There is a sentence that falls in every second anxiety treatment and ruins everything: “You have to go through it.”
It sounds like resolve, like candor, like the bitter medicine that works precisely because it is bitter. In fact it contains a theory of how fear disappears, and that theory is false. It is so thoroughly false that research abandoned it nearly twenty years ago—while it lives on in clinics, private practices, and self-help books as though nothing had happened.
This article is about what takes its place. And it is about a claim the reader will at first take for advertising: that walking into a situation avoided for years can be a real pleasure. Not relief afterward, not pride in hindsight, but exhilaration, delight, happiness, enjoyment—while it is happening.
The argument proceeds in six steps. First: the pleasure is not a promise but a documented fact—it occurs regularly and is destroyed just as regularly. Second: anxiety arises neither in the situation nor in the body, but in one’s thinking about both—in the judgment, the appraisal. Third: whoever fails to notice that judgment mistakenly takes the body to be the place where treatment must begin. Fourth: exposure research itself has demonstrated that its active ingredient lies in thinking. Fifth: from this follows a reversal of the order of treatment steps and the abandonment of the accompanying psychotherapist. And sixth: what is urged against this view.
The beginning lies twenty-five years back, on a farm track in the Rhineland-Palatinate countryside.
A Stick, a Pear Tree, and a Man Who Could Not Get Away
In the autumn of 2000, a psychotherapist from the Christoph Dornier Foundation for Clinical Psychology traveled to a man whom the journal Psychotherapie presented at the time, in a three-part series on appearance and reality in behavior therapy, under the name Helmut Mayer.1-3 He was forty years old and had suffered from agoraphobia—the fear of moving away from familiar surroundings—since a panic attack in the summer of 1982. His radius of movement was about five hundred meters. It had cost him his job and most of his life outside the house.
On the first day of treatment the two of them set out on foot. Beyond the line past which he could no longer have fled quickly enough, what he described as “the spiral of fear” came at him in very large strides. He wept. The psychotherapist stood calmly beside him. After two or three minutes it passed.
Then they walked on. And now something happened that appeared in no treatment protocol. The man saw a pear tree, fetched himself a stick, and knocked down a few pears, “which I did with pleasure.” They kept walking. And then comes the sentence this article is about: “It even slowly began to be rather fun, walking on my own the stretch of road I had been avoiding for years.”
In the afternoon he walked a good distance alone. That evening he felt, in his words, like a new man. He could have moved mountains.
Forty-eight hours later it was over. On the second day they drove: through the villages, into the county town, up to the shopping centers, into the industrial park, after the lunch break onto the autobahn, fifty kilometers, then through Koblenz, then on to the next town, then two more hours back across country. His verdict: “It was all too much, everything I experienced in that short time.” The next morning he broke off the treatment. His trust was gone.
Half a year later a second attempt, with a male colleague of the first psychotherapist, failed as well. At the end of that year the man wrote that he was back at the level of avoidance he had reached when the treatment ended.
Let us be precise about what this report proves and what it does not. It does not prove that exposure is ineffective. It proves something else, and for our question it is decisive: the pleasure was there. It came unbidden, it came on the very first day, it was noticed by the man himself and recorded in two independent formulations. It is therefore not a vendor’s marketing claim but an observation from a treatment that failed.
What it is not, is lasting.
Taking stock: anyone who regards the pleasure of overcoming anxiety as advertising must account for this report. The question is therefore not whether it exists, but why it so rarely survives. To answer that, one must first know what anxiety is made of.
Why the Same Symptoms Cost Money on a Roller Coaster
A person is standing in a department store. His heart is racing, he feels dizzy, his knees go weak, he has the sense that he is losing control of himself. He calls an ambulance. Two hours later he is discharged with unremarkable findings.
Five hundred meters away, another person waits in line for forty minutes in order to pay for exactly the same experience: racing heart, dizziness, weak knees, the sensation of surrendering control. He climbs out of the roller coaster and gets back in line.
The body is doing the same thing in both cases. What drives it is the same substance, adrenaline, and what it reports are the same sensations. The only difference is the judgment each man passes on what is going on inside him. One thinks: “I am dying.” The other thinks: “I am alive.”
Three things follow from this everyday observation, and they follow necessarily.
First, physical arousal is not the problem. If it were, the roller coaster would be a form of torture rather than a business model.
Second, the difference between panic and enjoyment does not lie in the situation. Nor does it lie in the body. It lies in a judgment the person passes on both—and that judgment is a thought.
Third, and this is what the title of this article rests on: change the judgment, and the same arousal flips from torment to delight. The pleasure, then, is not a bonus attached to anxiety treatment. It is the logical reverse side of the fear, released the moment the judgment is exchanged.
The insight is not new. Epictetus, a freed slave, fixed it in a single sentence some two thousand years ago: it is not things that disturb people, but their judgments about things.4 Psychology has reproduced the thought experimentally: Stanley Schachter and Jerome Singer showed in 1962 that one and the same artificially induced physical arousal is experienced as a different emotion depending on the circumstances in which it is interpreted.5 And from this David M. Clark developed, in 1986, the cognitive model of panic that still shapes the treatment of panic disorder: a panic attack arises when ordinary bodily sensations are catastrophically misread—palpitations as the herald of a heart attack, dizziness as the herald of collapse, light-headedness as the herald of insanity.6
The technical term for all this is cognition. It denotes nothing mysterious, simply thinking: a thought, a judgment, a conviction. And because the method at issue here works exclusively at that level, it is called cognitive psychotherapy.7
Taking stock: anxiety consists of two parts, an arousal and a judgment about it. The first part is the same in everyone and dangerous in no one. The second decides whether a person calls an ambulance or gets back in line. Anyone who wants to treat the anxiety must therefore know where to find it.
Why No One Notices His Own Thoughts
An objection presents itself immediately, and it comes regularly from patients themselves: “I’m not thinking anything. The fear is simply there.”
The objection is honest and it is wrong, and one has to hold both of those at once in order to get anywhere.
The thoughts in question are called, in the technical literature, automatic thoughts. The term is well chosen. They run off like the shifting of an experienced driver: fast, brief, effortless, without having to be produced deliberately, and above all without being retrievable afterward. Ask a driver what he was thinking as he pulled away and you get no answer. Not because he was thinking nothing, but because he no longer notices it.
What is perceived, therefore, is only the result: the pounding heart, the tightness, the urge to flee. And from this the sufferer draws the only conclusion available to him—that there is no thought at all, only the fear.
This conclusion is a fallacy, a textbook error of reasoning, and it has a name. One infers from non-perception to non-existence. The same trap catches anyone who denies speaking with an accent because he cannot hear his own, or who wears glasses and mistakes their distortion for the shape of the world because he does not see the lens he is looking through.
Now comes the step that matters, and the reader is asked to take it slowly, because it carries everything that follows.
A person who perceives only feeling, body, and behavior takes precisely those three levels to be the control level of his suffering. And whoever takes them for the control level will find every treatment plausible that intervenes there: at the feeling with reassurance, at the body with relaxation and breathing techniques, at the behavior with practice. Offer him instead a treatment that asks what he actually believes, and it strikes him as beside the point—it aims at a level he cannot see.
This gives behavioral exposure treatment an advantage that has nothing to do with its efficacy. It merely fits the patient’s own folk theory. It is vivid, it is filmable, it supplies dramatic television footage, and it answers the question “What are we doing now?” with a visible action rather than a conversation. Call this its plausibility advantage. It is considerable. It is not an efficacy advantage.
And it has an unpleasant consequence, which can be read in Helmut Mayer’s report. The patient agrees. He considers what is being done to him correct—“It must be right, what she’s doing with you here, it’s part of it”—because it matches his own theory. The agreement between treater and treated is not a mark of quality here. It is the same error, twice.
Taking stock: the most persistent misunderstanding in anxiety treatment arises not from stupidity but from a property of thinking: it works unnoticed. A person who cannot see the level on which his suffering is controlled will demand treatment on the levels he can see. And he will get it.
What Actually Happens During Exposure
A clarification is due at this point, so that no false impression arises. The guidelines name exposure treatment as the standard for anxiety, and they are right to do so: of all psychotherapeutic approaches to anxiety disorders, its efficacy is the best documented. Anyone who disputes that is arguing against the data.
The more interesting question is therefore not whether exposure works, but which component of it works. For a treatment always consists of several ingredients, and proof that the mixture works is not yet proof of which ingredient carries it.
The Old Answer: Habituation
The classical explanation was called habituation, or getting used to it. It holds that whoever endures long enough in the feared situation will find his physical arousal subsiding of its own accord, because no organism can sustain a permanent alarm; and as the arousal goes, so goes the fear. From this follows the practical instruction millions of patients have heard: stay in the situation until the fear declines, and on no account leave before.
Research has not borne this explanation out. Michelle Craske and her colleagues reviewed the evidence in 2008 and reached a conclusion that shook the founding assumption of exposure practice: neither the extent to which fear declines within a practice session nor the level of fear at the end of it predicts treatment outcome.8 What shows itself during the exercise is a performance in the moment—and a performance in the moment is not the same as what has been learned. Scoring well on a vocabulary quiz is not the same as having learned the language.
The New Answer: Disconfirming an Expectation
What does predict success is something else, and it carries an ungainly name: expectancy violation. What is meant is a simple process. The patient enters the situation with a specific prediction—“I will faint,” “I will lose control,” “They will laugh at me.” The prediction fails to come true. And the brain does not then erase the old association but forms a second, inhibitory one alongside it, which henceforth outvotes the first.9
The greater the distance between what was expected and what occurred, the more is learned. From this follows a rule that flatly contradicts the old one: what matters is not how long someone stays, but whether he went in with a clearly formulated expectation that can subsequently be disconfirmed.
Here too a direct comparison exists. Paul Salkovskis and colleagues set two procedures against each other in a pilot study: the same exercise, once with the rationale “stay until the fear subsides,” once with the rationale “find out whether the feared catastrophe occurs.”10 The rationale is the only difference—and it was not a matter of indifference.
With that the finding is clear, and it deserves to be stated plainly:
Exposure research has itself relocated the site of action to where cognitive psychotherapy always saw it: in the patient’s thinking. What is learned in the feared situation is not a bodily process of habituation but the refutation of a conviction.
And from this follows the conclusion that carries this article. Exposure is a delivery vehicle for a piece of information. It conveys the message “your prediction is wrong” into the patient’s head. Anyone who can deliver that message by another route does not need the vehicle in every case—any more than one needs a taxi to the library for a book one can read online.
Taking stock: the treatment with the best evidence does not work for the reason it was practiced for decades. It works because it disconfirms a conviction. That does not settle the dispute between cognitive and behavioral procedure—but it restates it.
Reversing the Order
Now it becomes possible to say precisely how the two procedures differ. It is not cognition versus behavior, and it is not practice versus no practice. It is the order.
The behavior-therapy order runs: experience first, insight second. The patient is brought into the situation, has an experience there, and the insight is supposed to follow from the experience. That can work. It presupposes, however, that the patient afterward evaluates the experience correctly—and no one has any control over whether he does.
The cognitive order runs the other way: insight first, experience second. First it is worked out what the patient actually believes, where he got it, what speaks for it and against it, and which error of reasoning it contains. Only when that work is done does he enter the situation—and not in order to endure it, but in order to check whether his new reasoning holds.
The difference in experience is enormous, and it is the heart of the matter. A man who passively endures the situation is undergoing an examination whose outcome he dreads. A man who thinks for himself and actively tests what happens is running an experiment whose outcome he expects. In the first case the patient is the laboratory animal. In the second he is the investigator.
And with that it also becomes clear where the practice takes place. It takes place where it takes place anyway: in everyday life. A person with an anxiety disorder meets the feared situations every day, whether he wants to or not—the elevator, the supermarket, the meeting, the phone call. He does not need to have any of it arranged for him. He need only walk in with corrected thinking.
Behavior therapy turns everyday life into an exercise. Cognitive psychotherapy turns the exercise back into everyday life.
What the Studies Actually Show
One should claim no more here than the evidence will bear, and the evidence is thinner than either side would like. Four findings are nonetheless pertinent.
First, the Oxford comparative trial in panic disorder. David Clark and colleagues assigned sixty-four patients to cognitive psychotherapy, applied relaxation, the antidepressant imipramine, or a waiting period. After three months, cognitive psychotherapy was superior to the other two treatments on most measures.11 For our question one detail matters more than the result: in every treatment arm, exposure took place as the patients’ homework—not as an accompanied exercise. The psychotherapist sat in the consulting room, not in the elevator.
Second, the abbreviated version. The same group tested in 1999 how much session time the effect actually requires. Five sessions with written self-study material between appointments proved equivalent to the twelve-session standard version; both achieved very large and practically identical effects, the shortened version with 6.5 hours of psychotherapist time including booster sessions.12 And the paper contains a secondary finding that is central here: the cognitive measures at the end of treatment predicted the patients’ condition twelve months later. It is not the number of sessions that carries the effect, but the precision with which the thinking was corrected.
Third, social phobia. Richard Mattick, Lorna Peters, and Christopher Clarke assigned forty-three patients in 1989 to three conditions: guided exposure, cognitive restructuring without exposure, and the combination of the two.13 The result is differentiated and should be reported exactly. The cognitive group and the combined group improved on all measures; the pure exposure group improved on the phobia measures but not on the attitudinal ones. Immediately after treatment, the purely cognitive group was weaker in actual approach behavior than the other two. By follow-up, however, it had continued to improve and had caught up with the exposure groups.
This pattern is remarkable because it is exactly what one would have to expect if the theory is correct: a person who has corrected his thinking but has not yet had the opportunity to apply it looks worse in the short run—and then, without further treatment, goes on improving on his own. He carries his tool with him.
Fourth, the question of the companion. Anjan Ghosh and Isaac Marks had agoraphobic patients carry out their exposure themselves, guided by a book, and compared them with psychotherapist-led treatment. The self-guided group did not differ significantly from the guided one.14 This too is no footnote: it shows that the presence of a professional during the exercise is not a necessary ingredient.
Taking stock: cognitive psychotherapy does not abolish the experience, it merely reverses the order. It supplies the insight and lets everyday life make a present of the experience—daily, without limit, and free of charge. The dispute is therefore not “practice or no practice,” but: who arranges the practice, and whose success is it?
Why Accompaniment Does Harm, Especially When It Goes Well
The question of the accompanying psychotherapist is now prepared, and it can be answered with a single argument.
A patient enters a feared situation. The catastrophe fails to occur. He leaves the situation. And now the question arises for him why nothing happened. Two explanations are on offer:
First: “My expectation was wrong. Nothing happens the rest of the time either.”
Second: “It went well because someone was with me.”
The technical term for choosing between such explanations is attribution. And the uncomfortable truth is that the anxiety patient reliably chooses the second. He does so neither out of malice nor out of stupidity, but because his disorder consists precisely in overestimating danger and underestimating his own capacity to cope. One cannot ask a person to judge against his disorder on the very question at issue.
The companion thereby becomes what the literature calls safety behavior: a measure that calms the person in the moment and for that very reason deprives him of the experience that would have cured him.15 Success in company does not demonstrate the harmlessness of the situation. It demonstrates the usefulness of the company.
That this is not speculation was tested experimentally by Salkovskis and colleagues. Patients with panic disorder and agoraphobia underwent a fifteen-minute exposure, one half instructed to drop their safety behavior, the other to maintain it. Those who dropped it showed, on repetition of the behavioral test within two days, a markedly stronger decline in catastrophic convictions and in anxiety.16 The same time, the same situation, a different yield—depending on whether the patient had kept a lifeline open.
And now open Helmut Mayer’s report again. On the second attempt at treatment he walks two kilometers alone across the fields to the neighboring village while the psychotherapist takes a detour. He shelters in a barn, waits, and writes: “But a feeling of happiness came over me as well. I, yes I, had managed it entirely on my own.”
Note when the happiness arrives. Not during the accompanied drive through Koblenz. But in the moment when no one was there.
Taking stock: the presence of the psychotherapist during the exercise is not an active ingredient. At best it is superfluous, and at worst harmful, because it offers the patient a convenient second explanation and thereby halves the yield of the experience.
Five Reasons It Is a Pleasure
It has been shown so far that the pleasure occurs and that everyday life is where it occurs. What remains is to explain what it feeds on. Five sources can be distinguished, and none of them needs the others.
First, the reinterpretation of arousal. A man who no longer believes that his racing heart is killing him experiences that same racing heart as what it is: excitement. And excitement is a state people seek out voluntarily. The road from the department store to the roller coaster runs through that realization.
Second, curiosity. A man testing a prediction wants to know how it turns out. A man enduring wants it to stop. The two states exclude each other, and only one of them is bearable. The report supplies the evidence: the new impressions on the forest path struck the patient as “somehow fascinating”—on the day, that is, when he was not yet overwhelmed.
Third, the disappearance of anticipatory anxiety. The greater part of the suffering in an anxiety disorder lies not in the situation but in the hours before it: in the brooding, the calculating, the search for excuses. This anticipatory anxiety hangs entirely on the prediction. Once the prediction is corrected, the lion’s share of the suffering falls away before the situation even arrives. One notices it first in being able to sleep again.
Fourth, undivided self-attribution. A success achieved alone belongs to one entirely. Albert Bandura coined the term self-efficacy for the resulting conviction that one can handle a task, and showed that it grows most strongly out of one’s own mastery.17 With a professional standing beside him, the yield is split—and the patient gets the smaller half.
Fifth, the satisfaction of being right. The patient has advanced a hypothesis: “Nothing will happen, and here is why.” It is confirmed. That is the pleasure of the experimenter, and it is peculiarly available to someone who has been overwhelmed by his own thinking for years. For the first time in a long while, he experiences his head working for him instead of against him.
This combination, incidentally, has a famous precedent. The often-cited example of the acrophobic Goethe, who climbed the tower of Strasbourg Cathedral and stayed at the top until the fear had gone, is regularly adduced as evidence for habituation. It is nothing of the kind. Goethe first analyzed his thoughts and gave himself an account of what he feared and why it was unfounded—he went up as an examiner, not as a sufferer. What befell him up there would today be called the earned success of the cognitive preparation and correction of thinking that Goethe carried out on himself as his own psychotherapist.
Taking stock: the pleasure is not a by-product of successful treatment but its distinguishing mark. Where overcoming anxiety is experienced purely as an ordeal, the cognitive work has in all probability been omitted.
How to Destroy That Pleasure
It is now possible to answer what happened to Helmut Mayer. Once one knows the foregoing, his report reads like a textbook of avoidable errors.2,3
He asked questions and got no answers. “I always had so many questions I wanted to ask, but they were never really answered,” he writes of the preliminary interview. He would have liked to know where his fear came from in the first place.
He was told why he would not be told. The reply was that one should not go rummaging around in the past, since that yielded no solutions; the fear was simply there and could only be brought under control by living through it again. This answer confuses two entirely different questions. The question about the prehistory may well be dispensable—the question about the present conviction is not. The patient asked the second and was fobbed off with a refusal of the first.
He was given a dose he could not process. The second day, with six and more hours of driving through unfamiliar cities, was wrong not because it was strenuous but because the patient lacked the tool with which to evaluate what he had experienced. He saw it himself: “As I now realize, it would have been better for me to break off the therapy for that day and work through the experiences I had gained.” He was asking for precisely the cognitive work that was withheld from him.
His catastrophic expectation was left standing. On the second attempt he asked the psychotherapist what would happen if he dropped dead along the way. The answer: “Then I’ll call the hearse.” That is the exact inversion of cognitive work. The patient’s conviction that he might drop dead was not examined but acknowledged with a joke and left unrefuted. He then marched off—believing he might die doing it. That is bravery. It is not psychotherapy.
And there was displeasure when he became independent. When, after the successful solo walk, he proposed stopping for the day so that his positive impressions would not tip into their opposite, and announced that he would carry on alone, the psychotherapist “became downright angry.” In that moment the patient was doing precisely what every good treatment declares to be its goal. It did not fit the program.
Carmen Heerdegen, the physician who commented on the report for Psychotherapie at the time, named the standard in a single sentence: it counts as a breach of good practice to conduct exposure with a patient without adequate cognitive preparation. Both psychotherapists, in her judgment, had merely taken the patient out without doing any cognitive work with him—something a neighbor could have managed no worse.2
The bitterest detail stands at the end of the report. After both failed treatments, the man builds himself a fear hierarchy: twenty stages, each about one hundred fifty meters apart, remaining at each stage “until I am completely free of anxiety,” and repeating that at least three times per stage.
Read that closely. He has taken over the doctrine of habituation so thoroughly that he now applies it against himself—including the very condition Craske later showed to be unsupported. They did not take his anxiety away from him. They left him a method for administering it.
Taking stock: in this case the pleasure was present, repeatedly and unbidden. It was destroyed not by severity but by the absence of cognitive preparation and explanation. A person who does not know why it went well can make nothing of the good outcome.
The Case Against This View
It would be convenient to stop here. It would also be dishonest, for there is a serious contrary finding, and it comes from the largest study addressing precisely this question.
A German multicenter trial led by Andrew Gloster and Hans-Ulrich Wittchen assigned 369 patients with panic disorder and agoraphobia to two versions of the same cognitive behavior therapy. The two were identical in content, structure, and length but for a single point: in one group the psychotherapist planned the exercises outside the consulting room and went along; in the other he likewise planned and discussed them but did not accompany the patient. The result: the accompanied version was superior to the merely instructed one in agoraphobic avoidance, in general functioning, and in panic attacks during the follow-up period.18
This finding cannot be talked away, and it should not be. Anyone arguing for doing without therapist-guided exposure must be measured against it.
Three observations are nonetheless in order, and the third is an interpretation, which shall remain marked as such.
First, the trial did not compare what is at issue here. Both conditions were exposure treatments. What was compared was accompanied against unaccompanied exposure—not exposure against the examination of the anxiety-producing convictions. The question of whether a patient whose expectation has first been carefully corrected still needs a companion was not the question the study asked.
Second, the authors themselves name the mediating mechanism: the accompanied patients entered the feared situations more often and for longer. The advantage therefore arose not because the professional’s presence taught anything, but because it set the patient in motion.
And third, as interpretation: that is exactly what one would expect when the conviction has not been worked on. A man who still believes the thing might kill him, and who has merely been instructed to do it anyway, will do it less often. A companion helps him the way a driver helps someone unwilling. The finding thus shows how expensive it is to leave the conviction standing—it does not show that correcting it is dispensable. That this reading is plausible does not yet make it a result; it is a construal, and the reader is free to disagree.
With that the decisive gap is named, and let it be said openly: the study that sets the examination of convictions without therapist-guided exposure against therapist-guided exposure at equal contact time has to this day not been carried out. Anyone who claims the question is settled is selling a conjecture as a result—whichever side he is on.
Two further qualifications belong here. The Mattick study cited above showed the purely cognitive group weaker in approach behavior immediately after treatment; anyone quoting only the follow-up point withholds half of it. And the view that safety behavior is harmful in every case has been contested in recent years; there is evidence that it does not always impair the effect of the exercise.19 A person who has brought his thinking and his expectations into line with the risks of real life through sound cognitive preparation will be able to ignore the safety offerings with equanimity, even when psychotherapists provide or press them on him.
Finally, suitability—and this point is no fine-print clause but the condition of the method. Whoever wants to change his thinking must take on the effort of thinking. He has to read, to write, to be contradicted, and to work on his own between contacts. Anyone who expects the change to be done for him is in the wrong place in a cognitive psychotherapy—not because he is unwelcome, but because the method would fail on him: “Without thinking for yourself there is no insight of your own—and no psychotherapeutic self-knowledge either.”20
What clinical experience has to say. The evidence base is one thing; observation across more than three decades of practice is another. It shall be stated here, but as what it is: not a controlled study, but clinical experience. It is unambiguous.
A person who has understood his errors of reasoning and wants to be rid of his anxiety goes out and practices alone—and gladly. He asks for no companion, because he needs none.
Where the anxiety, by contrast, performs a service the sufferer does not wish to give up—it keeps demands at bay, it secures him consideration or a disability pension, it excuses him—there even the most careful accompaniment achieves nothing. It merely prolongs the treatment.
This observation must not, however, serve as an excuse for failures. Anyone who explains a failed treatment by saying the patient simply did not want it has made his claim irrefutable and therefore worthless. It becomes testable only if both conditions are established before the treatment rather than after it: the understanding in the shared work on the thinking, the willingness in the assessment of suitability. That is precisely why psychodiagnostics stands at the beginning and not at the end.
Taking stock: the strongest contrary finding concerns a different question from the one at issue here, and the decisive comparative trial is missing. What is established: the active ingredient of behavioral exposure lies in thinking. What has not yet been investigated is the question: does a patient whose thinking has first been corrected still need an accompanied exercise at all?
The Price of the Pleasure
Let the claim in the title finally be brought to its proper measure, so that no one mistakes it for a promise of cure.
The pleasure does not lie in the session. Sessions of this kind are strenuous; what is contested there are convictions a person is attached to, and nobody enjoys losing a certainty he has lived with for twenty years. The pleasure lies afterward: in walking into the situation one has avoided, with a way of thinking that no longer expects a catastrophe there.
Nor is it to be had for nothing. It is merely not agonizing. The patient has to do the work, and no one can do it in his place—no psychotherapist, no drug, no program.
And in that lies the real yield, which reaches beyond the single fear. A person who has learned to formulate a conviction of his own, to test it, and if necessary to drop it, has not lost an anxiety but gained a method. He can apply it to the next situation, and to the one after that, without anyone having to be present. The goal of a good psychotherapy is not the cured patient. It is the patient who no longer needs it, because he has become his own psychotherapist.
There remains the man on the farm track. On that September day he experienced everything that makes a treatment successful: the refuted expectation, his own mastery, the happiness of walking alone, even the high spirits of knocking pears off a tree with a stick. What he lacked was a single sentence explaining his error of reasoning—why nothing he feared had happened.
Supplying that explanation is the whole art of cognitive psychotherapy. And it is the difference between cognitive psychotherapy and behavior therapy.21
Frequently Asked Questions About Cognitive Psychotherapy for Anxiety
Can Overcoming Anxiety Really Be a Pleasure?
Yes—when the sufferer no longer endures the situation but tests it. A man enduring wants it to stop. A man testing a conclusion of his own wants to know how it turns out. These are two different states, and only one of them is bearable. That the pleasure does in fact occur is documented even in failed treatments: patients report feelings of happiness precisely at the moment when they have managed something on their own.
Does Cognitive Psychotherapy for Anxiety Work Without Practice?
No. It works without the accompanying psychotherapist, not without the experience. A routine of avoidance ingrained over years persists as mere habit even once its reason has fallen away; it has to be overwritten, and that requires repetition. The dispute is therefore not “practice or no practice,” but: who arranges the practice, and whose success is it?
Does Exposure Work Through Habituation?
By current knowledge, no. Neither the extent to which fear declines within a practice session nor the level of fear at the end of it predicts treatment outcome. What does predict it is the disconfirmation of an expectation: the patient anticipates a catastrophe, the catastrophe fails to occur, and the brain forms a new, inhibitory association alongside the old one. That is a process in thinking.
Why Should the Psychotherapist’s Presence Do Harm?
Because it offers the patient a second explanation. If the catastrophe fails to occur while he is accompanied, he can attribute that failure to the accompaniment rather than to the untenability of his expectation. The companion thereby becomes a safety signal, and safety behavior sustains the anxiety disorder. It has been shown experimentally that patients who drop their safety behavior during the exercise lose their catastrophic convictions more markedly than those who retain it.
Why Do So Many People Take Behavior to Be the Right Point of Attack?
Because they do not notice their own thinking—their opining and judging. Automatic thoughts are brief, fast, and self-evident; what is perceived is only their result—palpitations, the urge to flee. From this it is concluded that there is no thought at all, only the fear. That is a fallacy inferring non-existence from non-perception, the same one that catches a man who denies speaking with an accent because he cannot hear his own.
Are There Studies on Cognitive Psychotherapy Without Exposure Exercises?
There are few, and they do not point in one direction. In social phobia, a group receiving cognitive restructuring without exposure improved in all areas and continued to gain up to follow-up, but was weaker in approach behavior immediately after treatment, because it had initially been denied the opportunity to practice. In panic disorder, cognitive psychotherapy was superior to applied relaxation and to the antidepressant. The decisive comparative trial—examination of convictions against therapist-guided exposure at equal contact time—has to this day not been carried out.
What Speaks Against Doing Without Therapist-Guided Exposure?
The weightiest objection comes from a German multicenter trial with 369 patients.18 There, accompanied exposure was superior to the merely instructed version in agoraphobic avoidance, in general functioning, and in panic attacks during the follow-up period, apparently mediated by the fact that the accompanied patients entered the feared situations more often and for longer. It is an open question whether the cognitive preparation was perhaps insufficient to enable the unaccompanied patients to become their own psychotherapists. This finding is discussed at length in the article.
For Whom Is This Approach Unsuitable?
For anyone unwilling to think for himself, to read, to write, and to work independently between contacts. Anyone who expects the change to be done for him is in the wrong place. That is not a restriction in the fine print but the condition under which cognitive psychotherapy works at all.
1 Seelen, G.: Behavior Therapy: Appearance and Reality. Part 1: “I went through the Christoph Dornier Clinic, and I’ve been broke ever since”. Psychotherapie. 12/31/2001.
2 Mayer, H.: Behavior Therapy: Appearance and Reality. Part 2: A Patient’s Account of Exposure Treatment for Agoraphobia and Panic Attacks. Psychotherapie. 12/31/2001.
3 Mayer, H.: Behavior Therapy: Appearance and Reality. Part 3: “Then I’ll call the hearse”. Psychotherapie. 12/31/2001.
4 Epictetus: Enchiridion, ch. 5. Edition used: Epiktet, Teles, Musonius: Ausgewählte Schriften. Griechisch - Deutsch. Edited and translated by Rainer Nickel. Zurich: Artemis Verlag, 1994, p. 15. [English wording after the public-domain translation by Elizabeth Carter, 1758: “Men are disturbed, not by things, but by the principles and notions which they form concerning things. Death, for instance, is not terrible, else it would have appeared so to Socrates. But the terror consists in our notion of death, that it is terrible. When, therefore, we are hindered, or disturbed, or grieved, let us never attribute it to others, but to ourselves—that is, to our own principles.”]
5 Schachter, S.; Singer, J.E.: Cognitive, social, and physiological determinants of emotional state. Psychological Review, 1962, 69(5), 379-399. doi.org/10.1037/h0046234. [Parts of the study have not replicated; what is drawn on here—that the interpretation of arousal co-determines the emotion experienced—is uncontested.]
6 Clark, D.M.: A cognitive approach to panic. Behaviour Research and Therapy, 1986, 24(4), 461-470. doi.org/10.1016/0005-7967(86)90011-2.
7 Luchmann, D.: What Is Cognitive Psychotherapy? Psychotherapie. 07/22/2026.
8 Craske, M.G.; Kircanski, K.; Zelikowsky, M.; Mystkowski, J.; Chowdhury, N.; Baker, A.: Optimizing inhibitory learning during exposure therapy. Behaviour Research and Therapy, 2008, 46(1), 5-27. doi.org/10.1016/j.brat.2007.10.003. [The authors state on p. 5 that neither the degree of fear reduction nor the fear level at the end of the session predicts therapeutic outcome.]
9 Craske, M.G.; Treanor, M.; Conway, C.C.; Zbozinek, T.; Vervliet, B.: Maximizing exposure therapy. An inhibitory learning approach. Behaviour Research and Therapy, 2014, 58, 10-23. doi.org/10.1016/j.brat.2014.04.006.
10 Salkovskis, P.M.; Hackmann, A.; Wells, A.; Gelder, M.G.; Clark, D.M.: Belief disconfirmation versus habituation approaches to situational exposure in panic disorder with agoraphobia. A pilot study. Behaviour Research and Therapy, 2007, 45(5), 877-885. doi.org/10.1016/j.brat.2006.02.008. [Per the abstract, p. 877: patients in the belief-disconfirmation condition improved significantly more on self-reported anxiety, panic, and situational avoidance, completed significantly more steps of a standardized behavioral walk while experiencing significantly less anxiety, and produced controlled effect sizes ranging from 1.7 to 2.7. Pilot study with a small sample; it supports the direction of the finding but does not carry it alone.]
11 Clark, D.M.; Salkovskis, P.M.; Hackmann, A.; Middleton, H.; Anastasiades, P.; Gelder, M.: A comparison of cognitive therapy, applied relaxation and imipramine in the treatment of panic disorder. The British Journal of Psychiatry, 1994, 164(6), 759-769. doi.org/10.1192/bjp.164.6.759. [n = 64. All treatment conditions included self-directed exposure as homework. At fifteen months, per p. 759, “cognitive therapy was again superior to both applied relaxation and imipramine.”]
12 Clark, D.M.; Salkovskis, P.M.; Hackmann, A.; Wells, A.; Ludgate, J.; Gelder, M.: Brief cognitive therapy for panic disorder. A randomized controlled trial. Journal of Consulting and Clinical Psychology, 1999, 67(4), 583-589. doi.org/10.1037/0022-006X.67.4.583.
13 Mattick, R.P.; Peters, L.; Clarke, J.C.: Exposure and cognitive restructuring for social phobia. A controlled study. Behavior Therapy, 1989, 20(1), 3-23. doi.org/10.1016/S0005-7894(89)80115-7. [n = 43, waiting-list controlled. The group receiving cognitive restructuring without exposure was inferior to the exposure groups in approach behavior immediately after treatment and continued to improve up to follow-up. The authors conclude on p. 21 that “cognitive restructuring without exposure instructions” can affect behavioral aspects of the disorder over the long term.]
14 Ghosh, A.; Marks, I.M.: Self-treatment of agoraphobia by exposure. Behavior Therapy, 1987, 18(1), 3-16. doi.org/10.1016/S0005-7894(87)80047-3.
15 Salkovskis, P.M.: The importance of behaviour in the maintenance of anxiety and panic. A cognitive account. Behavioural Psychotherapy, 1991, 19(1), 6-19. doi.org/10.1017/S0141347300011472. [The argument in the original: because safety-seeking behavior is experienced as preventive and is aimed at especially catastrophic outcomes such as death, illness, or humiliation, it makes spontaneous disconfirmation of the feared threat particularly unlikely.]
16 Salkovskis, P.M.; Clark, D.M.; Hackmann, A.; Wells, A.; Gelder, M.G.: An experimental investigation of the role of safety-seeking behaviours in the maintenance of panic disorder with agoraphobia. Behaviour Research and Therapy, 1999, 37(6), 559-574. doi.org/10.1016/S0005-7967(98)00153-3. [n = 18, fifteen-minute exposure, behavioral test repeated within two days.]
17 Bandura, A.: Self-efficacy. Toward a unifying theory of behavioral change. Psychological Review, 1977, 84(2), 191-215. doi.org/10.1037/0033-295X.84.2.191.
18 Gloster, A.T.; Wittchen, H.-U.; Einsle, F.; Lang, T.; Helbig-Lang, S.; Fydrich, T.; Fehm, L.; Hamm, A.O.; Richter, J.; Alpers, G.W.; Gerlach, A.L.; Ströhle, A.; Kircher, T.; Deckert, J.; Zwanzger, P.; Höfler, M.; Arolt, V.: Psychological treatment for panic disorder with agoraphobia. A randomized controlled trial to examine the role of therapist-guided exposure in situ in CBT. Journal of Consulting and Clinical Psychology, 2011, 79(3), 406-420. doi.org/10.1037/a0023584. [n = 369; accompanied condition n = 163, instructed condition n = 138, waiting list n = 68. The authors attribute the advantage to more frequent physical entry into the feared situations. Both conditions were exposure treatments; a condition without exposure was not tested.]
19 Rachman, S.; Radomsky, A.S.; Shafran, R.: Safety behaviour. A reconsideration. Behaviour Research and Therapy, 2008, 46(2), 163-173. doi.org/10.1016/j.brat.2007.11.008. [The evidence on the effects of safety behavior is mixed; for a summary see Blakey, S.M.; Abramowitz, J.S.: The effects of safety behaviors during exposure therapy for anxiety. Critical analysis from an inhibitory learning perspective. Clinical Psychology Review, 2016, 49, 1-15.]
20 Luchmann, D.: Why Cognitive Psychotherapy Is Demonstrably the Most Effective — Yet Helps the Fewest. Psychotherapie. 03/31/2026.
21 Luchmann, D.: Cognitive Behavior Therapy and Cognitive Psychotherapy — the Difference. Psychotherapie. 06/04/2026.
Written Cognitive Psychotherapy (WCP) by Dr. Dietmar Luchmann, LLC, provides assistance for self-help to enable the self-healing of psychological disorders:
1. Discover WCP
2. Take Suitability Assessment
3. Start Self-Therapy
Your Comment
Are you interested in contacting the author? Do you have remarks, suggestions, or additions regarding this article? Do you have personal therapy experiences? We welcome substantial feedback.