Initially, I matched patients according to age and sex. But I found no differences when I paired dissimilar individuals. The same held true for closet versus overt stutterers as well as for both type (I - IV) and severity of stuttering.

I then began to experiment with larger groups. First three people, then six, twelve, ..... Again, no difference could be seen for the variable of number of participants, except, of course, as the number grew larger, the amount of time necessary to achieve the desired therapeutic goals lengthened. I finally settled on fifteen patients as the maximum I would treat at once, and the length of each treatment session increased from one hour to three hours and then to an entire day. I found that the group structure coupled with a full day gave each patient ample opportunity to participate during the session, and maintained just enough stress that speaking was a challenge but not an impossibility. Patients learned from one another's mistakes and each time a patient spoke it was, by definition, public speaking.

After much trial and error, here is the treatment format that we found to be most effective, and that we now use in our workshops at The National Center for Stuttering.

We start by teaching patients to breathe passively from their mouths. I ask them to imagine themself with a cold, with their noses stuffed and able only to breathe from the mouth. I reinforce this by saing that they carry with them at all times the perfect model of the breath. It is the calm breathing they engage in when they are sitting quietly, doing nothing - except now it's through the mouth. I continue to talk to them and watch their mouths, chests, and abdominal walls for signs of passive outflows of air.

Once passive air flows from the mouth are established we move on to the use of passive airflows before monoysyllables and then multisyllables. At this point we introduce several feedback techniques that have been developed for monitoring Air Flow. By enabling patients to hear their own breathing patterns, these devices help them recognize the particular sound quality of a totally passive air flow. This quality of sound is called Flutter, and differs from person to person since everyone's respiratory system is shaped differently. Every patient must learn to recognize his own characteristic Flutter pattern. For some this takes no time at all; for others it takes more practice -- but whatever the time involved, no progress can be made in Air Flow Treatment without first attaining a consistent Flutter.

The simplest, and probably the most effective feedback technique involves a piece of rubber tubing about a foot long. One end of the tubing is placed directly in front of the speaker's lips and the other placed in his ear. Thus the patient, as he breathes, can listen to his breath. Another feedback device commonly used is a tape recorder. A special microphone is employed for all patients, a microphone capable of picking up minute Air Flows from the mouth and recording them. The microphone is placed in front of the mouth and the Air Flow is tape-recorded and played back for evaluation. The patient must learn to recognize his Flutter on the tape recorder.

The most important point about Flutter is that its presence indicates a passive air flow. When the air flow is no longer passive, flutter vanishes and is replaced by one of two classes of breathing sounds: pushed or squeezed flows. A pushed flow indicates that the patient is now actively aiding the outflow of air while a squeezed air flow is one that is produced when the vocal cords are already locked and the air is being forced between them. During Workshops, pushed and squeezed flows are recorded on the patient's recorder and played back for training purposes. Pushed and squeezed flows invariably lead to stuttering and are to be avoided at all costs.

Proper awareness of Flutter is crucial since fluency will always be present in the absence of stress. So if a patient is practicing at home alone, he will be comfortable and relaxed and probably completely fluent; at the office, on the other hand, he may have pressures that increase his stress level so that his stuttering returns. Thus the patient needs some form of external objective indication of the correctness of practice. Flutter provides just such an indication.

In Workshops Flutter is demonstrated through the use of the tape recorder and each of the participants is trained in methods of producing absolutely passive outflows. Much time is spent practicing this extremely basic and critical phase of the program.

When Flutter before single words is produced in a consistent fashion, we proceed to the production of Flutter before short phrases and then before short sentences. Two rules are taught to deal with the problem of the quick start. The first rule applies to the condition in which a phrase or sentence begins with a one-syllable word. The rule states: "When a sentence or phrase begins with a one-syllable word, we are to put a comma, a mental pause, between the first word and the rest of the sentence." The second rule relates to the situation wherein a phrase or sentence begins with a multisyllable word. This rule states: "When the first word is a multisyllable, we must say each syllable with equal slowing, much as if it were spoken to the rhythm of a slow me-tro-nome."

All patients are given special lists of phrases and sentences to practice. All practice is performed using the rubber tube to monitor for the presence or absence of Flutter and the slowed first word. In addition, periodic samples are recorded on a cassette and played back for evaluation.

During the Workshop, the technique is continually likened to a sport. The sport consists of two strokes: a passive outflow of air followed by a slowed first word. The presence of flutter and a slowed first word are the signs that the sport is being played correctly.

Since public speaking is usually described as painfully difficult, I developed a routine, in the first morning of treatment, of having each patient stand in front of his fellows at the end of the session and give a short speech. A few hours earlier, the thought of such an activity would have been an impossibility; now they were standing and speaking perfectly - without a trace of a stutter.

As each person's turn comes to speak, the others are required to subvocally practice with the patient. In this way, practice is fairly continuous. In addition, I continuously scan the room observing these subvocal practices to make certain that they are done correctly. If I see a misuse of air flow or a failure to slow the first word, I note it publicly and forcefully, thereby stressing the extreme importance of powerfully attending to technique.

I developed a one minute exercise called Contract to deal with the problem of attending to technique. After patients demonstrate both an understanding and ability to produce a passive air flow together with a slowed first word, they are required to recite a string of unrelated sentences out loud for one minute in front of an audience. Unrelated sentences are chosen initially because related ones would form a context which might distract the patient's attention away from his fledgling technique. Each sentence must be perfect, and if the speaker happens to stutter, he is required to pay a dollar for every block. My typical comment in announcing this is, "Now that you have shown that you can control your stuttering you must pay for the privilege of inflicting your struggle behavior on the world around you." Initially, contract is done for one minute a day. Later the duration is increased.

We have found that patients' attention to technique while under contract is outstanding. The slightest tendency for the mind to wander is effectively cancelled by the knowledge that if they stutter they have to pay up. All monies collected during Contract were given usually to the youngest individual at the Workshop who has been elected to buy me lunch. Alas, I often can't even raise enough money for desert.

I recall treating a young man from Houston whose father was a fabled Texas Oil Man. This nineteen-year-old received a monthly allowance of $3500. He came to the Workshop in his Turbo Porsche and when I proposed that he would have to pay a dollar a stutter, his response was, "Dr. Schwartz, that ain't no money!" To which I replied, "Roger, for you it's fifty dollars a stutter!" To which he replied, "More like it sir!".

Patients are required to practice Contract with someone at home for several months. If the Contract is done with a close friend or spouse, the money, instead of given to the individual, is simply to be thrown out the window. The thought of doing this clearly has the potential of upsetting all parties and increases the impulse for careful attention to technique. Patients learn and practice Contract at the Workshop and many have found it helpful in dealing with stressful situations long after the workshop is over.

I recall one patient calling the Center to tell me that he had gone on an interview for a position and as the interviewer began to ask the first question, the idea crossed the patient's mind that the situation was like a sort of Contract and that the instant he realized that, he went into what he called, "Contract Mode". I asked him what he meant by that and he described it as a psychological space he was in when under Contract - a space associated with a powerfully focused attention to technique. In this space, nothing could "throw" him. Needless to say, in so far as speech was concerned, the interview went perfectly.

Another exercise that has proved extremely effective is called Toughening. It is designed to make the stutterer resistant to the speed of the speech of those around him. There is a tendency for people to respond in kind. That is, if one is spoken to quickly, the tendency is to respond quickly. If the stutterer attempts to respond quickly, he will scarcely leave time for implementation of his technique. Time is required to let a small bit of air flow out passively from the mouth and to slow the first word. Toughening teaches the stutterer to take this time.

Like Contract, Toughening requires another person. The other person asks a question, which the patient then answers in a single sentence, employing the air flow technique. In the middle of the sentence, the assistant interrupts with a second question. The patient has to stop in mid-answer, generate another airflow, and, employing a complete sentence, respond - whereupon in mid-sentence he is again interrupted. This goes on for a minute. The patient tends to speed, discarding his technique in responding to the rapid-fire questioning. The goal is to retain the Passive Air Flow and continue to slow the first word regardless of the speed of the question.

I frequently tell patients at workshops that, in a sense, I wish they would all develop a peculiar form of paranoia. I wish they would believe that everyone in the world was being paid by me to toughen them. This would put them on their guard and make them highly resistant to the verbal speed demands of the world around them.