About the presenter: Barbara Dahm received her Bachelors and Masters degrees in Speech Pathology from Boston University in 1968 and 1971 respectively. In 1987 she founded the Communication Therapy Institute (CTI), in Kfar Shmaryahu, Israel. CTI is a private clinic that specializes in the treatment of stuttering. Ms. Dahm is a certified member of the American Speech, Hearing and Language Association, a certified speech clinician in Israel, and a licensed Speech-Language Pathologist in New Jersey. She is the author of Generating Fluent Speech: A Comprehensive Speech Processing Approach, published in its English version in the United States in January, 1997. Ms. Dahm has presented her approach to professionals and people who stutter in England, Germany and the United States and gives speech processing therapy workshops in Europe and the USA to professionals who are interested in learning this approach.

A Speech Processing Approach for the Treatment of Stuttering

Barbara Dahm
Kfar Shmaryahu, Israel

There have been countless definitions of stuttering. Until now, most have described stuttering in terms of the type, frequency and duration of stoppages in the forward flow of speech. They have referred to repetitions and prolongations of sounds and syllables, and blocks of airflow and voicing as the "core" behaviors of stuttering (Guitar, 1998). Most therapies for the treatment of stuttering have also focused on the disruptions in the flow of speech. Their goal is usually to accept, modify or eliminate these behaviors.

Recently a new model for explaining stuttering has been proposed. This model views stuttering within the perspective of a dynamic, multi-dimensional disorder (Smith and Kelly, 1997). According to this model, the observed stuttered events are the "surface structures". The dynamic, multi-factorial model looks at the fluency breakdown in terms of a complex integration of systems at multiple levels. This means that there are factors intrinsic to stuttering that are to be found below the surface of the speech itself. If such a model describes stuttering correctly, it would follow that effective therapy would also need to go below the surface in order to change the way the underlying dynamic speech processes function.

A Dynamic Therapeutic Approach

The speech processing approach for the treatment of stuttering is a state of the art therapy approach that goes below the surface of perceived behaviors in treating stuttering. Speech processing therapy relates to stuttering as a deviation in the normal functioning of a multi-dimensional and dynamic speech production system, instead of regarding it as a disfluent event. Stuttered speech is seen as the outcome of these deviations. This assumption is supported by Watson et al, (1994), who suggest that "Stuttering can be considered as one symptom of a defect in a fluent speech generating system that is diffusely represented in the central nervous system and that includes motoric, linguistic, and very likely cognitive processing" (p. 1226). Distortions in speech processing usually lead to perceived stuttering events, which include repetitions, prolongations, and blocks. However, the person who stutters may also experience the stuttering condition even when the listener can not perceive disruptions in the flow of the speech.

The focus of speech processing therapy is neither on accepting stuttering as an inevitable circumstance, nor on striving to make the speech fluent. The ultimate goal of therapy is to enable the stuttering speaker to use the same processes that normally fluent speakers use, thereby generating the same quality of fluent speech. The production of speech becomes a more dependable, comfortable and highly automatic activity. Instead of exerting effort to control speech, the speaker allows all of the components of the speech production system to function in parallel in a highly automatic way. This is a main condition for the generation of uninterrupted fluent speech (Levelt, 1993).

A Speech Production Model

In order to describe more clearly what the goals of speech processing therapy are, let us take a look at Levelt's model for the production of speech (1993). In his model the processes involved in the production of speech include a conceptualizer, a formulator and an articulator. The conceptualizer conceives an idea and selects, orders and keeps track of what was said before and so on. The formulator translates the preverbal message into a linguistic structure. This process, sometimes referred to as linguistic processing, involves developing a grammatical and phonetic plan. The end result of this activity is internal speech or a speech plan. The articulator refers to the motor execution of the speech plan by the processes of respiration, phonation and articulation.

The product of this process is overt speech. Levelt argues that conceptualizing requires the speaker's constant attention. However, both the formulation and motor processes are not subject to executive control. The speed and complexity of these processes demand that they be carried out automatically and without conscious awareness.

Deviations in an Interactive System

Having considered how the speech production system is meant to function, let us now take a look at how this system functions for the stuttering speaker. I want to state here that I am in no way attempting to explain the root cause or causes of stuttering i. e. the reason why a person begins to use speech processes that differ from those of normally fluent speakers. What I am describing are observations that were made in the clinical setting. It is the consistency of results when clients do or do not achieve specific therapy goals that gives strength to the importance of these clinical observations and is the foundation upon which speech processing therapy rests.

The speech processing approach does not relate to the various processes as separately functioning entities. It considers that all of the processes interact as speech is being produced. Therefore, if a speaker carries out any one of the processes incorrectly, it could affect the way the other processes function. These may, in turn, feedback to cause further errors in the way the originally distorted process is carried out or may affect the functioning of still another process. Since I have stated that we do not know the root cause of stuttering, we cannot attempt to identify which process is the one originally responsible for the overall defect in speech production. This model considers only the interactive nature of the speech production processes.

When people who stutter produce speech they often focus on the formulation of the internal speech. Instead of editing their ideas on the conceptual level, they consciously focus on the language formulation process and attempt to edit what they want to say and how they want to say it at this level. The way in which the formulation process is distorted may not be the same for all people who stutter. Clients have reported various approaches to linguistic processing. It also appears that the degree and type of formulation distortions vary from moment to moment within and among people who stutter.

For some stuttering speakers, there is a sequencing distortion. They report that they prepare the words that they want to say and then attempt to get them out all at one time. Their way of processing speech is to alternate between planning words and trying to get them out, instead of formulating a speech plan while speech is simultaneously being produced.

Other speakers use an avoidance mechanism. They are busily trying to identify the upcoming difficult words and sounds so that they can be avoided. This requires an inordinate amount of concentration and is contrary to the automatic development of a speech plan.

For most people who stutter there is also a pragmatic mechanism which interferes with the subconscious formulation of language. This means that they take conscious control over this process more often when the speaker is trying to present words to an active listener. This would explain why people who stutter usually find it easier to speak fluently when they are alone or speaking to completely non-judgmental listeners such as young children and animals.

The formulation of a speech plan is an internal process that occurs in the brain. Brain imaging studies are beginning to show anomalous brain function in adults who stutterer that is not related to cortical or sub-cortical structural anomalies. In addition the anomalous brain functions have been shown to involve cortical regions associated with motor and language processing (Watson & Freeman, 1997). Support for the theory that these differences exist also comes from the subjective reports of people who stutter and people who do not stutter. In fact, our clients report that there is a definite difference in the way they formulate and relate to the internal speech before and after therapy. Furthermore they say that the way they develop and relate to internal speech when producing speech before therapy is not the same as when talking silently to oneself. When the goals of therapy are achieved, the process is the same when thinking or talking to oneself and when talking to others.

There is a direct relationship between the linguistic process and the motor process used by the speaker. When the speaker concentrates on internal speech, there is a tendency to try to control articulation. Rather than allowing the articulators to function on a completely automatic mode, the speaker exerts some degree of control on the movement of the articulators in an attempt to form specific words or sounds. A mental link between the articulators and the phonetic plan is created.

We also see that there as an inverse relationship between the production of voicing and the movement of the articulators in the speech production of people who stutter. Fluent speakers produce voicing as the basic element in overt speech production. Once the voice is produced, the task of the articulators is to shape the voice. For people who stutter the major focus is on shaping the articulators. The production of voicing becomes a secondary function. Due to the connection between the musculature of the articulators and the larynx, certain types of stimulation to the articulators result in an inability to produce vocal fold vibrations.

Sometimes the speaker tries to set the vocal folds into vibration by pushing increased air pressure through the glottis. This may result in the temporary production of sound. However, voicing produced in this way does not allow the voicing to be continued as is necessary for fluent speech.

Normal voicing consists of consecutive acoustic segments called syllables. In the speech processing approach they are called "vocal" syllables to emphasize that they are produced by the voice regardless of the speech sounds that are formed while the syllable is being produced. The production of vocal syllables is the essential activity in speech production. As Perkins (1996) states, "(Vocal) syllables can occur without articulation, but audible speech cannot be produced without syllables.The reason is because voice is the sole component of the syllable; without a vocal air stream for energy, articulatory movements would produce no acoustic result." (p. 45). The task of the speaker is then to activate the muscles required to produce the voicing. At the same time the articulators function automatically in co-ordination with the subconscious development of a speech plan.

While it has been shown time after time that the adoption of the speech processes which are the focus of speech processing therapy most definitely do result in effortlessly produced fluent speech, establishing and adopting these processes requires changes that go beyond simply learning motor responses. Concepts, perceptions, attitudes, and thoughts all have an affect on both the motor and linguistic processes used by the speaker. It would be impossible to change the way one produces speech if change were not also made on the cognitive and affective levels. For this reason, throughout therapy, it is necessary to make changes in these areas at the same time that changes are being made in the processes used to produce speech. People who participate in this therapy approach work at developing a rational and clear understanding of how the speech production system is meant to function and how they unknowingly interfere with this system. They also work at developing a positive outlook regarding speaking, fluency disruptions, and their self-image as speakers.

Therapeutic Procedures

The principles of speech processing therapy are suitable for all people who stutter of any age. However, the framework of therapy and the instructional procedures will need to be adapted to suit the age, abilities and needs of the clients. The short-term goals of speech processing therapy include:

Note that none of these goals includes fluent speech. Fluency is a natural by-product of the achievement of the therapeutic goals. Other by-products include being able to produce speech with ease and comfort and greater self-confidence in one's ability to speak.

The short-term goals are established through a series of exercises, discussions and activities. This stage of therapy should be carried out as intensively as possible. For older children and adults the framework of a three-week intensive therapy program is highly recommended. However, when such a program is not possible, the same principles may be used in other time frames. In any case, until the client comprehends the goals of therapy and demonstrates the ability to use the desired speech processes, therapy sessions should be as frequent as possible.

The long-term therapy goals include:

After completing a three-week intensive therapy program clients will need to continue a self-management program. It is recommended that such a program be supplemented with follow-up therapy sessions in either an individual or group framework. After they are able to produce speech correctly and with ease, clients who have participated in less-intensive therapy programs will be able to reduce the frequency and length of therapy sessions depending on their individual needs. Therapy is terminated when the client feels confident and positive about being able to communicate effectively in everyday speaking situations.

Results of Therapy

Speech processing therapy does not cure stuttering. People who stutter appear to have a tendency to use processes for the production of speech that result in stuttering. However, speech processing therapy helps clients to learn how to counterbalance this tendency. Some clients develop a spontaneous habit of producing speech in the correct way at all times. Their speech is essentially fluent. Others incorporate this new way of producing speech some of the time. Their speech fluency is greatly improved.

Since the goal of therapy is to enable a dynamic system to function effectively, results of therapy can not be viewed in terms of complete success or failure. It is my belief that no client can experience failure when participating in speech processing therapy. However, there are certainly various degrees of success and various time frames for achieving the desired degree of success. Clients who have participated in this approach express the same general feeling. The goals of speech processing therapy are logical, achievable and effective for the treatment of stuttering. To quote a recent client, "Speech processing therapy has shown me the way. It has transformed speaking from a troublesome task into a pleasurable challenge."


Guitar, B., (1998) Stuttering: An integrated approach to its nature and treatment. Baltimore, Md: Williams & Wilkins

Dahm, B., (1997) Generating fluent speech: A comprehensive speech processing approach. Eau Claire, WI: Thinking Publications.

Levelt, W. (1989). Speaking: From intention to articulation. Cambridge, Mass: M.I.T. Press.

Perkins, W.H. (1996). Stuttering and science, California: Singular Publishing Group.

Smith, A. and Kelly, E. 1997). Stuttering: A dynamic multifactorial model. In Curlee, R. and Siegel, G. (Ed.) Nature and treatment of stuttering: new directions, 2nd ed., (p.204-217) Needham Heigts, MA: Allyn & Bacon.

Watson, B., Freeman, F., Devous, M., Chapman, S., Finitzo, T., and Pool, K. (1994). Linguistic performance and regional cerebral blood flow in persons who stutter. Journal of Speech and Hearing Research, 37, 1221-1228.

Watson, B and Freeman, F.J. 1997). Brain Imaging Contributions. In Curlee, R. and Siegel, G. (Ed.) Nature and treatment of stuttering: new directions, 2nd ed., (p.143-165) Needham Heights, MA: Allyn & Bacon.

September 29, 1998