The preceding article, "What Is Recovery? A Conceptual Model and Explication," presents a view of mental illness that clinicians will encounter in their work and must know about. It is a way of looking at persons with mental illness—their rights and the courses their lives and their diseases take—and working with them to restore dignity and self-respect in the face of stigma. Such a view promotes "collaboration with care providers" and "consumer empowerment." That is all for the good.

The article, however, is not a scientific document. There are no data here. It is, rather, a hopeful exposition and a forceful advocacy that does not claim to be objective. No evidence is presented that providing treatment in accordance with this "recovery model," as opposed to treatment along with psychoeducation, group support, family work, and the like, will have much effect on the real course of a mental illness. Indeed, at times this approach might not merely lack therapeutic usefulness but could even interfere with treatment.

The article and the recovery model acknowledge the disease element in mental illness but then tend to minimize its role. In a medical model, human beings are caught in disease processes inflicted on them. (The word "patient" is derived from the Greek word for suffering. Patients suffer from and struggle with these processes as the disease takes over their thinking, feeling, and behavior.) The model regards a medical illness as something that happens to a person's body and mind over which he or she has no direct control and is certainly not fully responsible for. The illness produces results that are not freely chosen. Embedded in their illness, patients lose their freedom to the disease.

Myocardial infarctions, injuries, infections, metabolic disorders, and the like are also disease processes that rob patients of their freedom. In the medical model, psychiatric illnesses such as schizophrenia, depression, obsessive-compulsive disorder, bipolar disorder, panic disorder, and alcohol or other substance use disorders are regarded similarly, except that in psychotic illnesses, and other illnesses, the disease process can subvert a patient's thinking process so that the disease-driven symptoms may be rationalized and not seen as ego-alien. The patient's self has been completely taken over by the disease.

Patients with schizophrenia do not freely choose to be tormented by voices, depressed patients by depression, obsessive-compulsive patients by compulsions, and panic disorder patients by panic attacks. Nor do alcoholic or drug-addicted patients freely choose to lose control, harm their livers or other aspects of health, destroy their marriages and family ties, or lose their jobs and careers. They just cannot help it. They have lost their freedom in that regard, and the doctor or other health care provider intervenes in an attempt to help overcome the disorder and restore the freedom not to have to suffer so, and not to lose health, family, or career.

Freedom is not simply freedom from external restraint. In a deeper sense it involves freedom from the internal impositions of the disease process and its consequences. To gain freedom from internal threats, a patient needs the help of health care providers and usually turns to them. Paradoxically, "external restraints" that may be imposed by health care providers can operate in the direction of the deeper freedom.

A colleague told me that one of his staff had a patient who claimed to have guns and ammunition and said he was going to kill someone. The treating physician applied the Tarasoff principle—the duty to protect or warn—and notified the police in an effort to protect the patient from his own homicidal impulses. The police found the weapons and brought the patient to the hospital. With treatment, the patient improved and was discharged.

My colleague was concerned about how this would affect his Tarasoff action, the imposition of an external restraint. I told him that involuntary outpatient commitment might be in order here—another external restraint, but one that would ensure that the patient remained in treatment so as to continue to be free of his homicidal compulsion. How does the recovery model, with its emphasis on "empowerment" and "collaboration" pertain here?

The overwhelming majority of people with serious mental illness are not immediately dangerous to others or themselves. However, a similar problem applies, equally important and more pervasive. A delusional young woman believes we have secret conversations in which I am more agreeable than I am in real life, and she is sure we have these conversations because she hears my voice (which she hallucinates). I tell her to make a note of the times when this happens so that I can bill her at the end of the month. She laughs. She apparently has an island of sanity somewhere within her that knows this is absurd. Several days later I call that laugh to her attention. She replies that we are on television, being watched, and she must pretend it is absurd. I tell her that her laugh was spontaneous, not calculated, and she says that it would have to appear so. I say that I shall give her parents back all the money they paid if there is a television camera here. She grins and replies that I would never have made that statement if I hadn't hidden the camera so cleverly that I knew she couldn't find it.

How can we speak about "empowerment" and "collaboration" with the patient's reasonable self when it can be so subverted by the disease? Would that not really be empowering and collaborating with the disease process that imprisons the patient's self with hallucinations and delusions, rather than liberating the rationality and health of the patient within?

I recall a patient who had been addicted to cocaine and had been abstinent for two years. One summer—his wife away on business, his children away at camp, I away on the traditional August break—he went out and scored some cocaine and used it. When I came back he told me about this incident and said it was the expression of his freedom. His "prison keepers" were away and now he was free. I told him I suppose one could look at it that way. He admitted then that one could also look at it as though the compulsion to take cocaine, the disease of addiction, was always there, lurking in the wings, waiting for its moment to emerge. When the prison keepers were no longer there it emerged strongly and his self gave in and found rationalizations. What seemed a loss of freedom due to his prison keepers was now, more deeply, the loss of freedom to the disease because of the absence of prison keepers. Were not his prison keepers actually persons who were facilitating his deeper freedom?

How do we understand freedom and "empowerment" here? On which occasion was he freer, more empowered? Is freedom only the absence of external restraint? Or is it even more, at times, the result of empowerment against internal processes in such instances? And therefore with what must we "collaborate"?

The authors of the paper on the model of recovery speak of a goal of eliminating involuntary commitments. This objective sounds good until one actually thinks seriously about how mentally ill people who are decisionally incapable will fare out on the streets. The same hospitalization and medication that they object to might free patients with schizophrenia from their tormenting voices and delusions and return to them their freedom. Compelling alcoholic and drug-addicted physicians to enter impaired physicians programs with psychoeducation, group therapy and support, treatment, and monitoring might restore their deeper freedom, save their livers and their families, and keep their lives and careers from going down the drain.

Here, of course, we could run into dangers. The line of thinking in the previous paragraphs can lead and has led to terrible injustices. Life is not filled with simple solutions but with two-edged swords. Doctors and other health care providers do not always know best. Even before the Holocaust, many Nazis believed that certain groups of people had "lives not worthy to be lived." People with mental and physical illnesses and mental retardation were therefore simply murdered. In the United States, Justice Oliver Wendell Holmes favored sterilization of mentally retarded persons, saying, "Three generations of imbeciles are enough."

We might find such dangers in some of the clinical research conducted by well-intentioned people, especially research on patients who are incapable of making decisions, where the researchers hoped to find a way to help them, or at least the next generation of patients. Because such research has the potential for causing more harm than good, governments all over the world, especially in the United States, have developed strict patient protection laws, regulations, and guidelines.

True, there are dangers in going too far with parens patriae, but there are dangers in going too far in the other direction. Parens patriae has its place, as effectively discussed by Appelbaum (1) and Talbott (2).

Dr. Peyser is on the faculty and attending staff at Mt. Sinai-New York University Medical Center and St. Luke's-Roosevelt Hospital Center in New York City, where he also has a private practice.