Richard L. Rubens, Ph.D.



Beginning in the early 1940's, W. Ronald D. Fairbairn developed a unique psychoanalytic theory that anticipated and laid the groundwork for some of the most important current theoretical advancements in psychoanalysis. At the heart of Fairbairn's theory was a notion of endopsychic structure based directly on the vicissitudes of human object relatedness --in a way so radically different from other theories of his time that it is only now, a half-century later, that his ideas are finally having their appropriately profound influence on the general spectrum of psychoanalytic thinking.

In an earlier paper (Rubens, 1984), I advanced the position that Fairbairn had not been studied as widely and thoroughly as might be expected due to the extent to which his ideas depart from classical analytic theory. While increasingly many psychoanalysts had been drawn to Fairbairn's insights into the nature of human interactions and their implications for clinical practice, surprisingly few allowed themselves even to realize the extent to which these insights were based on a radically novel understanding of the human psyche --and fewer still could recognize and acknowledge the full implications of his departures.

It was my contention that it was Fairbairn's complete rejection of Freud's structural theory (and the drive model it embodied) that explained this almost phobic avoidance of the deeper implications of Fairbairn's ideas. The theory of structure is the key issue in defining psychoanalysis in general, and in distinguishing between psychoanalytic theories in particular. Thus, to accept Fairbairn's theory in the fullness of its structural divergence from Freud was to abandon Freud in too radical a way for many psychoanalysts. Also, most psychoanalysts had been so habitually attached to speaking in terms of Freud's tripartite division of the psyche into id, ego, and superego that they failed to notice that this structural theory was based on metapsychological assumptions that they themselves no longer in fact adhered to.

In recent years, there has been a growing awareness of the viability --and even necessity-- of alternatives to the metapsychological assumptions embodied in Freud's structural theory. This change is expressed in the perspective developed by Greenberg and Mitchell (1983) that there are two, very different basic models on which psychoanalytic theories are based:


The most significant tension in the history of psychoanalytic ideas has been the dialectic between the original Freudian model, which takes as its starting point the instinctual drives, and a comprehensive model initiated in the works of Fairbairn and Sullivan, which evolve structure solely from the individual's relations with other people. Accordingly, we designate the original model the drive/structure model and the alternative perspective the relational/structure model. (p. 20)


Mitchell (1988, p. 18) describes Fairbairn as one of the "purest representatives" of this relational/structure model.

Although a very large percentage of modern psychoanalysts actually have underlying assumptions far more consistent with those of the relational/structure model, there remains a tremendous inertia toward preserving a connection to the drive/structure model --or, at least, utilizing the terminology of that model.

The typical use which has previously been made of Fairbairn's ideas has been to note their relevance to early development and to those conditions most directly deriving from these stages (i.e., schizoid, narcissistic, and borderline states), while maintaining that the later developments can still be satisfactorily described employing the traditional drive/structure model. Even British object-relations theorists such as Winnicott (1965) have attempted to retain their connection to classical theory through just this sort of adherence to the importance of the drive/structure model in later development. Mitchell (1988) provides a brilliant discussion of the shortcomings of this manoeuvre, which he terms "developmental tilt." (pp. 136 ff.)

Fairbairn himself, while radically departing from Freud's metapsychological assumptions, was nevertheless guilty of employing terms taken too directly from the language of drive theory. He repeatedly utilized terms like "ego" and "libidinal" in crucial positions in his theories, although they bear virtually no similarity to their original meanings in Freud. Even his use of the term "object" is misleading, since it does not begin to convey how extensively it departs from the drive/structure model's concept of object. Although careful to redefine his use of such terms, Fairbairn's use of the language of drive theory did introduce a great deal of confusion into the understanding of his work --and a considerable opportunity for avoiding the full impact of its novelty.

Nevertheless, Fairbairn did succeed in completely abandoning Freud's structural model. Moreover, in a still more radical way, he developed a new structural theory based on a very different notion of the psyche and of the underlying meaning and role of structure within it. It is only in recent years that psychoanalysis has finally begun to incorporate directly the full implications and novelty of Fairbairn's theoretical innovations.

This paper will attempt to explore the actual extent of Fairbairn's departure from traditional notions of psychic structure by a detailed explication of his own theory of endopsychic structure in light of the assumptions out of which it was developed and the clinical implications which derive from it.


The Basic Nature of the Self


Fairbairn viewed people as being object-related by their very nature. For him, the fundamental unit of consideration was that of a self in relation to an other --and the nature of the relationship in between. Personhood, in the external world, essentially and definitionally involves relationship with other people. Internally considered, the self therefore is to be understood as always existing in and defined in terms of the relationships it has, remembers, desires, or creates. In the relational/structure model of Fairbairn, the shape of the self grows and changes from its experience in relationships, while at the same time the nature of the relationships it has are being shaped and changed by that self.

Fairbairn's theory gives appropriately great weight to the significance of intrapsychic functioning. Unlike some interpersonal theories, it is no way guilty of naively reducing the study of the human psyche to a mere examination of external relationships. His relational/structure model provides room for the most extensive and rich of notions of inner world. Furthemore, as will be discussed below, Fairbairn viewed the self not simply as the result of experience, but rather as the precondition for it. In an irreducible way, the self is the pre-existent starting point for all experience and provides continuity in all that develops later --coloring and shaping all subsequent experience. On the other hand, Fairbairn firmly maintained that it was in relationship to others that the self expresses its selfhood and is shaped in the course of its development. Fairbairn's theory of self is, therefore, "relational" in precisely the way described by Mitchell (1988), in which


the interpersonal and the intrapsychic realms create, interpenetrate, and transform each other in a subtle and complex manner. (p. 9)


It is the self in its relationship to the other that constitutes the only meaningful unit of consideration for Fairbairn. This unit of self, other, and the relationship in between becomes the pattern for Fairbairn's understanding of the form of all subsystems within the self.


The Inseparability of Energy and Structure


Central to Freud's conception of the organization of the psyche is the primary existence of an energic, chaotic entity, the id, the fundamental principle of which being the immediate and indiscriminate discharge of its stimulus-related and endogenous excitation, and the subsequent evolution of a highly structured ego, adaptively derived to mediate contact between the psyche's energic underpinnings in the id and the realities of the external world (Freud, 1900, 1923, 1933). In this way, Freud separated the structure for achieving self expression from that energy within the self which strives to be expressed.

Fairbairn adopted as his most fundamental postulate the notion that structure and energy were inseparable: "both structure divorced from energy and energy divorced from structure are meaningless concepts" (Fairbairn, 1952, p. 149). The structure is that which gives form to the energy, and the energy does not exist without a particular form. For him, "impulses" (a term he characteristically set off in quotation marks to indicate his discomfort with this notion of energy treated as through it possessed some independent and separate existence)


cannot be considered apart from the endopsychic structures which they energize and the object relationships which they enable these structures to establish; and, equally, "instincts" cannot profitably be considered as anything more than forms of energy which constitute the dynamic of such endopsychic structures (p. 85).


In Fairbairn's system, the structure for achieving self expression is inextricably interrelated with that which strives for expression. The self is simultaneously structure and energy, inseparable and mutually inter-defining.


The Object-Related Nature of the Self


Even in Freud's late description of the id (1933), the reservoir of energy within the psyche was seen as seeking at all times the reduction of tension through the immediate and indiscriminate discharge of its energy. This pattern was termed by Freud the pleasure principle. In it, there is virtually no consideration of the object towards which this discharge takes place. The pleasure principle was seen by Freud as being developmentally prior to operation in accordance with the reality principle --a mode more co-ordinated with the specific nature of the world of external objects and involving delay of gratification, planning, and purposive awareness of cause and effect and of future consequence.

Fairbairn (1952, pp. 149f.) understood Freud's position to be a direct consequence of his divorcing of energy from structure, for what goal could there be for structureless, directionless energy other than indiscriminate discharge for the purpose of homeostasis. For Fairbairn, having initially postulated the inseparability of energy and structure, it followed that the goal (or aim) of self-expression could no longer be viewed as mere tension reduction (the discharge of energy, ending the "unpleasure" of excitation and thereby definitionally resulting in pleasure) with little or no reference to the object by means of which this discharge is accomplished. Rather he completely inverted Freud's position, maintaining that relationship with the object was itself the goal, and that the pleasure involved was a secondary consequence. Thus he wrote that, "The function of libidinal pleasure is essentially to provide a signpost to the object" (1952, p. 33), and that "The real libidinal aim is the establishment of satisfactory relationship with objects" (p. 138).

To Fairbairn, the pleasure principle, rather than being the universal first principle of self expression, "represents a deterioration of behaviour" (1952, p. 139). The rightful mode of libidinal expression, at all developmental levels, is more closely related to that described by Freud as the reality principle, at least in so far as this expression is seen as always purposively intending towards relationship with objects in some realistic way, rather than towards pleasure itself:


Explicit pleasure-seeking has as its essential aim the relieving of the tension of libidinal need for the mere sake of relieving this tension. Such a process does, of course, occur commonly enough; but, since libidinal need is object need, simple tension-relieving implies some failure of object-relationships (p. 140).

Central to this theory is the concept that human beings do not naturally operate with the goal of reducing tensions, but rather with the goal of self expression in relationships with other human beings. This view of fundamental human motivation is one of Fairbairn's most important contributions to contemporary relational theory.


Unitary and Dynamic Origin of the Psyche


Fairbairn maintained that the genesis of the human psyche lay in "an original and single dynamic ego-structure present at the beginning" (1952, p. 148); or, as he wrote elsewhere, "The pristine personality of the child consists of a unitary dynamic ego" (1954, p. 107). The individual elements of these statements are important enough to the theory to merit expansion and explication.

It is first necessary to note again that Fairbairn's use of the term "ego" is in no way equivalent to Freud's structural use of the term. Rather, it refers to the entirety of the psychic self. In adopting this connotation of "ego," Fairbairn is closely paralleling Freud's use of the term prior to his writing The Ego and the Id. As Strachey (1961) points out, Freud in this period used the term to apply to the whole of a person's self. Nevertheless, it would be better if Fairbairn had substituted "self" for "ego" to distinguish his usage from Freud's. To minimize any possibility for confusion, and to emphasize the differences inherent in Fairbairn's conception, I have utilized "self" rather than "ego" wherever practical.

That Fairbairn refers to this primitive state as a "dynamic ego-structure" or "dynamic ego" follows directly from his postulate of the inseparability of energy and structure. He could not posit, as had Freud, an unstructured supply of energy out of which an adaptive structure would subsequently develop. Rather he insisted on the innate structural integrity of the self: the self was a "singular" and "unitary" whole. Further, this self was the a priori condition of life experience: "original" and "pristine", it existed from the very outset and was not in any way dependent upon experience for its existence.

Combining these notions with Fairbairn's idea that psychic energy is object-seeking, the resulting conception of the psyche is that of a self-generated, unitary center of definition and energy, with the potential for, and the drive toward, self-expression outward into the object world, and the potential for experiencing that world, its own self-expression, and the resulting interaction between the two.


The Nature of Endopsychic Structure


The self as it has been described above requires no further structural development. It begins in a condition of wholeness, already capable of and actively involved in the self-defining processes of self-expression and of experience. While this assertion naturally does not imply that the capacities of this primitive self are fully matured, it does insist that they are all present at least in seminal form.

Fairbairn acknowledged that structural differentiation in fact does occur within the psyche --and even that it is unavoidable and universal (1954, p. 107). The substructures resulting from such differentiations he saw as modelled after the self as a whole: each is comprised of an element of self in energic, affective relationship with an element of the object world. He termed these resultant substructures of the self "endopsychic structures."

Fairbairn noted (1952, Chapter 4) that certain unavoidable features of early human experience lead universally to the establishment of two such endopsychic structures: the first formed around the experience of the self in intolerably exciting relationship, and the second formed around intolerably rejecting relationship.

He understood that each of these subsystems of the self represents a particular crystallization of what originally was the growing and continually self-defining process of the self as a whole. Whereas the original self is in ongoing and essentially unbounded relationship with the outside world as a whole, such an endopsychic structure is a particularized aspect of that self, in specific relationship with a particular aspect of the object world. Fairbairn eventually came to realize (1952, p. 158) that it was the entirety of such a subsystem which constituted the endopsychic structure set up within the self. The first of the two such endopsychic structures referred to in the preceding paragraph will here be termed the Libidinal Self, as Fairbairn never developed an explicit terminology to refer to the entirety of the subsystem composed of what he termed the Libidinal Ego in specific relationship to what he called the Exciting Object. Similarly, the second subsystem will be termed the Antilibidinal Self (following Fairbairn's later terminology for the Internal Saboteur and its Rejecting Object).

The third element in Fairbairn's picture of the structurally differentiated psyche will here be termed the Central Self, consisting of Fairbairn's Central Ego in relationship with the Idealized Object. This entity is what remains of the original self after the other two parts have been separated off. Because of this unique aspect of its origin, as well as for other differences discussed below, the Central Self is not an "endopsychic structure" in the same sense as the other two entities.

The fact that Fairbairn's model of endopsychic structure is tripartite naturally invites comparisons to Freud's structural model --and, of course, certain congruence is to be expected, since both metapsychological models attempt to describe the same clinical phenomena. Nevertheless, Fairbairn repeatedly rejected such comparisons (1952, pp. 106 f., 148, etc.).

Freud's ego rather closely corresponds to the "ego" component of Fairbairn's Central Self, in that the ego is the organization of purposive self-expression and experience in relationship with the external world. It was viewed by Freud as a derivative structure, however, and not as the original structure Fairbairn viewed as the source of all other endopsychic structures. It must be agreed, that, as Kernberg (1980, p. 81) maintains, the ego psychologists' notion of an undifferentiated ego-id matrix existing prior to the emergence of either individual structure furthers the Freudian model in a direction more consonant with that of Fairbairn. Nevertheless, the ego-psychological viewpoint still posits the eventual developmental necessity of the progressive structural differentiation of the ego from the id. In so doing, it clearly differs from Fairbairn's understanding of structure. Furthermore, the metapsychological foundations of the ego-psychological view still rest on a drive/structure model --albeit one that recognizes the central importance of relationship in achieving this end-- whereas Fairbairn's metapsychology is founded on the need for self-expression in relationship.

The differences become more striking in comparisons drawn with the other two endopsychic structures. The Libidinal Ego, while certainly id-like in many aspects of its functioning, is consistently viewed by Fairbairn as existing in dynamic relationship with the Exciting Object; and the Libidinal Self which is constituted by this relationship is a proper subsystem of the Self, in that it is specifically object-related in a manner foreign to the concept of the id. The Libidinal Self represents a particularized relation of a specific aspect of the self in relationship with a specific aspect of the object world, and not the more generalized, freely displaceable and mutable energic center which the id is conceived as being. The superego is somewhat related to the Rejecting Object of the Antilibidinal Self, although not coterminous with it.

The Rejecting Object does contain the more archaic elements of the superego, although the moral aspects of superego functioning are related more to the relationship with the Idealized Object which occurs in the Central Self and to what Fairbairn discussed as the mechanism of the moral defense. Moreover, the superego concept emphasizes the object component of the Antilibidinal Self, and not the Antilibidinal Ego component --it therefore being necessary to include the ego's relationship with the superego to make a more appropriate comparison.

The ego-psychological branch of object relations theory (most ably represented by Jacobson and Kernberg) has attempted, with considerable success, to transform Freud's metapsychology in a direction more consonant with the insights of Fairbairn. Yet it is not possible fully to incorporate Fairbairn's insights without abandoning central tenets of Freud's metapsychology, contrary to the claim to this effect made by Kernberg (1980).

Freud's structural model simply is not the same as Fairbairn's system of Central, Libidinal, and Antilibidinal Selves. Nor do the modifications introduced by Ego Psychology suffice to make Fairbairn's system subsumable under their revised drive/structure model. In the first place, the "self-component" of endopsychic structures is not the equivalent of "what we would now call a self-representation", as Kernberg claims (1980, p. 81). One of the most brilliant of Fairbairn's insights lies precisely in his recognition that the self --and not some ideational representation (for who, in that case, would be the one doing the representing?)-- has as its primary, innate function active expression in the form of relationship with the object world --and not, until the intervention of some pathological process, with some ideational representation thereof! To alter this conception is to eschew the most essential thrust of Fairbairn's theory.

It is precisely Kernberg's refusal to acknowledge this difference which leads him to cite the criticism put forth by Winnicott & Khan (1953) of Fairbairn's concept of primary identification (which he described as a relationship between the self and object which has not been differentiated from it):


If the object is not differentiated it cannot operate as an object. What Fairbairn is referring to then is an infant with needs, but with no "mechanism" by which to implement them, an infant not "seeking" an object, but seeking de-tension, libido seeking satisfaction, instinct tension seeking a return to a state of rest or un-excitement; which brings us back to Freud(p. 332).


The self in Fairbairn's theory is a living, growing, self-defining center which he viewed as the point of origin of human psychic process; and, it follows directly from this most basic of principles that it is possible for such a self to have relationships with other human beings, even though they have not yet representationally differentiated as objects separate from the self. Initially this self relates to the world with little basis in experience for self-object differentiation. Nevertheless, it does express itself and experience the world in a manner that is precisely the prototype for all later activities of the self. To assert that this brings Fairbairn's theory back to the pleasure principle of Freud is totally to miss his point.

It is an actual fragment of the self, and not a representation of it, which comprises the essence of an endopsychic structure in Fairbairn's theory. As a subsystem of the self, such a structure is a purposive entity with its own energy. It is not reducible, as Kernberg (1980) suggests it is, to self and object representations energized by "an activation of affects reflecting...drives in the context of internal object relations" (p. 80). Such a view is quite closely related to Freud's drive/structure model, modified to include the notion of the expression of drive derivatives in object-relational constellations --but is not at all the same as Fairbairn's relational/structure model.

The Libidinal and Antilibidinal Selves differ from the original self in only two ways. The first difference is that each is a crystallization of what in the original self was a more freely developing potentiality. Whereas the original self (and later the Central Self, in a more limited way) was free to experience the world and express itself in relationships to that world, the subsidiary selves carry within them a pre-existing template (based on the experiences out of which they were formed) for particularized relationships with specific aspects of the world. As in the case of the Central Self, the Libidinal and Antilibidinal Selves continue to seek experience and self-expression through relationship. In the case of the Libidinal and Antilibidinal Selves, however, this process is sharply restricted by the fact that the particularized crystallization involved in the formation of each structure tends to permit only that experience and expression which is fundamentally consonant with the specific template involved. Thus, while there is a certain amount of growth within these subsidiary self systems, it is minimal. This limitation on the growth and change of the Libidinal and Antilibidinal Selves is more potently enforced by the factor which is the second way in which they differ from the original self, and later from the Central Self; they were created in an act of repression and at all times continue under the pressure of this repression.



Structure as Pathology


Virtually all psychoanalytic theories have accepted a metaphor for psychic growth which has been borrowed from biology: growth is defined as movement through progressive levels of structural differentiation and complexity. This metaphor is manifest in Freud's notion that psychic growth (and health) involves the differentiation of an ego, structurally separate from the id, and later a superego, precipitated out from the ego. It also stands at the root of the generally accepted belief that the self-object differentiation implies structural differentiation within the psyche --and the unspoken underlying assumption that the process of self and object representation is a structural one.

In what is his most radical departure from the mainstream of psychoanalytic thought, Fairbairn maintained that, far from being the necessary condition for psychic growth, structural differentiation was a defensive and pathological process in human development.

Fairbairn discussed at great length the process by which the psyche of the infant, due to some intolerable inability to cope with the unsatisfying aspects of experience, internalized this experience in such a way ultimately as to eventuate in the establishment of certain endopsychic structures. The creation of such structures involves the splitting of the self and the repression of that part of the self which has been thus split off.

Repression is the key element in the creation of endopsychic structure, because it is the mechanism by which the self becomes split. Experience which is integrable into the self results simply in memory or in the gradual alteration of the nature of the self as a whole. It is only when such experience is unintegrable --when it is so intolerable as not to permit of consciousness (which after all, is that which is "knowable together", i.e. integrable)-- that it must be subjected to repression. When that which is thus in need of repressing is importantly a part of the self, which is to say, when it is relationally so intrinsic to the life of that self that it is part of the definition of that self, then the act of repression must be understood as a splitting of the self. Repression and splitting in this structural sense are merely different perspectives on the identical operation. A particular aspect of the self, defined by its particular affective and purposive relationship with a particularized object, and reflecting a fundamental aspect of self-definition within the psyche, too intrinsic and powerful to be abandoned and too intolerable and unacceptable to be integrated into the whole --this fully functional, albeit crystallized, subsystem of the self is what becomes an endopsychic structure by virtue of the act of its repression. If it were not repressed, it would continue to exist within the conscious, integrable matrix of the self and there would be no splitting of that self and consequently no formation of endopsychic structure.

Fairbairn came to this understanding in stages. At first, differing from what he viewed as Freud's mistaken notion that what was subject to repression was either intolerably unpleasant memories or intolerably guilty impulses, Fairbairn (1952, p.62) developed the idea that it was intolerably "bad" objects that were subject to repression. He later altered this view:


It becomes necessary to adopt the view that repression is exercised not only against internalized objects (which incidentally are only meaningful when regarded in the light of endopsychic structures) but also against ego-structures which seek relationships with these internal objects. This view implies that there must be a splitting of the ego to account for repression (1952, p. 168).


Although he repeatedly referred separately to the repression of objects and the splitting of the self, it is clear from the above citation that he understood the two to be inextricably bound together in a manner that clearly justifies the use of the notion employed in this paper that it is the entire subsystem of the self (including both the object and what he termed the "ego" --or self-- element) that is repressed in the very act which creates its existence as endopsychic structure.

Thus it was that Fairbairn arrived at the notion that existence as a structure within the self means existence as a split-off subsystem of the self, created and maintained by repression, and owing its existence to the self's inability to deal with some important aspect of its experience which it found to be intolerable. He termed the process of establishing such structures "schizoid" because the splitting and repression by which it is constituted invariably diminish the self's capacity for growth and expression, and are, therefore, pathological.

The Libidinal and Antilibidinal Selves, by their very existence, limit the range and depth of the conscious functioning open to the Central Self. Both of these endopsychic structures press continuously for the recreation of experience of the sort which occasioned their creation, which experience always has two determining characteristics: it is equally experienced as intolerably "bad" (which, in Fairbairn's terminology, means unsatisfying), and it is equally experienced as being needed by the self absolutely for survival.

It is in this way that Fairbairn accounted for the clinically ubiquitous phenomenon of the repetition compulsion. There exists, at the very structural foundation of these subsidiary selves, an attachment to some negative aspect of experience which is felt as vital to the definition of the self (at least in the specific particularization thereof involved in each subsystem). The raison d'etre of these endopsychic structures is to continue living out these "bad" relationships. Much as the original self sought to express psychic existence of the whole person, such a subsystem seeks at all times to express itself and have experience in accordance with the template based on the formative intolerable experience which defines its existence. Thus the existence of such an endopsychic structure leads to the seeking of relationships that will be consonant with the specific neurotic paradigms of early experience, to the distortion of current relationships so that they can be experienced in accordance with such paradigms, and to the patterning of activity in the world so as to be expressive of such a relationship --and, in so doing, restricting the freer, more situationally appropriate expression of the self and experience of the world. It is important to note that this theory is not only more parsimonious than Freud's appeals to explanations based on mastery, masochism, and, finally, a death instinct, but that it also provides a direct explanation for the clinically observed sense of loss that is involved when patients, as the result of a successful psychoanalytic process, begin to relinquish their tenacious adherence to such patterns. The loss is twofold: most obviously, it involves the loss of the object component, which is felt as having made possible the particular internal relationship; and, perhaps more importantly, albeit less obviously, it involves a sense of loss of self, in so far as part of the self had been defined in the crystallization around the particular paradigm.

The fact that the Libidinal and Antilibidinal Selves always exist under repression further contributes to their pathological nature. Although in Fairbairn's view these structures are at least minimally able to grow and evolve through progressive accretion and overlay of later experience (in so far as the experience is fundamentally consonant with the defining paradigm), the isolating effort of the repression results in an inertia that is not readily overcome. Central to the nature of this repression --and the resistance it subsequently offers to growth and change-- is the attachment which has just been described. The self chooses to encapsulate and crystallize these aspects of itself and of its relationships rather than to be at risk for their loss. This maintenance of the internal world as a closed system is what Fairbairn (1958) ultimately described as "the greatest of all sources of resistance" (p. 380). Furthermore,


A real relationship with an external object is a relationship in an open system; but, in so far as the inner world assumes the form of a closed system, a relationship with an external object is only possible in terms of transference, viz., on condition that the external object is treated as an object within the closed system of inner reality (p. 381).


The splits which create endopsychic structures are, of course, variable in their extent and depth, depending on the nature of the relationships out of which they developed (which involve the specific strengths and weaknesses --constitutional and developmental-- of the child, as well as those of the parent, and of the vicissitudes of their interactions). The more profound the splits, the more extensive and the more deeply repressed the subsidiary selves they engender, the greater will be the pathological effect on the Central Self. Just as this Central Self is what remains after the splitting off of the Libidinal and Antilibidinal Selves, so too will the Central Self's ongoing experience and expression be diminished by the tendency of the subsidiary selves to limit and to transform subsequent experience and expression according to the closed systems of their defining paradigms. The more extensive the portion of the self which has been repressed, the less that will be available for open, ongoing interaction with the world.

Not only the quantity of the Central Self's experience and expression is diminished by the extent of the subsidiary selves, but also the quality of its relating to the world is similarly diminished. The more severe the tendency to experience the external world in accordance with the subsidiary selves, the more impoverished and idealized becomes the nature of the objects with which the Central Self relates. It is in this light that the objects of the Central Self become the Idealized Object, rather than the actual objects of external reality --which is to say that all of the complexity and imperfection must be abstracted out and subsumed into the experience of the subsidiary selves. This position is fully in harmony with the clinical observation that all idealizations invariably are based on the denial of some experienced imperfection, inadequacy, or "badness'.

The upshot of Fairbairn's theory is that healthy development is not dependent upon the establishment of endopsychic structures, but rather that such internal structural differentiation is a clearly pathological, albeit unavoidable, schizoid phenomenon which, to varying extents, diminishes the functioning of all human beings. As Fairbairn (1952) concluded,


Psychology may be said to resolve itself into a study of the relationships of the individual to his objects, whilst, in similar terms, psychopathology may be said to resolve itself more specifically into a study of the relationships of the ego to its internalized objects (p. 60).


On the other hand,


The chief aim of psychoanalytical treatment is to promote a maximum "synthesis" of the structures into which the original ego has been split (Fairbairn, 1958, p. 380).


Non-Structuring Internalization


Perhaps the most confused issue in Fairbairn's writings is the question of internalization. This confusion results from the fact that he used that concept of internalization in two distinctly different ways, while never acknowledging the difference existed.

The first sense of internalization is the one which Fairbairn clearly delineated in his theory and which has been discussed in detail in the preceding two sections of this paper. It is that form of internalization which eventuates in the formation of repressed endopsychic structures. For the purpose of clarifying the distinction which Fairbairn did not make explicit, this process will here be called structuring internalization.

As noted above, it is only intolerably "bad" experience that gives rise to structuring internalization. It is to just such structuring internalization that Fairbairn is referring in his major theoretical disagreement with Melanie Klein: whereas she had posited the internalization of both good and bad objects. Fairbairn (1952, Chapters 3, 4, and 7) repeatedly disagreed, insisting that it was only bad objects that were internalized. "It is difficult to find any adequate motive for the internalization of objects which are satisfying and "good" (Fairbairn , 1952, p. 93). Fairbairn's assertion here is that good objects are never structurally internalized, which follows directly from the fact that there would be no explanation for the repression (which is the essential ingredient of the formation of endopsychic structure) were it not for the intolerable "badness" of the experience with an object.

In apparent contradiction to this strongly propounded position, Fairbairn elsewhere (1952) writes of the internalization of "good" objects. He made it clear, however, that the internalized "good" object is the Idealized Object of the Central Self, which is a system in which none of the components is under structural repression. The apparent contradiction thus is easily resolved by the recognition that "good" objects, while they are internalized are never subjected to structure generating repression. This process, in which there occurs no repression, and therefore no self-splitting and no formation of endopsychic structure, will here be termed non-structuring internalization. Thus, it can be true that only "bad" objects are involved in structuring internalization, while it also can be true that "good" objects are internalized, but only in the non-structuring sense.

It is obvious that a human being needs to be able to internalize aspects of his experience in the world in order to grow and thrive. There must be learning that takes place as the result of both positive and negative interactions, and this learning must be integrated into the self in some meaningful way. While Fairbairn did not explicitly write about the nature of growth process, implicitly it is contained in the notion of non-structuring internalization. To understand Fairbairn's position on the nature of the process of non-structuring internalization, it is necessary to extrapolate from certain other of his previously discussed positions.

The most central principle, deriving from the definition of non-structuring internalization, is that such a process cannot lead to repression. Clearly, there is no need for the self to repress segments of its experience which are "good", or even which are "bad" in a tolerable way. Rather, such experience must be integrable into the self in a manner which remains conscious and openly available.

Secondly, it should be clear that such a process cannot lead to the formation of endopsychic structure. Rather, non-structuring internalization must be viewed as resulting in memory, or in the conscious organization of experience. The progressive development of a personal Weltansicht --viewed from any of what is an unlimited range of possible perspectives, be it that of Kant's categories of experience, Kohlberg's moral schema of development, or any other dimension of developmental progression-- implies learning, memory, organization, and synthesis, but not structural differentiation. Even the all important development of self-object differentiation does not, of necessity, imply the structural differentiation of the self, but rather the progressive recognition of the separateness of that self from the external world with which it interacts, and a progressive organization of the self's awareness of its own nature and potential. In addition, it must be remembered that, for Fairbairn, any fragmentation of the self cannot be viewed as a developmental arrest, but rather must be seen as some pathological miscarriage of development.

A further extrapolation can be made from another disagreement between Fairbairn and Klein. Fairbairn (1952) wrote,


As it seems to me, Melanie Klein has never satisfactorily explained how phantasies of incorporating objects orally can give rise to the establishment of internal objects as endopsychic structures --and, unless they are such structures, they cannot be properly spoken of as internal objects at all, since otherwise they will remain mere figments of phantasy (p. 154).


It is clear from this position that non-structuring internalization does not result in the establishment of any "entity" within the self, but rather results in an alteration of the integration of the self, or in the production of a thought, memory or fantasy within the self.

Kernberg (1976) presented a schema for the nature of internalization which is relevant to the present discussion. He wrote:


All processes of internalization of object relations refer to the internalization of units of affective state, object-representation, and self-representation. Following Erikson...I considered introjection, identification, and ego identity as a progressive sequence of such internalization processes. In the case of introjection, object- and self-representations are not yet fully differentiated from each other, and their affect is primitive, intense and diffuse. In the case of identification, not only is there a well-established separation between self- and object-representations, but there is an internalization of a role aspect of the relationship, that is, of a socially recognized function that is being actualized in the self-object interaction. The affective state is less intense, less diffuse, and...the spectrum of affect dispositions is broadened and deepened...Ego identity may be thought of as the supraordinate integration of identifications into a dynamic, unified structure (pp. 75 f.).


Although in Fairbairn's theory the notion of structure is radically different and a relational/structure model is employed rather than a drive/structure model, what is being described phenomenologically in both theories is closely related. There is a high degree of correspondence between Fairbairn's non-structuring internalization and Kernberg's concept of ego identity. Both theories recognize that there is a continuity of self experience and expression which is involved in such internalization which results in progressively higher levels of synthesis and integration. The opposite is true with respect to structuring internalization, which like Kernberg's introjection, refers to a level of functioning in which discontinuity and unintegrability result in a pathological form of internalization involving the splitting of the self and the radical formation of structure. Kernberg wrote of this process of introjection and the structures resulting from it that,


The persistence of "nonmetabolized" early introjections is the outcome of a pathological fixation of severely disturbed, early object relations, a fixation which is intimately related to the pathological development of splitting(1976, p. 34).


Kernberg's intermediate mode of internalization, the important issue of identification, is less obviously but just as certainly related to Fairbairn's non-structuring internalization. Kernberg described normal identification as follows:


(1) a partial modification of the total self-concept under the influence of a new self-representation, (2) some degree of integration of both self- and object- representations into autonomous ego functioning in the form of neutralized character traits, and (3) some degree of reorganization of the individual's behavior patterns under the influence of the newly introduced identificatory structure (1976, p. 78).


Once again it is crucial to note the emphasis on continuity and integration within the larger unity of the self, as opposed to any sense of structural isolation within that whole. Even in what Kernberg termed pathological identification, it is clear that the correspondence is to non-structuring internalization, although in this case the process takes place largely in relation to either the Libidinal or Antilibidinal Self rather than to the Central Self. This fact accounts for the rigidity and crystallization Kernberg observed to be characteristic of such internalizations.


The final outcome of pathological identification processes is character pathology. The more rigid and neurotic the character traits are, the more they reveal that a past pathogenic internalized object relation (representing a particular conflict) has become "frozen" into a character pattern (1976, p. 79).


While such identifications take place under the influence of pathological endopsychic structures and can slowly alter the nature of these structures, they do not eventuate in any further formation of such structures.

Kernberg (1976, 1980) was one of the first important theorists who explored and acknowledged the importance of Fairbairn's theories, and it is clear that he integrated into his theory many valuable aspects of Fairbairn's thought. Most centrally, Kernberg accepted the notion that internalizations, on all levels, have the basic form which Fairbairn suggested --an element of self, an element of object, and the affective, purposive relationship between them. It is also clear that Kernberg agrees that higher forms of internalization involve less disjunction in the self and more integration and continuity. It remains as a fundamental difference, however, that Kernberg integrates these insights into a drive/structure model, whereas Fairbairn was intentionally departing from such a model. Moreover, Kernberg, as virtually every other psychoanalytic theorist, maintains that the progressively higher levels of internalization involve increasing levels of internal structure. In contradistinction, Fairbairn demonstrated how it is not necessary to view the higher levels of internalization as creating structure at all. Rather, he showed that there was a conceptual advantage to differentiating structuring internalization, which is invariably pathological, from non-structuring internalization, which is defined by its continuity with, and potential for, integration into the self as a whole. While Kernberg obviously agrees with Fairbairn's observations concerning the phenomenological differences involved in these different levels of internalization, he does not adopt Fairbairn's conclusions about the nature of structure itself. Thus, despite the similarities, there are profound differences between them when it comes to crucial issues like the internalization of good experience and the metapsychological understanding of the self in which these questions occur.

The vicissitudes of these forms of internalization and their interrelationships are at the heart of Fairbairn's developmental notion of the movement from infantile to mature dependence, the central issue in which being the move away from primary identification (which, it is interesting to note, is the same issue of self-object differentiation which is central to Kernberg's hierarchy of forms of internalization).


The Growth of the Self


Fairbairn chose to discuss the development of the self in terms of levels of dependency. In so doing, he was emphasizing his contention that all meaningful human activity --from its most primitive to its very highest expression-- is at all times involved with relationship, be it with actual people in the external world or with the memory or fantasy of people in the inner world; and that the primary and ultimate goal of this activity, even in the neonate, is self-expression in relationship.

Views of healthy, adult development almost invariably include a positive notion of interdependence with significant others, and particularly the intense closeness and inter-relatedness with loved ones. In such love relationships, it is clearly acknowledged that it is a virtue to be the sort of person who can both 'be depended on' and be able to 'depend on' one's partner. Fairbairn, in labelling the highest level of development mature dependence, was choosing to emphasize the importance of human inter-relatedness and interdependence.

Dependency, in its pejorative sense, was associated by Fairbairn with the concept of infantile dependence. In doing so he was assigning the pathology not to the dependency itself, but rather to its infantile character.

Central to Fairbairn's notion of infantile dependence, and almost synonymous with it, (1952, p.42) is his concept of primary identification. In primary identification, the infant relates to an other whom he does not experience as separate or different from himself. It is clear that what is taking place does represent a form of relating --complete with a sense of intentionality and expression of the subject involved. Nevertheless, it is equally apparent that the subject is not aware in any differentiated way of the other person as being separate and apart from him. Fairbairn's contention was that the reality of both sides of this situation needs to be accepted: there is a relationship occurring, and self-object differentiation is not present (to a greater or lesser extent).

Although Fairbairn was completely insistent that the infant was object related from birth, he acknowledged that the infantile dependent relatedness of the earliest stages had specifically primitive characteristics: 1) it is unconditional; 2) the quality of need is absolute --if the infant's needs are not met, it will die; 3) the infant is not aware of any sense of option or choice of object --there is no experience of alternative, and the failure of the relationship to meet needs is tantamount to death.

The process of psychological maturation, in Fairbairn's scheme of the movement from infantile to mature dependence, consists of the gradual "abandonment of relationships based on primary identification in favor of relationships with differentiated objects."(1952, p.42) The key element in this change is the progressive differentiation of the object from the self: "The more mature a relationship is, the less it is characterized by primary identification." (1952, p.34 n.)

Fairbairn was clear that this process is a continuous one, ranging through various levels of self-object differentiation. At its most infantile level, there is no sense of separation between self and other --and thus there can be no awareness of any concept of self or other. As the infant has experience in the world, it gradually begins to organize and awareness of self and a concomitant awareness of other.

Although Fairbairn did not speak to the point, his system has obvious implications for the understanding of the highest levels of self-other differentiation. This process does not cease with the establishment of the notion that there is a discontinuity between one's self and others (physically as well as psychologically), but rather involves progressive levels of organization of the meaning of this differentiation and of the nature of the objects being differentiated. Ultimately, it is possible to utilize this schema to explore differences in the most mature levels of emotional development. For example, it is possible to see even moral development as an issue of learning to understand others as differentiated to the point of being ends in themselves (cf. Kant, 1785) and having an equally valid claim on shaping and defining their own experience and meaning.

The state of mature dependence implies a recognition of the separateness of individuals, even while they are involved in the most intimate and interdependent of relationships. Separateness thus in no way implies isolation, or even disconnection. Rather, separateness hinges on the recognition of the existence of the selfhood of the other, ultimately conceived of in a form that is not subsumable by one's own selfhood. It is the recognition that the other is a center of experience and intentionality, feeling and will, thought and purposiveness. In other words, it involves the acknowledgment of the unique individuality of the other in a way that is in no way diminished by the existence of the relationship between the self and that other. It should be clear that perhaps the most salient practical touchstone for this sort of separateness will be the recognition and acceptance of individual responsibility.

Between the stages of infantile dependence and mature dependence, Fairbairn envisioned a stage which he termed quasi-independence. It should be clear, from what has been noted above, that this term is designed, in part, as a negative comment on the traditional emphasis placed on independence in most developmental theories. Nevertheless, it also is designed to convey a sense of the struggle at this level to move out of the state of infantile dependence in a way that is still very much attached to that very state. (For this reason, Fairbairn also referred to this stage as "transitional.") The state of quasi-independence is ultimately doomed to failure, because it consists of an attempt to change an earlier state without relinquishing the essential tenets of that state. It is that state out of which neuroses, as classically conceived, arise; and thus it is fitting that it be predicated on a situation of conflict between the preservation and abandonment, the expression and inhibition, of an infantile state of affairs.

It is essential to realize that Fairbairn's entire conception of how the self grows is in no way predicated upon the process of structural differentiation. The self's growing awareness of individuation and separateness is based on integrated development of the whole of that self. As the individual achieves progressively higher levels of organization and interpretation of his experience, he functions with an increasing level of self-object differentiation, and moves from operation in an infantile dependent mode towards a progressively more adult mode of mature dependence. This movement represents the growth of the self as a whole, proceeding through the process of non-structuring internalization, and not through the establishment of divisions or structures within the self. This latter process of structuring internalization has been shown to be essential to the development of psychopathology, but not to the healthy development of the self.




It has been shown that Fairbairn's structural model of the psyche is in no way the same as Freud's drive/structure model. Fairbairn's theory is achetypally a relational/structure model. Based on the assumption that the fundamental human motivation is for self-expression in relationship, it is a theory that takes as the fundamental structural building block the constellation of self, other, and relationship between. Substructures of the self naturally are seen as conforming to this same pattern. Furthermore, the theory is predicated on a radically different notion of the nature of structure itself.

Fairbairn's insistence that structure implies pathology and that wholeness and integration imply health is unique among psychoanalytic theories. It presupposes a notion of the self that is in itself a radical departure. For Fairbairn, the self is not reducible to a self-concept, or a self-representation, or a system of reflected appraisals. It is a self-generating center of origin which, while it is shaped and changed in relation to its objects (or, more accurately, its "others") and does in part define itself in terms of those relationships, has an expressive, experiencing existence separate from, and prior to, these relationships.

There is room in Fairbairn's theory to accommodate identifications and representations of self and objects, as there is room to accommodate systems of reflected appraisals. These can be viewed as aspects of the self's experience of itself and its world. The major innovative insight of Fairbairn was that these phenomena do not in any way require structural differentiation of the self. Rather, he made a clear and crucially useful distinction between these non-structuring internalizations, which are far more related to memory and the progressive organization of experience (and which do involve representations of self and object), and the internalizations which involve actual segments of the self (not representations thereof) and that therefore create real structures within the self-crystallized subsystems which function within the self with a dissociated life of their own.






Fairbairn, W. R. D. (1952). Psychological Studies of the Personality. London: Routledge & Kegan Paul.


Fairbairn, W. R. D. (1954). Observations on the nature of hysterical states. Brit. J. Med. Psychol., 27:105-125.


Fairbairn, W. R. D. (1958). On the nature and aims of psychoanalytic treatment. Int. J. Psychoanal., 39:374-385.


Freud, S. (1900). The interpretation of dreams. S.E., 4 & 5.


Freud, S. (1923). The ego and the id. S.E., 19.


Freud, S. (1933). New introductory lectures on psychoanalysis. S.E., 22.


Greenberg, J. R. & Mitchell, S. A. (1983). Object Relations in Psychoanalytic Theory. Cambridge, Massachusets: Harvard University Press.


Kant, I. (1785). Foundations of the Metaphysics of Morals. Eng. trans., Beck, L. (1959). New York: Liberal Arts Press.


Kernberg, O. F. (1976). Object Relations Theory and Clinical Psychoanalysis. New York: Jason Aronson.


Kernberg, O. F. (1980). Internal World and External Reality. New York: Jason Aronson.


Mitchell, S. A. (1988). Relational Concepts in Psychoanalysis: An Integration. Cambridge, Massachusetts: Harvard University Press.


Rubens, R. L. (1984). The meaning of structure in Fairbairn. Int. Rev. Psychoanal., 11:429-440.


Rubens, R. L. (1990). Psychoanalysis and the tragic sense of life. New Issues in Psychology.


Rubens, R. L. (currently in preparation). Fairbairn's developmental theory.


Strachey. J. (1961). Editor's introduction to "The ego and the id" by S. Freud. S.E. 19:3-11.


Winnicott, D. W. & Khan, M. M. R. (1953). Book review of Psychological Studies of the Personality by W. R. D. Fairbairn. Int. J. Psychoanal., 34:329-333.


Winnicott, D. W. (1965). The Maturational Process and the Facilitating Environment. New York: Int. Univ. Press.


Return to Richard L. Rubens PUBLICATIONS Page.