Learning Center
Plans & pricing Sign in
Sign Out



									(1995). Psychoanalytic Review, 82:229-247
    Projective identification and Countertransference Interventions:
                           Since Feeling is First
                                             Karen Maroda
since feeling is first
who pays any attention
to the syntax of things
will never wholly kiss you
— e.e. cummings

Projective identification cannot be discussed without first granting that it is ill-defined and often
misinterpreted. Yet it holds a certain mystique for most of us who do therapy. We keep discussing
and redefining projective identification because we want to truly grasp what is happening in a
session when we are deeply, and perhaps suddenly, seized by an intense feeling that we were not
expecting to have. We know that our affective experience is a response to the patient we are
treating, but often we are uncertain about the reason for, or the origin of, our feeling. Has the
patient stimulated something in us that was lying dormant, waiting to be awakened, or is he simply
―depositing‖ his feelings in us? Is the experience of intense affect as precipitous as it often seems
or is it the culmination of a series of communications between therapist and patient?

We are also curious about the patient's intent. Is the patient trying to communicate with us, trying
to rid himself of unbearable feelings, or trying to hurt or seduce us? Is there a primary motive or is
the projection of strong affect usually multi-determined? Is the patient communicating with us and
thereby facilitating the treatment, or is he trying to defeat us by forcing us to experience and
possibly act on the most destructive feelings that he can muster? Recent discussions of projective
identification focus on these questions in an attempt to solve the riddle of projective identification,
as well as to formulate techniques that respond to its nature and purpose.

When Melanie Klein coined the term projective identification in 1946, she was referring to the
child's fantasy of ridding himself of unwanted feelings by assigning them to someone else. This
early notion of projective identification was purely intrapsychic and did not include the stimulation
of feeling in another person. It was not meant to be a communication. Rather the child simply
thought the unwanted attribute away and was able to ease his mind through this defensive act.
However, most current definitions of projective identification focus on the treatment setting and
refer to the stimulation of strong affect in the analyst or therapist. As with many psychoanalytic
terms, the name remained the same, but the concept changed dramatically as it evolved from an
intrapsychic to an interpersonal event. And, with this metamorphosis, the question of motivation
became more complex. Originally, the only motivation was ―to get rid of‖ an emotion. But, as an
event between two people, projective identification held the promise of communication.

In 1966, Malin and Grotstein, elaborating on the seminal contribution of Bion (1959), emphasized
the notion of normalcy with regard to projective identification, citing it as a part of the growth
process in everyone. According to them, all children project onto their parents the feelings that they
cannot yet manage. The task for the parent is to be capable of that which the child is not. That is,
the parent needs ―to contain‖ (Bion 1959) the child's projected affect so that the child can identify
with the parent and internalize this activity. Malin and Grotstein point out that the child is then
ready to move on to higher levels of functioning, which will undoubtedly involve repeated instances
of projective identification, offering the child the opportunity to use the adult as a sort of conduit for
achieving affective mastery. In accordance with their view of projective identification, adults who

have not accepted and integrated their own affective experience continue to use projective
identification to manage their unacceptable emotions. This deficit in affect integration produces
―impaired affect tolerance and an inability to use affects as self-signals,‖ according to Stolorow and
colleagues (1987, p. 72).

Malin and Grotstein are well aware that being stuck in a primary mode of projective identification
indicates a high level of pathology

   -   230 –

that is found mostly in patients suffering from psychoses and from borderline and some narcissistic
personality disorders. They rightly point out that the mechanism itself is normal in that it occurs
naturally in children and represents an attempt at growth and development. Earlier, Bion (1959)
made the point that it can and will also be used by normal adults when they are overwhelmed by
their own affective experiences and, as such, is a part of all communication. It becomes
pathological when someone reaches adulthood without having attained a reasonable degree of
success in integrating his or her affect. And, if this person enters therapy, it will necessarily bring
the full force and power of an adult mind and a lifetime of repressed aggression to bear on the

Tansey and Burke (1989) echo the views expressed by Malin and Grotstein, and outline a step-by-
step program for therapists that emphasizes the communicative and constructive aspects of
projective identification, as part of their approach to understanding countertransference. They
encourage therapists to see their patients as unconsciously trying to communicate split off affect.
Therapists should be empathic both toward the patient and themselves, as they will inevitably
experience strong countertransference reactions to their patients that are likely to be unpleasant
and difficult. Tansey and Burke advise therapists to accept the projected feelings willingly, to
experience the intended affect, to sort out what is coming from the patient (as contrasted with what
is idiosyncratic to the therapist) and, finally, to use the insights to make enlightening
interpretations. They caution against disclosing the countertransference, advising ―silent use‖ of it
for the purpose of guiding interpretations. This, incidentally, is a position widely supported by those
who write about projective identification. Though it is beyond the scope of this article to review the
overall literature on countertransference, current views on countertransference, as it relates to
projective identification and related therapeutic interventions, are discussed.

Ogden (1982) sees projective identification as both a form of communication and a form of
defense. He believes that it is a genuine challenge for therapists to contain and respond to
projective identifications, and notes how often treatments fail because of a therapist's inability to
cope with the feelings that are stimulated in him.

- 231 -

Failure to adequately process a projective identification is reflected in the therapist's response in
one of two ways: either by his mounting a rigid defense against awareness of the feelings
engendered, or allowing the feeling or the defense against it to be translated into action. (p. 32)

Ogden's examples of the therapist ―translating into action‖ include socializing with the patient,
giving or receiving gifts, and breaching confidentiality. He acknowledges that interpretation is often
unproductive, and advises therapists to be tolerant and understanding when a patient is
functioning at a preverbal level. His only suggestion regarding technique with such patients is as

Under such circumstances, the therapist must rely on noninterpretive interventions and
management of the therapy to convey his silently formulated understandings of that which the
patient has unconsciously asked him to both contain and return to him in a form that he can utilize.
(p. 74)

Grinberg (1979) was one of the early writers on the link between projective identification and
countertransference, but he views projective identification and the therapist's response to it as
being a function primarily of the patient. Calling the therapist's response ―projective
counteridentification,‖ he describes the therapist as a passive recipient of the projective
identification which is ―projected, lodged, or forced‖ into him. Grinberg views projective
identification as a rather malignant process that threatens to destroy the treatment. As he says:

One might say that what was projected, by means of the psychopathic modality of projective
identification, operates within the object as a parasitic superego which omnipotently induces the
analyst's ego to act or feel what the patient wanted him to act or feel in his unconscious fantasy. (p.

Like many of his analytic colleagues, Grinberg believes that interpretation of the patient's wish to
have the analyst feel what he feels is the technique of choice when confronted with projective

Kernberg (1987) says that projective identification is a primitive defensive operation ―based on an
ego structure centered on splitting as its essential defense‖ (p. 798). He differentiates between
projection and projective identification on the basis of affective stimulation in the therapist: only
projective identification produces affect in the analyst. While not seeking to pursue the varied views
concerning the differences that may or may not exist between projection and projective
identification, my use of the term projective identification is similar to Kernberg's, referring only to
those times when intense, unexplained, and ego-dystonic affect is stimulated, usually repeatedly,
in the therapist or analyst. (This does not mean that therapists cannot and do not use projective
identification toward their patients as well. But the focus of most writing, and of this article, is on the
patient's projective identifications rather than the therapist's.)

Kernberg is quick to recognize that interventions other than interpretation are often required when
dealing with the patient's projective identifications, but he also cautions analysts against
―countertransference acting-out.‖ Although he is vague about what analysts should use instead of
interpretation, he is very clear about why interpretation often fails. He says, ―The patient typically
resists the analyst's efforts at interpretation because of the dread of what had to be projected in the
first place‖ (p. 801) — a point that might seem obvious, yet is rarely made. Logically, patients who
are able to accept and acknowledge interpretations regarding projective identification are on the
verge of knowing and accepting the feelings that were formerly split off.

But what about the patient who is far from ready to cognitively recognize and accept the split off
emotions? What is the analytic therapist to do? It seems to me that the whole transference-
countertransference mess that often follows a projective identification is made worse by therapists
who insist on beating their beleaguered patients over the head with interpretations they are
convinced are correct. And, as the patient becomes increasingly agitated and unreceptive to these
interpretations (even when they are correct), some therapists end up compounding their error: they
insist that the patient is resisting their interpretations, which is no doubt due to a desire on the part
of the patient to defeat the therapist and ruin the treatment.

Following along this line of thought, Finell (1986) believes that the term projective identification is
often used by therapists to avoid responsibility for their own countertransference reactions. Rather
than owning up to their feelings in response to an admittedly difficult patient, some therapists are
fond of speaking of their patients, à la Grinberg, as malevolent souls who force experiences of
hopelessness, despair, lust, and rage down their therapists' throats. Objecting to this tendency to
―dump‖ on the patient, Finell states:

The interaction between two personalities communicating on a projective-introjective level is open
to many different influences. No two analysts will react identically to the same patient. … I believe it
is an oversimplification to propose that any therapist would react identically to a particular patient.
Neat theoretical conceptualizations may satisfy the analyst's need for intellectualized closure, but
the richness and nuances of the therapeutic interaction are lost. The ultimate value of a clinical
concept lies in the degree to which it fosters accurate understanding and analytic skill. Projective
identification runs the risk of shutting down rather than opening up the therapist's attitude to the
patient and the clinical material. (p. 106)

In his discussion of Finell's ideas, Whipple (1986) considers the concept of projective identification
to be too useful to discard. But he admits that it is often used to exonerate the analyst: it helps him
to avoid the real issue of countertransference and, perhaps even more importantly, the patient can
be blamed for the therapist's untoward reactions.

It seems to me that both Finell and Whipple's points are well made, and I agree with Whipple when
he says that it does not make sense to throw the term out because it has been used defensively in
the past. Getting rid of the term projective identification will not stop therapists from being
defensive, nor will it stop patients from trying anything they can to communicate their feelings to us.
It is regrettable that the term that is used to describe a patient's desire to communicate disavowed
affect is so intellectualized, if not pejorative. Something more descriptive would be helpful (but,
alas, Schaefer [1976] and others have been making this point about psychoanalytic language for
some time). Realistically, a new term is not likely to be adopted. What can be done, however, is to
focus on the function of projective identification in the interpersonal and clinical sense, as well as
the therapist's response to it, instead of intellectualizing or creating new jargon (such as projective
introjection, projective counteridentification, and introjective identification) to define what we think
we are talking about.

It seems safe to say that most, if not all, analytic clinicians accept the existence of the mechanism
we call projective identification.

- 234 -

It may even be safe to say that the current literature emphasizing the communicative and
constructive aspects of projective identification, rather than the pathology and potential for
destruction, also reflect the views of a majority of clinicians. If this is true, then most therapists
believe that a patient tries to communicate disavowed affect so that the therapist can experience it,
understand it, and find a way to live with it — the idea being that if the therapist can achieve this
affective task, then so can the patient.

So, the problem for us as analytic therapists is this: How do we let a patient know that his or her
communication has been received? The challenge for the therapist is to acknowledge receipt of
the patient's message in terms that the patient will understand. If the patient does not get our
message, then the required cycle of communicated affect, that begins with projective identification
and ideally ends with affective integration, will not be complete. Failure on the therapist's part to
participate in affective communication often results in frustration for the patient and renewed efforts
on his or her part to send a new message, with the hope that the next one will be received and
acknowledged. And when the patient, believing that the message has not been heard, feels forced
to repeat himself, subsequent messages are usually stronger and louder. A patient who continues
to ―up the ante‖ by being more and more provocative is desperately trying to get a response from
his or her therapist that is not forthcoming.

The question remains: What response is truly most therapeutic? Traditionally, we have believed
that most of our communications to patients should take the form of interpretation. Yet virtually
everyone who writes about patients who use splitting and projective identification notes that these
patients cannot use an interpretation because they are typically or temporarily regressed at a
preverbal level. Kernberg is not alone in saying that if a patient could acknowledge and discuss his
disavowed affect, then he would have no need for projective identification in the first place. Ogden
(1982) notes that the patient not only rejects interpretation but sometimes experiences it as
―dangerous and inassimilable.‖ Balint (1968) in his classic discussion of the ―basic fault,‖ says,

At the Oedipal — and even at some of the so-called preoedipal levels, a proper interpretation,
which makes a repressed conflict conscious and thereby resolves a resistance or undoes a split,
gets the patient's free associations going again; at the level of the basic fault this does not
necessarily happen. The interpretation is either experienced as interference, cruelty, unwarranted
demand or unfair impingement, as a hostile act, or a sign of affection, or is felt so lifeless, in fact
dead, that it has no effect at all. (p. 175)

Lomas (1987) elaborates on what he calls the ―limits of interpretation,‖ noting how inadequate it is
as an all-purpose intervention of choice with all patients, and calling for therapists to expand their
repertoires for greater effectiveness.

If we accept that interpretation does not work well with patients who rely heavily on projective
identification, or with any patient who, regardless of diagnosis, is communicating through projective
identification at a particular moment in time, then how are we to let the patient know that we
understand? If we believe the argument regarding the futility of using verbal interventions with a
patient who is regressed to a period before language was acquired, then our interventions must
change accordingly. (Of course, therapists do respond on a preverbal level, through their facial
expressions and other body language, to regressed patients. However, this is often unintentional
and it is not the same as directly providing a vocal and deliberate expression of affect on the
therapist's part.)

Casement (1985) believes that projective identification is a form of affective communication, and
he notes how important it is for the patient to know that the therapist has, indeed, received his
affective message. He does not, however, suggest that an affective response from the therapist is
the best way, or even an acceptable way, to communicate to the patient that his or her affective
message has been received.

My own clinical experience has convinced me that actually expressing the affect experienced in
response to the patient is often the most therapeutic intervention possible. What I am proposing
regarding technique is that the therapist or analyst freely, yet responsibly, respond to the patient by
disclosing the strong affect he or she is experiencing in response to that patient, provided that the
patient appears to be unconsciously ―provoking‖ such disclosure. When I speak of disclosing or
expressing the affect, I am referring to an affective response from the therapist, a show of genuine
emotion, rather than an intellectual statement of the therapist's feelings. For example, to say to a
patient, ―I am angry with you right now,‖ while remaining cool, calm, and collected, would not
qualify as an expression of therapist affect. The affect must be present and congruent with the
therapist's description of his or her emotional state. Naturally, using this method effectively
depends heavily on the therapist making the correct reading regarding the patient's need for an
affective response. But I believe this is more easily done than one might imagine. The basis for
deciding whether or not to disclose affect to a patient is done by assessing the degree to which the
therapist is being stimulated through projective identification.

I believe that a strong stimulation, particularly if it is repetitive, is meant to stimulate an affective
response in the therapist, in the interests of completing the cycle of affective communication
between patient and therapist. This potentially enables the patient to experience the therapist's
struggle to manage and express the affect that was too dangerous and overwhelming for the
patient. It also lets the patient know that he or she is having a strong impact on the therapist,
something that represents both a wish and a fear for the patient who is using projective
identification. On one hand, he desperately needs his therapist to understand and experience his
affect, both to prove to him that his therapist can do this and to confirm his impact on the therapist.
On the other hand, the patient's worst fear is that he will be too dangerous and too powerful — that
if the therapist actually does experience his affect then one or both of them will be overwhelmed
and the relationship will be destroyed.

The challenge for the therapist is to show and express feeling without losing control, something
that the patient is convinced is impossible. This truly provides a model for identification purposes
that the patient can use in life. For a patient to observe his therapist experiencing and
constructively expressing his or her affect means that someday the patient may be able to do the
same. I believe that this is what patients are looking for when they stimulate strong affects in us.
The implicit message from the patient is this: ―I can't experience the full power of this emotion and
function. Can you?‖ The fact that this resembles an affective game of ―hot potato‖ is not lost on
most therapists. Very often, the intensity of the patient's split off affect is quite difficult to bear and
presents a considerable challenge to the therapist. To the extent that the patient's affect stimulates
existing conflicts or complimentary narcissistic vulnerabilities in the therapist, the situation
becomes even more complex and difficult, and certainly requires a willingness on the part of the
therapist to take responsibility for his or her emotional responses.

Yet I am convinced from my own clinical experience that responding with an expression of
countertransference affect, provided that I am reasonably in control and can be responsible with it,
is the most effective way to intervene with patients who cannot make use of interpretation. And I
often find that an insight-producing interpretation eventually follows an emotional confrontation with
a patient, and usually comes from the patient. It is as if the affective exchange, once completed
successfully, clears the heaviness and sense of oppression and hopelessness from the session,
creating an opening for new growth and development. The patient, who seemed so impossible and
so infantile, along with the analyst or therapist, who may have been behaving in a similar manner,
are suddenly able to talk to one another once they have openly felt with each other. This is
contrasted with the frustrating stalemates that often occur when a provocative patient only
produces an increasing determination in the analyst or therapist not to show emotion, in the belief
that somehow this constitutes being defeated by the patient.
To clarify the variety of ways in which the patient may unconsciously stimulate the therapist, I
would like to give a few examples.

Jane, a bright, hard-working, 30-year-old woman sought therapy to work out her problems and to
be a ―better wife.‖ Her husband, a mental health professional, had referred her after convincing her
that she was not only depressed (which was quite obvious) but also confused, uncommunicative,
and unable to take responsibility in the marital relationship. Jane said her marital problems,
including her husband's abuse of drugs and alcohol, were all ―her fault.‖ Each session she would
relate some current argument with her husband that demonstrated her inadequacy and ―wrong-
thinking‖ in the relationship. It soon became quite evident that her husband was ―gaslighting‖ her,
convincing her that any negative perception she had of him was really a distortion based on her
own pathology. He would also severely criticize her and verbally bully her into submission,
whereupon she would become severely depressed and self-blaming.

After a few of these stories, told by this rather docile, self-effacing, and suffering patient, I found
myself responding first with anger and, ultimately, rage. Initially, I was angry at the husband for his
psychological abuse of my patient, particularly since his professional training gave him such an
unfair advantage over her. But I eventually became angry at my patient, too, for being so passive
and masochistic. I was outraged at this man's treatment of his wife and her willingness to accept
his behavior as normal. I worked at containing my emotions and gave interpretations and asked
questions that explored Jane's feelings rather than imposing my own. She said that perhaps her
husband was a bit harsh at times, but that she only felt guilt and shame because she was so
inadequate. She could not blame him for getting fed up with her.

I finally reached a point where I could not stand it anymore and I let her know how angry I felt, at
both her and her husband, when she would describe their arguments to me. I showed anger,
telling her that I thought her husband was terribly abusive and that I couldn't believe that she didn't
get angry at him. When I expressed my anger, I was careful not to criticize and humiliate her for
being such a willing victim, which would have left me treating her not much better than her
husband did. Rather, I emphasized how angry I was at her husband for treating anyone the way he
treated her and expressed my disbelief and anguish at seeing that she did not seem able to stop
him. At first she weakly denied both her anger at him and the extent of her pain, yet her facial
expression registered relief and pleasure at my expression of rage and disgust — so I knew I was
getting somewhere. Eventually Jane owned these feelings for herself and even stood up to her
husband, admitting fault when it was hers, but steadfastly refusing to take responsibility for his
feelings and his behavior. I doubt that this evolution would have occurred if I had not been willing
to express Jane's split-off anger for her until she was able to feel and communicate it herself.

From discussions with colleagues and reading case histories in the literature, it seems that many
therapists or analysts report having experienced a positive outcome as the result of a strong
affective exchange in treatment, though not necessarily because it was planned. What seems to
often happen is that the therapist is mired down with a difficult patient who finally succeeds in
provoking the therapist to the point of disclosure, even though the therapist was working to prevent
this. Having had an emotional outburst toward the patient, the therapist fears having made things
worse, only to discover that it actually made things better, even if it scared the patient a little.
However, most therapists are not fond of feeling overwhelmed and perhaps a bit out of control, and
may even have vowed not to let something like this happen again, viewing the positive outcome as
idiosyncratic or just plain lucky.

The position advocated here is that success at such moments had nothing to do with luck but was
a function of giving the patient exactly what he or she needed. Rather than avoiding the exchange
of emotion with patients, the therapist can work with his or her own feelings to gain control and
finesse in expressing them for the patient's benefit. This type of intervention makes sense if you,
as the therapist, believe: that the patient is trying to elicit a response from you in the interests of
furthering rather than sabotaging the treatment, that the patient will know what you are feeling
whether or not you freely express and take responsibility for it, and that psychoanalytic
psychotherapy or psychoanalysis are interpersonal as well as intrapsychic events that require
insight into the relationship that exists between patient and therapist.

Before giving an example, I want to emphasize that the technique suggested here depends on the
therapist responding honestly, directly, and in a reasonably controlled manner, to the affect that is
being stimulated, as opposed to what the patient may verbally be clamoring for. Though an
adequate presentation of technique is too lengthy to be presented here (I have outlined the
technique for disclosing and analyzing the countertransference elsewhere [Maroda, 1991]), I think
it important to note the paradoxical effect that often occurs in treatment, e.g. patient demands for
love stimulate hate, clinging stimulates anger and the wish to flee, self-hatred often stimulates
tenderness and affection, etc. It is therapeutic to reveal the emotion that has actually been
stimulated, even if this is uncomfortable for the therapist and in conflict with what the patient claims
to want or need.

Each therapist's responses will be somewhat different to a particular patient, yet they are also likely
to be similar enough to others' past responses to be valuable in terms of reality testing for the
patient. Over-reactions and other idiosyncratic responses by an individual clinician are preferably
admitted to the patient, rather than suppressed in the hope that the patient will not know what the
therapist or analyst is actually feeling if it is not stated. However, the personal history that accounts
for the analyst's idiosyncratic response is preferably not given unless absolutely necessary for the
sake of understanding. The first priority is one of communicating the patient's disavowed affect to
him and any technical considerations should address this and contribute to this effort. It goes
without saying that any violation of traditional therapeutic boundaries, including sexual or
aggressive acts on the part of the analyst or therapist, are not considered as acceptable
countertransference interventions.

One of the basic notions inherent in applying this technique is that the patient has not yet found
someone who can honestly tolerate and express the patient's disavowed affect, leaving the patient
doomed to reenacting certain scenes again and again until someone can intervene in an honest,
responsible, yet emotional way. The following example sheds light on the difficulty in tolerating the
very strong affects characteristic of adult projective identifications, and the potential positive
outcome when countertransference affect is expressed.

Mary, a highly intelligent and enormously intense person, formed a very early and strong
attachment to me. Though in her late 30's when she began her treatment, she was alone because
of her inability to sustain an intimate relationship or close friendship. She was narcissistically
vulnerable with borderline features. She worked at being soft-spoken and reasonable, yet was
known by her friends and co-workers to be excessively critical, perfectionistically demanding, and
prone to fits of hurt and rage when disappointed. She rarely expressed any deep feeling
voluntarily, showing intense emotion only when she lost control.

She had been raised in a religious Catholic family and firmly believed that her sexual and
aggressive feelings, as well as her desire for power, recognition, and love, were unChristian and
shameful. Forced by this upbringing to deny her own nature, she relied heavily on projective
identification in her communications with everyone. I, of course, was no exception.

By the third year of a four-session-per-week analytic treatment, she had made enough progress to
openly declare her love for me and her desire to have me with her forever. At this point, however,
she was just beginning to take in the reality that this would never happen. Though she showed
some progress in her ability to begin to face the separation that was inevitable, she came to this
realization kicking and screaming. As the reality of our relationship became more apparent, I could
literally observe her emotionally and physically trying to contain the hurt and rage that she was
feeling. Convinced that I would reject her violently if I saw the extent of her rage, she fought to
restrain it.

An extremely prolonged period of discord and unpleasantness resulted. Mary was distraught and
inconsolable every day that she came to her sessions. When she entered my office she typically
threw me an angry glance, then sat down and glared silently at me. Eventually she would begin
talking, usually telling me how horribly unhappy her life had become and how the treatment had
only made her worse. She said she had accomplished nothing more than repeating the painful
past of loving her mother and being unable to inspire her love, finally leaving home feeling
abandoned and unwanted. She alternated between criticizing and blaming me for her
unhappiness, pitifully crying and demanding that I do something to help her.

At these times I had tried all manner of intervention. I interpreted to her that she felt overwhelmed
by the feelings of the past that she was now experiencing with me — and that it seemed so unfair
for me to ―withhold‖ expressions of love and affection, just as her mother had done. She agreed
with this interpretation, but this had little positive effect, and sometimes even fueled her rage at
feeling cheated. I also interpreted her fears of separation and loss, including the fear that I would
reject and abandon her if she looked for love elsewhere. This interpretation had more impact on
her, and would stimulate some reflection on her part, yet she would inevitably return to being
clinging and demanding.

I also let her know that I was sorry to see her in so much anguish, yet could not possibly meet her
demands that I soothe, comfort, and love her. Occasionally, she would respond positively to some
expression on my part of empathy, which forced her to stop accusing me of being indifferent to her.
But only for a moment. She would return for her next session newly armed for combat, having
successfully shrugged off the effects of the previous day's attempts at compassion and

For myself, I began to show noticeable signs of fatigue and stress. I felt I had a good grasp of what
was happening, which served to protect me from insanity. Yet, the constant wear and tear of
Mary's meagerly disguised rage was unmistakably takying its toll. Clearly I had underestimated her
capacity for unrelenting anguish. Finally, after several weeks, I began to break down. I was
becoming depressed and no longer wanted to go to my office in the morning. In addition, I was
developing psychosomatic symptoms. I had headaches, backaches, slept poorly, and generally
didn't feel well. Mary predicted that I would eventually tire of her and end the relationship. I began
to wonder how I was going to keep that prophecy from coming true.

On a day when I was feeling particularly exhausted and had noticed a tender, enlarged spot on my
neck, I called my internist, wondering if something wasn't really wrong. He examined me and told
me that I was exhausted and needed a vacation. I said, ―But what about my neck?‖ and he
explained that I had managed to produce a knot in a major muscle that runs up the neck. And this
knot had created the soreness and the headaches. For me that was the last straw. I left my
internist's office telling myself that something had to change, and soon. Enough was enough.

That same afternoon I saw Mary and, not coincidentally, she began her session by saying that she
was having a recurrence of a gastrointestinal problem that she had had prior to beginning
treatment. Having already had two surgeries for this problem, she expressed her fear that she
would again ―twist her guts all up‖ with anguish. She demanded that I do something to insure that
this didn't happen. My response was to say that, oddly enough, I had just come from seeing my
doctor and that it seemed obvious to me that our relationship was making us both sick. Mary
looked startled but stared at me quietly and intently. I went on to say that there was nothing really
wrong with me at the present time, but that my body was certainly registering the high level of
distress that I felt in relation to her. I added that this was clearly true for her as well. I let her know
that I was upset and that I could take no more. With exasperation in my voice, I told her that I
couldn't tolerate anymore of her criticism, demands, and despair, and that I was beginning to hate
spending time with her.

By way of taking responsibility for my own responses to her, I also acknowledged that more than
once I had distanced to protect myself, which only fed her fears of abandonment and her rage at
me for not doing enough or being available enough. I said I didn't want to distance from her, but
that sometimes I couldn't help it. I could only take so much.

This case example is quite an involved and complex one that could be discussed further at great
length, but the point I want to make here is that only when I said, ―That's it, I've had enough‖ and
showed the intense anger, hatred and frustration that went with those words, did I have a favorable
impact on Mary. She needed me to experience and protect both of us from her self-destructive
impulses. She knew she was feeling depressed, rejected, hopeless, and alone. (What she couldn't
know she felt, and thus needed me to feel for her, was her hatred of herself and me and her desire
to destroy this relationship that was making her feel so sick and unlovable.) Rather than admit to
herself and to me that she in some ways wanted to make this happen, she presented herself in
such a way to me that it seemed likely that I would make it happen, because I would be at my wit's
end and feel I had no other choice. She needed me to express all the feelings she was having
including, and especially, the wish to end the relationship precipitously for the purpose of ending
our struggle and to free herself from a relationship that could not possibly meet her needs.

The situation with Mary represented both a reenactment of the past, and a plea for me to respond
in a way that no one else ever had. Over the years, unable to achieve emotional closeness with
her mother, Mary settled for eliciting sympathy and being placated by her mother as a substitute for
love. Her mother was fond of saying ―Poor Mary, you have it so hard‖ which only perpetuated
Mary's poor self-esteem and dependency. Her mother offered this quasipity, encouraged her to eat
too much, and paid attention to her when she threw tantrums, and this became the ―currency‖ of
their relationship, rather than genuine closeness and caring. Mary perpetuated this in her adult life,
settling for people giving her her own way, yet unconsciously knowing they hated her for it, as her
mother probably had also.

I believe that Mary came to hate her mother for this situation, yet, like her mother, she felt too guilty
to acknowledge this. Rather she disguised her rage and wish to destroy and leave her mother
through clinging, remaining overweight, and failing to establish intimate relationships with others. I
believe that Mary continued to stimulate hate and the desire to flee in others in the hope that
someone would finally express and acknowledge the hatred that she disavowed, thus freeing her.
Once this was accomplished in her therapy she was able to understand what had happened and
gain insight into her current behavior. My experience with Mary and other regressed patients has
proven to me that interpretations are not heard by patients using projective identification without
first receiving an expression of my feelings toward them. As Stolorow et al. (1987) say,

… the tendency for affective experiences to create a disorganized … self-state is seen to originate
from early faulty affect attunement, with a lack of mutual sharing and acceptance of affect states.
(p. 72, emphasis mine)

I believe that an essential aspect of the patient's need for an expression of the therapist's affect is
related to his or her unheeded desire to be heard and responded to by the parents. I have
never treated a patient who frequently used projective identification who did not basically
feel powerless to have an emotional impact on others. Convinced of their powerlessness, such
patients often project this experience onto their therapists, but with the secret hope that the
therapist will prove them wrong by responding with strong feeling. Since I believe that this
knowledge of interpersonal power and strength is so important, I do not share the concern that
many people have about the therapist's idiosyncracies and pathology. My feeling is that the
therapist has a duty to take responsibility for his or her own feelings in the interpersonal
therapeutic setting and, if it is done well, a little pathology will not ruin things. No one in the
patient's world will ever meet the ideal of mental health and I think it an unrealistic and
unnecessary standard for therapists to expect themselves to meet. Rather, I think we have a
responsibility to maintain the therapeutic boundaries and to be as honest as we can with our
patients and ourselves, particularly as it concerns our limitations.

As a last note on the need for the therapist's emotional expression in response to the patient's
projective identifications, I would like to quote Grotstein (1981), someone I admire for his
willingness to be dramatic and even poetic in his description of the analytic situation. He explains
why the patient needs to know that he has succeeded in stimulating feeling: There are
certain feelings which are so constructed that they seem to be beyond words and may, therefore,
have been before words were first experienced. Powerful feelings are more often than not
expressed by giving another person the experience of how one feels. Throughout the course of
human history, dialogue and confession seem to have been the dominant forms of emotional
release. All human beings seem to have the need to be shriven, that is, to be relieved of the
burden of unknown, unknowable feelings by being able to express them, literally as well as
figuratively into the flesh … of the other so that this other person can know how one felt. The sadist
and murderer desire to see the look of agony on their victim's face so as to be sure that the
murderer's own tortured experience can be transmitted through the network of projective
identification to the victim whose agonized face completes the communication… How else can a
beleaguered patient know that his analyst understands than if he suffers the experience which the
patient lacks the words to describe? (p. 202)

Though it is not an easy thing to do, especially when we know that our patients want us to be in
pain, I believe that the most therapeutic act is one of submitting to the emotional experience
that they want us to have. To help our patients who cannot bear what is inside them, it is likely
that we have no recourse other than to bear it for them, letting them know how and why it is
happening and to see what they have wrought.

Upon being presented with their own disavowed feeling, most patients respond with a dramatic
recognition of what has taken place that goes beyond either their or my ability to describe it. What
follows is often an expression of deep relief, as well as remorse for having caused pain to the
therapist or others in the patient's life. Often the knowledge of having caused such pain to others
has also been disavowed and the therapist's affective expression opens not only new avenues of
understanding and insight, but also of compassion and empathy.

Balint, M. (1968) The Basic Fault. London, Tavistock. [→]
Bion, W. R. (1959) Attacks on linking. Int. J. Psycho-Anal., 18: 170-189. [→]
Casement, P. (1985) On Learning from the Patient. London, Tavistock.
Finell, J. S. (1986) The merits and problems with the concept of projective identification.
Psychoanal. Rev., 73: 103-120. [→]
Grinberg, L. (1979) Countertransference and Projective Identification. In L. Epstein and A. Feiner
(Eds.), Countertransference: The Therapist's Contribution to the Therapeutic Situation (pp. 169-
191). New Jersey, Jason Aronson.
Grotstein, J. (1981) Splitting and Projective Identification. New York, Jason Aronson.
Kernberg, O. (1987) Projection and Projective Identification: Developmental and Clinical Aspects.
J. Amer. Psychoanal. Assn., 35: 795-819. [→]
Klein, M. (1946) Notes on Some Schizoid Mechanisms. In Envy and Gratitutle and Other Works-
1966-1963. (pp. 1-24). New York, Delacorte.
Malin, A., & Grotstein, J. (1966) Projective Identification in the Therapeutic Process. Int. J. Psycho-
Anal. 56: 163-177. [→]
Lomas, P. (1987) The Limits of Interpretation. What's Wrong With Psychoanalysis? New York,
Penguin Books.
Maroda, K. (1991) The Power of the Countertransference: Innovations in Analytic Technique.
London, John Wiley & Sons.
Ogden, T. (1982) Projective Identification and Psychotherapeutic Technique. New York, Jason
Stolorow, R. et al. (1987) Psychoanalytic Treatment: An Intersubjective Approach. The Analytic
Tansey, M., & Burke, W. (1989) Understanding Countertransference, From Projective Identification
to Empathy. New Jersey, The Analytic Press.
Whipple, D. (1986) Discussion of ―The Merits and Problems With the Concept of Projective
Identification‖ by Janet Finell. Psychoanal. Rev. 73: 121-128. [→]


To top