Chapter 1
The Defensive and Restorative Traditions
The term regression has been used in many different ways since Freudâs early (1900) discussion of the formal, temporal and topographical aspects of regression in dreaming. Two dominant traditionsâhere termed the âdefensiveâ and ârestorativeâ traditionsâhave evolved, along with several integrative traditions, in the psychoanalytic literature.
The individual retreats, in a defensive regression, from a situation of unconscious danger to an earlier fixation of relative safety and pleasure. He replaces current wishes and dangers with the more manageable, if still problematic, ones from an earlier phase of life. The process of regression is likened metaphorically to the flow of a stream that washes back on itself (Freud, 1905) or to the retreat of an army to an old camp (Freud, 1917a).
In a restorative regression, by contrast, the subject experiences a primary yearning to re-establish early conditions of merger with a life-sustaining other who is felt to be an environment. This is not a defensive process. It is mobilized by feelings of safety, not of danger. Such regressions can be likened metaphorically to a merger with Thalassa, the primal sea spirit (Ferenczi, 1938), a harmonious mix-up in the motherâinfant dyad (Balint, 1968), or the inhale of the otherâs pride and joy as a kind of oxygen (Kohut, 1984).
In Balint, Winnicott and Kohutâs writing we find similar attempts to integrate the defensive and restorative regression traditions into a comprehensive model of mental functioning. They rely, in spite of their differences, on a common âdevelopmental tiltâ approach (Mitchell, 1988). Dependency needs are located in the pre-Oedipal years of development. Restorative regressions are thus thought to imply a return to pre-Oedipal needs. Sexual and aggressive conflicts, and retreats from them, are considered Oedipal in nature. Defensive regressions are thus thought to imply a higher level of developmental achievement.
In spite of its explanatory appeal, however, the developmental tilt highlights problems internal to the concept of regression. Above all, it raises the problem of the genetic fallacy (Hartmann, 1955). We find ubiquitous relational needs reduced to their original manifestations in early life in spite of widespread clinical findings that such needs are pervasive among so-called pre-Oedipal and Oedipal patients across the life cycle. The reduction suggests an implicit a morality of maturity vis-Ă -vis dependency needs. The internal break down of this solution re-opens an inquiry into the problems of regression.
The Tradition of Defensive Regression
The psychoanalytic concept of regression first appears in Freudâs Interpretation of Dreams (1900). âWe call it âregression,ââ he wrote, âwhen in a dream an idea is turned back into the sensory image from which it was originally derivedâ (p. 543). According to the theory, the process integrates three different kinds of regression. There is a âformalâ regression from complex ideas to primitive sensory images, a âtopographicalâ regression from the preconscious to the conscious systems, and a âtemporalâ regression from present to past psychic structures. The three kinds of regression were held to be âone at bottom,â because, âwhat is older in time is more primitive in form and in psychical topography lies nearer to the perceptual endâ (p. 548).
Freud developed the concept further in his Three Essays on the Theory of Sexuality (1905). Here he introduces a psychic forceâthe libidoâthat can progress or regress along an organically determined line of development. The libido is defined as a quantitative variable representing âprocessesâ and âtransformationsâ in sexual excitation (p. 83). It flows from a narcissistic reservoir of self-cathexis into oral, anal and genital object investments. Along the way, it forms fixation points of relative safety and gratification. If it encounters obstaclesâfor example, danger situationsâin its forward path, it can regress, washing back like a blocked âstream,â into the âchannelsâ of its old fixations (p. 170).
He later suggested that the ego marches in step with the libido (Freud, 1916â1917). If the libido behaves like a stream, the ego behaves like an army. It establishes camps along the path of libidinal development to which it can retreat, as needed, from conflict on the front line.
The concept of regression became an essential diagnostic tool. In hysteria, for example, there was thought to be a regression to the âprimary incestuous objectsâ of the OedipalOedipal phase (Freud, 1916â1917, p. 342). In the obsessional neuroses, by contrast, there was held to be a deeper regression from Oedipal danger to an anal-sadistic fixation wherein the libidinal impulse was âobliged ⌠to disguise itself as a sadistic impulsionâ (pp. 343â344). Paranoia, in this theory, reflected an even deeper regression from object love to narcissism. It left a dry riverbed of object libidoânot enough to form a transference to the analystsârendering the patient inaccessible by the psychoanalytic method (Freud, 1912, 1914).
Ego psychology expanded in various ways upon the work of regression in human development. We find, for instance, normal and pathological regressions along the developmental lines of id, ego and superego (A. Freud, 1965), progressive and regressive ego adaptations (Hartman, 1939), regression in psychic differentiation and integration (Hartmann, Kris, & Loewenstein, 1946), regression in the service of the ego (Kris, 1936) and, more recently, the anti-regressive function of the ego (Sandler & Sandler, 1994b).
âPsychoanalysis,â Gill (1954) wrote, in a statement that, for many, captured the essence of clinical practice, âis that technique which, employed by a neutral analyst, results in the development of a regressive transference neurosis and the ultimate resolution of this neurosis by techniques of interpretation aloneâ (p. 775). Implicitly excluded from analysis were those patients considered to be too regressed to tolerate the analystâs neutrality or, in the most withdrawn and psychotic cases, to form a neurotic transference.
At the same time, we also find in ego psychology a certain complication of the concept of regression: Its potential adaptive-progressive value emerges, such that the concept no longer refers so clearly to a return to the past. Hartmann (1939) wrote, for instance, that even when mature forms of mentation fail, and a regressive adaptation presents itself to fill the gap, âno pure form of a previous developmental stage emergesâ (p. 50). The regressive mentation has, paradoxically, a novel quality in the present. It is a solution to a new problem and, in its novelty of function, becomes part of new psychic structure.
Loewald (1981) seemed to acknowledge the latent conceptual problems when he wrote that, âwords and concepts (such as regression and the rest) reveal themselves as flimsy or enigmatic or ambiguous if they are not merely used as shortcuts and as currency in the exchange of thoughts but are given weight and consideration in their own rightâ (p. 22). He argued in favor of retaining the concept as a shortcut. One might counter-argue, however, that it is by exploring latent conceptual flimsiness that space opens for new thought.
Although Melanie Klein departed from the ego psychological reading of Freud, she too framed regression as a defensive retreat from situations of internal conflict and danger. In her writing, however, the standard picture of regression is complicated by regressive-progressive oscillation between the paranoid-schizoid and depressive positions:
For if persecutory fear, and correspondingly schizoid mechanisms, are too strong, the ego is not capable of working through to the depressive position. This forces the ego to regress to the paranoid-schizoid position and reinforces earlier persecutory fears and schizoid phenomena.
[Klein, 1946, p. 15]
The oscillation between positions is not phase specific. It can occur within any phase of life. It does not, that is, imply an erotogenic regression. This complicates the picture. It arguably opens the door for a re-imagining of regressions as shifts that occur within phases of the life cycle. I pursue this direction in Chapters 4 and 5.
Klein also implicitly challenged the concept of regression in her contention that the Oedipus complex manifests in oral phase of the first year of life (Klein, 1928, 1945). âThe frustration experienced at the motherâs breast,â she wrote, âleads both boy and girl to turn away from it and stimulates the infantâs desire for oral gratification from the penis of the fatherâ (Klein, 1945, p. 408). It is a matter of controversy whether the infant has the mental capacity to form a phantasy of the fatherâs penis. Whether or not Kleinâs assertion is correct, it suggests that the Oedipal situation is not bounded within the confines of the phallic-genital phase, but, on the contrary, can occur in so-called pre-Oedipal phases. It is, in essence, an issue of triangular love relations that can occur in any phase.
Britton (1992) explores this opening in his contention that the Oedipal situation is ânever finished,â but has to be âre-worked in each new life situation, at each stage of development, and with each major addition to experience or knowledgeâ (p. 38). We do not regress to the Oedipal situation; we rework it over and over again.
The Tradition of Restorative Regression
In 1914, Ferenczi attempted an interpretation and elaboration of Freudâs libido theory. Ferencziâs reading, published in 1938 under the title Thalassa, indulged in what he admitted to be a loose, but arguably fruitful, play with mythical and biological metaphors. He argued that libidinal striving expresses a universal longing to return an âaquaticâ mode of existence, which the infant first experienced in the womb, and which was first experienced by proto-organisms in the ocean. There is a trace of the sea in the womb, so to speak, and one is always trying to get back to the state of repose experienced there. âThe human being is dominated,â he wrote, âfrom the moment of birth onwards by a continuous regressive trend toward the reestablishment of the intrauterine situationâ (p. 20).
For Freud and the ego psychologists, psychosexual development followed a path from primary narcissism to differentiation and ultimately genital investment in a separate object. Ferenczi held, by contrast, that genitality was not an achievement of the differentiated ego, but, on the contrary, the expression of a deep regressive longing for merger. âThe purpose ⌠of the sex act ⌠can be none other than an attempt on the part of the ego ⌠to return to the motherâs womb, where there is no ⌠painful disharmony between ego and environmentâ (p. 18).
Just as Freud appealed to the metaphor of the stream in formulating the libido theory, Ferenczi appealed to the metaphor of Thalassa, the primeval sea spirit who spawned the first life forms. The motherâs body symbolizes the primeval ocean from which life emerged, and to which it longs to returnâto a state of harmony prior to the trauma of separation. The libido is not just pleasure seeking, then, but also Thalassa-seeking.
Balint (1959, 1968) acknowledges the influence of Ferencziâs Thalassa more than any other theorist in the restorative regression tradition. He draws explicitly on the metaphor of Thalassa in his depiction of the motherâinfant dyad:
The child safely held in its motherâs arms, the intrauterine, and the thalassal existenceâare to be considered as symbolically identical; the friendly expanses of my theory are then but wish-fulfilling memories of these states; conversely, these states activate a strong attraction for regression.
(1959, p. 96)
âBasic faultâ patients, that is those who suffered early traumas of missatunement in the motherâinfant dyad, require a restorative regression to the thalassal harmony of primary love:
The patient must be allowed to regress ⌠to the particular form of object relationship which caused the original deficiency state. Only after that can the patient âbegin anew,â that is, develop new patterns of object relationship.
(pp. 165â166)
As for the analyst, he
must do everything in his power not to become, or to behave as, a separate, sharply countered object. In other words, he must allow his patients to relate to, or exist with, him as if he were one of the primary substances. This means he should be willing to carry the patient, not actively, but like water carries the swimmer ⌠that is, to be used without too much resistance against being used.
(p. 167)
The regression here is not a defensive retreat. To interpret it as such would, in fact, constitute a resistance, on the analystâs part, to the terror of self-loss in a merger with the patient. It must be recognized and met as a primary striving to facilitate the patientâs growth.
Winnicott and Kohut avoid any such explicit link to Ferencziâs Thalassa. Kohut (1959), in fact, dismissed it for its speculative, non-psychoanalytic character. One can, nonetheless, discern a congruent theoretical trend in their conceptualization of restorative regression.
Winnicott (1954a, 1954b, 1955) described primary regression, for instance, in the analysis of borderlines, to a state of dependency in which the analyst becomes a good-enough mother in a holding environment. âIn so far as the patient is regressed,â he wrote, âthe couch is the analyst, the pillows are breasts, the analyst is the mother at a certain past eraâ (italics in original) (1954a, p. 288). The analystâmotherâs provision constitutes a holding environment in which spontaneous gestures of omnipotence, destruction and play can emerge and be survived without too much impingement or retaliation. The relaxation of false self-structuresâthe legacy of impingement in the original motherâinfant dyadâallows for the emergence of a more vigorous and authentic self.
Kohut (1971, 1977, 1984) argued, along similar but different lines, for the remobilization and working through of regressive selfobject strivings in the analytic relationship. As the term, a selfobject striving involves a merger between oneself and the object. The merger is not a retreat from differentiation, but a source of nourishment and growth. In the most archaic expression of the mirror transference, for example, âthe analyst is experienced as an extension of the grandiose self and he is referred to only insofar as he has become the carrier of the grandiosity and exhibitionism of the analysandâs grandiose selfâ (1971, p. 114). There is a diffusion of boundaries such that the analystâs unobtrusive and affirming responsiveness become part of the patientâs self-structure.
Where Ferenczi and Balint drew on oceanic imagery, and Winnicott on the imagery of the motherâs lap, Kohut (1984) enlists the environmental imagery of oxygen. âA move from dependence (symbiosis) to independence (autonomy) in the psychological sphere is no more possible, let alone desirable,â he wrote, âthan a corresponding move from a life dependent on oxygen to a life independent of it in the biological sphereâ (p. 47). In the revival of thwarted regressive selfobject strivings, the analyst becomes like the airânot a separate objectâbut a vital substance absorbed into the patientâs self-structure.
The Developmental Tilt
The defensive and restorative regression traditions fit togetherâ awkwardlyâin what Mitchell (1988) called the âdevelopmental tiltâ solution. A diagnostic response, it attempts to differentiate between the supposedly more infantile disorders of relational trauma, deficit and arrest (e.g., borderline, schizoid and narcissistic personalities) and the allegedly more mature disorders of conflict (e.g., hysteria and obsessional neuroses). It suggests that relational needs are primary in infancy and drive conflicts central in later childhood. Early formed relational defenses/adaptations become the uneven ground, in Mitchellâs metaphor. Upon this uneven ground are built the leaning toward drive conflict.
Balint, Winnicott and Kohut present different variants on the developmental tilt solution. The classical hysteric and obsessional neurotic (see Freud, 1916â1917, pp. 342â423) remain, for each of them, the prime examples of the relatively well structured conflict disorders. The âbasic faultâ disorders (Balint, 1968), the narcissistic personality and behavior disorders (Kohut, 1971; Kohut & Wolf, 1978), and the borderline patient (Winnicott, 1965) become, in different ways, privileged examples of the early deficit disorders.
The earlier disorders are thought to reflect relational trauma in the parentâchild dyad and are therefore believed to necessitate a restorative regression in the analytic setting. The aim is to provide an unobtrusive environment in which development can resume. The higher disorders are thought to reflect greater structuralization. The patient has achieved a relatively stable, differentiated ego. He relates to his objects as separate objects. Typically, he has entered into an Oedipal conflict and stayed there or regressed defensively. To work with this more sophisticated patient, one must establish a therapeutic alliance (Zetzel, 1956; Greenson, 1965), that is, a collaborative identification between the patientâs observing ego and the analystâs analytic function (Sterba, 1934). It is not the primary vehicle of therapeutic action; rather, like the unobjectionable positive transference (Freud, 1912), of which it is the theoretical heir, it is a prerequisite for the successful analysis of motive, conflict and defense. The aim here is ego expansion and integration.
In his later writing, Kohut (1977, 1984) questioned the validity of this diagnostic app...