Before I go any further, I must touch on two questions that take us to the heart of the difficulties entailed by this topic. First: how does narcissism as we are here proposing it relate to autoeroticism, which we have elsewhere described as an early form of libido? Second: if we attribute a primary libido-cathexis to the ego, why is there any need to differentiate sexual libido from non-sexual energy in the ego drives? Wouldn't the supposition of a single, unified psychic energy spare us all the difficulties associated with trying to distinguish between ego-drive energy and ego-libido, between ego-libido and object-libido?
As to the first question, I say this: it is a necessary hypothesis that there is no entity present in the individual from the very beginning that is equatable with the ego; the ego has to be developed. Autoerotic drives, however, are primal; therefore something else must supervene in addition to autoeroticism, a new psychic process, in order to produce narcissism.
Any psychoanalyst called upon to give a definitive answer to the second question is bound to feel distinctly uncomfortable. One balks at the idea of abandoning empirical observation for the sake of sterile theoretical disputes, but nonetheless we cannot shirk the obligation to try to resolve the issue. Notions such as that of an ego-libido or an ego-drive energy etc. are undoubtedly neither particularly easy to grasp nor sufficiently weighty in content; a speculative theory of the relevant relationships would want above all to establish a sharply defined concept as a basis for everything else. But in my view that is precisely the difference between a speculative theory and a science founded on the interpretation of empirical facts. The latter will not envy speculation its privilege of resting upon neat and tidy foundations of unassailable logic, but will gladly make do with nebulously evanescent, scarcely conceivable basic ideas, hoping to grasp them more clearly as they develop, and willing if need be to exchange them for others. For these ideas are not the foundation upon which the entire science rests; instead, it rests solely upon observation. They are not the substructure but the superstructure of the whole edifice, and can be replaced or discarded without harm. We are currently seeing the same sort of thing happen in physics, moreover, whose fundamental ideas about matter, centres of force, attraction and such like are scarcely less precarious than their counterparts in psychoanalysis.
The value of the concepts ‘ego-libido’ and ‘object-libido’ resides in the fact that they derive from thorough study of the intimate characteristics of neurotic and psychotic processes. The separation of the libido into one that pertains to the ego, and one that becomes attached to objects, is a necessary corollary of a primary hypothesis that differentiated between sexual drives and ego drives. This, at any rate, was the conclusion that I was driven to by analysis of both of the pure forms of transference neurosis (hysteria and obsessional neurosis), and I know only that all attempts to account for these phenomena by other means have utterly failed.
Given the complete lack of any guiding theory of drives, it is legitimate, not to say imperative, first to take a hypothesis of some kind and test it thoroughly and rigorously until it either fails, or proves valid. Now, quite a number of things tend to support the hypothesis of a primal separation of sexual drives and other kinds of drives, i.e. ego drives, not least its efficacy in the analysis of transference neuroses. I admit that this factor on its own would not be unambiguous, for it might well be a question here of indifferent psychic energy that turns into libido only through the process of object-cathexis. For one thing, however, this conceptual distinction corresponds to the distinction so commonly encountered in ordinary life between hunger and love. For another thing, biological considerations lend support to the hypothesis. The individual really does lead a double existence both as an end in himself, and as a link in a chain that he serves against his will, or at any rate regardless of his will. He even supposes sexuality to be one of his own designs – whereas on an alternative view he appears as a mere appendage of his germ-plasm,12 to whose purposes he devotes all his energies in return for the reward of a mere sensation of pleasure. On this view, he is but the mortal vehicle of a – perhaps – immortal essence; like the lord of an entailed estate, he is but the temporary occupant of an institution that will outlast him. The separation of the sexual drives from the ego drives would simply mirror this dual function of the individual. Thirdly, one has to bear in mind that all our tentative psychological theories will need to be grounded at some point in organic systems. It will then very likely transpire that it is particular substances and chemical processes that are responsible for the workings of sexuality, and which make it possible for the life of the individual to carry over into the life of the species. We take full account of this likelihood by substituting particular psychological forces for particular chemical substances.
Precisely because I am normally at pains to keep psychology separate from all that is alien to it, including the mode of thinking characteristic of biology, I wish to concede quite explicitly at this point that the hypothesis of separate ego drives and sexual drives, i.e. the libido theory, is essentially biologically based, and is grounded scarcely at all in psychology. I shall therefore also be consistent enough to drop this hypothesis if a better and more serviceable theory of drives were to emerge from psychoanalytical work itself-though this has not so far proved to be the case. It might then turn out that – at the deepest possible level and at the remotest possible distance – sexual energy, the libido, originated as a part of the energy inherently active in the psyche that then separated off through differentiation. But such a proposition is of little relevance. It concerns things that are so far removed from the problems raised by our clinical observations, and so limited in their contribution to our knowledge, that it is no more worth contesting than it is worth applying in practice. Any such primal oneness is perhaps just as irrelevant to our analytical interests as the primal kinship of all the races of man is to the Probate Officer seeking proof of kinship between an heir and a testator. All these speculations get us nowhere. And as we cannot wait until the definitive theory of drives is handed to us on a plate by some other science, it is far more expedient to try to see what light can be thrown on these fundamental biological puzzles by a synthesis of psychological phenomena. By all means let us acquaint ourselves with the possibility of error, but let us not be deterred from rigorously following up the first hypothesis we mentioned,13 viz. that of an antagonism between ego drives and sexual drives thrust upon us by our analysis of the transference neuroses, and thereby discovering whether it can be developed in a fruitful and consistent way, and whether it can be applied to other disorders as well, e.g. schizophrenia.
Things would be different, of course, if it were proven that the libido theory had already come to grief in failing to explain this latter disorder. C. G. Jung has made precisely this claim (1912) and has thereby forced me to set out the considerations above, which I would much rather have been spared. I should have preferred to follow through to its conclusion the path already taken in my analysis of the Schreber case, without going into its underlying assumptions. Jung's claim, however, is premature at the very least. His reasoning is scant. He bases his argument in the first place on my own supposed admission that in the face of the difficulties of the Schreber analysis I felt driven to modify the libido concept, that is to say, to abandon the notion of its having a sexual content and to regard the libido as being part and parcel of psychic interest14 in general. As to rectifying this misconception, Ferenczi (1913) has already said all that needs to be said in his thorough critique of Jung's book.15 I can only agree with Ferenczi, and repeat that I have never voiced any such renunciation of the libido theory. A further argument of Jung's, asserting that there was no reason to think that the loss of the normal reality-function16 could be caused solely by withdrawal of the libido, is not an argument at all, but an assertion of dogma; it ‘begs the question17 and pre-empts debate, whereas the question whether and how such a thing might be possible really does need to be explored. In his next major wor
k (1913), Jung touches briefly on the solution that I pointed to quite some time ago: ‘Now in all this we admittedly also need to take account of the fact – something incidentally that Freud refers to in his account of the Schreber case – that introversion of the Libido sexualis leads to a cathexis of the “ego”, which is conceivably what causes this reality-loss effect to appear. The possibility that the psychology of reality-loss might be explained in this way is indeed an enticing one.’ Unfortunately, however, Jung does not explore this possibility very far. Only a few lines later he dismisses it with the comment that on such a basis ‘the psychology of an ascetic anchorite would emerge successfully, but not a dementia praecox’. To show how little this inapt analogy can contribute to a resolution of the issue, we need only remark that such an anchorite in his ‘eagerness to eradicate every trace of sexual interest’ (though only in the popular sense of the word ‘sexual’) does not even need to exhibit any pathogenic libido lodgement. Though he may have completely averted his sexual interest from human beings, he can easily have sublimated it into a heightened interest in the divine or natural or animal realm without falling victim to an introversion of his libido onto his fantasies, or a reversion of his libido to his ego. This analogy appears to disregard from the very outset any possibility of differentiating between interest arising from erotic sources, and that arising from others. If we also bear in mind that the researches of the Swiss school, however commendable, have elucidated only two features of dementia praecox – the existence of complexes familiar to the healthy as well as to neurotics, and the similarity between patients' fantasies and folk myths – whilst for the rest proving unable to throw any light on the actual mechanism of the disorder, then we can readily reject Jung's claim that the libido theory has been proved a ‘failure’ by its inability to solve the problem of dementia praecox, and is therefore finished in respect of other neuroses too.
II
Any direct study of narcissism seems to me to be prevented by a number of special difficulties. The principal means of approaching the matter is likely to remain the analysis of paraphrenias. Just as the transference neuroses have enabled us to trace the libidinal drive-impulses, so, too, dementia praecox and paranoia will afford us insight into the psychology of the ego. Once again, our understanding of the normal in all its seeming simplicity has to be derived from the pathological with all its warped and coarsened features. All the same, a few other paths to a better understanding of narcissism do remain open to us: the study of organic illness, of hypochondria, and of the love-life of the sexes; and I shall now discuss each of these in turn.
In considering the influence of organic illness on the distribution of the libido, I am following a suggestion made to me in conversation by Sándor Ferenczi. It is universally known, indeed it seems self-evident to us, that anyone tormented by organic pain and dire discomfort abandons all interest in the things of the external world, except in so far as they bear on his suffering. Closer observation shows us that he also withdraws all libidinal interest from his love-objects; that so long as he suffers, he ceases loving. The banality of this fact need not prevent us from translating it into the language of the libido theory. We would then say: the patient retracts his libido-cathexes into his ego, and redeploys them once he is well. “The sole abode of his soul forsooth’, says Wilhelm Busch of a toothache-stricken poet, ‘is the small black hole in his molar tooth.’ Libido and ego-interest share the same fate in this regard, and are once again indistinguishable from each other. The notorious egoism of the ill covers both. We find this egoism so self-evident because we know for certain that in similar circumstances we would behave in exactly the same way. In its own way, comedy, too, exploits this phenomenon whereby physical ailments sweep away even the most passionate inclinations, and replace them with utter indifference.
Like illness, the sleep state, too, involves a narcissistic process whereby the libido is withdrawn from its various positions18 and focused on the self or, to be more precise, on the sole desire for sleep. The egoism of dreams probably fits in very well in this context. If nothing else, we see examples in both cases of alterations in libido distribution as a result of ego-alteration.19
Hypochondria, like organic illness, expresses itself in painful and distressing physical sensations, and matches it, too, in the effect it has on libido distribution. The hypochondriac withdraws both his interest and – particularly markedly – his libido from the objects of the external world, and concentrates both of them on the organ that concerns him. But a disparity between hypochondria and organic illness forces itself on our attention here: in the latter case the painful sensations are grounded in demonstrable physical changes, whereas in the former they seem not to be. However, it would be fully in accord with our conception of neurotic processes as a whole if we were to venture the view that the message given out by hypochondria must indeed be quite right, and that it, too, must surely involve organic changes. But what would these changes consist in?
We are going to let ourselves be guided here by our knowledge that physical sensations of an unpleasant kind, comparable to those encountered in hypochondria, are also present in the other neuroses. I have already on an earlier occasion mentioned my inclination to regard hypochondria as the third ‘actual’ neurosis20 alongside neurasthenia and anxiety neuros's.21 It is probably not going too far to suppose that an element of hypochondria may also routinely be present in the other neuroses; the finest example of this is probably to be seen in anxiety neurosis and its overlying hysteria. It is of course the genital organ in its various states of excitation that constitutes the most familiar exemplar of an organ at once painfully sensitive and physically changed in some way, yet not in any ordinary sense of the word morbid. In such circumstances it becomes engorged with blood, swollen, moist, and the locus of manifold sensations. Let us use the term erogeneity to describe the process whereby a part of the body transmits sexual stimuli to the psyche; let us also bring to mind that our reflections on the theory of sexuality have long since accustomed us to the view that certain other parts of the body – the erogenous zones – might be able to substitute for the genitals and behave in a similar way to them: there is then just one further step that we must dare to take at this point. We can venture to regard erogeneity as a general property of all the organs, and we can then speak of it as increasing or diminishing in intensity in any particular part of the body. Any such variation in the erogeneity of the organs might be paralleled by a change in libido-cathexis within the ego. It is in such factors perhaps that we need to search for whatever it is that we might consider the basis of hypochondria, and that is capable of having the same effect on libido distribution as when the organs are affected by physical illness.
We note that if we continue thinking along these lines we shall come face to face with the problem not only of hypochondria, but also of the other ‘actual’ neuroses, neurasthenia and anxiety neurosis. Let us therefore call a halt at this juncture: the purposes of a purely psychological study are not served by straying so far into the realm of physiological research. We might simply mention that on the available evidence it seems reasonable to suppose that hypochondria stands in a similar relationship to paraphrenia as the other ‘actual’ neuroses do to hysteria and obsessional neurosis, that is to say that it depends on ego-libido just as the others depend on object-libido; on this supposition, hypochondriac fear is the counterpart on the ego-libido side to neurotic fear. Furthermore, given that in the case of the transference neuroses we are already familiar with the idea that the mechanism of the onset of illness and of symptom-formation - the progression from introversion to regression – can be linked to a heavy build-up of object-libido,22 then we may also feel more inclined to embrace the idea of a heavy build-up of ego-libido, and relate it to the phenomena of hypochondria and paraphrenia.
Naturally enough, our thirst for knowledge prompts us at this point to ask why such a build-up of libido in the ego has to be experienced as unpleasurable. I should like
to make do here with the reply that unpleasure is routinely the form in which increased tension expresses itself, and that therefore what happens here, as elsewhere, is that a certain quantity of the physical process transmutes into the psychic quality of unpleasure; though it may then well be the case that what determines the degree of unpleasure is not the absolute magnitude of that physical process, but rather some particular function of it. From this vantage point one may even dare to approach the question as to where the compulsion comes from in the first place that makes the psyche transcend the boundaries of narcissism and invest the libido in objects. Again, the logical answer in terms of our overall train of thought would be that the compulsion arises when the libidinal cathexis of the ego has exceeded a certain level. A strong ego affords some protection against falling ill; but in the end we must necessarily start loving if we are not to fall ill, and we must necessarily fall ill if refusal23 makes us incapable of loving – rather along the lines of the model offered by Heinrich Heine when he envisions the psychogenesis of Creation: