The Art of Being a Failure as a Therapist

I read Jay Haley’s The Art of Being a Failure as a Therapist many years ago. I remember being struck by how funny it was—and by how little of it felt dated. Reading it again now, I have much the same reaction.

Psychiatry has a peculiar library. Papers about receptors, medications, or technologies may be overtaken by new evidence. Papers about the habits of clinicians often age much more slowly. Our instruments change faster than our temptations.

Haley’s article appeared in the American Journal of Orthopsychiatry in July 1969. It is brief, sardonic, and constructed as a set of instructions for how to fail consistently in psychotherapy. The joke is obvious: anyone hoping to succeed should do the opposite. More than half a century later, that inversion still produces a useful clinical ethic.

Take the patient’s problem seriously

Haley begins with one of the most reliable routes to failure: decide that the problem bringing a patient to treatment is not the important problem. Reclassify it as a mere symptom, redirect attention toward the clinician’s preferred theory, and avoid the awkward obligation to help with what is actually causing distress.

This does not mean that symptoms have no history, function, or context. They often do. Nor does it mean that the patient’s first explanation must be the final formulation. It means that depth is not an excuse for irrelevance. A sophisticated account of why someone suffers is not sufficient if it never returns to the suffering that brought the person through the door.

Our instruments change faster than our temptations.

Do not confuse a diagnosis with a plan

Haley next turns to diagnostic language: terms that sound precise, establish professional authority, and still leave the clinician with no idea what to do. Psychiatry has changed since 1969, but this problem has not disappeared. A diagnosis can organize evidence, support communication, and open access to treatment. It can also create the illusion that naming a pattern is the same as understanding a person.

The test is practical. What does the diagnosis help us notice? What decision does it inform? What would make us reconsider it? If it cannot improve formulation, treatment, or prognosis, the label may be doing more for the clinician’s confidence than for the patient’s care.

Never make the patient fit the method

Another of Haley’s rules for failure is loyalty to one method regardless of the problem. When the patient does not respond as the method predicts, declare the patient unsuitable and preserve the method intact.

Every era has its favored version of this mistake. Medication can become the answer to every problem, or the presumed cause of every problem. Psychotherapy can be treated as universally sufficient, or dismissed as merely supportive. Neuromodulation can be presented as either a machine-driven cure or an intervention to be feared because it is unfamiliar. The camps change; the structure of the error does not.

Expertise matters. So does having a coherent approach. But clinical allegiance should be provisional. A treatment earns its place by fitting the person, the problem, the evidence, and the patient’s goals—not by protecting the clinician’s professional identity.

Define change in a way that can be seen

To fail reliably, Haley advises, avoid clear goals and avoid evaluating outcomes. If no one states what treatment is meant to change, no one can ask whether it has worked.

This remains one of his sharpest observations. Psychiatry now has more rating scales, dashboards, and measurement-based tools than Haley could have imagined. That is progress, but measurement can still become ritual. A score may improve while a person remains unable to work, sleep, think, love, exercise, or make an ordinary decision without dread. Conversely, a scale can miss a change that has restored someone’s life.

Good care makes the desired change explicit and revisits it. What would improvement look like in daily life? What burden is treatment itself creating? What would tell us to continue, adjust, combine approaches, or stop? These are not administrative questions. They are the means by which treatment remains accountable to the person receiving it.

Remember that patients live somewhere

Haley warns that failure becomes easier when a therapist ignores the patient’s actual world. Focus exclusively on the interior life and avoid the family, work, school, neighborhood, finances, sleep, illness, and relationships in which symptoms are sustained—or recovery becomes possible.

Psychiatry cannot reduce people to their environments any more than it can reduce them to neurotransmitters. Yet a treatment plan that ignores the conditions of a person’s life is not more biological or more psychological. It is simply incomplete. Sometimes the most important clinical fact is not hidden. It is the impossible job, the frightened spouse, the chronic pain, the isolation, the caregiving burden, or the absence of any safe place to recover.

Duration is not evidence of depth

Haley also mocks the idea that only years of treatment can produce real change, and that improvement should be regarded with suspicion. The point is not that long-term therapy is inherently excessive. For many people, sustained treatment is valuable or necessary. The point is that duration is not, by itself, a measure of seriousness, skill, or depth.

A clinician should be able to explain why treatment continues, what it is accomplishing, and what a transition would require. Care can become an end in itself. Concern about relapse can become fear of recovery. Continuity is a virtue only when it supports autonomy rather than quietly replacing it.

A lesson for every physician

Although Haley addresses therapists, his satire belongs in medicine more broadly. Any physician can convert a patient’s lived problem into a preferred abstraction. Any specialty can mistake a classification for an explanation, repeat an intervention because it is familiar, or define success by the value that is easiest to measure. We can all protect a theory by blaming the patient when reality fails to cooperate.

The opposite posture is not therapeutic passivity. It is disciplined humility: listen closely, formulate clearly, act deliberately, and remain willing to be corrected by what happens next. The clinician is not outside the system being observed. Our assumptions, language, timing, and relationship to uncertainty can help treatment work—and can help it fail.

New tools, old failure modes

Haley’s argument may be even more relevant in a period of rapidly expanding psychiatric technology. Biomarkers, digital phenotyping, artificial intelligence, pharmacogenomics, and increasingly precise neuromodulation may all improve care. They can also give old mistakes a modern interface.

A more precise tool does not decide which outcome matters. An algorithm does not know whether its recommendation fits the person sitting in front of us. A sophisticated treatment can still fail if it answers a question the patient did not ask, ignores the world in which the patient lives, or continues without evidence of meaningful benefit.

The antidote is not suspicion of innovation. It is to insist that innovation remain answerable to the same basic questions Haley’s satire exposes.

That, to me, is why the article still feels so current. It is not evidence that psychiatry has made no progress since 1969. It is evidence that the failure modes Haley identified are not merely historical or technical. They are human. Every generation of clinicians inherits them anew.

Good clinical work therefore requires more than knowing what success should look like. It requires the less comfortable question Haley leaves behind: How might I be helping this treatment fail?