Sigmund Freud
Sigmund Freud; drawing by David Levine

It has become a commonplace among the general public, and perhaps among journalists in particular, that some things are Freudian symbols. In the popular usage of this phrase, the word Freudian is synonymous with “sexual,” the word “sexual” is synonymous with “genital,” while the question “What does the word ‘symbol’ mean?” is answered, if it is asked at all, by saying that it is something that stands for something else. A pipe, a sword, a gun are held to be Freudian symbols on the ground that Freud reputedly said that they stand for penises. In view of this widespread’ assumption, I thought it might be appropriate to start this essay by investigating whether the general public has in fact correctly divined and understood an essential part of Freud’s thinking or whether some distortion of his ideas has taken place during their passage from the learned literature into the popular press.

Now, in one very simple sense, a distortion certainly has occurred, Freud himself was not the psychoanalyst who first attached central importance to sexual symbolism. As he himself said, he arrived late at a full realization of the importance of symbolism of any kind. The first edition of The Interpretation of Dreams (1900) contains only a few pages devoted to the subject of symbolism in general and only one dream exemplifying sexual symbolism. These few pages came at the end of a section entitled “Considerations of Representability,” and the matter under discussion was not primarily “How are sexual ideas symbolized in dreams?” but two more general questions: How are thoughts, which are not necessarily visual, converted into the visual imagery characteristic of dreams? And how can forbidden, taboo ideas be expressed in dreams in such a way that the dreamer, when he awakens, will not understand them? (Sexual ideas being only one of several classes of forbidden ideas which the “dream work” has to translate into imagery that will get past the censor.)

Freud did, however, eventually get around to attaching great importance to symbolism in general and to sexual symbolism in particular, successive editions of The Interpretation of Dreams containing additions referring to sexual symbolism. But it was not until the fourth edition (1914) that Freud included a section concerned exclusively with symbolism. However, even in the present definitive Standard Edition, this section consists of only 55 pages in a work running to 623 pages.

Freud’s reasons for taking fourteen years to attaching central importance to the idea with which his name has become most widely associated are expressed clearly in the opening paragraphs of this section, and I should like now to quote extensively from them.

The analysis of this last, biographical, dream [i.e., the one discussed at the end of the previous section] is clear evidence that I recognized the presence of symbolism in dreams from the very beginning. But it was only by degrees and as my experience increased that I arrived at a full appreciation of its extent and significance, and I did so under the influence of the contributions of Wilhelm Stekel (1911), about whom a few words will not be out of place here. [1925]

That writer, who has perhaps damaged psycho-analysis as much as he has benefited it, brought forward a large number of unsuspected translations of symbols; to begin with they were met with scepticism, but later they were for the most part confirmed and had to be accepted. I shall not be belittling the value of Stekel’s services if I add that the sceptical reserve with which his proposals were received was not without justification. For the examples by which he supported his interpretations were often unconvincing, and he made use of a method which must be rejected as scientifically untrustworthy. Stekel arrived at his interpretations of symbols by way of intuition, thanks to a peculiar gift for the direct understanding of them. But the existence of such a gift cannot be counted upon generally, its effectiveness is exempt from all criticism and consequently its findings have no claim to credibility….

Advances in psycho-analytic experience have brought to our notice patients who have shown a direct understanding of dream-symbolism of this kind to a surprising extent. They were often sufferers from dementia praecox [i.e., what we nowadays call schizophrenia], so that for a time there was an inclination to suspect every dreamer who had this grasp of symbols of being a victim of that disease. But such is not the case. It is a question of a personal gift or peculiarity which has no visible pathological significance.

I should like now to make two glosses on this quotation. First, it is apparent that in symbolism Freud encountered a phenomenon which was resistive to, and indeed incompatible with, his ideal of founding a psychology based on natural-scientific principles. Freud’s original and indeed lifelong Grand Design was to construct a science of mind which would be analogous to the physical sciences, which used concepts such as force and energy, which was strictly determinist, and in which all explanations were in terms of causation. But here in symbolism he encountered phenomena which required explanations in terms of meaning, not of cause, and in which convincing explanations could be arrived at by “a method which must be rejected as scientifically untrustworthy” and by persons who possessed not a training in the rigors of the scientific method but a “peculiar gift for the direct understanding” of symbolic equations. It must have been all very embarrassing for Freud and one has to sympathize with his resistance against recognizing the importance of symbolism.


It should, however, be remembered that Freud displayed similar resistance against two other ideas which have become generally associated with his name: the idea that human beings are as affected by their fantasies as by their actual experiences (vide his distress at realizing that his female patients couldn’t all have been seduced by their fathers); and the idea that psychoanalytical treatment is essentially a matter of transference—“finally every conflict has to be fought out in the sphere of transference” (Freud, 1912).

Second, it is evident that Freud must himself have had patients who possessed the “peculiar gift” of being able to translate symbols intuitively, that it would have suited his theoretical book if they had all turned out to be mad, but that in fact they often weren’t—but note his phrase “no visible pathological significance.” It followed from this embarrassing observation that symbolism in the Freudian sense could not be either a pathological phenomenon capable of a causative explanation or one of the so-called “primary processes” which Freud postulated were characteristic of unconscious mental activity.

Freud held that there are two types or principles of mental functioning, one discursive, verbal, conforming to the laws of grammar and formal logic, and characteristic of conscious thinking, the other nondiscursive, condensive, iconic, ignorant of the categories of space and time, and characteristic of unconscious thinking. It would have been methodologically convenient and economical if Freudian symbolism could have been included in the latter, and most analysts other than Freud have in fact done so, but the existence of peculiar people like Stekel—and also, as Freud himself pointed out, of artists and jokers—makes it impossible to do so.

Having described Freud’s rather reluctant acceptance of the fact that there is such a thing as what has come to be called Freudian symbolism, I must turn to the question whether this symbolism is exclusively sexual, as popularizations of psychoanalysis would have one believe. The answer to this question is equivocal. According to Freud’s Tenth Introductory Lecture (1916-1917), “The range of things which are given symbolic representation in dreams is not wide; the human body as a whole, parents, children, brothers and sisters, birth, death, nakedness—and something else besides.” This “something else besides” turns out to be “the field of sexual life—the genitals, sexual processes, sexual intercourse. The great majority of symbols in dreams are sexual symbols. And here a strange disproportion is revealed. The topics I have mentioned are few, but the symbols for them are extremely numerous.” Freud then goes on to list over thirty symbols for the male genitals and over twenty for the female.

Reading this lecture one is indeed left with the overwhelming impression that in Freud’s view Freudian symbolism is predominantly sexual, but I must confess that I think that Freud reached the conclusion that “the great majority of symbols in dreams are sexual” by a mixture of logical error and intellectual sleight of hand.

The logical error consists in failing to appreciate that the topics which he designates “few” and the “symbols” which he designates as “numerous” are not of the same logical type, and cannot therefore be compared numerically with one another. The topics mentioned—birth, death, sex, etc.—are general ideas arrived at by abstraction, the symbols mentioned—umbrellas, revolvers, Zeppelins, cupboards, apples, etc., etc.—are specific objects, or, to be pedantic, concepts of a lower level of abstraction. It is therefore hardly surprising that there are more symbols that topics symbolized. The entities classified must of necessity be more numerous than the classes into which they are classified.

The intellectual sleight of hand consists of a) asserting that the range of things which are given symbolic representation is not wide, when in fact the list of things which he says are symbolizable embraces almost the whole range of human experience, apart perhaps from work, intellectual activity, and some kinds of play, and b) detaching the sexual life from the other members of the list of things deemed symbolizable and treating it as though it has no intrinsic connection with them. But this is, of course, not so. To go through his list: we only have a body because our parents once had sexual intercourse; we only have children because we have had intercourse; we only have brothers and sisters because our parents had inter-course more than once; birth and death are the first and last members of the series birth, copulation, and death; nakedness has obvious connections with both the sexual life and with the interfaces between the self as private and public being, as biological and social being.


What, it seems to me, Freud should have said was that the range of things which are given symbolic representation embraces all aspects of man’s biological life cycle, and that scrutiny of dreams reveals that human beings are much more preoccupied with their biological destiny and with their intimate personal relationships than most of them realize. Freud, strangely, seems always to have assumed that it is normal for human beings to be oblivious and obtuse about the importance of this emotional, poetic, mythopoeic, imaginative aspect of human experience, and that people who are aware of it are “special cases,” with gifts that are “peculiar” in both senses of the word. As I have written elsewhere (1962) Freud’s conception of the normal person was cast “in the mold of the scientist at work.” Hence, incidentally, his ambivalence toward artists, whom he both admired for their natural understanding of things that he had had to learn the hard way, and disparaged by trying to demonstrate that they were all neurotic. To be fair to Freud, I must add that late in his life he abandoned this attempt: “Before the problem of the creative artist analysis must, alas, lay down its arms” (Freud, 1928).

In parenthesis I must mention that I believe that Freud’s insistence in 1917 that the great majority of symbols are sexual was polemical, and was an attempt to preserve the scientific purity of psychoanalysis from contamination by Jungian ideas about archetypes, which he regarded as mystical and irrational. He wanted psycho-analysis to remain grounded in biology, and genitals are more down to earth than mandalas.

The same concealed polemical motive can also be discerned in Ernest Jones’s paper “The Theory of Symbolism” (1916), which remains to this day the classic statement of the psychoanalytical position, and which only a few analysts—in Britain Marion Milner (1952), Hanna Segal (1957), and myself (1956)—have challenged. In this paper Jones makes two points which I should like to discuss in some detail, one because it has been largely responsible for the difficulties in communication which have notoriously always existed between psychoanalysts on the one hand and anthropologists and linguists on the other hand, and the second because it is highly germane to the general theme of symbols and sentiments.

The first is Jones’s argument, or rather assertion, that there is such a thing as “true symbolism,” which is psychoanalytical and predominantly sexual and can be differentiated from what he calls “symbolism in its widest sense.” According to Jones’s distinction words, emblems, tokens, badges, charms, conventionalized gestures, etc., are not true symbols, even though they “represent some other idea from which they derive a significance not inherent in themselves,” and are only loosely and vaguely called symbols by those ignorant of psychoanalysis. “The thesis will here be maintained that true symbolism, in the strict sense, is to be distinguished from other forms of indirect representation.”

On the face of it, the claim that only Freudian symbols are true symbols is both arrogant and parochial. One would not, after all, allow a mathematician to get away with the assertion that the only true symbols are algebraic, or a linguist with the assertion that the only true symbols are words. But in Jones’s defense it must be mentioned that in 1916 it was generally believed that the concept of evolution could be applied to human societies, and that it was, therefore, legitimate to equate phylogenetically the “infantile fantasies” of civilized neurotics and the dreams of healthy civilized adults with the rites, folklore, myths, religions, etc., of primitives. As a result of this evolutionary assumption, Jones could without personal arrogance make remarks in 1916 which ring false in 1973, e.g., “Much more significant for the genesis of symbolism is the phylogenetic fact that in primitive civilizations an importance was attached to sexual organs and functions that to us appears absolutely monstrous,” and “The tendency of the primitive mind—as observed in children, in savages, in wit, dreams, insanity, and other products of unconscious functioning—to identify different objects and to fuse together different ideas, to note the resemblances and not the differences, is a universal and most characteristic feature….”

Now, given this assumption that civilized Western Europeans are more evolved than lesser breeds, Jones was, of course, justified in postulating that symbols encountered in dreams and neurotic symptoms had some sort of evolutionary priority over other more sophisticated kinds of symbols such as words, emblems, or crests, the use of which demands a considerable degree of ontogenetic and phylogenetic development. Jones’s position assumes, incidentally, that civilized children dream and fantasize before they learn how to talk—which may well be true—and that primitives think while awake in a way that resembles the way in which civilized people think while asleep—which I presume isn’t.

According to Jones true psycho-analytical symbols “represent ideas of the self and the immediate blood relatives or of the phenomena of birth, love, and death. In other words, they represent the most primitive ideas and interests imaginable.” Note again the curious use of the word “primitive” in a context which seems to demand “basic” or “fundamental.” “The two cardinal characteristics of symbolism in this strict sense are 1) that the process is completely unconscious… and 2) that the affect investing the symbolized idea has not, insofar as the symbolism is concerned, proved capable of that modification in quality denoted by the term ‘sublimation.”‘

I shall now consider these two cardinal characteristics separately. First, the idea that the process underlying symbolism is completely unconscious. This statement is, I think, capable of two interpretations. It could mean either 1) that an innate and unconscious sense of similarities determines whether particular objects are usable as symbols for other particular objects, or 2) that the person using a true symbol is unconscious of its true meaning. Both interpretations can be found in the psychoanalytical literature, the first to explain the apparent universal or constant meaning of certain symbols, the second to explain the difference between true symbols and metaphors. To make clearer what I am getting at, and to rescue us from the morass of abstractions into which we have fallen, I shall now give four illustrative examples, all of which depend on the fact that the human mind is capable of conceiving a similarity between a woman’s genitals and a violin or other stringed instrument. The two clinical examples are not my own; I have pinched them from Hannah Segal.

  1. A female violinist was asked, while having a psychotic breakdown, why she no longer played the violin. She answered, quite seriously, “But one doesn’t masturbated in public,” thereby implying that for her at that moment her violin was her genitals. The symbolism was fully conscious, but in a curious way, since she had temporarily lost insight into the fact that a symbol only represents something other than itself. She equated the symbol with its referent and could, therefore, no longer perform in public. I would recall here my earlier quotation from Freud in which he expressed his original suspicion that people with a direct understanding of symbolism were insane.
  2. Another female violinist was having psychoanalytical treatment on account of, inter alia, inhibitions about performing in public. During the course of her treatment she reported various dreams in which she appeared to represent masturbatory wishes by violin-playing, and after these dreams had been interpreted to her, she resumed playing in public. In this case the symbolism was initially completely unconscious; only in dreams did she equate violins and female genitals, and her analyst had to draw her attention to the equation. But after she had recognized her unconscious tendency to fuse, or rather confuse, violins and genitals—and after, presumably, her sense of guilt about masturbation had been reduced—she could again perform in public.
  3. The French novelist Honoré de Balzac, who died in 1850 and cannot, therefore, have read any Freud, is reputed once to have remarked that the love-making of many men resembled a gorilla trying to play a violin. In order to have been able to make this remark, Balzac must have been able to conceive of a similarity between men of a certain type and gorillas and of another similarity between women and violins. Since he was awake when he said it, he must have been fully conscious of these similarities and have assumed confidently that his audience could also become conscious of them. He was in fact talking metaphorically, in perfect confidence that his audience would neither assume that he thought that clumsy men are in fact gorillas or that women are in fact violins, nor that they would fail to appreciate that clumsy men can be likened to gorillas and that women can be likened to violins.

According to Jones’s formulations, such metaphorical statements are not truly symbolic, since the process is not completely unconscious, and yet it is difficult to see in what way the process of symbolism differs when it occurs consciously and emerges as a metaphor and unconsciously and emerges as a dream image. It seems to me to be more economical and logical to say that the process is identical in both cases and that symbolism, or rather the capacity to symbolize, is a general mental capacity which can be used consciously or unconsciously, while awake or asleep, neurotically or creatively, with or without insight into its implications.

  1. The sculptor Ossip Zadkine, who was born in 1890 and who may, therefore, well have read Freud, and whose work is characterized by “a flair for metaphor and transformation” and “a vocabulary in which voids replace solids, concavities replace convexities” (Jean Cassou in his introduction to the Arts Council exhibition, 1961), has produced a number of works in which human bodies and stringed instruments, mostly guitars and cellos, are fused. In some of these the strings of the instrument pass through the genital area of the female torso, leaving the viewer in no doubt that Zadkine is equating stringed instruments and women as sexual beings.

Here again the symbolism is patently conscious; sculptors aren’t asleep while they sculpt and must be presumed to reflect upon what they are doing. But since Zadkine could have read Freud, one would have to possess biographical information about him before deciding whether he spontaneously equated stringed instruments with women or whether he was self-consciously exploiting the intellectual concept of Freudian symbolism.

Personally, I suspect the former, but it is, I understand, a fact that artists who work for advertising agencies do consciously exploit the idea of Freudian symbols into advertisements, hoping thereby to add unconscious sex appeal to the wares they are trying to sell. Whether such gambits are effective remains dubious and there is, I understand, some evidence that they may on occasion misfire. The slogan “There’s a Tiger in my Tank” is said to have frightened some motorists off buying the gasoline it advertised. Surrealist painters also used Freudian symbolism self-consciously.*

Jones’s second cardinal characteristic of true symbolism is “that the affect investing the symbolized idea has not…proved capable of that modification in quality denoted by the term ‘sublimation’.” In other words, true symbols retain the emotional tone that is appropriate to the referent symbolized, while other classes of “symbolism in its widest sense” possess an emotional tone of some other quality, this other quality being only definable by reference to the concept of sublimation.

The implication here is that images and actions are only true symbols if they are accompanied by, or evoke, the feelings naturally aroused by the phenomena of birth, love, and death, but are not true symbols if they arouse some other more “sublime” affect. According to this view, violins were true symbols for the psychotic violinist I mentioned earlier since she conceived herself to be masturbating if she played one. They were also true symbols for the inhibited, neurotic violinist when she was dreaming about them but ceased to be when, after recovery, she could again play them. They were not true symbols for Balzac when he made his remark about gorillas playing them, since he and his audience were, presumably, amused and not sexually aroused by it. And Zadkine must be presumed to have been having aesthetic not sexual emotions while sculpting his torsos.

It seems to me that this second cardinal distinction between true and other symbols is open to three objections. First, Jones has really made a distinction between two different kinds of affect, unmodified and modified, not between two different kinds of symbol. Secondly, there are too many intermediate phenomena which resist classification on the basis of Jones’s distinction; for instance, obscene language, pornography, erotic art, intentional and unintentional double entendres. And thirdly, the essential nature of the distinction between unmodified and modified, sublimated affects remains obscure and mysterious.

Indeed much of the obscurity and esotericism of contemporary psychoanalytical theory derives precisely from the difficulties it has encountered in trying to formulate and explain the difference between instinctual and sublimated activities. According to Heinz Hartmann, the doyen of American psychoanalysis, sublimations occupy “a conflict-free area of the ego” and use “de-eroticized, de-aggressified energy.” I have no doubt that he knows what he means by this formulation—and, in fact, I do too, though I don’t subscribe to it—but it is hardly a formulation readily translatable into the language of any of the other humane sciences.

To summarize:

  1. The concept of Freudian symbolism has passed into general circulation and refers to the idea that many and perhaps most objects are sexual symbols.

  2. Although the idea is popularly believed to have originated from Freud’s The Interpretation of Dreams, Freud did not in fact attach great importance to symbolism, whether sexual or otherwise, until it was forced upon him by Stekel and a number of peculiarly gifted patients.

  3. Freudian symbols are not in fact exclusively sexual. They represent, to quote Jones again, “ideas of the self and the immediate blood relatives or of the phenomena of birth, love, and death,” in fact, everything that comprises man’s biological destiny.

  4. Psychoanalytical theory has attempted to pre-empt the concept of symbolism by asserting that the only true symbols are those which analysts encounter when interpreting dreams and neurotic symptoms.

  5. By so doing, psychoanalytical theory has offended against common usage and has created well-nigh insuperable barriers between itself and other disciplines.

  6. To my mind, these barriers could be lowered by recognizing that symbolization is a general capacity of the mind, which can be used both by the discursive, syntactical form of thinking characteristic of waking, intellectual activity, and by the nondiscursive, condensive form of thinking characteristic of our dream life; and by recognizing that these two types of thinking are not necessarily opposed to one another but can work in harness, as is exemplified by the activities of creative artists, jokers, and peculiar people like Stekel.

  7. Since Freudian ideas have, after a fashion, passed into general currency, self-conscious, contrived, and fanciful manipulations of the idea of Freudian symbolism have become possible. These manipulations are meretricious, since they can be passed off as products of the creative imagination.

This Issue

January 24, 1974