Dale Boesky
Dr. Boesky is Training and Supervising Analyst, Michigan Psychoanalytic Institute and Former Editor-in-Chief of the Psychoanalytic Quarterly.
One of Charles Brenner's many virtues has been his clear-eyed scrutiny of our ambiguous theoretic lexicon and his astringent foraging in our theoretic attic. He has repeatedly demonstrated that some of our terms which had seemed to be clear enough were actually ambiguous. His present paper expresses more clearly than ever his wish to strip away unnecessary and anachronistic accretions of meanings in our terms and theories. He has challenged us in the past and now does so again by stimulating us to join him in asking whether we still need certain conceptual tools. Are these tools and terms doing the job we think they are? Are we losing something by perpetuating their use? In this case he concludes that it is time to seriously consider abandoning the concepts of id, ego, and superego because these terms erroneously separate and disconnect the components of conflict in the human mind.
First I will remind readers of what Brenner does not doubt. He is still convinced of the value of the notion of conflict. Nor does he doubt the nature of the four components of conflict mediated by compromise formation. It will be recalled that these are the drive derivatives, the unpleasure evoked by the drive derivatives as these relate to the calamities of childhood, the defenses, and self-punitive trends. These four elements of conflict he considers still essential. Brenner then proposes major replacements for the psychic agencies. We should think of the drive derivatives rather than the id, of a person in the ordinary sense of a whole, entire human being instead of the ego, and of the calamity of parental disapproval instead of the superego. His major and perhaps most consequential and complicated doubt is about the value of structural theory altogether, because it divides the mind into separate agencies or structures on the basis of a false and concrete view of literal separation between these major structures.
He told us that the view of the mind divided into three agencies is predicated on a group of simplified and erroneous assumptions which share one feature: static separation as contrasted with dynamic and systemic interaction. My own view is that Brenner is doing us a great service because he is dealing with a problem which has never been adequately appreciated; namely, what is the most advantageous way to describe the interactions encompassed by the hierarchical organizations and suborganizations of human mental functioning? It is one of the most important problems that Freud struggled with throughout his writings. To illustrate what I have in mind I will remind readers that we need to clarify what part of the mind it is that compromises the elements of conflict into the best available compromise formation. Consider further that important symptoms rarely constitute only one compromise formation, but are actually compromises between compromises in which complex ensembles of interactions are deployed.
As far as I can tell, Brenner does not distinguish the terms agency and structure in this paper although he is obviously well aware of it. My views about his proposals will be more clear if I state at the outset that it is clarifying to distinguish the theory of separate agencies from the theory of psychic structure which has emerged in recent years. The term structure has always caused confusion. The term psychic structure does not appear in the general index of Freud's writings. In fact no one seems to know who coined the term structural theory in the first place, and many authors have pointed out the tendency to reify and concretize the notion of the three psychic agencies as though they were federal office buildings in the human mind. Psychic structure really has never meant anything other than psychic functioning wherein that functioning was theoretically conveniently assembled into stable groups of functions. I agree with David Beres (1965) who said that the structural theory should have been called the functional theory in the first place. Since a major aspect of Brenner's criticisms depends on a clear understanding of just what degree of false separation resides in the concepts of id, ego, and superego rather than in their misapplication, I must now ask the reader's indulgence to discuss some tedious but essential distinctions.
I think we must have a very clear idea of psychic structure in mind in order to evaluate Brenner's present views of structural theory. The term structure is only a metaphor. We use it to connote abiding configurations of functions; we use it to connote the continuity and stability of a functional organization. One price we pay for using the word structure is that because we are using a noun to connote dynamic processes we are pushed without realizing it toward reification. The same problem is inherent in our use of the noun mind which should really be a gerund form minding. Minding and structuring would be more accurate terms to describe mental functioning.
If we grant that we should still retain the notion of conflict in which at least one component of the conflict is unconscious, and that drive derivatives, unpleasure, defense, and self-punitive trends are still to be considered a valid description of the condition of the human mind, are we not still required to find some way to account for the universal similarity, stability, and continuity of these abiding components of mental conflict which we are retaining? In other words, if we say we want to organize our observations of conflict with the concept of compromise formation, we are at the same time saying that we must have a simultaneous assumption of some form of psychic structure. The abiding configurations of each of the four components of conflict (i.e., the drive derivatives, unpleasant affects, defenses, and moral considerations) are each psychic structures. They are just as much psychic structures as are the three major agencies of the Freudian psychic apparatus, but they are on a lower level of abstraction.
A horizontal or cross-sectional view of the functional integration of the three agencies expresses the fact that structural theory is a systems theory. Brenner was right to remind us that it is imperative to remember that in a systems theory anything that occurs in one part of a system affects the entire system, and that no part of the system can ever function without entailing the participation of the entire system. It has been my understanding that this has been an ever increasing feature of the best way to define structural theory in recent years, due in no small part to the writings of Brenner himself. In fact I agree with Brenner that no special part of the mind deals only with defense, or only with reality, or only in a mature and reasonable way. That is because our clinical experience requires that we utilize a systems theory in which everything in one part of the system is affected by and in turn affects everything else in the system. Brenner's new proposals in this sense are also proposals for a systems theory. I wish especially to emphasize the fact that this systemic interdependence between the id, superego, and ego is not only functional or horizontal. Everything that occurs developmentally in the longitudinal sense also affects the entire system. The three agencies are conceptually linked then in a horizontal-functional sense and in a longitudinal-developmental sense. One of the advantages of the terms id, ego, and superego has been that it has allowed us to artificially but conveniently separate these functional organizations for purposes of discussion and investigation of the developmental fate of each of the three major functional components.
I have previously (1988) suggested that the 1923 definitions of the three agencies were misleading because they ignored the dual interrelationships of functional inseparability and developmental continuities in the id, ego, and superego model. This is one of the important differences between the 1923 tripartite structural theory and the current view of structural theory. Modern structural theory places central emphasis on the developmental consequences for the present interactions of each agency with the other on their specific prior developmental pathways and interactions. I wonder if Brenner's new formulations will allow us to account for developmental vicissitudes in a better way. In the form that we are discussing today, the new proposals seem to float without developmental anchors. Compromise formations give us a better cross-sectional or horizontal view of local conflicts but the three agencies give us a better longitudinal view of macrointeractions which is advantageous for the study of the developmental vicissitudes of the components of conflict.
If we omit the requirement that an organization of mental functions must have both functional and developmental systemic qualities as some authors have, it becomes all too easy to call anything that endures in the mind a structure. A memorized list of nonsense syllables is no more a structure than is a group of memories. As a result of this omission we have seen the nomination of too many conceptual entities for the status of "structure," such as the confusing term microstructure. Thus the word structure has suffered the fate of the word instinct at the beginning of this century. Just as we formerly observed an ever expanding number of "instincts" then, so in recent decades there has been a hyperinflation of new "structures." Brenner is right to insist that no function of the mind such as perception can be allocated to just one of the agencies and that none of the agencies ever acts alone. Freud implied that in his New Introductory Lectures when he said: "here is another warning. . . . In thinking of this division of the personality into an ego, a superego and an id, you will not, of course, have pictured sharp frontiers like the artificial ones drawn in political geography" (1933, p. 79). Analogous to the reification of the agencies is the error of a concretized and schematic view of the possibility of endowing the agencies with separate functions rather than as functional centers of special emphasis. They were never intended by Freud to be viewed as three separate parts of the mind. It is precisely the issue of conceptualizing hierarchically stratified levels of organization that has eluded us to date. The history of the evolution of structural theory can be described as the attempt to refine our view of the interaction of interdependent functional organizations with suborganizations.
It is conflict that requires self-deception. It is self-deception that leads to the transformation and disguise of motives and meanings which we pursue in our daily work. We are asked to continue to believe that conflict is at the center of our need to deceive ourselves about our motives. What is it then that accounts for the stability of conflicts both in normal and pathological conditions? How do Brenner's new proposals account for functional stability in the human mind in a manner that will replace that explanation in our prior theory? Aside from a new terminology, are Brenner's proposals still not linked to the structural notion of a commitment to describe the human mind in terms of its dynamic functions?
If there is no way to account for functional stability in the human mind, how are we to account for the causes and consequences of conflict? If we retain the idea that we need some way to conceptualize functional and developmental continuities in the human mind, then we are still dealing with a structural theory. How can there be enduring conflicts and stable compromise formations whether normal or pathological without psychic structure? To summarize my first question more succinctly, would Brenner agree that his new proposals still constitute a structural model, albeit with different "agencies"? It would appear that his proposals offer fewer chances for reification and a greater emphasis on the fact that this is truly a systems theory; but is it not still a structural theory? In the event that he is saying that we still need some kind of systems theory of psychic functional groups, but also proposing that we refine and clarify the nature of those groupings, I am in firm agreement with him, especially about his first major doubt which is about the confusing notion of the id. He said, and I can understand, that space did not allow him to deal with the conceptual problems of the id at this occasion. I will have to say something about those issues to make my own further questions more coherent.
There has been confusion and disagreement for many years about the best way to account for the functions of the id. The id originally was the frontier concept between the mind and the body because it was the agency of the drives that arose from bodily processes. The id was also a frontier concept because it was supposed to have ideational and representational content as the repository of the repressed unconscious. It was also considered to have a second nonmental dimension, represented by its inputs from other bodily processes such as coenaesthetic sensations, and most importantly for purposes of our present discussion, sensations of pleasure and pain. Thus the id was assigned both perceptual and memory capacities and Freud never spelled out how those capacities were to be distinguished from the perceptual and memory functions of the ego.
Max Schur (1966) discussed many of these same conceptual problems in his 1966 monograph about the id. In agreement with Merton Gill (1963), who had made a pioneering contribution to the explication of these problems with the id concept in his 1963 monograph, Schur said that since the id was supposed to contain repressed ideational content it was necessary to view it as structured in some way. It was not just a seething cauldron of bodily drive energies awaiting processing in the id refinery into the mental fuel of the psychic apparatus. It was also supposed to contain ideas called id wishes. Since a wish implies conflict and ideational content it implies some form of psychic structure. But if the id was structured, how was it to be distinguished from the early ego both in the sense of development and function? Yet if we wanted to be consistent in our frames of reference we were driven to reducing the id to an energy concept without ideational content. An energy id would be nonideational but also nonmental; that too was an untenable position. This was the state of affairs in the early sixties, and Brenner is helping us to see that this confusion is still with us.
Nor have I yet considered other puzzles about our views of the relationships between the early ego and the id. The developmental line from reflex discharge to wish fulfillment implies that the id somehow undergoes development even though we also sometimes say that the id never changes. How are we to account for developmental and maturational changes in the id if it is not structured in some way? But if it is structured how does it differ from the ego of the child? A related problem is Freud's original view of compromise in the id itself on the basis of instinctual fusion, a notion which few analysts still utilize.
Arlow (Panel, 1963) put his finger on the key issue when he asked almost thirty years ago: "What are the functions of the id?" Gill and Schur also concluded that if the id is structured it is only a primitive ego, and if it is not structured it has no functions. I think it might also have been with these problems in mind that Brenner has proposed that it is finally time to relinquish the notion of an id. Certainly this was foreshadowed in his 1982 book when he proposed that we speak of drive derivatives rather than drives. That proposal had the advantage of mentalizing the id once and for all, but it still carried forward the blurring of the distinctions between subordinate and supraordinate compromise formations. That is because drive derivatives are also defined as wishes, so we seem to confront a confusion in frames of reference because wishes are themselves compromise formations. The drive derivative seems to be at one point only a component of a large compromise formation such as a symptom, and at another it seems to be a wish and wishes themselves are compromise formations. This is actually a spurious problem. Compromise formations are rarely simple. Compromise formations in clinical practice are highly complex, hierarchically organized groups of subcompromise formations. Brenner correctly points out that it is also useful to view ego functions as compromise formations. He gave perception as an example. Again, perception itself is a compromise formation but as part of a symptom it is a component of a larger compromise formation.
Brenner pointed out the same issues in his clarifying redefinition of the superego concept in 1982. The superego is itself a compromise formation as well as a component of other compromise formations. Let us take stock. We started with the three concretized agencies. I suggested that an additional problem with the concept of the id was that it was really no different than a very primitive or early ego. I summarized certain of Brenner's points. He said that each individual function of the ego was itself a compromise formation and that we could view the superego also as a compromise formation rather than as a separate agency. That was because the id, ego, and superego were merely hierarchically stratified ensembles of compromise formations which were in every instance reducible to the familiar four elements of conflict. There was no sense in separating the agencies because it would force us to disregard the fact that no one part of the mind is dedicated exclusively to rationality and correct perceptions of the external world. If that is an argument for revising our theory I have another question. Every defensive function, self-punitive trend, and unpleasure can be shown to be just as much of a compromise formation as the drive derivative. So a symptom viewed as a supraordinate compromise formation is not a quartet of irreducible components. Each of the components is itself also a smaller compromise formation. We seem to teeter here on the edge of infinite regress. I shall return to this problem when I consider the criticism that Brenner is reducing everything in the mind to compromise formations.
There is another problem to consider first. With our original three agencies we at least had a person whose mind was supraordinate to his id, ego, and superego. I say that because I wonder if there is not evident here a confusion in levels of abstraction or what is sometimes called a category error. Just as one could not logically speak previously of a mind in conflict with its defenses, which are a suborganization on a different level of abstraction, so as Brenner knows one cannot speak of the mind of a person in conflict with his own drivederivatives because it would entail confusing supraordinate and subordinate functional organizations. Yet if the person replaces the ego that seems to be the consequence. If our old structural theory left us with an ego that was a homunculus, do the new proposals shrivel a person into one part of the mind?
The ego operated on a gradient of levels both participating in the creation of compromise formations and mediating between the compromise formations and integrating them. It is unclear how Brenner's person can replace the prior ego in this regard. Moreover, this leaves us with a vast orchestra of compromise formations without a conductor. What part of the mind in this new model does the compromising? I don't think we can yet dispense with the notions of the organizing, integrative, and synthetic functions of the ego. How else do we account for the creation, maintenance, and coordination of the various compromise formations?
The principal reason for abandoning the structural theory according to Brenner is because it gives a misleading view of the role of conflict in normal mental functioning. Every aspect of mental life, whether it is normal or pathological, has been shown to be a compromise formation. Therefore the ego itself is dynamically the same as a neurotic symptom or a parapraxis. It will be recalled that this is the same reason he gave in his 1982 book for changing our definition of the superego. The superego itself was a macrocompromise formation, if you will, or a group of compromise formations sharing in common moral considerations. Since there is no special part of the mind which is always mature, realistic, and uninfluenced by the drives, why not discard the notion of an ego altogether? Since there was no special part of the mind used only for defense, why not discard the mechanisms of defense? Perception, reality testing, memory--these are all compromise formations. Everything can be shown to be a compromise formation. That is, of course, one of the criticisms one hears most often about these views. Is there nothing in the human mind that is not a compromise formation?
I think it would be clarifying in that regard to introduce a distinction between two uses of the term compromise formation. The common and prevalent usage has been to designate an "entity," called a compromise formation, that describes the actual existence of a special style of functioning in the human mind. I propose a second definition. Compromise formation may also be defined as a theoretic construction; it is an organizing matrix or agenda that derives from the structural model. If one accepts the view that there is something akin to a regulatory principle in the inherent tendency of the human mind to deal with unconscious conflicts by pursuing maximum pleasure and minimum unpleasure, in this way then one should always look for those components of conflict and organize one's observations in accordance with that framework. Brenner does retain the name of one of the original regulatory principles, but speaks of the affects of pleasure and unpleasure rather than in terms of psychic energy. Nevertheless the Brenner pleasure principle has an unwieldy theoretic assignment because it is the only element in his theory which stands outside of the compromise formations themselves and which seems to evoke the creation of these compromise formations. Until now we have not found an adequate theoretic home for the regulatory principles, and this is still so for Brenner's "energy free" pleasure principle because it is not itself an affect and is on a different level of abstraction than the compromise formations. I wonder also whether Brenner would agree that we need some way to account for both the creation and subsequent management of the myriad compromise formations in the mind which would logically be in a different frame of reference than the compromise formations.
The first definition of compromise formation allows Brenner to claim that everything in the mind has been shown to be a compromise formation. I would suggest using the second definition to say that everything in the mind can be shown to be a compromise formation, because it allows for the possibility that there are things in the mind besides compromise formations. If we fail to distinguish these two definitions we have a tautology. We first say we will organize whatever we observe into the four categories of conflict that comprise compromise formations, and then we will say that everything we observed has been proven to be a compromise formation instead of saying this was a useful way to organize our understanding. It is a bit like saying that it is remarkable how well shoes fit feet, as though they weren't designed to do that very thing.
Structural theory, as Brenner describes it in the earlier passages in his paper (pp. 478-479), gives a misleading account of the importance of distinguishing pathology from normality on the basis that we should monitor changes in our patients with the hopeless and misleading expectation that their conflicts will disappear if we do the right thing. I was surprised to learn that there were still a large number of analysts who have been led by structural theory to expect the smashing or disappearance of their patients' conflicts. My own view, thanks in large measure to the work of Brenner himself, has been to view the results of treatment in terms of a beneficial change in the nature of the patient's methods of dealing with conflicts rather than to expect the disappearance of conflicts. Most psychoanalysts no longer share the expectation that conflicts should be smashed or disappear as a consequence of successful psychoanalytic treatment. Certainly Freud retracted his own overly ambitious therapeutic expectations, as we know from his sober reassessment in his paper "Analysis Terminable and Interminable" (Freud, 1937).
Time will not permit the discussion of another question about the pros and cons of substituting a person for the ego. That question in turn pertains to the as yet insoluble problems of consciousness, subjectivity, and the elusive, vexing problem in structural theory of accounting for the sense of "I" or "me." Since we recognized that the ego as a metapsychological system could never be equated with the self, we have had this gap in our theory which remains unremedied with these new proposals. Ironically, the self as a metapsychologic system in self psychology suffers the same problem.
At this point I wish to address what I view to be a point which is embedded rather than explicit in Brenner's new views. This is the problem of defining what criteria to use in deciding whether to give up or to continue using certain theoretic terms. Many psychoanalytic terms can only be defined in their historic context, many others are defined only very loosely, and it is not uncommon to find psychoanalysts disagreeing about whether certain terms continue to deserve further use. Just to indicate what I have in mind, let us consider the terms sublimation, acting out, character, and countertransference. Even analysts who have long since discarded the concept of psychic energy and libido continue to use the term sublimation. This is an example of preserving the original shell of a term and discarding the connotational core of the term. There are deeper and well known problems which transcend the dynamic evolution of any language. The entire problem of the relation between one's theory and technique is complex, and to put it mildly, nonlinear. We know too little about how each of us actually uses theory while working with patients. We want to know what the advantages of a new proposal are, we want to know the disadvantages of the old, but we also want to know that changing a theory will have demonstrable and consequential significance. It is not enough to say we should discard a term because it describes something that is also described by another term. There can be important advantages to viewing clinical phenomena in more than one way. Now at this point I have a difference with Brenner. I think he has argued persuasively to demonstrate that in actual practice we should often use structural terms at a much lower level of abstraction than the frame of reference of id, ego, and superego, which is at the most abstract level of our theories. That alone does not necessarily mean that the very abstract terms are worthless, it only means that some of us have been misusing these terms. The word ego does not force any of us to think that successful psychoanalytic treatment destroys pathogenic conflicts--that mistake is rooted elsewhere. Now it is one thing to say that the components of a conflict have provenance in functional and developmental groupings of three major types, which we call id, ego, and superego, and another thing altogether to speak of three noncommunicating, separate agencies. It is to Brenner's credit that he insists on showing us how misleading it is to concretize our views of structural interaction. It is a different matter to insist that we stop utilizing a variety of levels of abstraction for purposes of theory discussion, theory formulation, theory revision, and theory validation and disconfirmation.
Brenner has made an eloquent, continuing effort in a series of papers and books over some decades to eliminate terminologic redundancy. He has been among our best known theoretic minimalists. He deserves more credit than any other single author I can think of for our recent healthy tendency to dispense with jargon in our literature. Although this has been highly useful I have some questions about the possibility that he might have been better advised to insist on the ambiguity of certain terms than to insist that because of definitional imprecision that the concepts denoted by these terms be abandoned. To illustrate what I have in mind I will remind readers of his arguing that the concept of the therapeutic alliance should be deleted principally because it was a form of transference rather than a separate entity. More recently I have had an exchange of views with him about the concept of resistance which he feels to be unnecessary for the same reason because he believes that resistance is only another term for the transference. He has also argued that we should delete the defense mechanisms, the id, the superego, and now the ego. In every one of these instances he has argued forcefully against fictitious entities that are perpetuated by anachronistic terms that are no longer useful. I think this raises the question about what is useful and to whom is it useful.
The topic of eliminating redundancy in a theoretic system touches on conceptual and epistemologic problems that I am not qualified to discuss adequately, but I have one point to make. We are dealing here with two quite different reasons for arriving at a decision to delete an idea or a term. The first reason is that the term describes a fictitious entity. The second is that we should delete a term because it is an alternative description of the same phenomenon denoted by another term. I am in complete agreement with Brenner's wish to avoid perpetuating conceptual fictions. Only time will tell whether his new proposals will achieve that goal. The second form of deletion is different. To say, for example, that resistance is indeed only another name for transference contradicts the view that it is useful to view resistance as a special organization of the transference. At issue is the difference between defining similarities and noting distinctions. It is a well-known problem in the history of taxonomy. We do not want to invent fictitious species. On the other hand it is misleading to say that a man and a dog are the same thing because they are both mammals.
Phobias are another example. Brenner has said that the term phobia is not useful because it is essentially a compromise formation no different in its dynamic structure than other symptoms except for the one common element in all the phobias of the use of avoidance (1982, pp. 157-158). But it is precisely the element of avoidance which it is useful to highlight. The element of avoidance allows us to see that the phobias as a group entail externalization and alloplastic implications when compared to obsessional symptoms. This realization has implications for a better future understanding of the phobias. Perhaps the difference here is that Brenner prefers to see how similar things are and I prefer here to see the differences. Yet it seems to me that we must always have both points of view in that there is a valuable dialectic tension between the two. I think it can matter a lot to make distinctions between two things which appear at first glance to be alike. Many things in nature are similar but not identical. It matters a lot whether the same water molecule exists in the form of a liquid, solid, or a gas. It matters a lot whether the water is in fog, a river, a drinking glass, or in tears. The tension between comparisons and distinctions is a vital part of any scientific enterprise. Brenner's new proposals force us to think hard about some comfortable assumptions. Has he succeeded in persuading us to seriously consider that we relinquish the idea of the mind consisting of separate structures, the id, ego, and superego? I would say that he has, but that such a misunderstanding does not warrant dispensing with the structural theory if indeed he is even suggesting that.
In conclusion I believe the most valuable feature of Brenner's challenging proposals is that it forces us to address the neglected problem of how best to account for the interaction of larger and smaller structural organizations. Once again, thanks to him we will be compelled to find a better way to view the mind in conflict.
Beres, D. (1965), Structure and function in psychoanalysis. Internat. J. Psycho-Anal., 46:53-63.
Boesky, D. (1988), The concept of psychic structure. J. Amer. Psychoanal. Assn., 36: 113-136.
Brenner, C. (1982), The Mind in Conflict. New York: International Universities Press.
Freud, S. (1923), The Ego and the Id. Standard Edition, 19. London: Hogarth Press, 1961.
------ (1933), New Introductory Lectures on Psychoanalysis. Standard Edition, 22:3-182. London: Hogarth Press, 1964.
------ (1937), Analysis terminable and interminable. Standard Edition, 23:211-253. London: Hogarth Press, 1964.
Gill, M. (1963), Topography and Systems in Psychoanalytic Theory. New York: International Universities Press.
Panel (1963), The concept of the id. E. Marcovitz, reporter. J. Amer. Psychoanal. Assn., 11:151-160.
Schur, M. (1966), The Id and the Regulatory Principles of Functioning. New York: International Universities Press.
