The Psychoanalytic Study of the Child

ISSN: 0079-7308 (Print) 2474-3356 (Online) Journal homepage: http://www.tandfonline.com/loi/upsc20

Hospitalism

Rene A. Spitz

To cite this article: Rene A. Spitz (1945) Hospitalism, The Psychoanalytic Study of the Child, 1:1,
53-74, DOI: 10.1080/00797308.1945.11823126

To link to this article: https://doi.org/10.1080/00797308.1945.11823126

Published online: 13 Feb 2017.

Submit your article to this journal

Article views: 5

View related articles

Citing articles: 137 View citing articles

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=upsc20

Download by: [University of Rochester] Date: 15 January 2018, At: 04:08

 HOSPITALISM
An Inquiry into the Genesis of Psychiatric Conditions in
Early Childhood!
By RENE A. SPITZ, M.D. (New York)
Downloaded by [University of Rochester] at 04:08 15 January 2018

"En la Casa de Ninos Expositos el nino se va

poniendo triste y muchos de ellos mueren de
tristeza.' ,
(1760, from the diary of a Spanish bishop.)

I. The Problem

The term hospitalism designates a vitiated condition of the body

due to long confinement in a hospital, or the morbid condition of the
atmosphere of a hospital. The term has been increasingly preempted
to specify the evil effect of institutional care on infants, placed in
institutions from an early age, particularly from the psychiatric point
of view." This study is especially concerned with the effect of con-
tinuous institutional care of infants under one year of age, for reasons
other than sickness. The model of such institutions is the foundling
home.
Medical men and administrators have long been aware of the
shortcomings of such charitable institutions. At the beginning of
our century one of the great foundling homes in Germany had a
mortality rate of 71.5% in infants in the first year of life (1) .3 In
1915 Chapin(2) enumerated ten asylums in the larger cities of the
United States, mainly on the Eastern seaboard, in which the death rates
of infants admitted during their first year of life varied from 31.7%
to 75% by the end of their second year. In a discussion in the same

1. Preliminary report.
2. Hospitalism tends to be confused with hospitalization, the temporary confinement
of a seriously ill person to a hospital.
3. Numbers in parentheses refer to the bibliography at the end of the paper.

55
 54 RENIt A. SPITZ

year before the American Pediatric Association(3), Dr. Knox of

Baltimore stated that in the institutions of that city 9010 of the infants
died by the end of their first year. He believed that the remaining
1010 probably were saved because they had been taken out of the
institution in time. Dr. Shaw of Albany remarked in the same dis-
cussion that the mortality rate of Randalls Island Hospital was prob-
ably 10010.
Conditions have since greatly changed. At present the best
American institutions, such as Bellevue Hospital, New York City,
Downloaded by [University of Rochester] at 04:08 15 January 2018

register a mortality rate of less than 1010( 4), which compares fav-
orably with the mortality rate of the rest of the country. While these
and similar results were being achieved both here and in Europe,
physicians and administrators were soon faced with a new problem:
they discovered that institutionalized children practically without ex-
ception developed subsequent psychiatric disturbances and became
asocial, delinquent, feeble-minded, psychotic, or problem children.
Probably the high mortality rate in the preceding period had obscured
this consequence. Now that the children survived, the other draw-
backs of institutionalization became apparent. They led in this
country to the Widespread substitution of institutional care by foster
home care.
The first investigation of the factors involved in the psychiatric conse-
quences of institutional care of infants in their first year was made in 1933
in Austria by H. Durfee and K. Wolf(S). Further contributions to the prob-
lem were made by 1. G. Lowrey(6), 1. Bender and H. Yarnell(7), H. Bakwin
(4), and W. Goldfarb(8-11). The results of all these investigations are
roughly similar:
Bakwin found greatly increased susceptibility to infection in spite of high
hygienic and nutritional standards. Durfee and Wolf found that children under
three months show no demonstrable impairment in consequence of institution-
alization; but that children who had been institutionalized for more than eight
months during their first year show such severe psychiatric disturbances that
they cannot be tested. Bender, Goldfarb and Lowrey found that after three
years of institutionalization the changes effected are irreversible. Lowrey found
that whereas the impairment of children hospitalized during their first year
seems irremediable, that of children hospitalized in the second or third year
can be corrected.
Two factors, both already stressed by Durfee and Wolf, are made re-
sponsible by most of the authors for the psychological injury suffered by these
children.
First: Lack of stimulation. The worst offenders were the best equipped
and most hygienic institutions, which succeeded in sterilizing the surroundings
of the child from germs but which at the same time sterilized the child's
 HOSPITALISM 55

psyche. Even the most destitute of homes offers more mental stimulation than
the usual hospital ward.
Second: The presence or absence of the child's mother. Stimulation by
the mother will always be more intensive than even that of the best trained
nursery personnel (12 ). Those institutions in which the mothers were present
had better results than those where only trained child nurses were employed.
The presence of the mothers could compensate even for numerous other short-
comings.
We believe that further study is needed to isolate clearly the
various factors operative in the deterioration subsequent to prolonged
Downloaded by [University of Rochester] at 04:08 15 January 2018

care in institutions. The number of infants studied by Bakwin, Durfee-

Wolf and Lowrey in single institutions is very small and Bender-
Yarnell, and Goldfarb did not observe infants in the first twelve
months of life. We are not questioning here whether institutions should
be preferred to foster homes, a subject now hardly ever discussed-
the decision can by implication be deduced from the results of the
studies of the Iowa group in their extensive research on the "Nature
Versus Nurture" controversy(13-18). It may seem surprising
that in the course of this controversy no investigation has covered the
field of the first year of life in institutions.' All Iowa investigators
studied either children in foster homes or children over one year of
age, using their findings for retrospective interpretations." They did
not have at their disposal a method of investigation that would permit
the evaluation and quantification of development, mental or other-
wise, during the. first year of life. Their only instrument is the I.Q.,
which is unreliable (21), and not applicable during the first year.
However, the baby tests worked out by Hetzer and Wolf(22) fill
the gap, providing not only a quotient for intelligence but also quanti-
fiable data for development as a whole, such as indication of Develop-
mental Age and of a Developmental Quotient. They provide, further-
more, quantifiable data on six distinct sectors of personality, namely:
development of perception, body mastery, social relations, memory,
relations to inanimate objects, and intelligence (which in the first
4. Woodworth(19) in discussing the results of the Child Welfare Research Station
of the State University of Iowa makes the following critical remarks (p. 71): "The
causes of the inferior showing of orphanage children are obviously open to debate. . . .
It would seem that a survey and comparative study of institutional homes for children
would be instructive. . . ."
~. jonest zo) takes exception to this method as follows: "It seems probable that we
shall turn from retrospective surveys of conditions assumed to have had a prior influence,
and shall prefer to deal with the current and cumulative effects of specific environmental
factors. It may also be expected that our interest will shift to some extent from mass
statistical studies • • . to investigations of the dynamics of the growth process in
individuals."
 56 RENt. A. SPITZ

year is limited to understanding of relations between and insight into

the functions of objects).
With the help of these data ("dimensions"), a profile (per-
sonality curve) is constructed from which relevant conclusions can
be drawn and with the help of which children can be compared with
one another. Averages of development in anyone sector or in all
of them can be established for given environments. Finally, the
relevant progresses of one and the same child in the several sectors
of its personality can be followed up. The profiles present a cross-
Downloaded by [University of Rochester] at 04:08 15 January 2018

section of infantile development at any given moment; but they also

can be combined .nto longitudinal curves of the developmental pro-
gress of the child's total personality as wen as of the various sectors
of the personality.
The aim of my research is to isolate and investigate the pathogenic
factors responsible for the favorable or unfavorable outcome of in-
fantile development. A psychiatric approach might seem desirable;
however, infant psychiatry is a discipline not yet existent: its advance-
ment is one of the aims of the present study.

II. Material6

With this purpose in mind a long-term study of 164 children

was undertaken." In view of the findings of previous investigations
this study was largely limited to the first year of life, and confined to
two institutions, in order to embrace the total population of both
6. It is interesting to note that independently of our approach to this problem
(mapped out and begun in 1936) Woodworth(19) recommends a research program
on extremely similar lines as being desirable for the better understanding of the problem
of heredity and environment:
"Orphanages. Present belief based on a certain amount of evidence regards the
orphanages as an unfavorable environment for the child but the causes are not well under-
stood. Two general projects may be suggested.
(a) A survey of institutional homes for children with a view to discovering the
variations in their equipment and personnel and in their treatment of the children, with
some estimate of the results achieved.
(b) Experimental studies in selected orphanages which retain their children for a
considerable time, with a view to testing out the effects of specific environmental factors.
For example, the amount of contact of the child with adults could be increased for certain
chi!dren fo~ the purpose of seeing whether this factor is important in mental development.
It 1S conceivable that an orphanage could be run so as to become a decidedly favorable
environment for the gray-ring child, but at present we do not know how this result could
be accomplished." .
7. I wish to thank K. Wolf, Ph.D., for her help in the experiments carried out in
"Nursery" and in private homes, and for her collaboration in the statistical evaluation of
the results.
 HOSPITALISM 57

(130 infants). Since the two institutions were situated in different

countries of the Western hemisphere, a basis of comparison was
established by investigating non-institutionalized children of the same
age group in their parents' homes in both countries. A total of 34 of
these were observed. We thus have four environments:
TABLE 1.

Corresponding Corresponding
Institution Institution private
Environment private
Downloaded by [University of Rochester] at 04:08 15 January 2018

No. 11 background 2 No.2 background

Number of 69 11 61 23
Children

III. Procedure.
In each case an anamnesis was made which whenever possible
included data on the child's mother; and in each case the Hetzer-
Wolf baby tests were administered. Problems cropping up in the
course of our investigations for which the test situation did not
provide answers were subjected to special experiments elaborated for
the purpose. Such problems referred, for instance, to attitude and
behavior in response to stimuli offered by inanimate objects, by social
situations, etc. All observations of unusual or unexpected behavior
of a child were carefully protocoled and studied.
A large number of tests, all the experiments and some of the
special situations were filmed on 16/mm film. A total of 31,500
feet of film preserve the results of our investigation to-date. In the
analysis of the movies the following method was applied: Behavior
was filmed at sound speed, i.e., 24 frames per second. This makes
it possible to slow action down during projection to nearly one-third
of the original speed so that it can be studied in slow motion. A pro-
jector with additional handdrive also permits study of the films frame
by frame, if necessary, to reverse action and to repeat projection of
every detail as often as required. Simultaneously the written protocols
of the experiments are studied and the two observations compared.
1. Institution No. 1 will from here on be called "Nursery" institution No. 2
"Foundling Home". "
2. The small number of children observed in this particular environment was justified
by the fact that it has been previously studied extensively by other workers; our only
aim was to correlate our results with theirs. However, during the course of one year
each child was tested at least at regular monthly intervals.
 58 REN£ A. SPITZ

IV. Results.

For the purpose of orientation we established the average of the

Developmental Quotients for the first third of the first year of life
for each of the environments investigated. We contrasted these
averages with those for the last third of the first year. This compari-
son gives us a first hint of the significance of environmental influences
for development.
Downloaded by [University of Rochester] at 04:08 15 January 2018

Type of Cultural Developmental Quotients

Environment and Social Average of Average of
Background first four months last four months

Professional 133 131

Parental
Home Village 107 108
Population

"Nursery" 101.5 105

Institution
"Foundling 124 72
home"

TABLE II.
. Children of the first category come from professional homes in a large city;
their Developmental Quotient, high from the start, remains high in the course
of development.
Children in the second category come from an isolated fishing village of
499 inhabitants, where conditions of nutrition, housing, hygienic and medical
care are very poor indeed; their Developmental Quotient in the first four
months is much lower and remains at a lower level than that of the previous
category.
In the third category, "Nursery", the children were handicapped from birth
by the circumstances of their origin, which will be discussed below. At the
outset their Developmental Quotient is even somewhat lower than that of the
village babies; in the course of their development they gain slightly.
In the fourth category, "Foundling Home", the children are of an un-
selected urban (Latin) background. Their Developmental Quotient on
admission is below that of our best category but much higher than that of the
other two. The picture changes completely by the end of the first year, when
their Developmental Quotient sinks to the astonishingly low level of 72.

Thus the children in the first three environments were at the end
of their first year on the whole well-developed and normal, whether
they were raised in their progressive middle-class family homes (where
 IIDSPITALISM 59

obviously optimal circumstances prevailed and the children were

well in advance of average development), or in an institution or a
village home, where the development was not brilliant but still
reached a perfectly normal and satisfactory average. The chil-
dren in the fourth environment, though starting at almost as high a
level as the best of the others, had spectacularly deteriorated.
The children in Foundling Home showed all the manifestations of
hospitalism, both physical and mental. In spite of the fact that
hygiene and precautions against contagion were impeccable, the chil-
Downloaded by [University of Rochester] at 04:08 15 January 2018

dren showed, from the third month on;: extreme susceptibility to in-
fection and illness of any kind. There was hardly a child in whose
case history we did not find reference to otitis media, or morbilli, or
varicella, or eczema, or intestinal disease of one kind or another. No
figures could be elicited on general mortality; but during my stay an
epidemic of measles swept the institution, with staggeringly high
mortality figures, notwithstanding liberal administration of convales-
cent serum and globulins, as well as excellent hygienic conditions. Of
a total of 88 children up to the age of 20, 23 died. It is striking
to compare the mortality among the 45 children up to 10 years,
to that of the 43 children ranging from 1.0 to 20 years: usually,
the incidence of measles is low in the younger age group, but among
those infected the mortality is higher than that in the older age group;
since in the case of Foundling Home every child was infected, the
question of incidence does not enter; however, contrary to expectation,
the mortality was much higher in the older age group. In the younger
group, 6 died, i.e., approximately 13%. In the older group, 17 died,
i.e., close to 4070. The significance of these figures becomes apparent
when we realize that the mortality from measles during the first year
of life in the community in question, outside the institution, was
less than 0 %.
In view of the damage sustained in all personality sectors of the
children during their stay in this institution we believe it licit to
assume that their vitality (whatever that may be), their resistance to
disease, was also progressively sapped. In the ward of the children
ranging from 18 months to 20 years only two of the twenty-six surviv-
ing children speak a couple of words. The same two are able to
walk. A third child is beginning to walk. Hardly any of them can eat
alone. Cleanliness habits have not been acquired and all are incon-
tinent.
 60 RENE A. SPITZ

In sharp contrast to this is the picture offered by the oldest

inmates in Nursery, ranging from 8 to 12 months. The problem here
is not whether the children walk or talk by the end of the first year;
the problem with these lO-month-olds is how to tame the healthy
toddlers' curiosity and enterprise. They climb up the bars of the cots
after the manner of South Sea Islanders climbing palms. Special
measures to guard them from harm have had to be taken after one
lO-month-old actually succeeded in diving right over the more than
two-foot railing of the cot. They vocalize freely and some of them
actually speak a word or two. And all of them understand the signifi-
Downloaded by [University of Rochester] at 04:08 15 January 2018

cance of simple social gestures. When released from their cots, all
walk with support and a number walk without it.
What are the differences between the two institutions that result
in the one turning out normally acceptable children and the other
showing such appalling effects?

A. Simiiaritiesi"
1. Background of the children.
Nursery is a penal institution in which delinquent girls are
sequestered. When, as is often the case, they are pregnant on ad-
mission, they are delivered in a neighboring maternity hospital and
after the lying-in period their children are cared for in Nursery from
birth to the end of their first year. The background of these children
provides for a markedly negative selection since the mothers are mostly
delinquent minors as a result of social maladjustment or feeble-
mindedness, or because they are psychically defective, psychopathic, or
criminal. Psychic normalcy and adequate social adjustment is almost
excluded.
The other institution is a foundling home pure and simple.
A certain number of the children housed have a background not much
better than that of the Nursery children; but a sufficiently relevant
number come from socially well-adjusted, normal mothers whose only
handicap is inability to support themselves and their children (which
is no sign of maladjustment in women of Latin background). This is
expressed in the average of the Developmental Quotients of the two
institutions during the first 4 months, as shown in Table II.

8. Under this heading we enumerate not only actual similarities but also differences
t~at are of no et!ological significance for the deterioration in Foundling Home. These
differences cornpnse two groups: differences of no importance whatever, and differences
that actually favor the development of children in Foundling Home.
 HOSPITALISM 61

The background of the children in the two institutions does

therefore not favor Nursery; on the contrary, it shows a very marked
advantage for Foundling Home.

2. Housing Conditions.
Both institutions are situated outside the city, in large spacious
,gardens. In both hygienk conditions are carefully maintained. In
both infants at birth and during the first 6 weeks are segregated from
the older babies in a special newborns' ward, to which admittance
Downloaded by [University of Rochester] at 04:08 15 January 2018

is only permitted in a freshly sterilized smock after hands are washed.

In both institutions infants are transferred from the newborns' ward
after 2 or 3 months to the older babies' wards, where they are placed
in individual cubicles which in Nursery are completely glass enclosed,
in Foundling Home glass enclosed on three sides and open at the end.
In Foundling Home the children remain in their cubicles up to 15 to
18 months; in Nursery they are transferred after the 6th month to
rooms containing four to five cots each.
One-half of the children in Foundling Home are located in a
dimly lighted part of the ward; the other half, in the full light of
large windows facing southeast, with plenty of sun coming in. In
Nursery, all the children have well-lighted cubicles. In both institu-
tions the walls are painted in a light neutral color, giving a white
impression in Nursery, a gray-green impression in Foundling Home.
In both, the children are placed in white painted cots. Nursery is
financially the far better provided one: we usually find here a small
metal table with the paraphernalia of child care, as well as a chair,
in each cubicle; whereas in Foundling Home it is the exception if a
low stool is to be found in the cubicles, which usually contain nothing
but the child's cot.

3. Food.
In both institutions adequate food is excellently prepared and
varied according to the needs of the individual child at each age;
bottles from which children are fed are sterilized. In both institutions
a large percentage of the younger children are breast-fed. In Nursery
this percentage is smaller, so that in most cases a formula is soon
added, and in many cases weaning takes place early. In Foundling
Home all children are breast-fed as a matter of principle as long as
they are under 3 months unless disease makes a deviation from this
rule necessary.
 62 RENE A. SPITZ

4. Clothing.
Clothing is practically the same in both institutions. The children
have adequate pastel-colored dresses and blankets. The temperature
in the rooms is appropriate. We have not seen any shivering child
in either set-up.
5. Medical Care.
Foundling Home is visited by the head physician and the medical
Downloaded by [University of Rochester] at 04:08 15 January 2018

staff at least once a day, often twice, and during these rounds the chart
of each child is inspected as well as the child itself. For special
ailments a laryngologist and other specialists are available; they also
make daily rounds. In Nursery no daily rounds are made, as they
are not necessary. The physician sees the children when called.
Up to this point it appears that there is very little significant
difference between the children of the two institutions. Foundling
Home shows, if anything, a slight advantage over Nursery in the
matter of selection of admitted children, of breast-feeding and of
medical care. It is in the items that now follow that fundamental
differences become visible.

B. Differences:
1. Toys.
In Nursery it is the exception when a child is without one or
several toys. In Foundling Home my first impression was that not a
single child had a toy. This impression was later corrected. In the
course of time, possibly in reaction to our presence, more and more
toys appeared, some of them quite intelligently fastened by a string
above the baby's head so that he could reach it. By the time we left a
large percentage of the children in Foundling Home had a toy.
2. Visual Radius.
In Nursery the corridor running between the cubicles, though
rigorously white and without particular adornment, gives a friendly
impression of warmth. This is probably because trees, landscape
and sky are visible from both sides and because a bustling activity
of mothers carrying their children, tending them, feeding them, play-
ing. with them, chatting with each other with babies in their arms,
is usually present. The cubicles of the children are enclosed but the
glass panes of the partitions reach low enough for every child to
 HOSPITALISM 63

be able at any time to observe everything going on all around. He

can see into the corridor as soon as he lifts himself on his elbows.
He can look out of the windows, and can see babies in the other
cubicles by just turning his head; witness the fact that whenever the
experimenter plays with a baby in one of the cubicles the babies in
the two adjoining cubicles look on fascinated, try to participate in
the game, knock at the panes of the partition, and often begin to
cry if no attention is paid to them. Most of the cots are provided with
widely-spaced bars that are no obstacle to vision. After the age of 6
Downloaded by [University of Rochester] at 04:08 15 January 2018

months, when the child is transferred to the wards of the older babies,
the visual field is enriched as a number of babies are then together in
the same room, and accordingly play with each other.
In Foundling Home the corridor into which the cubicles open,
though full of light on one side at least, is bleak and deserted, except
at feeding time when five to eight nurses file in and look after the
children's needs. Most of the time nothing goes on to attract the
babies' attention. A special' routine of Foundling Home consists in
hanging bed sheets over the foot and the side railing of each cot.
The .cot itself is approximately 18 inches high. The side railings are
about 20 inches high; the foot and head railings are approximately
28 inches high. Thus, when bed sheets are hung over the railings,
the child lying in the cot is effectively screened from the world. He is
completely separated from the other cubicles, since the glass panes of
the wooden partitions begin 6 to 8 inches higher than even the head
railing of the cot. The result of this system is that each baby lies in
solitary confinement up to the time when he is able to stand up in his
bed, and that the only object he can see is the ceiling.

3. Radius of Locomotion.
In Nursery the radius of locomotion is circumscribed by the
space available in the cot, which up to about 10 months provides a
fairly satisfactory range.
Theoretically the same would apply to Foundling Home. But
in practice this is not the case for, probably owing to the lack of stimu-
lation, the babies lie supine in their cots for many months and a
hollow is worn into their mattresses. By the time they reach the
age when they might turn from back to side (approximately the
7th month) this hollow confines their activity to such a degree that
they are effectively prevented from turning in any direction. As a
result we find most babies, even at 10 and 12 months, lying on their
 64 RENE A. SPITZ

backs and playing with the only object at their disposal, their own
hands and feet.
4. Personnel.
In Foundling Home there is a head nurse and five assistant nurses
for a total of forty-five babies. These nurses have the entire care
of the children on their hands, except for the babies so young that
they are breast-fed. The latter are cared for to a certain extent by
their own mothers or by wetnurses; but after a few months they are
Downloaded by [University of Rochester] at 04:08 15 January 2018

removed to the single cubicles of the general ward, where they share
with at least seven other children the ministrations of one nurse. It
is obvious that the amount of care one nurse can give to an individual
child when she has eight children to manage is small indeed. These
nurses are unusually motherly, baby-loving women; but of course the
babies of Foundling Home nevertheless lack all human contact for
most of the day.
Nursery is run by a head nurse and her three assistants, whose
duties do not include the care of the children, but consist mainly in
teaching the children's mothers in child care, and in supervising them.
The children are fed, nursed and cared for by their own mothers or,
in those cases where the mother is separated from her child for any
reason, by the mother of another child, or by a pregnant girl who in
this way acquires the necessary experience for the care of her own
future baby. Thus in Nursery each child has the full-time care of
his own mother, or at least that of the substitute which the very able
head nurse tries to change about until she finds someone who really
likes the child.
V. Discussion.
To say that every child in Nursery has a full-time mother is an
understatement, from a psychological point of view. However modern
a penal institution may be, and however constructive and permissive
its reeducative policies, the deprivation it imposes upon delinquent
girls is extensive. Their opportunities for an outlet for their interests,
ambitions, activity, are very much impoverished. The former sexual
satisfactions as well as the satisfactions of competitive activity in
the sexual field, are suddenly stopped: regulations prohibit flashy
dresses, vivid nail polish, or extravagant hair-do's. The kind of social
life in which the girls could show off has vanished. This is especially
traumatic as these girls become delinquent because they have not
been able to sublimate their sexual drives, to find substitute gratifica-
 HOSPITALISM 65

tions, and therefore do not possess a pattern for relinquishing pleasure

when frustrated. In addition, they do not have compensation in rela-
tions with family and friends, as formerly they had. These factors,
combined with the loss of personal liberty, the deprivation of private
property and the regimentation of the penal institution, all add up to
a severe narcissistic trauma from the time of admission; and they
continue to affect the narcissistic and libidinal sectors during the whole
period of confinement.
Luckily there remain a few safety valves for their emotions:
Downloaded by [University of Rochester] at 04:08 15 January 2018

( 1) the relationship with wardens, matrons and nurses; (2) with

fellow prisoners; (3) with the child. In the relationship with the
wardens, matrons and nurses, who obviously represent parent figures,
much of the prisoner's aggression and resentment is bound. Much
of it finds an outlet in the love and hate relationship to fellow pris-
oners, where all the phenomena of sibling rivalry are revived.
The child, however, becomes for them the representative of their
sexuality, a product created by them, an object they own, which they
can dress up and adorn, on which they can lavish their tenderness and
pride, and of whose accomplishments, performance and appearance
they can boast. This is manifested in the constant competition among
them as to who has the better dressed, more advanced, more intelligent,
better looking, the heavier, bigger, more active--in a word, the better
baby." For their own persons they have more or less given up the
competition for love, but they are intensely jealous of the attention
given to their children by the matrons, wardens, and fellow prisoners.
It would take an exacting experimenter to invent an experiment
with conditions as diametrically opposed in regard to the mother-child
relationship as they are in these two institutions. Nursery provides
each child with a mother to the nth degree, a mother who gives the
child everything a good mother does and, beyond that, everything
else she has!". Foundling Home does not give the child a mother,
nor even a substitute-mother, but only an eighth of a nurse.

9. The psychoanalytically oriented reader of course realizes that for these girls in
prison the child has become a hardly disguised phallic substitute. However, for the
purposes of this article I have carefully avoided any extensive psychoanalytic interpreta-
tion, be it ever so tempting, and limited myself as closely as possible to results of direct
observations of behavior. At numerous other points it would be not only possible but
natural to apply analytic concepts; that is reserved for future publication.
10. For the non-psychoanalytically oriented reader we note that this intense mother-
child relationship is not equivalent to a relationship based on love of the child. The
mere fact that the child is used as a phallic substitute implies what a large part uncon-
scious hostility plays in the picture.
 66 RENE A. SPITZ

We are now in a position to approach more closely and with

better understanding the results obtained by each of the two institu-
tions. We have already cited a few: we mentioned that the Develop-
mental Quotient of Nursery achieves a normal average of about 105
at the end of the first year, whereas that of the Foundling Home sinks
to 72; and we mentioned the striking difference of the children in the
two institutions at first sight. Let us first consider the point at which
the developments in the two institutions deviate.
Downloaded by [University of Rochester] at 04:08 15 January 2018

On admission the children of Foundling Home have a much better

average than the children of Nursery; their hereditary equipment is
better than that of the children of delinquent minors. But while
Foundling Home shows a rapid fall of the developmental index,
Nursery shows a steady rise. They cross between the 4th and 5th
months, and from that point on the curve of the average Develop-
mental Quotient of the Foundling Home drops downward with in-
creasing rapidity, never again to rise (Curve I).
The point where the two curves cross is significant. The time
when the children in Foundling Home are weaned is the beginning of
the 4th month. The time lag of one month in the sinking of the
index below normal is explained by the fact that the Quotient repre-
sents a cross-section including all sectors of development, and that
attempts at compensation are made in some of the other sectors.
However, when we consider the sector of Body Mastery (Curve
II) which according to WolfH is most indicative for the mother-child
relationship, we find that the curves of the children in Nursery cross
the Body Mastery curve of the Foundling Home children between the
3rd and 4th month. The inference is obvious. As soon as the babies
in Foundling Home are weaned the modest human contacts which
they have had during nursing at the breast stop, and their development
falls below normal.
One might be inclined to speculate as to whether the further
deterioration of the children in Foundling Home is not due to other
factors also, such as the perceptual and motor deprivations from
which they suffer. It might be argued that the better achievement of
the Nursery children is due to the fact that they were better provided
for in regard to toys and other perceptual stimuli. We shall therefore
analyze somewhat more closely the nature of deprivations in per-
ceptual and locomotor stimulation.

11. K. Wolf, "Body Mastery of the Child as an index for the Emotional Relationship
between Mother and Child" (in preparation).
 }{OSPLTALlSM 67

First of all it should be kept in mind that the nature of the

inanimate perceptual stimulus, whether it is a toy or any other object,
has only a very minor importance for a child under 12 months. At
this age the child is not yet capable of distinguishing the real purpose
of an object. He is only able to use it in a manner adequate to his
own functional needs (23) . Our thesis is that perception is a function
of libidinal cathexis and therefore the result of the intervention of
an emotion of one kind or another .12. Emotions are provided for the
COMPARISON of DEVELOPMENT
Downloaded by [University of Rochester] at 04:08 15 January 2018

Average in
Developmental "NURSERY"and "FOUNDLINGHOME··
Quotient during the first five months.
130

II "Foundlinghome"
g•:.:........ "N ursery ..
120

Curve 11
110

100

90 .....:.:lW::&:t:~
AGe: 2-3 3-4 4-5 2-3 3-4 4-5 Months
12. This is stating in psychoanalytic terms the conviction of most modern psycholo-
gists, beginning with Compayre(24) and shared by such familiar authorities in child
psychology as Stern( 25) and Buhler( 26), and in animal psychology, Tolman (27) .
 68 RENE A. SPITZ

child through the intervention of a human partner, i.e., by the mother

or her substitute. A progressive development of emotional inter-
change with the mother provides the child with perceptive experiences
of its environment. The child learns to grasp by nursing at the
mother's breast and by combining the emotional satisfaction of that
experience with tactile perceptions. He learns to distinguish animate
objects from inanimate ones by the spectacle provided by his mother's
face(28) in situations fraught with emotional satisfaction. The inter-
change between mother and child is loaded with emotional factors
Downloaded by [University of Rochester] at 04:08 15 January 2018

and it is in this interchange that the child learns to play. He be-

comes acquainted with his surroundings through the mother's carry-
ing him around; through her help he learns security in locomotion
as well as in every other respect. This security is reinforced by her
being at his beck and call. In these emotional relations with the
mother the child is introduced to learning, and later to imitation. We
have previously mentioned that the motherless children in Foundling
Home are unable to speak, to feed themselves, or to acquire habits of
cleanliness: it is the security provided by the mother in the field of
locomotion, the emotional bait offered by the mother calling her child,
that "teaches" him to walk. When this is lacking, even children two
to three years old cannot walk.
The children in Foundling Home have, theoretically, as much
radius of locomotion as the children in Nursery. They did not at
first have toys, but they could have exerted their grasping and tactile
activity on the blankets, on their clothes, even on the bars of the
cots. We have seen children in Nursery without toys; they are the
exception-but the lack of material is not enough to hamper them
in the acquisition of locomotor and grasping skills. The presence of
a mother or her substitute is sufficient to compensate for all the other
deprivations.
It is true that the children in Foundling Home are condemned to
solitary confinement in their cots. But we do not think that it is the
lack of perceptual stimulation in general that counts in their depriva-
tion. We believe that they suffer because their perceptual world is
emptied of human partners, that their isolation cuts them off from
any stirnulation by any persons who could signify mother-representa-
tives for the child at this age.is The result, as Curve III shows,
is a complete restriction of psychiccapacity by the end of the first year.
13. This statement is to be developed further in a forthcoming article on "The
Beginning of the Social Relations of the Child".
 HOSPITALISM 69

Average
Deve lopmental
Quotient
130
COMPARISON of DEVELOPMENT
in
"NURSERY"and "FOUNDUNGHOME"

• "Foundlinghome"
l'ZO
iii "Nursery"
Downloaded by [University of Rochester] at 04:08 15 January 2018

Curve III

1\0
::::~;::~:. :;f.:::::;:; ~~~~I~~~~Il ~ ~ ~ ~ ~ ~ ~
lltl ~r~
100
tWi
;tW, ~. :.I(~:i.·; :l." ;:~.! :. ·:.! 1;:" . :1i~.!;"
II ..

w.w. _ _
.:.:.:::.:::.:
..
::.:.:::.

90

80

70

60 -I:A~=-:'~
AGE: 2-3 3-4 4-5 5-6 6-7 7-8 8 10 10-/2 Months
 70 RENE A. SPITZ

This restriction of psychic capacity is not a temporary phenomenon.

It is, as can be seen from the curve, a progressive process. How much
this deterioration could have been arrested if the children were taken
out of the institution at the end of the first year is an open question.
The fact that they remain in Foundling Home probably furthers this
progressive process. By the end of the second year the Development~l
Quotient sinks to 45, which corresponds to a mental age of approxi-
mately 10 months, and would qualify these children as imbeciles.
The curve of the children in Nursery does not deviate significantly
Downloaded by [University of Rochester] at 04:08 15 January 2018

from the normal. The curve sinks at two points, between the 6th and
7th, and between the 10th and 12th months. These deviations are
within the normal range; their significance will be discussed in a
separate article. It has nothing to do with the influence of institu-
tions, for the curve of the village group is nearly identical.

VI. Provisional Conclusions.

The contrasting pictures of these two institutions show the sig-

nificance of the mother-child relationship for the development of the
child during the first year. Deprivations in other fields, such as per-
ceptual and locomotor radius, can all be compensated by adequate
mother-child relations. "Adequate" is not here a vague general term.
The examples chosen represent the two extremes of the scale.
The children in Foundling Home do have a mother-for a time,
in the beginning-but they must share her immediately with at least
one other child, and from 3 months on, with seven other children.
The quantitative factor here is evident. There is a point under which
the mother-child relations cannot be restricted during the child's first
year. without inflicting irreparable damage. On the other hand, the
exaggerated mother-child relationship in Nursery introduces a dif-
ferent quantitative factor. To anyone familiar with the field it is
surprising that Nursery should achieve such excellent results, for we
know that institutional care is destructive for children during their first
year; but in Nursery the destructive factors have been compensated by
the increased intensity of the mother-child relationship.
These findings should not be construed as a recommendation for over-
protection of children. In principle the libidinal situation of Nursery is
almost as undesirable as the other extreme in Foundling Home. Neither in
the nursery of a penal institution nor in a foundling home for parentless children
can the normal libidinal situation that obtains in a family home be expected.
The two institutions have here been chosen as experimental set-ups for the
 HOSPITALISM 71

purpose of examining variations in libidinal factors ranging from extreme frus-

tration to extreme gratification. That the extreme frustration practised in
Foundling Home has deplorable consequences has been shown; the extreme
gratification in Nursery can be tolerated by the children housed there for two
reasons:
(1) The mothers have the benefit of the intelligent guidance of the
head nurse -and her assistants, and the worst exaggerations are thus corrected.
(2) Children during their first year of life can stand the ill effects of
such a situation much better than at a later age. In this respect Nursery has
wisely limited the duration of the children's stay to the first twelve months.
Downloaded by [University of Rochester] at 04:08 15 January 2018

For children older than this we should consider a libidinal set-up such as that
in Nursery very dangerous indeed.

DEVELOPMENT in NURSERY DEVELOPMENT in VILLAGE GROUP

Average
Developmental
Quotient
130

Curve IV Curve V
120
roo-'

110
- I--

100

90
AGE:
-~.
~~~Z;;;
~;;;;;E; IIIIIII
2-3 3-4 4-5 5-6 6-77-88-1010·12
- F

2-3 3'4 4·5 5'66-1'7'88-\010-12 Month$

VII. Further Problems.

This is the first of a series of publications on the results of a

research project on infancy that we are conducting. As such it is a pre-
liminary report. It is not intended to show more than the most general
outline of the results of early institutional care, giving at the same
time a hint of the approach we use. The series of other problems
on which this investigation has shed some light, as well as the formu-
lation of those problems that could be recognized as such only in the
course of the investigation, have not been touched upon in our present
study and can only summarily be touched upon; they are headings, as
it were, of the chapters of our future program of publication.
 72 RENt A. SPITZ

Apart from the severe developmental retardation, the most. strik-

ing single factor observed in Foundling Home was the change ill the
pattern of the reaction to strangers in the last third of the first
year(29}. The usual behavior was replaced by something that co~ld
vary from extreme friendliness to any human partner combined with
anxious avoidance of inanimate objects, to a generalized anxiety ex-
pressed in blood-curdling screams which could go on indefinitely. It
is evident that these deviant behavior patterns require a more thor-
ough and extensive discussion than our present study would have per-
Downloaded by [University of Rochester] at 04:08 15 January 2018

mitted.
We also observed extraordinary deviations from the normal in the
time of appearance and disappearance of familiar developmental pat-
terns; and certain phenomena unknown in the normal child, such as
bizarre stereotyped motor patterns distinctly reminiscent of the stereo-
typy in catatonic motility. These and other phenomena observed in
Foundling Home require an extensive discussion in order to determine
which are to be classified as maturation phenomena, (which appear
even under the most unfavorable circumstances, and which appear
with commensurate retardation when retardation is general); or
which can be considered as the first symptoms of the development of
serious psychiatric disturbances. In connection with this problem
a more thorough discussion of the rapidity with which the Develop-
mental Quotients recede in Foundling Home is intended.
Another study is to deal with the problems created by the enor-
mous over-protection practised in Nursery.
And finally the 'rationale of the one institutional routine as against
that of the other will have to be discussed in greater detail. This study
will offer the possibility of deciding how to compensate for unavoid-
able changes in the environment of children orphaned at an early age.
It will also shed some light on the social consequences of the progres-
sive disruption of home life caused by the increase of female labor
and by the demands of war; we might state that we foresee in the
course of events a corresponding increase in asociality, in the number
of problem and delinquent children, of mental defectives, and of psy-
chotics.
It will be necessary to take into consideration in our institutions,
in our charitable activities, in our social legislation, the overwhelming
and unique importance of adequate and satisfactory mother-child
relationship during the first year, if we want to decrease the unavoid-
able and irreparable psychiatric consequences deriving from neglect
during this period.
 HOSPITALISM 75

BIBLIOGRAPHY

1. Schlossman, A., "Zur Frage der Siiuglingssterblichkeit", Miinchner Med.

Wochenschrift, 67, 1920.
2. Chapin, H. D., "Are Institutions for Infants Necessary?", Tournai of
American Medical Association, January, 1915.
3. Chapin, H. D., "A Plea for Accurate Statistics in Infants' Institutions",
Archives of Pediatrics, October, 1915.
4. Bakwin, H., "loneliness in Infants", American Tournai of Diseases of
Children, 63, 1942, pp. 30-40.
Downloaded by [University of Rochester] at 04:08 15 January 2018

5. Durfee, H. and Wolf, K., "Anstaltspflege und Entwicklung im ersten

Lebensjahr", Zeitschrift fur Kinderjorscbung, 42/3, 1933.
6. lowrey, 1. G., "Personality Distortion and Early Institutional Care",
American Tournai of Orthopsychiatry, X, 3, 1940, pp. 576-585.
7. Bender, 1. and Yarnell, H., "An Observation Nursery: a Study of 250
Children in the Psychiatric Division of Bellevue Hospital", American
Tournai of Psychiatry, 97, 1941, pp. 1158-1174.
8. Goldfarb, W., "Infant Rearing as a Factor in Foster Home Placement",
American Tournai of Orthopsychiatry, XIV, 1944, pp. 162-167.
9. Goldfarb, W., "Effects of Early Institutional Care on Adolescent Per-
sonality: Rorschach Data", American Tournai of Orthopsychiatry, XIV,
1944, pp. 441-447.
10. Goldfarb, W., "Effects of Early Institutional Care on Adolescent Per-
sonality", [ournal of Experimental Education, 12, 1943, pp. 106-129.
11. Goldfarb, W. and Klopfer, B., "Rorschach Characteristics of Institutional
Children", Rorschach Research Exchange, 8, 1944, pp. 92-100.
12. Ripin, R., "A Study of the Infant's Feeding Reactions During the First
Six Months of life", Archives of Psychology, 116, 1930, p. 38.
13. Skeels, H. M., "Mental Development of Children in Foster Homes",
Tournai of Consulting Psychology, 2, 1938, pp. 33-43.
14. Skeels, H. M., "Some Iowa Studies of the Mental Growth of Children in
Relation to Differentials of the Environment: A Summary", 39th Year-
book, National Society for the Study of Education, II, 1940, pp. 281-308.
15. Skeels, H. M., Updegraff, R, Wellman, B. 1., and Williams, H. M., "A
Study of Environmental Stimulation; an Orphanage Preschool Project",
University of Iowa Studies in Child Welfare, 15, 4, 1938.
16. Skodak, M., "Children in Foster Homes", University of Iowa Studies in
Child Welfare, 16, 1, 1939.
17. Stoddard, G. D., "Intellectual Development of the Child: an Answer to
the Critics of the Iowa Studies", School and Society, 51, 1940, pp. 529·
536.
18. Updegraff, R, "The Determination of a Reliable Intelligence Quotient
for the Young Child", Journal of Genetic Psychology, 41, 1932, pp. 152-
166.
19. Woodworth, R S., "Heredity and Environment", Bulletin 47, Social
Science Research Council, 1941.
20. Jones, H. E., "Personal Reactions of the Yearbook Committee", 39th
Yearbook, National Society for the Study of Education, I, 1940, pp.
454-456.
 74 RENE A. SPITZ

21. Simpson, B. R., "The Wandering 1. Q.", Journal of Psychology, 7, 1939.

pp. 351-367.
22. Hetzer, H. and Wolf, K., "Baby Tests", Zeitschrift fiir Psycbologie,
107, 1928.
23. Buhler, Ch., Kindheit und [ugend, Leipzig, 1931, p. 67.
24. Cornpayre, G., L'eoolution [ntellectaelle et morale de l'enfant, Paris, 1893.
25. Stern, Wm., Psychology of Early Cbildbood, London, 1930.
26. BUhler, K., Die geistige Entwicklung des Kindes, 4th ed., Jena, 1942;.
p. 106 and p. 116.
27. Tolman, E. c., Purposive Behavior, New York, 1932, P: 27 ff.
28. Gesell, A. and Ilg, P., Feeding Behavior of Infants, Phila., 1937, p. 21.
Downloaded by [University of Rochester] at 04:08 15 January 2018

29. Gesell, A. and Thompson, H., Infant Behavior, its Genesis and Growth,
New York, 1934. p. 208.