The Mind as Conflict and Compromise Formation

At the time my medical education began the structural theory was in its infancy. Fewer than ten years had passed since the publication of The Ego and the Id and the application to clinical work of the ideas contained in it and in Inhibitions, Symptoms and Anxiety had barely begun to be felt even in Vienna and Berlin, let alone in the United States. The contributions of Alexander, Fenichel, A. Freud, Hartmann, Kris, and Waelder to ego analysis were still in the future. For most analysts it was to be years before defense was thought of in any terms other than repression. Despite what Freud had so recently written, neurotic anxiety was still thought to be repressed libido that had gone bad rather than a signal for defense. The topographic theory held sway in the minds of most of the leaders in the field of analysis and it continued to do so for many years.

The superiority of the structural theory to its predecessor was easily apparent, however, to those like myself who were new to the field and who had the good fortune to be able to learn from the eminent teachers who had fled to our shores from the political convulsions that engulfed Europe in the thirties and forties. We grew up with the structural theory. We learned early the relation between defense and signal anxiety and that defenses are to be analyzed. I said, early, but perhaps that is an exaggeration. It would probably be more accurate to say that we learned about signal anxiety and defense analysis after not too long a time. Certainly, we learned about such things before we were irrevocably committed to the topographic theory as our predecessors had been.

Like my contemporaries, then, I grew up with the structural theory. I viewed its introduction for what it was: a great step forward in our understanding of how the mind develops and functions. I studied it, I applied it, I wrote about it, and I taught it. For very many years I believed it to be true, scientifically speaking; that is, I believed it to be the best theory of how the mind works that could be based on available psychoanalytic data.

Recently I have had some doubts. I have no doubt about the importance in mental life of conflict resulting from drive derivatives that are associated with unpleasure. I have no doubt about the importance of the ideational content of the unpleasure, that is, of the calamities of childhood that Freud outlined in Inhibitions, Symptoms and Anxiety (1926). I have no doubt about the importance of analyzing defenses and self-punitive trends. All of these are features of clinical, analytic work to which attention was drawn by the introduction of the structural theory. All of them are essential parts of the technique of analysis. My doubts don't have to do with the elements of conflict as they appear in clinical work or with the nature of their interaction. My doubts concern the question whether the facts as we know them today support the theory that there is a structure or agency of the mind, the id, that consists of drive derivatives; that it is separate from another agency of the mind, the ego, which has other functions, including defense; and that both are separate from still another structure, the superego.

The principal purpose of this paper is to acquaint its readers with the reasons for my doubts with the hope that doing so will raise questions in their minds as well. I shall consider that my purpose has been successfully accomplished if those who read are persuaded to give serious consideration to the possibility--please note my emphasis on the word possibility--that the time has come to relinquish the idea that the mind is best understood as consisting of the separate structures id, ego, and superego.

The idea that the mind is composed of separate structures, agencies, or systems occupied a major position in Freud's theories from first to last. Its first version is to be found in chapter 7 of The Interpretation of Dreams (1900). There Freud postulated two systems which he called the system Ucs. and the system Pcs.-Cs. The Ucs., he believed, functioned according to the primary process, was made up of wishes dating from childhood, and was inaccessible to consciousness. The Pcs.-Cs. by contrast functioned according to the secondary process and was normally in control of planned behavior, emotional expression, and consciousness. Two decades later, in The Ego and the Id (1923), Freud revised these aspects of his theory of the mind and postulated instead three structures, familiar to us as the ego, the superego, and the id. It is worth noting that both sets of structures, or systems, corresponded to Freud's understanding at the time of the nature of pathogenic conflict. In 1900, and for many years thereafter, Freud understood pathogenic conflict to be between what is accessible to consciousness and what is inaccessible to consciousness. By 1923 he had come to realize that this formulation is only a rough approximation, since, for one thing, defenses are often inaccessible to consciousness and, for another, what is defended against may be, at least in part, accessible to consciousness.

In order to keep his theory in conformity with clinical data, therefore, Freud dropped accessibility to consciousness as the touchstone by which the instinctual aspect of a conflict is to be distinguished from the defensive aspect and substituted other criteria.

Instead of the Ucs. he proposed an agency or structure to be called the id. As its name indicates, it is the demonic part of the mind, the strange, unfamiliar part that one can hardly, if at all, recognize as part of oneself. It functions in such a way as to drive one to seek instinctual gratification, including gratification of wishes of which one has no conscious knowledge and/or which one strenuously rejects or denies. Like the Ucs., then, the id is made up of drive derivatives.

The difference between id and Ucs. is largely termino-logic, with the exception, noted previously, that the change in name takes account of the fact that, in situations of conflict, aspects of drive derivatives that are strenuously defended against may, at least at times, be accessible to consciousness.

In place of the Pcs.-Cs. Freud proposed two structures or agencies that he named the ego and the superego. As its name indicates, the ego includes what one as a person is familiar with about oneself, what one consciously recognizes as oneself. In addition, it includes much that is inaccessible to consciousness, in particular unconscious defenses or, as Freud later (1926) called them, defense mechanisms, that serve the purpose of warding off unwanted drive derivatives. Thus in situations of mental conflict the id (childhood sexual wishes) is to be placed on one side of the conflict and the ego (defense mechanisms) is to be placed on the other.

The ego that Freud described in 1923 and later is not characterized only by its functioning in situations of conflict over repressed, childhood sexual wishes, however. He also defined, or described, the ego as the integrated or coherent part of the personality, as having or embodying a concern for logic and consistency, and particularly as being tied to the outer world, that is, to the environment. Whereas the id is presumed to have no concern with the environment, the ego has a special allegiance to it in consequence of its various functions.

As Freud put it, "the ego seeks to bring the influence of the external world to bear upon the id.... For the ego, perception plays the part which in the id falls to the drives" (Freud, 1923, p. 25).

The third of the three agencies of the structural theory is the superego, the agency that has to do with morality. Freud's reason for distinguishing it as a special agency had again to do with what he had observed about mental conflict. There are conflicts in which moral strictures are themselves defended against; conflicts, that is, between ego and superego, conflicts that lead to an unconscious sense of guilt or need for punishment.

What I consider to be of particular importance in evaluating the validity of the structural theory is what has just been said about the id and the ego. The structural theory divides the mind into agencies ( = structures) that are distinguished from one another not just on the basis of being opposed to one another in situations of mental conflict, but additionally, and importantly, on the basis of having different relations to external reality and to the drives. According to the structural theory the id is concerned only with gratifying drive derivatives ( = wishes), without regard either to the limitations of external reality or to possible consequences. The ego, by contrast, is conceived of as being tied to external reality and bound by the need to take into account the limitations that external reality imposes and the possible consequences that may ensue from actions that involve the environment. The ego is conceived of as being logical, consistent, and coherent, all in contrast to the id, which is conceived of as its opposite in these respects.

In retrospect it seems that certain analytic findings since 1926 might have been expected to raise questions about the validity of separating ego from id on other grounds than the opposition between drive derivative and defense in situations of mental conflict. For example, A. Freud (1936) wrote that the structures of the mind are distinguishable from one another only in situations of conflict between them. In the absence of conflict, in her opinion, they are indistinguishable from one another. Clearly such a view is incompatible with the theory that says that ego and id, at least, are to be separated from one another, and that they have each a different and readily distinguishable relation to external reality.

As early as 1935 Kris identified certain aspects of mental functioning that he called regression in the service of the ego. His observations were elaborated by later authors (Arlow and Brenner, 1964; Brenner, 1968) who showed that aspects of mental functioning attributed to the ego by the structural theory indicate that what the theory designates as the ego is by no means as consistent, as integrated, as mature, and as immune from primary process functioning as the ego is supposed to be.

The findings responsible for my own doubts about the validity of the structural theory, however, have to do with the role of conflict in normal mental functioning. Every aspect of conscious mental life that is emotionally significant, whether normal or pathological, has been shown to be a compromise formation. That aspect of mental functioning that the structural theory calls the ego, it seems, is dynamically the same as a neurotic symptom. Every normal or pathological thought, fantasy, wish, action, mood, and character trait results from the interaction of the libidinal and aggressive wishes of childhood; the highly unpleasurable ideas of object loss, loss of love, and castration that are associated with the wishes in question; and the defenses and moral strictures that serve the function of eliminating the unpleasure or, if they cannot eliminate it, as least of reducing it as much as possible.

The wishes attributed by the structural theory to a special structure, the id, the unpleasure and defenses attributed to another special structure, the ego, and the moral strictures attributed to still another structure, the superego, are in constant and ubiquitous interaction ( = conflict). There is no special rational part of the mind that takes realistic account of external reality without being motivated by the libidinal and aggressive wishes of childhood and the unpleasure associated with those wishes. There is no part that is mature, integrated, and free of conflict, as the structural theory assumes is the case.

The structural theory maintains that, beginning quite early in childhood, normal mental functioning depends on the existence of a strong, well-functioning, special mental structure or agency, the ego.

The ego must be strong and well in control of the id ( = childhood drive derivatives) if the mind is to function normally. According to the structural theory it is only when the ego is weak relative to the id that conflict ensues and a greater or less degree of disturbance of mental functioning results. Such a theory is obviously at odds with the fact that every thought, even every perception (Arlow, 1969) gratifies one or more drive derivatives, in however disguised a way.

Granted that the structural theory does not take adequate account of the demonstrable fact that mental conflict is as characteristic of normal mental functioning as it is of abnormal mental functioning, what, if any, disadvantages ensue in one's clinical work when one is guided by the idea, basic to the structural theory, that the mind is made up of separate agencies ( = id, ego, and superego)? Are the disadvantages great enough to justify changing the theory? After all, this idea has been a basic part of psychoanalytic theory since the beginning. First in the form of the topographic theory, with its systems Ucs. and Pcs.-Cs., and later, with necessary and useful changes, in the form of the structural theory. In the latter form it has remained unchallenged since its introduction some seventy years ago. How one to justify changing it now in any major way?

I believe that the answer is this: If, in clinical work, one equates ego with defense, the structural theory seems to fit the facts very well.

Day-to-day work with patients is mostly concerned with analyzing and interpreting the components of psychic conflict. The aspect of what, in the structural theory, is called ego functioning, on which attention is focused in analytic practice, is that of defense and in much of one's clinical work one can safely neglect the fact that the concept, ego, embraces far more than just defense. Not always, however. For example, how is one to evaluate progress in analysis? According to the structural theory, progress means diminution and, eventually, disappearance of conflict. To use familiar terminology, the goal of treatment is the resolution of conflict. The ego is supposed to become stronger as well as more mature, conflict is supposed to disappear, and the ego to deal with the id wishes that were involved in conflict by judgmental repudiation rather than by using one or another defense mechanism. Conversely, if conflict persists, if it does not substantially diminish or resolve ( = disappear), one must conclude, according to the structural theory, that no substantial analytic progress has occurred. Since the fact is, however, that conflict over what were originally pathogenic drive derivatives is still obvious and still active in the mind of every patient who by all other criteria has made substantial analytic progress, the structural theory is clearly not adequate as a guide in judging progress, something that is of considerable practical importance in clinical work. Put in other words, the structural theory gives one an incorrect idea of what it is that psychoanalysis can achieve. It is at odds with the clinically observable fact that the conflicts that give rise to symptoms persist after psychoanalysis has cured or relieved those symptoms.

The structural theory asserts that psychoanalysis puts an end to ( = resolves) conflicts, whereas, in fact, what a successful psychoanalysis actually achieves is an alteration of conflict in the direction of normality, an alteration that results in a normal compromise formation in place of the pathological one that was formerly present (Brenner, 1976, 1982).

An additional drawback to the structural theory in clinical practice is apparent with respect to the moral functioning of the mind. According to the structural theory this aspect of mental functioning is attributed to a separate structure ( = agency), the superego. This has the great virtue of calling special attention to what deserves special recognition, both because of its importance in clinical work and its importance in mental functioning in general. The role of morality and of conflicts involving moral demands and strictures is unquestionably a major topic in psychopathology. Yet attributing morality to a special agency has led to distorted views of both the origins and the functioning of morality (Brenner, 1982) and has in consequence greatly hampered the assessment and the analysis of many conflicts to which the moral functioning of the mind gives rise. As Hoffman (1992) pointed out, very few analysts have discussed the technical aspects of superego analysis. It is at present largely an unexplored area of technique, just as defense analysis was before the publication of Inhibitions, Symptoms and Anxiety (Freud, 1926).

It seems, then, first, that the assumption that the mind is composed of separate agencies is at odds with current knowledge about the ubiquity of mental conflict and, second, that this assumption entails significant practical disadvantages. This being the case, two questions arise: "What changes are necessary to bring the present theory into consonance with what we currently know about mental functioning, especially with the ubiquity of conflict?" and, "Is there substantial practical advantage to making such changes?"

The basic premises of the revisions I propose are simple. Mental functioning is regulated by the pleasure-unpleasure principle. People seek pleasure and avoid unpleasure. Their efforts to achieve both these aims are what result in mental conflict and in the compromise formations that are what one experiences as one's own mental life and what one observes as the mental lives of others.

One can best understand conflict as consisting of four interacting components: drive derivatives, unpleasure associated with the drive derivatives in question, defense, and moral demands and prohibitions. What is to be said about each?

What I have to say about the drive derivatives is this. One must first of all be clear about the distinction between drive derivative and drive. A drive derivative is a wish for pleasurable libidinal and/or aggressive satisfaction. Every drive derivative is unique for the individual whose wish it is. Every drive derivative has to do with particular persons in that individual's life, with particular actions, particular bodily sensations, and so forth. Each person's drive derivatives are best observed or, better, inferred with the help of the psychoanalytic method applied to that person. That is to say, the most reliable way of learning about anyone's libidinal and aggressive wishes is through psychoanalysis of the person in question. The concept of a drive is, by contrast, impersonal and general rather than personal and specific. A drive is what one can say about everyone's wishes for pleasurable gratification, regardless of individual endowment and life experience. The concept of drive is the generalization made from many observations of individual drive derivatives.

Current psychoanalytic theory ( = the structural theory) postulates two drives, libidinal and aggressive. From as early in life as it is possible to apply the psychoanalytic method both can be seen to be involved in each person's pleasure-seeking wishes. One never sees evidence of pleasure-seeking wishes in thought or behavior that are purely libidinal without any trace of aggression, nor does one see evidence of the reverse (i.e., of wishes for aggressive satisfaction without associated libidinal desires). Indeed, the fact that the two are always found together is of great importance in clinical practice. Object relations are always ambivalent, however strenuously a patient may insist that this is not the case.

As I have noted elsewhere (Brenner, 1982), Freud's view was that there is no pleasure associated with satisfaction of aggression. As he put it, aggression, which he believed to be the manifestation in mental life of a universal, protoplasmic death drive, operates silently in mental life. Only libidinal satisfaction is accompanied by pleasure, Freud thought. It was in line with this belief that Freud attributed different roles to the two drives in mental conflict. It was his view that conflict is occasioned by libidinal wishes only and that the role of aggression in conflict is limited to self-punitive and self-destructive tendencies.

At present the role of aggression is viewed quite differently. Aggression and libido are considered to be the same, both with respect to pleasure premium and with respect to their roles in conflict (Hartmann, Kris, and Loewenstein, 1949). The satisfaction of both libidinal and aggressive wishes is accompanied by pleasure, so far as one can judge from psychoanalytic data, and aggressive wishes can give rise to conflict no less than libidinal ones can.

These aspects of the theory of drives and drive derivatives are as important to the new theory I am proposing as they are and have been to the structural theory. The difference lies in the fact that in the structural theory the drives and those of their derivatives that are associated with a sufficient degree of unpleasure are considered to constitute a special agency of the mind, the id. As such they are believed to be barred from participation in the normal aspects of mental life, though they are of prime importance in those abnormal aspects of mental life that are called neurotic symptoms and neurotic character traits, as well as in those borderline mental activities known as dreams, as jokes, and as the slips and errors of daily life. It is on this point that the new theory diverges from the old. As noted earlier, current knowledge concerning the ubiquity of conflict in mental life speaks against dividing the mind into an id and an ego that are separate from one another, that function according to different modes, and that are opposed to one another in certain fundamental ways.

The nature and significance of the changes proposed by the new theory will be clearest if I shift at this point to a discussion of the ego.

What does the new theory propose to put in the place occupied by the ego in the structural theory?

Instead of positing "an ego," the new theory speaks simply of an individual, of a person, and of that person's mind. Not of "the self," in any such technical sense as many analysts use that word today, but simply of a person or individual in the ordinary, colloquial meaning of the words. When unpleasure arises in association with a drive derivative of childhood origin, a person's mind functions in such a way as to minimize the unpleasure while at the same time permitting as much gratification to the drive derivative as is compatible with not too much unpleasure. Whatever helps minimize unpleasure associated with a drive derivative is a defense. There are no special defense mechanisms; that is, there are no mental mechanisms or activities that serve the function of defense and of nothing else. Whatever thoughts or behavior serve to diminish unpleasure are defenses (Brenner, 1982).

In the conflict or conflicts that give rise to every compromise formation, defense must play a role. To put the matter a bit loosely, every compromise formation has its defensive aspect.

It is worth noting that not all defense is directed primarily against gratification of drive derivatives. Every defense serves to diminish unpleasure, but not every defense is anti-instinctual. In counterphobic reactions, for example, it is the unpleasurable affect of anxiety, not the drive derivative, that is denied (Fenichel, 1941, p. 64). To dramatize: a counterphobic person says, "I'm not afraid of (symbolic) castration. I enjoy running the risk of (symbolic) castration." Similarly, a euphoric person says, in effect, "I'm not unhappy. I'm as happy as a lark."

What about the unpleasure that is part of every conflict? Like all affects, the unpleasure has an ideational content. In the most general terms, the content is that one or more of the calamities of childhood either is a fact of life or impending. If the calamity is experienced as a fact of life, the unpleasure is to be classified as depressive affect ( = misery). If it is experienced as impending, it is to be classified as anxiety. Sometimes the two are so blended as to be indistinguishable. More often, so far as one can judge from analytic data, they are distinguishable even when they are intertwined (Brenner, 1982).

The calamities are the familiar ones listed by Freud (1926). They are object loss, loss of love, genital loss or injury ( = castration), and punishment. The last of these, punishment, can include any or all of the first three. Punishment is what a disapproving parent metes out and parental disapproval, in a child's mind, can involve any combination of object loss, loss of love, and castration. It is the effort, most urgent during the oedipal phase of development, to win or regain its parents' love that gives rise to the various compromise formations that go to make up what is called the superego in the structural theory. In the mind of a young child, what is right or wrong, morally speaking, is what its parents say (or are believed to be saying) is right or wrong. They are the supreme arbiters of morality at that early age and it is their approval that a child must seek and their disapproval it must avoid or neutralize.

In the theory I propose, therefore, what was a special category in the structural theory, namely, the superego, becomes instead one of the calamities of childhood, the calamity of parental disapproval. As such it blends with the calamity of loss of love. It is not identical with object loss, because death or absence of a parent is not identical with parental disapproval. In a child's mind, a parent may disapprove, but be very much present, just as, in a child's mind, a parent may be absent ( = dead and gone) and consequently impotent rather than fearfully powerful. The point is that when a child believes a parent disapproves ( = thinks the child is bad, naughty, doing wrong) the consequence is believed to be one or more of the other calamities, namely, object loss, loss of love, or castration.

In our society, at least, the idea of a parent judging one or more of a child's wishes to be bad ( = morally wrong) plays a very important role in every child's mental life and development. As Rothstein (1994) has shown, a very closely related idea, namely, that a parent judges one or more of a child's wishes to be, not morally wrong, but shameful or contemptible, can also play a very important role in mental life and development. Shame and guilt are closely related dynamically because both are consequences of parental disapproval at an early stage of development. There are a number of topics that deserve discussion in connection with this proposed revision of psychoanalytic theory. One has to do with what have until now been labeled as various aspects of ego functioning other than defense. How does one account for reality testing, to choose but one example, or for the need for logic and coherence that the structural theory attributes to the integrative ( = synthetic) function of the ego, to choose another?

As to the first, the theory I am proposing dispenses with the idea that one part of the mind, called the ego, has to do with one's relation to reality ( = environment), while the other, the id, does not. I agree with Biven (1986) that one perceives what one can perceive, that is, whatever one's sensory apparatus permits one to perceive, and that one deceives oneself about what one has perceived in accordance with the pleasure-unpleasure principle. To put the matter in other words, every perception is itself a compromise formation. As Arlow (1969) pointed out, every conclusion one draws from the data of perception is multiply determined.

As for the ego's integrative function, a regard for logic and coherence is not, in fact, an innate characteristic of the mind, as the structural theory assumes it to be. It is the result of a long and painful development over the course of millennia, a development that must be transmitted anew to each new member of society. A regard for logic is one of the phenomena of acculturation, not something that is inborn, and even now there are glaring exceptions among the most intelligent and sophisticated ( = the most acculturated) people. Like all similar cultural phenomena, a concern for logic and coherence is, for each individual, a compromise formation that conforms to the pleasure-unpleasure principle.

Another topic might be sadomasochism. While Freud thought that we cannot subsume masochistic functioning under the pleasure principle, today we think we can. We believe that unpleasure is sought partly as symbolic gratification, partly as atonement ( = to win love and to avoid disapproval), and partly to avoid worse unpleasure (Brenner, 1959). This idea has had a substantial effect on clinical practice. Patients are no longer considered unanalyzable simply because they are "moral masochists." There is a theoretical consequence as well. Once one accepts the explanation of masochism outlined above, there is no need to postulate a special repetition compulsion.

Still another topic for discussion in connection with the proposed change in theory is the difference between health and illness in mental life (i.e., the difference between what is normal and what is pathological). Such a discussion would include the related topic of how psychoanalysis cures. When Freud wrote, "Wo Es war soll Ich werden," he meant that the drive derivatives involved in pathologic conflict should be governed by the secondary process rather than by the primary process, not that the drive derivatives should disappear. It was his opinion that such a thing is impossible. He meant only that they should be governable by the ego rather than in conflict with it. What he called judgmental repudiation should prevail rather than defense.

At the same time a different idea was also contained in his theory, namely, that as long as primal repression remained intact, as long as there was no return of the repressed from repression, one's mind functioned normally. Once some intolerable drive derivative had burst forth from repression, however, and given rise to conflict and compromise formation, then it could be dealt with in one of two ways: reestablishment of repression or making it subject to the secondary process ( = judgmental repudiation), which is what psychoanalysis does, in Freud's opinion.

In this connection one should bear in mind that Freud chose the words primary and secondary as reminders that he considered the one to precede the other chronologically in the course of development.

The one is the more infantile and immature, the other, the more adult and mature. Thus it is fair to say that Freud tried to account for the differences between illness and health in mental life by positing different agencies, one of which is mature and the other immature.

Freud thought of childish mental functioning as pathological when it affects conscious thought and behavior in waking, adult life. His idea was that, although childhood drive derivatives never disappear, they don't give rise to mental illness if they are dealt with in a mature fashion, without conflict. If they are dealt with immaturely, however, if they are dealt with in accordance with primary process functioning, then they are pathogenic.

Unfortunately the facts are not that simple. While it is clear that Freud was right in believing that childhood drive derivatives persist throughout one's lifetime, the fact is that the conflicts to which they gave rise in childhood also persist throughout life. Childish ways of thinking ( = primary process thought) are as much a part of normal mental functioning as they are of pathological mental functioning.

Health and illness in mental life are not distinguishable on the basis of maturity versus immaturity of thought processes. The distinction is more complex and less clear than Freud and others would like to think (Brenner, 1982). If a compromise formation allows for an adequate amount of pleasurable gratification of drive derivatives, if it does not arouse too much unpleasure, if it does not entail too much inhibition of function as a result of defense, and if it does not involve too much by way of self-injury and suffering as punishment for moral transgression, it qualifies as normal. If not, a compromise formation deserves to be called pathological. No simpler or more precise distinction is possible at present between normality and pathology in mental life.

To summarize, the principal change that I believe is required to bring the psychoanalytic theory of the mind into consonance with current knowledge, is that the mind be no longer divided into agencies or structures. The change is not an easy one to make. It is difficult to give up such familiar concepts as those of psychic structure, and to relinquish id, ego, and superego as conceptual tools. They are concepts that have served generations of analysts well. I have tried to show that they no longer do so, and that newer concepts will serve better. As Freud himself said more than once, scientific theories are like scaffolding. When they are no longer useful, they must be replaced by ones that better fit the data that one is attempting to order and explain.

As long as conflict was viewed as essentially pathological or, at the least, potentially pathogenic, it was possible to conceive of the mind as consisting of two agencies of which one, the id, was constantly or periodically threatening to arouse mental conflict, while the other, the ego, had as its function the task of subduing and regulating drive derivatives in such a way as to remain conflict free. Should the ego become involved in pathogenic conflict, in this view, the task of analysis is to resolve the conflict(s) responsible for the symptoms, that is, to make the conflicts go away. Today we know that conflicts are never resolved in that sense, that they are as much a part of normal mental functioning as they are of pathological mental functioning. In the light of this newer knowledge I believe that the idea of separate mental agencies is untenable.

Moreover, the realization that conflict is both normal and ubiquitous has significant consequences for analytic technique and for the understanding of that aspect of mental functioning that, in the structural theory, is subsumed under the heading, the superego (Brenner, 1982).

Finally it should be emphasized that the theoretical revisions that I am proposing bear the same relation to the structural theory as did that theory to the earlier set of generalizations that we call the topographic theory. They are evolutionary in nature rather than revolutionary, to use terms that are widely employed today. They derive from the analytic situation and they have substantial consequences for analytic practice. They are changes that are required by data currently obtainable through the use of the psychoanalytic method. I hope I have been able to persuade the reader that the question of the validity of making those changes at this time warrants serious consideration.

Arlow, J. A. (1969), "Fantasy, memory, and reality testing." Psychoanalysis Quarterly, 38:28-51.

------ Brenner, C. (1964), Psychoanalytic Concepts and the Structural Theory. New York: International Universities Press.

Biven, L. (1986), "Compromise formation and cognition: A revision of the psychoanalytic theory of perception. In: Psychoanalysis: The Science of Mental Conflict, ed. A. D. Richards & M. S. Willick. Hillsdale, NJ: Analytic Press, pp. 129-151.

Brenner, C. (1959), "The masochistic character: Genesis and treatment." Journal of the American Psychoanalytic Association, 7:197-226.

------ (1968), "Archaic features of ego functioning." International Journal of Psychoanalysis, 49:426-430.

------ (1976), Psychoanalytic Technique and Psychic Conflict. New York: International Universities Press.

------ (1982), The Mind in Conflict. New York: International Universities Press.

Fenichel, O. (1941), Problems of Psychoanalytic Technique. Albany, NY: The Psychoanalytic Quarterly.

Freud, A. (1936), The Ego and the Mechanisms of Defence. New York: International Universities Press, 1966.

Freud, S. (1900), The Interpretation of Dreams. Standard Edition, 4 & 5. London: Hogarth Press, 1953.

------ (1923), The ego and the id. Standard Edition, 19:1-60. London: Hogarth Press, 1961.

------ (1926), Inhibitions, symptoms and anxiety. Standard Edition, 20:77-172. London: Hogarth Press, 1959.

Hartmann, H., Kris, E., & Loewenstein, R. (1949), "Notes on the theory of aggression." The Psychoanalytic Study of the Child, 3/4:9-36. New York: International Universities Press.

Hoffman, L. (1992), "On the clinical utility of the concept of depressive affect as signal affect." Journal of the American Psychoanalytic Association. 40:405-424.

Kris, E. (1935), "The psychology of caricature". In: Psychoanalytic Explorations in Art. Madison, CT: International Universities Press, 1952.

Rothstein, A. (1994), "Shame and the superego: Clinical and theoretical considerations". The Psychoanalytic Study of the Child, 49:in press. New Haven, CT: Yale University Press.

Dr. Brenner is a Training and Supervising Analyst at the New York Psychoanalytic Institute.

I am indebted to two of my colleagues, Dr. Arnold M. Richards and Dr. Yale Kramer, who separately called my attention to the theoretical implications of the recognition that conflict and compromise formation are ubiquitous in mental life. They have generously granted me the opportunity to develop more fully the consequences of what they were the first to recognize clearly.

To read some commentaries and discussions of this article, go to the abstract page, or the Table of Contents.