In psychoanalysis, the patient's self is experienced as being made whole by means of functions performed by the analyst. More encompassing than a simple transference, this selfobject transference reveals that the transference object (in or out of the analytic situation) is a necessary extension of the self, a separate person yet experienced by the patient in some ways as part of the self. This selfobject transference can be idealizing or mirroring. The other person is seen as a mirror reflecting back the patient's grandiose false self (This other person adores me because I'm special). Or the object is idealized so that, vicariously, the patient's self is felt to attain whatever it has been lacking: I'm special because I am friends with this great person. Another transference paradigm is the twinship transference. At first seen by Kohut as a variant of the mirroring transference, twinship was later conceived as developmental line toward coherence in its own right: (I'm just like this person I adore, and he has the same feelings, attitudes, and values as me.)

Intersubjectivists[2] feel that there are two dimensions of transference, repetitive and reparative. The repetitive aspect is similar to Freud's notion of the transference as a "stereotype plate" that guides perception. And the reparative aspect is similar to Kohut's notion of selfobject transferences. This bidimensional transference occurs as the patient unconsciously seeks to repeat an old object relationship while at the same time hoping that the new version of the self will be an improved, healed version of the old one.

Lately a unitary notion of the transference neurosis has has changed. Looking at various interlocking aspects of transferences (note the plural) some  feel the concept of the transference is obsolete[3] and now support a multifaceted view. This cuts the Gordian knot: Freud himself was never fully able to reconcile two aspects of his theory -- transference causing repetition compulsions with important objects -- versus -- transference as a resistance to exploring those very dynamics of conflict and repetition.

Westen and Gabbard[4] try to integrate several aspects of transference into a coherent theory that jibes with how psychoanalysis is actually practiced. They consider (1) the role of the analyst in eliciting transference by means of "anonymity" (a.k.a. the analytic "incognito") and (2) the nature of "real" versus "transferential" components of the therapeutic relationship. Referring to recent advances in cognitive psychology, they conclude by the end of the paper that analytic "anonymity" is not the driving force behind most transference reactions (in fact, it is not even a cognitive possibility). Optimally, "A useful analytic stance is one that allows the patient's enduring dynamics to dominate the analytic field."

And rather than "anonymity" being pivotal, they find that other features of the analytic situation cause the distinction between "real" and "transferential" perceptions -- features related to power, intimacy and attachment, and sexuality. The asymmetrical conversational patterns of psychoanalytic therapy inherently confers the role of power on the analyst. In addition, analytic reserve is conducive to the patient's feeling the analyst rejects intimacy with the patient. And, for all gender combinations of patient and analyst, lying down in another's presence evokes sexual associations and feelings of sexual vulnerability.

In regard to "real" versus "transferential" components of the therapeutic relationship, they contrast two poles of the debate: One pole[5] sees the alliance as a relatively nonneurotic and conflict-free relationship of doctor and patient; the other side[6] argues that "the alliance" can actually be parsed into a variety of transferences, that "even what Freud[7] (1912, 8) referred to as the 'unobjectionable positive transference' may conceal a variety of resistances lurking behind a mask of compliance." (p 127).

What differentiates "transference" and "therapeutic alliance" is that the former catches our attention -- and should -- because it is anomalous, idiosyncratic, and tells us something....what differs between the two is the extent to which the patient's associations are culturally patterned and hence culturally shared, or are more idiosyncratically organized and hence become figure rather than ground. (p 128). 

But they do not suggest that any aspect of the relationship be exempt from analysis. As Stein[8] noted, "what may look like the 'real relationship' is likely to be complex and overdetermined." Clinically significant transference patterns involve an "object-relational constellation," consisting of self representations, object-representations molded by prior experience, "affects and motives associated with this self-in-relation constellation (in which self- and object representations can also be reversed) ... patterns of interaction, and defenses against ... unpleasant affect." (p 117).

Analysis of countertransference enactments and working-through play a large part in both diagnosis and treatment. Patients may "actualize an internal scenario" that results in "the analyst's being drawn into playing a role scripted by the patient's internal world." The patient may emit subtle interpersonal signals that pressure the analyst "to respond in a manner similar to that of the past objects projected onto the analyst."

Projective identification is seen by object relations theorists as the patient projecting aspects of the self into the analyst. But since the patient doesn't project into an empty container, there may need to "be a 'hook' in the analyst that facilitates the introjection" of the projected contents (p 101). Renick[9] similarly focuses on transference enactments and the interplay of transference and countertransference as both a reflection of the patient's dynamics and an opportunity to rework them. But he notes that skillful use of the tool of transference analysis requires the therapist to pay attention not just to distortions of process in the therapy but also to the larger role they play in the patient's life outside the treatment.

Westen and Gabbard take the position that transference reactions are neither cognitive constructions of the patient-analyst dyad nor co-constructions of relatively equal partners but, rather, that transference reactions are constructs from the sum of the following:  the "patient's enduring dispositions to react in particular ways under particular conditions" -- plus -- real "features of the analytic situation and the analyst" unrelated to "anonymity" -- plus -- the interactions between the analyst and the patient (p 99). 

They end on a cautionary note (p 129): "Nor do we have reason to assume that the patient's most important dynamics will invariably present themselves 'in the transference,' any more than we have reason to assume that anything that presents itself  'in the transference' is likely to be therapeutically valuable." In short, beware oversimplified notions of transference.

Problems with the diagnostic utility of the concept

How sensitive is the concept of transference, and how specific is it? Are these medical diagnostic distinctions even clinically meaningful in psychodiagnosis? Consider the following dynamic formulation paraphrased from a paper on transference: 

Ms K learned early to connect wishes for her father's attention with his negative reactions to her assertiveness. Therefore, in relation to nurturing and authority figures, she came to defend against expressing her needs directly in ways that might reawaken her father's rejection. Her inability to see the analyst as a different kind of parental figure thus arose from her conflict between the wish for nurture, her anger over needing it, and the fear of rejection. Her sham passivity reflected a defensive solution that left her feeling like an impostor when she received positive attention and ugly when she did not. 

Such formulations may be very sensitive. This one certainly captures a version of Ms K's pathology, but it is not very specific. Such formulations populate the psychoanalytic literature, and you can almost just fill in the blanks. 

Mr A probably learned early to connect wishes for his mother's ___________ with her negative reactions to his __________. Therefore, he came to defend against interpreting responses from parental or authority figures in ways that might reawaken her _________. His inability to see the analyst as a different kind of parental figure thus arose from his conflict between the wish for ___________ and the fear of ________, and reflected a defensive solution that left him ___________.

This is the "too generic" school of formulating transference, and while its productions are always applicable, the contrasting school has its own pitfalls: In the "clever school," as we might call it, each formulation is, yes, unique -- as any individual is unique. But the causal connection between past and present (and the patient's resulting attitude toward the therapist and other important objects) is so idiosyncratically tethered to the case report that what it illustrates is not general enough for the reader to relate to his or her own clinical practice. Such formulations are presented in the literature as evidence of some fine point of psychoanalytic theory but often they seem applicable only to an n = 1.

Important to bear in mind about case reports of transference is this paradox: If a formulation is the generic type, it's not going to be very interesting. If it's a clever one-off formulation, it may be interesting but it probably won't help you deal with or even recognize transference in your own clinical practice.

The point is, say Westen and Gabbard, watch for transferences in the plural: "...transference patterns that emerge over the course of a treatment always have a 'day residue' (the influence of current experience in the analytic process). This residue sometimes provides 'noise' that may lead the analyst" astray in formulating the source and nature of the patient's behavior. If the patient is reacting in the present to a real event in the analytic relationship, this is not transference. Transference involves the stirring up of "enduring patterns of thought, feeling,  motivation, affect regulation, or behavior...." (p 113)

Clinically significant transference paradigms tend to "involve the activation of an object-relational constellation [including] self-representations; representations of others influenced by prior experience; affects and motives associated with this self-in-relation constellation" along with patterns of interpersonal interaction and defenses against warded-off affects. (p 117) But once the transference neurosis is established, it becomes an intrapsychic structure unto itself. Bird -- most concisely -- describes it from the inside: 

When I think of transference, I think of feelings, reactions, and of a repetition of past events; but when I think of transference neurosis, I think literally of a neurosis. A transference neurosis is merely a new edition of the patient's original neurosis, but with me in it. (p 281)

Transference in its universal sense is the displacing of feelings and attachments from one object to another, and of repeating the past in the present. In this process two separate identities -- the patient's parent and the analyst -- are merged, but the patient's own identity and the analyst's identity remain clear and separate.

Apropos the notion of "day residue" in the transference, Kern[10] says that elements of the transference neurosis have all of Freud's components of dreamwork (condensation, displacement, plastic representation, secondary revision) and are analyzable as compromise formations in exactly the same way as dreams. "Thus whether the analyzable content ... is symptom on one occasion or transference on the other, we are, at bottom, addressing the same process. ...a newly constructed psychodrama for two." (p 346)

Next:  theoretical speculation.