Evidence Against the Doctrine of “Symptom Substitution”
One traditional criticism levelled at hypnotherapy by psychoanalytic therapists was that its benefits must be temporary unless it attempted (as in regression) to “get to the root” of the problem by analysing its remote childhood causes. This theory was first propounded by Freud, on the basis of a tiny handful of cases,
[Hypnosis] could be employed in certain cases only and not others; with some much could be achieved by it, and with others very little, one never knew why. But worse than its capricious nature was the lack of permanence in the results; after a time, if one heard from the patient again, the old malady reappeared or had been replaced by another. (Freud, 1920: 157)
Indeed, Freud and his followers tended to argue that any therapy which attempted to directly remove symptoms without analysing their unconscious root cause, in the so-called “Oedipus Complex”, would lead to temporary improvement while leaving the client vulnerable to relapse and deterioration. As all symptoms, on the original psychoanalytic model, were viewed as disguised representations of unconscious complexes, new symptoms were expected to occur which would symbolise repressed material in different ways a “return of the repressed.”
In fact, this was merely a supposition made by Freud, an objection based on theory rather than observed facts. However, it was not until the introduction of behaviour therapy that hypnotherapy found its first major ally in disputing this hypothesis. However, as Marks rightly points out, it seems absurd that this presupposition is confined to psychotherapy which is symptom-focused whereas nobody seems to “fear the dragon of symptom substitution” when prescribing common psychiatric medication, such as tranquilisers (Marks, 1981, p. 237). In cases where medication is used to remove symptoms, there is clearly no evidence of new substitute symptoms erupting from the unconscious mind in symbolic form, contrary to the prediction made by psychoanalytic theory. Hence, in the opening salvos of the “psychotherapy wars”, Eysenck announced that contrary to the assumptions of Freud, symptom substitution was essentially a myth.
How about the return of symptoms? I have made a thorough search of the literature dealing with behaviour therapy with this particular point in view. Many psychoanalytically trained therapists using these methods have been specially on the outlook for the return of symptoms, or the emergence of alternative ones; yet neither they nor any of the other practitioners have found anything of this kind to happen except in the most rare and unusual cases. […] relapses occur, as indeed one would expect in terms of learning theory under certain circumstances, but they quickly yield to repeat treatment. […] Nor would it be true that alternative symptoms emerge; quite the contrary happens. The disappearance of the very annoying symptom promotes peace in the home, allays anxieties, and leads to an all-round improvement in character and behaviour. […] Once the symptom is removed, the patient is cured; when there are multiple symptoms, as there usually are, removal of one symptom facilitates removal of the others, and removal of all the symptoms completes the cure. (Eysenck, 1960: 12-13)
Even when psychoanalysts turned to the empirical evaluation of this theory, their own results were negative,
Mowrer, having accepted, as we have seen, Freud’s conclusion with regard to the meaning and function of symptoms, was considerably embarrassed by his own empirical finding that “symptomatic” treatment of enuresis was not only 100 percent successful with regard to the symptoms, but was not followed by symptom substitution in a single case! (Yates, 1958, in Eysenck, 1960: 22)
Likewise, when Azrin and Nunn carried out direct habit reversal treatment with over 300 subjects, they reported remarkable success in breaking habits such as nail-biting, hair-pulling, stammering and tics, of the kind traditionally treated by Freudian psychoanalysis. However, despite changing the habit symptoms directly, without attempting to interpret their “unconscious root”, they found no evidence whatsoever of symptom substitution (1977: 32). Freud himself forwarded no evidence whatsoever for this theory other than anecdotes based on his own limited clinical experience with a small sample of clients. It is inconsistent with the clinical experience of most modern therapists. However, it has seeped into popular culture and you will find many clients who have internalised this view in the form of a superstition or a “myth” about therapy.
Behaviour therapists passionately argued that symptom substitution was a superstition and only likely to occur if the client had been led to expect it to happen. Of course, in some cases clients may solve one problem while neglecting another, but it is rare that this would lead to new symptoms. Lazarus carried out a detailed follow-up study of 112 clients who were treated by him using behaviour therapy. He could find indications of symptom substitution in only five or six cases (5%) and even these were classed as “tenuous.” Similarly, Kroger & Fezler confidently assert that there is “no evidence” that psychodynamic symptom substitution exists (1976: 79). Even Weitzman, a psychoanalytic therapist, in an article openly critical of behaviour therapy, accepts,
It has been pointed out, from both camps, that analytic theory requires that symptom substitution or recurrence must attend a symptomatic treatment which, by definition, does not affect the dynamic sources of the symptoms. The evidence is rather impressive that neither substitution nor recurrence typically follows treatment by systematic desensitization. When occasional recurrences are reported, they are described as being of low intensity and, apparently, never catastrophic. (Weitzman, 1967: 301)
Drawing on evidence from reviews by behavioural researchers including Bandura, Lazarus, Paul and Wolpe, Rimm & Masters conclude,
Reviews of empirical findings (including case histories and controlled experiments) indicate that the evidence is overwhelmingly against symptom substitution. (Rimm & Masters, 1974, p. 10)
Evidence from behaviour therapy unequivocally demonstrated that this psychoanalytic theory was false, and that people did improve as a result of direct symptom removal, without analysis of their past. Indeed, the results of behaviour therapy were often much quicker and more reliable than anything that could be hoped for from psychoanalytic approaches. As one behavioural hypnotherapist, explains,
Once these changes start to occur, they will become self-perpetuating. You will realise you can cope with what once seemed formidable problems, and so you will approach other situations with far greater expectations of success. The therapeutic effects of hypnosis and self-hypnosis are undoubtedly ongoing and permanent. (Jackson, 1990: 30)
Indeed, as Freud himself had originally stated (1895), when any symptom is removed clients will generally develop more ego-strength, growing in confidence in a way that tends to make them improve across the board in other areas of their life. For example, when people successfully quit smoking, it is rare (though not impossible) for them to substitute some other negative behaviour. Most people feel more empowered and tend to improve in other areas of their life as well, creating a kind of positive “domino effect.”
The notion of “symptom substitution” is a particularly insidious one as it discourages clients from making practical changes that are well within their sphere of control. This is most notable in depressed clients who typically suffer from a lack of initiative and motivation and are further de-motivated by the excuse that the myth of symptom substation offers. They may complain that there is “no point” fixing one problem until they have solved their “underlying” character problem. In most cases, this is not a realistic goal, however, and it is many small changes which improve the quality of life for most people. Notably, psychoanalysis has been blamed for worsening the condition of some depressed clients. Likewise, after reviewing relevant outcome data from a wide range of independent studies, Bandura was tempted to speculate that the predictions of “dire consequences” resulting from symptom substitution were little more than scare-mongering by psychoanalytic therapists, attempting to stifle innovations in symptom-focused treatment (Bandura, 1969, p. 48).
The distorted “grain of truth” in the theory of symptom substitution, as Eysenck (1960: 13) notes, is that where the client suffers from an (autonomic) emotional reaction and their (psychomotor) behaviour is reconditioned without addressing their underlying mood, they may relapse or seek another behaviour to alleviate their inner distress. For example, someone who bites their nails to cope with stress may relapse or begin grinding their teeth instead if this habit is suppressed directly, unless they are also alleviated of the emotional arousal associated with stress, e.g., by desensitisation therapy.
Thus, there is no axiom of behaviourism which precludes the substitution of one maladaptive behaviour for another. But from a practical point of view, it is a phenomenon only rarely observed. (Rimm & Masters, 1974, p. 10)
This is a far cry, moreover, from the Freudian notion of “symptom substitution” due to unconscious dynamics, and only a seriously incompetent therapist would attempt to remove a self-comforting habit without also addressing the associated emotions. It isn’t a question of removing the symptom and its cause, but rather one of removing all of the symptoms from a mutually inter-dependant and self-maintaining cluster.