A theory should explain the observable facts and relationships, but should also enable us to derive testable predictions. The most important prediction that should be derivable from a theory on the cause of a defect is, naturally, how to repair the defect; thus the most important prediction a causal theory of stuttering should allow is how to treat it most effectively. If the theory presented in Chapter 2 is true, then persistent developmental stuttering should be treatable best by a change of attention allocation during speech. which can be achieved in the following way (no rank order):
1. Desensitization: The speaker’s attention should not be burdened with thinking on stuttering, with fear of stuttering, with the attempt to avoid or weaken it deliberately, or with concern over the listeners’ response to stuttering symptoms. Therefore, an important step is to accept one’s own stuttering, and to accept oneself as a stutterer, i.e., as a person who has a tendency to stutter, and who will possibly have this tendency throughout life.
2. Spontaneous speech: The speaker’s attention should not be burdened with excessive speech planning. That means, first, no internal word substitutions or reformulations in order to avoid feared initial sounds. Second, particularly for children: incremental sentence planning, step by step, with pauses at appropriate positions. Pausing is helpful also for adults if long sentences are needed to produce. Third, for parents and siblings: no time pressure, no pressure to speak perfectly, and, whenever possible, a relaxed atmosphere allowing the child to talk spontaneously and without fear.
3. Attention to auditory feedback: A sufficient portion of attention should be directed to the auditory feedback of speech. The speaker should consciously listen to his/her own voice and words, particularly to the ends of words and to short, unstressed function words – only listen, but not search for errors or disfluencies. Electronic speech aids using altered auditory feedback can help to practice listening to one’s own speech (see Section 3.1) and should be applied with this aim. The other and better way is to speak with a powerful, sonorous voice – which can besides help to enhance the speaker’s self-confidence – and to develop the habit of listening to the sound of one’s own voice or of consciously feeling the vocal fold oscillation during speech. Eventually, as Prestes et al. (2016) proposed, treatment “may include acoustically controlled auditory training, aiming for an improvement in hearing that may lead to an improvement in speech fluency.”
4. Attention to breathing: Adults and particularly children who frequently stutter immediately at the onset of sentences or clauses, that is, after inhalation, should consciously perceiving the moment when inhalation is complete so that exhalation and phonation can start. They should not direct attention to speech, before inhalation is complete, and should take care to speak with exhalation only (better to start phonation shortly after exhalation has started; see Air Flow Technique). However, it’s only about attention; it is not necessary to change breathing, except in cases of significantly incorrect breathing.
5. Repair techniques: If speaking has been blocked within an utterance, repeat the word preceding the blocked word, and listen to the repetition attentively, particularly to its end. Keep listening to your voice and continue speaking. Do not use the repetition to build momentum – better to make a short break after the repetition. If a block has occurred immediately at utterance onset: Breathe in shortly and not too deeply, start breathing out, and only at this moment – better a little later than too early – start speaking and listen to your voice.
Briefly said, the speaker’s attention should be directed not too much to speech planning, but sufficiently to auditory feedback and, if stuttering frequently occurs after inhalation, to the perception of breath. Attention should not be directed to a deliberate avoidance of stuttering in any way, also not to the volitional control of articulation or breathing.
I wanted to formulate the above points very clear, even if some successful and meritorious therapists will be angry with me. I do not want to cheapen somebody’s work and effort. But first, the above points result from my theory and from my own experience as a stutterer – but the theory may be wrong, and my experience may only be relevant for myself. Second, I do not consider that current therapies are ineffective, by contrast, the benefit of several treatment programs was ascertained in some studies (e.g., Euler et al., 2009, 2014; Ingham et al., 2001, 2015). The question, however, is: Why do these therapies work? That is, not least, a matter of theory – namely of the theories behind the therapies. Even if a therapy is effective, the validity of the appertaining theory does not follow from this fact.
For example, in the Kassel Stuttering Therapy (the most advanced fluency-shaping program in Germany), prolonged speech and gentle voice onsets are central in treatment. However, Neumann et al. (2003) examined clients of the Kassel Stuttering Therapy with fMRI before and after treatment, and found significantly greater activations in many auditory cortex areas during overt reading after therapy compared to before therapy. Even in a follow-up examination two years after therapy, the superior temporal gyrus (BA22) was found to be significantly greater activated bilaterally during overt reading in the former clients, as a group, than before therapy – and this, although no effect on auditory behavior or auditory processing is intended with the Kassel Stuttering Therapy. Similarly, greater activations on BA22 bilaterally were found during a monologue in a group of clients who had completed the first phase of the Modified Phonation Interval (MPI) Therapy (Ingham et al., 2003). Likewise, no alteration of auditory behavior or auditory processing is intended in this treatment approach. Hence the question arises: Is it actually the prolonged speech, gentle voice onsets, or the avoidance of short phonated intervals respectively, that works in those therapies – or is it an alteration in auditory perception, perhaps an improved processing of auditory feedback?
It is quite natural that the application of a fluency-shaping technique influences auditory attention during speech and, by that, the processing of auditory feedback. On the one hand, prolonged speech or particularly the prolongation of short phonated intervals enhances the temporal percentage of vocalized speech sounds (vowels and voiced consonants), which might intensify the perception of one’s own voice – either by itself or because it sounds unfamiliar and, by that, draws the speaker’s attention more to the auditory channel, especially at short, unstressed, tendentially uncared syllables (remember that, after the present theory, feedback disruptions often occur at "unimportant" words prior to the main content words, with stuttering resulting at the onset of these content words). Likewise, gentle voice onsets may draw the speaker’s attention to the auditory channel, since they sound unfamiliar (effect of altered auditory feedback, see Section 3.1) and to the self-perception of breathing, since exhalation at word onset must be deliberately controlled. On the other hand, the proper performance of prolonged speech and gentle voice onset in everyday talking must be monitored via auditory feedback, which makes the speaker listen to his/her voice (read more).
Furthermore, gentle voice onsets require to consciously control exhalation, which draws a portion of the speaker’s attention to breathing with the effect that feedback deficits are avoided in this field. The same might be true for all breathing techniques applied in the treatment of stuttering: It may be not so much the alteration of breathing, that reduces stuttering – many people breathe in a less-than-ideal manner and do not stutter – but more a better perception of breathing.
In sum: It may rather be a better allocation of the speaker’s attention required by a speech- or breathing technique than the technique itself , that enhances speech fluency in some current therapy programs.
What I propose for the treatment of stuttering – speaking with a sonorous voice, listening to one’s own speech, and making pauses – is not new. I can only report about the German history here: As early as in the second half of the 19th century, the in thous days famous speech therapist Rudolf Denhardt recommended, among others, speaking in a sonorous voice with emphases on vowels. In the 30s of the 20th century, Oskar Hausdörfer who ran a speech therapy institute in Breslau, taught his clients they should “sound” (i.e., speak with a sonorous voice, with emphases on vowels similar as in singing) and “listen” while speaking, and they should not try to control their “mouth movements” by the will (Hausdörfer, 1933). In the 60s, the Briton Ronald Muirden came to a similar approach: Stutterers should speak with a sonorous voice (with resonance) and in short phrases (short utterance units) with pauses between them (Muirden, 1971; published in German in 1983). Denhardt, Hausdörfer and Muirden had stuttered themselves and derived their therapeutic approaches from their own experiences. Peter Reitzes propagated “pausing” in the U.S. (Reitzes, 2006).
Hausdörfer’s and Muirden’s advices have been the core of a self-help approach called Natural Method in the German-speaking countries for 35 years. The McGuire Programme, a British self-help program, applies costal breathing as the main method, but speaking with a powerful voice and pausing also seem to play an important role in the program, as can be seen in this video. These few examples may suffice to show that the acceptance of the proposed theory would not lead to a revolution in stuttering therapy – many treatment methods have proven themselves in practice for decades – but, perhaps, the philosophy behind the therapeutic approaches may change: The aim would no longer be to gain control over one’s own speech movements, but to give up conscious control and to behave in a way that allows the automatic speech control to work without disturbance.
As already mentioned in Section 1.1, Kalveram (1983) discovered that a DAF of about 40ms (below the threshold of conscious perception) results in a prolongation of long stressed syllables. He referred to the effect as “audio-phonatory coupling” – a feedback-based online control of vowel duration in long stressed syllables, i.e., in syllables that are stressed by speaking them a little longer (see also Kalveram & Jäncke, 1989; Kalveram & Natke, 1998). Online control means that the time of vowel offset depends on the auditory feedback of vowel onset. The online control of short syllables is impossible because of the time the signal needs; vowel onset is fed back too late for controlling vowel offset in short syllables. Therefore, short syllables can only be feedforward-controlled.
Since audio-phonatory coupling works in stressed syllables only, it can hardly play a role in the causation of stuttering: If stuttering was the result of a failing audio-phonatory coupling in stressed syllables, so that the start of the always subsequent syllable was interfered, then stuttering should typically occur on unstressed syllables – but the contrary is the case (Brown, 1938; Hahn, 1942).
However, audio-phonatory coupling may be the cause of the fluency-enhancing effect of prolonged speech, including the avoidance of short phonated intervals in the MPI therapy: If short phonated syllables are avoided, then speaking is completely online-controlled by audio-phonatory coupling, that is, the termination of the current syllable and the start of the next syllable are automatically controlled on the basis of auditory feedback (not to confuse with the self-monitoring for error detection that is an offline control and interferes only if a mismatch is detected).
Taken together, prolonged speech may lead to fluency in two ways: by drawing the speaker’s attention to the auditory channel (avoiding feedback disruptions in the monitoring system especially on short, unstressed, tendentially uncared syllables) and by audio-phonatory coupling (feedback-based online control of articulation). It makes the client speak fluently in a short time and helps him to develop self-confidence. However, speaking on the basis of online feedback control is not normal: Only in the period of learning, sensorimotor sequences are mainly feedback-controlled (see Section 1.1). Therefore, in order to avoid relapse after therapy when speech again becomes more spontaneous and automatic, the client, together with prolonging short syllables, should develop a habit of listening to his/her voice while speaking.