About the presenter: Woody Starkweather received his Ph.D. from Southern Illinois University in 1970. He is currently a Professor of Communication Sciences and the author of over 50 articles and chapters, and 8 books on stuttering. He is also Co-Director, with Janet Givens, of the Birch Tree Foundation, immediate Past President of the Internation Fluency Association, and a member of the Steering Committee of ASHA's Special Interest Division on Fluency Disorders. His most recent book, Stuttering, was selected as one of the "Best Academic Books of 1997," the first time a speech pathology book has been selected for this honor.

Relapse: A Misnomer?

by C. Woodruff Starkweather
Pennsylvania, USA

Any clinician who has worked extensively with adult stutterers has encountered the problem we call relapse, the tendency for the stutterer to begin to stutter again after treatment has helped the person learn ways to talk with little or no stuttering. Only preschool children seem immune from this tendency to revert to previously learned patterns of stuttering. What happens when the painstakingly achieved gains of therapy are eroded by trickles of old behavior or washed away altogether by a flood of feeling?

I spent my first ten postdoctoral years performing nonavoidance therapy, the next five performing fluency shaping, then the next ten performing an integrated version of the two. In all of these years, I found myself frequently battling the tendency for clients to slip back into old patterns. The tendency for relapse was always strongest and the reversals most dramatic in fluency-shaping, but the tendency was also present, although in a milder and more gradual form, in nonavoidance therapy. Finally, with the use of what Gordon Blood called the "dental" model, requiring periodic checkups, it seemed that I had at least gotten a handle on relapse. But the "dental" model didn't really deal with relapse; it simply recognizes that relapse is part of the problem.

In 1995 I began training in experiential therapy and realized that I had been using elements of it for many years. With this new approach, relapse seems to have stopped being a problem. In fact, now the opposite seems to occur. Clients treated with experiential therapy are given tools that help them continue to recover during their day-to-day experiences with stuttering, and not surprisingly they seem to continue to improve during vacations or other long periods away from treatment. But of course, this therapy is new, and there may not have been enough time for relapse to occur. Although I doubt it, they may all start to stutter again some day.

Looking back on the way I used to do therapy, and contrasting it with the experiential therapy I use now, I think that what we have always called relapse isn't relapse at all but an entirely different, and simpler, phenomenon. The term "relapse," according to the dictionary, suggests that the person treated backslides, loses faith, or just doesn't keep up his or her end of the therapeutic deal. A more charitable reading might be that the disorder, having been removed or reduced through treatment, reasserts itself in some way. In some medical conditions, the term most commonly refers to a situation in which the patient has been treated, but the disease has not been entirely destroyed. For example, in the case of an infectious disease, when only some of the bacterial or viral agents are destroyed, those that remain can regain a foothold . In the case of cancer, the malignant cells may not have been completely destroyed or removed and can begin to grow and proliferate once again. In these medical conditions, it is the original disease that reasserts itself, sometimes in a stronger form. Most "relapses" in stuttering are not at all like this.

Is it that stuttering stubbornly reasserts itself after treatment, or is it simply that the techniques themselves were not completely useful in the first place? The fluency shaping techniques, for example, require the client to learn how to talk in an entirely new way and to continue to talk in this new way in all situations. Some stutterers find a path to better speech in this method (although in my experience only when they supplement fluency shaping with additional techniques), but when people who have learned the fluency-shaping targets begin to stutter again, it is not because they have become lazy, nor has their stuttering reasserted itself. Instead, the person has simply become weary of the effort involved in trying to maintain a nonspontaneous, unnatural form of speaking, a way that feels false to them. They do not feel like themselves. The disorder has not reasserted itself; it was never dealt with in the first place. The technique itself is the problem. It is too burdensome to use and feels unnatural. Attempts to render it more natural-sounding may accomplish that, essentially superficial, goal, but they do not and cannot address the more profound problem that the new way of speaking does not feel right to the stutterer. This is not a relapse. This is simply the person coming to recognize that the new way of talking is as bad or worse than the old way of talking. It is simply another trick, still effortful, still false in some way. It may be a more sophisticated trick, i.e., more difficult for the listener to detect, but still a trick -- a way of talking differerently so as not stutter.

Van Riper, Sheehan, and many other authors recognized long ago that fluency shaping was stuttering in a new form, as have many clients. It has persisted nonetheless through hype and advertising and, not insignificantly because there are occasional stutterers who profit from it. Annie Glenn and Nicholas Tunbridge are the two best known examples, although both acknowledge that it was not the fluency shaping alone that made their recovery possible, but a combination of fluency-shaping with other work on the deeper issues. Still, in both these cases, and in many others, fluency shaping has been an important first step.

There are many reasons for "relapse" when the stutterer is taught a new way of talking. The stutterer may grow tired of talking with the intense concentration that the new way requires. Normal speech is free and spontaneous, as nearly all stutterers know from their own periods of normal speech. Sometimes too the targets that are so easy to achieve in the clinical environment are completely inaccessible when the situation is difficult. They work, but only when the stutterer doesn't need them. It seems right to abandon such techniques. Sometimes the person can't seem to remember to use them when engaged in meaningful conversation. It isn't difficult to use them when "practicing," but as soon as a genuine spontaneous conversation is encountered, the new way of talking seems unavailable. In this case too it seems right to abando n the technique. Why frustrate yourself trying to use it if you can't use it when you need it. Normal speech is spontaneous, nonspontaneous speech is abnormal.

None of these "relapses" are backsliding, nor do they have anything to do with the disorder returning or reasserting itself. Instead, they are the client discovering that the techniques are unsatisfactory and deciding not to use them. A decision to abandon treatment that is genuinely insufficient is not a "relapse." And using the term "relapse" to describe it heaps blame on someone who has made an intelligent and reasonable decision. It is not the stutterer's fault that the treatment was insufficient. The fault lies in the treatment.

There is also a tendency for the stuttering behaviors to return in the stuttering modification or nonavoidance therapies. The effect seems far less powerful than in the fluency shaping techniques, but it is there nonetheless. I know of two types. The first type is not unlike the examples given above for fluency-shaping. In stuttering modification therapies, the client does not have to learn to talk in a whole new way, but he does need to learn to stutter differently. So there is less effort involved than in fluency-shaping, but there is nonetheless some concentration that needs to be applied. Perhaps more important, there is a need to confront a painful reality. Typically, the client, delighting in his new found fluency, and perhaps assuming that the awful problem is completely behind him, simply stops thinking about it. He feels as though he has been cured of the malady. Why should be bother to do this or that exercise. Stuttering is behind him.

Soon small "microstutters" begin to occur and are at first ignored because they seem so inconsequential. The memory of the truly horrendous stuttering is still there to compare with these little sticky words. But the microstutters give rise to equally small avoidance behaviors, which may occur almost without thought. The use of avoidance behaviors, even though they are very small ones, increases the fear that motivates them, as Bandura (1969) pointed out many years ago, leading to slightly larger, more effortful stuttering, which in turn leads to larger avoidance behaviors, and so on, until, if not checked by the person, the full panoply of stuttering behaviors can return. This seems more like a relapse in that an element of the original disorder (avoidance) reappears, and this pattern has the same sense of insidious invasion that the regrowth of cancer cells has.

But again, it seems wrong to call this a relapse. What has in fact happened is that old fears, fears that were not removed in the first place, have prompted the stutter er to try to hide or avoid the stuttering, although in an apparently minor way, as a result of which the microstutters began. Then the person begins again to use avoidance behaviors. It is the use of the avoidance behaviors that caused the cycle of fear --> avoidance --> fear --> avoidance to recur. If the microstutters were simply accepted as current reality, or if voluntary stuttering were used to prevent the development of new fears, the microstutters might occur, but the slide into "relapse" would not.

Of course, in the case of nonavoidance therapy, or of integrated therapy, the possibility that fears or microstutters may occur can be dealt with in advance by the clinician and the stutterer can be taught a defense against this form of recurrence. Thus it is not so much a relapse as it is a failure of the clinician to prepare the client for a well known aspect of the disorder, or if this has been done, a failure of the client to heed warnings. A better solution, however, would be to deal with the fears themselves as part of therapy. So, again, this pattern seems more like a failure of the therapy than a "relapse."

Another form of recurrence, also misnamed relapse, can occur in either the fluency-shaping or the stuttering modification forms of therapy. In this form, the person reencounters a situation which has acquired the ability to evoke stuttering as a result of previous experiences. Often the original learning is very old, going back to struggles and forcing learned when the stutterer was very young. Neither the fluency-shaping nor the stuttering modification therapies can adequately prepare a person for such deeply seated emotional reactions, reactions which were, after all, acquired when the person was very young.

Suppose, for example, that a child who stutters is called upon to read aloud in front of his classmates, and the child, knowing that he stutters and frightened of the ridicule and humiliation he might receive, leaves the room, feigning illness and perhaps feeling some real nausea because of the fear. As a result, let us suppose, he succeeds in avoiding the assignment. This, and other similar patterns of avoidance used in similar situations, plus occasional situations in which he could not avoid and was indeed humiliated, leave the child with a deep fear of reading aloud in front of a group. As he grows older the need to talk in this particular way is gradually reduced, and by adulthood it is rare to have to read aloud in front of a group. As an adult, the person seeks therapy, learns a number of ways to speak with reduced stuttering, either through fluency-shaping or through stuttering modification. But at no time has he dealt with the fear of reading aloud in front of a group of people. It is unfinished business. If by chance he encounters such a situation, even well after therapy has been completed, he will be very likely to re-experience the old fears he felt as a schoolboy, and he will be very likely to stutter again. The problem here is also not one of relapse, but rather that the old unfinished business has still not been dealt with by either of the therapies. The healing has simply never taken place.

Or, to choose another example, a child whose parents believed, when they saw their child exhibiting stuttering behaviors, that the child was unable, and might always be unable, to communicate well. In my work with parents, this is the most common reaction to a child's disfluent speech. Whether the parents try to hide this reaction from the child or not, the child seems to end up with the same attitude of self-defeat -- he believes he is a "poor communicator." Stuttering does not need to detract from one's ability to communicate, as the successful careers of numerous stuttering salesmen, teachers, preachers, lawyers, and businessmen can attest. But the person has a deeply held belief to the contrary, a belief acquired in childhood, either through the verbal and nonverbal reactions of the child's parents, or in some cases, acquired on his own. Some children test this belief and find that it isn't true, that they can communicate very effectively whether they stutter or not. But sadly, many do not check it out in this way. This kind of belief, accepted without sufficient examination, is called an "introject" (Perls et al., 1994). It is another kind of "unfinished business." If not dealt with through therapy, this false belief may persist for many years, influencing the person in many harmful ways.

The solution to this is to resurrect and focus on as much unfinished business as can be found. This is not as difficult as it might appear because unfinished business makes its presence known in many different ways, and a clinician trained in experiential techniques knows how to find and finish it. But when unfinished business is not dealt with, the result may be what appears to be a relapse. But it is not appropriate to call it by that term. The disorder has not reasserted itself. Instead, the therapy was inadequate to the task.


In fluency-shaping, there is much unnecessary effort used to speak "fluently", and many clients just realize that the cure is worse than the disease. Also, the urge to avoid is not dealt with at all, and in fact it may be fertilized by the use of "targets" designed to promote speech free of stuttering, so it is not surprising that when a situation recalls previous experiences in which avoidance behavior has been performed the old stuttering pattern will occur, often dramatically. The stuttering modification, or disfluency-shaping, approaches usually deal well with behaviors acquired as a result of avoidance conditioning, but they may not deal with deeper emotional consequences of stuttering, such as shame, or unfinished business such as anger at schoolmates who teased, or a teacher who permitted teasing. When these old feelings remain present, they are likely to foster a return of stuttering behaviors. We can prevent these "relapses" by dealing effectively with these problems during therapy.


Bandura, A., The Modification of Behavior. New York: Holt, 1969.

Perls, F., Hefferline, R., and Goodman, P., Gestalt Therapy: Excitement and Growth in the Human Personality. Highland, NY: The Gestalt Journal Press, 1994.

September 23, 1998