Sigmund FreudThis Page Left Intentionally Blank- 1 – [621203]

Freud, S. (1923). The Ego and the Id. The Standard Edition of theComplete Psychological Works of Sigmund Freud, Volume XIX (1923-1925): The Ego and the Id and Other Works, 1-66The Ego and the Id

Sigmund FreudThis Page Left Intentionally Blank- 1 –
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

Editor's Introduction to "The Ego and the Id"James Strachey(a) German Editions:1923 Das Ich Und Das Es Leipzig, Vienna and Zurich: InternationalerPsycho-analytischer Verlag. Pp. 77.1925 Das Ich Und Das Es G.S., 6, 351-405.1931 Das Ich Und Das Es Theoretische Schriften, 338-91.1940 Das Ich Und Das Es G.W., 13, 237-289.(b) English Translation::The Ego and the Id 1927 London: Hogarth Press and Institute of Psycho-Analysis. Pp. 88. (Tr. Joan Riviere.)The present is a very considerably modified version of the one published in1927.This book appeared in the third week of April, 1923, though it had been inFreud's mind since at least the previous July (Jones, 1957, 104). On September26, 1922, at the Seventh International Psycho-Analytical Congress, which washeld in Berlin and was the last he ever attended, he read a short paper with thetitle ‘Etwas vom Unbewussten [Some Remarks on the Unconscious]’, in whichhe foreshadowed the contents of the book. An abstract of this paper (which wasnever itself published) appeared that autumn in the Int. Zeitschrift Psychoanal.,5 (4), 486, and, although there is no certainty that it was written by Freudhimself, it is worth while recording it:‘Some Remarks on the Unconscious’‘The speaker repeated the familiar history of the development of the concept“‘unconscious” in psycho-analysis. “Unconscious” was in the first instance apurely descriptive term which accordingly included what is temporarily latent.The dynamic view————————————— This page can be read in German in GESAMMELTE WERKE Vol 18S,Page 730 A translation was published in the Int. J. Psycho-Anal. the next year, 4 (3),367. (The date of the reading of the paper is there misprinted ‘Sept. 25’.) It isreprinted here in a slightly modified form.- 3 -1[PEP]1
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

of the process of repression made it necessary, however, to give the unconsciousa systematic sense, so that the unconscious had to be equated with the repressed.What is latent and only temporarily unconscious received the name of“preconscious” and, from the systematic point of view, was brought into closeproximity to the conscious. The double meaning of the term “unconscious”undoubtedly involved disadvantages, though they were of little significance andwere difficult to avoid. It has turned out, however, that it is not practicable toregard the repressed as coinciding with the unconscious and the ego with thepreconscious and conscious. The speaker discussed the two facts which showthat in the ego too there is an unconscious, which behaves dynamically like therepressed unconscious: the two facts of a resistance proceeding from the egoduring analysis and of an unconscious sense of guilt. He announced that in abook which was shortly to appear—The Ego and the Id—he had made anattempt to estimate the influence which these new discoveries must have uponour view of the unconscious.’The Ego and the Id is the last of Freud's major theoretical works. It offers adescription of the mind and its workings which is at first sight new and evenrevolutionary; and indeed all psycho-analytic writings that date from after itspublication bear the unmistakable imprint of its effects—at least in regard totheir terminology. But, in spite of all its fresh insights and fresh syntheses, wecan trace, as so often with Freud's apparent innovations, the seeds of his newideas in earlier, and sometimes in far earlier, writings.The forerunners of the present general picture of the mind had beensuccessively the ‘Project’ of 1895 (Freud, 1950a), the seventh chapter of TheInterpretation of Dreams (1900a) and the metapsychological papers of 1915. Inall of these, the interrelated problems of mental functioning and mental structurewere inevitably considered, though with varying stress upon the two aspects ofthe question. The historical accident that psycho-analysis had its origin inconnection with the study of hysteria led at once to the hypothesis of repression(or, more generally, of defence) as a mental function, and this in turn to atopographical hypothesis—to a picture of the mind as including two portions,one repressed and the other repressing. The quality of ‘consciousness’ wasevidently closely involved in————————————— This page can be read in German in GESAMMELTE WERKE Vol 18S,Page 730- 4 -[PEP]
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

these hypotheses; and it was easy to equate the repressed part of the mind withwhat was ‘unconscious’ and the repressing part with what was ‘conscious’.Freud's earlier pictorial diagrams of the mind, in The Interpretation of Dreams(Standard Ed., 5, 537-41) and in his letter to Fliess of December 6, 1896(Freud, 1950a, Letter 52), were representations of this view of the position.And this apparently simple scheme underlay all of Freud's earlier theoreticalideas: functionally, a repressed force endeavouring to make its way into activitybut held in check by a repressing force, and structurally, an ‘unconscious’opposed by an ‘ego’.Nevertheless, complications soon became manifest. It was quickly seen thatthe word ‘unconscious’ was being used in two senses: the ‘descriptive’ sense(which merely attributed a particular quality to a mental state) and the‘dynamic’ sense (which attributed a particular function to a mental state). Thisdistinction was already stated, though not in these terms, in The Interpretationof Dreams (Standard Ed., 5, 614-15). It was stated much more clearly in theEnglish paper written for the Society for Psychical Research (1912g, StandardEd., 12, 262). But from the first another, more obscure notion was alreadyinvolved (as was plainly shown by the pictorial diagrams)—the notion of‘systems’ in the mind. This implied a topographical or structural division of themind based on something more than function, a division into portions to which itwas possible to attribute a number of differentiating characteristics and methodsof operating. Some such idea was no doubt already implied in the phrase ‘theunconscious’, which appeared very early (e.g. in a footnote to the Studies onHysteria, 1895d, Standard Ed., 2, 76). The concept of a ‘system’ becameexplicit in The Interpretation of Dreams (1900a), Standard Ed., 5, 536-7.From the terms in which it was there introduced, topographical imagery was atonce suggested, though Freud gave a warning against taking this literally. Therewere a number of these ‘systems’ (mnemic, perceptual, and so on) and amongthem ‘the unconscious’ (Standard Ed., 541), which ‘for simplicity's sake’ wasto be designated as ‘the system Ucs.’. In these earlier passages all that wasovertly meant by this unconscious system was the repressed, until we reach thefinal section of The Interpretation of Dreams (Standard Ed., 5, 611 ff.), wheresomething with a much wider scope was indicated. Thereafter the questionremained in abeyance- 5 –
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

until the S.P.R. paper (1912g) already referred to, where (besides the cleardifferentiation between the descriptive and dynamic uses of the term‘unconscious’), in the last sentences of the paper, a third, ‘systematic’, use wasdefined. It may be noted that in this passage (Standard Ed., 12, 266), it was onlyfor this ‘systematic’ unconscious that Freud proposed to use the symbol ‘Ucs.’.All this seems very straightforward, but, oddly enough, the picture was blurredonce more in the metapsychological paper on ‘The Unconscious’ (1915e). InSection II of that paper (Standard Ed., 14, 172 ff.) there were no longer threeuses of the term ‘unconscious’ but only two. The ‘dynamic’ use disappeared,and was presumably subsumed into the ‘systematic’ one, which was still to becalled the ‘Ucs.’, though it now included the repressed. Finally, in Chapter I ofthe present work (as well as in Lecture XXXI of the New IntroductoryLectures, 1933a) Freud reverted to the threefold distinction and classification,though at the end of the chapter he applied the abbreviation ‘Ucs.’, inadvertentlyperhaps, to all three kinds of ‘unconscious’ (p. 18).But the question now arose whether, as applied to a system, the term‘unconscious’ was at all appropriate. In the structural picture of the mind whathad from the first been most clearly differentiated from ‘the unconscious’ hadbeen ‘the ego’. And it now began to appear that the ego itself ought partly to bedescribed as ‘unconscious’. This was pointed out in Beyond the PleasurePrinciple, in a sentence which read in the first edition (1920g): ‘It may be thatmuch of the ego is itself unconscious; only a part of it, probably, is covered bythe term “pre-conscious”.’ In the second edition, a year later, this sentence wasaltered to: ‘It is certain that much of the ego is itself unconscious …; only asmall part of it is covered by the term “preconscious”’. And this discovery andthe grounds for it were stated with still greater insistence in the first chapter ofthe present work.It had thus become apparent that, alike as regards ‘the————————————— The two terms seem to be definitely equated in Beyond the PleasurePrinciple (1920g), Standard Ed., 18, 20. [I.e. not merely in the descriptive but also in the dynamic sense.] Freud had actually already spoken in the opening sentence of his secondpaper on ‘The Neuro-Psychoses of Defence’ (1896b) of the psychicalmechanism of defence as being ‘unconscious’.- 6 -1
23123
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

unconscious’ and as regards ‘the ego’, the criterion of consciousness was nolonger helpful in building up a structural picture of the mind. Freud accordinglyabandoned the use of consciousness in this capacity: ‘being conscious’ washenceforward to be regarded simply as a quality which might or might not beattached to a mental state. The old ‘descriptive’ sense of the term was in fact allthat remained. The new terminology which he now introduced had a highlyclarifying effect and so made further clinical advances possible. But it did not initself involve any fundamental changes in Freud's views on mental structure andfunctioning. Indeed, the three newly presented entities, the id, the ego and thesuper-ego, all had lengthy past histories (two of them under other names) andthese will be worth examining.The term ‘das Es’, as Freud himself explains below (p. 23), was derived inthe first instance from Georg Groddeck, a physician practising at Baden-Baden,who had recently become attached to psycho-analysis and with whose wide-ranging ideas Freud felt much sympathy. Groddeck seems in turn to have derived‘das Es’ from his own teacher, Ernst Schweninger, a well-known Germanphysician of an earlier generation. But, as Freud also points out, the use of theword certainly goes back to Nietzsche. In any case, the term was adapted byFreud to a different and more precise meaning than Groddeck's. It cleared upand in part replaced the ill-defined uses of the earlier terms ‘the unconscious’,‘the Ucs.’ and ‘the systematic unconscious’.The position in regard to ‘das Ich’ is a good deal less clear. The term had ofcourse been in familiar use before the days of Freud; but the precise sense whichhe himself attached to it in his earlier writings is not unambiguous. It seemspossible to detect two main uses: one in which the term distinguishes a person'sself as a whole (including, perhaps, his body) from————————————— There was to begin with a good deal of discussion over the choice of anEnglish equivalent. ‘The id’ was eventually decided upon in preference to ‘theit’, so as to be parallel with the long-established ‘ego’. The symbol ‘Ucs.’ disappears after the present work, except for a singleoccurrence in New Introductory Lectures (1933a), Standard Edition, 22, 72,and a single belated one in Moses and Monotheism (1939a), Essay III, Part 1(E), where oddly enough it is used in the ‘descriptive’ sense. Freud continuedto use the term ‘the unconscious’, though with diminishing frequency, as asynonym for ‘the id’.- 7 -1212
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

other people, and the other in which it denotes a particular part of the mindcharacterized by special attributes and functions. It is in this second sense thatthe term was used in the elaborate account of the ‘ego’ in Freud's early ‘Project’of 1895 (Freud, 1950a, Part I, Section 14); and it is in this same sense that it isused in the anatomy of the mind in The Ego and the Id. But in some of hisintervening works, particularly in connection with narcissism, the ‘ego’ seems tocorrespond rather to the ‘self’. It is not always easy, however, to draw a linebetween these two senses of the word.What is quite certain, however, is that, after the isolated attempt in the‘Project’ of 1895 at a detailed analysis of the structure and functioning of theego, Freud left the subject almost untouched for some fifteen years. His interestwas concentrated on his investigations of the unconscious and its instincts,particularly the sexual ones, and in the part they played in normal and abnormalmental behaviour. The fact that repressive forces played an equally importantpart was, of course, never overlooked and was always insisted on; but thecloser examination of them was left to the future. It was enough for the momentto give them the inclusive name of ‘the ego’.There were two indications of a change, both round about the year 1910. In apaper on psychogenic disturbances of vision (1910i), there comes what seems tobe a first mention of ‘ego-instincts’ (Standard Ed., 11, 214), which combine thefunctions of repression with those of self-preservation. The other and moreimportant development was the hypothesis of narcissism which was firstproposed in 1909 and which led the way to a detailed examination of the egoand its functions in a variety of connections—in the study on Leonardo (1910c),in the Schreber case history (1911c), in the paper on the two principles ofmental functioning (1911b), in the paper on ‘Narcissism’ itself (1914c) and inthe metapsychological paper on ‘The Unconscious’ (1915e). In this last work,however, a————————————— In a few places in the Standard Edition where the sense seemed to demandit, ‘das Ich’ has been translated by ‘the self’. There is a passage inCivilization and its Discontents (1930a), towards the beginning of the fourthparagraph of Chapter I, in which Freud himself explicitly equates ‘das Selbst’and ‘das Ich’. And, in the course of a discussion of the moral responsibility fordreams (1925i), p. 133 below, he makes a clear distinction between the twouses of the German word ‘Ich’.- 8 -1
1
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

further development occurred: what had been described as the ego now becamethe ‘system’ Cs. (Pcs.). It is this system which is the progenitor of the ‘ego’ aswe have it in the new and corrected terminology, from which, as we have seen,the confusing connection with the quality of ‘consciousness’ has been removed.The functions of the system Cs. (Pcs.), as enumerated in ‘The Unconscious’,Standard Ed., 14, 188, include such activities as censorship, reality-testing, andso on, all of which are now assigned to the ‘ego’. There is one particularfunction, however, whose examination was to lead to momentous results—theself-critical faculty. This and the correlated ‘sense of guilt’ attracted Freud'sinterest from early days, chiefly in connection with the obsessional neurosis. Histheory that obsessions are ‘transformed self-reproaches’ for sexual pleasureenjoyed in childhood was fully explained in Section II of his second paper on‘The Neuro-Psychoses of Defence’ (1896b) after being outlined somewhatearlier in his letters to Fliess. That the self-reproaches may be unconscious wasalready implied at this stage, and was stated specifically in the paper on‘Obsessive Actions and Religious Practices’ (1907b), Standard Ed., 9, 123. Itwas only with the concept of narcissism, however, that light could be thrown onthe actual mechanism of these self-reproaches. In Section III of his paper onnarcissism (1914c) Freud began by suggesting that the narcissism of infancy isreplaced in the adult by devotion to an ideal ego set up within himself. He thenput forward the notion that there may be ‘a special psychical agency’ whose taskit is to watch the actual ego and measure it by the ideal ego or ego ideal—heseemed to use the terms indiscriminately (Standard Ed., 14, 95). He attributeda number of functions to this agency, including the normal conscience, thedreamcensorship and certain paranoic delusions. In the paper on ‘Mourning andMelancholia’ (1917e [1915]) he further made this agency responsible forpathological states of mourning (Standard Ed., 14, 247) and insisted moredefinitely that it is something apart from the rest of the ego, and this was madestill more clear in Group Psychology (1921c). It must be noticed, however, thathere the distinction between the ‘ego ideal’ itself and the ‘agency’ concernedwith its enforcement had been dropped: the————————————— These abbreviations (like the ‘Ucs.’) go back to The Interpretation ofDreams (1900a), Standard Ed., 5, 540 n. Actually all of them are already used(in the systematic sense) in the Fliess correspondence (Letter 64 and Draft N)on May 31, 1897 (Freud, 1950a). Some remarks on the ‘synthetic’ function of the ego will be found in NewIntroductory Lectures, Standard Edition, 22, 76 and footnote 3. So too in the Introductory Lectures (1916-17), ibid., 16, 428-9.- 9 -12
3
123
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

‘agency’ was specifically called the ‘ego ideal’ (Standard Ed., 18, 109-10). Itis as an equivalent to the ‘ego ideal’ that ‘das Über-Ich’ makes its firstappearance (p. 28 below), though its aspect as an enforcing or prohibitingagency predominates later. Indeed, after The Ego and the Id and the two or threeshorter works immediately following it, the ‘ego ideal’ disappears almostcompletely as a technical term. It makes a brief re-emergence in a couple ofsentences in the New Introductory Lectures (1933a), Lecture XXXI; but herewe find a return to the original distinction, for ‘an important function’ attributedto the super-ego is to act as ‘the vehicle of the ego ideal by which the egomeasures itself’—almost the exact terms in which the ego ideal was firstintroduced in the paper on narcissism (Standard Ed., 14, 93).But this distinction may seem to be an artificial one when we turn to Freud'saccount of the genesis of the super-ego. This account (in Chapter III) is no doubtthe part of the book second in importance only to the main thesis of the threefolddivision of the mind. The super-ego is there shown to be derived from atransformation of the child's earliest object-cathexes into identifications: it takesthe place of the Oedipus complex. This mechanism (the replacement of anobject-cathexis by an identification and the introjection of the former object) hadbeen first applied by Freud (in his study of Leonardo, 1910c) to the explanationof one type of homosexuality, in which a boy replaces his love for his mother byidentifying himself with her (Standard Ed., 11, 100). He next applied the samenotion to states of depression in ‘Mourning and Melancholia’ (1917e), StandardEd., 14, 249. Further and more elaborate discussions of these various kinds ofidentifications and introjections were pursued in Chapters VII, VIII and XI ofGroup Psychology (1921c), but it was only in the present work that Freudarrived at his final views on the derivation of the super-ego from the child'searliest object-relations.Having once established his new account of the anatomy of the mind, Freudwas in a position to examine its implications, and this he already does in thelater pages of the book—the————————————— Jones (1957, 305 n.) remarks that the term had been used earlier byMünsterberg (1908), though, he adds, it was in a different sense and it isunlikely that Freud had come across the passage.- 10 -1
1
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

relation between the divisions of the mind and the two classes of instincts, andthe interrelations between the divisions of the mind themselves, with specialreference to the sense of guilt. But many of these questions, and in particular thelast one, were to form the subject of other writings which followed in rapidsuccession. See, for instance, ‘The Economic Problem of Masochism’ (1924c),‘The Dissolution of the Oedipus Complex’ (1924d), the two papers on neurosisand psychosis (1924b and 1924e), and the one on the anatomical distinctionbetween the sexes (1925j), all in the present volume, as well as the still moreimportant Inhibitions, Symptoms and Anxiety (1926d), published only a littlelater. Finally, a further long discussion of the super-ego, together with aninteresting examination of the proper use of the terms ‘super-ego’, ‘conscience’,‘sense of guilt’, ‘need for punishment’ and ‘remorse’ will be found in ChaptersVII and VIII of Civilization and its Discontents (1930a).Extracts from the earlier (1927) translation of this work were included inRickman's General Selection from the Works of Sigmund Freud (1937, 245-74).- 11 –
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

The Ego and the Id[Preface]The present discussions are a further development of some trains of thoughtwhich I opened up in Beyond the Pleasure Principle (1920g), and to which, as Iremarked there, my attitude was one of a kind of benevolent curiosity. In thefollowing pages these thoughts are linked to various facts of analyticobservation and an attempt is made to arrive at new conclusions from thisconjunction; in the present work, however, there are no fresh borrowings frombiology, and on that account it stands closer to psycho-analysis than doesBeyond the Pleasure Principle. It is more in the nature of a synthesis than of aspeculation and seems to have had an ambitious aim in view. I am conscious,however, that it does not go beyond the roughest outline and with that limitation Iam perfectly content.In these pages things are touched on which have not yet been the subject ofpsycho-analytic consideration, and it has not been possible to avoid trenchingupon some theories which have been put forward by non-analysts or by formeranalysts on their retreat from analysis. I have elsewhere always been ready toacknowledge what I owe to other workers; but in this instance I feel burdened byno such debt of gratitude. If psycho-analysis has not hitherto shown itsappreciation of certain things, this has never been because it overlooked theirachievement or sought to deny their importance, but because it followed aparticular path, which had not yet led so far. And finally, when it has reachedthem, things have a different look to it from what they have to others.————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 237 [Standard Ed., 18, 59.]- 12 -1
[PEP]1
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

I Consciousness and What is UnconsciousIn this introductory chapter there is nothing new to be said and it will not bepossible to avoid repeating what has often been said before.The division of the psychical into what is conscious and what is unconsciousis the fundamental premiss of psycho-analysis; and it alone makes it possible forpsycho-analysis to understand the pathological processes in mental life, whichare as common as they are important, and to find a place for them in theframework of science. To put it once more, in a different way: psycho-analysiscannot situate the essence of the psychical in consciousness, but is obliged toregard consciousness as a quality of the psychical, which may be present inaddition to other qualities or may be absent.If I could suppose that everyone interested in psychology would read thisbook, I should also be prepared to find that at this point some of my readerswould already stop short and would go no further; for here we have the firstshibboleth of psycho-analysis. To most people who have been educated inphilosophy the idea of anything psychical which is not also conscious is soinconceivable that it seems to them absurd and refutable simply by logic. Ibelieve this is only because they have never studied the relevant phenomena ofhypnosis and dreams, which—quite apart from pathological manifestations—necessitate this view. Their psychology of consciousness is incapable of solvingthe problems of dreams and hypnosis.‘Being conscious’ is in the first place a purely descriptive term, resting onperception of the most immediate and certain————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 239 [‘Bewusst sein’ (in two words) in the original. Similarly in Chapter II of LayAnalysis (1926e), Standard Ed., 20, 197. ‘Bewusstsein’ is the regular Germanword for ‘consciousness’, and printing it in two words emphasizes the fact that‘bewusst’ is in its form a passive participle—‘being conscioused’. TheEnglish ‘conscious’ is capable of an active or a passive use; but in thesediscussions it is always to be taken as passive. Cf. a footnote at the end of theEditor's Note to Freud's metapsychological paper on ‘The Unconscious’,Standard Ed., 14, 165.]- 13 -1[PEP]1
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

character. Experience goes on to show that a psychical element (for instance, anidea) is not as a rule conscious for a protracted length of time. On the contrary, astate of consciousness is characteristically very transitory; an idea that isconscious now is no longer so a moment later, although it can become so againunder certain conditions that are easily brought about. In the interval the ideawas—we do not know what. We can say that it was latent, and by this we meanthat it was capable of becoming conscious at any time. Or, if we say that is wasunconscious, we shall also be giving a correct description of it. Here‘unconscious’ coincides with ‘latent and capable of becoming conscious’. Thephilosophers would no doubt object: ‘No, the term “unconscious” is notapplicable here; so long as the idea was in a state of latency it was not anythingpsychical at all.’ To contradict them at this point would lead to nothing moreprofitable than a verbal dispute.But we have arrived at the term or concept of the unconscious along anotherpath, by considering certain experiences in which mental dynamics play a part.We have found—that is, we have been obliged to assume—that very powerfulmental processes or ideas exist (and here a quantitative or economic factorcomes into question for the first time) which can produce all the effects inmental life that ordinary ideas do (including effects that can in their turn becomeconscious as ideas), though they themselves do not become conscious. It isunnecessary to repeat in detail here what has been explained so often before. Itis enough to say that at this point psycho-analytic theory steps in and asserts thatthe reason why such ideas cannot become conscious is that a certain forceopposes them, that otherwise they could become conscious, and that it wouldthen be apparent how little they differ from other elements which are admittedlypsychical. The fact that in the technique of psycho-analysis a means has beenfound by which the opposing force can be removed and the ideas in questionmade conscious renders this theory irrefutable. The state in which the ideasexisted before being made conscious is called by us repression, and we assertthat the force which instituted the repression and maintains it is perceived asresistance during the work of analysis.————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 240 [See, for instance, ‘A Note on the Unconscious’ (1912g), Standard Ed., 12,262 and 264.]- 14 -1
[PEP]1
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

Thus we obtain our concept of the unconscious from the theory of repression.The repressed is the prototype of the unconscious for us. We see, however, thatwe have two kinds of unconscious—the one which is latent but capable ofbecoming conscious, and the one which is repressed and which is not, in itselfand without more ado, capable of becoming conscious. This piece of insight intopsychical dynamics cannot fail to affect terminology and description. The latent,which is unconscious only descriptively, not in the dynamic sense, we callpreconscious; we restrict the term unconscious to the dynamically unconsciousrepressed; so that now we have three terms, conscious (Cs.), preconscious(Pcs.), and unconscious (Ucs.), whose sense is no longer purely descriptive.The Pcs. is presumably a great deal closer to the Cs. than is the Ucs., and sincewe have called the Ucs. psychical we shall with even less hesitation call thelatent Pcs. psychical. But why do we not rather, instead of this, remain inagreement with the philosophers and, in a consistent way, distinguish the Pcs. aswell as the Ucs. from the conscious psychical? The philosophers would thenpropose that the Pcs. and the Ucs. should be described as two species or stagesof the ‘psychoid’, and harmony would be established. But endless difficulties inexposition would follow; and the one important fact, that these two kinds of‘psychoid’ coincide in almost every other respect with what is admittedlypsychical, would be forced into the background in the interests of a prejudicedating from a period in which these psychoids, or the most important part ofthem, were still unknown.We can now play about comfortably with our three terms, Cs., Pcs., and Ucs.,so long as we do not forget that in the descriptive sense there are two kinds ofunconscious, but in the dynamic sense only one. For purposes of exposition thisdistinction can in some cases be ignored, but in others it is of courseindispensable. At the same time, we have become more or less accustomed tothis ambiguity of the unconscious and have managed pretty well with it. As faras I can see, it is impossible to avoid this ambiguity; the distinction betweenconscious and unconscious is in the last resort a question of perception, whichmust be answered ‘yes’ or ‘no’, and the act of perception itself tells us nothingof the reason why a thing is or————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 241 [Some comments on this sentence will be found in Appendix A (p. 60).]- 15 -1[PEP]1
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

is not perceived. No one has a right to complain because the actual phenomenonexpresses the dynamic factor ambiguously.In the further course of psycho-analytic work, however, even————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 242 This may be compared so far with my ‘Note on the Unconscious in Psycho-Analysis’ (1912g). [Cf. also Sections I and II of the meta-psychological paperon ‘The Unconscious’ (1915e).] A new turn taken by criticisms of theunconscious deserves consideration at this point. Some investigators, who donot refuse to recognize the facts of psycho-analysis but who are unwilling toaccept the unconscious, find a way out of the difficulty in the fact, which noone contests, that in consciousness (regarded as a phenomenon) it is possibleto distinguish a great variety of gradations in intensity or clarity. Just as thereare processes which are very vividly, glaringly, and tangibly conscious, so wealso experience others which are only faintly, hardly even noticeablyconscious; those that are most faintly conscious are, it is argued, the ones towhich psycho-analysis wishes to apply the unsuitable name ‘unconscious’.These too, however (the argument proceeds), are conscious or ‘inconsciousness’, and can be made fully and intensely conscious if sufficientattention is paid to them.In so far as it is possible to influence by arguments the decision of a questionof this kind which depends either on convention or on emotional factors, wemay make the following comments. The reference to gradations of clarity inconsciousness is in no way conclusive and has no more evidential value thansuch analogous statements as: ‘There are so very many gradations inillumination—from the most glaring and dazzling light to the dimmestglimmer—therefore there is no such thing as darkness at all’; or, ‘There arevarying degrees of vitality, therefore there is no such thing as death.’ Suchstatements may in a certain way have a meaning, but for practical purposesthey are worthless. This will be seen if one tries to draw particularconclusions from them, such as, ‘there is therefore no need to strike a light’, or,‘therefore all organisms are immortal’. Further, to include ‘what isunnoticeable’ under the concept of ‘what is conscious’ is simply to play havocwith the one and only piece of direct and certain knowledge that we have aboutthe mind. And after all, a consciousness of which one knows nothing seems tome a good deal more absurd than something mental that is unconscious.Finally, this attempt to equate what is unnoticed with what is unconscious isobviously made without taking into account the dynamic conditions involved,which were the decisive factors in forming the dsycho-analytic view. For itignores two facts: first, that it is exceedingly difficult and requires very greateffort to concentrate enough attention on something unnoticed of this kind; andsecondly, that when this has been achieved the thought which was previouslyunnoticed is not recognized by consciousness, but often seems entirely alienand opposed to it and is promptly disavowed by it. Thus, seeking refuge fromthe unconscious in what is scarcely noticed or unnoticed is after all only aderivative of the preconceived belief which regards the identity of thepsychical and the conscious as settled once and for all.- 16 -1[PEP]1
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

these distinctions have proved to be inadequate and, for practical purposes,insufficient. This has become clear in more ways than one; but the decisiveinstance is as follows. We have formed the idea that in each individual there is acoherent organization of mental processes; and we call this his ego. It is to thisego that consciousness is attached; the ego controls the approaches to motility—that is, to the discharge of excitations into the external world; it is the mentalagency which supervises all its own constituent processes, and which goes tosleep at night, though even then it exercises the censorship on dreams. From thisego proceed the repressions, too, by means of which it is sought to excludecertain trends in the mind not merely from consciousness but also from otherforms of effectiveness and activity. In analysis these trends which have been shutout stand in opposition to the ego, and the analysis is faced with the task ofremoving the resistances which the ego displays against concerning itself withthe repressed. Now we find during analysis that, when we put certain tasksbefore the patient, he gets into difficulties; his associations fail when they shouldbe coming near the repressed. We then tell him that he is dominated by aresistance; but he is quite unaware of the fact, and, even if he guesses from hisunpleasurable feelings that a resistance is now at work in him, he does not knowwhat it is or how to describe it. Since, however, there can be no question butthat this resistance emanates from his ego and belongs to it, we find ourselves inan unforeseen situation. We have come upon something in the ego itself which isalso unconscious, which behaves exactly like the repressed—that is, whichproduces powerful effects without itself being conscious and which requiresspecial work before it can be made conscious. From the point of view ofanalytic practice, the consequence of this discovery is that we land in endlessobscurities and difficulties if we keep to our habitual forms of expression andtry, for instance, to derive neuroses from a conflict between the conscious andthe unconscious. We shall have to substitute for this antithesis another, takenfrom our insight into the structural conditions of the mind—the antithesisbetween the coherent ego and the repressed which is split off from it.For our conception of the unconscious, however, the consequences of ourdiscovery are even more important. Dynamic————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 243 Cf. Beyond the Pleasure Principle (1920g) [Standard Ed., 18, 19].- 17 -1[PEP]1
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

considerations caused us to make our first correction; our insight into thestructure of the mind leads to the second. We recognize that the Ucs. does notcoincide with the repressed; it is still true that all that is repressed is Ucs., butnot all that is Ucs. is repressed. A part of the ego, too—and Heaven knows howimportant a part—may be Ucs., undoubtedly is Ucs. And this Ucs. belonging tothe ego is not latent like the Pcs.; for if it were, it could not be activated withoutbecoming Cs., and the process of making it conscious would not encounter suchgreat difficulties. When we find ourselves thus confronted by the necessity ofpostulating a third Ucs., which is not repressed, we must admit that thecharacteristic of being unconscious begins to lose significance for us. It becomesa quality which can have many meanings, a quality which we are unable tomake, as we should have hoped to do, the basis of far-reaching and inevitableconclusions. Nevertheless we must beware of ignoring this characteristic, forthe property of being conscious or not is in the last resort our one beacon-light inthe darkness of depth-psychology.————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 244 [This had already been stated not only in Beyond the Pleasure Principle(loc. cit.) but earlier, in ‘The Unconscious’ (1915e), Standard Ed., 14, 192-3.Indeed, it was implied in a remark at the beginning of the second paper on ‘TheNeuro-Psychoses of Defence’ (1896b).]- 18 -1
[PEP]1
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

II The Ego and the IdPathological research has directed our interest too exclusively to therepressed. We should like to learn more about the ego, now that we know that it,too, can be unconscious in the proper sense of the word. Hitherto the only guidewe have had during our investigations has been the distinguishing mark of beingconscious or unconscious; we have finally come to see how ambiguous this canbe.Now all our knowledge is invariably bound up with consciousness. We cancome to know even the Ucs. only by making it conscious. But stop, how is thatpossible? What does it mean when we say ‘making something conscious’? Howcan that come about?We already know the point from which we have to start in this connection.We have said that consciousness is the surface of the mental apparatus; that is,we have ascribed it as a function to a system which is spatially the first onereached from the external world—and spatially not only in the functional sensebut, on this occasion, also in the sense of anatomical dissection. Ourinvestigations too must take this perceiving surface as a starting-point.All perceptions which are received from without (sense-perceptions) andfrom within—what we call sensations and feelings—are Cs. from the start. Butwhat about those internal processes which we may—roughly and inexactly—sum up under the name of thought-processes? They represent displacements ofmental energy which are effected somewhere in the interior of the apparatus asthis energy proceeds on its way towards action. Do they advance to the surface,which causes consciousness to be generated? Or does consciousness make itsway to them? This is clearly one of the difficulties that arise when one begins totake the spatial or ‘topographical’ idea of mental life seriously. Both thesepossibilities are equally unimaginable; there must be a third alternative.————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 246 Beyond the Pleasure Principle [Standard Ed., 18, 26]. [This had been discussed at greater length in the second section of ‘TheUnconscious’ (1915e), Standard Ed., 18, 173-6.]- 19 -1
2[PEP]12
1Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

I have already, in another place, suggested that the real difference between aUcs. and a Pcs. idea (thought) consists in this: that the former is carried out onsome material which remains unknown, whereas the latter (the Pcs.) is inaddition brought into connection with word-presentations. This is the firstattempt to indicate distinguishing marks for the two systems, the Pcs. and theUcs., other than their relation to consciousness. The question, ‘How does a thingbecome conscious?’ would thus be more advantageously stated: ‘How does athing become preconscious?’ And the answer would be: ‘Through becomingconnected with the word-presentations corresponding to it.’These word-presentations are residues of memories; they were at one timeperceptions, and like all mnemic residues they can become conscious again.Before we concern ourselves further with their nature, it dawns upon us like anew discovery that only something which has once been a Cs. perception canbecome conscious, and that anything arising from within (apart from feelings)that seeks to become conscious must try to transform itself into externalperceptions: this becomes possible by means of memory-traces.We think of the mnemic residues as being contained in systems which aredirectly adjacent to the system Pcpt.-Cs., so that the cathexes of those residuescan readily extend from within on to the elements of the latter system. Weimmediately think here of hallucinations, and of the fact that the most vividmemory is always distinguishable both from a hallucination and from an externalperception; but it will also occur to us at once that when a memory is revivedthe cathexis remains in the mnemic system, whereas a hallucination, which is notdistinguishable from a perception, can arise when the cathexis does not merelyspread over from the memory-trace on to the Pcpt. element, but passes over to itentirely.Verbal residues are derived primarily from auditory perceptions, so that thesystem Pcs. has, as it were, a special————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 247 ‘The Unconscious’ [Standard Ed., 201 ff.]. [Cf. Chapter VII (B) of The Interpretation of Dreams (1900a), StandardEd., 5, 538.] [This view had been expressed by Breuer in his theoretical contribution toStudies on Hysteria (1895d), Standard Ed., 2, 188.] [Freud had arrived at this conclusion in his monograph on aphasia (1891b)on the basis of pathological findings (Standard Ed., 92-4). The point isrepresented in the diagram reproduced from that work in Appendix C to thepaper on ‘The Unconscious’, Standard Ed., 14, 214.]- 20 -1
234[PEP]1234
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

sensory source. The visual components of word-presentations are secondary,acquired through reading, and may to begin with be left on one side; so may themotor images of words, which, except with deaf-mutes, play the part of auxiliaryindications. In essence a word is after all the mnemic residue of a word that hasbeen heard.We must not be led, in the interests of simplification perhaps, to forget theimportance of optical mnemic residues, when they are of things, or to deny thatit is possible for thought-processes to become conscious through a reversion tovisual residues, and that in many people this seems to be the favoured method.The study of dreams and of preconscious phantasies as shown in Varendonck'sobservations can give us an idea of the special character of this visual thinking.We learn that what becomes conscious in it is as a rule only the concretesubject-matter of the thought, and that the relations between the various elementsof this subject-matter, which is what specially characterizes thoughts, cannot begiven visual expression. Thinking in pictures is, therefore, only a veryincomplete form of becoming conscious. In some way, too, it stands nearer tounconscious processes than does thinking in words, and it is unquestionablyolder than the latter both ontogenetically and phylogenetically.To return to our argument: if, therefore, this is the way in which somethingthat is in itself unconscious becomes preconscious, the question how we makesomething that is repressed (pre)conscious would be answered as follows. It isdone by supplying Pcs. intermediate links through the work of analysis.Consciousness remains where it is, therefore; but, on the other hand, the Ucs.does not rise into the Cs.Whereas the relation of external perceptions to the ego is quite perspicuous,that of internal perceptions to the ego requires special investigation. It givesrise once more to a doubt whether we are really right in referring the whole ofconsciousness to the single superficial system Pcpt.-Cs.Internal perceptions yield sensations of processes arising in the most diverseand certainly also in the deepest strata of the————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 248 [Cf. Varendonck (1921), a book to which Freud contributed an introduction(1921b).]- 21 -1
[PEP]1
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

mental apparatus. Very little is known about these sensations and feelings; thosebelonging to the pleasure-unpleasure series may still be regarded as the bestexamples of them. They are more primordial, more elementary, than perceptionsarising externally and they can come about even when consciousness is clouded.I have elsewhere expressed my views about their greater economicsignificance and the metapsychological reasons for this. These sensations aremultilocular, like external perceptions; they may come from different placessimultaneously and may thus have different or even opposite qualities.Sensations of a pleasurable nature have not anything inherently impellingabout them, whereas unpleasurable ones have it in the highest degree. The latterimpel towards change, towards discharge, and that is why we interpretunpleasure as implying a heightening and pleasure a lowering of energiccathexis. Let us call what becomes conscious as pleasure and unpleasure aquantitative and qualitative ‘something’ in the course of mental events; thequestion then is whether this ‘something’ can become conscious in the placewhere it is, or whether it must first be transmitted to the system Pcpt.Clinical experience decides for the latter. It shows us that this ‘something’behaves like a repressed impulse. It can exert driving force without the egonoticing the compulsion. Not until there is resistance to the compulsion, a hold-up in the discharge-reaction, does the ‘something’ at once become conscious asunpleasure. In the same way that tensions arising from physical needs can remainunconscious, so also can pain—a thing intermediate between external andinternal perception, which behaves like an internal perception even when itssource is in the external world. It remains true, therefore, that sensations andfeelings, too, only become conscious through reaching the system Pcpt.; if theway forward is barred, they do not come into being as sensations, although the‘something’ that corresponds to them in the course of excitation is the same as ifthey did. We then come to speak, in a condensed and not entirely correct manner,of ‘unconscious feelings’, keeping up an analogy with unconscious ideas whichis not altogether justifiable. Actually the difference is that, whereas with Ucs.ideas connecting links must be created before they can be————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 249 [Beyond the Pleasure Principle (1920g), Standard Ed., 18, 29.] [Standard Ed., 8.]- 22 -12
[PEP]12
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

brought into the Cs., with feelings, which are themselves transmitted directly,this does not occur. In other words: the distinction between Cs. and Pcs. has nomeaning where feelings are concerned; the Pcs. here drops out—and feelingsare either conscious or unconscious. Even when they are attached to word-presentations, their becoming conscious is not due to that circumstance, but theybecome so directly.The part played by word-presentations now becomes perfectly clear. By theirinterposition internal thought-processes are made into perceptions. It is like ademonstration of the theorem that all knowledge has its origin in externalperception. When a hypercathexis of the process of thinking takes place, thoughtsare actually perceived—as if they came from without—and are consequentlyheld to be true.After this clarifying of the relations between external and internal perceptionand the superficial system Pcpt.-Cs., we can go on to work out our idea of theego. It starts out, as we see, from the system Pcpt., which is its nucleus, andbegins by embracing the Pcs., which is adjacent to the mnemic residues. But, aswe have learnt, the ego is also unconscious.Now I think we shall gain a great deal by following the suggestion of a writerwho, from personal motives, vainly asserts that he has nothing to do with therigours of pure science. I am speaking of Georg Groddeck, who is never tired ofinsisting that what we call our ego behaves essentially passively in life, and that,as he expresses it, we are ‘lived’ by unknown and uncontrollable forces. Wehave all had impressions of the same kind, even though they may not haveoverwhelmed us to the exclusion of all others, and we need feel no hesitation infinding a place for Groddeck's discovery in the structure of science. I propose totake it into account by calling the entity which starts out from the system Pcpt.and begins by being Pcs. the ‘ego’, and by following Groddeck in calling theother part of the mind, into which this entity extends and which behaves asthough it were Ucs., the ‘id’.————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 250 [Cf. Section III of ‘The Unconscious’ (1915e), Standard Ed., 14, 177-8.] Groddeck (1923). [See Editor's Introduction, p. 7.]—Groddeck himself no doubt followed theexample of Nietzsche, who habitually used this grammatical term for whateverin our nature is impersonal and, so to speak, subject to natural law.- 23 -1
23[PEP]123
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

We shall soon see whether we can derive any advantage from this view forpurposes either of description or of understanding. We shall now look upon anindividual as a psychical id, unknown and unconscious, upon whose surfacerests the ego, developed from its nucleus the Pcpt. system. If we make an effortto represent this pictorially, we may add that the ego does not completelyenvelop the id, but only does so to the extent to which the system Pcpt. forms its[the ego's] surface, more or less as the germinal disc rests upon the ovum. Theego is not sharply separated from the id; its lower portion merges into it.But the repressed merges into the id as well, and is merely a part of it. Therepressed is only cut off sharply from the ego by the resistances of repression; itcan communicate with the ego through the id. We at once realize that almost allthe lines of demarcation we have drawn at the instigation of pathology relateonly to the superficial strata of the mental apparatus—the only ones known to us.The state of things which we have been describing can be representeddiagrammatically (Fig. 1); though it must be remarked that the form chosen hasno pretensions to any special applicability, but is merely intended to serve forpurposes of exposition.FIG. 1
————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 251 [Compare the slightly different diagram near the end of Lecture XXXI of theNew Introductory Lectures (1933a). The entirely different one in TheInterpretation of Dreams (1900a), Standard Ed., 5, 541, and its predecessorin a letter to Fliess of December 6, 1896 (Freud, 1950a Letter 52), areconcerned with function as well as structure.]- 24 -1
[PEP]1
1Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

We might add, perhaps, that the ego wears a ‘cap of hearing’—on one sideonly, as we learn from cerebral anatomy. It might be said to wear it awry.It is easy to see that the ego is that part of the id which has been modified bythe direct influence of the external world through the medium of the Pcpt.-Cs.; ina sense it is an extension of the surface-differentiation. Moreover, the ego seeksto bring the influence of the external world to bear upon the id and itstendencies, and endeavours to substitute the reality principle for the pleasureprinciple which reigns unrestrictedly in the id. For the ego, perception plays thepart which in the id falls to instinct. The ego represents what may be calledreason and common sense, in contrast to the id, which contains the passions. Allthis falls into line with popular distinctions which we are all familiar with; atthe same time, however, it is only to be regarded as holding good on the averageor ‘ideally’.The functional importance of the ego is manifested in the fact that normallycontrol over the approaches to motility devolves upon it. Thus in its relation tothe id it is like a man on horseback, who has to hold in check the superiorstrength of the horse; with this difference, that the rider tries to do so with hisown strength while the ego uses borrowed forces. The analogy may be carried alittle further. Often a rider, if he is not to be parted from his horse, is obliged toguide it where it wants to go; so in the same way the ego is in the habit oftransforming the id's will into action as if it were its own.Another factor, besides the influence of the system Pcpt., seems to haveplayed a part in bringing about the formation of the ego and its differentiationfrom the id. A person's own body, and above all its surface, is a place fromwhich both external and internal perceptions may spring. It is seen like any otherobject, but to the touch it yields two kinds of sensations, one of which may beequivalent to an internal perception. Psychophysiology has fully discussed themanner in which a person's own body attains its special position among otherobjects in the world of perception. Pain, too, seems to play a part in the process,and the way in which we gain new knowledge of our organs————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 252 [‘Hörkappe.’ I.e. the auditory lobe. Cf. footnote , p. 20 above.] [This analogy appears as an association to one of Freud's dreams in TheInterpretation of Dreams, Standard Ed., 4, 231.]- 25 -1
2
[PEP]142
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

during painful illnesses is perhaps a model of the way by which in general wearrive at the idea of our body.The ego is first and foremost a bodily ego; it is not merely a surface entity,but is itself the projection of a surface. If we wish to find an anatomicalanalogy for it we can best identify it with the ‘cortical homunculus’ of theanatomists, which stands on its head in the cortex, sticks up its heels, facesbackwards and, as we know, has its speech-area on the left-hand side.The relation of the ego to consciousness has been entered into repeatedly; yetthere are some important facts in this connection which remain to be describedhere. Accustomed as we are to taking our social or ethical scale of values alongwith us wherever we go, we feel no surprise at hearing that the scene of theactivities of the lower passions is in the unconscious; we expect, moreover, thatthe higher any mental function ranks in our scale of values the more easily it willfind access to consciousness assured to it. Here, however, psycho-analyticexperience disappoints us. On the one hand, we have evidence that even subtleand difficult intellectual operations which ordinarily require strenuous reflectioncan equally be carried out preconsciously and without coming intoconsciousness. Instances of this are quite incontestable; they may occur, forexample, during the state of sleep, as is shown when someone finds,immediately after waking, that he knows the solution to a difficult mathematicalor other problem with which he had been wrestling in vain the day before.There is another phenomenon, however, which is far stranger. In our analyseswe discover that there are people in whom the faculties of self-criticism andconscience—mental activities, that is, that rank as extremely high ones—areunconscious and unconsciously produce effects of the greatest importance; theexample of resistance remaining unconscious during analysis is————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 253 [I.e. the ego is ultimately derived from bodily sensations, chiefly from thosespringing from the surface of the body. It may thus be regarded as a mentalprojection of the surface of the body, besides, as we have seen above,representing the superficies of the mental apparatus.—This footnote firstappeared in the English translation of 1927, in which it was described ashaving been authorized by Freud. It does not appear in the German editions.] I was quite recently told an instance of this which was, in fact, brought up asan objection against my description of the ‘dream-work’. [Cf. TheInterpretation of Dreams, Standard Ed., 4, 64, and 5, 564.]- 26 -1
2[PEP]12
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

therefore by no means unique. But this new discovery, which compels us, inspite of our better critical judgement, to speak of an ‘unconscious sense ofguilt’, bewilders us far more than the other and sets us fresh problems,especially when we gradually come to see that in a great number of neuroses anunconscious sense of guilt of this kind plays a decisive economic part and putsthe most powerful obstacles in the way of recovery. If we come back oncemore to our scale of values, we shall have to say that not only what is lowest butalso what is highest in the ego can be unconscious. It is as if we were thussupplied with a proof of what we have just asserted of the conscious ego: that itis first and foremost a body-ego.————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 254 [This phrase had already appeared in Freud's paper on ‘Obsessive Actionsand Religious Practices’ (1907b), Standard Ed., 9, 123. The notion was,however, foreshadowed much earlier, in Section II of the first paper on ‘TheNeuro-Psychoses of Defence’ (1894a).] [This is further discussed below, p. 49 ff.]- 27 -12[PEP]12
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

III The Ego and the Super-Ego (Ego Ideal)If the ego were merely the part of the id modified by the influence of theperceptual system, the representative in the mind of the real external world, weshould have a simple state of things to deal with. But there is a furthercomplication.The considerations that led us to assume the existence of a grade in the ego, adifferentiation within the ego, which may be called the ‘ego ideal’ or ‘super-ego’, have been stated elsewhere. They still hold good. The fact that this partof the ego is less firmly connected with consciousness is the novelty which callsfor explanation.At this point we must widen our range a little. We succeeded in explainingthe painful disorder of melancholia by supposing that [in those suffering from it]an object which was lost has been set up again inside the ego—that is, that anobject-cathexis has been replaced by an identification. At that time, however,we did not appreciate the full significance of this process and did not know howcommon and how typical it is. Since then we have come to understand that thiskind of substitution has a great share in determining the form taken by the egoand that it makes an essential contribution towards building up what is called its‘character’.————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 256 [See Editor's Introduction, pp. 9-10.] Cf. ‘On Narcissism: an Introduction’(1914c), and Group Psychology and the Analysis of the Ego (1921c). Except that I seem to have been mistaken in ascribing the function of ‘reality-testing’ to this super-ego—a point which needs correction. [See GroupPsychology (1921c), Standard Ed., 18, 114 and n. 2, and the Editor's Note tothe metapsychological paper on dreams (1917d), 14, 220.] It would fit inperfectly with the relations of the ego to the world of perception if reality-testing remained a task of the ego itself. Some earlier suggestions about a‘nucleus of the ego’, never very definitely formulated, also require to be putright, since the system Pcpt.-Cs. alone can be regarded as the nucleus of theego. [In Beyond the Pleasure Principle (1920g) Freud had spoken of theunconscious part of the ego as its nucleus (Standard Ed., 18, 19); and in hislater paper on ‘Humour’ (1927d) he referred to the super-ego as the nucleus ofthe ego.] ‘Mourning and Melancholia’ (1917e) [Standard Ed., 14, 249]. [Some references to other passages in which Freud has discussed character-formation will be found in an Editor's footnote at the end of the paper on‘Character and Anal Erotism’ (1908b), Standard Ed., 9, 175.]- 28 -1234[PEP]12
34
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

At the very beginning, in the individual's primitive oral phase,object-cathexis and identification are no doubt indistinguishable from eachother. We can only suppose that later on object-cathexes proceed from the id,which feels erotic trends as needs. The ego, which to begin with is still feeble,becomes aware of the object-cathexes, and either acquiesces in them or tries tofend them off by the process of repression.When it happens that a person has to give up a sexual object, there quite oftenensues an alteration of his ego which can only be described as a setting up of theobject inside the ego, as it occurs in melancholia; the exact nature of thissubstitution is as yet unknown to us. It may be that by this introjection, which is akind of regression to the mechanism of the oral phase, the ego makes it easier forthe object to be given up or renders that process possible. It may be that thisidentification is the sole condition under which the id can give up its objects. Atany rate the process, especially in the early phases of development, is a veryfrequent one, and it makes it possible to suppose that the character of the ego is aprecipitate of abandoned object-cathexes and that it contains the history of thoseobject-choices. It must, of course, be admitted from the outset that there arevarying degrees of capacity for resistance, which decide the extent to which aperson's character fends off or accepts the influences of the history of his eroticobject-choices. In women who have had many experiences in love there seemsto be no difficulty in finding vestiges of their object-cathexes in the traits of theircharacter. We must also take into consideration cases of simultaneousobject-cathexis and identification—cases, that————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 257 [Cf. Chapter VII of Group Psychology (1921c), Standard Ed., 18, 105.] An interesting parallel to the replacement of object-choice by identificationis to be found in the belief of primitive peoples, and in the prohibitions basedupon it, that the attributes of animals which are incorporated as nourishmentpersist as part of the character of those who eat them. As is well known, thisbelief is one of the roots of cannibalism and its effects have continued throughthe series of usages of the totem meal down to Holy Communion. [Cf. Totemand Taboo (1912-13), Standard Ed., 13, 82, 142, 154-5, etc.] Theconsequences ascribed by this belief to oral mastery of the object do in factfollow in the case of the later sexual object-choice.- 29 -12
[PEP]12
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

is, in which the alteration in character occurs before the object has been givenup. In such cases the alteration in character has been able to survive the object-relation and in a certain sense to conserve it.From another point of view it may be said that this transformation of an eroticobject-choice into an alteration of the ego is also a method by which the ego canobtain control over the id and deepen its relations with it—at the cost, it is true,of acquiescing to a large extent in the id's experiences. When the ego assumesthe features of the object, it is forcing itself, so to speak, upon the id as alove-object and is trying to make good the id's loss by saying: ‘Look, you canlove me too—I am so like the object.’The transformation of object-libido into narcissistic libido which thus takesplace obviously implies an abandonment of sexual aims, a desexualization—akind of sublimation, therefore. Indeed, the question arises, and deserves carefulconsideration, whether this is not the universal road to sublimation, whether allsublimation does not take place through the mediation of the ego, which beginsby changing sexual object-libido into narcissistic libido and then, perhaps, goeson to give it another aim. We shall later on have to consider whether otherinstinctual vicissitudes may not also result from this transformation, whether, forinstance, it may not bring about a defusion of the various instincts that are fusedtogether.Although it is a digression from our aim, we cannot avoid giving our attentionfor a moment longer to the ego's object-identifications. If they obtain the upperhand and become too numerous, unduly powerful and incompatible with oneanother, a pathological outcome will not be far off. It may come to a disruptionof the ego in consequence of the different identifications becoming cut off fromone another by resistances; perhaps————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 258 Now that we have distinguished between the ego and the id, we mustrecognize the id as the great reservoir of libido indicated in my paper onnarcissism (1914c) [Standard Ed., 14, 75]. The libido which flows into theego owing to the identifications described above brings about its ‘secondarynarcissism’. [The point is elaborated below on p. 46.] [Freud returns to the subject of this paragraph below, on pp. 45 and 54. Theconcept of the fusion and defusion of instincts is explained on pp. 41-2. Theterms had been introduced already in an encyclopaedia article (1923a),Standard Ed., 18, 258.]- 30 -12[PEP]12
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

the secret of the cases of what is described as ‘multiple personality’ is that thedifferent identifications seize hold of consciousness in turn. Even when things donot go so far as this, there remains the question of conflicts between the variousidentifications into which the ego comes apart, conflicts which cannot after allbe described as entirely pathological.But, whatever the character's later capacity for resisting the influences ofabandoned object-cathexes may turn out to be, the effects of the firstidentifications made in earliest childhood will be general and lasting. This leadsus back to the origin of the ego ideal; for behind it there lies hidden anindividual's first and most important identification, his identification with thefather in his own personal prehistory. This is apparently not in the first instancethe consequence or outcome of an object-cathexis; it is a direct and immediateidentification and takes place earlier than any object-cathexis. But the object-choices belonging to the first sexual period and relating to the father and motherseem normally to find their outcome in an identification of this kind, and wouldthus reinforce the primary one.The whole subject, however, is so complicated that it will be necessary to gointo it in greater detail. The intricacy of the problem is due to two factors: thetriangular character of the Oedipus situation and the constitutional bisexuality ofeach individual.In its simplified form the case of a male child may be described as follows.At a very early age the little boy develops an object-cathexis for his mother,which originally related to the mother's breast and is the prototype of an object-choice on the anaclitic model; the boy deals with his father by identifyinghimself with him. For a time these two relationships proceed————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 259 Perhaps it would be safer to say ‘with the parents’; for before a child hasarrived at definite knowledge of the difference between the sexes, the lack of apenis, it does not distinguish in value between its father and its mother. Irecently came across the instance of a young married woman whose storyshowed that, after noticing the lack of a penis in herself, she had supposed it tobe absent not in all women, but only in those whom she regarded as inferior,and had still supposed that her mother possessed one. [Cf. a footnote to ‘TheInfantile Genital Organization’ (1923e), p. 145 below.]—In order to simplifymy presentation I shall discuss only identification with the father. [See the beginning of Chapter VII of Group Psychology (1921c), StandardEd., 18, 105.] [See the paper on narcissism (1914c), Standard Ed., 14, 87 ff.]- 31 -12
3[PEP]1
23
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

side by side, until the boy's sexual wishes in regard to his mother become moreintense and his father is perceived as an obstacle to them; from this the Oedipuscomplex originates. His identification with his father then takes on a hostilecolouring and changes into a wish to get rid of his father in order to take hisplace with his mother. Henceforward his relation to his father is ambivalent; itseems as if the ambivalence inherent in the identification from the beginning hadbecome manifest. An ambivalent attitude to his father and an object-relation of asolely affectionate kind to his mother make up the content of the simple positiveOedipus complex in a boy.Along with the demolition of the Oedipus complex, the boy's object-cathexisof his mother must be given up. Its place may be filled by one of two things:either an identification with his mother or an intensification of his identificationwith his father. We are accustomed to regard the latter outcome as the morenormal; it permits the affectionate relation to the mother to be in a measureretained. In this way the dissolution of the Oedipus complex would consolidatethe masculinity in a boy's character. In a precisely analogous way, the outcomeof the Oedipus attitude in a little girl may be an intensification of heridentification with her mother (or the setting up of such an identification for thefirst time)—a result which will fix the child's feminine character.These identifications are not what we should have expected [from theprevious account (p. 29)], since they do not introduce the abandoned object intothe ego; but this alternative outcome may also occur, and is easier to observe ingirls than in boys. Analysis very often shows that a little girl, after she has had torelinquish her father as a love-object, will bring her masculinity intoprominence and identify herself with her father (that is, with the object whichhas been lost), instead of with her mother. This will clearly depend on whetherthe masculinity in her disposition—whatever that may consist in—is strongenough.————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 260 Cf. Group Psychology (1921c), loc. cit. [Cf. the paper bearing this title (1924d) in which Freud discussed thequestion more fully. (P. 173 below.)] [The idea that the outcome of the Oedipus complex was ‘preciselyanalogous’ in girls and boys was abandoned by Freud not long after this. See‘Some Psychical Consequences of the Anatomical Distinction between theSexes’ (1925j), p. 248 below.]- 32 -1
23
[PEP]123
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

It would appear, therefore, that in both sexes the relative strength of themasculine and feminine sexual dispositions is what determines whether theoutcome of the Oedipus situation shall be an identification with the father orwith the mother. This is one of the ways in which bisexuality takes a hand in thesubsequent vicissitudes of the Oedipus complex. The other way is even moreimportant. For one gets an impression that the simple Oedipus complex is by nomeans its commonest form, but rather represents a simplification orschematization which, to be sure, is often enough justified for practical purposesCloser study usually discloses the more complete Oedipus complex, which istwofold, positive and negative, and is due to the bisexuality originally present inchildren: that is to say, a boy has not merely an ambivalent attitude towards hisfather and an affectionate object-choice towards his mother, but at the same timehe also behaves like a girl and displays an affectionate feminine attitude to hisfather and a corresponding jealousy and hostility towards his mother. It is thiscomplicating element introduced by bisexuality that makes it so difficult toobtain a clear view of the facts in connection with the earliest object-choicesand identifications, and still more difficult to describe them intelligibly. It mayeven be that the ambivalence displayed in the relations to the parents should beattributed entirely to bisexuality and that it is not, as I have represented above,developed out of identification in consequence of rivalry.In my opinion it is advisable in general, and quite especially where neuroticsare concerned, to assume the existence of the complete Oedipus complex.Analytic experience then shows that in a number of cases one or the otherconstituent disappears, except for barely distinguishable traces; so that the————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 261 [Freud's belief in the importance of bisexuality went back a very long way.In the first edition of the Three Essays (1905d), for instance, he wrote:‘Without taking bisexuality into account I think it would scarcely be possible toarrive at an understanding of the sexual manifestations that are actually to beobserved in men and women.’ (Standard Ed., 7, 220.) But still earlier we finda passage in a letter to Fliess (who influenced him greatly on this subject)which seems almost to foreshadow the present paragraph (Freud, 1950a, Letter113, of August 1, 1899): ‘Bisexuality! I am sure you are right about it. And Iam accustoming myself to regarding every sexual act as an event between fourindividuals.’]- 33 -1[PEP]1
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

result is a series with the normal positive Oedipus complex at one end and theinverted negative one at the other, while its intermediate members exhibit thecomplete form with one or other of its two components preponderating. At thedissolution of the Oedipus complex the four trends of which it consists willgroup themselves in such a way as to produce a father-identification and amother-identification. The father-identification will preserve the object-relationto the mother which belonged to the positive complex and will at the same timereplace the object-relation to the father which belonged to the inverted complex:and the same will be true, mutatis mutandis, of the mother-identification. Therelative intensity of the two identifications in any individual will reflect thepreponderance in him of one or other of the two sexual dispositions.The broad general outcome of the sexual phase dominated by the Oedipuscomplex may, therefore, be taken to be the forming of a precipitate in the ego,consisting of these two identifications in some way united with each other.This modification of the ego retains its special position; it confronts the othercontents of the ego as an ego ideal or super-ego.The super-ego is, however, not simply a residue of the earliest object-choices of the id; it also represents an energetic reaction-formation against thosechoices. Its relation to the ego is not exhausted by the precept: ‘You ought to belike this (like your father).’ It also comprises the prohibition: ‘You may not belike this (like your father)—that is, you may not do all that he does; some thingsare his prerogative.’ This double aspect of the ego ideal derives from the factthat the ego ideal had the task of repressing the Oedipus complex; indeed, it is tothat revolutionary event that it owes its existence. Clearly the repression of theOedipus complex was no easy task. The child's parents, and especially hisfather, were perceived as the obstacle to a realization of his Oedipus wishes; sohis infantile ego fortified itself for the carrying out of the repression by erectingthis same obstacle within itself. It borrowed strength to do this, so to speak,from the father, and this loan was an extraordinarily momentous act. The super-ego retains the character of the father, while the more powerful the Oedipuscomplex was and the more rapidly it succumbed to repression (under theinfluence of authority, religious teaching, schooling and reading),————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 262- 34 -[PEP]
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

the stricter will be the domination of the super-ego over the ego later on—in theform of conscience or perhaps of an unconscious sense of guilt. I shall presently[p. 48] bring forward a suggestion about the source of its power to dominate inthis way —the source, that is, of its compulsive character which manifests itselfin the form of a categorical imperative.If we consider once more the origin of the super-ego as we have described it,we shall recognize that it is the outcome of two highly important factors, one of abiological and the other of a historical nature: namely, the lengthy duration inman of his childhood helplessness and dependence, and the fact of his Oedipuscomplex, the repression of which we have shown to be connected with theinterruption of libidinal development by the latency period and so with thediphasic onset of man's sexual life. According to one psycho-analytichypothesis, the last-mentioned phenomenon, which seems to be peculiar toman, is a heritage of the cultural development necessitated by the glacial epoch.We see, then, that the differentiation of the super-ego from the ego is no matter ofchance; it represents the most important characteristics of the development bothof the individual and of the species; indeed, by giving permanent expression tothe influence of the parents it perpetuates the existence of the factors to which itowes its origin.Psycho-analysis has been reproached time after time with ignoring the higher,moral, supra-personal side of human nature. The reproach is doubly unjust, bothhistorically and methodologically. For, in the first place, we have from the verybeginning attributed the function of instigating repression to the moral andaesthetic trends in the ego, and secondly, there has been a general refusal torecognize that psycho-analytic————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 263 [In the German editions this sentence reads as follows: ‘If we consider oncemore the origin of the super-ego as we have described it, we shall recognizethat it is the outcome of two highly important biological factors: namely, thelengthy duration in man of his childhood helplessness and dependence, and thefact of his Oedipus complex, which we have traced back to the interruption oflibidinal development by the latency period and so to the diphasic origin ofman's sexual life.’ The slightly different version given in the text above wasinserted by Freud's express orders in the English translation in 1927. For somereason the emendations were not included in the later German editions.] [The idea was put forward by Ferenczi (1913c). Freud seems to accept itrather more definitely near the end of Chapter X of Inhibitions, Symptoms andAnxiety (1926d), Standard Ed., 20, 155.]- 35 -12
[PEP]1
2
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

research could not, like a philosophical system, produce a complete and ready-made theoretical structure, but had to find its way step by step along the pathtowards understanding the intricacies of the mind by making an analyticdissection of both normal and abnormal phenomena. So long as we had toconcern ourselves with the study of what is repressed in mental life, there wasno need for us to share in any agitated apprehensions as to the whereabouts ofthe higher side of man. But now that we have embarked upon the analysis of theego we can give an answer to all those whose moral sense has been shocked andwho have complained that there must surely be a higher nature in man: ‘Verytrue,’ we can say, ‘and here we have that higher nature, in this ego ideal orsuper-ego, the representative of our relation to our parents. When we were littlechildren we knew these higher natures, we admired them and feared them; andlater we took them into ourselves.’The ego ideal is therefore the heir of the Oedipus complex, and thus it is alsothe expression of the most powerful impulses and most important libidinalvicissitudes of the id. By setting up this ego ideal, the ego has mastered theOedipus complex and at the same time placed itself in subjection to the id.Whereas the ego is essentially the representative of the external world, ofreality, the super-ego stands in contrast to it as the representative of the internalworld, of the id. Conflicts between the ego and the ideal will, as we are nowprepared to find, ultimately reflect the contrast between what is real and what ispsychical, between the external world and the internal world.Through the forming of the ideal, what biology and the vicissitudes of thehuman species have created in the id and left behind in it is taken over by theego and re-experienced in relation to itself as an individual. Owing to the way inwhich the ego ideal is formed, it has the most abundant links with thephylogenetic acquisition of each individual—his archaic heritage. What hasbelonged to the lowest part of the mental life of each of us is changed, throughthe formation of the ideal, into what is highest in the human mind by our scale ofvalues. It would be vain, however, to attempt to localize the ego ideal, even inthe sense in which we have localized the ego, or to————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 264 [The super-ego is accordingly not included in the diagram on p. 24.Nevertheless it is given a place in the later diagram in Lecture XXXI of theNew Introductory Lectures (1933a).]- 36 -1[PEP]1
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

work it into any of the analogies with the help of which we have tried to picturethe relation between the ego and the id.It is easy to show that the ego ideal answers to everything that is expected ofthe higher nature of man. As a substitute for a longing for the father, it containsthe germ from which all religions have evolved. The self-judgement whichdeclares that the ego falls short of its ideal produces the religious sense ofhumility to which the believer appeals in his longing. As a child grows up, therole of father is carried on by teachers and others in authority; their injunctionsand prohibitions remain powerful in the ego ideal and continue, in the form ofconscience, to exercise the moral censorship. The tension between the demandsof conscience and the actual performances of the ego is experienced as a senseof guilt. Social feelings rest on identifications with other people, on the basis ofhaving the same ego ideal.Religion, morality, and a social sense—the chief elements in the higher sideof man—were originally one and the same thing. According to the hypothesiswhich I put forward in Totem and Taboo they were acquired phylogeneticallyout of the father-complex: religion and moral restraint through the process ofmastering the Oedipus complex itself, and social feeling through the necessityfor overcoming the rivalry that then remained between the members of theyounger generation. The male sex seems to have taken the lead in all these moralacquisitions; and they seem to have then been transmitted to women by cross-inheritance. Even to-day the social feelings arise in the individual as asuperstructure built upon impulses of jealous rivalry against his brothers andsisters. Since the hostility cannot be satisfied, an identification with the formerrival develops. The study of mild cases of homosexuality confirms the suspicionthat in this instance, too, the identification is a substitute for an affectionateobject-choice which has taken the place of the aggressive, hostile attitude.With the mention of phylogenesis, however, fresh problems arise, from whichone is tempted to draw cautiously back. But————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 265 I am at the moment putting science and art on one side. [Freud (1912-13), Standard Ed., 13, 146 ff.] Cf. Group Psychology (1921c) [Standard Ed., 18, 120] and ‘Some NeuroticMechanisms in Jealousy, Paranoia and Homosexuality’ (1922b) [Standard Ed.,231].- 37 -12
3[PEP]123
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

there is no help for it, the attempt must be made—in spite of a fear that it willlay bare the inadequacy of our whole effort. The question is: which was it, theego of primitive man or his id, that acquired religion and morality in those earlydays out of the father-complex? If it was his ego, why do we not speak simply ofthese things being inherited by the ego? If it was the id, how does that agree withthe character of the id? Or are we wrong in carrying the differentiation betweenego, superego, and id back into such early times? Or should we not honestlyconfess that our whole conception of the processes in the ego is of no help inunderstanding phylogenesis and cannot be applied to it?Let us answer first what is easiest to answer. The differentiation between egoand id must be attributed not only to primitive man but even to much simplerorganisms, for it is the inevitable expression of the influence of the externalworld. The super-ego, according to our hypothesis, actually originated from theexperiences that led to totemism. The question whether it was the ego or the idthat experienced and acquired these things soon comes to nothing. Reflection atonce shows us that no external vicissitudes can be experienced or undergone bythe id, except by way of the ego, which is the representative of the externalworld to the id. Nevertheless it is not possible to speak of direct inheritance inthe ego. It is here that the gulf between an actual individual and the concept of aspecies becomes evident. Moreover, one must not take the difference betweenego and id in too hard-and-fast a sense, nor forget that the ego is a speciallydifferentiated part of the id [p. 25]. The experiences of the ego seem at first tobe lost for inheritance; but, when they have been repeated often enough and withsufficient strength in many individuals in successive generations, they transformthemselves, so to say, into experiences of the id, the impressions of which arepreserved by heredity. Thus in the id, which is capable of being inherited, areharboured residues of the existences of countless egos; and, when the ego formsits super-ego out of the id, it may perhaps only be reviving shapes of formeregos and be bringing them to resurrection.The way in which the super-ego came into being explains how it is that theearly conflicts of the ego with the object-cathexes of the id can be continued inconflicts with their heir,————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 266- 38 -[PEP]
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

the super-ego. If the ego has not succeeded in properly mastering the Oedipuscomplex, the energic cathexis of the latter, springing from the id, will come intooperation once more in the reaction-formation of the ego ideal. The abundantcommunication between the ideal and these Ucs. instinctual impulses solves thepuzzle of how it is that the ideal itself can to a great extent remain unconsciousand inaccessible to the ego. The struggle which once raged in the deepest strataof the mind, and was not brought to an end by rapid sublimation andidentification, is now continued in a higher region, like the Battle of the Huns inKaulbach's painting.————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 267 [This was the battle, usually known as the Battle of Châlons, in which, in451, Attila was defeated by the Romans and Visigoths. Wilhelm von Kaulbach(1805-1874) made it the subject of one of his mural decorations, originallypainted for the Neues Museum in Berlin. In this the dead warriors arerepresented as continuing their fight in the sky above the battlefield, inaccordance with a legend that can be traced back to the fifth century Neo-Platonist, Damascius.]- 39 -1[PEP]1
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

IV The Two Classes of InstinctsWe have already said that, if the differentiation we have made of the mindinto an id, an ego, and a super-ego represents any advance in our knowledge, itought to enable us to understand more thoroughly the dynamic relations withinthe mind and to describe them more clearly. We have also already concluded [p.25] that the ego is especially under the influence of perception, and that,speaking broadly, perceptions may be said to have the same significance for theego as instincts have for the id. At the same time the ego is subject to theinfluence of the instincts, too, like the id, of which it is, as we know, only aspecially modified part.I have lately developed a view of the instincts which I shall here hold to andtake as the basis of my further discussions. According to this view we have todistinguish two classes of instincts, one of which, the sexual instincts or Eros, isby far the more conspicuous and accessible to study. It comprises not merely theuninhibited sexual instinct proper and the instinctual impulses of an aim-inhibited or sublimated nature derived from it, but also the self-preservativeinstinct, which must be assigned to the ego and which at the beginning of ouranalytic work we had good reason for contrasting with the sexualobject-instincts. The second class of instincts was not so easy to point to; in theend we came to recognize sadism as its representative. On the basis oftheoretical considerations, supported by biology, we put forward the hypothesisof a death instinct, the task of which is to lead organic life back into theinanimate state; on the other hand, we supposed that Eros, by bringing about amore and more far-reaching combination of the particles into which livingsubstance is dispersed, aims at complicating life and at the same time, of course,at preserving it. Acting in this way, both the instincts would be conservative inthe strictest sense of the word, since both would be endeavouring to re-establisha state of things that was disturbed by the emergence of life. The emergence oflife would thus be————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 268 Beyond the Pleasure Principle [1920g].- 40 -1
[PEP]1
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

the cause of the continuance of life and also at the same time of the strivingtowards death; and life itself would be a conflict and compromise between thesetwo trends. The problem of the origin of life would remain a cosmological one;and the problem of the goal and purpose of life would be answereddualistically.On this view, a special physiological process (of anabolism or catabolism)would be associated with each of the two classes of instincts; both kinds ofinstinct would be active in every particle of living substance, though in unequalproportions, so that some one substance might be the principal representative ofEros.This hypothesis throws no light whatever upon the manner in which the twoclasses of instincts are fused, blended, and alloyed with each other; but that thistakes place regularly and very extensively is an assumption indispensable to ourconception. It appears that, as a result of the combination of unicellularorganisms into multicellular forms of life, the death instinct of the single cell cansuccessfully be neutralized and the destructive impulses be diverted on to theexternal world through the instrumentality of a special organ. This special organwould seem to be the muscular apparatus; and the death instinct would thus seemto express itself—though probably only in part—as an instinct of destructiondirected against the external world and other organisms.Once we have admitted the idea of a fusion of the two classes of instinctswith each other, the possibility of a—more or less complete—‘defusion’ ofthem forces itself upon us. The sadistic component of the sexual instinct wouldbe a classical example of a serviceable instinctual fusion; and the sadism whichhas made itself independent as a perversion would be typical of a defusion,though not of one carried to extremes. From this point we obtain a view of agreat domain of facts which has not before been considered in this light. Weperceive that for purposes of discharge the instinct of destruction is habituallybrought into the service of Eros; we suspect that the epileptic fit is a product andindication of an instinctual defusion; and we come to————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 269 [Cf. footnote , p. 46 below.] [Freud returns to this in ‘The Economic Problem of Masochism’, p. 163below.] [Cf. above, p. 30. What follows in regard to sadism is hinted at in Beyondthe Pleasure Principle, Standard Ed., 18, 54.] [Cf. Freud's later paper on Dostoevsky's fits (1928b).]- 41 -1
234[PEP]12234
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

understand that instinctual defusion and the marked emergence of the deathinstinct call for particular consideration among the effects of some severeneuroses—for instance, the obsessional neuroses. Making a swift generalization,we might conjecture that the essence of a regression of libido (e.g. from thegenital to the sadistic-anal phase) lies in a defusion of instincts, just as,conversely, the advance from the earlier phase to the definitive genital onewould be conditioned by an accession of erotic components. The question alsoarises whether ordinary ambivalence, which is so often unusually strong in theconstitutional disposition to neurosis, should not be regarded as the product of adefusion; ambivalence, however, is such a fundamental phenomenon that it moreprobably represents an instinctual fusion that has not been completed.It is natural that we should turn with interest to enquire whether there may notbe instructive connections to be traced between the structures we have assumedto exist—the ego, the super-ego and the id—on the one hand and the two classesof instincts on the other; and, further, whether the pleasure principle whichdominates mental processes can be shown to have any constant relation both tothe two classes of instincts and to these differentiations which we have drawn inthe mind. But before we discuss this, we must clear away a doubt which arisesconcerning the terms in which the problem itself is stated. There is, it is true, nodoubt about the pleasure principle, and the differentiation within the ego hasgood clinical justification; but the distinction between the two classes ofinstincts does not seem sufficiently assured and it is possible that facts ofclinical analysis may be found which will do away with its pretension.One such fact there appears to be. For the opposition between the two classesof instincts we may put the polarity of love and hate. There is no difficulty infinding a representative of Eros; but we must be grateful that we can find arepresentative of the elusive death instinct in the instinct of destruction, to whichhate points the way. Now, clinical observation shows————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 270 [Freud recurs to this point in Inhibitions, Symptoms and Anxiety (1926d),Standard Ed., 20, 114.] [For what follows, see the earlier discussion of the relation between loveand hate in ‘Instincts and their Vicissitudes’ (1915c), Standard Ed., 14, 136-40, as well as the later one in Chapters V and VI of Civilization and itsDiscontents (1930a).]- 42 -1
2[PEP]12
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

not only that love is with unexpected regularity accompanied by hate(ambivalence), and not only that in human relationships hate is frequently aforerunner of love, but also that in a number of circumstances hate changes intolove and love into hate. If this change is more than a mere succession in time—if, that is, one of them actually turns into the other—then clearly the ground is cutaway from under a distinction so fundamental as that between erotic instinctsand death instincts, one which presupposes physiological processes running inopposite directions.Now the case in which someone first loves and then hates the same person(or the reverse) because that person has given him cause for doing so, hasobviously nothing to do with our problem. Nor has the other case, in whichfeelings of love that have not yet become manifest express themselves to beginwith by hostility and aggressive tendencies; for it may be that here thedestructive component in the object-cathexis has hurried on ahead and is onlylater on joined by the erotic one. But we know of several instances in thepsychology of the neuroses in which it is more plausible to suppose that atransformation does take place. In paranoia persecutoria the patient fends off anexcessively strong homosexual attachment to some particular person in a specialway; and as a result this person whom he loved most becomes a persecutor,against whom the patient directs an often dangerous aggressiveness. Here wehave a right to interpolate a previous phase which has transformed the love intohate. In the case of the origin of homosexuality, and of desexualized socialfeelings as well, analytic investigation has only recently taught us to recognizethat violent feelings of rivalry are present which lead to aggressive inclinations,and that it is only after these have been surmounted that the formerly hated objectbecomes the loved one or gives rise to an identification. The question ariseswhether in these instances we are to assume a direct transformation of hate intolove. It is clear that here the changes are purely internal and an alteration in thebehaviour of the object plays no part in them.There is another possible mechanism, however, which we have come toknow of by analytic investigation of the processes concerned in the change inparanoia. An ambivalent attitude is present from the outset and thetransformation is effected by————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 271 [See footnote , p. 37.]- 43 -1
[PEP]13
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

means of a reactive displacement of cathexis, energy being withdrawn from theerotic impulse and added to the hostile one.Not quite the same thing but something like it happens when the hostilerivalry leading to homosexuality is overcome. The hostile attitude has noprospect of satisfaction; consequently—for economic reasons, that is—it isreplaced by a loving attitude for which there is more prospect of satisfaction—that is, possibility of discharge. So we see that we are not obliged in any ofthese cases to assume a direct transformation of hate into love, which would beincompatible with the qualitative distinction between the two classes ofinstincts.It will be noticed, however, that by introducing this other mechanism ofchanging love into hate, we have tacitly made another assumption whichdeserves to be stated explicitly. We have reckoned as though there existed in themind—whether in the ego or in the id—a displaceable energy, which,indifferent in itself, can be added to a qualitatively differentiated erotic ordestructive impulse, and augment its total cathexis. Without assuming theexistence of a displaceable energy of this kind we can make no headway. Theonly question is where it comes from, what it belongs to, and what it signifies.The problem of the quality of instinctual impulses and of its persistencethroughout their various vicissitudes is still very obscure and has hardly beenattacked up to the present. In the sexual component instincts, which areespecially accessible to observation, it is possible to perceive a few processeswhich are in the same category as what we are discussing. We see, for instance,that some degree of communication exists between the component instincts, thatan instinct deriving from one particular erotogenic source can make over itsintensity to reinforce another component instinct originating from another source,that the satisfaction of one instinct can take the place of the satisfaction ofanother, and more facts of the same nature—which must encourage us to ventureupon certain hypotheses.In the present discussion, moreover, I am only putting forward a hypothesis; Ihave no proof to offer. It seems a plausible view that this displaceable andindifferent energy, which is no doubt active both in the ego and in the id,proceeds from the narcissistic store of libido—that it is desexualized Eros. (Theerotic instincts appear to be altogether more plastic, more————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 272 [The word here, and two paragraphs below, translated ‘neutral’ in theuncorrected printings of the present volume, is ‘indifferent’ in the Germanoriginal. ‘Indifferent’ is in fact a better rendering: the term ‘neutral energy’ hasrecently been commonly used in what seems a dissimilar sense. Actually theword and the whole notion had already been put forward by Freud in his paperon ‘Narcissism’ (1914c), Standard Edition, 14, 78. There the word is rightlytranslated ‘indifferent’.]- 44 -1
[PEP]1
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

readily diverted and displaced than the destructive instincts.) From this we caneasily go on to assume that this displaceable libido is employed in the service ofthe pleasure principle to obviate blockages and to facilitate discharge. In thisconnection it is easy to observe a certain indifference as to the path along whichthe discharge takes place, so long as it takes place somehow. We know this trait;it is characteristic of the cathectic processes in the id. It is found in eroticcathexes, where a peculiar indifference in regard to the object displays itself;and it is especially evident in the transferences arising in analysis, whichdevelop inevitably, irrespective of the persons who are their object. Not longago Rank [1913] published some good examples of the way in which neuroticacts of revenge can be directed against the wrong people. Such behaviour on thepart of the unconscious reminds one of the comic story of the three villagetailors, one of whom had to be hanged because the only village blacksmith hadcommitted a capital offence. Punishment must be exacted even if it does not fallupon the guilty. It was in studying the dream-work that we first came upon thiskind of looseness in the displacements brought about by the primary process. Inthat case it was the objects that were thus relegated to a position of no more thansecondary importance, just as in the case we are now discussing it is the paths ofdischarge. It would be characteristic of the ego to be more particular about thechoice both of an object and of a path of discharge.If this displaceable energy is desexualized libido, it may also be described assublimated energy; for it would still retain the main purpose of Eros—that ofuniting and binding—in so far as it helps towards establishing the unity, ortendency to unity, which is particularly characteristic of the ego. Ifthought-processes in the wider sense are to be included among thesedisplacements, then the activity of thinking is also supplied from the sublimationof erotic motive forces.Here we arrive again at the possibility which has already been discussed [p.30] that sublimation may take place regularly through the mediation of the ego.The other case will be recollected, in which the ego deals with the first object-cathexes of the id (and certainly with later ones too) by taking over the libidofrom them into itself and binding it to the alteration of————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 273 [The story was told by Freud in the last chapter of his book on jokes(1905c), Standard Ed., 8, 206, and in the eleventh of the IntroductoryLectures (1916-17), ibid. Standard Edition, 15, 174-5.]- 45 -1
[PEP]1
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

the ego produced by means of identification. The transformation [of eroticlibido] into ego-libido of course involves an abandonment of sexual aims, adesexualization. In any case this throws light upon an important function of theego in its relation to Eros. By thus getting hold of the libido from the object-cathexes, setting itself up as sole love-object, and de-sexualizing or sublimatingthe libido of the id, the ego is working in opposition to the purposes of Eros andplacing itself at the service of the opposing instinctual impulses. It has toacquiesce in some of the other object-cathexes of the id; it has, so to speak, toparticipate in them. We shall come back later to another possible consequenceof this activity of the ego [p. 54].This would seem to imply an important amplification of the theory ofnarcissism. At the very beginning, all the libido is accumulated in the id, whilethe ego is still in process of formation or is still feeble. The id sends part of thislibido out into erotic object-cathexes, whereupon the ego, now grown stronger,tries to get hold of this object-libido and to force itself on the id as alove-object. The narcissism of the ego is thus a secondary one, which has beenwithdrawn from objects.Over and over again we find, when we are able to trace instinctual impulsesback, that they reveal themselves as derivatives of Eros. If it were not for theconsiderations put forward in Beyond the Pleasure Principle, and ultimately forthe sadistic constituents which have attached themselves to Eros, we shouldhave difficulty in holding to our fundamental dualistic point of view. But sincewe cannot escape that view, we are driven to conclude that the death instinctsare by their nature mute and that the clamour of life proceeds for the most partfrom Eros.And from the struggle against Eros! It can hardly be doubted that the pleasureprinciple serves the id as a compass in its struggle against the libido—the forcethat introduces disturbances————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 274 [See Appendix B (p. 63) for a discussion of this.] [The consistency with which Freud held to a dualistic classification of theinstincts will be seen from his long footnote at the end of Chapter VI of Beyondthe Pleasure Principle (1920g), Standard Ed., 18, 460-1, and from thehistorical sketch in the Editor's Note to ‘Instincts and their Vicissitudes’(1915c), Standard Ed., 14, 113-16.] In fact, on our view it is through the agency of Eros that the destructiveinstincts that are directed towards the external world have been diverted fromthe self.- 46 -123[PEP]123
1Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

into the process of life. If it is true that Fechner's principle of constancygoverns life, which thus consists of a continuous descent towards death, it is theclaims of Eros, of the sexual instincts, which, in the form of instinctual needs,hold up the falling level and introduce fresh tensions. The id, guided by thepleasure principle—that is, by the perception of unpleasure—fends off thesetensions in various ways. It does so in the first place by complying as swiftly aspossible with the demands of the non-desexualized libido—by striving for thesatisfaction of the directly sexual trends. But it does so in a far morecomprehensive fashion in relation to one particular form of satisfaction in whichall component demands converge—by discharge of the sexual substances, whichare saturated vehicles, so to speak, of the erotic tensions. The ejection of thesexual substances in the sexual act corresponds in a sense to the separation ofsoma and germ-plasm. This accounts for the likeness of the condition thatfollows complete sexual satisfaction to dying, and for the fact that deathcoincides with the act of copulation in some of the lower animals. Thesecreatures die in the act of reproduction because, after Eros has been eliminatedthrough the process of satisfaction, the death instinct has a free hand foraccomplishing its purposes. Finally, as we have seen, the ego, by sublimatingsome of the libido for itself and its purposes, assists the id in its work ofmastering the tensions.————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 275 [Cf. Beyond the Pleasure Principle, Standard Ed., 18, 8-10.] [Freud's views on the part played by the ‘sexual substances’ will be found inSection 2 of the third of his Three Essays (1905d), Standard Ed., 7, 212-16.]- 47 -1
2
[PEP]12
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

V The Dependent Relationships of the EgoThe complexity of our subject-matter must be an excuse for the fact that noneof the chapter-headings of this book quite correspond to their contents, and thatin turning to new aspects of the topic we are constantly harking back to mattersthat have already been dealt with.Thus we have said repeatedly that the ego is formed to a great extent out ofidentifications which take the place of abandoned cathexes by the id; that thefirst of these identifications always behave as a special agency in the ego andstand apart from the ego in the form of a super-ego, while later on, as it growsstronger, the ego may become more resistant to the influences of suchidentifications. The super-ego owes its special position in the ego, or in relationto the ego, to a factor which must be considered from two sides: on the one handit was the first identification and one which took place while the ego was stillfeeble, and on the other hand it is the heir to the Oedipus complex and has thusintroduced the most momentous objects into the ego. The super-ego's relation tothe later alterations of the ego is roughly similar to that of the primary sexualphase of childhood to later sexual life after puberty. Although it is accessible toall later influences, it nevertheless preserves throughout life the character givento it by its derivation from the father-complex—namely, the capacity to standapart from the ego and to master it. It is a memorial of the former weakness anddependence of the ego, and the mature ego remains subject to its domination. Asthe child was once under a compulsion to obey its parents, so the ego submits tothe categorical imperative of its super-ego.But the derivation of the super-ego from the first object-cathexes of the id,from the Oedipus complex, signifies even more for it. This derivation, as wehave already shown [p. 36 ff.], brings it into relation with the phylogeneticacquisitions of the id and makes it a reincarnation of former ego-structureswhich have left their precipitates behind in the id. Thus the super-ego————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 277- 48 -[PEP]
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

is always close to the id and can act as its representative vis-à-vis the ego. Itreaches deep down into the id and for that reason is farther from consciousnessthan the ego is.We shall best appreciate these relations by turning to certain clinical facts,which have long since lost their novelty but which still await theoreticaldiscussion.There are certain people who behave in a quite peculiar fashion during thework of analysis. When one speaks hopefully to them or expresses satisfactionwith the progress of the treatment, they show signs of discontent and theircondition invariably becomes worse. One begins by regarding this as defianceand as an attempt to prove their superiority to the physician, but later one comesto take a deeper and juster view. One becomes convinced, not only that suchpeople cannot endure any praise or appreciation, but that they react inversely tothe progress of the treatment. Every partial solution that ought to result, and inother people does result, in an improvement or a temporary suspension ofsymptoms produces in them for the time being an exacerbation of their illness;they get worse during the treatment instead of getting better. They exhibit what isknown as a ‘negative therapeutic reaction’.There is no doubt that there is something in these people that sets itselfagainst their recovery, and its approach is dreaded as though it were a danger.We are accustomed to say that the need for illness has got the upper hand in themover the desire for recovery. If we analyse this resistance in the usual way—then, even after allowance has been made for an attitude of defiance towards thephysician and for fixation to the various forms of gain from illness, the greaterpart of it is still left over; and this reveals itself as the most powerful of allobstacles to recovery, more powerful than the familiar ones of narcissisticinaccessibility, a negative attitude towards the physician and clinging to the gainfrom illness.In the end we come to see that we are dealing with what may be called a‘moral’ factor, a sense of guilt, which is finding its satisfaction in the illness andrefuses to give up the punishment of suffering. We shall be right in regarding thisdisheartening explanation as final. But as far as the patient is concerned this————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 278 It may be said that the psycho-analytic or metapsychological ego stands on itshead no less than the anatomical ego—the ‘cortical homunculus’ [p. 26].- 49 -1
[PEP]1
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

sense of guilt is dumb; it does not tell him he is guilty; he does not feel guilty, hefeels ill. This sense of guilt expresses itself only as a resistance to recoverywhich it is extremely difficult to overcome. It is also particularly difficult toconvince the patient that this motive lies behind his continuing to be ill; he holdsfast to the more obvious explanation that treatment by analysis is not the rightremedy for his case.The description we have given applies to the most extreme instances of thisstate of affairs, but in a lesser measure this factor has to be reckoned with invery many cases, perhaps in all comparatively severe cases of neurosis. In factit may be precisely this element in the situation, the attitude of the ego ideal, thatdetermines the severity of a neurotic illness. We shall not hesitate, therefore, todiscuss rather more fully the way in which the sense of guilt expresses itselfunder different conditions.An interpretation of the normal, conscious sense of guilt————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 279 The battle with the obstacle of an unconscious sense of guilt is not made easyfor the analyst. Nothing can be done against it directly, and nothing indirectlybut the slow procedure of unmasking its unconscious repressed roots, and ofthus gradually changing it into a conscious sense of guilt. One has a specialopportunity for influencing it when this Ucs. sense of guilt is a ‘borrowed’ one—when it is the product of an identification with some other person who wasonce the object of an erotic cathexis. A sense of guilt that has been adopted inthis way is often the sole remaining trace of the abandoned love-relation andnot at all easy to recognize as such. (The likeness between this process andwhat happens in melancholia is unmistakable.) If one can unmask this formerobject-cathexis behind the Ucs. sense of guilt, the therapeutic success is oftenbrilliant, but otherwise the outcome of one's efforts is by no means certain. Itdepends principally on the intensity of the sense of guilt; there is often nocounteracting force of a similar order of strength which the treatment canoppose to it. Perhaps it may depend, too, on whether the personality of theanalyst allows of the patient's putting him in the place of his ego ideal, and thisinvolves a temptation for the analyst to play the part of prophet, saviour andredeemer to the patient. Since the rules of analysis are diametrically opposedto the physician's making use of his personality in any such manner, it must behonestly confessed that here we have another limitation to the effectiveness ofanalysis; after all, analysis does not set out to make pathological reactionsimpossible, but to give the patient's ego freedom to decide one way or theother.—[Freud returned to this topic in his paper on ‘The Economic Problemof Masochism’ (1924c), p. 166 below, where he discussed the distinctionbetween the unconscious sense of guilt and moral masochism. See alsoChapters VII and VIII of Civilization and its Discontents (1930a).]- 50 -1
[PEP]1
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

(conscience) presents no difficulties; it is based on the tension between the egoand the ego ideal and is the expression of a condemnation of the ego by itscritical agency. The feelings of inferiority so well known in neurotics arepresumably not far removed from it. In two very familiar maladies the sense ofguilt is over-strongly conscious; in them the ego ideal displays particularseverity and often rages against the ego in a cruel fashion. The attitude of the egoideal in these two conditions, obsessional neurosis and melancholia, presents,alongside of this similarity, differences that are no less significant.In certain forms of obsessional neurosis the sense of guilt is over-noisy butcannot justify itself to the ego. Consequently the patient's ego rebels against theimputation of guilt and seeks the physician's support in repudiating it. It wouldbe folly to acquiesce in this, for to do so would have no effect. Analysiseventually shows that the super-ego is being influenced by processes that haveremained unknown to the ego. It is possible to discover the repressed impulseswhich are really at the bottom of the sense of guilt. Thus in this case the super-ego knew more than the ego about the unconscious id.In melancholia the impression that the super-ego has obtained a hold uponconsciousness is even stronger. But here the ego ventures no objection; it admitsits guilt and submits to the punishment. We understand the difference. Inobsessional neurosis what were in question were objectionable impulses whichremained outside the ego, while in melancholia the object to which the super-ego's wrath applies has been taken into the ego through identification.It is certainly not clear why the sense of guilt reaches such an extraordinarystrength in these two neurotic disorders; but the main problem presented in thisstate of affairs lies in another direction. We shall postpone discussion of it untilwe have dealt with the other cases in which the sense of guilt remainsunconscious. [See p. 53.]It is essentially in hysteria and in states of a hysterical type that this is found.Here the mechanism by which the sense of guilt remains unconscious is easy todiscover. The hysterical ego fends off a distressing perception with which thecriticisms of its super-ego threaten it, in the same way in which it is in the habitof fending off an unendurable object-cathexis—by an act of repression. It is theego, therefore, that is responsible for the————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 280- 51 -[PEP]
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

sense of guilt remaining unconscious. We know that as a rule the ego carries outrepressions in the service and at the behest of its super-ego; but this is a case inwhich it has turned the same weapon against its harsh taskmaster. In obsessionalneurosis, as we know, the phenomena of reaction-formation predominate; buthere [in hysteria] the ego succeeds only in keeping at a distance the material towhich the sense of guilt refers.One may go further and venture the hypothesis that a great part of the sense ofguilt must normally remain unconscious, because the origin of conscience isintimately connected with the Oedipus complex, which belongs to theunconscious. If anyone were inclined to put forward the paradoxical propositionthat the normal man is not only far more immoral than he believes but also farmore moral than he knows, psycho-analysis, on whose findings the first half ofthe assertion rests, would have no objection to raise against the second half.It was a surprise to find that an increase in this Ucs. sense of guilt can turnpeople into criminals. But it is undoubtedly a fact. In many criminals, especiallyyouthful ones, it is possible to detect a very powerful sense of guilt whichexisted before the crime, and is therefore not its result but its motive. It is as if itwas a relief to be able to fasten this unconscious sense of guilt on to somethingreal and immediate.In all these situations the super-ego displays its independence of theconscious ego and its intimate relations with the unconscious id. Having regard,now, to the importance we have ascribed to preconscious verbal residues in theego [p. 20 f.], the question arises whether it can be the case that the super-ego, inso far as it is Ucs., consists in such word-presentations and, if it does not, whatelse it consists in. Our tentative answer will be that it is as impossible for thesuper-ego as for the ego to disclaim its origin from things heard: for it is a partof the ego and remains accessible to consciousness by way of these word-presentations (concepts, abstractions). But the cathectic energy does not reachthese contents of the super-ego from————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 281 This proposition is only apparently a paradox; it simply states that humannature has a far greater extent, both for good and for evil, than it thinks it has—i.e. than its ego is aware of through conscious perception. [A full discussion of this (together with some other references) will be foundin Part III of Freud's paper on ‘Some Character Types’ (1916d), Standard Ed.,14, 332-3.]- 52 -12
[PEP]12
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

auditory perception (instruction or reading) but from sources in the id.The question which we put off answering [see p. 51] runs as follows: How isit that the super-ego manifests itself essentially as a sense of guilt (or rather, ascriticism—for the sense of guilt is the perception in the ego answering to thiscriticism) and moreover develops such extraordinary harshness and severitytowards the ego? If we turn to melancholia first, we find that the excessivelystrong super-ego which has obtained a hold upon consciousness rages againstthe ego with merciless violence, as if it had taken possession of the whole of thesadism available in the person concerned. Following our view of sadism, weshould say that the destructive component had entrenched itself in the super-egoand turned against the ego. What is now holding sway in the super-ego is, as itwere, a pure culture of the death instinct, and in fact it often enough succeeds indriving the ego into death, if the latter does not fend off its tyrant in time by thechange round into mania.The reproaches of conscience in certain forms of obsessional neurosis are asdistressing and tormenting, but here the situation is less perspicuous. It isnoteworthy that the obsessional neurotic, in contrast to the melancholic, never infact takes the step of self-destruction; it is as though he were immune against thedanger of suicide, and he is far better protected from it than the hysteric. We cansee that what guarantees the safety of the ego is the fact that the object has beenretained. In obsessional neurosis it has become possible, through a regression tothe pregenital organization, for the love-impulses to transform themselves intoimpulses of aggression against the object. Here again the instinct of destructionhas been set free and it seeks to destroy the object, or at least it appears to havethat intention. These purposes have not been adopted by the ego and it strugglesagainst them with reaction-formations and precautionary measures; they remainin the id. The super-ego, however, behaves as if the ego were responsible forthem and shows at the same time by the seriousness with which it chastises thesedestructive intentions that they are no mere semblance evoked by regression butan actual substitution of hate for love. Helpless in both directions, the egodefends itself vainly, alike against the instigations of the murderous id andagainst the reproaches of the punishing conscience. It succeeds in holding incheck at————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 282- 53 -[PEP]
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

least the most brutal actions of both sides; the first outcome is interminable self-torment, and eventually there follows a systematic torturing of the object, in sofar as it is within reach.The dangerous death instincts are dealt with in the individual in variousways: in part they are rendered harmless by being fused with erotic components,in part they are diverted towards the external world in the form of aggression,while to a large extent they undoubtedly continue their internal work unhindered.How is it then that in melancholia the super-ego can become a kind of gathering-place for the death instincts?From the point of view of instinctual control, of morality, it may be said ofthe id that it is totally non-moral, of the ego that it strives to be moral, and of thesuper-ego that it can be super-moral and then become as cruel as only the id canbe. It is remarkable that the more a man checks his aggressiveness towards theexterior the more severe—that is aggressive—he becomes in his ego ideal. Theordinary view sees the situation the other way round: the standard set up by theego ideal seems to be the motive for the suppression of aggressiveness. The factremains, however, as we have stated it: the more a man controls hisaggressiveness, the more intense becomes his ideal's inclination toaggressiveness against his ego. It is like a displacement, a turning round uponhis own ego. But even ordinary normal morality has a harshly restraining,cruelly prohibiting quality. It is from this, indeed, that the conception arises of ahigher being who deals out punishment inexorably.I cannot go further in my consideration of these questions without introducinga fresh hypothesis. The super-ego arises, as we know, from an identificationwith the father taken as a model. Every such identification is in the nature of adesexualization or even of a sublimation. It now seems as though when atransformation of this kind takes place, an instinctual defusion occurs at the sametime [p. 30]. After sublimation the erotic component no longer has the power tobind the whole of the destructiveness that was combined with it, and this isreleased in the form of an inclination to aggression and————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 284 [Freud returned to this paradox in Section B of ‘Some Additional Notes onDream-Interpretation as a Whole’ (1925i), p. 134 below, and also in ‘TheEconomic Problem of Masochism’ (1924c), p. 170 below, and discussed itmore fully in Chapter VII of Civilization and its Discontents (1930a).]- 54 -1
[PEP]1
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

destruction. This defusion would be the source of the general character ofharshness and cruelty exhibited by the ideal—its dictatorial ‘Thou shalt’.Let us again consider obsessional neurosis for a moment. The state of affairsis different here. The defusion of love into aggressiveness has not been effectedby the work of the ego, but is the result of a regression which has come about inthe id. But this process has extended beyond the id to the super-ego, which nowincreases its severity towards the innocent ego. It would seem, however, that inthis case, no less than in that of melancholia, the ego, having gained control overthe libido by means of identification, is punished for doing so by the super-egothrough the instrumentality of the aggressiveness which was mixed with thelibido.Our ideas about the ego are beginning to clear, and its various relationshipsare gaining distinctness. We now see the ego in its strength and in itsweaknesses. It is entrusted with important functions. By virtue of its relation tothe perceptual system it gives mental processes an order in time and submitsthem to ‘reality-testing’. By interposing the processes of thinking, it secures apostponement of motor discharges and controls the access to motility. This lastpower is, to be sure, a question more of form than of fact; in the matter of actionthe ego's position is like that of a constitutional monarch, without whosesanction no law can be passed but who hesitates long before imposing his vetoon any measure put forward by Parliament. All the experiences of life thatoriginate from without enrich the ego; the id, however, is its second externalworld, which it strives to bring into subjection to itself. It withdraws libidofrom the id and transforms the object-cathexes of the id into ego-structures. Withthe aid of the super-ego, in a manner that is still obscure to us, it draws upon theexperiences of past ages stored in the id [p. 38].There are two paths by which the contents of the id can penetrate into the ego.The one is direct, the other leads by way of the ego ideal; which of these twopaths they take may, for some mental activities, be of decisive importance. Theego develops from perceiving instincts to controlling them, from————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 285 [Cf. ‘The Unconscious’ (1915e), Standard Ed., 14, 188.] [Cf. ‘Formulations on the Two Principles of Mental Functioning’ (1911b),Standard Ed., 12, 221.]- 55 -12
[PEP]12
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

obeying instincts to inhibiting them. In this achievement a large share is taken bythe ego ideal, which indeed is partly a reaction-formation against the instinctualprocesses of the id. Psycho-analysis is an instrument to enable the ego toachieve a progressive conquest of the id.From the other point of view, however, we see this same ego as a poorcreature owing service to three masters and consequently menaced by threedangers: from the external world, from the libido of the id, and from the severityof the super-ego. Three kinds of anxiety correspond to these three dangers, sinceanxiety is the expression of a retreat from danger. As a frontier-creature, the egotries to mediate between the world and the id, to make the id pliable to theworld and, by means of its muscular activity, to make the world fall in with thewishes of the id. In point of fact it behaves like the physician during an analytictreatment: it offers itself, with the attention it pays to the real world, as alibidinal object to the id, and aims at attaching the id's libido to itself. It is notonly a helper to the id; it is also a submissive slave who courts his master'slove. Whenever possible, it tries to remain on good terms with the id; it clothesthe id's Ucs. commands with its Pcs. rationalizations; it pretends that the id isshowing obedience to the admonitions of reality, even when in fact it isremaining obstinate and unyielding; it disguises the id's conflicts with realityand, if possible, its conflicts with the super-ego too. In its position midwaybetween the id and reality, it only too often yields to the temptation to becomesycophantic, opportunist and lying, like a politician who sees the truth but wantsto keep his place in popular favour.Towards the two classes of instincts the ego's attitude is not impartial.Through its work of identification and sublimation it gives the death instincts inthe id assistance in gaining control over the libido, but in so doing it runs therisk of becoming the object of the death instincts and of itself perishing. In orderto be able to help in this way it has had itself to become filled with libido; itthus itself becomes the representative of Eros and thenceforward desires to liveand to be loved.But since the ego's work of sublimation results in a defusion of the instinctsand a liberation of the aggressive instincts in the super-ego, its struggle againstthe libido exposes it to the danger of maltreatment and death. In suffering underthe attacks of the super-ego or perhaps even succumbing to them, the ego is————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 286- 56 -[PEP]
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

meeting with a fate like that of the protista which are destroyed by the productsof decomposition that they themselves have created. From the economic pointof view the morality that functions in the super-ego seems to be a similarproduct of decomposition.Among the dependent relationships in which the ego stands, that to the super-ego is perhaps the most interesting.The ego is the actual seat of anxiety. Threatened by dangers from threedirections, it develops the flight-reflex by withdrawing its own cathexis from themenacing perception or from the similarly regarded process in the id, andemitting it as anxiety. This primitive reaction is later replaced by the carrying-out of protective cathexes (the mechanism of the phobias). What it is that the egofears from the external and from the libidinal danger cannot be specified; weknow that the fear is of being overwhelmed or annihilated, but it cannot begrasped analytically. The ego is simply obeying the warning of the pleasureprinciple. On the other hand, we can tell what is hidden behind the ego's dreadof the super-ego, the fear of conscience. The superior being, which turned intothe ego ideal, once threatened castration, and this dread of castration is probablythe nucleus round which the subsequent fear of conscience has gathered; it is thisdread that persists as the fear of conscience.The high-sounding phrase, ‘every fear is ultimately the fear of death’, hashardly any meaning, and at any rate cannot be justified. It seems to me, on thecontrary, perfectly correct to————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 287 [Freud had discussed these animalculae in Beyond the Pleasure Principle,Standard Ed., 18, 48. They would probably now be described as ‘protozoa’rather than ‘protista’.] [What follows on the subject of anxiety must be read in connection withFreud's revised views as stated in Inhibitions, Symptoms and Anxiety(1926d), where most of the points raised here are further discussed.] [The notion of the ego being ‘overwhelmed’ (of an ‘Uberwältigung’) occursvery early in Freud's writings. See, for instance, a mention of it in Part II of hisfirst paper on ‘The Neuro-Psychoses of Defence’ (1894a). But it plays aprominent part in his discussion of the mechanism of the neuroses in Draft K ofJanuary 1, 1896, in the Fliess correspondence (Freud, 1950a). There is anevident connection here with the ‘traumatic situation’ of Inhibitions,Symptoms and Anxiety (1926d). See also Essay III in Moses and Monotheism(1939a), Standard Edition, 23, 77-8.] [‘Gewissensangst.’ An Editor's footnote on the use of this word will befound in Chapter VII on Inhibitions, Symptoms and Anxiety, Standard Ed., 20,128.] [Cf. Stekel (1908, 5).]- 57 -12345[PEP]12345
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

distinguish the fear of death from dread of an object (realistic anxiety) and fromneurotic libidinal anxiety. It presents a difficult problem to psycho-analysis, fordeath is an abstract concept with a negative content for which no unconsciouscorrelative can be found. It would seem that the mechanism of the fear of deathcan only be that the ego relinquishes its narcissistic libidinal cathexis in a verylarge measure—that is, that it gives up itself, just as it gives up some externalobject in other cases in which it feels anxiety. I believe that the fear of death issomething that occurs between the ego and the super-ego.We know that the fear of death makes its appearance under two conditions(which, moreover, are entirely analogous to situations in which other kinds ofanxiety develop), namely, as a reaction to an external danger and as an internalprocess, as for instance in melancholia. Once again a neurotic manifestation mayhelp us to understand a normal one.The fear of death in melancholia only admits of one explanation: that the egogives itself up because it feels itself hated and persecuted by the super-ego,instead of loved. To the ego, therefore, living means the same as being loved—being loved by the super-ego, which here again appears as the representativeof the id. The super-ego fulfils the same function of protecting and saving thatwas fulfilled in earlier days by the father and later by Providence or Destiny.But, when the ego finds itself in an excessive real danger which it believes itselfunable to overcome by its own strength, it is bound to draw the same conclusion.It sees itself deserted by all protecting forces and lets itself die. Here, moreover,is once again the same situation as that which underlay the first great anxiety-state of birth and the infantile anxiety of longing—the anxiety due to separationfrom the protecting mother.These considerations make it possible to regard the fear of death, like thefear of conscience, as a development of the fear of castration. The greatsignificance which the sense of guilt has in the neuroses makes it conceivablethat common neurotic————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 288 [Some discussion of the appearance of this notion here will be found in theEditor's Introduction to Inhibitions, Symptoms and Anxiety, Standard Ed., 20,85-6.] [This foreshadows the ‘separation anxiety’ discussed in Inhibitions,Symptoms and Anxiety (1926d), Standard Ed., 20, 151.]- 58 -12[PEP]12
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

anxiety is reinforced in severe cases by the generating of anxiety between theego and the super-ego (fear of castration, of conscience, of death).The id, to which we finally come back, has no means of showing the egoeither love or hate. It cannot say what it wants; it has achieved no unified will.Eros and the death instinct struggle within it; we have seen with what weaponsthe one group of instincts defends itself against the other. It would be possible topicture the id as under the domination of the mute but powerful death instincts,which desire to be at peace and (prompted by the pleasure principle) to putEros, the mischief-maker, to rest; but perhaps that might be to undervalue thepart played by Eros.————————————— This page can be read in German in GESAMMELTE WERKE Vol 13,Page 289- 59 -[PEP]
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

Appendix A to "The Ego and the Id"James StracheyThe Descriptive and the Dynamic UnconsciousA Curious point arises out of two sentences both of which appear on p. 15above. The Editor's attention was drawn to it in a private communication fromDr. Ernest Jones, who had come across it in the course of examining Freud'scorrespondence.On October 28, 1923, a few months after this work appeared, Ferenczi wroteto Freud in these terms: ‘…Nevertheless I venture to put a question to you …since there is a passage in The Ego and the Id which, without your solution, I donot understand.… On p. 13 I find the following: “… that in the descriptivesense there are two kinds of unconscious, but in the dynamic sense only one.”Since, however, you write on p. 12 that the latent unconscious is unconsciousonly descriptively, not in the dynamic sense, I had thought that it was preciselythe dynamic line of approach that called for the hypothesis of there being twosorts of Ucs., while description knows only Cs. and Ucs.’To this Freud replied on October 30, 1923: ‘… Your question about thepassage on p. 13 of The Ego and the Id has positively horrified me. Whatappears there gives a directly opposite sense to p. 12; and in the sentence on p.13 “descriptive” and “dynamic” have simply been transposed.’A little consideration of this startling affair suggests, however, that Ferenczi'scriticism was based on a misunderstanding and that Freud was over-hasty inaccepting it. The confusions which underlie Ferenczi's remarks are not veryeasily sorted out, and a rather lengthy argument is inevitable. Since, however,others besides Ferenczi may fall into the same error, it seems worth while to tryto clear the matter up.We will start off with the first half of Freud's later sentence: ‘in thedescriptive sense there are two kinds of unconscious.’ The meaning of thisseems perfectly clear: the term ‘unconscious’ in its descriptive sense covers twothings—the latent unconscious and the repressed unconscious. Freud might,however, have expressed the idea even more clearly. Instead of————————————— Of the German edition. Both sentences are on p. 15 here.- 60 -11
1
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

‘two kinds of unconscious [zweierlei Unbewusstes]’ he might have saidexplicitly that in the descriptive sense there are ‘two kinds of things that areunconscious’. And in fact Ferenczi evidently misunderstood the words: he tookthem to be saying that the term ‘descriptively unconscious’ had two differentmeanings. This, as he rightly saw, could not be so: the term unconscious, useddescriptively, could only have one meaning—that the thing it was applied to wasnot conscious. In logical terminology, he thought Freud was speaking of theconnotation of the term whereas he was actually speaking of its denotation.We now proceed to the second half of Freud's later sentence: ‘but in thedynamic sense [there is] only one [kind of unconscious]’. Here again themeaning seems perfectly clear: the term ‘unconscious’ in its dynamic sensecovers only one thing—the repressed unconscious. This is once more astatement about the denotation of the term; though even if it had been about itsconnotation it would still be true—the term ‘dynamic unconscious’ can onlyhave one meaning. Ferenczi, however, objects to it, on the ground that ‘it wasprecisely the dynamic line of approach that called for the hypothesis of therebeing two sorts of Ucs.’. Ferenczi was once more misunderstanding Freud. Hetook him to be saying that if we consider the term ‘unconscious’, bearingdynamic factors in mind, we see that it has only one meaning—which would, ofcourse, have been the opposite of everything that Freud was arguing. Whereaswhat Freud really meant was that all the things that are unconscious dynamically(i.e. that are repressed) fall into one class.—The position is made a little moreconfused by Ferenczi's using the symbol ‘Ucs.’ to mean ‘unconscious’ in thedescriptive sense—a slip which Freud himself makes by implication on p. 18.Thus this later sentence of Freud's seems altogether immune from criticism initself. But is it, as Ferenczi suggests and as Freud himself seems to agree,incompatible with the earlier sentence? This earlier sentence speaks of the latentunconscious as being ‘unconscious only descriptively, not in the dynamic sense’.Ferenczi appears to have thought that this contradicts the later statement that ‘inthe descriptive sense there are two kinds of unconscious’. But the two statementsdo not contradict each other: the fact that the latent unconscious is onlydescriptively unconscious does not in the least imply that it is the only thing thatis descriptively unconscious.- 61 –
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

There is, indeed, a passage in Lecture XXXI of Freud's New IntroductoryLectures, written some ten years later than the present work, in which the wholeof this argument is repeated in very similar terms. In that passage it is explainedmore than once that in the descriptive sense both the preconscious and therepressed are unconscious, but that in the dynamic sense the term is restricted tothe repressed.It must be pointed out that this interchange of letters took place only a veryfew days after Freud had undergone an extremely severe operation. He was notyet able to write (his reply was dictated), and he was probably in no conditionto weigh the argument thoroughly. It seems likely that on reflection he realizedthat Ferenczi's discovery was a mare's nest, for the passage was never altered inthe later editions of the book.- 62 –
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

Appendix B to "The Ego and the Id"James StracheyThe Great Reservoir of LibidoThere is considerable difficulty over this matter, which is mentioned in thefirst footnote on p. 30 and discussed at greater length on p. 46.The analogy seems to have made its first appearance in a new section addedto the third edition of the Three Essays (1905d), which was published in 1915but had been prepared by Freud in the autumn of 1914. The passage runs asfollows (Standard Ed., 7, 218): ‘Narcissistic or ego libido seems to be the greatreservoir from which the object-cathexes are sent out and into which they arewithdrawn once more; the narcissistic libidinal cathexis of the ego is theoriginal state of things, realized in earliest childhood, and is merely covered bythe later extrusions of libido, but in essentials persists behind them.’The same notion had, however, been expressed earlier in another favouriteanalogy of Freud's, which appears sometimes as an alternative and sometimesalongside the ‘great reservoir’. This earlier passage is in the paper onnarcissism itself (1914c), which was written by Freud in the early part of thesame year, 1914 (Standard Ed., 14, 75): ‘Thus we form the idea of there beingan original libidinal cathexis of the ego, from which some is later given off toobjects, but which fundamentally persists and is related to the object-cathexismuch as the body of an amoeba is related to the pseudopodia which it puts out.’The two analogies appear together in a semi-popular paper written at the endof 1916 for a Hungarian periodical (‘A Difficulty in the Path of Psycho-Analysis’, 1917a, Standard Ed., 17, 139): ‘The ego is a great reservoir fromwhich the libido that is destined for objects flows out and into which it flowsback from those objects … As an illustration of this state of things we may thinkof an amoeba, whose viscous substance puts out pseudopodia …’The amoeba appears once more in Lecture XXVI of the————————————— This analogy had appeared already in a rudimentary form in the third essay inTotem and Taboo, which was first published early in 1913. (Standard Ed., 13,89.)- 63 -1
1
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

Introductory Lectures (1916-17), dating from 1917, and the reservoir inBeyond the Pleasure Principle (1920g), Standard Ed., 18, 51: ‘Psycho-analysis …came to the conclusion that the ego is the true and original reservoirof libido, and that it is only from that reservoir that libido is extended on toobjects.’Freud included a very similar passage in an encyclopaedia article which hewrote in the summer of 1922 (1923a, Standard Ed., 18, 257), and then almostimmediately afterwards came the announcement of the id, and what appears likea drastic correction of the earlier statements: ‘Now that we have distinguishedbetween the ego and the id, we must recognize the id as the great reservoir oflibido …’ And again: ‘At the very beginning, all the libido is accumulated in theid, while the ego is still in process of formation or is still feeble. The id sendspart of this libido out into erotic object-cathexes, whereupon the ego, nowgrown stronger, tries to get hold of this object-libido and to force itself on the idas a love-object. The narcissism of the ego is thus a secondary one, which hasbeen withdrawn from objects.’ (Pp. 30 n. and 46 above.)This new position seems quite clearly intelligible, and it is therefore a littledisturbing to come upon the following sentence, written only a year or so afterThe Ego and the Id, in the Autobiographical Study (1925d [1924]), StandardEd., 20, 56: ‘All through the subject's life his ego remains the great reservoir ofhis libido, from which object-cathexes are sent out and into which the libido canstream back again from the objects.’The sentence, it is true, occurs in the course of a historical sketch of thedevelopment of psycho-analytic theory; but there is no indication of the changeof view announced in The Ego and the Id. And, finally, we find this passage inone of Freud's very last writings, in Chapter II of the Outline of Psycho-Analysis (1940a), written in 1938: ‘It is hard to say anything of the behaviour ofthe libido in the id and in the super-ego. All that we know about it relates to theego, in which at first the whole available quota of libido is stored up. We callthis state absolute primary narcissism. It lasts till the ego begins to cathect theideas of objects with libido, to transform narcissistic libido into object-libido.Throughout the whole of life the ego remains the great reservoir, from whichlibidinal cathexes are————————————— An almost identical statement is made in Lecture XXXII of the NewIntroductory Lectures (1933a). Standard Edition, 22, 103. But see also ibid.,22, 77: ‘The object-cathexes spring from the instinctual demands of the id.’- 64 -1
1
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

sent out to objects and into which they are also once more withdrawn, just as anamoeba behaves with its pseudopodia.’Do these later passages imply that Freud had retracted the opinions heexpressed in the present work? It seems difficult to believe it, and there are twopoints that may help towards a reconciliation of the apparently conflictingviews. The first is a very small one. The analogy of the ‘reservoir’ is from itsvery nature an ambiguous one: a reservoir can be regarded either as a waterstorage tank or as a source of water supply. There is no great difficulty inapplying the image in both senses both to the ego and to the id, and it wouldcertainly have clarified the various passages that have been quoted—and inparticular the footnote on p. 30—if Freud had shown more precisely whichpicture was in his mind.The second point is of greater importance. In the New Introductory Lectures,only a few pages after the passage referred to in the footnote above, in thecourse of a discussion of masochism, Freud writes: ‘If it is true of thedestructive instinct as well that the ego—but what we have in mind here israther the id, the whole person—originally includes all the instinctual impulses…’ The parenthesis points, of course, to a primitive state of things in which theid and the ego are still undifferentiated. And there is a similar, but moredefinite, remark in the Outline, this time two paragraphs before the passagealready quoted: ‘We picture an initial state as one in which the total availableenergy of Eros, which henceforward we shall speak of as “libido”, is present inthe still undifferentiated ego-id…’ If we take this as being the true essence ofFreud's theory, the apparent contradiction in his expression of it is diminished.This ‘ego-id’ was originally the ‘great reservoir of libido’ in the sense of beinga storage tank. After differentiation had occurred, the id would continue as astorage tank but, when it began sending out cathexes (whether to objects or to thenow differentiated ego) it would in addition be a source of supply. But the samewould be true of the ego as well, for it would be a storage tank of narcissisticlibido as well as, on one view, a source of supply for object-cathexes.This last point leads us, however, to a further question, on which it seemsinevitable to suppose that Freud held different views at different times. In TheEgo and the Id (p. 46) ‘at the————————————— This is, of course, a familiar view of Freud's.- 65 -1
1
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).

very beginning, all the libido is accumulated in the id’; then ‘the id sends part ofthis libido out into erotic object-cathexes’, which the ego tries to get control ofby forcing itself on the id as a love-object: ‘the narcissism of the ego is thus asecondary one.’ But in the Outline, ‘at first the whole available quota of libidois stored up in the ego’, ‘we call this state the absolutely primary narcissism’and ‘it lasts until the ego begins to cathect the ideas of objects with libido’. Twodifferent processes seem to be envisaged in these two accounts. In the first theoriginal object-cathexes are thought of as going out direct from the id, and onlyreaching the ego indirectly; in the second the whole of the libido is thought of asgoing from the id to the ego and only reaching the objects indirectly. The twoprocesses do not seem incompatible, and it is possible that both may occur; buton this question Freud is silent.- 66 –
Copyrighted Material. For use only by UPENN. Reproduction prohibited. Usage subject to PEP terms & conditions (see terms.pep-web.org).