Professional Documents
Culture Documents
PSYCHOANALYSIS AND
SCHIZOPHRENIA:
A CAUTIONARY TALE
instances, the very scientists who have given the world the most pen-
etrating insights have also made serious errors, errors that could not
be corrected until shifts in theory or advances in technology enabled
them or others to gather the data with which to correct them.
Errors in our own f ield, of course, start with Freud himself. When
he realized that his early belief in the universality of childhood sexual
seduction in cases of hysteria was not correct, he was able to develop
one of his most important insights—the discovery of infantile sexu-
ality and its representation in unconscious, repressed sexual fantasies.
Twenty years later, he corrected his theory of anxiety, a revision that
opened up new investigations of the processes of defense.
Freud’s corrections and revisions required not a new technology but
a reevaluation of existing clinical data, as well as the collection of
new data based on revisions of his technique. Since then we have con-
stantly been revising our views of the nature and genesis of psycho-
pathology as we gather and reevaluate clinical observations and data.
More recently, however, psychoanalysis has been confronted with
data of a different kind from the f ield of neurobiology—data that have
28
forced us to revise our formulations of the etiology of mental illnesses
such as schizophrenia, bipolar disorder, OCD, and infantile autism
(Marcus 1992). Even as we accept these revisions, it will prove useful
to examine the sources of our incorrect theories of etiology in order
both to understand how we came to believe in them and to alert our-
selves to the possibility that we might be making similar mistakes today.
There is perhaps no better example of a psychoanalytic theory that
has turned out to be incorrect than the once prevalent theory regarding
the etiology of schizophrenia. This theory, or perhaps I should say con-
viction, about the cause of schizophrenia had three components. It was
believed, f irst, that the damage to the ego and its functions took place
primarily in the f irst one or two years of life; second, that the major
cause of these ego impairments was to be found in profoundly inade-
quate caretaking; and, third, that in most cases there was no fundamen-
tal organic process responsible for the illness. As has been true of other
mistaken scientif ic ideas, the correction of these ideas had to await
the development of a technology capable of studying the brain in ways
hitherto impossible—technologies that are now about twenty-f ive
years old in the f ield of neuroscience. While it is true that powerful
genetic data obtained through family, twin, and adopted-away studies
had existed even prior to this new neurobiological information, those
PSYCHOANALYSIS AND SCHIZOPHRENIA
studies were frequently ignored by analysts or placed far into the back-
ground of the etiological equation.
The neurobiological research of the past twenty-f ive years offers
compelling evidence that the etiology of schizophrenia is, to put it suc-
cinctly, biological rather than psychogenic. Although its exact cause, as
well as its pathogenesis, has yet to be determined, most researchers
believe that a combination of genetic predisposition and intrauterine
and birth insults are necessary for the illness to occur. If there are
environmental factors after birth that contribute to the cause of schizo-
phrenia, there is very little evidence that they take place in the f irst
one or two years of life (see Willick 1990, 1993).1
It should be noted that there are still a number of psychoanalysts
who, though they now accept the contribution of a genetic predispo-
sition, believe that a destructive early environment is involved in the
later expression of the disorder. Robbins (1993) believes that good early
mothering can prevent the emergence of the illness in otherwise predis-
posed children. Volkan (1997) proposes that schizophrenia occurs in
individuals who have an infantile psychotic self “formed during the
29
early interaction of mother and child . . . ” (p. 8).
What I intend to do in telling this cautionary tale is to f irst
review some of the forces that have led so many analysts to formu-
late these theories of the etiology of schizophrenia. Who were the
analysts who led the way to such formulations? What ideas inf lu-
enced them in that direction? And what sort of clinical data did they
use?
My purpose here is not to criticize analysts for past errors. As I
have said, errors are an inevitable part of any scientif ic endeavor. What
is important is to inquire into the reasons for them. After a brief
review of some of these reasons, I will turn to an examination of
current psychoanalytic theories about the cause of other serious psychi-
atric disorders in order to ask the question, Are we making the same
mistakes again?
1
Many studies seem to indicate that environmental factors after birth may not
be of etiological importance. A few studies (Tienari 1991; Tienari, Sorri, and Lahti
1987; Jones et al. 1994) do, however, implicate family dysfunction or poorer-than-
average mothering in the expression of the illness. Even these, however, do not
necessarily concern caretaking during the first year or two of life. In addition, the rate
of schizophrenia among African-Caribbeans in Trinidad is markedly lower than the
high rate among second-generation African-Caribbeans in London (Bhugra et al.
1996). No one knows precisely what the environmental risk factors are.
Martin S. Willick
Ego Psychologists
Both Jacobson (1953, 1954, 1966) and Bak adhered to the view that
early ego development in schizophrenia was more disturbed than in
illnesses they considered less severe. Bak did so in comparing schizo-
phrenia to the perversions, while Jacobson did so in comparing it to
manic-depressive illness. “Manic depressives,” Jacobson (1953) wrote,
“seem, at some point, to have reached a higher level in the differenti-
ation and integration of the psychic systems. Consequently, the acute
regressive process during their episodes does not go so far as in schizo-
phrenics and is of a different type” (p. 51). She also made a number of
distinctions about schizophrenia and manic depressive illness based on
the type of regressive symptoms they manifested. Many of these mani-
festations, such as the loss of ego boundaries, and the presence of thought
insertion and thought broadcasting, are now seen to be present in both
types of patients and are therefore not useful in distinguishing between
the illnesses or between degrees of regression. Thus, both Jacobson and
Bak went along with some form of the idea that there is a chronological
timetable for the development of various psychiatric illnesses.
The proposition of a chronological timetable can to be found also
32
in Kernberg’s writings in the seventies (1975, 1976). He believes that
there are two major tasks for the early ego: the f irst is to differentiate
self- and object representations; the second is to integrate them. By
proposing that borderlines accomplish the f irst task but not the second,
while “psychotics” accomplish neither, he promotes the idea that such
defects in ego structure in schizophrenia occur during the first year of life,
while those in borderline conditions occur later in child development.
Mahler’s work (Mahler and Furer 1968; Mahler, Pine, and Bergman
1975), in laying out a developmental sequence of stages from autistic
through to symbiotic to separation-individuation on the way to object
constancy, fostered the idea that the more severe the pathology, the
earlier were its roots in development.
By and large, ego psychologists believed that symptoms such as
loss of ego boundaries or loss of self-object differentiation were mani-
festations of regression to very early normal functioning. They were
then led to the conclusion that such severe ego impairments must be
caused by some basic faults in the very laying down of psychic
structure during infancy. It was therefore not diff icult for a clinician
like Greenacre (1953) to state that “the matrix of these severe distur-
bances . . . lay in disturbances in that period at the very dawn of the ego,
roughly around six months and a little later . . .” (p. 10).
PSYCHOANALYSIS AND SCHIZOPHRENIA
Interpersonalists
The observation that most of the British object relations theorists
who wrote about schizophrenia lacked actual experience with schizo-
phrenic patients cannot be applied to the American analysts who can be
grouped together under the name of the Washington school of psychi-
atry. Proponents of the interpersonal point of view, they published
extensive case histories of schizophrenic patients treated four or f ive
times weekly, mostly in a hospital setting.
Many analysts from this school would acknowledge that Sullivan
(1953, 1956, 1962) provided a good deal of the foundation for their for-
PSYCHOANALYSIS AND SCHIZOPHRENIA
4
From his writings it appears that Sullivan was not only an astute clinician who
did not accept easy formulations about the etiology of the disorder, but was an
empathic therapist who treated his patients with great respect. As a matter of fact, as
early as the mid-1920s he established a ward to treat young male schizophrenic
patients at the Shepard and Enoch Pratt Hospital in Maryland. The guiding principle
was to treat these patients as human beings capable of relating to others despite their
severe impairments. He did believe that in some cases of schizophrenia an heredi-
tarily determined organic process was responsible (1956, p. 309).
5
In Schizophrenia and the Need-Fear Dilemma (Burnham 1969) these con-
clusions are represented in this way: “We believe that among the various types of
schizophrenia there is at least one group of persons whose inborn psychic apparatus
is normal but whose disordered early object relations interfered with normal devel-
opment. It seems likely that another group has suffered compounding of genetically
given weaknesses as a result of unfavorable early experiences” (pp. 15–16); “We
postulate that the nature of the schizophrenic’s early object relations interferes with
the normal processes of differentiation and integration” (p. 31). These conclusions
were drawn from the study of staff-patient relationships in a group of hospitalized
chronic schizophrenic patients and from the long-term intensive treatment that was
carried out at that time.
Martin S. Willick
6
While such intensive work with nonmedicated, actively psychotic patients must
surely be distressing, it is difficult to know whether all therapists would experience
so much uncertainty about their essential self or identity as Searles so frequently
describes. Prominent in Searles’s countertransference feelings are many that seem to
be quite idiosyncratic and raise serious questions about the accuracy of the conclu-
sions he drew from them.
Martin S. Willick
is that the disturbance had its roots in the earliest years of growth and
development. A corollary to this idea is that if regressive symptoms
are evident, or what are considered to be primitive defenses and object
relations, we are justif ied in concluding that impairments took place
very early in life. The second idea is that we can fairly reliably use
transference-countertransference interactions in the analytic situation
to understand the nature of the early object relations of the patient. The
third is that normal infant observation studies can help us understand
how early diff iculties can lead to adult psychopathology.
The more serious the disorder, the earlier its etiological roots. The
idea that the more serious the disorder, the earlier in life may be found
its major psychogenetic inf luences was promulgated by Freud himself.
Although Freud never attributed f ixations or impairments in devel-
opment in schizophrenia to failures in the nurturing environment in the
earliest years of life, his concepts of f ixation and regression were to
have a profound effect on theories of pathogenesis (see Willick 1990).
He furthered the idea of a chronological timetable for f ixation points
and later regression by stating that dispositional f ixations to schizo-
39
phrenia and paranoia were to be found in the earliest stages of libidinal
development, in contrast to those found in the neuroses (Freud 1913).
Despite his cautionary comments about the possibility of basic ego dis-
turbances in the psychoses (Freud 1911), the foundations were laid for a
general theory of phase-specif icity holding that the more serious the
pathology, the earlier in life did the impairments in development occur.
These ideas were elaborated by Abraham (1908, 1924), who proposed
that the psychosexual phases could be subdivided, with f ixations at each
subphase leading to different illnesses, and by Tausk (1919), whose
investigation of the delusion of an inf luencing machine in a woman with
schizophrenia led him to conclude that early self-object differentiation
had been impaired in this illness.
Thus, a very powerful concept came to the fore in the thinking of
many analysts. It seemed reasonable to them that by examining the
symptomatic manifestations in patients with schizophrenia one could
see regressions not only of the libidinal drive, but also of ego, superego
functions, and object relations. It seemed to follow, then, that there had
to have been diff iculties in the actual early phases of development,
whether one used the earlier designation of fixation at a very early
infantile level or later concepts such as developmental arrest or pro-
found vulnerability to regression to that early level at times of stress.
Martin S. Willick
8
It was also problematic that Mahler chose to label the first two months of life the
“autistic” phase because in truth, other than the fact that the infant may sleep a great
deal and does not appear to be affectively engaged with its caretaker, these features
bear little resemblance to those of a child with autism. More recent work (Stern
1985) has demonstrated that there is a good deal of object-relatedness quite early,
as well as the fact that there seems to be much more capacity for differentiating self
from other than had been realized.
9
It is more likely that the neurological abnormalities in early infantile autism are
responsible for the inability to use the mother as a beacon of emotional orientation
rather than their being defenses employed by the ego. Since Mahler’s work, an abun-
dance of studies have shown that autism is a neurological illness whose etiology is
mainly genetic, most forms of which are accompanied by significant mental retarda-
tion (Piven and Folstein 1994). Studies such as Mahler’s further strengthened the
mistaken etiological concept that autism and “childhood schizophrenia” are caused
by failures in the mother-child interaction.
Martin S. Willick
Now fast forward to the year 2000. Are we making the same kind of
mistakes again? I do not really know the answer to this question, but
I would like to extend this cautionary tale to our current work. I believe
it is fair to say that the very ideas that led to erroneous notions of the
etiological determinants in schizophrenia are still very much with
us and may be causing us to be too certain about the causes of other
disorders.
As I have said, the mistakes that concern me are of two kinds.
One is the reluctance that many analysts still feel in acknowledging
biological predispositions in major psychiatric syndromes other than
schizophrenia. The other involves the etiological role accorded various
disturbances in the mother-child relationship during the preverbal
period. These are often viewed as crucial determinants for the devel-
opment of borderline conditions and very severe character disorders
such as schizoid and paranoid personality.
42
As to the f irst concern: Because I have become somewhat knowl-
edgeable about the use of medications, I have often been called on by
colleagues to consult with patients who have been in analysis for a
number of years but have not suff iciently responded to the treatment.
Most of these patients have been suffering from moderate low-grade
chronic depression, low self-esteem, and severe self-critical feelings,
and are often considered masochistic. In clinical psychiatry today
they are classif ied as having dysthymia. Some, but by no means all,
respond to antidepressant medications. In addition, studies show
that some of the personality traits found in depressed patients are
often altered when medication is successful in lifting the depression
(Black and Sheline 1997). The main point I wish to make is that in
discussing these patients with their analysts I often found a reluc-
tance to prescribe medication because suff icient “dynamic” conf licts
seemed to be present to explain the illness. In other words, these
analysts were working with an etiological model that made them doubt
that medication was needed. Recall that it was not so long ago that
psychoanalysts treating schizophrenia at hospitals like Chestnut Lodge
and Austin Riggs acted similarly (McGlashan and Nayfack 1988).
I should state here that for a number of psychiatric disorders the
decision to use medication need not rest on an understanding of
etiology. It is, rather, a clinical judgment supported not by convictions
PSYCHOANALYSIS AND SCHIZOPHRENIA
about etiology but by clinical empirical studies that have shown the
usefulness of medication for particular disorders. I should state also
that in this “heyday of biological psychiatry” mistakes are often
made in the other direction as well—some patients are mistakenly
given medication out of the conviction of a biological etiology, and
psychotherapy is neglected.
More broadly, I believe that analysts are still reluctant to consider
the role of biological predispositions or even abnormalities in deter-
mining “choice of neurosis,” a problem we are in no way near solving.
Thus, while we can note many crucial dynamic conf licts, even common
ones, in people suffering, say, from panic disorder, OCD, or eating dis-
orders, these conf licts may not be suff icient to cause expression of the
syndrome. The signif icant etiological factor in many cases of schizoid
and paranoid personality disorder may turn out to be not a def iciency
in the mother-child dyad but the patient’s position on a biologically
based schizophrenia spectrum.
The second kind of potential mistake that concerns me involves
serious character disorders such as the borderline conditions. In a
43
previous publication (Willick 1990) I pointed out that when analysts,
in the 1970s and 1980s, began to investigate the borderline group of
disorders, they often stated that their etiology could be traced back
to diff iculties in childhood development occurring later than in the
psychoses but earlier than in the neuroses. Thus, they were proposing
a developmental continuum on which inadequate caretaking played a
major role in pathogenesis. At times their ideas along these lines were
quite specif ic, as in the following statement by Modell (1968): “For
those children who have been subjected to massive neglect . . . the
outcome is likely to be a form of childhood schizophrenia. . . . What
can be reconstructed from the history of adult schizophrenia and
borderline patients is not the absence of a maternal environment
but a relative failure of the maternal environment” (p. 84).
Led by Kernberg’s seminal work on borderline personality organi-
zation, many analysts stresssed impairments in ego structure, object
relations, and defense organization occurring before object constancy
could be achieved. For many analysts it made sense that diff iculties
in the mother-child interaction during the early part of the separation-
individuation phase were the crucible for borderline functioning in
adulthood (Masterson 1976; Frosch 1983; Buie and Adler 1982; Adler
1985; Volkan 1987). Needless to say, the fact that such theories of
Martin S. Willick
etiology are not correct for the major psychoses does not mean that
they are not correct for the borderline group of disorders. Nevertheless,
I believe that the same three ideas that earlier drove the etiological con-
cepts for schizophrenia have been operating here as well. Increasingly,
these ideas are contributing also to conclusions about patients with
other serious forms of character disorder. I am concerned about the
adequacy of these formulations, especially as they invoke the mother-
child interaction during the preverbal period. One reason for this
concern is my belief that we cannot reach such conclusions from our
work in the analytic or psychotherapeutic setting. Second, I agree with
Gunderson (1984) that “developmental theories that attempt to explain
the pathogenesis of borderline personality disorder on the basis of
a phase-specif ic, parent-child interactional failure are bound to be
inadequate. The nature of the borderline patient’s psychopathology is
such that the shaping inf luences extend to multiple phases of develop-
ment” (p. 42). This was also the conclusion that my coauthors and I
put forward in our book about borderline patients (Abend, Porder, and
Willick 1983). A similar caution about the origins of borderline condi-
44
tions is to be found in the remarks of Fonagy (1996a), one of the most
respected analysts doing infant observation research. “The identif ica-
tion of what are presumed to be primitive modes of mental functioning
in individuals with severe personality disorders . . . is often regarded as
evidence for the persistence or regressive recurrence of early patho-
genic developmental experiences. Yet, even if horizontal splitting or
identity diffusion were representatives of early modes of thought, an
issue highly controversial in any case, . . . their reemergence in adult
mental functioning may be linked to later persistent trauma” (p. 405).
Example One
An interesting example of a formulation of very early ego impair-
ment is to be found in Gerald Adler’s Borderline Psychopathology and
Its Treatment (1985.) He believes that Kernberg’s formulation, which
traces the roots of the disorder to the very young child’s inability to
integrate self- and object representations, does not suff iciently explain
the terror of utter aloneness, feelings of inner emptiness, and a sense
of panic that feels like annihilation of the self, symptoms he believes
are the cardinal ones of this disorder. His own theory of the cause of
these symptoms and of the disturbance is that even prior to the integra-
tion of self- and object representations there is a relative developmental
PSYCHOANALYSIS AND SCHIZOPHRENIA
The second was that along with this rage came fears that he might “fall
into the chest of the therapist and disappear,” or that if the therapist
leaned forward in his chair he might fall into the patient’s chest and be
totally absorbed. Adler interprets the presence of “merger” fantasies
and the failure of evocative memory as ref lections of developmental
failure in the f irst year. His conclusions about the failure of “evocative
memory” in the clinical setting are derived in part, just as they were
for some of those who investigated schizophrenia, from the application
of normal child observation studies, in this case those of Mahler, Pine,
and Bergman (1975) and Fraiberg (1969).Since Adler felt that his patient
had a very serious borderline disorder, he assumed that the roots of
the disturbance had to lie in the preverbal period.
Example Two
Another example, more recent, is taken from an issue of the
Journal of Clinical Psychoanalysis devoted to the topic “The Talking
Cure and the Preverbal Period: Developmental Facts or Genetic
Fallacies?” In this example, an analyst uses her own countertrans-
46
ference response to a patient, along with an application of current
infant research, to draw conclusions about the early roots of the
patient’s diff iculties (Lally 1996).
The analyst reports on the f irst twelve sessions of an analysis in
which the patient, a young women in her late twenties, is afraid to lie
on the couch: “As we talked about her fears regarding analysis during
the second session, I started to have visual images of caring for an
infant. My off ice felt like ‘the nursery,’ the couch was ‘the crib.’ I was
aware of the intense desire to soothe her anxiety. Her anxieties struck
me like infant cries of distress. As she revealed her fears, and as
I understood them and interpreted them, I felt like a mother minister-
ing to a distressed baby. . . . I felt as if I were using words to rock, pat,
and envelope her, to soothe her inner distress” (p. 11). “I was distinct-
ly aware of feeling as if I were with a crying baby, talking as adults
do in such situations, ref lecting back the feeling state of the infant as
they see it” (p. 13).
Interestingly, this patient did not seem to be very disturbed, although
the analyst did feel she was like a “fragile bird or lost waif.” What we
learn about her history is that the patient’s mother was frequently
hospitalized for depression beginning in the patient’s infancy. She
was often cared for by an aunt, and she had many memories of lonely
PSYCHOANALYSIS AND SCHIZOPHRENIA
hours at the aunt’s house, waiting for her father to arrive. During her
mother’s absence, the patient felt desperate with loneliness and self-
blame. She believed that if she were very good, she would be able to
bring her mother home.
In order to understand her intense countertransference reaction,
the analyst turns to the infant observation research of Stern, Emde,
Brazelton, and Bebe and Lachmann, especially the latters’ work on
reciprocal responsiveness. She believes this research bolsters her con-
viction about the usefulness of her countertransference reactions in
understanding her patient and offers justif ication for her therapeutic
interventions. It is clear that she believes (although this was not the
focus of her presentation) that the mother, because of her depression,
was unable to be adequately empathic toward her infant. “I deciphered,”
writes the analyst, “an aspect of Joan’s experience in infancy, when her
depressed mother was unable to be responsive to her ” (p. 21).
In this example, conclusions are drawn about the patient’s
preverbal development from both the transference-countertransference
interaction and from the application of infant-mother observation
47
research. Remember that this patient had many memories of lonely
hours at the aunt’s, waiting for her father to arrive, feeling desperate
with loneliness and self-blame. Her analyst’s emotional response, while
perhaps intuitively empathic toward the patient’s pain, could just
as well have ref lected those later childhood years when the patient,
as noted, had the capacity for self-blame.
Not present in this formulation is the idea that since the patient is
so disturbed there must have been early failures. As a matter of fact,
there is nothing in the presentation of the case that would indicate that
the patient was “borderline” or had signif icant impairments in her
object relations or ego functioning, a fact pointed out by a discussant
of the paper (Olesker 1996).
Discussion
In both these examples we have etiological formulations that trace
the patient’s psychopathology to various disturbances in the develop-
ment of object relations in infancy, much as was done earlier with schizo-
phrenia. Indeed, had either of these patients been schizophrenic, many
analysts writing in the 1950s, 1960s, or 1970s might well have argued
for the causative role of the failure of soothing or holding introjects in
the first case, and the lack of reciprocal responsiveness in the second.
Martin S. Willick
With regard to the second case, I have heard a number of colleagues cite
the presence of serious depression in the mother during a child’s infancy
as playing an important role in the development of schizophrenia.
I think we all realize that we analysts, perhaps like other scientists,
tend to look for data that will conf irm our own theories. If one is a self
psychologist, there will be ample data to reveal lack of empathy in the
mother. If one is a classical analyst, there will be enough to indicate that
conf lict and compromise formation, especially around the drives and
their derivatives, are decisive. And so on. Likewise, if one is commit-
ted to seeing the effects on the individual of the profoundly important
preverbal period, there will be ample material in case studies to support
that proposition. McGlashan (1984), who conducted a long-term follow-
up study of patients treated for psychoses at Chestnut Lodge from 1950
through 1975, says that when social workers did the intakes on schizo-
phrenic patients during those years, they found the things they were
looking for—impairments in early relationships with caretakers (per-
sonal communication).
It is important to bear in mind that our recent knowledge about
48
schizophrenia has come not from the clinical situation but from other
ways of studying the disorder—brain imaging and genetic epidemiolo-
gy studies of families. The same can be said about recent advances
in our knowledge of bipolar disorder, autism, ADHD, and OCD. There
is also some indication that some borderline patients may have a form
of af fective illness and others a form of thinking disturbance; that
is to say, some may have a biological diathesis that interacts with
the environmental milieu, early or even later in life (see, e.g., Siever
and Davis 1991).
Let us, however, leave biological considerations aside and turn our
attention to the psychological milieu. If biological predispositions are
not playing a role, we must have some other explanation to account
for the etiology of the more severe character disorders and borderline
conditions, as against the neuroses or less severe disturbances. It seems
reasonable to assume that these sicker patients have had greater diff i-
culties earlier in life, when basic ego structure and function are devel-
oping. Surely all of us believe that not only are the f irst two years of
life extremely important, but that the mother-child dyad is one of the
most crucial determinants of the infant’s development during that time.
Nevertheless, if history is to be a guide, we must proceed cautiously
when we formulate etiological theories invoking the inf luence of those
PSYCHOANALYSIS AND SCHIZOPHRENIA
early years. First, we may not be able to conf irm such theories from
our work in the clinical situation. Second, we must be careful when we
use countertransference responses to reconstruct very early conf licts
in object relations. Third, we should be cautious in moving too easily
back and forth from normal child observation studies to the clinical
situation. (For similar caveats regarding the conclusions that can be
drawn from infant observation, see Wolff 1996.) There is reason to
believe that sexual and/or physical abuse occurring after the f irst
two years of life might be the crucial determinants for serious psycho-
pathology. In other cases, such as the two cited here as examples,
cumulative trauma or ongoing diff iculties in caretakers may be more
important than what takes place during the preverbal period.
To address questions about the role of preverbal development in the
etiology of various disturbances, we may have to turn from the clinical
situation to prospective studies that follow infants through childhood
and into the adult years. We have had a serious problem in psycho-
analysis due to the absence of any system for testing and then accept-
ing or dismissing hypotheses. As Cooper (1993) has said, “Without
49
empirical studies we have no way ever to discard a hypothesis. . . .
Psychoanalysts have leaned heavily on clinical experience to give
us conf idence in our activities, although the history of medicine is
replete with tenaciously held false beliefs based on clinical experience”
(pp. 383–384).
An example of prospective empirical studies in psychoanalysis is
the research being undertaken by Fonagy and his colleagues (1996b) on
the implications of the quality of early infant-mother relationship for
later development. Working with the Adult Attachment Interview and
the Strange Situation assessment, Fonagy seems able to predict which
children will develop insecure attachments at one year. Some effects
seem to be present at f ive years as well. Now what are needed are long-
term follow-up studies of these same children to see what they will look
like as adults. Whatever these studies show, we must keep in mind what
we all know—that development is not static. Subsequent to the prever-
bal period, we enter into later phases with all of their conf licts, and
these too can have a profound inf luence on personality development.
The inf luence of later phases of development was taken up by
McDevitt (1997) in a twenty-f ive-year follow-up of the subjects of
Mahler’s original child observation studies. McDevitt questions the
“validity of inferences about inner life, about intrapsychic experiences
Martin S. Willick
and meanings, that are drawn from largely behavioral data” during
infancy (p. 123). In one case followed up after many years, the inf lu-
ence of conf lict and compromise formation taking place after the
preverbal period is carefully described by McDevitt.
As we work toward an integration of data from the psychoanalytic
situation, from normal infant and child observation studies, and from
neurobiological research, we will have to broaden our understanding
of the complexities of the concept of etiology. It may very well turn out
that in some psychiatric syndromes, like PTSD, psychological trauma
has produced actual brain changes and abnormalities. A recent example
of the possible effect of impaired early mothering on the brains of
infants and toddlers comes from Dawson et al. (1997). They found that
at the age of 13–15 months, the infants of depressed mothers exhibit
reduced left frontal EEG activity. The long-term consequences of such
changes remain to be seen.
For analysts who still believe that the mother’s inadequacies during
the f irst year or two of life contribute to the development of schizo-
phrenia or other psychoses, such f indings will support their belief that
50
the presence of biological abnormalities in the adult do not necessarily
indicate that the etiology of the disturbance is primarily “biological” or
“organic.” However, just as with the results of the studies of “attach-
ment” disorders, the long-term consequences of neurobiological
changes in the brains of infants have yet to be determined. As I have
mentioned, we now have the opportunity to achieve an integration of
the results of studies from our own f ield, whether conducted on the
couch or in the nursery, with those of neurobiological research.
withhold medication in the belief that one must get to the psychological
roots of the problems, roots that might be obscured by medication.
However, with cases in which medication is already being used,
or in which it is clearly not necessary, do our theories of the origins of
the disturbance necessarily get in the way? This is diff icult to answer.
Evidently, analysts from very different persuasions, with very different
views of pathogenesis, are convinced of the correctness of their formu-
lations and are guided in their treatments by that conviction. But since
we do not yet know what is truly mutative in the psychotherapeutic
experience, it may very well be that for many patients the analyst’s
views of etiology are indeed not so crucial. I am reminded of a case
I presented at a panel on conf lict and def icit. My patient had been
unable to talk in a previous treatment, and I was therefore being more
actively helpful and supportive, talking to her at the beginning of every
session because she could not begin. One of the panelists, a self
psychologist, not only complimented me on my work, but indicated
that I had treated the patient much as he would have. He said, how-
ever, that my intuitive need to alter my therapeutic stance should have
51
led me to a conclusion regarding the cause of her psychopathology—
namely, the inadequacy of the mother’s empathic responsiveness. I will
often alter my therapeutic stance, not out of some conviction about
the origins of the patient’s psychopathology, but based on what I think
will facilitate the psychotherapeutic or psychoanalytic process.
I know there are many analysts who analyze transference, and con-
f lict and defense in all their ramif ications, without necessarily thinking
about etiological factors. However, even for some of them, as well as
for myself, reconstructions of patterns of relationship reenacted in the
transference are often an important part of the therapeutic work. I have
in general, I must say, become much more cautious in providing such
reconstructions. And, when they concern the preverbal period, I believe
we enter into a much more diff icult area, one we have visited before,
often to our detriment, an area in which verif ication is extremely hard
to come by.
REFERENCES
ABEND, S.M., PORDER, M.S., & WILLICK, M.S. (1983). Borderline Patients:
Psychoanalytic Perspectives. New York: International Universities
Press.
ABRAHAM, K. (1908). The psycho-sexual differences between hysteria and
dementia praecox. In Selected Papers. New York: Basic Books, 1960,
Martin S. Willick
pp. 64–69.
——— (1924). A short study of the development of the libido, viewed in the
light of mental disorders. In Selected Papers. New York: Basic Books,
1960, pp. 418–501.
ADLER, G. (1985). Borderline Psychopathology and Its Treatment. New York:
Aronson.
BAK, R. (1954). The schizophrenic defense against aggression. International
Journal of Psycho-Analysis 35:129–134.
——— (1971). Object relationships in schizophrenia and perversion.
International Journal of Psycho-Analysis 52:235–242.
BATESON, G. (1969). Double-bind—1969. Paper presented at the Annual
Meeting of the American Psychological Association, Washington, DC.
——— JACKSON, D.D., HALEY, J., & WEAKLAND, J.H. (1956). Toward a theory
of schizophrenia. Behavioral Science 1:251–264.
BHUGRA, D., HILWIG, M., HOSSEIN, B., MARCEAU, H., NEEHALL, J., LEFF, J., MALLET,
R., & DER, G. (1996). First-contact incidence rates of schizophrenia in
Trinidad and one year follow-up. British Journal of Psychiatry
169:587–592.
BION, W.R., (1967). Second Thoughts: Selected Papers on Psychoanalysis.
52 New York: Aronson.
BLACK, K. J., & SHELINE, Y.I. (1997). Personality disorder scores improve with
effective pharmacotherapy of depression. Journal of Affective Disorders
43:11–18.
BOWEN, M. (1960). A family concept of schizophrenia. In The Etiology of
Schizophrenia, ed. D.D. Jackson. New York: Basic Books, pp. 346–372.
——— DYSINGER, R.H., & BASMANIA, B. (1959). The role of the father in fami-
lies with a schizophrenic patient. American Journal of Psychiatry
115:1017–1020.
BUIE, D., & ADLER, G. (1982). The definitive treatment of the borderline per-
sonality. International Journal of Psychoanalytic Psychotherapy
9:51–87.
BURNHAM, D.L., GLADSTONE, A.I., & GIBSON, R.W. (1969). Schizophrenia and
the Need-Fear Dilemma. New York: International Universities Press.
COOPER, A.M. (1993). Discussion: On empirical research. Journal of the
American Psychoanalytic Association 41(suppl.):381–399.
DAWSON, G., FREY, K., PANAGIOTIDES, H., OSTERLING, J., & HESSL, D. (1997).
Infants of depressed mothers exhibit atypical brain activity: A replica-
tion and extension of previous findings Journal of Child Psychology and
Psychiatry 18:179–186.
DUNBAR, F. (1948). Psychosomatic Diagnosis. New York: Harper & Row.
FAIRBAIRN, W.R.D. (1954). On Object Relations Theory of the Personality. New
York: Basic Books.
FONAGY, P. (1996a). Commentaries on Peter Wolff’s paper. Journal of the
PSYCHOANALYSIS AND SCHIZOPHRENIA
sities Press.
PIVEN, J., & FOLSTEIN, S. (1994). The genetics of autism. In The Neurobiology
of Autism, ed. M.L. Bauman & T.L. Kemper. Baltimore: Johns Hopkins
Press.
REDLICH, F.C. (1952). The concept of schizophrenia and its implications
for therapy. In Psychotherapy with Schizophrenics, ed. E.B. Brody
& F.C. Redlich. New York: International Universities Press.
ROBBINS, M. (1993). Experiences of Schizophrenia. New York: Guilford Press.
ROSENFELD, H.A. (1952). Transference-phenomena and transference-analysis
in an acute catatonic schizophrenic patient. International Journal of
Psycho-Analysis 33:457–464.
——— (1965). Psychotic States: A Psychoanalytic Approach. London:
Hogarth Press / Institute of Psychoanalysis.
SEARLES, H. (1965). Collected Papers on Schizophrenia and Related Subjects.
New York: International Universities Press.
SEGAL, H. (1964). Introduction to the Work of Melanie Klein. 2nd ed. New
York: Basic Books.
SIEVER, L.J., & DAVIS, K.L. (1991). A psychobiological perspective on the per-
sonality disorders. American Journal of Psychiatry 148:1647–1658.
STERN, D. (1985). The Interpersonal World of the Infant. New York: Basic 55
Books.
SULLIVAN, H.S. (1953). The Interpersonal Theory of Psychiatry. New York:
Norton.
——— (1956). Clinical Studies in Psychiatry. New York: Norton.
——— (1962). Schizophrenia as a Human Process. New York: Norton.
TAUSK, V. (1919). On the origin of the “influencing machine” in schizophre-
nia. In The Psychoanalytic Reader, ed. R. Fliess. London: Hogarth Press,
1950, pp. 31–64.
TIENARI, P. (1991). Interaction between genetic vulnerability and family envi-
ronment: The Finnish adoptive family study of schizophrenia. Acta
Psychiatrica Scandinavica 84:460–465.
——— SORRI, A., & LAHTI, I. (1987). Genetic and psychosocial factors in
Schizophrenia: The Finnish adoptive family study. Schizophrenia
Bulletin 13:477–484.
VOLKAN, V.D. (1987). Six Steps in the Treatment of Borderline Personality
Organization. Northvale, NJ: Aronson.
——— (1997). Laying the groundwork. In The Seed of Madness: Constitution,
Environment, and Fantasy in the Organization of the Psychotic Core, ed.
V.D. Volkan & S. Akhtar. Madison, CT: International Universities Press.
W EISS , E., & ENGLISH, O.S. (1957). Psychosomatic Medicine. 3rd ed.
Philadelphia: Saunders.
WILL, O.A. (1961). Process, psychotherapy, and schizophrenia. In Psycho-
therapy of the Psychoses, ed. A. Burton. New York: Basic Books.
Martin S. Willick