A man sat on the roof of his house as flood waters deepened all around. He called out to God to rescue him. Looking skyward, he noticed a helicopter approaching with a ladder extending down. "Climb up," shouted the pilot. "No," said the man firmly, "I'll wait for God to save me." Suddenly a man rowed toward him in a boat. "Get in," implored the boatsmen. "No," insisted the man firmly, "I'll wait for God to save me." After the man drowned, he demanded to speak directly to God. "You saw me in peril," he said accusingly, "and didn't respond." "Didn't respond?" God replied, "I sent a helicopter, I sent a boat..."

Why do some clients not get better? People come to psychotherapy with troublesome symptoms and often clients, like the drowned man above, seem to persist in self-defeating behavior despite offers of what could be effective therapeutic help. When we study this issue, the wonder is that successful behavior change happens as often as it does.

Most maladaptive behavior patterns begin in childhood as adaptive  responses to troublesome situations. Suppose, for example, a man threatens to punish his son severely if he does not clean his room. The likelihood is that the boy will, a) clean his room, b) feel anxious and, c)  silently resent his father's intimidation. If this pattern is repeated, the boy may grow up to be a worriedly compliant man who fears authority and tries desperately to anticipate the demands of others. As an adult he will be at risk for being grimly compliant, anxious, and overly controlled in his feeling life. People like this are often at risk for acting out (e.g. binge drinking--as feelings "erupt" after prolonged over-control.)

Suppose that this person comes into psychotherapy after his binge drinking has caused social disruption. Having practiced for fifty years how to manage his anxiety by complying with authority - imagine how this man might react to his therapist who explains what  the man must do to succeed in his treatment program. This person will, in all likelihood, respond to treatment in the same way he has responded to life. He will grimly comply with every expectation his therapist has of him and will try to be the perfect patient. 

And if his therapists are fooled into believing that his abstinence is  true sobriety and his compliance is genuine conversion, the therapists will "reward" the client's behavior by declaring his progress to be outstanding. Psychotherapy will then have produced an even more effective and better practiced excessively compliant client who will be at greater risk for experiencing in the future, the same problems that originally precipitated treatment.   

A man and his brother go to a psychiatrist. The man says to the psychiatrist, "My brother has a problem - he thinks he's a dog." The brother, meanwhile, is barking and growling. The psychiatrist says, "This is indeed a serious problem." He turns to the brother and says, "Please lie down on the couch." "Oh no," says the man, "He's not allowed on the couch."

Even if a client manages to change his behavior patterns during residential treatment, he will have to face powerful, sometimes unconscious challenges to this changed behavior; challenges that will tempt the client to resume past problematic behavior. Like the man in the above story who seeks help for his brother but keeps him off the furniture, people may have mixed feelings about behavior change. This can best be illustrated by observing the complex interaction between humans and candy machines….

Routinely our interactions with candy machines are very rewarding: put the coins in, pull the lever, and out comes the candy. Yet if this routine changes, profound events occur. Imagine, for example, that you deposit your coins, pull the lever, and no candy emerges. The machine has changed its behavior. Your quite human response to the machine's "behavior change" is to pull the lever harder and faster - as long as it takes to convince you that no candy is forthcoming.

Observe, here, an important behavioral fact: previously rewarded behavior (e.g. pulling the lever on the candy machine) will initially increase in frequency - not decrease - when it is no longer rewarded. So if people around the client have found the past behavior rewarding ("I relied on the way he would always do what I needed." or "He was more fun when he drank!"), a change in that behavior will prompt them to pull harder and faster on the lever - attempting to elicit the past behavior it found so rewarding.

Thus a client will first have to change a powerful and long-standing pathological personality style and then stay unaffected by conscious or unconscious promptings from people in his post-treatment world to revert to his old ways. These two tasks are so difficult; it is truly a wondrous thing that so many clients manage to make healthy and lasting behavior changes.

And then there's the man who kept acting like a chicken, clucking all the time and doing other chicken-like things. "Why don't you just cut this out?"  asked his friend. "I can't," the man replied, "My family needs the eggs."

Stephen Montana, Ph.D. is the Director of Clinical Services at Saint Luke Institute.