A Strange Survival Niche of Psychoanalysis: a Test for Screening Off Air Force Pilots With Crashing Disposition
Table of Contents
*II General Features of DMT
*III Empirical Support versus Logical Analysis
*IV Survey of the Criticism
*V The Truth-Value of Psychoanalytic Theory
*VI Hermeneutic Psychoanalysis
*VII The Scoring Code of DMT
*VIII Identification with the Aggressor
*IX Repression of Memories
*X Repression of Emotions and Drives
*XI Repression and DMT
*XII Isolation in Psychoanalytic Theory and in DMT
*XIII False Denial as Pre-Freudian and Normal Human Behaviour in Pre‑Freudian Times
*XIV Freudian Denial
*XV Reaction Formation
*XVI The Logical Relations Entailed by DMT and the Theory of Percept Genesis
*XVII A Non-Motivational Theory of the DMT Phenomena
*XVIII The Fundament of Psychoanalytic Experimentation
*XIX The Framework of the Subliminal Experiments
*XX The Aetiology of Schizophrenia, Homosexuality, Stuttering, and Depression: the Stimulus Pictures of an Experiment
*XXI Methodological Parasitism and Pair Comparison Instead of Matrix Comparison
*XXII Methodological Parasitism and Unconscious Perception of Cancer
*XXIII Direct Access to an Unconscious Oral and Regressive Fantasy: Statistical Flaws
*XXIV Direct Access to an Unconscious Oral and Regressive Fantasy: Semantic Flaws
*XXV Some Additional Comments
*XXVI General Outlook
The death of psychoanalysis in the near future has been predicted many times, and most such predictions have turned out to be false. Nevertheless, a new situation has evolved since the 1990s. Prior to this decade psychoanalysts would hardly ever try to refute criticism. They would merely attribute personal defects to the critic: either he was ignorant or he was the victim of unconscious pathological motives. But since 1990 a growing proportion of psychoanalysts have realised the necessity of producing some kind of defence. This change of psychoanalysts‘ own behaviour is a hard fact, even if the logic of their counter arguments may be poor.
At least for Sweden another hard fact also holds. Before 2007 the Swedish section of the international Freudian Association (IPA) had got the right to bestow the title „licensed psychologist“ to students who had taken its courses. But in 2007 this right was withdrawn, because the IPA courses were deemed to be of too low quality and too poorly adapted to the needs of mental clinics. The loss of this right seems to be the first step toward the exit of psychoanalysis from Sweden.
The exit proceeds with a snail’s pace, and psychoanalysts have invented numerous counter strategies aimed at further slowing down the pace. During autumn 2010 a „philosophical“ discussion of three psychoanalysts was broadcast in Sweden. Among other things they blamed contemporary people for wanting everything to proceed at a high speed. Supposedly there is something wrong with persons who want their suffering to disappear quickly, and who have no patience (!) to wait for a number of years.
If this ideology is applied to people who have been seriously injured by traffic accidents, or suffer from a kidney stone, or from strong toothache, it is easily seen to be a cruel ideology. And there as definitely some psychological diseases that are equally painful.
*II. General Features of DMT
One of the strangest survival niches is a selection test for air force pilots, Defence Mechanism Test (DMT), which was invented in the 1950s. Still today (2011) it is applied in Sweden and Norway. Not very long ago it was also applied in Denmark.
Over the years the constructors and their co-workers had used skilled sales promotion. They had participated in broadcast scientific programs and claimed that the test could identify pilot candidates who were more or less predestined to crash. They used strong rhetorical language, e.g.: „their names are on the death list“.
During a period of some years in the late 1960s the test was administered to all applicants for Air Force pilot training. But it was not used for selection: everyone who passed all other tests was accepted for training regardless of his DMT score. The scores of the applicants during this non-selection period were preserved in a computer of the Air Force. (So were also the scores during later periods, but this is not relevant.)
Obviously, the scores obtained during the early non-selection period would offer an excellent opportunity to measure the predictive validity of the test. Did the pilots with „dangerous“ scores actually crash more often than the others? Or did they show other negative and related features?
Unfortunately, an unidentified person has later deleted all DMT scores – whatever his motives may have been. As a consequence, no researchers outside the DMT team have had the opportunity to compare the predictions with the actual performance.
DMT exists in several versions, and it is impossible to disclose which version was applied in the 1960s, though it was definitely not the version described in Ulf Kragh’s book of 1985. The first exposure of a picture may range from 2 to 10 milliseconds, but this feature of no significance for the criticism that will be presented below. Likewise the number of defence mechanisms disclosed may range from 5 to 10.
If it can be shown that the reasoning around the original 5 defence mechanisms is downright arbitrary and flawed, then the test as a whole could hardly be valid even if no deficiency pertained to the 5 newcomers. Besides, it would strain our imagination to think that the constructor of the test had produced 5 flawless newcomers, though he had nevertheless retained the original 5 flawed mechanisms.
I have chosen to present and criticise the version described by Bert Westerlundh in his book Aggression, Anxiety and Defence (1976), because it seems to be closer to the original version.
*III. Empirical Support versus Logical Analysis
However much we may agree that in the last hand the validity of a test is dependent on its agreement with empirical facts, empirical studies are expensive and time-consuming. Moreover, different studies may yield contradictory results. In a number of studies published between 1960 and 1986 and studying 14 groups, Kragh and Neuman claimed to have found validities averaging 0,34. But the only Swedish researcher that did not belong to the specific DMT group found 0,07. Norwegian researchers have found correlations agreeing with the figures of Kragh and Neuman, while British researchers have found low or even negative correlations. The following two tables are cited from Sjöberg & Källmén & Scharnberg (1997). Even those who are not accustomed to read statistical tables will notice that that the figures in table 2 are conspicuously lower than those in table 1.
Table 1. Reported co-variations and explained variance between DMT variables and practical results from military training and performance in Sweden
|Study||Correlation||Determination in per cent|
|Neuman (1968a, 1968b)||0,35||12|
|Neuman (1968a, 1968b)||0,46||21|
|Neuman (1968a, 1968b)||0,4||16|
Table 2. Reported co-variations and determinations between DMT variables and practical results in air force and attack diver training outside of Sweden
|Study||Correlation||Determination in per cent|
|Stoker (1982)||– 0,26||6,8|
|Martinusson & Torjussen (1993)||0,19||3,6|
Sometimes there are strong rational reasons why it is extremely unlikely that a certain test could be valid. Such reasons do not provide stronger evidential power than clear-cut empirical facts. But many empirical results are published in such a form that it cannot be seen exactly what empirical facts were obtained, and exactly how they were obtained.
In the case to be examined in this paper, an interval of 24 years intervened between the invention of the test and the first empirical study performed by a researcher outside the DMT group. Since psychological tests of doubtful validity are continually invented, published, recommended and applied, such a large time-lag is not desirable. Moreover, in the case of DMT the time-lag is even greater, because still today the test is applied as if it were valid. It seems that the task of cleaning the academic and practical worlds of invalid tests by means of experiment-like procedures is not highly efficacious.
It is worthwhile to ask whether there were cheaper and more speedy methods. I suggest that logical analysis of DMT and percept genetic theory would have been much less expensive and time-consuming, and that this approach could have been performed as soon as the test was constructed. Logical analysis would have revealed that there are strong reasons to believe that DMT is invalid, and there are strong reasons to take a close look at papers that purport to present empirical support: what information is really stated in these papers and what significant information is missing?
*IV. Survey of the Criticism
The Defence Mechanism Test (DMT) is allegedly based on both subliminal perception and psychoanalytic theory. To justify the test and the scoring code, the theory of percept-genesis was invented. Its first assumption is that percepts develop over a brief period of time. Its second assumption is that the early stages of this process are strongly influenced by psychoanalytic defence mechanisms. From these two assumptions the Swedish DMT theorists concluded that tachistoscopically exposing an individual to visual stimuli of duration of a few milliseconds, would reveal his or her the defence mechanisms.
Four fundamental errors will be exposed:
(a) Psychoanalytic theory does not agree with reality. Therefore, the application of this theory within other domains that those originally intended, could only by a miraculous coincidence yield true results.
(b) The scoring code of DMT does not follow from psychoanalytic theory. Instead it runs counter to this theory.
(c) The theory of percept genesis cannot be derived from psychoanalytic theory.
(d) Both DMT and the theory of percept-genesis are internally inconsistent.
*V. The Truth-Value of Psychoanalytic Theory
During almost a century it was widely agreed that Freud and his followers have gathered an immense wealth of observations; that these observations are of an entirely new character; that they differ from what other psychiatrists and psychologists have gathered in a way that could well be compared to the difference between what can be seen by the naked eye and in the microscope, respectively; that no pre-Freudian theory can explain these observations; and that the observations constitute sensible reason (though not necessarily altogether sufficient reason) to believe that psychoanalytic theories are true.
It was also agreed that every psychoanalyst bases every interpretations on an immense wealth of observations.
It is odd that such views could become so widespread, because a glance in any psychoanalytic paper will reveal extremely few observations, and nearly all of these are extremely shallow and trivial. And it can be proved that those few observations that are not shallow, are fabricated out of thin air. Moreover, literally all interpretations are based on extremely few observations. In addition, even if all observations in the writings by Freud and his followers were true, they would provide no support for any psychoanalytic theory. Freud and his followers were pseudo-scientists who were well aware of the fact that they were lying.
The true state of affairs has been so well documented by so many scientists, that there is no reason to repeat the documentation here.
Freud recurrently asserted that his theories were proved on the couch. But his most clear account of the nature of these proofs are found in two footnotes in the case study of Dora: If the psychoanalyst presents an interpretation and the patient responds “I don’t know” or “I haven’t thought of that”, then such answers constitute valid proof that the interpretation is true. Freud makes a telling addition: no other kind of evidence can be obtained.
Thirty-two years later he published Constructions in Psychoanalysis, in which he set himself the task of proving that psychoanalytic interpretations are not arbitrary. But the result is hardly different from what he stated in the Dora article.
*VI. Hermeneutic Psychoanalysis
It is on the surface for anybody to see that all Freud’s interpretations are postulations about causal relations, and that Freud claims that these causal relations definitely exist in the real world. The same is true of the interpretations produced by later generations of psychoanalysts.
But in the 1960s some analysts tried to save psychoanalysis by means of what they called a hermeneutic re-interpretation of psychoanalysis. According to this innovation, the postulated causal relations do not at all exist in the real world. They are a kind of fictions that will enable the patient to „continue living by relying on them“. They constitute a special kind of consolation lies.
After Grünbaum’s meticulous scrutiny in The Foundations of Psychoanalysis (1984) the hermeneutic re-interpretation has completely collapsed. Some further comments may nevertheless be appropriate.
It would seem that if a certain statement is not asserted to have any truth-value, it would be illogical for non-psychoanalysts to request empirical support of it. But this is irrelevant. Instead it must definitely be requested from a consolation lie that it consoles. But it would be absurd to suggest that those interpretations Freud tried to force on Dora could at all have consoled her. And the same thing is true of the interpretations found in the writings by other analysts. They are much more likely to make patients depressed. In addition, some of them are clearly aimed at producing violent outbursts of impotent rage.
Hermeneutic analysts force their interpretations on their patients, even when they observe their harmful effect. It seems rather obvious that the hermeneutic variant was never intended to have any consequences as regards what happens in the consultation room. They were solely constructed as weapon to be used in external debates with sceptical non-analysts.
If psychoanalytic defence mechanisms such as “denial” and “reaction formation” are fictive phenomena, then it is impossible that they can be observed in any individual, and with any means. In particular, their presence in a specific individual cannot be established by tachistoscopical experimentation. And no DMT score of any air pilot could predict his proneness to crash.
*VII. The Scoring Code of DMT
As I stated above, it is no longer possible to find out what version of DMT was used during the non-selection episode. I shall follow Westerlundh’s presentation, because we do know that the 5 defence mechanisms listed by him were included in the original version.
In the DMT procedure, the testee is tachistoscopically presented with a stimulus picture 12 times at logarithmically increasing durations from 10 to 500 milliseconds. The testee does not know that the same picture is presented on all occasions. After each exposure the subject must report and draw what he saw. At close and normal inspection, the pictures are not ambiguous. But they are constructed to facilitate faulty perception at less close inspection.
DMT does not measure the degree of correct perception. Instead, certain specific non-veracious patterns are taken as evidence of specific psychoanalytic defence mechanisms.
Each picture depicts a central “hero” (H) and a peripheral person (PP) who is “threatening” to the “hero”. The interpretation code is as follows:
Repression: The testee sees H or PP or both (or something at their place) as being stiff, petrified (e.g. a statue), masked, an animal, a plant, a non-living object etc. [Exceptions: Death’s head, skeleton, a doll, monster, “the man looks like an ape”, etc.]
Isolation: H and PP are separated from each other.
Denial: Existence of threat denied or made slight.
Reaction formation: Although objectively PP is threatening H, PP is perceived as helpful and positive toward H.
Identification with the aggressor: Although objectively PP is threatening H, H is perceived as threatening PP.
These items will be discussed in the order and to the extent that I deem to be most illuminating.
*VIII. Identification with the Aggressor
According to psychoanalytic theory, this defence mechanism can be manifested in two and only two situations. (a) The victim (V) may try to reduce his anxiety for the aggressor (A) by persuading himself that V is really A. And on the basis of this illusion V may assist A in attacking V. (b) V’s anxiety reducing technique may consist of imitating certain other characteristics or behaviours of A; e.g., if A has a moustache, V may also grow a moustache.
In DMT it is assumed that the testee identifies himself with H. This is by no means proved. But even if we take this assumption at face value: to deliver a counter attack or a counter threat against PP would according to psychoanalytic theory illustrate the very opposite of identification with the aggressor. Yet this is the kind of behaviour which the DMT scoring code takes as evidence of this defence mechanism.
Several additional aspects around this mechanism are remarkable. In Sigmund Freud’s Gesammelte Werke (the original German text) an entire volume of 1099 pages is devoted to the indexes. There are 139 entries for Identifizierung including Identifizierung mit (“identification”, “identification with”). Nonetheless, there is not a single entry for Identifizierung mit dem Angreifer (nor for “Angreifer”).
It is generally known that Freud in 1896 believed that hysteria and other neuroses were caused by sexual seduction during pre-school age. In his three seduction papers he makes it clear that he himself was the one who invented this causal interpretation, while the patients denied that they had been seduced. Nevertheless, in a letter from the same year Freud explains hysterical grimaces as an imitation of those grimaces which the seducer had shown during the sexual assault. In order to overcome her anxiety, the victim supposedly identifies herself with the perpetrator, thereby persuading herself that she need not be afraid because she is not the one who is being abused but the one who is abusing.
Anyone who has seen hysterical grimaces must seriously doubt that more than an extremely small fraction of active paedophiles could have shown this variety of grimaces.
An even stronger objection derives from the fact that Freud rejected the seduction theory in September 1897. If his patients had not been seduced at all, it is logically impossible that they could have imitated the grimaces of any sexual seducer.
But Scharnberg examined psychoanalytic methodology and found that it is a fundamental rule that any statement, which was once deemed to be proved, will remain proved, even if all supporting evidence is later declared to be non-existent, and no new evidence is substituted.
In 1936 Anna Freud published what was for generations considered a standard book on defence mechanisms. She devotes nine pages to “identification with the aggressor”, and she presents two illustrative examples. But it is a noteworthy fact that she supplies no example from any psychoanalytic writing, and neither one from her own clinical practice. Both her examples had merely been presented orally by August Aichhorn. This is in itself an astonishing selection, since Aichhorn had a limited training in psychoanalysis, and had primarily applied Freud’s theories in an educational, not a clinical context.
Moreover, Aichhorn’s (1972) examples are too strange to be believed. Numerous psychoanalysts have published examples in which they found the concealed true state of things in a way reminding of Sherlock Holmes’s deductions. But no one can match the successes Aichhorn claims to have achieved. Repeatedly he finds the solution already during the first consultation, and therapeutic improvement follows immediately.
*IX. Repression of Memories
Repression occupies two places in psychoanalytic theory, depending on whether the repressed thing is a memory, or else an emotion or a drive.
According to the third seduction paper, Freud had in 1896 conclusively established by meticulous study of 18 hysterical patients, that their symptoms were caused by repression of events of sexual seduction during pre-school age. Freud had also constructed a technique for lifting repressions. All 18 patients eventually recalled the original causal events in their entirety. And concomitantly with this recall all symptoms of all patients disappeared.
It is irrelevant that the seduction theory was later retracted, because still in 1917 Freud asserted that symptoms will disappear if and only if hitherto repressed causal events have re-emerged into the conscious. The only difference between his position in 1896 and 1917 is the nature of the repressed event.
Despite intensive search, no one has found any patient who was improved by Freud’s treatment. And apart from therapists who combine psychoanalysis with recovered memory therapy, it seems that no contemporary psychoanalyst will agree that his treatment will produce recall of hitherto repressed memories.
One of the best sources revealing what kinds of observations Freud had made is the 101 case studies which Felix Gattel published in 1898. All Gattel’s observations are shallow and devoid of any evidential power. Moreover, there is strong reason to believe that each and every patient whom Gattel gave the diagnosis “hysteria”, had a purely somatogenic disease.
Nevertheless, Freud accused Gattel of having plagiarised Freud’s discoveries in this book. This is a clear indication that Freud’s own observations were equally shallow and equally devoid of evidential power.
Actually, it is no longer controversial that all assertions by Freud and his followers that they had observed lifted repression, are not true.
*X. Repression of Emotions and Drives
Only a small part of the writings by Freud and all his followers is concerned with repression of emotions and drives. There is one exception: the doctrine of negative transference. Psychoanalysts of all schools agree that patients, who are otherwise not aggressive, will have numerous violent outbursts against their psychoanalyst. They also agree that the analyst is merely an innocent bystander, who had done nothing that could justify or even explain the patient’s aggressive response. Instead they claim that these outbursts consist of hitherto unconscious emotions that the patient had hitherto felt toward his parents.
But the myth of the non-aggressive psychoanalyst is amply refuted. The dialogues included in the writings by the early analysts reveal hyper-aggressive behaviour by the analyst, together with the astonishing unawareness of the nature of the analyst’s own influence. Illustrative examples are Freud’s case study of Dora and his letter 3.1.1897 about G. de B., and so are several Books by Edmund Bergler.
Today much more sophisticated techniques for exasperating a patient into fury exist, than those applied by Freud or Bergler. And as regards the invention of such techniques, psychoanalysis has made genuine progress. But after Haley’s scrutiny of the theory of negative transference and his examination of the actual behaviour of analysts, it can no longer be denied that the patients’ outbursts foremost constitute the natural response to specific provocative measures.
Scharnberg analysed eleven audio-recorded psychoanalytic sessions and disclosed, inter alia, the psychoanalyst’s lack of awareness of his own hyper-aggressive behaviour.
*XI. Repression and DMT
The Swedish DMT theorists never explained whether the relevant parts of DMT were related to repression of memories, or to repression of emotions and drives. But from the two preceding sections it can be seen that the psychoanalytic statements about both varieties of repression are false. Hence, a test that allegedly discloses psychoanalytic repression, cannot be valid.
On the basis of the descriptions of repression in psychoanalytic literature, it is incomprehensible why it should be an indication of repression, if “the testee sees H or PP or both (or something at their place) as being stiff, petrified (e.g. a statue), masked, an animal, a plant, a non-living object etc. [Exceptions: Death’s head, skeleton, a doll, monster, “the man looks like an ape”, etc.]”
Most psychoanalysts have maintained that hysterics primarily suffer from repression. They have also claimed that repression is least prominent in psychotic schizophrenics; allegedly the psychotics’ defence mechanisms have to a large extent broken down. Allegedly, rudiments of more complex defence mechanisms, such as projection and regression, may be found in schizophrenics.
But it is an empirical fact that it is not the hysteric but the psychotic schizophrenic who believes himself to have a dog’s pawns instead of hands; to be a cat, a ram, Switzerland, a radio; not to be born but to have been produced by a machine; to dream of her parents and siblings as stony statues who fall together as a heap of sand if touched. Scharnberg gathered these examples from the literature. Besides, Fenichel explains stiff and lifeless positions in catatonics as deriving from the uterine existence. Other comments of his would rather suggest projection if a machine is involved.
Clearly, the scoring code for repression is glaringly discrepant with psychoanalytic theory.
*XII. Isolation in Psychoanalytic Theory and in DMT
No clear description of isolation can be found anywhere in Freud’s own writings. But we can be rather sure that he would agree with the following account by Otto Fenichel. Note that Fenichel was the indisputable theoretic leader of the Freudian movement after Freud’s death.
“The most important special case of this defense mechanism is the isolation of an idea from the emotional cathexis that originally was connected with it. In discussing the most exciting events, the patient remains calm but may then develop at quite another point an incomprehensible emotion, without being aware of the fact that the emotion has been displaced. Extremely objectionable ideational contents, like murder or incest wishes, may become conscious in the form of obsessions, because the obsessional neurotic is able to feel these ideas as mere thoughts, securely isolated from motility. The emptiness of affect, which is so characteristic for certain compulsion neurotics and which creates a serious difficulty in treating them, is based on isolation of this type.
Of practical importance is the patient who hinders any therapeutic effect of his analysis by carrying on the whole analysis ‘in isolation’.”
Many objections can be raised against these two excerpts. At least since 1960 a large number of psychoanalysts have denied that their variety of treatment will cure neurotic symptoms (although they have indulged in much double-talk about this topic). But if the treatment itself does not produce any therapeutic effect, a patient could not hinder therapeutic effect, whether by isolation or by anything else.
Psychoanalytic isolation is invariably concerned with separation of two or more reactions or behaviours by the same individual. It is not concerned with separation of two external entities; and neither with the separation of an individual from parts of his environment. Hence, the DMT code as regards to isolation bears no comprehensible relation to psychoanalytic theory.
*XIII. False Denial as Pre-Freudian and Normal Human Behaviour in Pre‑Freudian Times
In contrast to most psychoanalytic defence mechanisms, a phenomenon akin to psychoanalytic “denial” really exists. It is a good guess that it has existed for at least 2000 years. But at the present stage of scientific methodology it would hardly be possible to gather the best evidence. The task can literally be compared to finding a needle in a haystack. However, in the near future computer techniques might well be developed, so that private letters, private diaries, memoirs, case-records of legal trials, and many other kinds of documents, can be „sifted“ in order to find some instances of the pattern searched for. It is a realistic expectation that sifting of millions of documents will yield a very small number of instances.
Until such techniques can be applied, the most adequate approach is to search for instances in fictional writings: novels, short-stories and plays. Obviously, the fact that a certain pattern of behaviour is described in one or in many novels, does not prove that human individuals really behave in this way. But I think that numerous persons will easily recognise the pattern I shall illustrate.
Because of specific motives people may sometimes add or delete a negation from a formulation which without this negation corresponds to their true view. They may be fully aware of what they do: the militant dictator preparing a war may assure that he only wants peace. But another (sub‑)pattern is more relevant here: the person who tries to persuade himself that his real view is the formulation he utters, although it is actually what would result if a negation were added or subtracted. In such cases the style of the „modified“ utterance might seem peculiar and strained.
Nicholas Nickleby by Charles Dickens was published in 1839, that is, 17 years before Freud was born. When Nicholas’s mother meets his friend Smike for the first time, she assures that he is welcome, and she does so in so many and so exaggerated words, that it is evident that she means the very opposite of what she says. It is equally evident that she is defending herself against what she really feels:
“I am sure, my dear Nicholas, […] I am sure any friend of yours has, as indeed he naturally ought to have, and must have, of course, you know – a great claim upon me, and of course it is a very great pleasure to me to be introduced to anybody you take interest in – there can be no doubt about that: none at all; not the least in the world” (Dickens, 1982:225, in chapter 35).
Proust’s large novel has in various countries been published in three, seven, or some twenty volumes. In seven-volume-editions (such as the Swedish) we shall find the relevant scene in the 7th volume. Two males outside the entrance of a homosexual sado-masochist brothel are torn between desire and shame. One of them repeats every other minute: “You don’t care about that, do you?”. The original French text is “Quoi! Après tout s’en fiche?” (1982:126).
This part of the novel was published in 1927, while Freud’s article on negation (or rather on denial) was published in 1925. Nevertheless, Proust could not have been inspired by Freud. Freud’s examples are counter-intuitive, and prove little more than his own aim of deceiving his patients. Proust’s account agrees with general lay experience.
Among other things the Danish author Astrid Ehrencron-Kidde wrote many short-stories about a man given the name Martin Willén. In the edition published in 1941 this man, now 70 years old, looks back at his life and recalls the tragic death of three of his siblings and of three other persons who were very close to him. Only one of these six persons was approaching middle-age, while four were still teenagers.
In one of the short-stories, The Tarantula, there is an excellent scene in which the child Martin asks his mother if it is true that his older sister Elvira is going to marry his father’s brother and follow him to his African farm. Evidently the mother tries to persuade herself that this will be a good thing, since she knows that she cannot stop it. But she really feels (and, it will turn out, rightly so) that this marriage will be a disaster.
At first I was very scared. But then I thought that she had been joking. Nevertheless, to be sure I went to my mother and asked her if it was true that Elvira would leave us and be Uncle Richard’s wife.
“Yes”, mother said and sighed and looked very grave. But a moment later she smiled. “What can we do about that, Martin,” she said in a joking tone of voice. “Elvira wants him! And, you know, Uncle Richard is a very good man, isn’t he Martin?”
Alas, if only she had not said the last thing! Because now I knew that neither my mother liked Uncle Richard. I heard it, it rang to me through the word, although they said the very opposite things. (Ehrencron-Kidde, 1941:98f.)
*XIV. Freudian Denial
Freud sometimes applies false denials (or false non-denials) as a persuasive technique. He writes, „I am, as far as I know, not ambitious“. But he can hardly have been unaware that not many people sought as strongly as he himself for honour and praise.
We are all familiar with the person who makes a deliberately false statement, and then claims that he has “forgotten” the facts that prove his statement. But in his paper Character and Anal Eroticism Freud carries this pattern one further step. What he allegedly had forgotten were not the facts constituting the evidence of the theory. It was those facts which originally lead him to invent the theory. In other words, he had forgotten some unimportant facts, and he left the question open whether or not he had also forgotten the important facts.
If the content of his text is completely spelled out, something like the following sentence will appear: “Do not conclude from the fact that I have forgotten the unimportant facts, that my doctrine is based on speculation.” The combination of the first and second half of the sentence is awkward. But if the prefix “un‑” is deleted the sentence will become logically coherent.
By contrast, I have been unable to find any instances in which Freud or his followers have attributed false denial to a patient, and this interpretation was supported. From Freud’s brief paper On Negation it can be gathered that, according to his own view, the following two examples have a particularly strong evidential power and are particularly illuminating as to the nature of “denial”.
A patient may say, after having recounted a dream, “You ask who this person in the dream can be. It’s not my mother”. – Freud claims that this statement is in itself proof that the figure is a symbol for the mother. “It is as though the patient had said: ‘It’s true that my mother came into my mind as I thought of this person, but I don’t feel inclined to let this association count’.” – Freud perceives no complication by the fact that the psychoanalyst may previously and repeatedly have interpreted every unknown female in the patient’s previous dreams as a mother symbol.
It is a frequent human experience, first, that headache is temporarily cured by a tablet; and, second, that, about half an hour before the effect of the tablet has faded, the thought pops up: “I haven’t had a headache for several hours.” The pharmacological explanation is rather obvious. When the effect is fading, the brain may receive some pain signals. At this stage they will not be sufficiently many for producing the headache again, but sufficiently many for producing a memory of the headache.
But Freud ignores these features and interprets the popping-up thought as a further instance and proof of psychoanalytic denial.
The DMT scoring code for denial („existence of threat in the stimulus picture is denied or made slight“) is neither in clear-cut agreement nor in clear-cut disagreement with psychoanalytic theory. Nevertheless, many circumstances could induce human beings to overlook numerous features, and many of them are conspicuously different from those postulated by psychoanalysts. This is true both in tachistoscopical and in everyday perception. And it is true whether the structure of the picture is consistent or inconsistent.
*XV. Reaction Formation
Many religious ascetics during the first millennium have experienced something akin to reaction formation, e.g. intense fright of sex or intensive repulsion against sexual drives. There is no need for laboratory studies to unearth whether such responses exist. They can even be observed among some contemporary psychotics. – Non-original love (a positive emotion) could also derive from the attempt to combat feelings of hate.
However, the specific psychoanalytic theory postulates much more than the pattern just described. It is a central constituent that the derivative emotion is claimed to be a disguised form of the original emotion, so that both emotions are somehow identical.
Besides, the DMT code would be acceptable only under a series of assumptions. (a) The testee identifies himself with H. (b) H feels genuine hate toward PP. (c) H also feels fear of PP. (d) Because of this fear H’s original hate is transformed into love felt by H. (e) PP’s helpfulness reflects the transformation of H’s own feeling.
I fail to see that the DMT code satisfies any of these five assumptions.
So far, it has been established that the derivation of the DMT code from psychoanalytic theory is arbitrary. Consequently, even if psychoanalytic theory were true, it would provide no support for the DMT code.
But I cannot help thinking that when the reasoning of a group of researchers is so extremely careless as that of the DMT theorists, then it would be a miracle if they would arrive at true results.
*XVI. The Logical Relations Entailed by DMT and the Theory of Percept Genesis
In the present chapter, whenever I assert that a statement is logically necessary, it is understood that it is so only on the condition that the theory of percept genesis and the additional theoretical assumptions underlying DMT are true.
It should be carefully noted that Kragh’s theory is not intended to be a description and explanation of phenomena which occur only in tachistoscopical situations. It is a theory about the nature of ordinary visual perception. Although certain particular aspects can best (and perhaps only) be disclosed in the DMT situation, the same processes are assumed to be present in all varieties of visual perception.
It is evident that human beings do not experience the specific DMT distortions during their everyday interaction with their environment. But let us try out the pattern that, on the one hand, such distortions occur during the early stages of the development of a potentially painful percept; and that, on the other hand, the final form of the percept that enters the conscious mind is free from these distortions. Then two consequences are logically necessary: (a) Some correcting process must have deleted the distortions. (b) The correction must have taken place after the defence mechanisms had distorted the percept, but before the percept was admitted to consciousness.
Kragh assumes (in accordance with psychoanalytic theory) that the aim of the distortions of the percept is to protect the conscious mind from pain. But then it is perplexing that both the distortions and the correction take place before the percept reaches the conscious mind. The second process cancels the effect of the first one. Hence, what the conscious mind actually receives, is exactly the same percept, and exactly the same amount of pain, as it would have received if there had been neither any distortion nor any correction.
When unearthing further errors in Kragh’s theory, I want to avoid giving several different meanings to the word “percept”. I shall therefore use the expression “message” as an omnibus term for (a) the package of information entering the perceptual apparatus; (b) the form the package has got after each transformation; and (c) the form finally sent to the conscious mind.
When a physical state of things impinges on the organism, an indeterminate “something” is received. This is the first form of the message. I have deliberately used a vague formulation (a “something”) in order not to commit my analysis to any particular variety of theories. This first stage is not logically necessary. But it is in agreement with all theories of visual perception entertained by any experimental psychologist at the present time.
Unless the message had in advance been sufficiently worked out to enable the assessment that it would indeed be painful to the conscious mind, the perceptual apparatus would have a poor ground for finding out whether there would be any need of distortion, and what aspects of the message should be concealed by distortion. Hence, a second stage must necessarily follow, during which the “something” is worked into a distinct form. In turn, a third and separate stage must assess the need of distortion. And a fourth stage must perform the distortion.
From what has already been shown, the message must run through a sequence of at least six consecutive stages: (a) A “something” is received by the organism. (b) This something is worked out into a distinct message. (c) The message is assessed as to whether it would be painful to the conscious mind. If the answer is negative, the further fate of the message lies outside the scope of DMT, the percept genetic theory, and my analysis. But if the answer is positive, the message proceeds to the next stage. (d) By means of Freudian defence mechanisms the message is distorted into a form which would be less painful to the conscious mind. (e) The distortions are removed and the previous distinct message is retrieved. (f) The distinct message enters the conscious mind.
All six activities, except the first one, are logically necessary if the DMT theory is true.
We are now approaching what I consider the most serious error in Kragh’s theory, and also the error that is most easy to detect. Note that my argument is independent of what brain substrata “handle” each stage or activity.
There would be some sense in the idea that we might be able to observe the nature of the distortions produced by defence mechanisms, if we could interfere with the smooth function of the brain substrata concerned with deleting the distortions and retrieving the first distinct message; for instance, by reducing the time the message spent in these substrata. But it is enigmatic why reduction of the time the message spent in stage (a) should have any influence upon whether the distorting influence of stage (d) is corrected or not before the message is given access to the conscious mind.
A corollary. Psychoanalytic writings from any period unambiguously reveal that interpretations are almost exclusively based on the patient’s words. Even tone of voice is hardly ever taken into account. But it follows from Kragh’s theory that any word spoken by the patient and heard by the analyst could at best correspond to the end result after the message had finished all six stages.
Whenever Freud or any of his followers attributed a defence mechanism to a patient, they did so on the basis of the final message entering the patient’s conscious mind, in the form it had after having been exposed the supposed Kraghian distortions and the supposed Kraghian corrections. Hence, it is not a prima facie plausible suggestion that the very same kind of defence mechanism should be active in almost the same way at two highly different levels: on the one hand, in the patient’s words, and, on the other hand, during the perceptual stages from the external stimuli start to hit the patient’s sense organs and until the patient perceives the external object.
Theories that are prima facie implausible have sometimes turned out to be true. But as regards prima facie implausible theories we are entitled to request genuine evidence, and it must be presented in a much more clear form than what is found in the relevant writings by Kragh and Neuman.
*XVII. A Non-Motivational Theory of the DMT Phenomena
For a century it was known that ultra-brief expositions of pictures will very often be falsely perceived, regardless of whether they are frightening or not. Hence, the most parsimonious explanation is that something else than the frightening nature is responsible for the distortions.
To unprejudiced persons who have seen the DMT pictures, it must seem enigmatic why they would be frightening to anyone. It must be even more incomprehensible why the distorted form should be less frightening. DMT psychologists would not be happy if I published some of these pictures, because this could interfere with the performance of future testees. I shall therefore choose an example from a subliminal experiment by Kline & Cooper. The supposedly frightening picture is “Pig with black spots and piglet feeding from her”, while the supposedly less frightening distortion is “Pig with spots and udders, but little pig may not be feeding but talking”.
Human perception is not very good under highly unusual conditions. The explanation might be that biological evolution has not favoured correct perception in some situations that must have been extremely infrequent during the entire existence of mankind, and in which correct perception could hardly have any survival value. Whether or not this explanation is true, proponents of DMT and the percept-genetic theory will have a formidable task in explaining why non-frightening pictures are distorted as much as frightening pictures.
In our “macrophysical” experience, most of us have sometimes mistaken word, pictures, and entire events. People, who are somewhat hard of hearing, will repeatedly hear words which were never said. If the words erroneously heard are compared with the words actually said, it may often be apparent that the falsely perceived words were constructed from correctly perceived fragments, together with fictive details filled in to obtain a coherent and meaningful pattern. Experimental research has shown the same process to be active in many contexts.
A visual picture may contain 20 details. At ultra-brief exposition a person may overlook 15, correctly perceive 2, and mistakenly perceive 3. Even if the physical picture is coherent, the set of details perceived may be contradictory. It is no bold hypothesis that the human perceptual apparatus may try to make sense of this set by “cutting a heal and a toe”, and by adding something too.
If the stimulus picture is contradictory or contains constituents which seldom or never occur together (e.g., a drawing by the Danish humorist Storm-P. in which the head of a man is substituted by a bicycle wheel), it would be unsurprising if earlier and briefer expositions might result in a more correct percept, than somewhat later and longer exposition. Perhaps the testee or experimental subject detected a smaller and less inconsistent set of details in the beginning.
*XVIII. The Fundament of Psychoanalytic Experimentation
In a series of chapters we shall examine some attempts to produce empirical and experimental support of psychoanalysis by means of subliminal perception. The period of psychoanalytic experimentation ended more than twenty years ago, so we are here not concerned with a survival niche. There is nevertheless reason to examine the approach by the American groups around Lloyd Silverman, Donald P. Spence, Norman Dixon and their co-workers. They were active at the same time as the Swedish group who developed and propagated DMT, and there were close relations between both groups.
The present chapter will be devoted to illustrate the fundamental flaws of all psychoanalytic experiments, whether or not they involve subliminal perception. The experimenter will invariably test a concrete prediction, which they claim to have derived from psychoanalytic theory. However, a common feature of all such predictions is that they have a high probability of being true on the basis of ordinary background knowledge, and that this probability will neither be enhanced nor reduced if psychoanalytic theory is false. – It may be less significant that often the prediction cannot be derived from psychoanalytic theory, and that it sometimes runs counter to this theory.
Let us start with a schematic example not involving psychoanalysis. Suppose I want to prove the astrological theory that people born in the Zodiacal sign of Leo are particularly prone to have cardiac diseases. I may compare two groups: Leo-born individuals taken from a medical clinic, and Capricorn-born members of a yachting association. Here the empirical evidence which truly belongs to one hypothesis (patients at a medical clinic will more often than members of a yachting association have cardiac diseases) is, as it were, embezzled and in turn used as pseudo-support for a quite different astrological hypothesis.
This is parasitism, a very important concept within scientific methodology. In Swedish you can form an excellent word, which would be awkward in most other languages: „snålskjutshypotes“, which literally means a hypothesis travelling as a free passenger.
The true and the false hypothesis are locally equivalent, because there are some predictions which both hypotheses will make.
Paul Kline, a proponent of psychoanalysis, examined hundreds of psychoanalytic experiments. In his books of 1972 he selected 18 experimental studies as having strong evidential power, and in the second enlarged edition of 1980 he added two more experiments. To illustrate the flaws of the approach as such, we shall in the present chapter take a closer look at one of these experiments, although it does not involve subliminal perception.
The most fundamental methodological rule of the interpretations and theories by Freud and his followers is the principle of similarity: the cause is similar to the effect. Hence, the cause of a symptom can be disclosed by finding – or by inventing – an event that is similar to the symptom.
Asthmatic attacks are often triggered off by bad smell, and a fart is an instance of bad smell. From these facts Freud concluded that the cause of asthma has a close relation to anal-eroticism.
In 1958 Stein & Ottenberg published a study that, according to both themselves and Kline, provided strong evidence for this anal aetiology of asthma. They presented asthmatics with a long list of smells, which Stein & Ottenberg had classified as oral, anal or genital (three psychoanalytic categories). They asked which of these would trigger off asthmatic attacks. They found overwhelming predominance of anal smells.
Scrutinising this study, let us first ask: which of the following four categories of smell are unpleasant to most human people: food, dirt, chemicals for removing dirt, perfume? Most people would agree that dirt and chemicals are unpleasant. (A biological explanation might be suggested: uncontrolled amounts of both have a negative survival value, while the same is not generally true of smells of food or perfume.)
Hence, it is unsurprising that asthmatic attacks are more often triggered off by smells of dirt and chemicals, than by smells of food or perfume. This is not a part of psychoanalytic theory, nor an empirical fact that supports some part of this theory.
But Stein & Ottenberg applied a legerdemain. They classified the smell of food as oral, and the smell of perfume as genital. What they deemed to be anal smells were either dirt or dirt-removing chemical.
It seems odd to classifying the smell of rotten fish and other rotten food as anal. This is actually an instance of producing evidence by classifying. Moreover, these experimenters did not derive their prediction from psychoanalysis, but from the background knowledge of ordinary laymen.
It is astonishing that pseudo-experiments of this variety were published at all, even in prestigious periodicals. The same is true of the subliminal experiments that will be analysed below.
*XIX. The Framework of the Subliminal Experiments
Common to the Swedish DMT group and the American group involving Lloyd Silverman were the attempts at finding empirical support (or pseudo-support) for psychoanalytic theory, their use of subliminal perception, and their strong focus on defence mechanism. It is no great difference that the Americans primarily were concerned with experiments, while the Swedes primarily were concerned with tests.
Another difference was remarkable. Both groups entertained (and allegedly found evidence for) opposite and contradictory doctrines in relation to defence mechanisms. The fundamental axiom of the Swedish researchers (and allegedly their primary empirical result) was that defence mechanisms would be more active and produce greater distortion, when the time of exposition is very brief. The fundamental axiom of the American group (and allegedly their primary empirical result) was that defence mechanisms would be less active and produce lesser distortion when the time of exposition is very brief. In other words, the Swedes intended to study the defence mechanisms themselves, while the Americans intended to study the unconscious itself, before its real properties had been distorted by the defence mechanisms.
This could have been a fruitful controversy. But both groups were more inclined to admire the results of each other, while ignoring the glaring contradictions between their positions.
*XX. The Aetiology of Schizophrenia, Homosexuality, Stuttering, and Depression: the Stimulus Pictures of an Experiment
Kline (1981:433) cites Silverman et al. (1973) and Silverman (1976) as being among the 20 studies which have supplied the most powerful support of psychoanalytic theory. He states that “any blanket rejection of Freudian theory as a whole simply flies in the face of the evidence.” I have nonetheless chosen to analyse Silverman et al. (1976), which is an extension of his earlier study.
Four stimulus pictures were shown tachistoscopically for 4 milliseconds to four samples of subjects. The stimulus pictures had four different characters: aggressive, incest, anal, neutral. The subjects consisted of schizophrenics, homosexuals, stutterers, and depressives. There were two dependent variables: (a) increase of main symptoms of each sample (various measures of thought disorders in schizophrenics; of homosexual arousal; of stuttering; and of some measures of depression); (b) “pathological non-verbal behaviour” (e.g. fingertapping).
The predictions were that the main symptoms as well as pathological nonverbal behaviour would increase after exposure to sample-relevant pictures, and only in relation to these. The sample relevancy joined schizophrenia and depression with aggression, homosexuality with incest, and stuttering with anality.
Silverman and his co-workers claim that there is agreement among psychoanalysts that these connections exist. This claim reveals their limited acquaintance with psychoanalytic literature. Schizophrenic females who have fantasies about having slept with their fathers, were known before psychoanalysis. And it would be a matter of routine to fill up 1000 pages with psychoanalytic quotations about the close relations between schizophrenia and incest. I shall merely mention Fenichel (1945:422, 437, 441), Bleuler (1955b:412), Schilder (1952:71, 72), and Rosen (1953:100f.).
The proof of the connection between stuttering and anal eroticism is based on the principle of similarity: individuals suffering from constipation and stuttering, respectively, have difficulty in “pressing out” faeces or speech.
The theoretical support of the incestuous aetiology of homosexuality is of the same arbitrary kind.
As regards the empirical support the team claims that 3 out of the 4 hypotheses about increase of main symptom. The non-supported hypothesis was the one about the depressives. Likewise the empirical support of the increase of non-verbal behaviour was confirmed for 3 groups, that is, for all groups except the stutterers. Indeed, Table 3 looks impressive, and its evidential power has been accepted by Dixon (1981:170).
To those who are not used to read statistical tables, the less a numeric figure is, the greater is the evidential power (unless there are hidden errors in the experiment). „n.s.“ means „not significant“ and indicate that the number is too large to provide any support. „NT“ means that this connection was not tested at all.
Two features are highly interesting. One is the nature of the pictures. The other is the strange pattern of what was not tested. Both will be scrutinised.
The pictures were provided with text. I shall describe all pictures and indicate the text in capitals.
Text on picture
|The aggressive stimulus:||An angry man about to stab a woman.|
(DESTROY MOTHER)The incest stimulus:A nude man and woman in a sexually suggestive pose.
(FUCK MUMMY) / FUCK DADDY, for male or female subjects, respectively.)The anal stimulus:A person of indeterminate sex seen from the rear and defecating.
(GO SHIT)The neutral stimulus:Two men.
One objection to this experimental design immediately suggests itself: What would be the reactions of the same persons to the same pictures, if there were no text? An important alternative hypothesis would be that the reactions would be same. And at the very least it should have been tested.
Males may look at playboy nudes for several minutes without perceiving the text. But Silverman and his co-workers took for granted that the homosexuals correctly perceived both the picture and the text, and that their reaction was to a considerable extent determined by the text. These arbitrary presuppositions would have been easy to test by deleting any text, or by substituting a neutral text such as „EAT MORE FRUIT“.
If a heterosexual individual had grown up in a society of 97% homosexuals, and had for decades encountered a homosexual orientation in nearly 100% of all love poems, love novels, love movies, sexual jokes, educational writings, nude magazines, and so on; then we might not be surprised if he or she became heterosexually aroused by a homosexual sexy picture.
When the Silverman experiments were conducted in the early and middle 1970s, their homosexual subjects had indeed grown up in a society in which love and sexuality were by and large equalled with heterosexuality. Hence, it would not be remarkable if some or all of them could be homosexually aroused by sexy heterosexual pictures.
This would be a significant alternative hypothesis. But these experimenters thought that they had ruled it out, and I shall not con conceal how they had done it. Together with other co-workers Silverman found in 1973 that a heterosexual control group did not become homosexually aroused after the so-called incest stimulus. – What if these subjects had become heterosexually aroused after seeing a homosexual couple in a sexually suggestive situation?
*XXI. Methodological Parasitism and Pair Comparison Instead of Matrix Comparison
The postulated unanimity of psychoanalysts as regards the aetiology of the four samples is much less than the Silverman team claims. After Freud’s death in 1939 Otto Fenichel was the undisputed theoretical leader within International Psychoanalytic Association. Moreover, no textbook has been used so often at psychoanalytic training institutes as The Psychoanalytic Theory of Neuroses, written by Fenichel and published in 1945. Fenichel does not restrict his account of the aetiology of stuttering to anal circumstances. On pp 311-317 he also states oral and phallic causes. But he applies foremost what Scharnberg (1993) called the gossip theory of (psychic) disease: the person stutters for the purpose of making people look at him – though his real wish is that they should look at his penis.
A reasonable alternative hypothesis to those stated by Silverman as well as those stated by Fenichel would be that stuttering may often increase during states of annoyance, distress, worry, embarrassment. From this alternative hypothesis we would expect increase of stuttering, inter alia, after watching a picture of defecation, but also after pictures of severe burns, serious car accidents, surgical operations etc. The aggressive picture that Silverman showed to the schizophrenics might likewise increase stuttering.
It is a recurrent psychoanalytic procedure to compare only pairs of groups, instead of comparing the reaction of all groups upon all pictures, that is, using a matrix as point of departure. Why did the Silverman team refrain from investigating the responses of the schizophrenics and the homosexuals to the anal picture? Or the reaction of the stutterers to the aggressive picture? Or the reaction of the depressives to the incest picture? They could not excuse this flaw by claiming that they had merely performed a preliminary study, because they have published several papers on the same topic.
This tendentious preference of pairs should as a matter of routine give rise to suspicion. It is not difficult to find a narrow interval within which the same predictions could be derived from ordinary lay knowledge and (really or allegedly) from psychoanalytic theory. Studying the outcome for all groups in relation to all pictures would decrease the possibility that psychoanalysis could be parasitic on lay knowledge.
It is no far-fetched guess that the Silverman team were somewhat aware that their derivations would fail to yield any support or pseudo-support to psychoanalysis, if they had used a full matrix; and likewise if they had deleted or changed the text.
As an autobiographic remark, I would be not be surprised if it turned out that the Silverman team had actually used all pictures with all groups, but had preferred to conceal some of their empirical results.
*XXII. Methodological Parasitism and Unconscious Perception of Cancer
Although the following study by Donald Spence does not use the tachistoscope, it is related to the experiments by the Silverman teams, in so far as it claims to have established subliminal perception of non-distorted factual states of things. Spence interviewed 62 female subjects who were waiting for the outcome of a test of cervical cancer. 59% of those who turned out actually to have cancer, but only 25% of those who did not, used the word “death” during the interview. Spence’s explanation of the more frequent use of this word among the real cancerous patients reveals that they unconsciously were aware of their true condition.
In chapter XIV we discussed the possibility that a limited number of diffuse pain signals might announce that headache was about to re-appear. In this situation two circumstances must be noted. The psychological aspects of headache are rather simple: it is pain in a particular body part. Furthermore, the subject had earlier in his life had a completely conscious experience of how headache feels. The subject had even had an attack of headache a few hours earlier.
By contrast, cancer is a complex phenomenon. While the oldest known description was made 3000 B.C., still today medical experts can very often not tell whether a certain patient has got this disease, except on the basis of laboratory tests.
While Spence’s argument is false, it is easy to overlook another aspect. Psychoanalysts and psychoanalytic theorists have always boasted of the extremely fine-grained nature of their observations. In actual fact, all psychoanalytic writings reveal the extremely crude and unrefined of these observations. And Spence constitutes a good example.
All patients whom he studied were equal in sure knowledge as to whether they had cancer. But were they also equal in suspicion? What if a majority of those who really had cancer, but only a minority of those who had not, had consciously felt various strange bodily sensations or pains? What if a majority of the former but only a minority of the latter had consciously wondered whether they had cancer? What if a majority of those who had so wondered but only a minority of those who had not, had used the word “death”?
Such distributions will be more easily to survey if exact numerical values are given for all majorities and minorities, e.g., 7/8 for the former and 1/8 for the latter. We shall then arrive at the distribution in Table 4. The patients who actually had or did not have cancer are indicated by C and not-C, respectively. Those who used or did not use the word „death“ are indicated by D and not-D, respectively.
Table 4 Table 5
Scharnberg’s Distribution Spence’s Distribution
The difference between the two distributions is conspicuously little. In both tables the larger numbers are found in the upper left and lower right cells: the majority of those who really had cancer used the word „death“, and the majority of those who did not have this disease did not use this word. And in both tables the minor numbers are found in the lower left and upper right cells: a minority of those who did not have cancer used the word, and a minority of those who had this disease did not use it. In other words, it seems that Spence’s empirical support for his own hypothesis is merely parasitic upon the hypothesis formulated by Scharnberg. We could hardly call it „Scharnberg’s hypothesis“ because it is merely borrowed from ordinary lay knowledge.
*XXIII. Direct Access to an Unconscious Oral and Regressive Fantasy: Statistical Flaws
The primary idea of Gordon & Spence is that they succeeded in  arousing  a fantasy that is  unconscious,  oral, and  regressive (= related to the infantile feeding situation). I shall describe their experimental design without psychoanalytic terminology. They started with excluded all unusually vigilant individuals. The remaining subjects were thrice divided into half groups. First they were divided into consolation eaters (C) and non-consolation eaters (non-C, indicated by a bar over C), and it was based on a questionnaire that dealt with conscious feelings and behavioural habits. In turn all subjects were in turn were given fake personality information, but it was so constructed to make half of them feel frustrated (F) and half of them self-confident (non-frustrated, indicated by a bar over F).
The third step was that all subjects were tachistoscopically presented with a stimulus for 1/150 seconds at low illumination, which was repeated 5 times. Half of them „saw“ the word MILK (M) and the other half saw a blank slide (non-M, indicated by a bar over M).
Still through the tachistoscope, but now at clearly visible duration and illumination, all subjects were presented with a series of 30 words, in the same random order for all subjects, and they would be asked to recall these words later.
Spence & Gordon classify the words into three categories:
“Infantile associates” (i.e., associates of the infantile feeding situation): baby, cry, formula, mother, sleep, suck, swallow, warm.
“Socialised milk associations”: butter, cold, cream, dairy, drink, glass, thirst, white.
“Buffer words”: book, comb, ear, game, idea, low, paper, pierce, porch, ring, road, speech, signal, watch.
When the subjects were asked to recall these words, two primary measures (a & b) and one subordinate measures (c) were used:
(a) the number of falsely recalled words (“importations”) of an oral regressive nature (I);
(b) the number of falsely recalled words (importations) which were not of an oral regressive nature (a bar over I);
(c) the number of correctly recalled oral regressive words minus the number of correctly recalled oral non-regressive words.
The outcome is shown in three geometric figures (pp. 111, 113, 115). It should not come as a surprise that here we are once more provided with comparison of pairs, so that the diagrams conceal most of the interrelations between the variables. In Table 6 and 7 I have converted the diagrams into numbers. Such numbers can only be approximate, but this is no obstacle because no great numerical precision is called for.
In addition I have displayed four variables in a matrix, so that genuine and exhaustive comparison can be made.
In Table 6 one cell is outstanding, and it is the right one from the writers‘ standpoint. Consolation eaters who had been frustrated and exposed to the milk stimulus falsely recalled an average of nearly 1 so-called oral regressive word. Each of the other 15 groups hardly deviates from zero.
In Table 7 one group is likewise outstanding, but this a false cell from the writers‘ view. Non-consolation eaters who had been frustrated and exposed to the blank stimulus correctly recalled more oral regressive words than any other group, and 76% more than the second highest group.
A long summary of no less than 497 words is attached to Spence & Gordon’s paper. Nevertheless, the second and embarrassing result is not even mentioned there.
The second result is neither mentioned in the extensive accounts of this experiment given by Dixon (1971:163ff., 1981:11ff) and Sjöbäck (1985:59ff.). Dixon fully accepts the evidential power of the study, and Sjöbäck supplies no hint as to the methodological errors.
*XXIV. Direct Access to an Unconscious Oral and Regressive Fantasy: Semantic Flaws
There is little reason to apply statistical tests of significance, because the writers‘ classification of the words is so arbitrary that nothing has been tested. Agreement with the questionnaire item “When I’m feeling blue I try to find something to eat” is no valid sign that the subject “use[s] food as substitute for affection”. The writers also took for granted before they started their study, that if a subject agrees with the just mentioned item, “an unconscious fantasy had been aroused which guided his behaviour”, and that “rejection would arouse a REPRESSED oral fantasy”.
What has the words “formula, sleep, swallow, warm” to do with the infantile feeding situation? And none of the other “infantile associates” has any exclusive relation to this situation. The “socialised milk associates” is likewise an arbitrary group.
But the greatest flaw is the classification of the importations: “bottle, milk, mouth, smell, taste” are taken to be oral regressive words (while “smoke” is a non-regressive word).
“Milk” (the mere repetition of the stimulus word) is not listed separately. It is quite possible that “juice” as the stimulus word would have had the same effect on recall and importations. But it would have a weaker persuasive effect on the reader. “Nipple”, given by a single subject, is the only word of some interest. Assuming that it was not a chance phenomenon, it would not be surprising if some consolation eaters were prone to use their mouth for other pleasures than eating, inter alia for kissing female breasts.
It should be noted that Spence & Gordon after nearly 30 pages admit in one single sentence, that their result might have a common sense explanation.
*XXV. Some Additional Comments
For more than a century it was known that hyper-brief exposition to visual stimuli may lead to distorted perception. It is unsurprising that several groups of proponents would use such phenomena to provide sham evidence of psychoanalytic theory.
There are also strong reason for examining the Swedish DMT groups and the American Silverman group together because of a number of reasons. The procedures of both groups are parasitic on facts that would be equally true whether psychoanalytic theory were true or false. But it should be noted that both groups acknowledge the results of each other, despite the fact that these results are in flagrant contradiction. The fundamental postulate of the American Silverman group is that hyper-brief exposition will enable the immediate observation of the unconscious processes themselves, before the observation is distorted by defence mechanisms. By contrast, the fundamental postulate of the Swedish group is that hyper-brief exposition will enable the immediate observation of the distorting processes of the defence mechanisms, before they are corrected by more rational processes.
I cannot help feeling that the Swedish postulate is curious. The physical stimuli or the physical „something“ impinging on the mechanism will have to be analysed sufficiently well to detect that it would be painful for the conscious mind to learn what it is. In order to protect the conscious mind from this pain, the defence mechanisms intrude and distort the something into a less painful but also less veracious message. However, when the defence mechanisms have accomplished their distortion, some other rational mechanisms detect the distortion and make the evaluation that the conscious mind would be mislead if it would receive the product of the defence mechanisms. Therefore the rational mechanisms undo the work of the defence mechanisms. The final outcome of the transformational chain is that the conscious mind will perceive the same painful thing that it would have perceived if no defence mechanisms had been involved.
This is certainly much ado about nothing.
It is known that if the same person take the DMT twice, the result may be opposite at the two times. A person may be tested before he is permitted to start training as a pilot. He may have undergone several hundreds of flying hours with the best performance. And just as a final „examination“ he may take the DMT once more – and then his career may be ruined, because the test result supposedly reveals that he is predestined to crash.
*XXVI. General Outlook
It took 24 years from DMT was constructed and until the first researcher outside the DMT group examined its validity. Eventually researchers inside and outside the DMT group, respectively, arrived at opposite results. The size and consistency of the discrepancy could not easily have emerged, unless the figures of the “insiders” were forged.
Moreover, the scores of the testee were for many years preserved in a central computer of the Swedish Air Force. These results would have been invaluable for disclosing the real value of DMT. Unfortunately, an unidentified person has deleted them.
Experimental studies were insufficient to get rid of DMT. But a logical analysis of DMT, the percept-genetic theory and the relevant parts of psychoanalytic theory could have been performed already in the late 1950s. And a logical analysis would have been much less expensive. And after the outcome of it had been obtained, it would be the obligation of the academic community not to accept Kragh’s and Neuman’s claims of empirical validity, until much more detailed empirical information from genuine empirical studies had been presented.
Suppose that the only alternative to accepting psychoanalytic theory would mean to be confronted with chaos of incomprehensible and isolated facts. In such a situation we might tolerate many internal defects of the theory. – But since it has turned out that there were never any observations to explain, there is no need for this theory.
If psychoanalysis is extrapolated to a specific selection test, the latter could only be valid by a miraculous coincidence. Furthermore, the derivation of the scoring code of DMT from psychoanalytic theory is arbitrary. Neither can the percept-genetic theory be derived from psychoanalysis.
But first and foremost, DMT and the theory of percept-genesis are internally inconsistent. It is incomprehensible why hyper-brief time of exposure of an external physical stimulus picture should lead to an unusually strong influence of defence mechanisms and, in turn, to other subsequent activities aiming at correcting their supposed effect.
 Ulf Kragh (1985): DMT: Defence Mechanism Test.
 Freud (GW‑XVI:43ff./SE‑XXII:255ff.)
 Freud (GW-I:37ff., 57ff., 75ff./SE-III:41ff., 157ff., 187ff.)
 Freud (GW-I:37ff., 57ff., 75ff./SE-III:41ff., 157ff., 187ff.)
 Letter of December, 17, 1896, quoted in Masson (1985:218)
 Scharnberg (1993, vol. I, p. 86f., 91f.)
 Anna Freud (1980: 85-94): Das Ich und die Abwehr Mechanismen.
 August Aichhorn (1959): Erziehungsberatung und Erziehungshilfe. Zwölf Vorträge über psychoanalytische Pädagogik.
 Freud (GW-XI:288f./SE-XVI: #page number to be added; around 278#)
 Felix Gattel (1898): Über die sexuellen Ursachen der Neurasthenie und Angstneurose. Gattel’s case histories are extensively analysed in Scharnberg (1993, vol. II, pp. 24-75), and all biographies of patients given the diagnosis “hysteria” are translated in toto into English.
 Freud (GW-V:161ff./SE-VII:1ff.)
 Quoted in Masson (1985:220f.)
 Bergler (1958, 1961, 1971)
 Jay Haley (1963)
 Scharnberg (1996, vol. II, chapter 57-68)
 Scharnberg (1984:122f.)
 The examples are found in Schilder (1952:81, 39), Bleuler (1955:428, 436, 451), Lindner (1946:149), Landis & Mettler (1964:166), Boss (1953:182), Laing (1960:48ff.).
 Fenichel (1945:424)
 Fenichel (1945:430)
 Fenichel (1945:156)
 Palladius (1912) which was written during the 5th century; furthermore, Lecky (1975) and Brundage (1987)
 Kline, P. & Cooper, C. (1977): A percept-genetic study of some defence mechanisms in the test PN.
 Sjöberg, Lennart & Källmén, Håkan & Scharnberg, Max (1997): Selection for stressful jobs: is the Defence Mechanism Test the so