Having explored the dubious history and origins of CBT, it is time to turn our attention to its theory, epistemology and methodology. Despite shrill appeals to science and reason, there has been much critique of CBT from an a priori perspective, three fine examples of which can be found here, here, and here. I will not repeat their points. Much of what I say here of CBT is applicable elsewhere in mainstream, Anglophone psychology. The failings of CBT in particular, and of psychology generally – and they are many and serious – are, in my view, both avoidable and instructive. I discuss them here in order to learn from them, with a view to constructing a better psychology (by which I do not mean merely a more refined CBT, or an empirical psychology with better-researched norms).
There is a disavowal at the heart of both CBT and empirical psychology more generally. Practitioners in both fields are aware, if only abstractly, that their objects of study, concepts and methodologies derive principally not from science, as such, but from various branches of philosophy, the social sciences and the humanities. Despite being aware of this, practitioners in both camps conduct their research as if this is not the case, as if one could do a ‘scientific’ psychology in complete ignorance of metaphysics and history. Many of the problems in both fields derive from this basic disavowal, and wherever there is disavowal, psychoanalytically speaking, there is perversion.
It is not difficult to notice, for instance, that so many of the key concepts of empirical psychology – ‘sensation’, ‘perception’, ‘emotion’, ‘cognition’ – are distinctly Latinate. To view this in further detail, the above concepts, for instance, are the translations and re-interpretations made by Catholic Scholastic philosophers of the Middle Ages of Greek terms and concepts, specifically those of Plato and Aristotle. A study or research program is only as strong as its weakest assumptions, which in the case of much of psychology, are literally those of medieval Catholic theology. This point, which may appear to be of purely historical significance, takes on some importance in CBT, as we shall see. Psychologists could give rigour to their discipline through a return to the humanities, to give some of these concepts some greater scrutiny. This would not constitute a science, but it may allow for some secure, a priori principles. As it stands, some empirical psychologists have made a Faustian bargain in which they have swapped the humanities for the science, but have ended up with something that has neither humanity nor science.
Thus it is precisely this – the historical, the philosophical, the a priori – that our psychological stamp collectors disavow. Thus, we observe an ever-more convoluted dependence on statistical analysis at the very moment that conceptual analysis has all but been abandoned in psychology. The accrual of statistics for junk concepts; rigid and dogmatic adherence to certain idiosyncrasies of method; a bland, tedious and artless attempt at ‘neutrality’ of writing style – these quirks are fetishised in psychology to the extent of being mistaken for science itself. This is academic psychology in its perverse, Sadean aspect – ‘One more statistical analysis of hypothetical constructs on a questionnaire, gentlemen, and we become scientists!’. This may produce publications, but not anything worth reading.
Of course, those who think that science consists of entirely evidence and numbers grasp nothing of science, evidence, or numbers. Yet this is the situation of psychology today. Psy-researchers correctly observe that mathematics is the hallmark of precision in the ‘hard’ sciences. Now, this sort of mathematical exactitude of formulae is entirely impossible and absurd in the psy-disciplines. Rather than confront this lack head-on, research tends to lapse into a physics envy, a grotesque and embarrassing aping of the ‘hard’ sciences, with statistics as a shabby prosthesis.
The notion of an unconscious part of the mind, which is beyond consciousness but nonetheless exerts an influence of subjectivity is all but ignored. Hence, in the construction of interventions such as CBT, there are presumptions of perfect self-reflexivity, of singularity of volition and will, of a lack of division (and a lack of lack itself) within a given person’s subjectivity. CBT also presumes that what is ‘good’, ‘enjoyable’, ‘pleasant’ and so forth are more or less self-evident, and that any rational person would pursue these things as ends. As assumptions, these are totally untenable, and render any research proceeding on these lines unempirical, and perhaps even insulting to would-be readers (and patients).
As one can see, the very foundations of CBT render its end products unviable for rigorous discourse. Nonetheless, if we turn to the basic elements of the theory, it can be summarised briefly as follows. Affects – depression and anxiety, for instance, but also the ‘positive’ ones – are not determined by events ‘external’ to the individual who experiences them, but rather, by a personal, subjective interpretation of said events. So far, so good. However, in case of pathology – again, depression and anxiety are the crucial examples – the individual’s subjectivity has been led astray by ‘cognitive distortions’. (An example may be a pessimistic depressive who says that ‘Things will never work out for me’. The proposition is refutable, and therefore ‘distorted’). According to Beck (and Ellis), these types of statements constitute errors in thinking that are responsible for the alleged pathologies of affect. The aim of CBT, therefore, is to systematically identify and remedy these unreasonable distortions, and indeed, to coach a subject into managing his or her own thinking through constant discipline and surveillance of mental life. The role of the clinician is always, implicitly or explicitly (depending on the sensitivities of the clinician) to be an expert in thought itself, an exemplar for the irrational patient, training the latter to think properly. Errors of thought can, according to Beck, be brought into line with certain norms and ideals, thus extinguishing pathological affects. In addition to this regimen of thought management and panel-beating of distorted cognitions, the clinician will generally prescribe behavioural exercises in the form of direct, specific advice, again, taking on a role of avowed expert. Sometimes, this treatment will take the form of a step-by-step manualised routine, which can then be followed in minute detail, standardised, and ‘empirically validated’ in the form of outcome studies.
If one takes Beck at his word – in his text from 1976, (reprinted since) for instance, Cognitive Therapy and the Emotional Disorders – this process is heavily directed and micromanaged by the clinician, not the patient. Echoing his ego psychology forebears, cognitive distortions amount to ‘maladaptive attitudes’ (p. 314-5), but whereas the ego psychologist is ‘chipping away’ at these indirectly, through interpretations, the CBT clinician tries to explicitly modify ‘unrealistic’ thoughts, to push the patient to ‘unlearn a bad habit’ (i.e. thinking in a manner the clinician finds irrational). Beck is absolutely clear about his system – if you are depressed, anxious or whatever, then your response, at a ‘cognitive’ level, at least, is ‘faulty’, and must be modified.
Plainly, this is a system of coercion and indoctrination. Perhaps it is implemented with sensitivity, with ‘empathy’, but it is a transparent attempt at thought control nonetheless. I intend to address its ethical and political dimensions later in this writing. For now, it is worth turning to its theoretical and ‘scientific’ aspects, since these, after all, provide the ideological laundering for the abhorrent ethics. So, let us ‘chip away’ at Beck’s exuberant irrationality, shall we?
One obviously shoddy assumption is the notion that CBT is an intervention into ‘cognition’. It manifestly is not, but rather, is an intervention into language, into discourse. A psychologist has no more seen or handled a patient’s ‘cognition’ than performed an exorcism of a demon. CBT is a regimen of language coaching – the distorted thinker must take his or her narrative, his subjective understanding of events, with all of its associations, slips, digressions, and fantastical elements, and reduce this into a series of refutable propositions. Indeed, patients are coached through ‘homework’ to express themselves in this way, to perform this act of subjective self-mutilation at the level of discourse. Banter, daydreams and the like – these are irrelevant to CBT. Yet if one accepts that one is dealing with language, and not with ‘cognition’, then it follows that the patient’s utterances are not mere statements of fact, refutable or otherwise, but an attempt at intersubjective dialogue. Language is always an exchange, and the act of exchange (much like gift-giving in Levi-Strauss) can be far more important than the specific thing exchanged. Moreover, whatever is being exchanged may have psychic resonances and implications far beyond the obvious. Nevertheless, CBT methodically ignores this complexity, choosing instead to first reduce the patient’s speech, then to brutally interpret along normative lines. It’s worth remembering here that, even for those most trained in thinking – say, philosophers and logicians, for instance – there is considerable dispute about what constitutes ‘good’ thinking (and good living, for that matter). In CBT, such considerations are excluded – the clinician is an expert in thinking (which is to say, a coach of a certain kind of speaking) even though he or she may never have read or understood a word of philosophy. It is akin to morbidly obese smokers taking over the dietetics and personal training industries. It may sound far-fetched to outsiders, but in psychology, this is standard practice.
Now, is there any evidence that propositional statements are the most significant in a person’s subjective, discursive life? That fantasy, dreams, politeness, gesture are entirely irrelevant? These are the things that CBT ignores. It behaves with a patient’s language rather like a marauding Viking, plundering what it can make use of, and sullying the rest. In effect, CBT coaches patients into silence, or worse, into expressly agreeing with the ‘positive’ and ‘rational’ opinion the clinician wishes them to hold about themselves. (Silence can, after all, be a form of resistance sometimes; in CBT, even this resistance is rejected. Patients do not have the right to remain silent; famously, Ellis, for instance, would throw out patients who declined to do their assigned ‘homework’.) This is well-illustrated in some examples given by Oliver James, in which, among numerous other failures, he documents one case of an overweight patient being told to pretend that ‘black was white’, and that she was conventionally attractive. CBT in these circumstances is an exercise in ‘adaptive self-deception’, that is, the marketing of lies.
In short, CBT’s claims about correcting cognition do not pass the most minimal scrutiny. Another pillar of the approach is the notion that thinking (of the ‘distorted’ kind) is a causal influence on emotion. Psychologists have long debated the relative primacy of intellect and affect. This is a clear indication of psychology’s Scholastic (and Platonic/Aristotelian) inheritance, because to set up causal relations of any kind between thought and emotion presupposes that they are separate in the first place. (The affect-intellect split was inaugurated in different ways by both Plato and Aristotle, and later taken up by the theologians, such as Augustine, where it overlayed the body-mind dichotomy). Again, this is another of CBT’s false assumptions. The very definition of an emotion always already possesses a ‘cognitive’ element to it. Emotion is never reducible to a bare physiological ‘state’. The setting up of false, fictitious dichotomies here (and then, ludicrously, establishing causal relations between them) was long ago refuted by the likes of Nietzsche, Bergson, Heidegger, Lacan…But alas, the experts of cognition have not read these people. Moreover, contemporary neuroscience does little to support CBT’s theory of emotions – the neuroscientific work of Damasio, or LeDoux, for example, suggests that there is no functional or neuroanatomical distinction between affect and intellect. Even when it superficially appears that this is the case – in Damasio’s patients who seemingly show a dissociation (through neural damage) between thought and feeling in Descartes’ Error – what appears to be a dissociation between these two functions shows itself, on closer inspection, to be a subtle but catastrophic failure in both. A treatment paradigm can hardly claim to be ‘evidence-based’ if its foundational ideas are nonsense, and the cognitive theory of emotion is demonstrably nonsensical.
Even if one accepted that CBT actually deals with cognitions, and that its model of how thought works is valid, one cannot but be struck by the bizarre hucksterism of its notion of rationality. ‘Rationality’ is itself a disputed term, but in any case, there is a distinction to be had between rationality of process and rationality of outcome. (Ironically, cognitive scientists are among those to have pointed out this distinction). Thinking one’s way through the world in the form of syllogisms might be formally ‘rational’, but socially and practically inept. (Arguably, such thinking is reminiscent of certain instances of psychosis or autism – as Les Murray once wrote in a poem about an autistic boy – ‘He lives in objectivity’). Yet again when assessing CBT, one sees that the ‘distorted’ thinking is not isolated to patients. The irony in this is that both the heart of CBT’s theory (‘wrong’ thinking) and the manner of its patronising presentation to patients (boxes and arrows, an ‘ABCs’ of cognition, etc), there is an unmistakeable view of patients as contemptibly stupid. Contempt masquerading as ‘reason’ – this is the tenor of CBT.
Even when illogical ideas are expressed by somebody – and most assuredly, they are, and not just by ‘mental health service users’ – taking a sledgehammer to them in the manner of Beck and Ellis is but one possible response. Dismissing such thoughts as intrinsically idiotic (‘You are catastrophising’, etc) is another means of inculcating silence or agreement in a vulnerable person. (Rather disturbingly, the silence or agreement is construed as treatment success in CBT). Other treatment approaches do not attempt to affirm illogical notions, but can nevertheless point them out to people without attempting to extinguish them. Language always presupposes an Other. Of course, if somebody utters something like, ‘My future is bleak’, this may be an erroneous ‘cognitive distortion’, as CBT would have it, yet it may also be an appeal to the other to enter into further discussion or to offer comfort; or it may be a repetition of discourse by some (specific and influential) other; or it may be a linguistic support for some underlying fantasy or enjoyment. In any case, to ‘correct’ it is necessarily to reduce it to silence, and prevent the patient from learning anything about it, or doing anything with it. It goes without saying that CBT is inimical to the millennia-old notion of the creative melancholic, for instance, who can make of his or her suffering something other than an occasion for positive-thinking slogans and techniques of distraction.
Which such illogicality and incoherence at the heart of the theory of CBT, it is little wonder that Beck and his followers have pinned their empirical hopes almost entirely to outcome studies. These studies have always been an exercise in playing with loaded dice, but even on their own rigged terms, most CBT outcome studies are worthless (lack adequate controls, methodology, statistical analysis, etc). But even when these things are not at issue, the outcome study is still of little merit. It derives from a reductionist and ham-fisted analogy between (physical) medical conditions, for which symptoms can often be straightforwardly quantified, and disorders of the mind. The problem with the latter is that, as the recent fiasco regarding the DSM-5 has shown, there is no coherent or rigorous definition of what might constitute a mental disorder (or, by extension, mental health), much less any reliable means of quantifying such a hypothetical entity. Psychiatric diagnosis under the DSM and ICD systems is both arbitrary, and, as some have argued, brazenly political. Outcome studies are, in the psy-disciplines, perhaps the clearest illustration of the flimsy house-of-cards epistemology at play. Subjects are assembled into groups on the basis of fictitious and arbitrary constructs (somehow devoid of comorbidity, which always somhow escapes the laboratory, but not the clinic). Taking matters to utter epistemological oblivion, these constructs are then instrumentalised into reified, quantifiable measures through the use of psychometrics, in which all subjectivity is done the conceptual violence of being rendered into numerically equivalent units. Subjects are then led through the fairytale of a ‘standardised’ treatment, as if individual, historical relationships between people could ever be ‘standardised’. Add statistics to the psychometrics, and this elaborate pea-and-thimble trick constitutes the jewel in CBT’s empirical crown, notwithstanding the fact that not a bit of it is valid in the least. It is a particularly useful methodology for CBT, since ‘disorder’ can be operationalized in terms of propositional statements. CBT can then be used for 10-20 sessions of persuading subjects not to utter such statements, and voila – treatment success! But, as with so many flimsy epistemologies, if the foundations collapse, so too does the rest, and on this basis, it is difficult to see how any adherence to CBT could be sustainable without the fragile ‘empirical’ edifice. Yet this is where disavowal kicks in – yes, practitioners say, we know that the numbers aren’t the end of the story, we know that diagnosis is flawed, we know that relationships cannot be sensibly be understood as standardised, but nevertheless – we will forget all of this, and do it anyway.
Thus, Randomly Controlled Trials (RCTs) have always been of severely restricted value when it comes to psychological problems, and psychological treatments. If CBT has succeeded at them better than other approaches, this is less evidence of its being ‘empirically validated’ than it is support for the notion that CBT is better at gaming the academic system, at reducing complex operations to an illusory quantification and standardisation.
It is true, of course, that many of the failings of CBT apply to other areas of psychology, and the continuity between CBT’s errors and those of psychology as a whole no doubt accounts, in part, for its success among academics. Again, going back to the medical analogy, psychologists often hold the mistaken belief that a patient’s symptoms can be abstracted from their person (or from the relationship – transferential – from which they are assessed and diagnosed). As if, much like influenza, a panic attack, phobia or obsession are merely incidental to a person’s life. (As it turns out, many of the most lethal and prevalent medical conditions are, in any case, often products of ‘lifestyle’ – heart disease, certain cancers, diabetes, etc). This twofold motion of separating the symptom from the subject who bears it, and then objectifying it with a view to silencing and modifying it, cannot but be an exercise in defacing subjectivity, intensifying pre-existing alienation. When viewed in detail, the ‘strategies’ in the CBT toolkit are so many disciplinary tactics for silence, alienation, and manipulation, whether they consist of positive-thinking hucksterism, silencing inconvenient propositional statements, encouraging self-regulation (which is to say, having a subject internalise regulation by an other), or regression to imaginary, narcissistic concerns (such as boosting of ‘self-image’, ‘self-esteem’, taking oneself qua object, and the like).
‘Science does not think’, said a certain philosopher. This is certainly the case for ‘scientific’ psychology, whether of the CBT sort, or Behaviourism, or psychometrics. Scientificity is assured by the repetition of certain operations and methodologies, precisely at the exclusion of individual judgement and subjectivity, which are anathema. Thus, many forms of risk assessment are undertaken on the basis of actuarial methods, excluding clinical opinion entirely. Behaviourism – now in a marriage of convenience with CBT – was explicitly designed to rely only on ‘direct’ observation, excluding the unobservable, the inferential, the subjective. Nevertheless, subjectivity is there at the start, in the concepts themselves, as well as in the conditions of visibility under which these concepts emerge, and thus, subjectivity is there at the end. So even with the most basic and ‘observable’ of cases, the Behaviourists need necessarily impose an interpretive framework onto the ‘direct’ observations to determine that which is stimulus, and that which is response. The actuarialist still has to extract from the data an operational framework that can be implemented in clinical settings, which is to say, at some stage, they are obliged to think, and thereby abandon their ‘science’, if only temporarily. They still interpret and draw inferences in the manner of a folk psychologist, but merely unconsciously (and badly). No doubt, subjectivity is a great inconvenience to a discipline with scientific pretensions, but the radical extent to which psychology has abandoned this inconvenience reminds one of the drunkard, searching for his keys exclusively beneath the lamp post, because that is where the light is.
Data never ‘speaks for itself’, and always requires subjective, human intervention to interpret it. In empirical psychology, this interpretation is often designed to ring-fence the paradigm du jour and protect it from embarrassment from inconvenient results. To take a recent example, when a recent outcome study of a relatively small sample of eating disordered patients compared CBT and psychoanalysis, to the benefit of the former, the empiricists did not hesitate to declare that, once again, the results showed that CBT as triumphant, and psychoanalysis a mere historical relic. When the Lancet published a similar comparison, but with a sample size three times as large as the other study, and with an outcome that did not favour CBT, our very same empiricists declared that there must simply have been something wrong with the data. Whilst this is but one case of hypocrisy, it illustrates a general rule in academic psychology, namely, that data that does not mesh with existing frameworks and positions is no data at all. It does not fit the paradigm, it cannot be assimilated, and is duly rejected. This is one of the reasons why the a priori is so important, and why our dedicated academics relentlessly ignore it, since the framework and its assumptions provide the clue to the ‘data’, and not the converse. This is diametrically opposed to the self-serving myth of scientist-practitioners whose theory is determined by data – the opposite is always the case. As in chess, theories, rules and frameworks circumscribe the moves of the game, determine what is a good or bad move, and what moves are inadmissible. This dogmatic adherence to rules in then adjudged ‘scientific’, the more so since the rules themselves are seldom questioned.
Subjectivity remains one of the great problems (and symptoms) of psychology. Attempts to objectify it – through questionnaires, for instance – inevitably distort what they purport to measure. (As another German philosopher once said, the finden is really an erfinden). If nothing else, it is clear that a rigorous psychology would be one that takes account of it.