identifications described above brings about its ‘secondary narcissism’.

[15] Perhaps it would be safer to say ‘with the parents’; for before a child has arrived at definite knowledge of the difference between the sexes, the missing penis, it does not distinguish in value between its father and its mother. I recently came across the instance of a young married woman whose story showed that, after noticing the lack of a penis in herself, she had supposed it to be absent not in all women, but only in those whom she regarded as inferior, and had still supposed that her mother possessed one.

In order to simplify my presentment I shall discuss only identification with the father.

[16] Cf. _Group Psychology and the Analysis of the Ego_, chap. vii.

[17] [This sentence represents a slight modification of the original text in accordance with direct instructions from the author.—TRANS.]

[18] I am at the moment putting science and art on one side.

[19] Cf. _Group Psychology and the Analysis of the Ego_; and ‘Certain Neurotic Mechanisms in Jealousy, Paranoia and Homosexuality’ (1922), _Collected Papers_, vol. ii.

[20] _Beyond the Pleasure Principle._

[21] In fact, according to our view it is through the agency of Eros that the destructive instincts that are directed towards the external world have been diverted from the self.

[22] It may be said that the psycho-analytical or metapsychological ego stands on its head no less than the anatomical ego—the ‘cortical homunculus’.

[23] The battle with the obstacle of an unconscious sense of guilt is not made easy for the analyst. Nothing can be done against it directly, and nothing indirectly but the slow procedure of unmasking its unconscious repressed roots, and of thus gradually changing it into a conscious sense of guilt. One has a special opportunity for influencing it when this Ucs sense of guilt is a ‘borrowed’ one, i.e. when it is the product of an identification with some other person who was once the object of an erotic cathexis. When the sense of guilt has been adopted in this way it is often the sole remaining trace of the abandoned love-relation and not at all easy to recognize as such. (The likeness between this process and what happens in melancholia is unmistakable.) If one can unmask this former object-cathexis behind the Ucs sense of guilt, the therapeutic success is often brilliant, but otherwise the outcome of one’s efforts is by no means certain. It depends principally on the intensity of the sense of guilt; there is often no counteracting force of similar strength which the treatment can put in motion against it. Perhaps it may depend, too, on whether the personality of the analyst allows of the patient’s putting him in the place of his ego-ideal, and this involves a temptation for the analyst to play the part of prophet, saviour, and redeemer to the patient. Since the rules of analysis are diametrically opposed to the physician’s making use of his personality in any such manner, it must be honestly confessed that here we have another limitation to the effectiveness of analysis; after all, analysis does not set out to abolish the possibility of morbid reactions, but to give the patient’s ego _freedom_ to choose one way or the other.

[24] This proposition is only apparently a paradox; it simply states that human nature has a far greater capacity, both for good and for evil, than it thinks it has, i.e. than it is aware of through the conscious perceptions of the ego.

[25] [The author’s views upon anxiety as given in the following paragraphs have been largely revised in his later work, _Hemmung, Symptom und Angst_.—TRANS.]