There are still a few words to be said concerning the relation of anxiety-neurosis to the other neuroses in respect of its incidence and its inner connections with them.

The purest cases of anxiety-neurosis are as a rule the most developed. These cases are found among young and sexually potent persons; they show a uniform ætiology and no very long duration.

More frequently, however, anxiety symptoms occur contemporaneously and in combination with those of neurasthenia, hysteria, obsessions and melancholia. If because of this clinical confusion we are to refrain from distinguishing anxiety-neurosis as a self-contained unity, then to be logical we should have in addition to renounce the distinction between hysteria and neurasthenia which has been acquired so laboriously.

For the analysis of 'mixed neuroses' I can advocate the following important formula: _Wherever a mixed_ _neurosis exists a combination {103} of several specific ætiologies may be discovered_.

The number of ætiological factors which conditions a mixed neurosis may occur quite fortuitously; for instance, if a fresh injurious factor adds its effect to those already existing, e.g. a woman who had always been hysterical begins at a certain point in her married life to experience coitus reservatus and then acquires anxiety-neurosis in addition to her hysteria; a man who had previously been in the habit of masturbating, and had become neurasthenic, becomes engaged to be married and is sexually roused during the intimacy with his fiancée, and then acquires anxiety-neurosis in addition to his neurasthenia.

In other cases the number of ætiological factors is no accident; on the contrary, some one of them has brought another into activity; for example, a woman whose husband practises coitus reservatus without regard to her satisfaction finds herself compelled to relieve by masturbation the distressing excitation aroused by the act, as a result of which she develops an anxiety-neurosis, but not in a pure form, showing at the same time symptoms of neurasthenia; another woman in the same injurious situation will have to battle with lewd thoughts and visions, against which she struggles to defend herself, and will in this way develop obsessional ideas as well as the anxiety-neurosis as a result of coitus interruptus; lastly, a third woman's husband will lose his attraction for her in consequence of coitus interruptus and she will develop an affection for another man which she carefully keeps secret, as a result of which we find a mixture of anxiety-neurosis and hysteria.

In a third category of mixed neuroses the interrelation of the symptoms is even closer, in that the very same ætiological determinant {104} regularly and simultaneously evokes both neuroses. Thus for example, the sudden sexual revelation which is active in cases of virginal anxiety always gives rise to hysteria as well as anxiety-neurosis; cases of voluntary abstinence are for the most part from the beginning united with true obsessional ideas; coitus interruptus never seems to me to be able to provoke a pure anxiety-neurosis in men, but always a combination of it with neurasthenia; and so on.

It follows from these considerations that the ætiological conditions for the incidence of the neuroses must be distinguished more clearly from their specific ætiological factors. The former, for example, coitus interruptus, masturbation and abstinence, are still capable of various interpretations; each one of them can produce many neuroses; only the ætiological factors into which they can be resolved, such as _inadequate relief_, _psychical inadequacy_, _defence by substitution_, have an unequivocal and specific relation to the ætiology of the individual great neuroses.

In its essence anxiety-neurosis presents the most interesting similarities to and differences from the other great neuroses, particularly neurasthenia and hysteria. It shares with neurasthenia its main characteristic: the source of the excitation, the inciting factor in the disturbance, is somatic in nature, whereas in hysteria and the obsessional neurosis it is psychical in nature. In other respects we see rather a kind of antithesis between the symptoms of neurasthenia and those of anxiety-neurosis which may be summed up in the words: impoverishment or accumulation of excitation respectively. This antithesis does not prevent the two neuroses from being combined with one another, but is nevertheless clearly {105} demonstrated by the fact that the most extreme forms of both neuroses are also the purest.

In symptomatology anxiety-neurosis and hysteria have many points in common, which need to be better appreciated. The appearance of symptoms either in a chronic form or as attacks, the paræsthesias grouped like aurœ, the hyperæsthesias and points of pressure which are found in certain surrogates of the anxiety-attack (dyspnœa and heart-attacks), the exacerbation (through conversion) of pains perhaps having an organic basis---these and other common features even permit the conjecture that much of what we attribute to hysteria may with more justification be laid at the door of anxiety-neurosis. If we go into the mechanism of the two neuroses so far as it has been possible to discover it up to the present, aspects come to light which suggest that anxiety-neurosis is actually the somatic counterpart of hysteria. In each of them there is an accumulation of excitation---which perhaps accounts for the similarity of the symptoms already described; in each of them there is a _psychical inadequacy as a consequence of which abnormal somatic processes come about_. In each of them there occurs a deflection of excitation to the somatic field instead of psychical assimilation of it; the difference is merely this, that in anxiety-neurosis the excitation (in the displacement of which the neurosis expresses itself) is purely somatic (the somatic sexual excitation), whereas in hysteria it is purely psychical (evoked by conflict). Little wonder then that hysteria and anxiety-neurosis are regularly combined with one another, as in 'virginal anxiety' or 'sexual hysteria', and that hysteria simply borrows a number of its symptoms from the anxiety-neurosis. These intimate relations between {106} anxiety-neurosis and hysteria provide a new argument for demanding the distinction of anxiety-neurosis from neurasthenia; for if this distinction is not admitted then neither are we justified in maintaining the distinction between neurasthenia and hysteria which was so laboriously come by and is so indispensable for the theory of the neuroses.

VI {107} A REPLY TO CRITICISMS ON THE ANXIETY- NEUROSIS[28] (1895)

In No. 2 of Mendel's _Neurologisches Zentralblatt_ 1895, I published a short paper in which I ventured an attempt to distinguish a series of nervous states from neurasthenia and give them independence under the name of 'Anxiety-Neurosis'.[29] I was led to do so by finding a constant combination of certain clinical and ætiological characters, which may always decide us to make such distinctions. I found, as E. Hecker[30] had done before me, that the neurotic symptoms in question could all be united under the head of 'pertaining to the expression of _anxiety_'; and I was able to add from my study of the ætiology of the neuroses that these component parts of the complex 'anxiety-neurosis' show special ætiological conditions which are almost the opposite of the ætiology of neurasthenia. My observations had shown me that in the ætiology of the neuroses (at least of the acquired and acquirable forms) sexual factors play a predominant part, hitherto far too little appreciated; so that the assertion: 'the ætiology of the neuroses lies in sexuality', with all its necessary inexactness _per excessum et defectum_, nevertheless comes nearer to the truth than the other doctrines ruling at the present time. A further statement to which {108} experience forced me was to the effect that the various sexual noxiæ were not to be found indifferently in the ætiology of all neuroses, but that there were evidently special relationships of particular noxiæ to particular neuroses. I thought I could assume, therefore, that I had discovered the specific causes of the different neuroses. I then sought to comprehend in a short formula the peculiarity of the sexual noxiæ which constitute the ætiology of anxiety-neurosis, and arrived at the statement, dependent on my idea of the sexual process (see pp. 98--9): Anxiety-neurosis is produced by anything which withholds somatic sexual tension from the psychical and interferes with its elaboration within the psychical field. If we refer to the concrete circumstances in which this factor expresses itself, we come to the conclusion that voluntary or involuntary abstinence, sexual intercourse with incomplete gratification, coitus interruptus, the deflection of psychical interest from sexuality and so on, are the specific ætiological factors of the states I have called anxiety-neurosis.

When I published the paper mentioned above, I in no way deceived myself about its power to arouse conviction. In the first place I knew that I had only given a scanty, incomplete presentation---one that was even in parts difficult to comprehend; it was perhaps just sufficient to arouse the reader's expectation. I had besides adduced hardly any examples and quoted no statistics; the technique of collecting the case-histories was not touched on; no provision was made to avoid misunderstandings; other objections than the most obvious ones were not considered; and concerning the theory itself, only the main theme was brought into prominence and its limitations were disregarded. Accordingly, everyone was justified in forming his own {109} judgement upon the conclusiveness of the whole proposition. I had to look forward, moreover, to yet another source of disagreement. I know very well that I have advanced nothing new by suggesting the 'sexual ætiology' of the neuroses, that undercurrents in medical literature acknowledging these facts have never been absent, and that the official medicine of the schools has actually been aware of them also. But the last has behaved, however, as if it knew nothing about it, making no use of its knowledge, and deducing nothing from it. Such conduct must surely have a deep-rooted cause, originating perhaps in a kind of aversion from looking into sexual matters, or in a reaction against older attempts at explanation which were regarded as outworn. At all events, anyone who ventures to make something credible to others who could without any trouble have discovered it for themselves, must be prepared to meet with opposition.

With such a state of affairs it would perhaps have been more expedient not to answer critical objections until I had expressed myself in greater detail upon the complicated theme itself and had made myself more intelligible. Nevertheless I cannot resist the motives which induce me immediately to join issue with a recent criticism of my theory of anxiety-neurosis. I do this because the writer, L. Löwenfeld (Munich), the author of _Pathologie und Therapie der Neurasthenie und Hysterie_, is one whose judgement probably has great weight with the medical public, and because of a mistaken conception which Löwenfeld imputes to me; and also because I wish at once to combat the idea that my theory is to be overthrown so very easily by the first casual and impromptu objections. {110}

Löwenfeld correctly discovers at a glance the essence of my work to be my assertion that anxiety symptoms are due to a specific and uniform ætiology of a sexual nature. If this cannot be established as a fact then the principal reason for distinguishing an independent anxiety-neurosis from neurasthenia also falls to the ground. There remains, however, one difficulty to which I called attention---namely, that anxiety symptoms have manifest relations also to hysteria; so that the decision as Löwenfeld would have it brings the distinction between hysteria and neurasthenia to grief; though this difficulty is obviated by reference to heredity as the common cause of all the neuroses, a point which will be discussed later.

Now by what arguments does Löwenfeld support the case against my theory?

1. I have emphasized as a point essential for comprehension of the anxiety-neurosis that the anxiety of that condition does not admit of a psychic derivation, that is, that the anxious expectation which constitutes the nucleus of the neurosis cannot be acquired through a single or repeated affect of fright which is justified psychically. Fright, as I have pointed out, may result in hysteria or a traumatic neurosis, but not in an anxiety-neurosis. It is easy to see that this negation is simply the obverse of my assertion (with a positive content): the anxiety of my neurosis represents a somatic sexual tension deflected from the psychical field where it would otherwise have made itself felt as libido.

Against this, Löwenfeld lays stress on the fact that in a series of cases 'anxiety states appear immediately or shortly after a mental shock (fright alone, or accidents accompanied by fright) and that these circumstances make the co-operation of sexual noxiæ of the kind {111} mentioned at least most improbable.' He gives in brief, as a particularly pregnant example, one observation of a patient out of many. This example relates to a woman aged thirty, married for four years, of tainted stock, who had undergone a first difficult confinement a year before. A few weeks after it her husband had an attack of illness which frightened her exceedingly and in her excitement she ran about the cold room in her chemise. From that time onward she was ill, first with anxiety and palpitations in the evenings, later with attacks of convulsive trembling, and in further sequence phobias and the like---the picture of a fully-developed anxiety-neurosis. 'Here', concludes Löwenfeld, 'the anxiety states are manifestly of psychic origin consequent on a single shock.'

I do not doubt that my honoured critic can produce many similar cases; I myself can supply a long series of analogous examples. Anyone who had not seen such very common cases of the outbreak of anxiety-neurosis following upon a mental shock could not regard himself as qualified to discuss the question of anxiety-neurosis. I will only observe in this connection that neither shock nor anxious expectation is necessarily always demonstrable in the ætiology of such cases; any other emotion will serve as well. If I hurriedly call to mind a few cases from recollection there occurs to me that of a man of forty-five who first had an anxiety-attack (with cardiac collapse) on receiving the news of the death of his aged father: from that time onwards a complete and typical anxiety-neurosis with agoraphobia developed; further, a young man who fell a victim to the same neurosis on account of the disagreements between his young wife and his mother, and developed the agoraphobia afresh after every domestic quarrel; a {112} student who was rather an idler and had his first anxiety-attacks during a period of hard cramming under the spur of paternal displeasure; a woman, herself childless, who fell ill in consequence of anxiety concerning the health of her small niece, and so on. As to the facts themselves which Löwenfeld employs against me there cannot be the slightest doubt.

It is otherwise with their interpretation! Are we then without further ceremony to accept the _post hoc ergo propter hoc_ conclusion and save ourselves all critical consideration of the raw material? Surely we are acquainted with examples enough in which the final exciting factor cannot keep up its reputation as _causa efficiens_ in the face of critical analysis? Take the relation between trauma and gout, as an example. The rôle of a trauma in provoking an attack of gout in the injured limb is probably not different from the part it plays in the ætiology of tabes and general paralysis of the insane; only in the case of gout everyone would see that it is absurd to say the trauma has 'caused' the gout, instead of provoked it. It should make us cautious when we meet with ætiological factors of such a kind---ordinary factors, I should like to call them---in the ætiology of the most varied forms of illness. Emotion, such as fright, is also an ordinary factor of this kind; fright can evoke chorea, apoplexy, paralysis agitans and what not, just as well as an anxiety-neurosis. Now I should certainly not continue to argue along these lines, that because of their prevalence the ordinary factors do not satisfy our requirements, and therefore there must be specific causes besides; it would beg the question at issue. But I am justified in concluding as follows: If a single specific cause can be proved in the ætiology of all or almost all cases of anxiety-neurosis, then our {113} view will not necessarily be erroneous because the outbreak of illness only occurs under the influence of some one or other ordinary factor, such as emotion.

So it was in my cases of anxiety-neurosis. The man who fell ill---inexplicably---after receiving the news of his father's death (I make this interpolated comment because this death was not unexpected and did not occur in unusual or tragic circumstances) had for eleven years practised coitus interruptus with his wife whom he usually endeavoured to satisfy; the young man who could not tolerate the quarrels between his wife and his mother had practised withdrawal with his young wife from the beginning, in order to save them the burden of offspring; the student who developed anxiety-neurosis through overwork, instead of cerebrasthenia as we should expect, had for three years had a relationship with a girl whom he dared not let become pregnant; the woman who had no children herself and fell ill of anxiety-neurosis owing to the illness of a niece was married to an impotent man and had never been sexually gratified; and so on. Not all these cases are equally clear nor prove my theory with equal force; but if I add them to the very considerable series of cases in which the ætiology shows nothing but the specific factor, they conform without a dissentient note to the theory I have put forward and make it possible for us to extend our ætiological comprehension beyond the boundaries which have existed hitherto.

If anyone wishes to prove to me that I have unduly neglected the significance of ordinary ætiological factors in the foregoing discussion, he must confront me with observations in which my specific factor is lacking, that is, with cases in which, although the _vita {114} sexualis_ is (more or less) normal, anxiety-neurosis breaks out after a mental shock. Now let us see whether Löwenfeld's case fulfils this condition. My worthy opponent has manifestly not clearly recognized this necessity in his own mind, otherwise he would not have left us so completely in the dark concerning the _vita sexualis_ of his patient. I will leave on one side the fact that the case of the lady of thirty is obviously complicated by an hysteria, about the psychical origin of which I should be the last to have any doubt; I naturally admit without protest the presence of an anxiety-neurosis alongside this hysteria. But before I make use of a case as evidence either for or against my theory of the sexual ætiology of the neuroses I must study the sexual conduct of the patient more thoroughly than Löwenfeld has done here. I should not be contented with the conclusion that, because the lady had the mental shock shortly after a confinement, coitus interruptus had probably played no part during the previous year and that therefore no sexual noxiæ had arisen here. I know of cases of women who become pregnant year after year and develop anxiety-neurosis because---_incredibile dictu_---all sexual relations ceased after pregnancy had set in; so that in spite of having many children they might have lived all the years of their married life in sexual privation. It is known to every physician that women do conceive from men whose potency is very feeble and who are not able to give them satisfaction; and finally there are many women who are afflicted with a congenital anxiety-neurosis, that is to say, are endowed with a _vita sexualis_ (or develop it without demonstrable external provocation) of the kind usually acquired by coitus interruptus and similar noxiæ---a consideration which upholders of the ætiology of heredity ought to take into account. In a number {115} of these women we are able eventually to discover a past history of a hysterical illness in their youth, which has affected the _vita sexualis_ ever since or has established a deflection of sexual tension away from the psychical field. Women with this kind of sexuality are incapable of real satisfaction, even in normal coitus, and develop anxiety-neurosis either spontaneously or after further effective factors have supervened. Which of all these factors mentioned could have contributed to Löwenfeld's case I do not know; but I repeat that this case is evidence against me only if the lady who responded to a single fright with an anxiety-neurosis had previously enjoyed a normal _vita sexualis_.

It is impossible to pursue ætiological research by means of the anamnesis of patients if we accept the patient's account as he gives it or are satisfied with what he volunteers. If syphilidologists allowed themselves to depend upon the declarations of their patients, when endeavouring to trace an initial infection of the genitalia back to an act of sexual intercourse, they would have to attribute an imposing number of chancres to a chill in persons who protest their virginity, and gynæcologists would have little difficulty in confirming the miracle of parthenogenesis among their unmarried clients. I hope that some day the idea will gain admittance that neuropathologists too may be subject to similar ætiological prejudices in the work of collecting the case-histories of the great neuroses.

2. Löwenfeld further says that he has repeatedly seen anxiety-states appear and disappear where a change in the sexual life had certainly not taken place, but where there were other factors in play. {116}

I too have had exactly the same experience, but I was not misled by it. I too have brought anxiety-attacks to an end by psychical treatment, improvement of the patient's general condition and so on. I have naturally not concluded from this that lack of treatment was the cause of anxiety-attacks. Not that I would credit Löwenfeld with a conclusion of this kind, to be sure; my joking remark is only intended to show that the state of affairs may easily be so complicated as to nullify his objection completely. I have not found any difficulty in harmonizing the fact brought forward here with the theory of a specific ætiology of anxiety-neurosis. It will readily be granted that ætiological factors exist which, in order to become effective, must act with a certain intensity (or quantity) and over a certain length of time, which therefore _summate_; the effects of alcohol are a standard example of illness produced by summation. Accordingly we must reckon with a space of time in which the specific ætiology is at work, but during which its effects are not yet manifest. During such a period the person is not yet ill, but he is disposed to a particular illness---in this case to anxiety-neurosis---and as soon as the 'ordinary' noxia supervenes it can excite the outbreak of the neurosis just as well as a further increase in operation of the specific noxia. We can also express this as follows: It is not sufficient for the specific ætiological factor to exist; a certain measure of it must also be reached, and in attaining this measure a degree of the specific noxia can be made up by an amount of 'ordinary' injurious factors. If the last is again eliminated, the level of the specific ætiological factor falls below the threshold; the symptoms recede again. The entire therapy of the neuroses rests upon the fact that the total load upon the nervous system, to which the latter is succumbing, {117} can be reduced below this level by very various influences on the combination of ætiological factors. No conclusions can be drawn from this state of things in regard to the presence or absence of a specific ætiology.

These considerations are surely indisputable and assured. Anyone who does not think them sufficient may permit the following argument to weigh with him. According to the views of Löwenfeld and many others the ætiology of anxiety states lies in heredity. Now heredity can certainly not be altered; if anxiety-neurosis is curable by treatment we (and Löwenfeld too) ought to conclude that heredity cannot account for its ætiology.

For the rest, I might have been spared defending myself against both of Löwenfeld's objections if my worthy opponent had bestowed greater attention upon my paper itself. Both objections are anticipated in it and answered (p. 92 ff.); I could only repeat here what I said there; I have even purposely analysed the same cases of illness afresh. Further, the ætiological formulæ upon which I had previously laid weight are contained in the text of my former paper. I will repeat them once again here. I maintain that: _There exists a specific ætiological factor for the anxiety-neurosis which in taking effect can be reinforced quantitatively by 'ordinary' injurious factors but cannot be replaced by them qualitatively_. Further: _This specific factor determines more than anything the type of neurosis; whether a neurotic illness occurs at all depends on the total load on the nervous system_ (_in relation to its capacity to carry the load_). As a rule neuroses are _overdetermined_; that is to say, several factors in their ætiology operate together. {118}

3. I shall have less trouble in refuting Löwenfeld's next comments because on the one hand they have little to do with my theory and on the other because they call attention to difficulties which I acknowledge. Löwenfeld says: 'The Freudian theory is totally insufficient to explain the appearance and non-appearance of individual anxiety-attacks. If anxiety states, i.e. the clinical symptoms of anxiety-neurosis, occurred solely through subcortical accumulation of somatic sexual excitation and through abnormal expenditure of the same, then everyone subject to anxiety states would have an attack from time to time so long as no alteration took place in his sexual life, just as the epileptic has his attack of _grand et petit mal_; but this, as every-day experience shows, is by no means so. Anxiety-attacks occur for the most part only on definite occasions. If the patient avoids these or knows how to paralyse their influence by means of any precaution, then he remains exempt from anxiety-attacks, whether he habitually practises coitus interruptus or abstinence or enjoys a normal _vita sexualis_.'

Now there is a great deal to be said about this. In the first place Löwenfeld forces my theory to a conclusion which is not inherent in it. That the process must be the same with an accumulation of somatic sexual excitation as with an accumulation of the stimuli resulting in an epileptic convulsion is a far too detailed proposition, for which I have given no occasion; and it is not the sole one which offers itself. I need only assume that the nervous system possesses the power to master a certain amount of somatic sexual excitation even when this is deflected from its purpose, and that disturbances then occur only if the _quantum_ of this excitation suddenly undergoes an increase---and Löwenfeld's claim would be quashed. I have not ventured {119} to extend my theory far in that direction; chiefly because I did not expect to find reliable support on the way. I will merely indicate that we should not conceive the production of sexual tension to be independent of its expenditure; that in normal sexual life this production, when stimulated by the sexual object, takes a form essentially different from that which it has in psychic repose, and so on.

It must be admitted that the conditions here are in all probability different from those obtaining in the tendency to epileptic convulsion, and that they are not yet wholly correlated with the theory of an accumulation of somatic sexual excitation.

To Löwenfeld's further assertion---that anxiety-attacks only appear under certain conditions and fail to appear when these are avoided, whatever the _vita sexualis_ of those concerned---we may urge in contradiction that Löwenfeld manifestly has in mind only the anxiety of _phobias_, as is shown by the examples appended to the part of his essay which I have quoted. He says not a word concerning spontaneous anxiety-attacks, taking the form of dizziness, palpitation, dyspnœa, trembling, sweating, etc. My theory, however, seems quite equal to explaining the appearance and non-appearance of these anxiety-attacks. The semblance of periodicity in the onset of anxiety states may be found in a great number of such cases of anxiety-neurosis, similar to that observed in epilepsy, only that here the mechanism of this periodicity is more perspicuous. On closer examination we find with great regularity an exciting sexual occurrence (that is, one capable of releasing somatic sexual excitation), recurring at definite and often quite constant intervals of time, to which the anxiety-attack is related. In abstinent women this part is played by menstrual {120} excitation, and in both men and women by recurrent nocturnal pollutions and above all by sexual intercourse itself (injurious when it is incomplete), which transfers its own periodicity to the anxiety-attacks resulting from it. If anxiety-attacks occur apart from the usual periodicity, it is usually possible to trace them back to an occasional cause of more rare and irregular incidence, to a single sexual experience, something read, a visual impression, or the like. The interval I referred to varies from a few hours up to two days: it is the same as that which in other persons is followed by the well-known sexual migraine, due to the same causes, which has an undoubted connection with the symptom-complex of anxiety-neurosis.

Besides these there are plenty of cases in which a single anxiety state is provoked by an ordinary factor, i.e. by excitement of any kind. The same mechanism of displacement thus holds good for the ætiology of a single anxiety-attack as for the causation of a whole neurosis. That the anxiety of phobias answers to different conditions is not very remarkable; phobias have a more complicated structure than purely somatic anxiety-attacks. In them the anxiety is connected with a definite ideational or conceptional content, and the chief condition for the development of this anxiety comes into being when this psychical content is aroused. The anxiety is then 'released', just as, for example, sexual tension is released by the awakening of libidinous ideas; this process, however, is not yet clearly elucidated in its relation to the theory of anxiety-neurosis.

I see no reason why I should try to conceal the gaps and weaknesses in my theory. The main point in the problem of phobias seems to me that _phobias do not occur at all when the vita sexualis is normal_, that {121} is, when the specific determinant is absent; by this specific determinant we mean a disturbance of the _vita sexualis_ by deflection of the somatic away from the psychical field. However obscure otherwise the mechanism of phobias may be, my theory is only to be gainsaid by evidence of phobias occurring together with a normal _vita sexualis_, or even together with some disturbance of it that is not specific.

4. I now pass on to a remark made by my esteemed critic which I cannot leave uncontradicted. I had written in my essay on anxiety-neurosis (loc. cit. p. 86):

'In many cases of anxiety-neurosis no ætiology is recognizable at all. It is remarkable that in such cases evidence of a grave hereditary taint is seldom difficult to establish. But where there are grounds for regarding the neurosis as an acquired one, careful enquiry directed to that end reveals a series of injurious conditions (noxiæ) and influences within the sexual life as important factors in the ætiology...' Löwenfeld quotes this paragraph and adds the following comment: 'According to this Freud always appears to regard a neurosis as "acquired" whenever exciting causes are discoverable.'

If this meaning is readily conveyed by my text then the latter gives a very distorted expression of my thoughts. Let me point out that in the preceding paragraphs I have shown myself far stricter than Löwenfeld in assessing the importance of exciting causes. If I myself were to explain the meaning of my statement I should do so by adding to it, after the conditional: '_But where we have ground for regarding the neurosis as acquired_', the phrase: '_because the evidence_ (mentioned in the previous sentence) _of hereditary taint is not forthcoming_.' The meaning is: I hold a case to be acquired in which heredity is not demonstrable. I behave in this respect just as everyone else does, {122} perhaps with the slight difference that others will also explain a case as conditioned by heredity where no heredity exists, so that they overlook the entire category of acquired neuroses. But this difference runs in my favour. I admit however that I am myself to blame for this misunderstanding on account of the way I expressed myself in the first sentence: 'an ætiology cannot be found at all.' I shall also certainly be told that I have created useless trouble for myself by searching for specific causes of the neuroses. It will be said that the real ætiology of anxiety-neurosis, as of the neuroses in general, is already known to be that of heredity, and two real causes cannot exist side by side. Am I prepared to deny the ætiological rôle of heredity? If not, all other ætiologies must be equally valid or equally invalid---merely exciting causes.

I do not share this view concerning the ætiological rôle of heredity and, since on my short paper on anxiety-neurosis I have dealt less with this theme than with any other, I shall try to some extent to make good this omission and to efface the impression that in the composition of my paper I had not given equal consideration to all the inter-related problems.

I think we can effect a presentation of the probably very complicated ætiological conditions which exist in the pathology of the neuroses, if we establish the following ætiological concepts:

(_a_) _Predisposition_, (_b_) _Specific Cause_, (c) _Contributory Cause_ and, as a term not equivalent to the former (_d_) _Exciting_ or _Releasing Cause_.

In order to satisfy all possibilities let us assume that we are dealing with ætiological factors capable of quantitative alterations, and consequently of increase or decrease. {123}

If we may use the conception of a compound ætiological formula which must be fulfilled if the effect is to take place, then we may designate as exciting or releasing cause that which last makes its appearance in the formula, so that it immediately precedes the manifestation of the effect. It is this temporal element alone which constitutes the essence of an incitement; each of the other factors can in individual cases play the part of an incitement, and this part can even alternate within this same ætiological conglomeration.

The factors which are to be described as predisposition are those in whose absence the effect would never come about; but which, however, are incapable of alone bringing about the effect, no matter to what degree they may be present. For the specific cause is lacking.

The _specific cause_ is one which is never absent when the effect actually takes place, and which also suffices, in the required quantity or intensity, to bring about the effect, provided that the predisposition is present as well.

As _contributory causes_ we may comprehend such factors as are not necessarily present every time nor able in any degree to produce the effect alone, but which co-operate with the predisposition and the specific ætiological cause to make up the ætiological formula.

The peculiar position of the contributory or auxiliary causes seems clear; but how are we to distinguish between predisposition and specific causes, since both are indispensable and no one of them alone is sufficient as a cause?

In these circumstances the following considerations would seem to make a decision possible. Among the 'indispensable causes' we find several which are also present in the ætiological formulæ of many other {124} conditions beside anxiety-neurosis, thus showing that they have no particular relation to individual disorders; _one_ of these causes, however, stands out with special prominence, in that it is found in no other or in very few ætiological formulæ, and this has a claim to be called the _specific_ cause of the disease in question. Further, predisposing factors and specific causes are particularly clearly distinguished in cases where the former have the quality of long duration and little alteration in their condition, whereas the specific cause corresponds to a factor which has recently come into action.

I will attempt to give an example of this complete ætiological scheme:

_Effect_: Phthisis pulmonum.

_Predisposition_: For the most part an hereditary disposition of the organs concerned.

_Specific Cause_: Koch's bacillus.

_Contributory Causes_: Everything that lowers resistance; emotion as well as infections or colds.

The scheme for the ætiology of anxiety-neurosis seems to me to run similarly, thus:

_Predisposition_: Heredity.

_Specific Cause_: A sexual factor in the sense of a deflection of sexual tension from the psychical field.

_Contributory Causes_: All 'ordinary' injurious factors: emotion, fright, as well as physical exhaustion through illness or over-exertion.

If I go into this ætiological formula for the anxiety-neurosis in greater detail I can add the following remarks: Whether a special personal disposition (which need not necessarily be ascribed to heredity) is unconditionally required in anxiety-neurosis or whether every normal person can develop that neurosis if there should be a quantitative increase of the specific factor, I am not able to decide {125} with certainty, but I incline strongly to the latter view.---Hereditary disposition is the most important determinant of anxiety-neurosis, but not an _indispensable_ one, since it is lacking in a series of borderline cases.---The specific sexual factor is in the very great majority of cases to be demonstrated with certainty; in a series of cases (congenital) it is not distinguishable from the hereditary determinant, but comes to expression with it; that is to say, these persons are endowed with a peculiarity of the _vita sexualis_ as a stigma (being psychically unequal to the task of mastering somatic sexual tension), which in other cases must be acquired before the patient can fall ill of this neurosis. In another series of border-line cases the specific factor is contained in a contributory one, for example, if the psychical inadequacy mentioned above comes about in consequence of exhaustion, etc. All these cases fall into shifting series, not discrete categories; through all of them, however, runs one common factor---the same behaviour in regard to sexual tension, and for most of them the distinction between predisposition, specific and contributory causes is valid, in conformity with the completion of the ætiological formula given above.

When I consult my experience for the purpose I cannot find that an antithetic relation exists between hereditary disposition and the specific sexual factor in anxiety-neurosis. On the contrary, the two ætiological factors reinforce and complement each other. The sexual factor is _as a rule_ effective only with those persons who are also endowed with an hereditary taint; heredity alone is not usually able to produce an anxiety-neurosis, but waits for the incidence of a sufficient quantity of the specific sexual noxia. The existence of an {126} hereditary factor does not spare us therefore the search for a specific factor, upon which, incidentally, all therapeutic interest also depends. For where are we to begin therapeutically with an hereditary ætiology? It has always been present in the patient and will continue in him until his end is come. Taken by itself it cannot explain either the episodic outbreak of a neurosis or the cessation of one through treatment. It is nothing but a condition of the neurosis---an incalculably important one, it is true; but one that is nevertheless overestimated, to the disadvantage of therapy and theoretical comprehension. One has only to think, by way of contrast, of the cases of familial nervous diseases (chorea chronica, Thomsen's disease and so on) in which heredity unites in itself all the ætiological determinants.

In conclusion I desire to repeat the few statements in which I am accustomed, as a first approximation to the truth, to express the inter-relation of the different ætiological factors:

1. Whether a neurotic illness _occurs at all_ depends upon a quantitative factor, upon the total load on the nervous system in relation to its capacity for resistance. Anything which can keep this factor below a certain threshold, or bring it back below it, is effective therapeutically, since the ætiological formula is thus kept unfulfilled.

What is meant by 'total load' or 'capacity for resistance' could be explained in greater detail on the basis of certain hypotheses concerning nerve function.

2. To what _extent_ the neurosis develops depends in the first place on the measure of hereditary taint. Heredity acts like a multiplier introduced into the circuit, which increases the deflection of the needle many times. {127}

3. But what _form_ the neurosis takes---the direction of deflection---is determined only by the specific ætiological factor arising in the sexual life.

I hope that on the whole, although I am myself conscious of the many still unsolved difficulties of the subject, my theory of anxiety-neurosis will prove more fruitful for the elucidation of the neuroses than Löwenfeld's attempt to account for the same facts by reference to 'a combination of neurasthenic and hysterical symptoms in the form of attacks'.

VII {128} OBSESSIONS AND PHOBIAS: THEIR PSYCHICAL MECHANISMS AND THEIR ÆTIOLOGY[31] (1895)

I shall begin by challenging two assertions often found repeated with regard to the syndromes: obsessions and phobias. I must state, first, that they should not be included under neurasthenia proper, since the patients afflicted with these symptoms are often neurasthenics, but as often not; and secondly, that we are not justified in regarding them as the effect of mental degeneration, because they are found in persons no more degenerate than the majority of neurotics in general, because at times they improve, and indeed at times we even succeed in curing them.[32]

Obsessions and phobias are separate neuroses, with a special mechanism and ætiology which I have succeeded in demonstrating in a certain number of cases, and which, I hope, will prove similar in a large number of new cases.

As regards classification of the subject, I propose to exclude a group of intense obsessions which are nothing but memories, unaltered images of important experiences. As an example, I will cite Pascal's obsession: he always thought he saw an abyss on his left hand 'after {129} he had nearly been thrown into the Seine in his coach'. These obsessions and phobias, which might be called _traumatic_, are allied to the symptoms of hysteria.

Apart from this group we must distinguish: (_a_) obsessions proper; (_b_) phobias. The essential difference between them is the following:

Two components are found in every obsession: (1) an idea that forces itself upon the patient; (2) an associated emotional state. Now in the group of phobias this emotional state is always one of 'morbid anxiety', while in true obsessions other emotional states, such as doubt, remorse, anger, may occur in the same capacity as fear does in the phobias. I will first attempt to explain the remarkable psychological mechanism of true obsessions, a mechanism quite different from that of the phobias.

I

In many true obsessions it is quite plain that the emotional state is the chief element, since this state persists unchanged while the idea associated with it varies. The girl in Case 1 quoted below, for example, felt remorse in some degree for all sorts of reasons---for having stolen, for having ill-treated her sisters, for having counterfeited money, etc. Persons who doubt have many doubts at the same time or consecutively. In them it is the emotional state which remains constant; the idea changes. In other cases the idea, too, seems fixed, as in Case 4, the girl who persecuted servants in the household with an incomprehensible hatred, constantly changing the individual object, however.

Now a careful psychological analysis of these cases shows that the emotional state, as such, is always justified. Case 1, the girl who suffered from remorse, had good reasons for it; the women in Case 3 who doubted their resistance against temptation knew very well {130} why. The girl in Case 4, who detested servants, was justified in complaining, etc. Only, and it is in these two characteristics that the pathological sign lies, (1) the emotional state persists continually, and (2) the associated idea is no longer the appropriate original one, ætiologically related to the obsession, but is one which replaces it, a substitute for it.

The proof of this is the fact that we can always find in the previous history of the patient, at the beginning of the obsession, the original idea that has been replaced. The replaced ideas all have common attributes; they correspond to really painful experiences in his sexual life which the person is striving to forget. He succeeds merely in replacing the irreconcilable idea by another ill-adapted to the emotional state, which for its part remains unchanged. It is this incongruity between the emotional state and the associated idea that accounts for the absurdity so characteristic of obsessions. I will now bring forward my observations and conclude with a tentative theory and explanation.

_Case 1_. A girl reproached herself for things which she knew were absurd, for having stolen, for having counterfeited money, for having dabbled in magic, etc. according to whatever she had been reading during the day.

_Reinstatement of the replaced idea_. She reproached herself with the onanism she had been practising secretly without being able to renounce it. She was cured by careful surveillance which prevented her masturbating.

_Case 2_. A young man, a medical student, suffered from an analogous obsession. He reproached himself for all sorts of immoral acts: for having killed his cousin, for having violated his sister, for having {131} set fire to a house, etc. He got to the point of having to turn round in the street to see if he had not killed the last passer-by.

_Reinstatement_. He had been much affected by reading in a quasi-medical book that onanism, to which he was addicted, destroyed one's morale.

_Case 3_. Several women complained of an impulsion to throw themselves out of the window, to cut their children with knives, scissors, etc.

_Reinstatement_. Typical temptation-obsessions.---They were women who, not being at all satisfied in marriage, had to struggle against the desires and voluptuous ideas that haunted them in the presence of men.

_Case 4_. A girl who was perfectly sane and very intelligent displayed an uncontrollable hatred against servants in the household. It had been provoked by an impertinent servant, and had been transferred from servant to servant, to a degree that rendered housekeeping impossible. The feeling was a mixture of hate and disgust. As a reason for it she stated that the vulgarity of these girls sullied her idea of love.

_Reinstatement_. This girl had accidentally been witness of an amorous scene in which her mother had taken part. She had hidden her face, had stopped up her ears, and had done her utmost to forget it, as it disgusted her and made her feel quite unable to remain with her mother whom she loved tenderly. She succeeded in her efforts; but her anger at the person who had profaned her idea of love continued to exist within her, and this emotional state soon linked itself to the image of a person who could be a substitute for her mother.

_Case 5_. A young girl had become almost completely isolated on account of an obsessive fear of incontinence of urine. She could no {132} longer leave her room or receive visitors without having urinated a number of times. When at home or entirely alone the fear did not trouble her.

_Reinstatement_. It was an obsession of temptation or mistrust. She did not mistrust her bladder, but her resistance against amorous inclinations. The origin of the obsession shows this well. She had once in a theatre, at the sight of a man who pleased her, felt an amorous desire, accompanied (as spontaneous pollutions in women always are) by the desire to urinate. She was obliged to leave the theatre, and from that moment on she was a prey to the fear of reproducing the same sensation, but the desire to urinate had replaced the amorous one. She was completely cured.

Although the cases enumerated show varying degrees of complexity, they have the following in common: the original (intolerable) idea has been replaced by another idea, the substituted idea. In the cases which I now append the original idea has also been replaced, but not by another idea; it has been replaced by acts or impulses which originally served as measures of relief or as _protective_ procedures, and are now incongruously associated with an emotional state that does not fit them, but that has persisted in its original form and was also originally justified.

_Case 6_. Obsession of arithmomania.---A woman became obliged to count the boards in the floor, the steps in the staircase, etc.---acts which she performed in a state of ridiculous distress.

_Reinstatement_. She had begun the counting in order to turn her mind from obsessive ideas of temptation. She had succeeded in so doing, but the impulse to count had replaced the original obsession. {133}

_Case 7_. Obsessive brooding and speculating.---A woman suffered from attacks of this obsession that ceased only when she was ill, and then gave place to hypochondriacal fears. The theme of her worry was always a part or function of her body, for example, respiration: Why must I breathe? What if I try not to breathe? etc.

_Reinstatement_. At the very beginning she had suffered from the fear of becoming insane, a hypochondriacal phobia common enough among women who are not satisfied by their husbands, as she was not. To assure herself that she was not going mad, that she still possessed her mental faculties, she had begun to catechize herself and busy herself with serious problems. This quieted her at first, but with time the habit of speculation replaced the phobia. For more than fifteen years periods of fear (pathophobia) and of obsessive speculating had alternated in her.

_Case 8_. _Folie du doute_. Doubting mania.---Several cases showed the typical symptoms of this obsession but were explained very simply. These persons had suffered or were still suffering from various obsessions, and the knowledge that the obsessions had disturbed all their acts and had many a time interrupted their train of thought provoked a legitimate doubt about the reliability of their memory. The confidence of each one of us would be shaken, and we should all of us have to re-read a letter or repeat a calculation if our attention had been distracted several times during the performance. Doubt is a quite logical result when obsessions are present.

_Case 9_. _Folie du doute_ (Hesitation).---The girl in Case 4 had become extremely slow in performing all ordinary everyday actions, particularly in her toilet. She took hours to tie her shoe-laces or to clean her finger-nails. As an explanation she stated that she could {134} not make her toilet while obsessive ideas occupied her, nor immediately after. As a result, she had become accustomed to wait a definite period after each return of the obsessive idea.

_Case 10_. _Folie du doute_. Fear of paper.---A young woman had suffered scruples after having written a letter; at the same time she collected all the pieces of paper she saw, which she explained by confessing to a love which she had formerly refused to admit. As she was constantly repeating her lover's name, she was seized with a fear that the name might have slipped off the end of her pen, that she might have written it upon some bit of paper in a pensive moment.[33]

_Case 11_. Mysophobia.---A woman who washed her hands constantly and touched door-handles only with her elbow.

_Reinstatement_. The case of Lady Macbeth. The washing was symbolic, designed to replace by physical purity the moral purity which she regretted having lost. She tormented herself with remorse for conjugal infidelity, the memory of which she had resolved to banish from her mind. She also washed her genitalia.

As regards the theory and explanation of this substitution, I will content myself with answering three questions that arise here.

1. How does this substitution come about?

It seems to be the expression of a special mental disposition. At least, a 'similar heredity' is often enough found in obsessional cases, as in hysteria. Thus the patient in my second case related to me that his father had suffered from similar symptoms. He once {135} introduced me to a second cousin who had obsessions and convulsive tic, and to his sister's daughter, aged eleven, who already showed obsessions (probably of remorse).

2. What is the motive for this substitution?

I think it may be regarded as a defensive reaction (_Abwehr_) of the ego against the intolerable idea. Among my patients several remember a deliberate effort to banish the idea or the painful recollection of the voice of conscience. (See Cases 3, 4, II.) In other cases the repugnance is banished by an unconscious process that has left no trace in the patients' memory.

3. Why does the emotional state that is attached to the obsessive idea persist instead of vanishing like other conditions of the ego?

This question may be answered by applying to it the theory of the genesis of hysterical symptoms developed by Breuer and myself.[34] I will here only remark that by the very fact of the substitution the disappearance of the emotional state is rendered impossible.

II

In addition to these two groups of true obsessions there is the class of 'phobias', which must now be considered. I have already mentioned the great difference between obsessions and phobias: that in the latter the emotion is always one of anxiety, fear. I might add that obsessions are varied and more individualized, phobias are more uniform and typical. But this distinction is not all-important.

Among the phobias two groups may be differentiated, according to the nature of the object feared: (1) common phobias, an exaggerated fear {136} of all those things that everyone detests or fears to some extent: such as night, solitude, death, illness, dangers in general, snakes, etc.; (2) specific phobias, the fear of special circumstances that inspire no fear in the normal man; for example, agoraphobia and the other phobias of locomotion. It is interesting to note that these phobias have not the obsessive feature that characterizes true obsessions and the common phobias. The emotional state appears in them only under special conditions which the patient carefully avoids.

The mechanism of phobias is entirely different from that of obsessions. Substitution is no longer the predominant feature in the former; psychological analysis reveals no intolerable replaced idea in them. Nothing is ever found but the _anxiety state_ which, by a sort of selection, brings up all the ideas adapted to become the subject of the phobia. In the case of agoraphobia, etc., we often find the recollection of a state of _panic_; and what the patient actually fears is a repetition of such an attack under those special conditions in which he believes he cannot escape it.

The fear of this emotional state, which underlies all phobias, is not derived from any memory whatever; we must ask what the source of this overpowering nervous state can be.

I hope to be able to demonstrate, on another occasion, that there is some reason for distinguishing a special neurosis, the anxiety-neurosis,[35] of which the chief symptom is this emotional state. I shall then enumerate its various symptoms and lay stress on those which differentiate this neurosis from neurasthenia, with which it is now confounded. Phobias, then, are a part of the {137} anxiety-neurosis, and are almost always accompanied by other symptoms belonging to it.

The anxiety-neurosis has a sexual origin too, as far as I can see, but it does not attach itself to ideas taken from sexual life; properly speaking, it has no psychical mechanism. Its specific cause is the accumulation of sexual tension, produced by abstinence or by frustrated sexual excitation (using the term as a general formula for the effects of coitus reservatus, of relative impotence in the husband, of excitation without satisfaction in engaged couples, of enforced abstinence, etc.).

It is under such conditions, extremely frequent in modern civilized society, especially among women, that anxiety-neurosis develops; phobias are a psychical manifestation of it.

In conclusion I will state that combinations of a phobia and an obsession proper may co-exist, and that indeed this is a very frequent occurrence. We may find that a phobia had developed at the beginning of the disease as a symptom of anxiety-neurosis. The thought-content of the phobia accompanying the state of fear may be replaced by another idea or perhaps by a _protective procedure_ that lessens the fear. Case 7 (obsessive speculating) presents a neat example of this group, a phobia coupled with a true obsession evolved by substitution.

VIII {138} HEREDITY AND THE ÆTIOLOGY OF THE NEUROSES[36] (1896)

I am addressing the pupils of J. M. Charcot especially to bring to their attention some objections to the theory concerning the ætiology of the neuroses which was passed on to us by our master.

The part ascribed to neurotic heredity in this theory is well known. It is the only true and indispensable cause of nervous disease; other ætiological factors may aspire only to the title of precipitating causes. Thus the master himself and his pupils, Guinon, Gilles de la Tourette, Janet and others, have promulgated this doctrine for the major neurosis, hysteria; I believe that the same opinion is held in France and to some extent everywhere for the other neuroses, although it has not been laid down quite so solemnly and decidedly for conditions analogous to hysteria.

For a long time I have entertained suspicions in this matter, but I had to wait to find corroborating facts in my daily medical experience. My objections are now of two kinds, arguments based upon facts and arguments of a more speculative order. I will begin with the former, arranging them according to the importance that I ascribe to them.

I. (_a_) Sometimes diseases which are often enough foreign to the sphere of neuropathology and which do not necessarily depend upon disease of the nervous system have been considered nervous and {139} indicative of neuropathic heredity. Thus, true facial neuralgias and many headaches have been considered nervous which are caused rather by post-infectious pathological changes and by suppuration in the nasal sinuses. I am convinced that these patients would profit if we were more often to leave the treatment of these conditions to rhinologists.

(_b_) All nervous diseases found in the patient's family, without reference to their frequency or their gravity, have been accepted as justifying the accusation of nervous hereditary taint. Does not this point of view seem to imply a sharp distinction between families that are free from any nervous predisposition and families that are subject to it without limit or restriction? And do not the facts speak rather in favour of the opposite opinion, namely, that there are transitions and degrees of nervous predisposition and that no family escapes it entirely?

(_c_) Our opinion as to the ætiological rôle of heredity in nervous disease must assuredly be the result of an impartial statistical study and not of a _petitio principii_. Until this study has been made, the existence of acquired neuropathies should be considered as probable as that of hereditary neuropathies. But if the existence of neuropathies acquired by persons not predisposed to them is possible, it can no longer be denied that some of the nervous diseases found among the relatives of our patients may have such an origin. They can then no longer be called upon as conclusive proofs of the hereditary predisposition that is imputed to the patient by virtue of his family history, for a retrospective diagnosis of the diseases of ancestors or of absent members of the family is only very rarely successful. {140}

(_d_) Those who are adherents of the view of Fournier and of Erb, concerning the part played by syphilis in the causation of tabes dorsalis and in progressive paralysis, have learned that powerful ætiological factors must be recognized the presence of which is indispensable in the pathogenesis of certain diseases that heredity alone could not cause. Charcot remained to the end, however, as I learned through a private letter from him, strictly opposed to Fournier's theory, which, nevertheless, gains ground day by day.

(_e_) There is no doubt that certain neuropathies may develop in a man who was perfectly sound and of untainted family. One observes this every day in Beard's neurasthenia; if neurasthenia were limited to predisposed people, it would never have gained the importance and prevalence with which we are familiar.

(_f_) In nervous pathology there is a so-called similar and a so-called dissimilar heredity. No fault will be found with the former concept; it is even very remarkable that in the affections that depend upon similar heredity (Thomsen's disease, Friedreich's disease, Huntington's chorea, the myopathies, etc.), no trace of any accessory ætiological factor is ever found. But the much more important concept of dissimilar heredity has gaps that must be filled out to arrive at a satisfactory solution of the ætiological problems. They consist of the fact that members of the same family prove to be afflicted with the most varied forms of nervous disease, functional and organic, without our being able to discover a law that determines the substitution of one disease for another or the order in which they follow one another in successive generations. There are in these families, beside the affected members, persons who remain sound. The theory of dissimilar {141} heredity tells us neither why one person carried the same hereditary taint without succumbing to it, nor why another selects, among the diseases that constitute the great family of neuropathies, one particular nervous affection instead of another, hysteria instead of epilepsy, insanity, etc. Since there is no such thing as chance in the pathogenesis of nervous disease any more than elsewhere, it must be admitted that it is not heredity that controls the choice of a neuropathy developing in a member of a predisposed family, and that there is ground for suspecting the existence of other ætiological factors of a less incomprehensible nature which deserve to rank as the specific ætiology of any such nervous disease. Without the existence of this special ætiological factor heredity would have been powerless; it would have lent itself to the production of a different neuropathy if the specific ætiology in a given case had been replaced by any other.

II\. These specific, determining causes of neuropathy have been too little investigated, for the vision of physicians has been dazzled by the imposing prospect of the hereditary factor in ætiology. Nevertheless, they are well worth making the object of careful study; although their pathogenic power may be in general only auxiliary to that of heredity, great practical interest centres on the understanding of this specific ætiology. It affords a point of attack for our therapeutic efforts, while hereditary predisposition, predetermined for the patient from birth, frustrates all our efforts by the strength of its position.

I have been engaged for years in a study of the ætiology of the major neuroses (functional nervous states analogous to hysteria), and in what follows I shall report the result of these studies. In order to {142} avoid all possible misunderstanding, I shall begin by making two observations on the nosography of the neuroses and on their ætiology in general.

I had to start my work with an innovation in nosography. I have found reason to place the obsessional neurosis upon the same footing as hysteria, as an independent and autonomous disease, although the majority of authors classify obsessions among the syndromes comprising mental degeneration or else confound them with neurasthenia. I have learned, by examining their mental mechanism, that these disorders are much more closely allied to hysteria than one would have believed.

Hysteria and the obsessional neurosis form the first group of the major neuroses that I studied. The second contains Beard's neurasthenia, which I split up into two functional conditions differentiated both ætiologically and symptomatologically: neurasthenia proper and anxiety-neurosis---a term, by the way, which does not altogether please me. I have given detailed reasons for this differentiation, which I consider necessary, in a paper published in 1895.[37]

As for the ætiology of the neuroses, I think we should recognize that the various ætiological factors, differing in their importance and in the way in which they are related to the effect they produce, may be arranged theoretically in three classes: (1) _Conditions_ that are indispensable for the production of the disease in question, but which are of a universal nature and occur as well in the ætiology of many other diseases. (2) _Concurrent causes_ that share with the conditions in group (1) the characteristic of entering into the causation of other diseases as well as into that of the disease in question, but which are not indispensable for the production of the latter {143} disease. (3) _Specific causes_ just as indispensable as the first conditions but appearing only in the ætiology of the disease for which they are specific.

Now, in the pathogenesis of the major neuroses heredity plays the part of a _condition_, potent in all cases and even indispensable in the majority of them. It cannot do without the assistance of the specific causes; but the importance of hereditary predisposition is demonstrated by the fact that the same specific causes operating on a sound person would produce no manifest pathological effect, while its presence in a predisposed person will precipitate a neurosis, the development of which will in intensity and extent be proportional to the degree of hereditary predisposition.

The action of heredity is comparable to that of a multiplier in an electric circuit, which increases the visible deviation of the needle but which cannot determine its direction.

There is another point to be noted in the relations between the hereditary condition and the specific causes of the neuroses. Experience shows, as might have been anticipated, that among the problems of ætiology that of the quantitative relationship of the ætiological factors to one another should not be neglected. But one would not have guessed the fact which seems to follow from my observations, that heredity and the specific causes may replace one another quantitatively, that the same pathological effect will be produced by the co-existence of a very grave specific ætiology and a moderate degree of predisposition as by that of a severe neuropathic heredity with a slight specific factor. So that it is merely a quite possible extreme in this series when one finds cases of neurosis in which a tangible degree of hereditary predisposition is looked for in {144} vain, provided that this deficiency is compensated for by a powerful specific factor.

As _concurrent or accessory causes_ all those 'ordinary' factors found elsewhere may be mentioned: mental excitement, physical exhaustion, acute illness, intoxication, traumatic accidents, intellectual overwork, etc. I would maintain that none of them, not even the last, enters regularly or necessarily into the ætiology of neuroses, well knowing that the enunciation of this opinion is in direct opposition to a theory regarded as universal and incontestable. Since Beard declared neurasthenia to be the product of our modern civilization he has found only believers; but it is impossible for me to accept this opinion. A painstaking study of the neuroses has taught me that the specific ætiology of the neuroses has escaped Beard's attention.

I do not wish to underestimate the ætiological importance of these every-day factors. They are very varied, of frequent occurrence, and most often blamed by the patient; they are more manifest than the _specific causes_, which are either concealed or not understood. They often fulfil the function of precipitating causes bringing the hitherto latent neurosis to the surface; and a practical interest attaches to them, since consideration of these ordinary causes may furnish a basis for a therapy that does not aim at a radical cure, but contents itself with reducing the disease to its previous latent state.

But a constant and intimate relation between a given ordinary cause and a given nervous affection cannot be established; mental excitement, for example, is found as often in the ætiology of hysteria, of obsessions, of neurasthenia, as in that of epilepsy, Parkinson's disease, diabetes, and numerous others. {145}

The ordinary concurrent causes can also replace the specific ætiology quantitatively, but can never completely supplant it. There are numbers of cases in which hereditary predisposition and the specific factor constitute the entire ætiology, the ordinary causes being absent. In other cases the indispensable ætiological factors are not strong enough in themselves to precipitate the neurosis and a state of apparent good health may be maintained for a long time, a state that is really one of neurotic predisposition. It requires only the added effect of an ordinary cause to make the neurosis become manifest. It must be noted, however, that in such circumstances the nature of the supervening ordinary factor, whether it be excitement, trauma, infectious disease, or any other, is entirely a matter of indifference; the pathological result does not vary in accordance with it, and the nature of the neurosis will always be determined by the pre-existing specific cause.

What then are the specific causes that produce neuroses? Is there only one or are there several? And can a constant ætiological relationship be established between a given cause and a given neurosis so that each major neurosis may be referred to a particular ætiology?

Supported by a painstaking investigation of the facts, I wish to affirm that this hypothesis agrees well with the truth, that each of the major neuroses mentioned has as its immediate cause a special disturbance of the nervous economy, and that these pathological functional changes _betray, as their common source, the sexual life of the person concerned, either a disturbance of his present sexual life or important events in his past life_.

This is, strictly speaking, not a new, unheard-of proposition. It has always been admitted that sexual disturbances were among the causes of {146} nervousness, but they were grouped together on a level with other precipitating causes as subordinate to heredity; their ætiological importance was restricted to a limited number of the cases studied. Physicians had even acquired the habit of not looking for them if the patient did not himself accuse them. The distinctive characteristic of my point of view lies in my raising the sexual factors to the rank of specific causes, in my recognizing their influence in all cases of neurosis, and finally in my finding a constant parallelism, proof of a special ætiological relationship, between the nature of the sexual factor and the kind of neurosis.

I am quite sure that this theory will provoke a storm of contradiction from my fellow physicians. But this is not a fitting occasion to present the data and observations that have imposed my conviction on me, nor to explain the strict meaning of the somewhat vague term 'disturbances of nervous economy'. This will be done, I hope as thoroughly as possible, in a work that I am preparing on the subject. In this paper I shall confine myself to the presentation of my results.

Pure neurasthenia, which after it has been differentiated from anxiety-neurosis presents a monotonous clinical picture (exhaustion, sense of pressure on the head, flatulent dyspepsia, constipation, spinal paræsthesias, sexual weakness, etc.), admits of only two specific ætiological factors, excessive onanism and spontaneous emissions.

The prolonged and intense effect of this pernicious form of sexual satisfaction alone suffices to produce neurasthenia, or else it stamps the subject with the specific neurasthenic habitus which later is made manifest under the influence of a chance accessory cause. I have also {147} met persons who showed the signs of a neurasthenic constitution, but in whom I did not succeed in demonstrating the ætiology just mentioned; I was able at least to prove, however, that in these patients the sexual function had never developed to a normal degree. They seemed to be endowed by heredity with a sexual organization analogous to that which arises in the neurasthenic as a result of onanism.

The anxiety-neurosis, the clinical picture of which is much richer (irritability, states of anxious expectation, phobias, more or less severe panics, fear, vertigo, trembling, sweats, congestion, dyspnœa, tachycardia, etc.; chronic diarrhœa, chronic locomotor vertigo, hyperæsthesia, insomnia, etc.)[38] is easily shown to be the specific result of various disturbances of sexual life, all having a common trait. Forced abstinence, frustrated sexual excitement (not gratified by sexual intercourse), incomplete or interrupted coitus (not attaining gratification), sexual efforts that exceed the psychical capacity of the person, etc., all these factors, all too common in modern life, seem to unite in disturbing the equilibrium of psychical and somatic functions in sexual activity, and in hindering the psychical co-operation necessary to relieve the nervous economy from sexual tension.

These considerations, which contain perhaps the nucleus of a theoretical explanation of the functional mechanism of the neurosis in question, arouse the suspicion that a complete and really scientific demonstration of the matter is not at the moment possible, and that the problem of the physiology of the sexual life will first have to be approached from a new angle. {148}

I will conclude by stating that the presence of hereditary predisposition is not indispensable in the pathogenesis of neurasthenia and of the anxiety-neurosis. This is the result of daily observation; if the hereditary factor is present, however, the development of the neurosis will be strongly influenced thereby.

In the second class of major neuroses, hysteria and the obsessional neurosis, the solution of the ætiological problem is surprisingly simple and uniform. I owe my conclusions to the use of the new psycho-analytic[39] method, the probing procedure of J. Breuer, a method that is somewhat subtle but irreplaceable, so fruitful has it proved to be in explaining obscure unconscious mental processes. By means of this method, which cannot be explained here,[40] hysterical symptoms are traced to their origin, which invariably proves to be an experience in the person's sexual life well adapted to produce a painful emotional reaction. Going back into the patient's life step by step, guided always by the structural connection between symptoms, memories, and associations, I finally came to the starting-point of the pathological process; and I had to realize that the same factor was at the bottom of all the cases subjected to analysis, namely, the effect of an agent that must be accepted as the specific cause of hysteria.

It is indeed a memory connected with the person's sexual life, but one that presents two extremely important features. The event, the unconscious image of which the patient has retained, is a premature sexual experience with actual stimulation of the genitalia, the result of sexual abuse practised by another person, and the period of life in {149} which this fateful event occurs is early childhood, up to the age of eight to ten, before the child has attained sexual maturity.

_A passive sexual experience before puberty:_ this is the specific ætiology of hysteria.

I will at once supplement these conclusions by some detailed facts and some explanatory remarks, so as to combat the doubt that I anticipate. I have been able to analyse thirteen cases of hysteria completely. Three of this number were true combinations of hysteria with an obsessional neurosis (I do not say hysteria with obsessions). The experience mentioned above was not lacking in a single case; it was present either as a brutal attempt committed by an adult or as a less sudden and less repugnant seduction, having however the same result. In seven cases out of the thirteen we were dealing with a liaison between children, sexual relations between a little girl and a boy slightly older, generally her brother who had himself been the victim of an earlier seduction. These liaisons were sometimes continued for years, up to puberty, the boy repeating upon the little girl without alteration those practices that he had himself experienced at the hands of a servant or governess; because of this origin they were often of a disgusting kind. In some cases there had been both assaults and an infantile liaison or repeated brutal abuse.

The date of the premature experience was variable: in two cases it went back to the second year (?) of the child; in my observations the age of predilection was the fourth or fifth year. It may be a matter of chance, but I have the impression from them that a passive sexual experience occurring after the age of eight or ten can no longer be the foundation of a neurosis. {150}

How can one be convinced of the truth of these confessions made in analysis and said to be memories preserved since early childhood, and how can we protect ourselves against the inclination to fabricate and the facility for invention ascribed to hysterics? I would charge myself with blameworthy credulity if I did not offer more convincing proofs. But the fact is that patients never relate these histories spontaneously, and never suddenly offer, in the course of the treatment, the complete recollection of such a scene to the physician. The mental image of the premature sexual experience is recalled only when most energetic pressure is exerted by the analytic procedure, against strong resistance; so that the recollection has to be extracted bit by bit from the patients, and while it comes back into consciousness they fall prey to emotions difficult to simulate.

If we remain uninfluenced by the patient's behaviour and are able to follow impartially, in detail, the psycho-analysis of a case of hysteria, we are finally convinced ourselves.

The premature experience in question has left a permanent imprint upon the history of the case, being represented in it by a mass of symptoms and peculiar features that permit of no other explanation; the delicate but firm inter-relationship of the structural elements of the neurosis compel us to accept this view; the therapeutic result is retarded if we do not go so deep as this; no other choice exists then except that of rejecting or believing the whole matter.

Is it to be believed that a premature sexual experience of this kind, undergone by a person whose sex is scarcely differentiated, may become the starting-point of a permanent mental abnormality like hysteria? And further, how would such a view harmonize with our present conception of the psychical mechanism of that neurosis? A satisfactory {151} reply can be given to the first question: it is just because the subject is a child that premature sexual stimulation produces little or no effect at the time, but a mental impression of it is retained. Later, at puberty, when the sexual organs have developed a degree of activity immeasurably greater than that of childhood, this unconscious mental impression is somehow or other reawakened. Owing to the changes produced by puberty the memory will exercise a power which was entirely lacking when the experience itself took place; _the memory will produce the same result as if it were an actual event_. We have, so to speak, _the subsequent effect of a sexual trauma_.

As far as I know, the reawakening of a sexual recollection after puberty, the event itself having occurred at a time antedating this period, affords the only psychological conjunction in which the effect of a memory surpasses that of the event itself. But it is an abnormal constellation, which attacks the mind at a weak point and necessarily produces a pathological result.

I believe that this inverse relation between the psychical effect of the recollection and that of the event contains the reason why the memory has remained unconscious.

We thus approach a very complex psychical problem, one which, however, if duly appreciated promises to throw a flood of light some day upon the most intricate problems of mental life.

The ideas set forth here, based upon the results of psycho-analysis, namely, that the memory of a premature sexual experience is always found as the specific cause of hysteria, do not agree with Janet's psychological theory of the neurosis, nor with any other; but they {152} harmonize perfectly with my own hypotheses, described elsewhere, on the 'defence neuroses'.

All events occurring after puberty to which an influence on the development of hysteria and the formation of its symptoms must be ascribed are only concurrent causes, '_agents provocateurs_' in the words of Charcot, to whom neuropathic heredity occupied the position that I claim for premature sexual experiences. These accessory factors are not subject to the strict rules that govern the specific causes; analysis demonstrates beyond question that they produce a pathological effect in hysteria only by virtue of their ability to awaken the unconscious mental impression of the infantile experience. It is also because of their connection with the original pathogenic impression and by its magnetism, so to speak, that the memory of them becomes unconscious in turn, and is able to assist the growth of a mental process withdrawn from the influence of conscious processes.

The obsessional neurosis arises from a specific cause closely analogous to that of hysteria. A premature sexual experience, which has occurred before puberty and the memory of which becomes active during or after that period, is also found here and the same observations and arguments made in reference to hysteria apply to cases of this neurosis (six cases, three pure forms). There is only one difference that seems of capital importance. At the base of the ætiology of hysteria we found a passive sexual experience undergone with indifference or with a slight degree of disgust or fright. In the obsessional neurosis we are concerned, on the other hand, with an experience that was pleasurable, with a sexual aggression inspired by desire (in the case of the boy) or with a pleasurable participation in {153} sexual acts (in the case of the little girl). The obsessive ideas, whose inner meaning is made recognizable by analysis, reduced, so to speak, to their simplest form, are nothing but reproaches that the patient makes to himself because of that premature sexual pleasure, but reproaches disguised by an unconscious psychical work of transformation and substitution.

The very fact that this sexual aggression occurs at such a tender age seems to betray the effect of a previous seduction, which would result in precocity of sexual desire. Analysis has confirmed this suspicion in the cases I have treated. In this way an interesting fact always present in these obsessional cases is explained, namely, the regular way in which the clinical picture is complicated by a certain number of purely hysterical symptoms.

The importance of the active sexual attitude as a cause of obsessions, and that of the passive attitude in the pathogenesis of hysteria, seems to give the reason of hysteria's close association with the female sex and of the preference of the male for the obsessional neurosis. One sometimes comes across a pair of neurotics who had a liaison in early childhood, the man suffering from obsessions, the woman from hysteria; if we are dealing with brother and sister we may mistake for the result of heredity what is really derived from premature sexual experience.

There are, unquestionably, pure and isolated cases of hysteria and obsessional neurosis, independent of neurasthenia and anxiety-neurosis, but they are not the rule. More frequently, the psychoneurosis appears as accessory to the neurasthenic neuroses, having been provoked by the latter and following their decline. This happens because the specific causes of the latter, current disturbances of the sexual life, act at the same time as accessory {154} causes of the psychoneuroses, the specific cause of which, the memory of an early sexual experience, they reawaken and revive.

As for neurotic heredity, I am far from being able to estimate accurately its influence in the ætiology of the neuroses. I admit that its presence is indispensable in serious cases; I doubt whether it is a _sine qua non_ in mild cases, but I am convinced that neurotic heredity alone cannot cause a psychoneurosis if the specific ætiology of the latter, premature sexual stimulation, be lacking. I even believe that the question which of the two neuroses, hysteria or obsessional neurosis, will develop in a given case is determined not by heredity but by the particular nature of the early sexual experience in childhood.

IX {155} FURTHER REMARKS ON THE DEFENCE NEURO-PSYCHOSES[41] (1896)

In a short paper[42] published in 1894 I included hysteria, obsessions and certain cases of acute hallucinatory confusion under one heading as 'Defence Neuro-Psychoses'. I did this because one point of view showed itself as applying in common to all these affections: their symptoms arise through the psychical mechanism of (unconscious) defence, that is, through an attempt to repress an intolerable idea which was in painful opposition to the patient's ego. In a book since published by Dr. J. Breuer and myself (_Studien über Hysterie_) I have been able by quoting clinical observations to elucidate and illustrate what is meant by this psychical process of 'defence' or 'repression'. Information may also be found in it concerning the toilsome but completely reliable method of psycho-analysis which I use in making these investigations and by which at the same time the investigations serve a therapeutic purpose.

My experiences in the last two years of work have strengthened me in my surmise that defence is the nucleus of the psychic mechanism of the neuroses under discussion and have also made it possible for me to give this psychological theory a clinical basis. To my own surprise I came unexpectedly upon a few simple but narrowly definable solutions {156} of the problems contained in the neuroses and in the following paper I will give a short preliminary account of them. I cannot in this kind of communication bring forward the evidence on which my statements are based, but I hope to fulfil this obligation in a more detailed presentation.

I. THE 'SPECIFIC' ÆTIOLOGY OF HYSTERIA

In earlier publications Breuer and I have already expressed the opinion that the symptoms of hysteria can be understood only by tracing them back to 'traumatic' experiences and that these psychical traumas are related to the patient's sexual life. What I have to add here, as a uniform result of my analysis of thirteen cases of hysteria, concerns on the one hand the nature of these sexual traumas, and on the other hand the period of life in which they occur. The occurrence at any time of life of an experience in some way touching on the sexual life, which then becomes pathogenic on account of the generation and suppression of a painful affect, does not suffice to bring about a hysteria. _These sexual traumas must on the contrary occur in early childhood_ (_before puberty_) _and they must consist in actual excitation of the genital organs_ (_coitus-like processes_).

I have found this specific determinant of hysteria---_sexual passivity in the pre-sexual period_---present in all the cases of hysteria analysed (including two men). I need here only allude to the great reduction in the importance of the factor of hereditary disposition which is effected by thus establishing accidental ætiological factors as a necessary condition; moreover, we are also thus provided with a clue to the reason for the much greater frequency of hysteria in the {157} female sex, which even in childhood is more likely to provoke sexual assaults.

The more obvious objections to this theory will probably be the following: that sexual assaults on small children occur too frequently for them to be of great ætiological importance when they can be confirmed; or else that such experiences must remain ineffective for the very reason that they happen to a sexually undeveloped person; further, that one must guard against instigating patients to ostensible reminiscences of this kind by a cross-examination and guard against believing the romances which they themselves invent. In reply to the latter objections we may ask the favour that on this obscure topic no one should pass too certain a judgement before he has himself practised the only method which can throw light upon it---psycho-analysis, the method of making conscious what was previously unconscious.[43] The essential element in the first-mentioned objections is disposed of when we remember that it is not the experience itself which acts traumatically, but the memory of it when this is re-animated after the subject has entered upon sexual maturity.

My thirteen cases of hysteria were all of them severe; all of them had symptoms of many years duration, a few after long and fruitless treatment in institutions. The infantile traumas which analysis discovered in these severe cases must without exception be described as grave sexual injuries; some of them were absolutely appalling. Most prominent among the people who were guilty of these abuses with all their serious consequences were nursemaids, governesses, or domestic {158} servants, to whose care children are all too thoughtlessly abandoned, and teachers and tutors appear regrettably often; in seven of the thirteen cases, however, the assaults were perpetrated by innocent childish assailants, mostly brothers, who had for years carried on some kind of sexual relation with somewhat younger sisters. The course of events was probably in all cases similar to that which we were able to follow with certainty in some individual cases---namely, the boy was first misused by a person of the female sex, by which his libido was prematurely awakened and then a few years later he committed a sexual aggression reproducing exactly the procedure to which he himself had been subjected.

I must exclude active masturbation from the list of sexual noxiæ occurring in early childhood that are pathogenic for hysteria. Although indeed it may be very frequently found in cases of hysteria, this is due to the circumstance that masturbation is much more often the consequence of abuse or seduction than is generally supposed. It is by no means rare for both of the young couple to fall ill of a defence neurosis at a later date, the brother with obsessions, the sister with hysteria, which naturally gives the appearance of a familial neurotic disposition. This pseudo-heredity, however, is explained now and then in surprising ways; on one occasion I was able to observe a brother, sister and a somewhat older male cousin who were all ill. I learnt from the analysis which I undertook with the brother that he suffered from self-reproaches for being the cause of his sister's illness; he had been seduced by his cousin, who in his turn---as was known in the family---had himself been the victim of his nursemaid. {159}

I cannot say for certain up to what age-limit sexual injury falls within the ætiology of hysteria; but I doubt whether sexual passivity after the eighth to tenth year can evoke repression in the absence of previous experiences of the same kind. The lower limit extends as far as memory itself, that is, therefore, to the tender age of from one and a half to two years! (Two cases.) In a number of my cases the sexual trauma (or series of traumas) occurred in the third or fourth year. I should not myself give credence to these singular revelations if they had not been proved worthy of belief by the part they played in the subsequent development of the neurosis. In every case a number of morbid symptoms, habits, and phobias is only to be accounted for by going back to these experiences of childhood, and the logical structure of the neurotic manifestations makes it impossible to reject these faithfully preserved memories which emerge from among the experiences of childhood. It would, indeed, be quite useless to question an hysteric outside analysis about these traumas in childhood; their traces are never to be found in conscious memory, only in the symptoms of illness.

All the experiences and excitements which in the period of life after puberty prepare the way for or occasion the outbreak of hysteria can be proved to act only because they awaken the memory-trace of those traumas in childhood; this memory-trace does not become conscious but leads to a liberation of the affect and to repression. In close accord with this view of the part played by the traumas occurring in later life is the circumstance that they are not subject to the strict conditions governing the traumas of childhood, but may vary in intensity and quality from actual sexual assaults to mere sexual overtures, witnessing sexual acts by others, or being told about {160} sexual matters.[44]

In my first paper on the defence neuroses it was left unexplained how the efforts of previously healthy persons to forget some traumatic experience of this kind could have the result of really producing the repression intended and of thereby opening the door to a defence neurosis. It could not lie in the nature of the experiences, because other people remain well in spite of similar incitements to illness. Hysteria could not therefore be fully explained as the effect of the trauma; one had to acknowledge that the disposition to hysterical reaction had already existed before the trauma.

Now the subsequent effect of an infantile sexual trauma can fill the place either totally or partially of this indefinite hysterical disposition. 'Repression' of the memory of a painful sexual experience in maturer years is possible only for those people in whom this experience can re-activate the memory-trace of an infantile trauma.[45] {161}

Obsessional ideas likewise presuppose a sexual experience in childhood (different in nature from that in hysteria). The ætiology of the two defence neuro-psychoses now presents the following relation to the ætiology of the two simple neuroses---neurasthenia and anxiety-neurosis. The latter are direct results of the sexual noxiæ themselves, as I have shown in a paper on the anxiety-neurosis (1895);[46] the two defence neuroses are the indirect consequences of sexual noxiæ occurring before the onset of sexual maturity, that is, consequences of the psychical memory-traces of these noxiæ. The current causes which produce neurasthenia and anxiety-neurosis frequently play at the same time the part of inciting causes in the defence neuroses; on the other hand, the specific causes of the defence neuroses---the traumas of childhood---can at the same time lay the foundations of a neurasthenia which will develop later. Finally, it not infrequently happens that a neurasthenia or anxiety-neurosis is maintained, not by the current sexual noxiæ, but only by the continuous effect of the memory of the childhood-traumas.