THE EFFECTS or METRONOME FACING ON THE _ AERODYNAMIC PATTERNS .OF STUTTEREDSPEECH; , _ Thesis for the Degree 0f.M-'.A-, - ' - MICHieAN STATE UNIVERSITY V _- ' -. . . _— BRENDA MAE NAVARRE ” ‘ 1975 ' ,g-cl 'I'I’F‘a A MSU _ RETURNING MATERIALS: P1ace in book drop to LIBRARJES remove this checkout from “ your record. FINES will , be charged if book is returned after the date stamped below. ABSTRACT THE EFFECTS OF METRONOME PACING ON THE AERODYNAMIC PATTERNS OF STUTTERED SPEECH BY Brenda Mae Navarre It was the purpose of this study to investigate changes in aerodynamic events for normals and stutterers while read— ing with metronome stimulation. Subjects were five adult secondary stutterers and five normal adult speakers matched for sex. To determine the effects of metronome pacing on aerodynamic events each subject was requested to read under two conditions. The first condition involved reading with the beat of a metronome. Under the second condition the subject was requested to read without using any devices to reduce stuttering. Aerodynamic analysis was used to evaluate physio- logical changes associated with the experimental condition. Four experimental questions were asked: (1) How is peak intraoral air pressure affected by metronome stimulation? (2) What changes occur in duration as a result of metronome stimulation? (3) How does air flow rate change during metronome stimulation? (4) What qualitative differences occur when comparing metronome-induced fluency and dysfluent production of the same phoneme? The results of this investigation revealed that both Brenda Mae Navarre stutterers and normals exhibited lower peak intraoral air pressure during conditions of rhythmic stimulation. Both stutterers and normals also exhibited longer peak pressure onsets, offsets, and total durations during metronome pacing. Air flow values increased for normal speakers but decreased for stutterers during metronome pacing. Finally, qualitative inspection of fluent productions of words stuttered in the no-metronome condition indicated that pressure onset slopes were much more gradual with the metronome. The results of this investigation were interpreted in light of Wingate's [1969] "modified vocalization" hypothesis which accounts for the effects of rhythmic stimulation with references to consistent, routine and predictable changes in physiological function. Implications for further research are also presented. THE EFFECTS OF METRONOME PACING ON THE AERODYNAMIC PATTERNS OF STUTTERED SPEECH BY Brenda Mae Navarre A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTERS OF ARTS Department of Audiology and Speech Sciences 1975 Q. H7177 ACKNOWLEDGMENTS My thanks to the members of my committee, Dr. Daniel S. Beasley and Dr. Linda L. Smith who gave their time on my behalf. A very special thanks to my thesis director, Dr. John M. Hutchinson, for his ideas, effort, and encourage- ment. I would like to also extend my graditude to my friends and family who offered me much support and help when it was needed. ii Chapter I II III IV APPENDICES NUOEHP‘ REFERENCES TABLE OF CONTENTS INTRODUCTION 0 o o o o o o ‘0“, o o o o o o 0 Historical Review . . . . . . . . . . . Theoretical Accounts . . . . . . . . . Aerodynamic Analysis . . . . . . . . . Statement of Problem . . . . . . . . . METHOD 0 O O O O O O O O O I O 0 O O O O 0 Subjects . . . . . . . . . . . . . . . Speech Material . . . . . . . . . . . . Experimental Conditions . . . . . . . . Instrumentation . . . . . . . . . . . . Procedure . . . . . . . . . . . . . . . Data Analysis . . . . . . . . . . . . . RESULTS 0 O O O I O O O O O O O O C O O 0 Peak Intraoral Air Pressure . . . . . . Duration . . . . . . . . . . . . . . . Air Flow Rate . . . . . . . . . . . . . Qualitative Results . . . . . . . . . . DISCUSSION . . . . . . . . . . . . . . . . SUMMARY AND CONCLUSIONS . . . . . . . . . Implications for Further Research . . . INDIVIDUAL PROFILES OF STUTTERING SUBJECTS EXPERIMENTAL PASSAGE . . . . . . . . . . . ORDER OF EXPERIMENTAL CONDITIONS . . . . . INSTRUCTIONS TO SUBJECTS . . . . . . . . . RAW DATA . . . . . . . . . . . . . . . . . iii Page \Ole-J 10 10 10 11 12 13 15 l7 l7 19 22 25 28 36 37 40 41 42 43 44 47 Table LIST OF TABLES Page Frequency of Stuttering Type Based on Aerodynamic Patterns . , . . . . . . . . 26 Summary of Slopes (in degrees) of Intra- oral Air Pressure for all Dysfluent Phonemes During the No-Metronome Condition and the Same Phoneme During Metronome Pacing . . . . . . . . . . . . . . . . . 27 iv Figure LIST OF FIGURES Schematic array of instrumentation used for recording the aerodynamic data Summary of mean intraoral air pressure and standard deviations . . . Summary of mean onset durations and standard deviations . . . . . Summary of mean offset durations and standard deviations . . . . . Summary of mean total durations and standard deviations . . . . . Summary of air flow rates and standard deviations . . . . . . . . . . Example of quantitative results Example of qualitative results Page 14 18 20 21 23 24 29 30 Accepted by the faculty of the Department of Audiology and Speech Science, College of Communication Arts, Michigan State University, in partial fulfillment of the requirements for the Master of Arts Degr e Thesis Committee: :// J 16 4y 1 Director '. n M. inson, Ph. D. ‘Dafiiel S. Beasley, Ph.D. \vavWéwua\k~ :éummv¥R> Linda L. Smith: Ph.D. vi INTRODUCTION Historical Review Historically, it has been well dccumented that rhythmic pacing produces fluent speech among dysfluent speakers. In his historical account of the use of pacing, Van Riper [1971] noted the increased therapeutic emphasis on timing procedures in the several stammering schools which were popular during the 1800's in the United States. However, controversy over the clinical use of rhythmic stimulation arose in the early Twentieth Century and these procedures were largely discon— tinued. Undoubtedly, several reasons may be offered for this decline in popularity: (1) the growing disrepute of stammer- ing schools, (2) the rather transient effects of the metronome, and (3) an inability to explain the ameliorative effects of rhythmic pacing (Hutchinson, 1974]. Despite the disuse of these procedures from a clinical standpoint, researchers continued to investigate the effects of rhythmic stimulation on stuttered speech. One of the first empirical confirmations of the pacing effect was provided by Johnson and Rosen [1937] who attempted to ascertain whether specific changes in stutterers' speech rate would affect changes in frequency of dysfluencies. It was noted that the greatest reduction in stuttering occurred by alteration of the speech pattern in accordance with some imposed and very definite rhythm, (singing, metronome, arm-swing, sing-song, and reading in chorus). Barber [1940] added considerably more information to our understanding of the rhythm affect by documenting a variety of rhythmic stimuli associated with a reduction in stuttering (bodily activities, speech rhythms, sensory rhythms, etc.). Moreover, the salutary effects of rhythm are recognized by stutterers as seen in the question- naire data of Bloodstein [1950] concerning conditions during which stuttering is reduced or absent. In recent times, there has been a rekindling of interest in the use of pacing procedures for clinical management of dysfluency. Hutchinson [1974] attributes part of this revived interest to refutation of the first two reasons cited previously for the early decline in pacing therapy. The problem of misuse of rhythmic stimulation by disreputable therapists has been overshadowed by growing application of behavior therapy principles to stuttering therapy and a more rigorous, scientific accounting of the effects of metronome conditioning. The second problem of minimal carry-over of the pacing effects has been eliminated by the technoligcal development of a miniature behind-the—ear metronome [Meyer and Mair, 1963]. Perhaps the most complete clinical data regarding the use of pacing has been provided by Brady [1969] who designed four experiments to study the metronome effect on stuttering. The results of the first study suggested that the mere slow- ing of speech rate is not the basis of the metronome effect, since the speed of the metronome was equal to each subject's reading rate during the control condition. In his second experiment the subjects read in synchrony with the metronome and read while performing subsidiary tasks to produce a dis- traction effect. Since fluency was much higher for the metronome condition it was concluded that "distraction" is not the basis of the metronome effect. A third experiment resulted in the conclusion that auditory, tactile and visual pacing had equal effect on reducing dysfluencies. The final experiment required the subjects to read to a rhythmic beat and an arrythmic beat. Although the subjects were able to perform better with the rhythmic pace, a similar decrease in stuttering occurred for the arrythmic beat, suggesting that metronome effect may not be entirely a function of the rhyth- mical pattern. These last results are in disagreement with Fransella and Beech [1967] who observed no notable fluency changes with arrhythmic stimulation. Brady [1971] was also responsible for developing one of the first complete therapy programs for use of the miniaturized metronome. The treatment involved a behavioral analysis of the disorder of stuttering since rather strong applications of learning theory principles are required to observe the behavioral changes exhibited with metronome pacing, as well as experimental studies on the effects of metronome pacing on stuttered speech. His clinical data revealed considerable treatment success. A group of 26 severe stutterers received metronome conditioning therapy, and of the 23 who completed the program, over 90 percent showed an increase in fluency which persisted for a period of six months to three years. Despite the general alleviation of the problems of dis- reputable therapeutic use and clinical carry-over, the third concern expressed earlier, that of a reasonable theoretical explanation, remains unresolved. However, several attempts to account for the pacing phenomenon have been offered and warrant brief examination. Theoretical Accounts Perhaps the first major theoretical explanation of reduced stuttering during pacing was the distraction hypothesis. Johnson and Rosen [1937] and Barber [1940] first applied the distraction construct to explain the effects of syllable- timed speech. They suggested that a rhythmical pace draws the stutterer's attention away from anticipatory dysfluencies and reduces anxiety associated with speaking. Bloodstein [1950, 1972] further amplified the "distraction" hypothesis by suggesting that unusual stimulation absorbs the stutterer's attention thereby reducing speech anxiety and associated dysfluency. He concluded: The power of such a concept (as distraction) is that it vastly simplifies our perception of stut- tering phenomena by reducing a great number of apparently unrelated observation to a single common denominator. In so doing, it performs a characteristic function of science, which strives to find more general explanations within which to encompass existing ones as special instances [1972, p. 490]. Unfortunately distraction is a weak theoretical construct because it was not an operational definition. Accordingly, assuming distraction by observing improved fluency places the theoretician in a difficult tautology. The second explanation of the positive effects of rhythmic stimulation may be termed the auditory-perceptual deficit hypothesis. This hypothesis was derived from the observation that delayed auditory feedback produced stuttering behaviors in normal speakers. In view of this observation some suggested that stuttering is a result of a distorted auditory feedback system. For example, Webster and Lubker [1968] proposed that the delay in auditory feedback is due to a difference in middle ear muscle reflex patterns between stutterers and normals. The auditory-perceptual deficit hypothesis lacks internal consistency because it does not account for several well known phenomena associated with stuttering. First, the adaptation effect is not well-explained by this theory since according to it, the improved fluency can only be the func- tion of a sudden remission of the auditory-perceptual problem upon conclusion of the first reading of a passage. Second, the reasons for improved fluency during such conditions as high intensity masking noise or singing are not obvious. Finally, efforts to document the physiological substrate for an auditory-perceptual deficit have not been conclusive. The final theoretical position has been termed the "modified vocalization" hypothesis and was first stated by Wingate [1969]. He summarized this change of vocalization by stating: Imposed rhythm can thus be said to induce a simpli- fication, routinization, and predictability of "melody" all of which could be powerful factors in producing this "artificial" fluency. . . .In short, the process is organized around the production of a melody which is actualized through emphasis on vocalization [p. 679]. Wingate thus brings forth the notion that rhythmic stimula- tion results in adjustment of peripheral speech events which become controlled so as to produce fluent vocalizations. Adams, et a1. [1973] have provided further data to support the modified vocalization hypothesis. Stutterers were required to read a passage first under normal conditions and secondly seeing and reading one word per second. Great increases in fluency were found to be associated with the pacing condition. These findings suggested that the slower reading rate allowed sufficient time to coordinate respira- tion, phonation and articulation and reduce the motoric complexity involved in contextual speech. At the present time, the modified vocalization hypothesis, though unproven, still stands as the only theoretical account which may be a valid explanation of rhythmic stimulation. To date, most of the literature cited in support of this position involved relatively crude molar frequency count of stuttering behavior based upon subjective impressions [Adams and Reis, 1971; Adams and Moore, 1971; Conture, 1972]. This is a serious problem inasmuch as it is difficult to develop sound theoretical explanations about cortical function and peripheral adjustments, without data concerning physiological events associated with the changes encountered during metronome stimulation. Abbs and Netsell [1973] have corroborated this view by stating that it is "hazardous to investigate the nature of central nervous system events, muscle activity patterns, or movement of speech structure without a consider- ation of peripheral mechanics" [p. 421]. Therefore, before any credibility can be lent to a modified vocalization explanation of metronome stimulation, it is imperative that the peripheral physiological events associated with "modified vocalization" be properly investigated. Some preliminary research [Hutchinson, 1974] has docu- mented that there may be no one—to-one relationship between a perceptual judgment of stuttering and its physiological correlate. This casts further doubt on the development of theoretical systems which do not involve physiological data. Therefore, it seems clear that before further speculation regarding the effects of metronome condition are offered, some peripheral physiological data should be obtained. Aerodynamic Analysis The experimental literature concerning application of physiological investigation strategies during stuttering is relatively sparse. However, one experimental procedure has emerged as a sensitive index of the physiological concomitants of stuttering—-aerodynamic analysis. The major investigation of aerodynamic changes during the moment of stuttering was reported by Hutchinson [1974]. This study involved an analysis of intraoral air pressure and air flow rate-events operative during stuttering. One-hundred and fifty-five dysfluencies were recorded on optical psciollograms and six aerodynamically distinct stuttering patterns were described. Type I was characterized by multiple peak intraoral air pressure which released at the onset of subsequent phonemes. Perceptually this was perceived as a syllable repetition. Type II consisted of a prolonged intraoral air pressure rise time and was auditorily perceived as a prolonged phoneme. The most common pattern, Type III, was characterized by multi- ple peak intraoral air pressures without air flow or voicing and no release of constriction until termination of the block. These patterns were auditorily identified as prolonged silent blocks. Type IV was comprised of a prolonged rate of air flow associated with the onset of an attempted phoneme and an excessive peak air flow rate during the actual moment of stuttering. Type V patterns consisted of an abrupt decrement in intraoral air pressure, air flow rate and the acoustic com- ponent. Type VI was characterized by swift build-up of intraoral air pressure with sustained pressure following for a relatively long period of time. Data of this nature provide greatly increased precision in our description of stuttering and they provide more substantive evidence for explanations of the physiological commands operative during stuttering. Statement of Problem The conclusions drawn to support a modified vocaliza- tion hypothesis have in fact been based on subjective impres- sions from perceptual data. Given the availability of sensi- tive aerodynamic measures to obtain such information, it is logical that the nature of any hypothesized vocalization change accompanying rhythmic stimulation be studied using an aerodynamic analysis strategy. Accordingly, the purpose of this study was to investigate the effects of rhythmic stimulation on selected aerodynamic parameters of speech for stutterers and nonstutterers. Specifically, four experimental questions were asked: 1. How is peak intraoral air pressure affected by metronome stimulation? 2. What changes occur in duration as a result of metronome stimulation? 3. How does air flow rate change during metronome stimulation? 4. What qualitative differences occur when comparing metronome-induced fluency and dysfluent production of the same phoneme? METHOD Subjects The subjects of the present study were five adult stutterers, (four males, 1 female) and five normal speakers, (four males, 1 female). The stutterers were all classed as secondary stutterers (phoneme, word, and situation fears; anticipatory avoidance and embarrassment of stuttering) by qualified speech pathologists as determined by inspection of clinical records. The stuttering subjects had a mean age of 25 years and a history of previous therapy ranging from approximately 60 to 350 hours. Three of the experimental subjects had previous metronome pacing experiences in therapy, ranging from S to 60 hours. Individual profiles for each stuttering subject were provided in Appendix A. All subjects included in the normal group exhibited normal speech behaviors and had no history of previous dysfluency problems. The mean age for this group was 26 years of age. Speech Material Each subject read a 12l-syllab1e passage constructed such that 12 English consonants / p, t, k, b, g, f, 0, s, v, 3, 2/, representing four consonant classes (voiceless stops, voiced stops, voiceless fricatives, voiced fricatives) appeared in a syllable-initial stressed position three times. (The phoneme /d/ was only represented twice.) The context 10 11 was further constrained such that each consonant to be studied (hereafter referred to as a target consonant) was preceded by a linguistic pause, vowel, nasal, or semivowel and followed by a vowel. These contextual constraints facilitated identification of the target consonants on the oscillographic traces and reduced the coarticulatory effects of abutting consonants, which have high pressure and high volume velocity values. The passage and associated target consonants appear in Appendix B. Experimental Conditions All subjects read the lZl-syllable passage in two conditions, control and experimental. In the control con— dition (no metronome) the talkers received no rhythmic stimulation and were instructed not to use any devices to reduce their stuttering. In the experimental condition (metronome) the talkers received a beat through headphones at a rate of 60 beats per minute. They were instructed to pace their reading of the passage such that each word was synchronized with a beat of the metronome. This metronome rate was chosen from the clinical observation by Brady [1969] that most stutterers are able to successfully synchronize their speech during oral reading at 60 beats per minute. The experimental and control conditions were presented in a randomized order, to control for potential order effects. The order in which each subject was tested can be seen in Appendix C. 12 Instrumentation The instrumentation used in the present study was similar to that described by Hutchinson [1973]. A catheter (#12, French) was utilized to obtain measurements of intra- oral air pressure. The catheter was inserted through the nasal passage until it was visible in the oropharynx. The opening of the catheter was perpendicular to the eggressive air flow to prevent spuriously high air pressure readings that can occur when air flow directly impinges on the orifice of the tube [Hardy,l965]. The catheter was attached to a pressure transducer (Stathamm, 131 TC). The signal from the transducer was amplified (Accudata 113 Bridge Amplifier) and recorded on one channel of an optical oscillograph (Visicorder 1508B). Prior to the initiation of each experi- mental session, a static calibration was accomplished using a U-tube water manometer. This procedure enabled the experi- menter to establish Specific galvonometer deflections on the optical oscillograph with known input pressures. The air flow rate data were obtained by using a large tightly fitting face mask coupled to a pneumotachograph (Hewlett-Packard, custom made). The pneumotachograph houses a screen that provides a resistance to air flow. As stated by Isshiki and Ringel [1964], "the principle of measuring a flow rate is based on the fact that the pressure drop across a resistance (mesh screen), which is caused by an air stream varies linearly with flow rate." In the present 13 investigation the pressure drop was sensed by a differential pressure transducer (Statham, PM 15), amplified (Honeywell Accudata 113 Bridge Amplifier) and recorded on a second channel of the optical oscillograph. Calibration of the flow rate was accomplished using a flowrater meter (Fisher and Porter, 10A1027A) in a fashion similar to that described for intraoral air pressure. To obtain an audio signal a high quality micrOphone (Electrovoice 635 A) was placed near the end of the pneumo- tachograph. The signal was amplified (Ampex 601, tape recorder) and simultaneously recorded on a third channel of the optical oscillograph and the tape recorder. A schematic representation of the instrumental array is presented in Figure l. The metronome beat for the experimental condition was obtained from an electronic metronome (RCA Technical Series, HM 91) which was set to produce a high amplitude pulse occurring 60 beats per minute. This pulse was then directly recorded (Ampex AG 440) onto a tape to be presented via headphones. Procedure Prior to each experimental session, the subject practiced pacing his speech during oral reading until he and the experi- menter were confident that mastery of the task had been achieved. The subject was then seated in the testing room where the catheter was inserted and the face mask positioned. l4 Im4HumuHHm5q mo mHQmem _ .m musmflm < E< omm\oot..mmH ON: .8 SSH r. 1...: “32 w {\IJ. 86> m 8 a oomcflmm m 8 n mzozomkmz 920205.92 OZ 31 Inherent within the modified vocalization hypothesis is the assumption that the stutterer "does something differ- ent" while phonating which permits a more normal onset of vocalization and establishes conditions appropriate for uninterrupted phonation, when voicing is phonetically required. However, beyond this cursory assumption, Wingate provided no suggestions regarding the actual physiological changes which characterize modified vocalization. Since the original formulation of the modified vocal- ization hypothesis, a growing body of experimental evidence had accumulated in support of this position. For example, Adams and Reis [1971] provided indirect evidence by demon- strating that dysfluency is much higher when the stutterer is asked to read a passage with numerous voiced—voiceless transitions as opposed to a passage specially constructed to minimize such transitions. Also, Adams and Hutchinson [1974] established a strong inverse relationship between vocal intensity and dysfluency. These results were discussed with reference to the obvious laryngeal adjustments required for changes in vocal intensity. Finally, recent physiological studies such as those of Conture [1974] and Freeman and Ushijima [1974] have documented clear laryngeal aberrations associated with stuttering which diminish during fluency. Depsite the growing evidence supporting a modified vocaliza- tion hypothesis, no one has yet provided a logical explana- tion of the salutary effects rhythmic stimulation has on laryngeal function. 32 It may be possible to provide some insight regarding these changes in vocalization by referring to a model of laryngeal behavior advanced by Halle and Steven [1971]. This model consists of three parameters of importance to this discussion. The first is the relationship of supraglottic pressure (Psup) to subglottic pressure ( The difference Psub)° between these vocal tract pressures was assigned to symbol AP. The second parameter was the width of glottal Opening (Ws). The third parameter was the relative stiffness or slackness of the vocal folds. Halle and Stevens demonstrated that with slack folds, relatively wide ranges of Ws and AP could be obtained and vocalization would not be impeded. However, for stiff vocal folds, those respective ranges within which vocalization can occur would be markedly reduced. Moreover, severe reductions in AP prevent the vocal folds from vibrating regardless of Ws. Interestingly, some of the documented vocal tract disturbances associated with stuttering may create condi- tions which, according to the Halle and Stevens model, would prevent vocal fold vibration. Both in the present study and that of Hutchinson [1974] there were documented instances where Psup became excessive during dysfluency. This would serve to reduce AP and small values of AP minimize the chance of vocal fold vibration. In a similar vein, Konig and von Leden [1961] documented the existence of rich autonomic nerve endings in the thyroarytenoids. This would 33 make the vocal folds a logical sight for tension under con- ditions of stress. The results of Conture [1974] and Freeman and Ushijima [1974] confirmed the presence of very stiff, tense vocal folds during the moment of stuttering. This too, would reduce the chances for vocal fold vibration, particularly if the aerodynamic conditions were such that AP was quite small. In short, the stutterer may have, even during fluent speech, vocal tract conditions which are perilously near a threshold for complete cessation of vocal fold activity. Such cessations in vocal fold activity may be fundamental to the stuttering [Adams, 1972]. It may be recalled that metronome simulation was generally associated with reductions in intraoral air pres- sure. Therefore a pervasive result of such stimulation is a larger AP which maximizes the chances for vocalization, regardless of the vocal folds. The work on laryngeal muscle activity during stuttering permits further consideration of the positive effects rhythmic stimulation has on fluency. Two basic laryngeal disturbances have been suggested: (1) There is a breakdown in the reciprocity of laryngeal abductor and adductor activity, (2) There is often much greater laryngeal muscle activity during stuttering. The first condition suggests a central nervous system programming asynchrony and the second a rather uncontrolled or unchecked surge of physiological effort manifested in increased laryngeal muscle action 34 potential. Perhaps these conditions are exacerbated by the stutterer's attempts to achieve very rapid physiological adjustments. For example, the angle values of the pressure onsets reported in this study suggest that the physiological events of stuttering may involve extremely rapid adjustments and high amplitude responses. If thisvis accurate, the metronome apparently alleviates this problem by producing very gradual, low amplitude pressure onsets. In short, the additional time afforded during the onset of an initial consonant during rhythmic stimulation may permit the stutterer to properly synchronize motoric commands and more successfully monitor the amplitude of the physiological events, particularly those involving laryngeal function. As mentioned in the Introduction of this paper, reduced reading rates have consistently been proven effective in reducing stuttering. Recently, Adams et a1. [1973] pro- vided one explanation for this phenomenon. They suggested that reading individual words at the rate of one per second obviated the need for rapid transitional movements across word boundaries: Reading in this manner encourages the cessation of speech movements after the production of every word. Therefore, coordination and transition were required only as the stutterers moved from sound to sound within a word. Coordination and transition across word boundaries were not needed [p. 674]. Certainly, if this assumption is true, it supports the simplification aspect of Wingate's modified vocalization 35 hypothesis and suggests that paced speech creates conditions more conducive to central nervous system coordination of the vocal tract events. However, Adams et al. [1973] based their conclusion on the results of reading tasks not involving a metronome signal. In the present study it was demonstrated that a metronome signal not only slows the overall reading rate but increases the duration of phonetic elements in each utterance. Therefore, the stutterer has not only the time afforded by an overall reduction in reading rate but also the prolonged time period of the initial phoneme to monitor, control, and adjust the vocal tract events thereby reducing the possibility of dysfluency. SUMMARY AND CONCLUS I ONS It was the purpose of this study to investigate changes in aerodynamic events for normals and stutterers while read- ing with metronome stimulation. SubjeCts were five adult secondary stutterers and five normal adult speakers matched for sex. To determine the effects of metronome pacing on physiological events each subject was requested to read under two conditions. The first condition involved reading with the beat of a metronome. Under the second condition the subject was requested to read without using any devices to reduce stuttering. Aerodynamic analysis was used to evaluate physiological changes. Four experimental questions were asked: (1) How is peak intraoral air pressure affected by metronome stimulation? (2) What changes occur in duration as a result of metronome stimulation? (3) How does air flow rate change during metronome stimulation? (4) What qualita- tive differences occur when comparing metronome-induced fluency and dysfluent production of the same phoneme? The salient results of this investigation may be summarized as follows: 1. Both stutterers and normals exhibited lower peak intraoral air pressures during conditions of rhythmic stimulation. 2. Both stutterers and normals exhibited longer peak 36 37 pressure onsets, offsets, and total durations during metronome pacing. 3. Air flow values increased for normal speakers but decreased for stutterers during metronome pacing. 4. Qualitative inspection of fluent productions of words stuttered in the no—metronome condition indicated that pressure onset slopes were much more gradual with the metronome. The results of this investigation support Wingate's [1969] "modified vocalization" hypothesis, inasmuch as con- sistant, routine and predictable Changes in physiological function occurred with rhythmic stimulation. In addition the results of this study were interpreted with reference to a model of vocal tract functioning presented by Halle and Stevens [1971]. Specifically, the lower peak pressures associated with metronome stimulation would, according to this model, facilitate vocal fold vibration. Finally, it was suggested that increased consonant durations noted during rhythmic pacing may permit the stutterer more time to coor- dinate properly the physiological events necessary for fluent speech. Implications for Further Research A considerable body of literature has documented that rhythmic stimulation increased fluency. Distraction and auditory perceptual deficits have lent little insight into the effects of metronome pacing. The results of the present 38 study support the modified vocalization theory since the data revealed consistent physiological changes occurring when speech was paired with rhythmic stimulation. However, two important improvements should be incorporated into a study of this nature. First, a larger number of subjects would verify the distinctive patterns of physiological changes that occurred with metronome stimulation. Secondly, controlling for the number of previous metronome pacing therapy hours each subject had received, would allow for a more accurate determination of both the qualitative and quantitative changes occurring. The findings of this study provide some additional impetus for further research. Given the ability to measure physiological occurances associated with metronome pacing, in a theoretical sense, it would prove interesting to observe differences resulting from emotional states and speaking situations. That is, how do a stutterer's physiological changes vary during a relaxed state versus an anxiety situ- ation. It might be possible to establish a hierarchy of situational fears by examining systematic differences in physiological behavior as a function of anxiety arousal. Another implication for research might involve a study of the effect different reading rates (determined by a set metronome beat) have on aerodynamic events. A similar study designed to observe physiological changes occurring when a speaker paces his reading without a metronome 39 could also prove interesting. Finally, a study such as the present could be extended to evaluate the physiological changes associated with fluency during delayed auditory feedback and masking, conditions also subsumed under the modified vocalization hypothesis. All of the above could easily be applied to children, adding considerable informa— tion to the minimal literature available concerning physio- logical events during stuttering and therapeutic techniques to control stuttering for younger clients. APPENDICES APPENDIX A INDIVIDUAL PROFILES OF STUTTERING SUBJECTS Stuttering severity was determined using Johnson, Darley and Spriestersbach's rating scale of stuttering. Subject 1. Subject 2. Subject 3. Subject 4. Subject 5. Stuttering was reported to have occurred since the age of 3. Stuttering severity was rated moderate to severe--stuttering on about 8 to 12 percent of words; disfluencies average about 2 seconds in duration; a few distracting sounds and facial grimaces; a few distracting associated movements. Subject awareness of stuttering was reported to have occurred at age 9. Stuttering severity was rated mild to moderate--stuttering on about 2 to 5 percent of words; tension noticeable but not very distracting; most disfluencies do not last longer than a full second; patterns of disfluency mostly simple; no distracting associated movements. Subject reported noticing disfluencies at age 10. Stuttering severity was rated moderate to severe-- stuttering on about 8 to 12 percent of words; consistently noticeable tension; disfluencies average about 2 seconds in duration; a few dis- tracting sounds and facial grimaces; a few dis- tracting associated movements. Subject reported having always thought of himself as a stutterer. Stuttering severity was rated severe--stuttering on about 12 to 25 percent of words; conspicuous tension; disfluencies average 3 to 4 seconds in duration; conspicuous distrac- ting sounds and facial grimace; conspicuous dis- tracting associated movement. Subject awareness of stuttering was reported to have occurred at age 6. Stuttering severity was rated moderate to severe--stuttering on about 8 to 12 percent of words; consistently noticeable tension; disfluencies average about 2 seconds in duration; a few distracting sounds and facial grimaces; a few distracting associated movements. 40 APPENDIX B PASSAGE READ BY SUBJECTS goday Tom thought ahout a tishing trip. The possibility of hatching geveral heautiful hass was a tery tantalizing tuggestion. helmos's Tackle Shop will put up a valuable gash reward for the biggest one taught. In nineteen heventy- four, a thirty pound tish won and a tery good friend of Tom's got the prize. hinally Tom couldn't regist the thought and headed for the dock. When he got there he delved into his tackle box for his favorite lure—-the testy minnow. 41 APPENDIX C RANDOMIZED ORDER OF EXPERIMENTAL CONDITION FOR NORMALS AND STUTTERERS Subject lst Condition 2nd Condition 1 Metronome No Metronome 2 Metronome No Metronome 3 No Metronome Metronome 4 Metronome No Metronome 5 No Metronome Metronome 42 APPENDIX D INSTRUCTIONS TO SUBJECTS Condition without Metronome: During this portion of the experiment, you will be requested to read a passage which will be placed before you. Read the passage without any timing techniques, such as tapping your foot, fingers, etc. or any other control devices you use to help reduce stuttering. At my signal, place your mouth tight against the face mask and we will begin. Do you have any questions? Condition with Metronome: During this portion of the experiment, you will be requested to read a passage placed before you. Through the headphones you will hear a beat. Listen carefully to the beat and then begin reading one word to each beat. If you get off the beat, let a couple of beats go by and then continue reading with proper synchrony. When you are ready, place your mouth tight against the face mask and read the passage. Do you have any questions? 43 Summary of mean intraoral air pressures (in H 0). All entries in the difference column t on in pressure in the metronome condition. APPENDIX E centimeters of represent reduc- Control Metronome Difference ------- Centimeters of H20--------- Stuttering Group Voiceless Fricatives 5.76 4.91 -O.69 Voiced Fricatives 4.07 3.53 -0.54 Voiceless Stops 6.46 5.76 -0.7 Voiced Stops 5.11 4.75 -0.36 Normal Group Voiceless Fricatives 7.3 6.23 -l.07 Voiced Fricatives 4.75 4.72 - .03 Voiceless Stops 7.7 6.39 -l.3l Voiced StOps 6.57 5.9 -0.67 44 45 Summary of the mean durations (in milliseconds) for onset, offset and total. All entries in the difference column represent longer durations in the metronome condition. Onset Control Metronome Difference ----------- Millisecond------------- Stuttering Group Voiceless Fricatives 112.5 - 250.45 +l37.95 Voiced Fricatives 99.0 181.66 + 82.56 Voiceless Stops 138.6 281.66 +143.06 Voiced Stops 112.22 248.97 +136.75 Normal Group Voiceless Fricatives 106.67 257.78 +151.11 Voiced Fricatives 86.19 203.11 +116.92 Voiceless Stops 102.67 283.11 +180.44 Voiced Stops 87.0 224.0 +137.0 Offset Stuttering Group Voiceless Fricatives 89.5 112.22 + 22.72 Voiced Fricatives 74.0 98.63 + 24.63 Voiceless Stops 94.14 105.22 + 11.08 Voiced Stops 72.22 87.69 + 15.47 Normal Group Voiceless Fricatives 82.66 96.88 + 14.22 Voiced Fricatives 66.36 91.36 + 25.0 Voiceless Stops 80.44 104.89 + 24.45 Voiced Stops 57.0 77.5 + 20.5 Total Stuttering Group Voiceless Fricatives 229.33 363.9 +134.57 Voiced Fricative 180.0 280.93 +100.93 Voiceless Stops 223.33 384.66 +161.33 Voiced Stops 182.22 336.15 +151.93 Normal Group Voiceless Fricatives 189.33 348.88 +159.55 Voiced Fricatives 167.72 295.56 +127.84 Voiceless Stops 183.11 371.56 +188.45 Voiced StOps 144.0 321.5 +l77.5 46 Summary of the mean air flow rates (in cc/second) for the four consonant classes. Control Metronome Difference ------------- cc/second------------- Stuttering Group Voiceless Fricatives 350.13 320.7 -29.41 Voiced Fricatives 216.29 209.95 - 6.34 Voiceless StOps 648.9 614.39 -34.51 Voiced Stops 232.9 213.19 —19.71 Normal Group Voiceless Fricatives 385.51 436.67 +51.l6 Voiced Fricatives 270.61 321.59 +50.98 Voiceless Stops 823.41 983.75 +160.34 Voiced Stops 310.15 399.68 +89.53 REFERENCES REFERENCES Abbs, J. H. and Netsell, R. An interpretation of jaw acceleration during speech as a muscle forcing function. J of Speech Hearing Res, 16, 421-425 (1973). Adams, M. R. Motor determinants of stuttering. Paper pre- sented at Ann. Conv. of the American Speech and Hearing Assoc., San Francisco (1972). and Hotckiss, J. C. Some reactions and responses of stutterers to a miniaturized metronome conditioning therapy. Beh. Ther. 4, 565-569 (1973). and Hutchinson, J. M. The effects of three levels of auditory masking on selected vocal characteristics and the frequency of disfluency of adult stutterers, J of Speech Hearing Res., 17, 682-688 (1974). ; Lewis, J. I. and Besozzi, T. E. The effect of reduced reading rate on stuttering frequency, J Speech Hearing Res., 16, 671-675 (1973). and Moore, W. The effects of auditory masking on the anxiety level, frequency of dysfluency and selected vocal characteristics of stutterers, J Speech Hearing Res., 14, 639-644 (1971). and Reis, R. The influence of the onset of phona- tion on the frequency of stuttering, J Speech Hearing Res., 14, 639-644 (1971). Barber, Virginia. Studies on the psychology of stuttering, J Speech and Disorders, 5, 29-42 (1940). Bloodstein, O. Hypothetical conditions under which stutter- ing is reduced or absent, J Speech Hearing Disorders, 15, 142-153 (1950). . The anticipatory struggle hypothesis: Implica- tions of research on the variability of stuttering, J Speech Hearing Res., 15, 487-500 (1972). Brady, J. P. Studies on the metronome effect on stuttering, Behav. Res. Ther., 7, 197-204 (1969). 47 48 . Metronome-conditioned speech retraining for stuttering, Behav. Ther., 2, 129-150 (1971). Conture, E. G. The effects of noise upon the speaking behavior of stutterers. Doctoral dissertation, University of Iowa (1972). : McCall, G.N. and Brewer, D. W. Laryngeal activity during the moment of stuttering: Some preliminary observations. Paper presented at Ann. Conv. of the American Speech and Hearing Assoc., November (1974). Fransella, F. and Beech, H. R. An experimental analysis of the effect of rhythm on stutterers, Behav. Res. and Ther., 3, 195-201 (1965). Freeman, F. J. and Ushijima, T. Laryngeal activity accompany- ing the moment of stuttering: A preliminary report of EMG investigations. A paper presented at the 87th meeting of the Acoustical Society of America, New York, April (1974). Halle, M. and Stevens, K. A note on laryngeal features, Haskins Laboratories Q., P. R., 101, 198-213 (1971). Hutchinson, J. M. Aerodynamic patterns of stuttered speech. Paper presented at Ann. Conv. of the American Speech and Hearing Assoc., October (1973). . Metronome-conditioned speech retraining for fluency problems. MSHA Journal, Vol. 10, 2 (1974). . Aerodynamic patterns of stuttered speech. In Webster, L. Michal and First, Lois C. (eds.) Vocal Tract Dynamics and Dysfluengy. New York: Speech and Hearing Institute, (1974). Hardy, J. C. Air flow and air pressure studied; in communi- cative problems in cleft palate, ASHA Reports, No. 5 (1965). Isshiki, N. and Ringel, R. Air flow during the productions of selected consontants, J Speech Hearing Res., 233-244 (1964). Jamison, D. Spontaneous recovery of the stuttering response is a function of the time following adaptation. In Johnson, W. (ed.) Stuttering in Children and Adults. Minneapolis: University of Minnesota Press, (1955). Johnson, W. and Rosen, L. Studies on the psychology of stuttering, J Speech Disorders, 2, 405-409 (1937). 49 Konig, Werner F., and von Leden, Hans. The peripheral nervous system of the human larynx, Archives of Otolaryngology, 74, 455-55 (1961). Malecot, A. The force of articulation of American stops and fricatives as a function of position,Phonetica,18, 95-102 (1968). Meyer, V. and Mair, J. M. A new technique to control stammering: A preliminary report, Behav. Res Ther. 1, 251-254 (1963). Prosek, R. A. and House, A. S. Intraoral air pressure as a feedback cue in consonant production, J Speech Hearing Res., 18, 133-147 (1975). Subteliny, J. P.; Worth, J. H. and Sakuda, M. Intraoral pressure and rate of flow during speech, J Speech Hearing Res., 498-515 (1966). Schwartz, M. F. The core of the stuttering block, J Speech Hearing Disorders, 39, 169-177 (1974). Van Riper, C. The Nature of Stuttering. New Jersey: _Prentice Hall, Inc., (1971). Warren, D. W. and Wood, M. T. Respiratory volumes in normal speech: A possible reason for intraoral pressure differences among voiced and voiceless consonants, J Accoustical Society of America, 466-469 (1969). Webster, R. L. and Lubker, B. B. Masking of auditory feed- back in stutterers' speech, J Speech Hearing Res., 11, 221-222 (1968). Wingate, M. E. Sound and pattern in artificial fluency, J Speech Hearing Res., 12, 677-686 (1969). ,,,,, 'flxi‘HIGAN STATF UNIvFRSITv LIBRM’HES lll ll) 1| l 7* H 3174 4364 l ‘ lull o 1!‘ I 31293 k