super-ego, the loss of love from that side which the ego values
as danger and answers with the fear signal. As the last transforma-
tion of this fear of the super-ego is the fear of death (life), fear of
the projection of the super-ego into the power of destiny as it appears.

Formerly I placed a certain value on the presentation that it is the
fixation withdrawn in repression which the application experiences
as conducting off the fear. This view now seems to me hardly worth
while. The difference lies here, that formerly I believed that fear
always arose automatically through an economic process, while the
present conception of fear as a signal intended to influence the
pleasure-pain factor makes us independent of this economic com-
pulsion. There is, of course, nothing to be said against the assump-
tion that the ego utilizes this very energy which is set free in the
diversion of repression to arouse the affect, but it becomes unim-
portant with what portion of the energy this takes place.

Another proposition which I have expressed now demands retest-
ing in the light of our new comprehension. This is the assertion that
the ego is the actual seat of fear.

In my opinion this will be shown to be correct. We have no occa-
sion to impute to the super-ego any kind of an expression of fear.
If, however, one speaks of a fear of the It there is no contradiction
implied but one must correct a clumsy expression. Fear is an affec-
tive state which naturally can be felt only by the ego. The It cannot

500

have fear like the ego; it has no organization and is unable to judge
a danger situation. On the other hand, it is a frequent occurrence
for processes to be prepared or completed in the It which give occa-
sion to the ego for the development of fear. In fact, the probably
earliest repressions with the majority of the later ones are motivated
through fear of the ego towards individual processes in the It. We
differentiate here, again with good reason, the two processes which oc-
cur in the It. One activates the danger situation for the ego and
moves it to give the fear signal for inhibition; and the other pre-
pares in the It a situation analogous to the birth trauma, in which it
comes automatically to a fear reaction. The two are brought nearer
to each other when one emphasizes that the second corresponds to
the first and original danger situation; the first, however, is one of
the later ones derived from its fear conditions. Or, referred to affec-
tions which actually occur, that the second process is realized in the
etiology of the actual neuroses while the first remains characteristic
for the psychoneuroses.

We now see that we do not need to depreciate former communica-
tions but merely to bring them into association with the more recent.
It is not to be denied that in abstinence, or abusive disturbance in the
course of sexual excitement, resulting in a diversion of the same from
its psychic elaboration, fear may originate directly from the libido;
that is, that condition of helplessness of the ego against an over great
requirement tension is produced, which as in birth proceeds from fear
development. We see that upon the foundation of these actual
neuroses psychoneuroses may easily develop; that is to say, the ego
makes an effort to spare the fear, which it has learned to keep for a
time in suspense and bind it by symptom formation. Probably analysis
of the traumatic war neuroses, which term indeed comprises very
different affections, would show that a number of these share the
characteristics of actual neuroses.

As we presented the development of the different danger situations
from the original birth prototype, it is far from our intention to
maintain that every later fear condition merely sets the predecessor
out of power. The progress of ego development contributes indeed
to the depreciation of the former fear situation and shoves it aside,
so that one may say that a certain fear condition must be assigned as
adequate to a definite age of development. The danger of psychic
helplessness fits in with the time of life of the immaturity of the ego,
just as the danger of loss of the object belongs to the early years of
childhood, the castration danger to the phallic stage, and fear of the
super-ego to the period of latency. But it can also happen that all
these danger situations and fear conditions coexist side by side and can

501

cause the ego to react with fear at periods later than those adequate
or that several of them may step into activity simultaneously. Pos-
sibly closer relations may exist between the actual danger situation
and the form of the subsequent neurosis.^*

In an earlier portion of these investigations we came upon the
importance of the castration danger in more than one neurotic affec-
tion, and we imparted the warning not to overemphasize this
factor, for it might not work out in the neuroses to which women
are known to be more predisposed. We now see that we are not in
danger if we explain castration fear as the sole motif for the defen-
sive processes which lead to the neurosis. In another place I have
explained how the development of the little girl is directed by the
castration complex to a tender object fixation. It is exactly in
women that the danger situation of the loss of the object appears
to be the most efficacious. We may adduce the small modification
in their fear condition, that it is no longer a matter of missing or of
the real loss of the object but of the loss of love by reason of the
object. As it is certain that hysteria has a great affinity for female-
ness as the compulsive neurosis has for maleness, the presumption
would be that the fear condition of the loss of love plays a rôle in
hysteria which is analogous to that of castration threat in the phobias,
and the fear of the super-ego in the compulsive neuroses.

^*Since the distinction between the ego and It our interest in the prob-
lems of suppression must experience a new animation. Hitherto we had
been satisfied to take into account the aspects of the process applied to the
ego, the shutting off from consciousness and from motility and the substi-
tute (symptom) formation. For the suppressed prompting itself we
assumed it to remain indefinitely long and unchanged in the unconscious.
Now our interest became directed to the fate of the repressed and we
had the foreboding that such an unchanged and inalterable continuity is
not self-evident and perhaps not even customary. The original prompting
is in any case inhibited through the repression and diverted from its aim.
But does the portion in the unconscious remain intact and does it show
itself resistant toward the changing and depreciating influences of life?
Do the old wishes, which analysis showed formerly to exist, persist? The
answer seems near at hand and to be trustworthy. The repressed old
wishes must persist in the unconscious, for we find their derivatives, the
symptoms, still active. But such an answer is not sufficient; it allows no
decision to be made between the two possibilities, whether the old wish
merely acts through its derivatives to which all its energy of fixation has
been transferred or whether in addition the wish itself persists as such.
If it had been its fate to be exhausted in the fixation of its derivatives, there
remains a third possibility that in the course of the neurosis it is reanimated
by regression, however unseasonable this may seem at present. One does
not need to regard these speculations as idle; much of the phenomena of
the morbid as well as the normal in the psychic life seems to demand such
questioning. In my studies of the decline of the Oedipus complex I have
called attention to the difference between the bare repression and the actual
elevation of an old wish.