Jones, Ernest (1926). The origin and structure of the super-ego. International Journal of Psychoanalysis, 7: 303–311.
From 1913 Freud was interested in the phenomenon of guilt (Freud, 1912–1913, 1916d, 1917e). His interest arose as a problem not fully explained by his early theory of anxiety and libidinal energy. He eventually solved this problem with his structural model (Freud, 1923b) and the notion of the “super-ego” arising as a formation of identity from the introjection of primary objects. For many of his colleagues, the super-ego was a difficult concept to reconcile with his earlier sexual theories and the Oedipus complex. Freud was driven to amplify his idea that the super-ego is heir to the Oedipus complex a year later (Freud, 1924d). He also believed he had settled a range of other problems with his theory of the super-ego.
Jones (1926*), in the paper republished here, subjected the concept to careful and sometimes sceptical scrutiny. He was not the only analyst to do so, but his was the most magisterial statement. He asked a number of questions, including: how can the same institution present itself to the id to be loved instead of the parents and also as an active force criticizing the ego? If the super-ego is the internalization of a primary object, would not the surrendered object be that of the opposite sex? If the super-ego is a moral agent derived from the loved parents, why does it promote its influence with immoral sadistic means?
Others (Abraham, 1924; Alexander, 1929; Ferenczi, 1925; James Glover, 1926; Klein, 1927, 1933; Rado, 1925; Reik, 1924) also had questions or alternative points of view. The evident harshness of the super-ego was widely considered. The serious threat that backs the power of the super-ego is castration but, as Stephen (1999 [1946]) dryly observed, it is unlikely that most children have witnessed an actual parent conducting a castration. Alexander (1929) reflected on the origins of the super-ego in the death instinct rather than from introjected real objects. Ferenczi (1925), Glover (1925) and Abraham (1924) were interested in the pre-genital aspects of morality, and the oral- and anal-sadistic elements in an agency of mind that derives from genital phallicism. The wide critical appraisal of the super-ego concept demonstrates the difficulty for loyal analysts at the time to accept such a radical departure in Freud’s views as the structural model of the id, ego, and super-ego. Stephen, in 1946, argued on the basis that maturity is a respect for reality, that the super-ego is a pathological construction since it is based on the unrealistic belief in imminent castration. Such a penetrating criticism, and its credibility, leaves the concept somewhat bruised (Stephen, 1999).
Jones’s paper typifies this unease about Freud’s concept and in a way is a model of how the questions could be asked of Freud. Despite the uncertainties and questions, the development of the theory of the super-ego has contributed to understanding masochism, guilt, character formation, psychosis, and other psychoanalytic puzzles. However, the question that remains is whether there are particular inconsistencies in the theory, inconsistencies which therefore underlie the wide range of items in psychoanalytic metapsychology that the structural model explained.
It is unusual for the super-ego to be challenged or debated now, and it appears to be a permitted area where various views on a concept may co-exist, as if unnoticed. Many of the problems were discussed over the years, but without Freud’s authoritative voice there appears to be no arbiter between competing views and, as in many theoretical developments, progress is shaped like an evolutionary tree rather than a linear progression of knowledge. Jones’s paper would be a good start for a review of the steady march towards pluralism in psychoanalytic theory, as well as a current critical reappraisal of the super-ego concept.
Abraham, K. (1973) [1924]. A short study of the development of the libido, viewed in the light of mental disorders. In: K. Abraham (Ed.), Selected Papers on Psychoanalysis (pp. 418–501). London: Hogarth.
Alexander, F. (1929). The need for punishment and the death-instinct. International Journal of Psychoanalysis, 10: 256–269.
Ferenczi, S. (1925). Psycho-analysis of sexual habits. International Journal of Psychoanalysis, 6: 372–404.
Freud, S. (1912–1913). Totem and Taboo. S.E., 13: 1–162. London: Hogarth Press.
Freud, S. (1916d). Some character-types met with in psycho-analytic work. S.E., 14: 311–333. London: Hogarth Press.
Freud, S. (1917e). Mourning and melancholia. S.E., 14: 237–260. London: Hogarth Press.
Freud, S. (1923b). The Ego and the Id. S.E., 19: 1–60. London: Hogarth Press.
Freud, S. (1924d). The dissolution of the Oedipus complex. S.E., 19: 173–179. London: Hogarth Press
Glover, E. (1925). Notes on oral character formation. International Journal of Psychoanalysis, 6: 131–154.
Glover, J. (1926). The conception of the ego. International Journal of Psychoanalysis, 7: 414–419.
Jones, E. (1926). The origin and structure of the super-ego. International Journal of Psychoanalysis, 7: 303–311.
Klein, M. (1927). Criminal tendencies in normal children. British Journal of Medical Psychology, 7: 177–192.
Klein, M. (1933). The early development of conscience in the child. In: S. Lorand (Ed.), Psychoanalysis Today (pp. 64–74). New York: Covici-Friede.
Rado, S. (1925). The economic principle in psycho-analytic technique. International Journal of Psychoanalysis, 6: 35–44.
Reik, T. (1924). Psycho-analysis of the unconscious sense of guilt. International Journal of Psychoanalysis, 5: 439–450.
Stephen, K. (1999) [1946]. Relations between the superego and the ego. Psychoanalysis and History, 2: 11–28.
Ernest Jones
It is desirable to state clearly at the outset that this paper is of a peculiarly tentative character. The special occasion for which it was written, and the exigency of a time limit, induced me to attack an intricate theme when my own opinions about it are the very reverse of mature. Indeed, the essential object of the present contribution is merely to define a little more closely some of the complex problems involved and to invite further discussion of them; I would attach only a very restricted validity to any positive suggestions that may emerge in the course of the present remarks. The subject itself is concerned with one of the most important contributions that Freud has made to the science of psycho-analysis he created, and the spectacle of the following attempt to apprehend his latest teachings will serve as well as any other to illustrate the ever-pioneering nature of Freud’s work and the fact that his mind remains the youngest and freshest of any among us.
The particular problem to be considered here is that of the origin and actual structure of the super-ego, that is, the nature and genesis of the various trends composing it. As Freud himself says, “In other matters—for instance, concerning the origin and function of the super-ego—a good deal remains insufficiently elucidated”.1
As to the validity and value of the conception itself there will be universal agreement, for the reasons Freud gave when he postulated it can be definitely confirmed in any character analysis, and perhaps in any properly completed analysis. Further, a number of formulations in regard to it would appear to be equally well established. Thus, the genesis of the super-ego is certainly connected with the passing of the Oedipus complex, and the nuclear and essential part of its composition may be regarded as the direct imprint made on the personality by the conflicts relating to this complex;2 Freud neatly designates it as the heir of the Oedipus complex.3 Much is known also about the relation of the super-ego to the outer world and to the other institutions of the mind. The function it exercises is perhaps its clearest feature. It is to criticise the ego and to cause pain to the latter whenever it tends to accept impulses proceeding from the repressed part of the id. In this connection we may note the improvement Freud has effected in the terminology relating to the idea of guilt. He would confine the expression “consciousness of guilt”, or “sense of guilt”, to the perception of guilt on the part of consciousness4 and substitute that of “need for punishment” (Strafbedürfnis) when it concerns the unconscious ego, reserving “criticism” for the operation performed by the super-ego. The relation of both these active and passive aspects of the phenomenon to consciousness, however, is a very variable one; either or both may be unconscious, the latter more often than the former.5
When, however, we leave these valuable broad generalizations and come to a closer study of the problems involved, a considerable number of awkward questions present themselves. To mention only a few at this point: How can we conceive of the same institution as being both an object that presents itself to the id to be loved instead of the parents6 and as an active force criticising the ego? If the super-ego arises from incorporating the abandoned love-object,7 how comes it that in fact it is more often derived from the parent of the same sex? If it is composed of elements taken from the “moral” non-sexual ego-instincts, as we should expect from the part it plays in the repression of the sexual incestuous ones, whence does it derive its sadistic, i.e. sexual, nature? These and many other apparent contradictions need to be resolved. Finally, there is every reason to think that the concept of the super-ego is a nodal point where we may expect all the obscure problems of Oedipus complex and narcissism on the one hand, and hate and sadism on the other, to meet.
Before taking up the problems concerning the origin and structure of the super-ego, it is necessary to say something about its general relations, particularly the topographical ones. Relation to the outer world, the ego and the id. The ego is the part of the id that is altered by the influence of the outer world, and the super-ego is a differentiated part of the ego,8 again one brought about under the influence of the outer world. On the one hand we read9 that the super-ego stands nearer to the id than does the ego, is independent of the latter and represents to it the demands of the id, though the id can also influence the ego directly as well as through the super-ego.10 On the other hand it is just through its connection with the outer world, the reality demands of which it represents, that the super-ego gains its power of affecting the ego.11 The full explanation seems to be that the super-ego in some obscure way combines influences from both the inner and the outer world, from the id and from external reality, and that these are then directed towards the ego.12
Relation to Consciousness.—Two statements of Freud’s bear on this point. The super-ego may be for the greater part unconscious and inaccessible to the ego.13 It dips deeply into the id and is therefore farther removed from consciousness than is the ego.14 It is probable that the super-ego may be partly conscious, partly preconscious and partly unconscious; further that its relation to consciousness varies at different times. That it should be as a rule less conscious than is the ego may be explained by its relation to outer reality, for this relation was far closer in the past (in infancy) than it is in the present.
Relation to Repression.—It is the ego, not the super-ego, that performs the act of repression, though it commonly does so in obedience to the demands of the latter.15 It is important, however, to note that, especially in hysteria, the ego can keep from consciousness, i.e. repress, the feeling of guilt provoked by the super-ego’s attack on it.16 It should be possible in the future to describe this in economic terms as a balance between different amounts of pleasure and pain.
Relation to External Love-object.—Freud writes:17 “If a sexual object has to be given up, there is not infrequently brought about in its place the change in the ego which one must describe, for instance in melancholia, as an erecting of the object within the ego”, and he adds “the nearer closer conditions of this replacement are not yet known to us.” Throughout he appears to assume that the super-ego, which we know to be the heir of the Oedipus complex, results in this way from the incorporating of the parental figure that had to be given up in its sexual connection. But the evidence is fairly extensive that, though the super-ego may be derived from either parent and is commonly enough derived from both, it is normally and predominantly derived, not from the love-object that has been abandoned, but from the parent of the same sex. With the boy, for instance, it is derived in the main from the father, and when it is derived from the mother the chances are great that he will be homosexual. Freud himself points out this paradox,18 but offers no explanation of it. The discussion of bisexuality that follows in the context doubtless explains the facts of there being two types and of their often being mixed, but it in no way accounts for the phenomenon of the more normal type in which the heterosexual person derives his super-ego from the parent of the same sex as himself. It would therefore seem that a necessary condition for the process of incorporation is that the object incorporated must have thwarted the love impulses of the subject.
If this reasoning is sound, then it can only be that the mechanism of super-ego formation normally follows the order which Freud has described in connection with the attitude of a homosexual towards his brothers,19 namely, that original rivalry of a hostile kind was replaced by a friendly object-choice, which in its turn was replaced by identification. Applying this to the Oedipus situation, and taking again the case of the boy, we must assume that the super-ego usually arises from identification with the father where the initial hostile rivalry had been replaced by homosexual love. In the less usual and less normal case, that of the homosexual man, there are two possibilities open. Either the same mechanism as that just suggested holds good, which means that, the feminine component of his bisexuality being predominant, he deals with his jealous rivalry of his mother by a passing object love followed by identification with her, or else the identification is a means of dealing with hatred proceeding from the fact, characteristic of this type, that his castration fears are more closely connected with the mother than with the father. Both explanations accord with the law that the super-ego is derived from a thwarting object. The two explanations differ in that with the first congenital sexuality would be the ultimate cause of the undue reaction to the mother, with the second this might or might not be so.
It will be seen that here stress is laid on hostility20 being the essential condition of super-ego formation. This one may relate to the predominantly sadistic nature of the later super-ego, a matter which will be discussed presently. To recapitulate for the sake of clearness: it is suggested that the super-ego is derived from the thwarting parent, irrespective of whether this happens to be the primary love-object or not; normally it is a secondary love-object, the parent of the same sex.
The replacement of object-cathexis by identification brings about a profound change in the libidinal situation. The image thus incorporated into the (super-) ego serves itself as an object to the libidinal impulses proceeding from the id, so that more of them are directed towards the ego as a whole than previously; this constitutes what Freud terms “secondary narcissism”.21 Along with this goes a desexualization of the impulses, a kind of sublimation, and this important process gives rise to interesting problems. Freud hints that it is due to the giving up of sexual aims implied in the change from allo-erotic into narcissistic libido. To quote his exact words: “The transformation of object-libido into narcissistic libido that takes place here evidently brings with it a giving up of sexual aims, a desexualization, i.e. a kind of sublimation.”22 But narcissistic libido is still sexual, as is even an impulse inhibited in its aim (affection), and both in moral masochism and in the obsessional neurosis we see that the impulses concerned with the super-ego need not be desexualized; it is plain, further, that there are all degrees of desexualization. So that there must be some further factors at work to account for this interesting change when it occurs.
Two further clues are provided elsewhere by Freud. In the first place he points out that the super-ego is not simply a residuum of the object-choices, but also signifies an energetic reaction-formation against them. “Its relation to the ego is not all comprehended in the exhortation ‘You ought to be like the father’; it also includes the prohibition ‘You may not be like the father’ i.e., you may not do everything he does; many things are his prerogative”.23 In other words, the super-ego consists in the incorporation only of the “moral”, thwarting, and asexual elements of the object. The alloerotic libido of the subject’s id somehow accomplishes the extraordinary feat of substituting this loveless image for the previous love-object; by some magic he manages to love with all the strength of his being just that which he had most reason to hate and fear. It is very possible, however, that from the wreckage of his own desires he is able by means of the identification with the father to save at least in a vicarious way the object-relation which the latter bears to the mother; if so, this vicarious gratification would have to be much deeper in the unconscious than the super-ego.
A second and more valuable clue is afforded by the following considerations. If we enquire into the actual composition of the super-ego, the most obvious constituent to be perceived is sadism,24 usually desexualized. It is presumably to be accounted for as a pregenital regression of the libido that is no longer allowed to be directed towards the love-object; we know that regression is a common sequel to frustration. But this is only the result of a reaction on the part of the endangered ego, which yields to the (castration) threat to its integrity and defends itself by repression of the incestuous impulses. This threat to the primary narcissism must also mobilize the non-sexual ego-instincts, notably hate and fear, and probably all those which I have grouped under the name of “repulsion instincts”.25 The problem that here arises is the relation of the two groups of instincts to each other—roughly speaking, of the hate group26 to the love group. In Das Ich und das Es27 Freud supposes that any previous connection between the two undergoes a process of “de-fusion”. He takes for granted the desexualization of the libidinal impulses as a necessary consequence of the secondary narcissism and suggests that as the result of this desexualization the libido loses its power to bind the aggressive tendencies, which are therefore set free; hence the cruelty of the super-ego. To me at least an alternative hypothesis which he had previously put forward in the Triebe und Triebschicksale essay28 appeals as more likely. In speaking of ambivalence he shows illuminatingly how the ego instincts and sexual instincts mutually influence each other, and how they can form a unity during the pregenital phases of libidinal organization. “When the sexual function is governed by the ego-instincts, as at the stage of the sadistic-anal organization, they impart the qualities of hate to the instinct’s aim as well. . .. This admixture of hate in love is to be traced in part to those preliminary stages of love which have not been wholly outgrown, and in part is based upon reactions of aversion and repudiation on the part of the ego-instincts. . .. In both cases, therefore, the admixture of hate may be traced to the source of the self-preservation instincts. When a love-relationship with a given object is broken off, it is not infrequently succeeded by hate, so that we receive the impression of a transformation of love into hate. This descriptive characterization is amplified by the view that, when this happens, the hate which is motivated by considerations of reality is reinforced by a regression of the love to the sadistic preliminary stage, so that the hate acquires an erotic character and the continuity of a love-relation is ensured”. One may ask whether this does not describe the changes that occur when the super-ego is formed. That would mean a fusion, rather than a de-fusion, of the two groups. And it may be that the secret of the desexualization of the libidinal impulses, perhaps also the preceding regression of them to the anal-sadistic level, will be found in the influence on them of the hate impulses (ego instincts in general). Whether this holds good for the desexualization and sublimation which Freud29 suggests occurs at every identification is, of course, another matter. On the other side the libido would give an erotic colouring to the ego impulses, so that the hate would come to partake of the quality of sadism and fuse with the sadism resulting from libidinal regression.
We may now attempt to describe schematically the changes that ensue on the passing of the Oedipus complex and the replacement of it by the super-ego.
A. Ego Instincts. These “reactive” instincts are all stimulated by the unfriendly situation in the outer world (parents) that leads to the repression of the incestuous wishes. The hatred for the rival, the half of the Oedipus complex which is presently to be resolved by homosexual identification, arises later in time than these wishes.
Fear. The fear of castration constitutes the kernel of the dread which the ego displays in regard to the super-ego,30 and this is evidently a displacement from the father. It is continued later as a sensitiveness to conscience, that is, as a sense of guilt.31
Hate. This is activated against whichever parent is felt to be the obstacle to the love impulses, whether that be the main love-object or not.
B. Sexual Instincts. As was indicated above, the ego defends itself against external danger by repressing the genital impulses directed towards the love-object. Regression to the anal-sadistic level ensues, but the relation of this process to the frustration and to the influence of the ego instincts is not clear. The libido is then re-distributed as follows:32
We thus see that the super-ego arises as a compromise between the desire to love and the desire to be loved. On the one hand it provides an object for the libidinal impulses of the id when the external object is no longer available, whereas on the other hand it represents the renouncing of incest which is the only condition under which the parents’ approval (i.e. affection) can be retained.
* Article citation:
Jones, E. (1926). The origin and structure of the super-ego. International Journal of Psychoanalysis, 7: 303–311.
1. Collected Papers, Vol. II, p. 250.
2. Das Ich und das Es, 1923, S. 40, 60, etc.
3. Das Ich und das Es, 1923, S. 43.
4 Das Ich und das Es, 1923, S. 68.
5. Collected Papers, Vol. II, p. 266.
6. Das Ich und das Es, S. 34.
7. Das Ich und das Es, S. 33, etc.
8. Das Ich und das Es, S. 27, 31.
9. Das Ich und das Es, S. 43, 61, 67.
10. Das Ich und das Es, S. 72.
11. Collected Papers, Vol. II, pp. 251, 253.
12. Collected Papers, Vol. II, pp. 253, 264.
13. Das Ich und das Es, S. 47.
14. Das Ich und das Es, S. 61.
15. Das Ich und das Es, S. 66.
16. Das Ich und das Es, S. 66.
17. Das Ich und das Es, S. 33.
18. Das Ich und das Es, S. 38.
19. Das Ich und das Es, S. 45.
20. Cf. Freud’s remarks on ambivalence in connection with melancholic identification (Collected Papers, Vol. IV, p. 161).
21. Das Ich und das Es, S. 34.
22. Das Ich und das Es, S. 34.
23. Das Ich und das Es, S. 40.
24. The finding is not surprising when one reflects how sadistic and persecutory even ordinary (outwardly directed) morality often is; in the formation of the super-ego we have an example of the “turning round upon the subject”, which Freud described in connection with sadism as one of the vicissitudes of instincts (Collected Papers, Vol. IV, p. 70). Cf. Das Ich und das Es, S. 70, 71.
25. Trans. of the VII. International Congress of Psychology, 1924, p. 231.
26. Freud’s “death instinct”. I find myself unable to operate with this philosophical concept in a purely clinical discussion.
27. S. 71.
28. Collected Papers, Vol. IV, p. 82.
29. Collected Papers, Vol. II, p. 273.
30. Das Ich und das Es, S. 75.
31. Das Ich und das Es, S. 75.
32. It is doubtful if one can apply the term desexualization to the first two of these four groups.