The most important question about which Brenner and I disagree is whether or not we can preserve the definition of unconscious conflict without some kind of notion of psychic structure. Brenner would be the first to agree that the concept of unconscious conflict is a bedrock definitional requirement for psychoanalysis. He and I would therefore also agree that pathological compromise formations are often enduring and stable. But this means that the components of the compromise formation remain continuously interactive and that requires the assumption of functional continuity. I do not see how one can have a notion of unconscious conflict without an assumption of functional continuity in the human mind. Functional continuity is essentially another term for psychic structure. How best to account for these functional continuities remains to be seen, but functional continuity is in this sense just as much a part of the idea of compromise formation as it is of structural theory. Brenner's new proposals actually depend on two concealed assumptions. The first assumption is that we can simply ignore the requirement that there be functional continuities in the human mind in order for compromise formations to develop, endure, or change. The second is that the theory of compromise formations which is a subset or component of modern structural theory can be simply detached from the rest of its theoretic linkages to structural theory without detrimental consequences regarding coherence and consistency.
Psychoanalysis is the science of the mind in conflict. Does this necessarily mean that we should define the domain of psychoanalysis as consisting only of the study of the components of compromise formation? Is there nothing else in the functioning of the human mind which should properly attract our attention? When the mind of a person "decides" to modify a compromise formation are there no advantages to clarifying such processes as memory, perception, and learning in so far as these processes are relevant to alteration in compromise formations? When we decide to advise psychoanalytic treatment we often make explicit predictions about the person's capacity to regulate and tolerate unpleasant affects. When we compare one person's ability to tolerate unpleasure to that of another person, are we only comparing their compromise formations?
In our daily work we deal primarily with understanding the meaning of the patient's communications. This work is for the most part on the lowest level of abstraction in theoretical terms. We are trying to observe and understand the most important conflicts of the patient at any one time in any one session. It is when we are trying to understand what the patient is feeling in a given hour that we can most easily report our views without resorting to such highly abstract terms as the ego or the superego. When we are at work in the hour we are mostly attempting to understand the meaning of what the patient is communicating. We can therefore discuss the hour without ever resorting to the terms ego or superego. When we make diagnostic and prognostic predictions we attempt to understand how the patient functions. These two enterprises are both essential, are interrelated in a complex manner and are currently guided by discontinuous theoretic assumptions on various levels of abstraction. Physics is another science currently seeking a unified theory but is at present required to make do with theoretic discontinuities. When we make reconstructions, when we formulate conjectures about pathogenesis related to developmental conflicts ("postdictions"), we also need structural concepts or a nomenclature of abiding functional capacities. Although the notions of meaning and function are also actually inseparable it is useful to discuss them in their differing contexts.
The analyst doesn't only try to understand what the patient is communicating in a given hour. There are times when the analyst must make an assessment for diagnostic purposes or to determine whether or not a patient is progressing in a satisfactory manner, or whether or not to consider if termination of the analysis might be indicated. It is in these evaluative functions that the analyst needs to use conceptual tools that are more generalized and abstract. The analyst in Brenner's terms might be said merely to be comparing the compromise formations at this later time in the treatment with the compromise formations at the outset of the treatment. The analyst would say, if he or she felt that the patient was ready to terminate, that there had been a favorable alteration in the patient's previously unhealthy compromise formations. For example, the patient now suffered less pain, could function with fewer inhibitions, could experience more pleasure, and required less self-punishment. But in such a view one is left without a group of concepts to allow us to even address the question as to what it was that effected these beneficial changes and what it is that causes these good changes to endure. Floating compromise formations do not provide a theoretic home for therapeutic change. Nor does a theory consisting only of floating compromise formations allow for adequate explanations of unfavorable change.
Brenner stated in his original paper that he would feel successful in his purpose if he could persuade us to give serious consideration to the possibility that the time has come to relinquish the idea that the mind is best understood as consisting of three separate agencies termed the id, ego, and superego. In that respect I certainly agree with him if by "separate" he means concretized agencies that could function ever or at all without interaction with the other two. That was the error that Freud himself pointed out some sixty years ago, as I indicated in my original discussion. Much depends on a clear understanding of the word separate. If "separate" means something like an agency of the government that can act alone without consulting another agency, like the FBI and the DEA, we are dealing with a straw man since most analysts have understood that such a concretized view would be incorrect. But separate in the sense that I use the term means something else. It means that it is theoretically useful to discuss abiding functional continuities as though these "separate" functional groupings could be frozen in a fictitious moment of time. An analogy would be to speak separately of the respiratory and circulatory systems that could never function separately in reality, but whose interaction is convenient to discuss "separately."
I have written elsewhere about the problems with our notions of psychic structure (Boesky, 1988). Here I will only say the following. The Rappaport and Gill (1959) definition of structure is anchored in the notion of slow rates of change. Among the problems with this definition is that anything that endures in the mind deserves to be called a structure. Examples of the overinclusiveness of this definition include the tendency to call object representations structures. Also, certain obsessional phenomena are quite enduring, but there are healthy creative acts that are highly mobile and evanescent. More important is the fact that the entire notion of psychic structure is a systems theory. Anything that takes place in one part of the system affects the entire system. There can be no separation of any of the "agencies" of the mind. The notion of psychic structure, as I said in that paper, is anchored in the functional interrelatedness of all mental activity. It is not only a question of developmental functional continuities, but also the inseparability of the functioning of various activities of the mind that epitomizes my understanding of the notion of psychic structure.
Actually one could say that the device of speaking "separately" is equally true of compromise formations wherein Brenner speaks of four "separate" interacting components of conflict: a drive derivative, unpleasure, defense, and moral considerations. If it is permissible to "separate" these actually inseparable components, why is it not permissible to "separate" the same components on a higher level of abstraction? When I speak of a higher level of abstraction I have in mind the idea that we cannot logically think that any one solitary compromise formation is not linked or potentially capable of linkage to all other compromise formations in the mind. So if we spoke of all of the drive derivatives of a particular patient or all of the defenses of a patient, in the plural sense of all of that person's compromise formations, we would be generalizing from the single to the multiple and from the less abstract to the more abstract level.
In this way one might even consider applying Brenner's own suggestion about drive derivatives to our view of the ego. Just as he speaks of the drive derivative of one compromise formation in contrast to the drives of that person or of all persons, it might be useful to suggest that we speak of ego derivatives to describe specific functions in a single compromise formation or of one person in contrast to the ego of that person or all persons.
Are we not in fact on Brenner's lower level of abstraction still dealing with functionally distinguishable groupings in the mind? Here I wish to raise the question of distinguishing between concretized, artificial separation of continuously interacting functional concentrations as opposed to the theoretic convenience of speaking of these interacting components of conflict on an ascending scale of levels of abstractions. This exemplifies Waelder's (1962) views about levels of theoretic abstraction from the lowest to the highest that have always characterized psychoanalytic theory (as well as that of other scientific disciplines). It is for this reason that I prefer the word structure instead of agency since the latter was associated with Freud's earlier rather than later writings about structural theory.
The adult patient in analysis modifies his or her compromise formations very slowly. Important groups of mental functions are responsible in these instances for the invention of a method, normal or pathological, for dealing with conflicts in such a way as to maximize pleasure and to minimize pain. At issue here are two other problems, the pleasure principle and the relation of our notion of conflict to levels of abstraction. The latter is crucial, the former merely terminological.
I think it would be useful if Brenner spoke of the pleasure affects rather than the pleasure principle, which is encumbered with conceptual linkages to energy and outmoded aspects of libido theory which I believe Brenner would agree are outmoded. Brenner as far as I can tell doesn't mean to advocate that we should continue to use the concept of the regulatory principles. It seems quite clear that he merely means to say that the mind operates in such a way that a person will seek a maximum of pleasure and a minimum of pain. But that presents us with another problem caused by ditching the structural theory. How does it happen that what once caused pleasure now produces pain? T. Shapiro (1992) raised this question but I only learned of that after I wrote my own original discussion of Brenner's paper. The question as to how the child learns that what once caused pleasure now causes pain is an example of what I have in mind when I assert that we gain from structural theory a way to explain the developmental conflicts that lead to problems in structural differentiation as well as other forms of pathogenesis. This confusing issue is also linked to current controversies about so-called "deficit" pathology versus "conflict" pathology, because such a polarized contrast is misleading. Are we to be limited only to compromise formations that float free of the history of this crucially important transformation of pleasure into pain?
Brenner made a most important observation in his response to my discussion on pages 523-524: "one has to go beyond the clinical data of pathogenic conflicts, as Freud did in defining ego and id, if one is to justify a theory of agencies as opposed to a theory based on the operation of the pleasure-unpleasure principle." Brenner seems to imply that his newer views are closer to direct observation of the clinical data and much less marred by the encumbrance of fictitious theoretic entities. One can, for example, rewrite the events of a single hour and the opinions of the analyst about that hour without once resorting to the terms ego or id. But the use of compromise formations as an organizing method to understand the events of a single hour is also dependent on quite abstract theoretic assumptions and is in no way a simple atheoretic formulation. For example, it is impossible to deal with compromise formations as a theoretic model without invoking the assumption of unconscious mental processes. It is also necessary to assume that psychic determinism is a useful theoretic premise. If we speak of compromise formations we are also assuming that thousands of prior observations concerning bodily needs made by hundreds of psychoanalysts over a period of almost a century validly support the theoretic assumptions that conflicts as expressed by derivatives in the patient's associations during the hour are useful to apply.
On a different level of abstraction we are often inclined to give special weight to communications that occur at the very beginning of the hour or at times related to impending absences of the analyst. I cite only these examples to illustrate that the use of compromise formations is also dependent on a hierarchical group of theories on varying levels of abstraction, some of which are on a low level of abstraction (the opening associations often herald an important emerging theme) to a high level of abstraction (unconscious mental activity, drives, and psychic determinism). The distance from clinical data and the utilization of abstract theory are issues at the core of very complex scientific and philosophic problems which remain to be refined. It is not clear if I have read Brenner correctly, but he does seem to say that he is not going beyond the clinical data whereas those who adhere to structural theory are going beyond the data. I know from discussions with him that he would agree that there is no such thing as an observation or datum that is theory free.
One of the valuable challenges Brenner presents has to do with clarifying the most advantageous manner to understand the multiple and complex interactions of conflicts in the mind. Brenner is really forcing us to realize what a muddle we have to resolve in order to account better for the current ambiguities in our notions of the hierarchy of levels of abstraction in psychoanalytic theory formation. The redundancy and confusion carried over by the concept of an id illustrates this problem. I will return to that question shortly. The most important issue in my own thinking is that it is timely to consider the advantages of making explicit a reformulation of the interaction of conflicts with each other. I refer here to the fact that in actual clinical practice in daily sessions we don't deal with single or solitary compromise formations, but with the interaction and integration of groups of different compromise formations.
Even in the example of a "solitary" compromise formation composed of a drive derivative (or wish), unpleasure evoked by the drive derivative, defenses, and moral considerations it is possible to observe that there are at least three separate subcompromise formations operating in concert. The wish expressed by the drive derivative is itself a compromise formation in that it is a wish rather than an energy manifestation. A wish is itself a compromise formation with a history. The defenses are also compromise formations as Brenner himself has demonstrated so persuasively. And so are the moral considerations.
Now this complexity characterizes just a theoretically isolated "solitary" compromise formation. The actual clinical situation is far more complex and almost always entails more complex compromise formations in conflict with other complex compromise formations rather than a simple ego, id, superego conflict or a drive derivative, unpleasure, defense, moral issue conflict.
The following example will illustrate what I have in mind. A male patient preferred to have intercourse lying side by side while facing his partner. He had been enuretic until age 18. An important component of his enuresis was his mother's seductive behavior with him as a child coupled with her tacitly infantilizing him. Analysis revealed that his obligatory sexual posture was a complicated compromise formation. He feared his strong wish to lie on his back while the woman mounted him because of the unconscious fantasy that he would then be a woman. He was also fearful of the missionary position because of the unconscious fantasy of retaliation by his male rivals. Each of the unconscious fantasies, the wish to be a woman and the dreaded punishment by male rivals, was itself a compromise formation arising from a variety of related conflicts. The obligatory preference for the side-by-side position was not simply a single compromise formation consisting of a single drive derivative, a defense, unpleasure, and moral considerations. Instead, the side-by-side compromise formation was a macrocompromise formation formed to resolve a conflict between two other opposing compromise formations. He would be neither male nor female if he were in the side-by-side position, but he would also be both male and female. Furthermore there were important linkages between his enuresis as a child and adolescent with his orgastic experience. It was only after years of treatment that he began to experience orgasm rather than a mere flowing ejaculatory experience. It was useful to think in terms of the developmental continuities of his enuresis and his adult sexual conflicts. I am sure that Brenner would agree with that view but without something other than disconnected or "floating" compromise formations for a theoretic map I am unclear about the theoretic underpinnings available to him for describing the pathogenesis as well as the therapeutic change in this man both dynamically and developmentally. One might in general say the same thing about the entire series of oscillations between past and present which one encounters in observing the transference.
We need some theoretic basis to account not only for the developmental but also the dynamic functional continuities expressed in the patient's communications. We also need a theoretic basis to account for the correspondence rules which govern the methodology for translating the patient's communications into the meanings which govern the understanding of those communications. A theory of compromise formations in contrast to structural theory accounts for different dimensions of these correspondence rules. Arlow (1979) has outlined some of these correspondence rules.
The configuration of the material, the form and sequence in which the associations appear, . . . the repetition and the convergence of certain themes . . . the repetition of similarities or opposites . . . material in context appearing in related sequence, multiple representations of the same theme, repetition in similarity, and a convergence of the data into one comprehensible hypothesis constitute the specific methodological approach in psychoanalysis used to validate insights . . . [p. 203].
These empiric "rules" derive from highly abstract theoretic ideas that have been replicated by hundreds of analysts with thousands of patients. At issue here is whether or not a theory limited exclusively to compromise formations can better account for the dynamic, functional continuities that are the basis for these correspondence rules than does structural theory.
Brenner asked me to give a clinical example to illustrate why we needed a notion of functionally distinguishable agencies. I hope that the above clinical example will serve. As I view this matter each of the several components of conflict in any compromise formation constitute a functionally distinguishable activity. I therefore wonder if it might be useful to consider a reformulation of the actual interface of the components of conflict. In the topographic model the conflict was said to be between the system Ucs. and the system Pcs. In the original structural model (Freud, 1923) the conflict was said to be between the ego and the id. Here I suggest that we think of the conflict as located between one compromise formation and another compromise formation. Each compromise formation is in turn a resolution of its own subsidiary components of conflict. The advantage of this proposal is to avoid the reductive compression of the actual complexity of clinical experience by the misleading reference to single compromise formations. At the same time we are better able to view the hierarchical layering of compromise formations in the clinical situation. I do not propose deleting or abandoning the view of the components of conflict as Brenner has defined them. I do agree with him about the confusion of the id concept. My proposal is that we also think in terms of conflicts between compromise formations in addition to the conflict between a drive derivative, a defense, and a moral consideration. I also think of these integrations of multiple compromise formations as both horizontal and vertical. That is to say, I think of groups of compromise formations on an ascending and descending level of complexity in the vertical sense of levels of abstraction as well as thinking of a varying number of compromise formations integrated on the same level of abstraction. I think of the ego as a still necessary concept to help us to describe the integration of all of these compromise formations in the mind (in addition to serving numerous other functions).
Brenner has questioned why it is that there is a logical contradiction to his proposal to replace the term ego with the word person. He said (p. 482) "Instead of positing 'an ego' the new theory speaks simply of an individual, of a person, and of that person's mind." He then said that when unpleasure arises 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. We now seem to have a "mind" rather than an ego. There is a difference in the frames of reference between the term mind and the term ego. The ego is only one part of the mind.
Second, the ego is one of the components of the mind which can only be observed by the psychoanalytic method. More accurately, the ego is a part of the mind which can be postulated with logical consistency to account for certain observations gathered in the psychoanalytic situation. Another confusion results from what has been called a category error in which apples are compared with oranges. A person can be troubled about his or her wishes but cannot be in conflict with his drive derivatives. The ego can't decide anything, only a person can do that. The entire ego cannot be in conflict with a drive derivative; only a part of the ego can be. The mind cannot be in conflict with a drive derivative at all; the mind is all the mental functioning of a person. Brenner says he sees no logical difficulty. "Can't a person both wish for and fear something at the same time?" Yes a person can. But are we to substitute the word person for ego?
I am also puzzled about the advantage of substituting the term mind for the term ego since we are now saying that the mind has the function of regulating pleasure and pain rather than the ego. The mind is just as much a group of functions as the ego. It is an atheoretic general term, but nevertheless it is a masked concretization in that it is a noun used to connote functioning instead of a verb. The mind in a general sense is simply all of the mental functions of a person. If it is permissible to speak of an abstraction such as the mind in such a way why it is less permissible to speak of "an ego"? There is actually in a literal sense no such thing as a "mind" in spatial-temporal terms.
That conundrum has been well known for a very long time. One can speak more accurately of "minding" than of a mind. I would therefore say that if I have understood Brenner correctly the question is not only one of logical inconsistency but of arbitrariness. It may be that there is also another type of confusion about frames of reference. The terms id, ego, and superego are in a very different frame of reference than the term mind. All that we can now know about unconscious mental processes is what we can observe in the psychoanalytic situation, and all that we can say about what we observe with that method must be expressed in that specific frame of reference. The term mind is not easily accommodated to unconscious mental functioning unless we revert to the topographic model. Shall we now resume speaking of the unconscious mind rather than about unconscious ego functioning?
In one instance I think that Brenner seems to have misunderstood me. He asked me (p. 523) which functional or developmental criteria characterize the id and which the ego. It was my intention in my discussion to state in some detail that there was no coherent difference between the early ego and the id and that the logical trap of an unstructured id versus a structured id rendered the id concept highly dubious, and that this had been pointed out some thirty years ago by Merton Gill and Max Schur. So I will merely say here that I am in agreement with Brenner on this point but that he and I have drawn different inferences about this problem. I see this confusion early ego and id as an important example of the need to revise our views of the hierarchical layering of compromise formations.
Brenner's challenging questions have helped me to see an important point about which I was unclear. He correctly quotes me as stressing in my discussion the developmental issues that I feel are better explained with structural theory. I have elsewhere (Boesky, 1988) discussed the dual criteria of the structural theory in terms of the functional inseparability of the interaction of the components of conflict and its longitudinal developmental continuities. I do not therefore claim only that developmental issues are clarified with structural theory. I also claim that the functional inseparability of the interaction of the components of conflict are better explained with structural theory. I referred to that above in the case illustration when considering the functional continuities in a variety of different compromise formations at one moment in time as well as developmentally.
I wish to emphasize one other point. If the new views of Brenner replace the structural theory, we have a problem about accounting for conflict. The notion of conflict is quite properly still central to Brenner's proposals at the present time. If we deal only with compromise formation as an organizing model, how do we explain the development and the persistence of conflict? Is it simply an integral manifestation of human mental functioning to be vulnerable to conflict just as it is part of the inborn apparatus of the mind to achieve maximum pleasure and minimal pain?
I wish also to compare modern structural theory with Brenner's new views. Both models depend on highly abstract theoretic assumptions; both models go beyond immediate direct clinical observation; both models consist of functionally distinguishable groups of mental activities expressing the various activities of the unconscious human mind. Where Brenner sees widespread and misleading concretized views of separate agencies to be required by the continuance of the terms ego and superego, I believe that one can speak of functionally distinguishable tendencies which express inseparable interactions between the components of conflict.
Now I want to respond to some questions Brenner asked about my perhaps misunderstanding or misreading him. Certainly he is right to say that he never said that there were no mechanisms of defense, that there were merely no special parts of the mind that were reserved for defense, and I certainly agree with him about that.
I would still say it would be more clear to distinguish the methods of defense and the achievements of the defensive methods so that we could explicitly preserve the valuable idea that it is consequential to know whether for example a defensive deployment eventuates in forgetting or disconnecting, whether for example the mental content remains accessible to conscious awareness or merely "isolated." I actually agree with Brenner that the phobias do not constitute a diagnostic group. I also very much agree that if you want to study defense it is useful to study avoidance in all its forms, including phobias. In fact I don't think Brenner goes quite far enough in challenging the view that phobias are a diagnostic group. I expect he would agree with me that our basic nosologic methodology altogether is badly in need of clarification. The very idea of a "diagnostic group" is open to the grotesque misunderstandings of DSM-III (1980). The notion of abstracting and coordinating numerous observations about the defensive style of a given patient is useful. It is valuable to know that there are dynamic and developmental continuities in a patient's phobic defensive style just as there are advantages to abstracting and coordinating views about a patient's tendency to prefer other defensive styles such as isolating or repressive styles. It is part of the psychology of "character" to speak in these terms. It is one of the advantages of structural theory that we can move up and down the ladder of abstraction and generalization in just this way. My remarks about Brenner's views on resistance derive from personal conversations and correspondence with him which I have evidently misconstrued.
I want to thank Brenner again for his challenging questions. He has once again helped me to question unclear aspects of my own thinking. His new views may very well prevail in time to come. What matters most at this point is that he has forced us to address the problem that we need a better way to account for the hierarchical organization of our theories. There are few analysts who have had the courage to publicly modify their views and fewer still who have had the resources and skill to present such a challenge so effectively.
American Psychiatric Association (1980), Diagnostic and Statistical Manual of Mental Disorders, 3rd ed. (DSM-III). Washington, DC: American Psychiatric Press.
Arlow, J. (1979), Genesis of interpretation. J. Amer. Psychoanal. Assn., 27 (Suppl.): 193-206.
Boesky, D, (1988), The concept of psychic structure. J. Amer. Psychoanal. Assn., 36:113-136.
Freud, S. (1923), The ego and the id. Standard Edition, 19:3-89. London: Hogarth Press, 1961.
Rappaport, D., & Gill, M. (1959), The points of view and assumptions of metapsychology. Internat. J. Psycho-Anal., 40:153-162.
Shapiro, T. (1992), Discussion of "Modern conflict theory or beyond the ego and the id." C. Brenner, M.D. Meeting of The New York Psychoanalytic Society, December 2.
Waelder, R. (1962), Psychoanalysis, Scientific Method and Philosophy. Hook Symposium. 1959, ed. S. Hook. New York: New York University Press. Book Review in Psychoanal. Quart., 10:617-637.
