PATHOLOGICAL MOURNING :

AND JOHN BOWLBY,

hf.~.

CHILDHOOD hlOURNING i

HIS IS the fifth in a series of papers in which I am exploring

T the theoretical implications of the behavior to be observed

when young children are removed from the mother figures to
whom they are attached and are placed with strangers. T h e thesis
I am advancing is twofold: “first, that once the child has formed
a tie to a mother-figure, which has ordinarily occurred by the
middle of the first year, its rupture leads to separation anxiety
and grief and sets in train processes of mourning; secondly, that
in the early years of life these mourning processes not infrequently
take a course unfavourable to future personality development and
thereby predispose to psychiatric illness.” Since in earlier papers
I have discussed the theory of separation anxiety and mourning,
it is to the second part of the thesis that this and later papers are
directed.
In the preceding paper “Processes of Mourning” (5) reasons
were given for dividing healthy mourning, whether it occurs in
human infants or adults or in. lower species, into three main
phases. Particular attention was given to the first phase, during
which the instinctual response systems binding the bereaved to
the lost object remain focused on the object, because during this

. T h e substance of this paper was given as the Third Alexander Lecture to the
Chicago Institute of Psychoanalysis on May 4, 1961.
I am much indebted to James Robertson for the observations on which I have
drawn. T h e enquiry was undertaken as part of the work of the Tavistock Child
Development Research Unit, which is supported by the National Health Service
and by grants from the Foundations Fund for Research in Psychiatry and the Ford
Foundation, to each of which our thanks are due.

500
 PATHOLOGICAL AND CHILDHOOD MOURNING 501

phase yearning and an angry effort to recover the lost object seem
to be the rule. T h e hypothesis was advanced that, so far from
being pathological, an open expression of sych angry striving to
recover the object is a sign of health and-that it enables the
bereaved gradually to relinquish the object. TVhat seems to char-
acterize much pathological mourning is an inability to accept and
express this striving; instead, it becomes repressed and unconscious
and so, insulated from change, persists.
If grief itself is to be regarded as a disease, as Engel (1 1) has
persuasively argued, how, it may be asked, are the terms “healthy”
and “pathological” best applied? Engel’s own analogies are useful.
If the experience of loss is likened to the experience of being
wounded or being burned, the processes of mourning that follow
loss can be likened to the processes of healing that follow a wound
or a burn. Such healing processes, we know, may take a course
which in time leads to full, or nearly full, function being restored:
or they may, on the contrary, take one of many courses each of
which has as its outcome an impairment of function of greater
or less degree. In the same way, processes of mourning may take
a favorable course that leads in time to restoration of function,
namely, to a renewal of the capacity to make and maintain love
relationships, or they may take a course that leaves this function
impaired in greater or less degree. Just as the terms healthy and
pathological are applicable to the different courses taken by heal-
ing processes, so may they be applied to the different courses run
by mourning processes.
Four main variants of pathological response that are shown
by bereaved adults will be considered. I shall begin with those clin-
ical conditions, often suffused with anxiety and depression, in
which a persistent and unconscious yearning to recover the lost
object is most evident. A main reason for starting here is that an
unconscious urge to recover the lost object is a dynamic that
probably is present or at least latent in all the other pathological
variants of mourning but, because more deeply repressed, is more
difficult to see. Moreover, little attention hitherto seems to have
been directed to it, and much in my thesis rests on a clear under-
standing of its role.
T h e second variant to be considered is intense and persistent
 502 JOHN BOIVLBY

anger and reproach expressed toward various objects, including

the self. Although such aggressive thought and action, concerned
as they often are with events of the past, seem pointless enough
to the outsider, I believe their function,. of which the subject is
not usually fully conscious, is always to achieve reunion with the
lost 0bject.l And this I believe to be the case even where feelings
of revenge and hatred for the lost object seem to be the chief
underlying motives, and even, too, where these are directed not
toward the lost object but toward the self.
T h e third variant is the way in which, instead of grieving,
some bereaved people become absorbed in caring for someone else
who has also been bereaved, an absorption that seems at times to
amount to a compulsion.
T h e fourth variant, denial that the object is permanently lost,
is a way of responding very different from the first three. I n each
of these variants the bereaved acknowledges the reality of loss; the
pathology lies in the repression of certain of the instinctual
response systems evoked by it. JVhere, on the other hand, reality
of the loss itself is denied such response systems are not evoked:
instead a main part of the pathology is at a cognitive level. Such
denials, however, rarely represent the complete picture. Probably
always they coexist with some awareness of the object’s loss. As
a result there comes into being a split in the ego of the kind Freud
was interested in from the beginning to the end of his professional
life (e.g., Studies on Hysteria [7] and “Splitting of the Ego in the
Defensive Process” [23]).
Although it is convenient to discuss each of these four main
variants of pathological mourning separately, it will be evident
that they are not mutually exclusive. On the contrary, they exist
.
together in a multitude of combinations and give rise to clinical
conditions of protean variety: anxiety and depressive illness,
1Throughout this series of papers I am using the concept of function in a way
that is nonteleological. Thus the function of aggressive thought and action follow-
ing loss is conceived on the same model as the physiologist conceives the function
of inflammation following infection. In one case the function is to achieve reunion,
in the other to repel bacteria: in both the concept of purpose is absent. “In order
to emphasize that the recognition and description of end-directedness does not carry
a commitment to Aristotelian teleology as an efficient causal principle.” the term
teleonomic has been proposed by Pittendrigh (34): it denotes any end-directed
system.
 PATHOLOGICAL AND CHILDHOOD MOURNING 503

fetishism, and hysterical and psychopathic behavior appear to be

among the commoner. Here, however, my concern is ivith patho-
logical process not with clinical syndrome. For this reason, instead
of attempting to elucidate all the features of any one clinical pic-
ture, which is the task most analytic authors set themselves, my
interest is to trace the manifold ways in which processes of mourn-
ing can affect the lives of individuals. Once this has been well
done, it seems likely, the pathology of each of the many clinical
syndromes will be easier to decipher.
One very important variant of pathological mourning is a
distortion of the process of identification with the object. Since
it is this variant that has been the center of theoretical attention
in the psychoanalytic literature, its omission requires explanation.
One reason is that the problems raised by identification, healthy
and pathological, are so complex that together they require a
series of papers to themselves. Another is that the primary data
on which we are drawing-the responses of young children to
s e p a r a t i o n 4 0 not lend themselves readily to the study of identi-
ficatory processes and their deviations. Above all, as just explained,
the task I set myself is different from that of most analytic authors
who have discussed mourning and its pathology. T h e reason that
Freud (19) became interested in identification with the lost object
is because he was grappling with the problem of how it is that
reproach, originally directed against the lost object, comes in
melancholic patients to be directed instead against the self.
Although my discussion will touch on this problem, it does not
seek to solve it. Instead, its concern is with the origin and function
of the reproach itself.
I n describing the four main variants of pathological mourn-
ing, illustrations drawn from a number of case reports will be
presented. In several of them, it will be noted, the loss that gave
rise to the mourning from which the patient is still suffering is
the loss of a parent that occurred in the patient’s childhood. That
this should be so is no accident. Apart from any bias I may have
to select such cases, there is much evidence, statistical and clinical,
that many though by no means all cases of pathological mourning
in adults have originated following a loss i n childhood. This evi-
504 JOHN BOWLBY

dence, which is of much consequence to my thesis, is reviewed in

part elsewhere (6).
I n this paper the main aim is, having first delineated certain
pathological patterns of mourning, to compare these pathological
patterns with patterns of response that are known to be typical
of infants and young children following loss. T h e result is a
conclusion of great significance. This is that the two sets of pat-
tern are substantially the same: in other words, the mourning
responses that are commonly seen in infancy and early childhood
bear many of the features which are the hallmarks of pathological
mourning in the adult.
- This discussion will throw into relief the problem of defense.
T h e mourning responses of infancy and early childhood, it will
be seen, can no more be understood without reference to defense
than can the pathological variants of mourning in adults. Proc-
esses of repression, splitting, and denial are all at work. Because,
however, the issues raised by them and by their relation to mourn-
ing are large and controversial, there is no room to do them justice
i n this paper. .
When describing the four patterns of pathological response
that are shown by bereaved adults I shall frequently draw on case
material reported by other analysts. T h e discussion of literature
dealing with theory, however, will be postponed till the end.
There are three main features of pathological mourning to which
I am drawing attention. They are: (1) the role of an unconscious
urge to recover the lost object; (2) the tendency for defensive
processes to come into action after a bereavement; and (3) the
frequency with which individuals prone to pathological mourning
have experienced loss of a parent in childhood. Only the literature
dealing with the first will be discussed here.

Pathological Variants of Alotirning

Unconscious Yearning for Lost Object
'CVhen in the previous paper I described the first phase of the
normal mourning process I emphasized, in addition to persevera-
tion of thought, feeling, and behavior oriented toward the lost
partner, two main features: the first is weeping, and the second
 PATHOLOGICAL AND CHILDHOOD AIOURNIKG 505

aggressive thought and action. Both it seemed had as their func-

tion the recovery of the object. TVhen in fact the object is only
temporarily lost responses of these kinds are useful and this, it
was suggested, accounts in large part of their. being so regularly
the initial reaction to loss. TVhen, however, the object proves to
be permanently lost they are no longer useful; in due course,
therefore, they may be expected to diminish and disappear, I n
healthy mourning this is what occurs, and the second and third
phases of mourning duly follow: despair and disorganization are
succeeded by reorganization and hope. I n pathological mourning,
on the other hand (or at least in some of its commonest forms),
they persist, albeit in disguised forms. T h e bereaved person
remains preoccupied in thought and action with the lost object,
not only organizing his life as though it were still recoverable
but continuing to weep for it and commonly also to display
dissatisfaction and ill temper with his friends and himself.
Considerations of these kinds led to the hypothesis being
advanced that an open expression of protest and of demand for
the object’s return is a necessary condition for a healthy outcome.
Such an outcome, i t seems, requires the expression both of the
yearning for the lost object, accompanied by sadness and crying,
and also of the anger and reproach that are felt toward the object
for its desertion. That the open expression of both these compo-
nents of ambivalence in healthy mourning is common is well
illustrated by the case of Mrs. B. which was chosen by Edith
Jacobson (26)as a contrast to her cases of pathological mourning.
Mrs. B. was a middle-aged woman in a state of profound
mourning following the loss of her husband. On the one hand,
there was evident a deep yearning for him: “She would weep
copiously when alone dwelling on her memories of the beloved
partner. She would feel the need again and again to visit the same
places in the mountains whose beauty they had enjoyed together.
..
. . . A beautiful view . now evoked outbursts of weeping in
association with her memories o€ the delight she had shared with
her husband.” On the other, was hostility directed against him
“of which she apparently had been and was fully aware,” and also
some hostility directed against herself in the shape of remorse for
some of her behavior toward him. Incompatible both with each
506 JOHN BOWLBY

other and with reality though these responses appear, and in some
respects indeed are, it is my hypothesis that if mourning is to run
a healthy course, it is necessary that they be experienced and
expressed. TVhen they are not, it is suggested, reality testing is
more likely to fail and the unrealistic demand for the object’s
return to live on at an unconscious level. Certainly this is charac-
teristic of many cases of pathological mourning.
A case of neurosis in an adolescent girl reported by Root
(39) has many features which‘appear to be fairly typical. T h e
condition, he believed, stemmed from unsatisfactory mourning
following the death in an accident of the patient’s mother when
the patient herself was aged ten. T h e main symptoms were those
of anxiety and depression, with inhibition, indecision, nausea and
sickness. “In the analysis,” it is reported, “she could not at first
comprehend, even intelIectually, that she missed her mother.” I n
her dreams and fantasies, on the other hand, mother constantly
appeared. In some of them there was a happy reunion with her
mother; i n others the picture of her mother in a sanatorium, or a
frightening scene of her mother “with the gory signs of head
and face injuries.” From such evidence we may conclude that,
unknown to herself, this girl was still in the first phase of mourn-
ing and still unconsciously striving for reunion with her lost
mother. This view is supported by a remark she made when well
on in the analysis: “she spoke of grieving as ‘letting her mother
die.’ ” Only after she could accomplish the task of mourning her
mother’s death, Root reports, did improvement in her condition
develop.
There are two points in this girl’s history which are of inter-
est. I n the first place her initial reaction to her mother’s death
appears to have been one of incredulity and denial, and there had
been n o expression of grief. At the same time there was much
concern for vicarious figures: “she could shed tears for an
orphaned beggar girl,” and, during the analysis, “often displaced
her sadness onto something else or felt sad for someone else”;
moreover, her marriage at seventeen had been to a man who, like
herself, had lost his mother when he was aged ten.
T h e second point refers to the patient’s early childhood. Her
mother had continued working and the patient had been looked
 PATHOLOGICAL AND CHILDHOOD MOURNING 507

after by maids. How many is not stated, but when one to whom
she had been attached left when she was seven, there seems to
have been a “tearful parting” which in the analysis “was easily
recalled with much emotion and weeping.” T h e question there-
fore arises whether perhaps this patient had had experiences in
the earlier years that had predisposed her to respond to her
mother’s death with a pathological form of mourning, amounting
to an “absence of grief.” This is what my theoretical position
leads me to predict.
Two of the three cases reported by Joan Fleming et al. (13)
show many resemblances to this case of Root’s. T h e first is of a
woman of thirty complaining of anxiety and depression precipi-
tated by failure in an examination and disappointment in a love
affair. At fifteen years she had left Germany as a refugee and three
years later had heard of her father’s death: the fate of her mother
remained uncertain. Although she insisted that she “never
mourned them and never seemed to miss them,” in her dreams
“she cried as she rushed to meet them.” Because she was tearful
when reminded of her parents, “she fought against this response
by becoming angry and condemning of the grief responses of
others.” Although in analysis she protested her ability to cope
without help, later she admitted: “Time stopped for me at fif-
teen.” T h e second patient, a man of twenty-three, showed many
of the same features.
I n both these patients the repression of yearning for the
lost object is clear enough, and i n neither does the analyst seem
to have had too great a difficult in enabling the patient once
more to experience it and thereby to get over it. I n other cases
the unconscious yearning is more deeply repressed and therefore
less evident. Thus Helene Deutsch (10) reports the case of a man
whose personality had developed in such a way that, but for a
fantasy about a dog, it would be very difficult to convince a skeptic
that unconscious yearning was really present.
Helene Deutsch’s patient was in his early thirties when,
without apparent neurotic difficulties, he came into analysis for
nontherapeutic reasons. T h e clinical picture was one of a wooden
and aff ectionless character. She describes how “he showed com-
plete blocking of affect without the slightest insight. I n his limit-
508 TOHN BOWLBY

less narcissism he viewed his lack of emotion as ‘extraordinary

control.’ H e had no love-relationships, no friendships, no real
interests of any sort. T o all kinds of experiences he showed the
same dull and apathetic reaction. There was no endeavor and no
disappointment. . . . There were no reactions of grief at the loss
of individuals near to him, no unfriendly feelings, and no aggres-
sive impulses.”
As regards history, we learn that his mother had died when
he was five years old and that he had reacted to her death without
feeling. Later he had repressed not only the memory of his mother
but also everything else preceding her death.
“From the meager childhood material brought out in the
slow, difficult analytic work,” Helene Deutsch continues, “one
could discover only negative and aggressive attitudes towards his
mother, especially during the forgotten period, which were obvi-
ously related to the birth of a younger brother. T h e only reaction
of longing for his dead mother betrayed itself in a fantasy, which
persisted through several years of his childhood. I n the fantasy
he left his bedroom door open in the hope that a large dog would
come to him, be very kind to him, and fulfill all his wishes.
Associated with this fantasy was a vivid childhood memory of a
bitch which had left her puppies alone and helpless, because she
had died shortly after their birth.”
I n each of the cases reported by Helene Deutsch, and espe-
cially this one, the bereaved person adopted a hostile, aggressive
attitude in place of yearning for the loved figure that is missed.
This common tendency has been commented upon by a number
of analysts, for example Lindemann (33) and Melanie Klein (28).
Edith Jacobson (26), referring to the same thing, notes that “dur-
ing treatment severely depressed patients . . . may even consciously
realize that, could they only be sad and weep, they would ‘feel
for the world’ again,” and that when in fact they are able to
yearn for what is lost ‘‘a relieving ‘sweet sadness’ may break
through.” It is my hypothesis that the yearning which these
patients are initially unable to experience is the muted expression
of a demand for the object’s return and of an urge to recover it
which is active at an unconscious level. To some extent their
anger, even if conscious, is to be understood in the same light.
 PATHOLOGICAL AND CHILDHOOD MOURNING 509

Although i n part it seems to represent a wish that the lost and

therefore pain-arousing object should be removed or destroyed, in
part it is the distorted expression of an angry desire that the object
return.

Unconscious Reproach against 0bject: Self-reproach

Ever since Freud published “Mourning and hfelancholia” there
has been awareness among analysts of the immense importance
that anger with the lost object plays in pathological mourning.
Indeed, Freud himself and many others have been so struck by
it that its presence has often been thought to be diagnostic of
pathological mourning. I believe this to be a mistake, and that
anger with the lost object, with others, and with the self is
extremely common even in healthy mourning. I n fact, only by
clear recognition of its presence and function in healthy mourn-
ing, I believe, are we able to understand its place in pathological
mourning.
I n addition to its very existence, there are a number of
other features of the anger following loss that have proved puz-
zling. Among them are its bitterness, its persistence, and its fre-
quent selection of inappropriate targets, including its tendency to
turn against the self. T h e difficulty in understanding these features
has given birth to many complexities of psychoanalytic theory. I n
the theoretical schema advanced here I hope it may be possible to
explain them more simply.
T h e cardinal feature of my hypothesis is that loss of object
almost always results in an effort to recover it, and this effort is
often the more successful when infused with a dash of aggression.
Furthermore, at reunion after temporary loss, reproaches leveled
against the object may be expected to insure that the object
becomes less prone to go away again. For example, there is
many a mother who has vowed never again to leave her young
child in strange surroundings after she has been exposed to the
reproaches he levels against her following his return home. Such
reproaches, more or less bitter, and an angry clinging to her are,
we now know, the rule following temporary loss in childhood.
I n most cases, it is clear, they perform a useful function: they are
part and parcel of the effort to undo the loss, to recover the lost
 510 JOHN BOWLBY

object and to insure that it never deserts again. Their appearance

after bereavement in a way that seems to be maladaptive is due,
I have suggested, to irretrievable loss of object being statistically
so rare that it has not been taken into account in the design of our
biological equipment.
Such then is the function of anger and reproach. T h e targets
to which they may be directed in a way that is more or less appro-
priate include the lost object, third parties, and the self. When
directed at third parties they are apt to take quasi-paranoid forms;2
doctors or others connected with the loss are accused, often
unjustly. TVhen directed against the self they give rise to self-
reproach. I wish to emphasize that each of these targets may be
legitimate. Sometimes the loss has been caused or at least facili-
tated by third parties; sometimes by acts of omission or commis-
sion the bereaved has himself contributed to the loss. Even when
in reality this is not so, it may not be altogether unreasonable to
.suppose it to be so. I n all such cases, although the reproaches may
be powerless to reverse the loss in question, they cannot be
’regarded as unrealistic.
It is a main characteristic of pathological reproaches, on the
other hand, toward whatever target they are directed, that they
are inappropriate to the person reproached. It was their inapplica-
bility to the self in cases of melancholia that struck Freud so
‘forcibly and led him to his famous hypothesis regarding the origin
of pathological self-reproaches. “If one listens patiently to a mel-
ancholic’s many and various self-accusations,” he writes, “one
cannot in the end avoid the impression that often the most violent
of them are hardly at all applicable to the patient himself, but
that with insignificant modifications they do fit someone else,
someone whom the patient loves or has loved or should love. . . .
So we find the key to the clinical picture: we perceive that the
self-reproaches are reproaches against a loved object which have
been shifted away from it on to the patient’s own ego” (19, p. 248).
2Although a response of this kind may result from the projection onto another
of the bereaved person’s orrn unconscious hostile wishes torrard the lost object, and
so be truly eligible for the adjective “paranoid,” I do not believe this is ahvays so.
At times I suspect such behavior to be no more than the fury of frustration at loss
being more or less irrationally expressed. For this reason I term it “quasi-paranoid.”
Lindemann (33) appears to make the same distinction.
 PATHOLOGICAL AND CHILDHOOD MOURNING 51 1

As a result the anger and reproach felt toward the loved object
cease to be conscious and instead persist unconsciously.
Although this hypothesis regarding the origin of pathological
self-reproach has been widely accepted by analysts, there are some
who have attenuated or even at times reversed it; for example,
Melanie Klein and her colleagues and also Weiss (44)have empha-
sized instead the degree to which reproach against the self is
appropriate and reproach against the object not so. T h e position
I am adopting is near to Freud’s; it expands his hypothesis by
putting it into a more generalized form.
As we have seen, depending on the circumstances of each
case and the degree of understanding of the bereaved, blame for
loss of object may reasonably be attributed by him to one or more
of several persons: the object, the self, and third parties. This
being so, in principle reproach directed toward any one person
may be shifted away from that person toward any other. In prac-
tice, of course, such shifts tend to be of particular kinds and are
usually designed to spare either the self or the object. T h e com-
monest are away from the self and toward third parties, and away
from the object and toward the self. XVhen the self is spared
blame, the clinical picture is paranoid; when the object is spared
it is depressive. I n each case, it is held, the pathology lies not i n
the existence of anger and reproach but in their displacement
away from an appropriate object toward an inappropriate one.
As a result angry and reproachful feeling toward the original
object become unconscious.
This leads to a consideration of the place of guilt. Within
the context sketched, guilt over loss is seen to be generated i n at
least three different ways. First, there is guilt that springs from a
more or less realistic appraisal of the part the bereaved himself
has played in the loss, and may be large or small according to
circumstances. This was clearly recognized by Freud: there are,
he remarks in discussing the self-reproaches of melancholics, “a
few genuine self-reproaches . . . scattered among those that have
been transposed back” (19, p. 248). Secondly, there is the rather
less realistic guilt that arises from the bereaved knowing consciously
or unconsciously that, even if he is not in reality responsible for
the loss, he has often wished the lost object away. h4uch neurotic
512 JOHN BOIZILBY

guilt over loss appears to be of this character. Finally, there is the

wholly irrational guilt that arises from the turning toward the
self of reproaches initially and more or less appropriately aimed
at the object. It was one of Freud’s great discoveries to recognize
that psychotic guilt is to be accounted for by a shift of this kind.
This picture of guilt and its role in mourning, it will be
noted, distinguishes it sharply from grief. In some formulations,
especially Melanie Klein’s, grief is so closely identified with guilt
that the two are almost indistinguishable. I n this formulation, on
the other hand, whereas grief is conceived as an “amalgam of
anxiety, anger and despair following the experience of irretriev-
able loss” (5), guilt is conceived as resulting from those compo-
nents of angry reproach which happen, whether appropriately or
inappropriately, to be directed toward the self. These different
formulations, it will be evident, lead to substantial differences in
the content of interpretations that are given to patients suffering
from pathological mourning.
It seems probable that it is the turning of anger and reproach
away from an appropriate object and toward an inappropriate
one, so that one of their main components becomes unconscious,
that accounts for their tendency in cases of pathological mourning
to persist long beyond the normal period. Experience suggests
that, if from the first they are openly expressed toward their
appropriate object, both yearning for the lost object and angry
reproach against it or others (including the self) tend gradually
to fade. I n so far as there is hope of recovery they have performed
their function: in so far as there is no hope this has been demon-
strated beyond peradventure. It is when yearning and reproach are
not openly expressed toward their appropriate object that they
persist. It is as though secretly and unconsciously hope remains
that strenuous enough effort to recover the lost object may still
succeed and bitter enough reproach against it for deserting may
still prevent repetition. Until the effort is made and the reproach
expressed these possibilities remain: and so displaced and uncon-
scious yearning and also angry reproach rumble on over the years
causing misery to everyone in their orbit.
Large part though displacement seems to play in accounting
for the persistence of angry reproach, another factor probably
 PATHOLOGICAL AND CHILDHOOD, MOURNING 513

enters also. It is not only when displaced and unconscious that

reproach remains ineffectively expressed. Inevitably when loss is
prolonged or permanent it remains so. This fact seems also to
contribute to its becoming so bitter. This is best seen by contrast-
ing the psychological situations of subjects exphsed to short and to
prolonged (or permanent) separations respectively.
TVhen separation is only temporary opportunity soon comes
for the expression of the angry reproach felt toward the lost
object. Once expressed the affectionate components of the rela-
tionship can become active once more and, although anger and
reproach may persist, they are constantly corrected and modified
by the presence of the object and the positive feelings it arouses.
I n a situation of prolonged or permanent loss on the other hand,
not only is opportunity to express anger and reproach missing
but the affectionate componeiits of the relationship inevitably
wane. As a result the anger and reproach, never directly expressed,
fester.
Thus it is seen that, our biological equipment being what
it is, a loss of any kind that is permanent carries with it a special
risk that the anger and reproach, always evoked by loss, will
persist instead of fading. Nonetheless this does not occur i n all
cases of mourning. As Melanie Klein, Edith Jacobson, and others
have pointed out, in the healthy mourning of adults loving mem-
ories of the lost object are strong enough to hold these angry
reproaches in check, to permit their limited expression, and so to
enable the bereaved to get over the loss. Evidently in cases of
pathological mourning special conditions are present which lead
this process to miscarry. One of them, it is well known, is the
existence in the previous attitude of the bereaved of unduly strong
tendencies toward both possessiveness and anger for the object.
Although this is a point repeatedly made by analysts who have
considered the matter, the genesis of this intense ambivalence has
not often been linked, as there is evidence it should be, with the
bereaved having suffered previous and perhaps repeated separa-
tions or rejections, especially in his childhood. This is a main
plank in my explanation of why it is that some individuals become
prone to respond to loss in a pathological way.
I believe the hypothesis here advanced provides a satisfactory
514 TOHN BOWLBY

explanation of the presence of angry reproach in pathological

mourning and thus of its presence too in depressive illness. Never-
theless, it casts light neither on the reasons for the hatred being
so often diverted away from the object and toward the self nor
on the processes by which that occurs. These I see as a separate set
of problems. Since their solution requires an examination of the-
ories of identification, I shall make no attempt in this series of
papers to tackle them.

Care of Vicarious Figures

I n the case repqrted by Root (39) and described earlier the girl,
although failing to express grief for her mother, had, by contrast,
frequently “felt sad for someone else.” For example, she used to
“shed tears for an orphaned beggar girl.”
A similar case was reported a little more fully by Helene
Deutsch (10) in the same paper from which the case of the man
with the dog fantasy has already been quoted. This patient was
“a middle-aged woman without symptoms but with a curious
disturbance in her emotional life. She was capable of the most
affectionate friendships and love-relationships . . . [and] had the
potentiality for positive and negative feelings but only under
conditions which subjected her and her love-objects to disappoint-
ment. . . . She found vicarious emotional expression through iden-
tification, especially with the sad experiences of others. T h e
patient was capable of suffering a severe depression because some-
thing unpleasant happened to somebody else. She reacted with the
most intense sorrow and sympathy, particularly in cases of illness
and death affecting her circle of friends.”
At the beginning of every analytic hour the patient wept
bitterly and for no apparent reason. During the course of treat-
ment, however, it became possible to discover “how the displaced
emotional discharges were related to early unresolved experiences.
T h e original grievous experience was not a death but a loss in the
divorce of her parents. I t gradually became clear,” Helene Deutsch
concluded, “that she had actually sought out the situations in
which she had an opportunity to share the unhappiness of others,
and that she even felt a certain envy because the misfortune had
happened to another and not to herself.”
 PATHOLOGICAL AND CHILDHOOD MOURNING 515

I n the lives of persons such as these, two characteristics stand

out. One is the way in which they appear to be dogged by ill-
fortune, and the other their compulsion to take pity on and to
care for other unfortunates like themselves.. Often they surround
themselves with a succession of “lame ducks.” Although much
charitable work of value springs from these sources, when the
motivation is uncontrolled it is apt to be as much a burden to the
receivers as to the bestower.
This tendency of a bereaved person to limit the expression
of his own grief and instead to concentrate on succoring some
other individual who has also been bereaved is a response that
calls on the mechanism of projective identification and is one
about which Greene (25) has recently written. This other indi-
vidual, Greene points out, serves as a vicarious figure3 onto whom
the bereaved person projects both awareness of loss and his own
affects of grief and helplessness. In addition, I believe, he projects
onto the vicarious figure both his yearning for the lost object and
his anger at its desertion of him. As such it is a special elaboration
of the patterns already discussed.
For an individual to fill the role of vicarious figure it is
necessary either that he himself should have been bereaved or
that he be perceived as in some other way both helpless and look-
ing for help. Those who are younger or sick are often picked;
occasionally it is an elderly person. In the case of children animals
are often selected. By devoting himself to the care of vicarious
figures a bereaved person may sometimes seem almost entirely to
avoid experiencing sorrow and suffering himself.
Greene suggests tentatively that this proxy response may be
a component even in normal mourning. I suspect this to be true:
Eliot (12) has described it as frequent among the immediate
responses to loss. There can be no doubt, however, that when it
plays a dominant role in the reaction the mourning process
becomes so distorted as to be pathological. In such cases sorrow,
yearning, and despair remain unexpressed; disorganization and
reorganization are avoided: salvation is sought through works;
3 Greene’s term is “vicarious object.” Since. however, the person in question
is acting as a substitute for the self and not the loved object, the term “vicarious
figure” is preferable.
516 JOHN BOWLBY

and behind a bright and busy exterior lurks depression. Whether

in this form it is usefully described as a variant of mourning or
whether it is better seen as an alternative to mourning is perhaps
a matter of taste.
Although from the psychodynamic point of view the response
of succoring a vicarious figure can be seen as part of a defense
against the frustration of trying to recover a permanently lost
object and the pain of sorrow and disorganization, it can some-
times be seen also from a sociodynamic point of view as filling a
role vacuum. In every social group roles are distributed asymmetri-
cally among individuals so that, when one is removed, disequilib-
rium results. I n such circumstances there is pressure on one or
other of those remaining to assume the role of the absentee. When
a father dies, mother or elder son are likely to assume his role;
when a mother dies or deserts it may be an elder daughter who
takes her place. In addition, therefore, to internally determined
motivation to assume the rights of someone who is dead or absent,
externally determined social pressures may encourage or even
demand it. Thus when an elder daughter on her mother’s death
assumes the role of “little mother” to her younger siblings, and
perhaps also “wife” to her father, and thereby experiences little
overt sorrow herself, it is often the outcome of pressures from the
family group as well as an expression of her own oedipal wishes.
I t is evident that both succoring a vicarious figure and filling
a role that is left vacant by the departure of another are important
aspects of identification with the lost object. How large a part they
play in it, however, remains to be determined.

Denial That Object Is Permanently Lost

I n 1927 i n a paper on “Fetishism” Freud (22) described two
young male patients, each of whom “had refused to acknowledge
the death of his father” who in each case had been lost in the
patient’s childhood. T h e two cases led Freud to develop his theory
of defense in a critical way; he did not, however, relate his findings
to the theory of mourning.
Since then a number of cases have been reported in ivhich,
secretly but nonetheless consciously, a patient has retained an
 PATHOLOGICAL AND CHILDHOOD MOURNING 517

active belief in the living existence of the figure who has been
lost, albeit combined with a public acceptance of the reality of
the loss. Since it is a state of mind that was demonstrated very
vividly by a mother, Mrs. Q., to whom I gave once-a-week treat-
ment over a long period, I shall draw on her case for illustration.4
After making good progress over a period of three years,
during which her very intense ambivalence toward each of her
parents (overtly positive to her mother and overtly negative to
her father) was partially analyzed, Mrs. Q. lost her father unex-
pectedly following an elective operation. Her first intimation of
anything untoward was a message from the hospital that he had
died suddenly in the night. During the ensuing months she was
again intensely depressed, again had ideas of suicide, and described
also symptoms of depersonalization. It was unfortunate that at
the time of the loss treatment was suspended owing to my absence,
and not until the anniversary of the loss was drawing near did
she confess to me certain feelings and ideas she had been hiding
but which throughout seem to have been conscious.
During the weeks following her father’s death, she now
told me, she had lived in the half-held conviction that the hospital
had made a mistake in identity and that any day they would
phone to say he was alive and ready to return home. Furthermore,
she had felt specially angry with me because of a belief that,
had I been available, I would have been able to exert an influence
on the hospital and so enabled her to recover him. Now, twelve
months later, these ideas and feelings persisted. She was still half-
expecting a message from the hospital, and she was still angry
with me for not approaching the authorities there. Secretly, more-
over, she was still making arrangements to greet her father on his
return. This explained why she had been so angry with her mother
4Mrs. Q. had been referred for help with her little boy of eighteen months
because he had been refusing to eat and was seriously underweight, and her own
emotional disturbance had seemed to be contributing to his condition. This had
been confirmed by psychiatric investigation which had found her to be in a sub-
acute agitated depression with ideas of suicide and of killing the baby. a condition
that appeared to have developed during the puerperium. For many years previously,
however, she had had isolated episodes when she went berserk and threw saucepans
and crockery about her flat. Othenvise she was a conscientious housewife, notably
house-proud, excessively shy, and attached to her own mother with pathological
intensity. She was thirty-five a t the time of referral.
 518 JOHN BOWLBY

for redecorating the flat in which the old people had lived together
and why too she had continued to postpone having her own flat
redecorated: it was vital, she felt, that when at last her father did
return he should find the places familiar. ,Nevertheless she was
not deluded. Side by side with the dynamic system concerned to
recover her absent father was a recognition that he was dead and
never would return.
Among similar patients described i n the literature, Krupp
(30) reports the case of another mother, also i n her early thirties.
Here again there were difficulties with the children, and it was
on this account that she had been referred to a social agency. At
the age of fifteen years this woman had lost her father overseas.
A few months later she had become promiscuous, but shortly
afterward had married. From what she said it seemed that, from
the first, she had refused to believe her father was dead and so
had experienced no grief. For the next eighteen years, moreover,
she had persisted in the illusion that her father might still be
alive; accordingly she was forever searching for him, for example,
i n restaurants and other public places. For a time she had had an
intensely emotional but platonic extramarital relationship with a
man who reminded her of her father; and at the time of referral
she was still excessively preoccupied in fantasy with him, to the
neglect of her'children. I n the course of casework, she began to
accept the idea that her father was really dead and for the first
1
time began to cry and go through the work of mourning. Follow-
ing this, fantasies of her extramarital relationship diminished and
she developed a more satisfactory relationship with her chiIdren.
I n the method adopted for meeting loss the mental state of
this patient closely resembles that of Mrs. Q.By both of them the
object is recognized to be missing but is judged nevertheless to be
recoverable. Given energy and perseverance, the husband or father
can be found again and reunion attained. There is loss admittedly,
but it is only temporary. I t is the permanence of the loss that is
denied.
A variant of this is when the bereaved develops a fantasy
that, despite everything, he is still in touch with the object. Pol-
lock (35) refers to three such cases, all adults, and each of whom
 PATHOLOGICAL AND CHILDHOOD MOURNING 519

had suffered death of a parent prior to the sixth b i r t h d a ~ One

.~
woman’s mother had died when she was four and a half years old.
“This woman actually retained her mother in heaven as a live
figure. Each night she spoke to her, and the patient knew that
mother heard her.” She had been unable to mourn, states Pollock,
because that would have meant that her mother was dead. Pollock
notes that “in all three instances the patients denied ever visiting
the cemetery where the deceased parent is buried, and [that] there
was a period of amnesia that extended from the moment when the
patient was told of the death of the parent until many months
later.”
Nothing is more important for psychopathology than an
understanding of the precise conditions internal to the patient
and external to him that promote splits of this kind to occur in
the ego. While Freud does not give particulars of the external
circumstances surrounding the deaths of the parents in the child-
hoods of his patients, in the cases reported by Pollock and Krupp
and in that of my own patient, Mrs. Q., they were known. I n all
five cases the loss had not only occurred in the absence of the
bereaved but had been sudden. If, as I suspect, these conditions
do in fact promote splits, it may be a step to understanding why
such splits are so apt to occur in childhood; for young children
are usually banished from the scene of death and, since they have
little foresight, the event of death is almost bound to be experi-
enced by them as sudden.
There is no problem, however, in understanding why these
splits, once begun, persist and why they are kept so secret. To
confess to another belief that the object is still alive is plainly to
court the danger of disillusion. Only by refusing to match the
belief with the beliefs of others can the illusion of the object’s
presence, immediate or potential, be retained. Every intimate
relationship which may tempt the sharing of precious memories
and hopes is therefore fraught with risk. No wonder patients of
this kind find analysis and all other close relationships such a
threat.
5One was two and a half years old when his father died and was later in an
orphanage for a time. A second was five years old when her father died. The third,
referred to above, was four and a half years old when her mother died. (Personal
communication from Dr. George Pollock.)
520 JOHN BOWLBY

As with the other variants of pathological mourning that

have been described, denial that the object is permanently lost
is consistent with and contributory to a number of different clini-
cal syndromes. Mrs. Q was predominantly depressive though she
had also much manifest anxiety and occasional outbursts of hysteri-
cal behavior. T h e mother reported by Krupp seems to have been
more psychopathic. One of Pollock’s patients suffered principally
from depression, another from anxiety, and a third alternated
between periods of each. All five were married but, except for
Mrs. Q., had severe marital problems.
Although in the literature there has been a tendency to link
denial especially with manic conditions, this clearly needs quali-
fication. Denial that the object is permanently lost can be com-
bined with. many other defense mechanisms and so lead to a
variety of clinical conditions. Depression and also wandering,
as in Krupp’s case, are common. Indeed, as Stcngel (41, 42, 43)
has shown so clearly, i t seems likely that a large proportion of
cases i n which pathological wandering is a feature belong here:
these people are searching, sometimes consciously, sometimes
unconsciously, for the lost figure they believe still exists some-
where. I t is when differentiation between self and object is blurred
or absent and is replaced by unconscious identification, I suspect,
that a manic condition develops. I n such cases there is often not
only strong identification with the lost object, but also, by means
of a demonstration of activity and joy, a brave effort to deny the
possibility that the lost object is really gone. This view is consist-
ent with that of many analysts, e.g., Helene Deutsch (9), Linde-
mann (33), and Lewin (32).
We are now in a position to compare the denial that the
object is lost with the repression of yearning with which w e began.
I n both types of case one part of the self believes the lost object
to be alive and recoverable. When there is denial of loss, it is
believed consciously; when there is repression of yearning, it is
believed unconsciously. Here lies the difference between the two.
It is a parameter that permits of many degrees of difference, from
complete unconsciousness of yearning, as in the case of Helene
Deutsch’s male patient, to complete consciousness as i n the case
of Mrs. Q. As a result cases fall on a continuum between the
 PATHOLOGICAL AND CHILDHOOD MOURNING 52 1

two extremes and present a variety of clinical pictures. Irrespec-

tive of which end they tend to lie toward, however, they share one
feature in common: grief is only partially experienced because
loss is only partially admitted.

Comparison of Pathological with Childhood Mourning

In a previous paper (4), a systematic comparison was made
between responses to loss as they are commonly exhibited on the
one hand by adults and on the other by infants and young chil-
dren. T h e conclusion reached ran as follows: “Since the evidence
makes it clear that at a descriptive level the responses are similar
in the two age groups, I believe it to be wiser methodologically
to assume that the underlying processes are similar *also, and to
postulate differences only when there is clear evidence for them.
T h a t certain differences between age groups exist I have little
doubt, since in infants and small children the outcome of experi-
ences of loss seems more frequently to take forms which lead to
an adverse psychological outcome. I n my jud-went, however, these
differences are best understood as being due to special variants of
the mourning process itself, and not to processes of a qualitatively
different kind.”
Having now examined the nature of the processes at work
in mourning and some of the common pathological variants of
these processes we are in a position once again to make a com-
parison. This time it will be between the pathological variants
described in the earlier part of this paper and the responses to
loss that are commonly to be seen in infancy and early childhood.
TYhen this is done the conclusion is reached that many of the
features that are characteristic of one or another pathological
variant of mourning i n adults are found to be almost the rule in
the ordinary mourning responses of young children. Yearning for
the lost object and also angry reproaches against it readily become
unconscious; while vicarious objects are often cared for. Only in
respect of denial of permanent loss is direct evidence lacking. Let
us consider them one by one.
I n the earlier paper I described some of the common patterns
of response that were to be observed when healthy young children
522 JOHN BOWLBY

between the ages of about six months and three or more years
are removed from their mother figures and placed with strangers.
I n particular I emphasized that yearning for the lost object is
not only common but also that it is both more intense and more
prolonged than is generally supposed. Its relative neglect is no
accident, however. On the one hand, the adults in charge of a
child find this yearning painful to perceive; on the other, it is
subject to processes of distortion and repression which quickly
obscure the scene. Thus Laura, aged two years four months, the
subject of Robertson’s film A Two-Year-Old Goes to HospitaZ,
“mould interpolate without emotion and as if irrelevantly the
words ‘I want my hfummy, where has my Mummy gone?’ into
remarks about something quite different” (37, 38). At other times
the wish was displaced to another object, as when she chanted,
“I want to see the steam roller, I want to see the steam roller,
I want to see my hlummy, I want to see the steam roller.”
Distortions of these kinds start early in the separation period
and progress fairly rapidly to a state in which it is only the sensi-
tive observer who will notice them. For example, it will be recalled
how a small boy, Patrick, aged just over three years, after he had
been i n the Hampstead Nurseries for some weeks, “would stand
somewhere in a corner moving his hands and lips with an abso-
lutely tragic expression on his face.” By this time the movement
of hands and lips had ceased to signify anything to anyone unfa-
miliar with the history of their development. Those who were
familiar with it, however, knew that the lip movements were the
remnants of a formula that earlier he had repeated over and over
again-of how his mother would come for him, would put on
his overcoat, and take him home with her; and that the hand
movements were the fragments of a mime in which he had dem-
onstrated how she would put on his overcoat, zip up the zipper,
and put on his pixie hat.s T h e yearning for the lost object per-
sisted but was in course of undergoing repression and so of becom-
ing unrecognizable. A record of this kind is of special value for
the light it throws on cases i n which the process has already pro-
gressed far, for example, the male patient reported by Helene
6 The material on both Laura and Patrick is reported in (4, pp. 22-23).
 PATHOLOGICAL AND CHILDHOOD MOURNIKG 523

Deutsch who had ceased to yearn for his mother and had instead
taken to leaving his bedroom door open in the hope that the large
dog would come to him. In all such cases the yearning persists;
but the object of it becomes obscured and changed, and the yearn-
ing itself may become repressed and unconscious.
JVhen eventually this repressed yearning for the mother comes
into the open again the full force of the affect and of the motiva-
tion of which it is a part becomes apparent. In the adult this will
happen in the course of successful analysis. Angry reproaches for
being deserted will alternate with sorrowful yearning for the
object lost, a sequence that is likely first to be experienced within
the transference. In the case of young children whose separation
has been of limited duration the affect and motivation will become
apparent soon after return home. If he has been away only a few
weeks, the usual sequence is as follows: initially he seems strangely
remote from his mother. In spite of the angry protests at her
departure and tearful yearning for her return that he showed
following separation, now that he has recovered her he seems
unmoved. He may fail to recognize her or simply avoid her;
alternatively he may respond to her as though he were an auto-
maton. Sooner or later, however, his remoteness ceases and is .
replaced by clinging, often accompanied by a torrent of tears and
reproaches. Thenceforward he insists on never letting his mother
out of his sight. Below the falade of remoteness the yearning for
the lost object has lived on as a dynamic, ready for active expres-
sion as soon as external conditions permit. It is a mental state
identical with that seen in the adult patient in all features save
its reversibility: whereas in the adult patient suffering from patho-
logical mourning w e know it to be very slow and difficult to
reverse, in the young child, provided the separation has been
limited to weeks only and has not been repeated, it usually reverses
spontaneously after a few hours or days.
. It is because the yearning for the lost object remains an active
dynamic that I find myself unable to agree with the conception
of the process recently proposed by Anna Freud (15). In her com-
ments on my earlier paper, while agreeing that the responses to
separation can usefully be divided into three phases, protest,
524 JOHN BOWLBY

despair, and detachment, she calls in question my use of a term7

that implies a defensive process. I n its place she proposes the term
“withdrawal,” because, she holds, it covers both the manifest
behavior and “the internal process of libido ,withdrawal by which
we believe this behavior to be caused.” This picture of the proc-
esses at work, however, does not square with the observed data.
Instead of the child’s attitude to his mother developing again
gradually and steadily as such a formulation would lead us to
expect, it sometimes erupts with a suddenness and intensity which
makes it plain that it has been potentially active all along. It is
on this evidence that the hypothesis is advanced that, when a
young child is in a phase of detachment, strong defensive processes
are at work: in the terminology commonly in use, there is counter-
cathexis not withdrawal of cathexis. I t is of interest that in an
earlier publication (jointly with Dorothy BurIingham [IS]) Anna
Freud has herself advanced a view of this kind. Since, however,
this issue is more closely connected with the theory of defense
than with the present topic, a full discussion of i t is postponed
to a later publication.
Not only is there this evidence of a persistent but unconscious
yearning.for the lost object during the detached phase of separa-
tion,8 but there is evidence also that after reunion a residue of
reproach larger than is suspected may remain. This was shown
dramatically by Laura six months after her return home, by which
time she had seemed to be her normal self. Late one evening when
her parents were viewing a rough version of the film, through a
mischance Laura came in to see the final sequences. When the
lights went up she was found to be agitated. Suddenly she flushed,
turned angrily to her mother and exclaimed, “Where was you all
the time, Mummy, where war you?” Then she burst into loud crying
and turned to her father for comfort. Not unnaturally her parents
were astonished by the strength of the feelings she revealed and
were disturbed also that she turned angrily away from her mother.
?Although in her comments Anna Freud is refemng to my earlier formulation
in which I employed the term “denial” instead of “detachment,” the change of
term does not materially alter the burden of her criticism.
8The evidence is clearest in children who have not yet progressed far into the
detached phase and whose feelings for their mother, therefore, emerge fairly readily
after reunion.
 PATHOLOGICAL AND CHILDHOOD MOURNING 525

This episode illustrates not only the length of time after a

separation that the feelings aroused by it may persist in latent
form but also how much anger is elicited. Here Laura’s anger
took the form of reproaches against her mother. At other times
the anger is directed toward third persons. An example occurred
just as Laura and her mother were departing from the hospital.
Before leaving the ward, Laura had insisted on taking all her
possessions home with her; even a tattered old book she refused
to leave behind. On the way out she accidentally dropped this
book and a nurse, trying to be helpful, picked it up. At this Laura
screamed in temper and snatched the book away-the fiercest
feeling she had shown during her whole stay.9 T h e function of the
outburst was clear: insistence that she was going home with her
mother and refusal to be detained in the hospital by the nurse.
I t is an anger comparable to that of bereaved persons who feel
that would-be comforters are trying to rob them of the object they
have lost. hlrs. Q.’s anger with me for not approaching the hospital
authorities to arrange for her father’s return is not very different.
I n these examples the function of the anger is clear; in
others it is more obscure. As I have repeatedly emphasized, young
children separated from their parents in a hospital or residential
nursery are notoriously destructive. Often it seems to the onlooker
to be both wanton and pointless. If the hypothesis I am advancing
is right, however, it is neither. It is simply an expression, gravely
intensified and distorted, of the child’s anger with his mother for
deserting him and anger with third parties for detaining him.
As such it is the equivalent in childhood of the intense, persistent,
and misdirected anger of the melancholic patient and, though
expressed in a very different way, of the affectionless psychopath
also.
So far, in drawing the comparison I have concentrated atten-
tion on the two variants of pathological mourning first described,
repressed yearning for the loved object and repressed reproaches
against it. Each, it is seen, is of regular occurrence in childhood.
T h e third and fourth variants occur also but are perhaps more
limited in incidence.
Of the third variant, care for a vicarious figure, once again
Q Accounts taken from Robertson (38).
526 JOHN BOWLBY

Laura provides an illustration. Although she herself cried little,

Laura was much concerned when other children cried. On one
such occasion a small boy was screaming piteously. Laura’s imme-
diate response was to become solicitous and to demand that the
boy’s mother be brought. A little self-righteously she exclaimed,
“I not crying, see!” and then emphatically, “Fetch that boy’s
Mummy!” A couple of days later she exhibited a related response,
directed toward another figure. T o a nurse who was talking to her
she insisted, “My hfummy’s crying for me-go fetch her!” Here
Laura’s grieving was projected onto the lost mother herself. This
is a mental process identical, I suspect, to that underlying a belief
that “lost souls” are searching for their loved ones (14), to which
further reference is made in the next section.
Care for a vicarious figure appears to be especially common
when siblings suffer loss together. One of them usually mothers
the others. Though probably the elder most frequently fills this
role, Robertson has recorded cases when it was the younger. For
example, another little girl, Paula, who like Laura was under two
and a half, when taken away by ambulance with her weeping
brother, two years her senior, put her arm round him and said,
“I’ll look after you.”
I t seems not unlikely that girls and women respond i n this
way more often than do boys and men. Both the illustrative adult
cases reported earlier were of women. That it is a response that
can be evoked early in a little girl’s life is brought home to us by
both Laura and Paula, and also by observations made in the
Hampstead Nurseries during the war (17, pp. 32-33). Plainly in
these children, as in adults, the care of a vicarious figure is a
response alternative to seeking care from the lost object for them-
selves.
This brings us to the fourth pathological variant of mourn-
ing, denial that the object is permanently lost. Because all the
children whom Robertson has observed had lost their mothers
only temporarily, he has had no opportunity to observe a develop-
ment of this kind; the same is true of my other colleague, Chris-
toph Heinicke. That denial can originate during the third year
of life, however, is attested by Freud in his paper on “Fetishism”
(22). One of the two young men, each of whom “had refused to
 PATHOLOGICAL AND CHILDHOOD MOURNING 527

acknowledge the death of his father,” is reported to have lost his

father when he was two years of age. (The other was ten.) Of the
patients reported by Pollock, all three had lost the parent in
question before the sixth birthday. There can be no doubt, there-
fore, that this variant too originates in childhood even though it
may do so less regularly than the first two variants.
Thus we can conclude that responses similar to the first two
pathological variants of adult mourning are the rule when young
children lose a loved object; that the third, care for a vicarious
figure, is common especially in girls; and that the fourth, denial
that the object is permanently lost, undoubtedly occurs though
with what frequency is not known. These are findings, I believe,
that go far to explain why losses occurring in infancy and early
childhood have a pathogenic potential. Should loss prove pro-
longed or permanent, it is now plain, there is great danger that
one or more of the pathological variants of mourning will not only
be set in train but, once present, will persist, and that as a con-
sequence the child’s object relations will come to be permanently
organized on this pattern. Even when loss is only temporary danger
remains. Once one or more of these pathological responses to loss
has been active in a personality there is increased likelihood of
its being evoked afresh following further loss. This, I believe,
accounts for the development of at least some personalities prone
to respond to losses in later life by pathological mourning.
Nevertheless not every child who experiences either perma-
nent or temporary loss g r o w up to be a disturbed person. T o
understand why in some cases the experience is pathogenic and
in others not so requires that the conditions which favor reversal
and those which promote an overt or latent persistence of the
pathological patterns be identified. Among those that are probably
relevant in cases of temporary loss are the length and frequency
of the separation, the conditions of care during it, and the attitude
of the mother toward the child after his return, especially perhaps
the extent to which she accepts and meets his demands for her
presence or, on the contrary, forbids and punishes them. T h e
phase of the child’s development and the nature of his previous
relationship with his mother, while obviously relevant, may each
be of less significance than is usually believed by analysts.
528 JOHN BOWLBY

Review of Literature
I n an early section of the previous paper (5) I discussed seven
themes in the theory of mourning about which there had been
controversy in the psychoanalytic literature.. In considering the
sixth, ways in which pathological mourning is believed to differ
from healthy, I gave reasons for not accepting the criteria for
pathological mourning that have commonly been advanced,
prominent among which has been anger and reproach against the
lost loved object; and in later parts of that paper and earlier
parts of this I have explored other criteria. It is time now to
consider what place has hitherto been given in the psycho-
analytic literature to the first of the new criteria proposed, the
inability to accept and express yearning for and angry reproaches
against the lost object.
A reading of the Iiterature shows that an unconscious effort
to recover the lost object has not previously been singled out as
being, in most pathological variants of mourning, the principal
dynamic at work. Naturally it has not escaped attention altogether,
and there are indeed many references to it. Nevertheless, although
a few are explicit, a majority are only implicit, and it is striking
in how many formulations it is virtually absent.
Among many possible reasons for this neglect three stand
out. One is the overwhelming emphasis given by Freud to identi-
fication with the lost object as the key concept in mourning, an
emphasis that has obscured processes which I believe to be more
fundamental. A second reason is the tendency of much analytic
theorizing to be concerned, not with motivation and with objects
and situations in the external world toward which motivation is
directed, but with quantities of excitation and with affects as con-
ditions that lead these quantities to become excessive. As a result
there are plentiful references to affects, such as unconscious sad-
ness and anger and unresolved grief, and correspondingly few to
unconscious striving to recover a lost object and to unconscious
reproach designed to discourage repetition of loss.
A third reason for this neglect is the weight given by Melanie
Klein and her school to the bereaved’s sense of guilt, arising, it is
held, from his belief that he was the agent of the loss, and to her
 PATHOLOGICAL AND CHILDHOOD AIOURNING 529

emphasis also on his desire in consequence to make reparation.

Although the concept of reparation bears some relation to the
concept of striving to recover the object, the two are different:
reparation springs from a sense of guilt, the effort to recover from
a sense of loss. Each is important, but they must be kept distinct.
Let us now consider how individual analysts have treated
this theme. I n his discussion of the dynamics of healthy mourning,
it will be recalled, Freud attached much importance to persistent
and insatiable yearning for the lost object. I n Inhibitions, Symp-
toms and Anxiety, for example, he emphasizes the unappeasable
nature of bereaved longing and attributes to it the painfulness of
mourning and the onset of defensive processes (21, pp. 171-172).
Yet in his discussion of the dynamics of pathological mourning
as it manifests itself in melancholia there is little reference to this
longing. Indeed, by suggesting that the object is abandoned, he
seems to take an almost opposite view. I n the case of melancholia,
he maintains, “the object-cathexis proved to have little power of
resistance and was brought to an end. But the free libido was not
displaced on to another object; it was withdrawn into the ego.
There [it] . .. served to establish an identificution of the ego with
the abandoned object” (19, p. 249). Ever since this time there has
been preoccupation with identification as the key concept i n
. mourning.
At about the same date Freud makes a brief reference to the
inability of the Wolf Man to mourn his sister (20, p. 23). This
sister was two years older than the patient and was for him “the
most dearly loved member of his family.” She had committed
suicide in her early twenties. Nevertheless, “when the news of his
sister’s death arrived, so the patient told me, he felt hardly a trace
of grief. H e had to force himself to show s i p s of sorrow, and was
able quite coolly to rejoice at having now become the sole heir
to the property.” Analysis, however, showed that soon afterward
there had been a “substitute for the missing outbursts of grief”:
he had made a pilgrimage to the burial place of a great poet, to
whom his sister had’been compared, and had shed bitter tears on
the grave. I n this account, it will be noticed, though Freud recog-
nizes the repressed and displaced grief, it is at no point linked
530 JOHN BOWLBY

to motivation. An unconscious desire to recover the object lost is

not referred to.lo
In his published papers the nearest Freud seems to have
come to a recognition of the role of striving to recover the lost
object is in the paper on “Fetishism” in which he referred to the
two young men each of whom had refused to acknowledge the
death of his father. Nevertheless, these cases belong to the cate-
gory of “denial that the object is permanently lost” rather than
to “unconscious urge to recover the object.” In a letter of 1921 to
IVeiss, however, Freud does seem to refer to it (44). Weiss had
reported on the case o€ a man of twenty-seven who had been sent
for analysis by his mother because of intense apathy. T h e patient
himself described how he felt himself “separated from the outside
world by an impenetrable isolating layer.” IVeiss had diagnosed
it as “a case of simple depression without melancholic trait.” Freud
concurred. “The affection is little studied,” he continues, “I
should say, as a surmise, that it is a matter of a simple fixation of
high degree on the mother, whom he rejects from time to time,
so that nothing then is left him.” From the rather meager details
given it appears that the dynamics of the case were the same as in
Helene Deutsch’s case, described earlier, of the man whose un-
conscious longing for his mother was manifest only in his fantasy
of the big dog visiting him.
Edith Jacobson’s position resembles Freud’s. Like him she
has, on the one hand, drawn attention to the role that yearning
to retain the lost object plays in normal sadness and grief and,
on the other, given it relatively little place in her account of
pathological mourning. Yet, as with Freud, it is a concept which
is readily compatible with her theoretical position. This is well
illustrated in her paper “Denial and Repression” (27), in which
she reports a case that seems, like TVeiss’s, to have much in com-
10 In view of Freud’s current interest in mourning and its vicissitudes, including
his remark in the Introductory Lectures that “there are neuroses which may be
described as morbid forms of grief,” it is curious that the patient’s pathological
grief response, which is described early in the case history, is not once referred to
again, and that it is accounted for by Freud without reference either to any previous
experience of loss the patient may have had or to the patient’s relation to his
mother and mother substitutes, about which some fragmentary and suggestive
information is given (20, pp. 13-15).
 PATHOLOGICAL AND CHILDHOOD MOURNING 53 1

mon with Helene Deutsch’s case of the man with the dog fantasy.
T h e man reported by Edith Jacobson is described as similarly
detached in all his relationships; in addition he maintained that
his only feeling toward his mother was one of “cold resentment.”
Progress in’his analysis, Edith Jacobson reports, was signaled by
“waves of affection toward his mother [which] were permitted to
appear and to replace his symptoms.” Though Edith Jacobson’s
terminology is one of affect not motivation, the one is easily trans-
lated into the other. I n terms of motivation, it is evident, these
“waves of affection toward his mother” are equivalent to a longing
for her; and it is this longing which had been unconscious until
released during the analysis.
Edith Jacobson’s case follows the pattern of Helene Deutsch‘s
i n another respect also. I n both cases the disturbance was thought
by the analyst to have had its origin in a separation experience
of early childhood. I n commenting on her case Edith Jacobson
expresses the opinion that “the trauma that had broken up his
relationship to his mother” had occurred when the patient was
three and a half years old. At this time his mother had had a rnis-
carriage. His father was absent and, although his mother remained
at home, she “had been physically sick and depressed for many
..
weeks after. . Left in the hands of a maid, he had felt lost and
confused.”
T h e theoretical position taken by Helene Deutsch in her
paper “Absence of Grief” (10) is not dissimilar to that of Edith
Jacobson: once again there is implicit recognition that uncon-
scious longing for the lost mother is at the heart of the psycho-
pathology. For example, when discussing the man with the dog
fantasy, Helene Deutsch remarks: “apart from this one revealing
fantasy there was no trace of longing or mourning for his mother”
(who, it will be remembered, had died when he was five years
old). Despite this reference in the case report to longing as an
unconscious motive, however, in the notes at the end of the paper
theory is cast in the terminology of affect. “I believe that every
unresolved grief is given expression in one form or another,” she
writes. I “am convinced that the unresolved process of mourning
as described by Freud must be expressed in full.” Referring to
“suppressed affect following loss,” she continues: “TVe must as-
532 JOHN BOWLBY

.
sume that the urge to realization succeeds under the impetus of
an unconscious source of affect-energy exactly as in the case of
the criminal who is at the mercy of his guilt feelings. I suspect
that many life stories which seem to be due to a masochistic
attitude are simply the result of such strivings for the realization
of unresolved affects.”
This mode of theorizing, as Helene Deutsch herself points
out, is not unlike that employed by Freud in his reference to the
IYolf Man’s inability to mourn his sister. It is a mode of theorizing,
moreover, that has been followed by both Landauer (31) and
Lindemann (33), each of whom has taken mourning and its
pathological variants as his focus of attention. I n each case a use
of the language of affects has had the result that in their formula-
tions the unconscious longing that is part of the urge to recover
the lost object, although implicit, has tended to be obscured.
More explicit references to the longing for the lost object
are to be found in a contribution by Anna Freud (14) and also
in a number of clinical writings that report on cases of compulsive
wandering, depression, and suicide. I n a paper “About Losing and
Being Lost” Anna Freud discusses the connection between wishes
to possess and tendencies to lose as they refer to parent figures
and to material possessions respectively. In the third part she
discusses dreams of the return of the dead during phases of mourn-
ing and likens them to “the well-known folklore of the ‘lost
souls,’ i.e., dead people who are supposed to find no rest and [to]
wander at night in search of their former loved ones. I n both
cases,” she comments, “a libidinal process in the mourner, i.e.
longing, appears as projected into the image of the lost object.”
So long as the ghost is believed searching for his lost loved ones
the bereaved continues mourning. I n other words, mourning ends
when conscious or unconscious searching stops.
T h e truth of this generalization is well illustrated in the
papers by Stengel on compulsive wandering, a symptom, he found,
that is commonly associated with pseudologia, episodic depression,
and impulses to commit suicide. In his reports of thirty-six cases
(42, 43) Stengel draws special attention to two closely connected
features. The first is the high frequency of serious disturbance
in the patients’ relation to their parents in childhood, in par-
 PATHOLOGICAL AND CHILDHOOD MOURNIXG 533

ticular losses due to death or separation. T h e second is the desire

to seek the lost parent that is often present during the actual
episodes of wandering. “Almost all these patients had suffered
consciously from the failure of the normal child-parent relation,”
he writes. “Many felt even in childhood that they had missed
something which could never be replaced. I n a number this feel-
ing became particularly acute during their periodic depressions,
i.e. at the time when the wandering compulsion arose. A few be-
came conscious of the desire to seek for the dead or absent parent.
Some imagined immediately before or in this state that the dead
parent was not really dead but alive, and perhaps to be met in
their wanderings” (41).
Though Stengel gives a central place to the unrealistic urge
to seek a lost parent, he does not recognize it as a sequel of child-
hood mourning having taken a pathological course. As a result,
whereas elsewhere in the literature we find attention to mourn-
ing and its pathology but without reference to the unconscious
urge to recover the object, in Stengel’s work we find attention to
the urge but without reference to mourning and its pathology.
Elsewhere in the clinical literature Gero (24), in describ-
ing two patients suffering from “neurotic depression bordering
on melancholia,” emphasizes both the unconscious longing for
mother and her care and also the unconscious rage against her
for not having provided it that was present in both of them.
Analytic progress, he found, turns on the patient becoming con-
scious “that he desires the breast-that is to say-the mother,””
and also on his recognition that his rage is directed against her
for having neglected him. I n both the cases he reports there was
evidence that they had in fact been neglected. One patient is
described as having been an unwanted child, “starved of love.”
T h e other had been sent to a residential nursery until he was
three years old: “he seems never. to have forgiven his mother for
this separation.”
Weiss (44) in his discussion of different forms of depressive
11 Gerii’s conception of the place of orality resembles my own: “The importance
of the oral experience in infancy,” he writes. “lies in my opinion in the mother-child
relationship.... The spccifically oral pleasure is only one factor in the experience
satisfying the infant’s need for warmth. touch, love and care.”
534 JOHN BOWLBY

illness also refers to the yearning for a former love object. I n his
view, however, it is not so much the lost mother who is longed
for but “the parent or sibling of the opposite sex.” Felix Deutsch
(8) reasons on the same lines.
I n Melanie Klein’s writing and that of many of her followers
the emphasis is so firmly on the sense of guilt and the need to
make reparation that the sense of loss and the striving to recover
the lost object tend to be obscured. Indeed in his comprehensive
discussion of the literature on the theory of depression, particu-
larly as it supports hlelanie Klein’s views, Rosenfeld (40) omits
all reference to it. Nevertheless it is not missing from Melanie
Klein’s own work. For example, she describes how a male patient,
unable to experience sorrow, during his analysis “increasingly
experienced the grief and longing for [his mother] which he had
repressed and denied from his early days onwards.” Earlier in the
same paper she refers to the need for the ego “to develop methods
of defence which are essentially directed against the ‘pining’ for
the lost object” (28).
These same concepts underlie her interpretation of the per-
sonality of Fabian, the hero of a story by the French novelist
Julian Green (29). Fabian is represented as an unhappy dissatis-
fied man, full of resentment, incapable of sympathy, able to make
only brief and frustrating liaisons with girls, full of insatiable
desire for wealth and success and intense envy for those who
possess more than himself. I n the course of the story Fabian is
enabled to assume the personalities of other people. I n one such
personality he murders the girl, Berthe, whom the man he has
now become was trying, unsuccessfully, to win. I n another he
becomes fascinated by the eyes of a second girl, Elise, who has for
long been hopelessly in love with the man he now is. Her eyes, he
finds, have “in them all the tragedy of a longing that can never
be satisfied”; and at that moment he realizes they are the eyes of
Fabian. Finally, he becomes himself again and, with his mother
sitting by his bedside, is overcome by longing to be loved by her
and to be able to express his love for her, which he can never do.’
In the light of her experience with patients like Fabian and
drawing on the other dramatic details of the story Melanie Klein ’

builds up an interpretation of Fabian’s personality and the mode

 PATHOLOGICAL AND CHILDHOOD MOURNIKC 535

of development that lay behind it. A main component of it, she

suggests, is that all his life he had been “unconsciously searching
. . . for the ideal mother whom he had lost,” and unconsciously
expressing resentment at the deprivation he had felt “in the
earliest feeding relation.” TVhile the resentment is expressed in
the murder of Berthe, which Melanie Klein suggests was “an
expression of the infantile impulses to murder the mother,” the
longing is to be found in the longing eyes of Elise, the girl whose
eyes were Fabian’s. Elise, she suggests, “represented the good part
of his self which was capable of longing and loving.” “The search
for the lost ideal self,” she continues, “inevitably includes the
search for lost ideal objects; for the good self is that part of the
personality which is felt to be in a loving relation to its good
objects. T h e prototype of such a relation is the bond between
the baby and his mother. In fact, when Fabian rejoins his lost
self, he also recovers his love for his mother.” I n this interpreta-
tion it will be seen, although Melanie Klein places more emphasis
on orality and the feeding relation and, by implication, the first
year of life than I would, her picture of the dynamics underlying
this type of personality resembles closely the picture I am advo-
cating and also the ones that lie behind the thinking of Edith
Jacobson, Helene Deutsch, and Gero. By each author unconscious
longing for the loved object is thought to coexist with unconscious
resentment against it for its earlier neglect. One of the external
factors to which Melanie Klein attributes Fabian’s condition is the
same also: “his mother’s lack of affection.”
There are at least two other analysts, both influenced by
Melanie Klein, in whose thinking about pathological mourning
the role of unconscious yearning for, and reproach against, the
object can be discerned; in neither case, however, are they placed
at the center of the stage.
Referring to pathological self-reproach, Anderson (2) writes:
“These self-reproaches have their obverse for there is a plaint, an
accusation against the dead person for having left and deserted
the mourner.”’* It is a complaint, he points out, that enters also
12 In “Mourning and Melancholia” Freud remarks of melancholic patients that
“their complaints are really ‘plaints’ in the old sense of the word” (19, p. 248).
He does not discuss their origin or function, however.
536 JOHN BOWLBY

into certain peasant mourning rituals. Nevertheless, in his case

histories this insight is little exploited.
Throughout his writing TVinnicott shows a deep appreciation
of the place of sadness and grief in psychiatric disturbance. I n
one of his papers (45) he describes his approach to “an intelligent
girl of twelve who had become nervous at school and enuretic
at night. No one seemed to have realized,” he remarks, “that she
was struggling with her grief at her favourite brother’s death.
This little brother had gone away supposedly for a week o r two
with an infectious fever, but he had not come home immediately
as he developed a pain that turned out to be due to a tuberculous
hip. T h e sister had been glad with the rest of the family that he
was placed in a good tuberculosis hospital. I n the course of time
he suffered much more pain, and when at last he died of gener-
alised tuberculosis, she had been glad again. It was a happy release,
they had all said.”
“Events,” comments Winnicott, “had taken place i n such
a way that she never experienced acute grief, and yet grief was
there, waiting for acknowledgment. I caught her with an unex-
pected ‘You were very fond of him, weren’t you?’ which produced
loss of control, and floods of tears. T h e result of this was a return
to normal at school, and a cessation of the enuresis at night.”
I n another paper (46) Winnicott describes three sessions he
had with a boy of nine who was given to compulsive stealing.
Disturbances in Philip’s home life had begun when he was two
years old: and at six he had been sent to an uncle and aunt while
a sister was born. During the first interview Winnicott asked
him if he was ever sad. To this Philip replied that the worst
sadness had happened “a long time ago,” a phrase that referred
to the time of his sister’s birth. “My mother went away,” he
continued. “I and my brother had to live by ourselves. We went
to stay with my aunt and uncle. T h e awful thing that happened
then was that I could see my mother cooking in her blue dress and
I would run up to her but when I got there she would suddenly
change and it would be my aunt in a different coloured dress.”
Although TVinnicott links the stealing to Philip’s desire to find
the mother he had lost, in his commentary he seems to place more
weight on the birth of the baby, Philip’s jealousy of it and envy
 PATHOLOGICAL AND CHILDHOOD hfOURNIh’G 537

of his mother’s capacity to have it, than on the real loss that Philip
had suffered. Philip’s hallucination of his mother cooking, it can
be inferred, was the result of a split in the ego of the kind
described in the section, Denial That the Object Is Permanently
Lost.
Although striving to recover the lost object has never been
at the center of attention in theoretical formulations regarding
pathological mourning, it is not without interest to note the place
that yearning of other kinds has been given in psychopathology
generally. Many analysts from Abraham (1) and Rado (36) to
Melanie Klein have emphasized a fixation on the mother’s breast
and a longing to recover it. Others, including Freud (22, 23),
Helene Deutsch (9), and Anny Angel (3) have emphasized the
longing in female patients for a penis. Another line of theorizing,
favored among others by Rado (36) and Weiss (44),has explored
the notion that the missing object is wanted not so much for itself
but for its value in restocking the ego with “narcissistic supplies.”
A review of these hypotheses in the light of further knowledge
of object relations and of processes of mourning, healthy and
pathological, may lead to their being reappraised.
T h e resemblance that the mourning processes of childhood
bear to pathological variants of mourning in the adult is discussed
in the literature of only one main school of psychoanalytic the-
orizing, that of Melanie Klein. Inevitably it is missing from the
literature of those schools which either overlook the possibility
or reject the idea that mourning occurs in infancy and childhood.
I n the work of Melanie Klein and her colleagues, on the other
hand, the reality of childhood mourning and of its relatedness
to adult mourning is not only recognized but made a central pillar
of theory construction. T h e way in which hlelanie Klein treats
the connection’is at many points very different from the one pre-
sented here, however. I n the first place, her ideas rest not on the
empirical studies of the actual behavior of young children but on
reconstructions. I n the second, the picture she has of the dynamics
of mourning, whether healthy or pathological, in the adult or
the child, derives from a theoretical schema that is based on the
death instinct, the primacy of persecutory and depressive anxiety,
and the overwhelming importance of orality. Since in earlier
538 .TOHN BOWLBY

papers I have given reasons for not accepting this framework, no

purpose would be served in describing rvlelanie Klein’s conception
further. Nevertheless, trained in this school of thought as I have
been, I remain deeply indebted to it for an emphasis on the reality
of childhood mourning and its implications for understanding
psychiatric illness.
Conclusion
I n this paper I have had two aims. T h e first has been to delineate
the dynamics of .four of the commoner pathological variants of
mourning; the second to demonstrate that patterns of response
that are typical of these pathological variants in adult life resem-
ble closely the patterns that are typical of mourning processes
when they are evoked in infancy and early childhood. T h e impli-
cations of this finding for the theory of childhood pathogenesis
will be clear.
Once again the ground traversed has proved controversial.
As a result much of the paper has been concerned with clarifying
the dynamics of pathological mourning, especially the roles in it
of unconscious yearning and unconscious reproach. I n presenting
clinical material, therefore, I have deliberately selected that which
throws into relief the parts played by these two powerful motives;
often, it has seemed to me, the original case report has not given
them sufficient weight. This I believe to be due partly to their
role in healthy mourning not having been fully appreciated and
partly to the preoccupation with affect, guilt, and processes of
identification, which has marked so much of the psychoanalytic
theorizing about pathological mourning. There is no evidence,
we have seen, that Freud recognized the part played by the urge
to recover the lost object in pathological mourning, despite the
fact that he recognized its role in healthy mourning and also
detected that self-reproach is often a turning toward the self of
reproaches originally aimed at the object.
T h e function of reproach directed against the object seems
to have been little discussed. T h e hypothesis advanced here is
that it is to discourage the object from deserting again. I t is
within this functional context also, I believe, that both the
extraordinary bitterness and the persistence of the anger that
 PATHOLOGICAL AND CHILDHOOD MOURNING 539

follows prolonged or permanent loss are to be understood. When

separation is only temporary opportunity soon comes for the anger
and reproach felt toward the deserting object to be expressed and
for love to be revived. When separation is prolonged or perma-
nent, anger and complaint fester while love lies dormant. Though
therapeutic technique is outside the scope of these papers, the
therapeutic implications of these hypotheses are not negligible.
What they imply is that a principal goal in the treatment of many
sorts of patient, including some who are anxious and depressed
and others who suffer from various forms of character disorder,
is to assist them to experience each one of the impulses that are
directed toward recovering and retaining the lost object and of
the ambivalent feelings associated with t h e m - o n the one side the
urge to recover, with its accompanying yearning, weeping, and
sadness; and on the other the complaints and reproaches with
their accompanying anger and bitterness.
If this account of the dynamics of pathological mourning is
valid, their resemblance to patterns of response commonly seen
following loss in early childhood is unmistakable. Although
Melanie Klein has also laid much emphasis on the relatedness of
childhood mourning to adult mourning, the theoretical context
within which she has drawn the comparison is one that in earlier
papers I have given reasons for rejecting. I n none of the details of
her exposition, therefore, do I follow her.
I n the case histories recounted in the course of the paper,
all but one derived from the writings of others, pathogenesis has
again and again been attributed by the analyst reporting them to
a loss, temporary or permanent, that occurred in the patient’s
early years. JVhat appears in each case to have happened is that
processes of mourning, having been evoked in early childhood,
took a pathological turn in a way typical of them in that period
of life; and that, over the years, the pathological processes have
either persisted substantially unchanged or, having temporarily
modified, have been evoked afresh by a further loss. T h e extensive
evidence that shows that loss of a parent in the early years is
especially frequent in persons who later suffer from psychiatric
illness is reviewed elsewhere (6).
Repeatedly during the discussion of pathological mourning it
540 JOHN BOWLBY

has been necessary to refer to defense. I have described repression

of yearning for the lost object, displacement of reproaches against
it, denial of the reality of loss, splits in the ego, the role of pro-
jective identification, and the use as defenses -both of aggression
and of the care of a vicarious figure. Loss, it is clear, can readily
evoke defense. Furthermore, observation of young children who
either temporarily or permanently lose their mothers furnishes
us with information of how some of these defenses originate. Do
these observations, we may ask, throw fresh light on the nature of
defensive processes? It is my belief they do, and it is to exploring
this possibility that further publications will be directed.
BIBLIOGRAPHY
1. Abraham, K. A short study of the development of the libido, viewed in the light
of mental disorders (1924). In: Selected Papers on Psycho-Analysis. London:
Hogarth Press, 1927. pp. 418-501.
2, Anderson, C. Aspects of pathological grief and mourning. Znt. J . Psychoanal.,
30:48-55, 1949.
3. Angel, A. Einige Bemerkungen fiber Optimismus. Znt. 2. Psychoanal., 20:191-
199, 1934.
4. Bowlby, J. Grief and mourning in infancy and early childhood. The Psycho-
analytic Study of the Child, 15:9-52. New York: International Universities Press,
1960.
5. Bowlby, J. Processes of mourning. Znt. J . Psychoanal., 42:317-340. 1961.
6. Bowlby. J. Childhood mourning and its implications for psychiatry. Amer. J.
Ps).chiat., 118:481-498, 1961.
7 . Breuer, J. k Freud, S. Studies on hysteria (1895). Standard Edition, 2. London:
Hogarth Press, 1955.
8. Deutsch, F. Applied Psychoanalysis. New York: Grune 8: Stratton. 1949.
9. Deutsch, H. Zur Psychologie der manisch-depressiven ZustZnde, insbesondere
der chronischen Hypomanie. Znt. 2. Psychoanal., 19:358-371, 1933.
10. Deutscli, H. Absence of grief. Psychoanal. Quart., 6:12-22, 1937.
11. Engel, G. Is grief a disease? Psychosom. A M . , 23:18-22, 1961.
12. Eliot, T. D. Bereavement: inevitable but not insurmountable. In: Family,
Afarriage and Parenthood, ed. H. Bedrer k R. Hill. Boston: Heath, 1955.
13. FIeming, J., Altschul, S., Zielinski, V.. 8; Forman, Af. The influence of parent
loss in childhood on personality development and ego structure. Read a t the
American Psychoanalytic Association, May, 1958.
14. Freud, A. About losing and being lost. Abstract in Znt. J . Psychoanal., 35283,
1954.
15. Freud, A. Discussion of Dr. John Bowlby's paper [Grief and Mourning in
Infancy and Early Childhood]. The Psychoanalytic Study of the Child, 15353-62.
New York: International Universities Press, 1960.
16. Freud, A. & Burlingham, D. War and Children. New York: International
Universities Press, 1943.
17. Freud, A. k Burlingham, D. Infants Without Families. New York: International
Universities Press, 1944.
18. Freud, S. Introductory Lectures o n Psycho-Analysis (1916-1917). London: Allen
k Unwin, 1922.
 PATHOLOGICAL AND CHILDHOOD MOURNING 541

19. Freud, S. Mourning and melancholia (1917). Standard Edition, 14243-260.

London: Hogarth Press, 1937.
20. Freud, S. From the history of an infantile neurosis (1918). Standard Edition,
17:3-122. London: Hogarth Press, 1955.
21. Freud, S. Inhibitions, symptoms and anxiety (1926). Standard Edition, 20:75-
174. London: Hogarth Press, 1959.
22. Freud, S . Fetishism (1927). Collected Papers, 5:198-204. London: Hogarth Press,
1950.
23. Freud, S. Splitting of the ego in the defensive process (1938). Collected Papers,
5352-375. London: Hogarth Press, 1950.
24. Gero, G. T h e construction of depression. Znt. J. Psychoanal., 17:423-461, 1936.
25. Greene, IV. A. Role of a vicarious object in the adaptation to loss: I. Use of a
vicarious object as a means of adjustment to separation from a significant
person. Psychosom. Ned., 20:344-350, 1938.
26. Jacobson, E. Normal and pathological moods: their nature and functions. T h e
PqchoanaZytic Study of the Child, 1273-113. New York: International Univer-
sities Press, 1957.
27.. Jacobson, E. Denial and repression. This Journal, 5:61-92, 1957.
28. Klein, M. hIourning and its relation to manic-depressive states (1940). In:
Contributions to Psycho-AnaZysis. London: Hogarth Press, 1948, pp. 311.338.
29. Klein, hI. On identification. In: New Directions in Psycho-Analysis, ed. M.
Rlein, P. Heimann, 8: R. E. hfoney-Ryrle. London: Tavistock, 1955, pp. 309-345.
30. Rrupp, G . R. T h e bereavement reaction. T h e Psychoanalytic Study of Society,
2:42-74. New York: International Universities Press, 1962.
31. Landauer, R. Aquivalente der Trauer. Znt. 2. Psychoanal., 11:194-205, 1926.
32. Lewin, B. D. The Psychoanalysis of Elation. New York: Norton, 1950.
33. Lindemann. E. Symptomatology and management of acute grief. Amer. J.
Psychiat., 101:141-148, 1944.
34. Pittendrigh, C. S. Adaptation, natural selection and behavior. In: Behavior
and Evolution, ed. A. Roe 8: G. G. Simpson. New Haven: Yale University
Press, 1958, pp. 390-416.
35. Pollock, G. H. Mourning and adaptation. Int. J. Psychoanal., 42341-361. 1961.
36. Rado, S. T h e problem of melancholia. Znt. J. Psychoanal., 9:420-438, 1928.
37. Robertson, J. A Two-Year-Old Goes to Hospital (film). London: Tavistock
Child Development Research Unit. New York: New York University Film
Library, 1933.
38. Robertson, J. Guide to the film A Two-Year-Old Goes to Hospital. London:
Tavistock Child Development Research Unit, 1938.
39. Root, N. A neurosis in adolescence. T h e Psychoanalytic Study of the Child,
12:320-334. New York: International Universities Press, 1957.
40. Rosenfeld, H. An investigation into the psycho-analytic theory of depression.
Itat. J. Psychoanal., 40:103-129, 1959.
41. Stengel, E. Studies on the psychopathology of compulsive wandering. Brit. J.
Med. Psychol., 18:250-254, 1939.
42. Stengel, E. On the aetiology of the fugue states. J. Ment. Sn'., 87:572-599, 1941.
43. Stengel, E. Further studies on pathological wandering. J . Ment. Sci., 89:224-241,
1943.
44. TVeiss, E. Clinical aspects of depression. Psychoanal. Quart., 1345-461, 1944.
45. Winnicott, D. JV. Shyness and nervous disorders in children (1938). In: T h e
Child and the Outside World, ed. J. Hardenberg. New York: Basic Books, 1957,
pp. 35-39.
46. IVinnicott, D. IV. Symptom tolerance in paediatrics: a case history (1953). In:
CoZIected Papers. New York: Basic Books, 1937, pp. 101-117.
Submitted February 9, 1962