THE ACOUSTEC AND AERODYNAMIC . cmacmzsncs 0F COMPENSATED mama . consonast ‘ ' ‘ Thesis for the Degree of M. A. MICHIGAN STATE UNIVERSITY . ; HOWARD DMD sex-mm- . 1974 - - ‘ “gala «anti—.1 am A u-tn- A' F . . -‘ _' ' " I-LV’ l _. ""1' ' . ’ ' ' :53;- .'.4' " . " ~.- ' -55 I 3‘99,’ .. .~ {Shh-f. . “a :..‘"'U” .,. .. .e-r l'u . 9(_ .' 9‘ -.._; than ...,: :4 d I \ 1 "'"fl-. 6.. ’fl'lh— " " WWIXWTFV'IE‘Z. S's”: QIJBmJJJJ MSU LIBRARIES W RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. .4lfil‘ ’l‘l'll. It‘ll. I III II || '1 1': I ‘| El“ ABSTRACT THE ACOUSTIC AND AERODYNAMIC CHARACTERISTICS OF COMPENSATED ENGLISH CONSONANTS BY Howard David Schwartz The purpose of this study was to investigate the acoustic and aerodynamic characteristics of compensatory articulation. Following the directions proposed in the literature for compensatory articulation, two subjects participated in two experimental conditions. Each subject read a list of 84 VCV combinations that included all plosives and fricatives appearing in three different vowel contexts (aCa, uCu, iCi). The normal production of the consonant was read five times followed by the compensated consonant. Three judges determined the best production from each group of five. Wide band spectrograms were made of the best production from each group. The acoustic variables that were examined included the closure/constric- tion duration and the duration of the VC and CV transitions. A second experiment was conducted that included simultaneous recording of intraoral air pressure, air flow rate, and Howard David Schwartz voicing as each subject read the original list of stimuli saying each VCV combination only one time. A conventional nasal catheter, pneumotachograph, and optical oscillograph system were used to record the data. Quantification of the aerodynamic patterns involved peak amplitude measures for air flow rate and intraoral air pressure as well as the duration of the pressure trace. Distinct aerodynamic patterns were identified and discussed. The results of the acoustic analysis indicated that in the case of the voiced plosive closure duration, frica- tive constriction duration, and fricative transition duration, the compensated forms of the consonant exhibited a decrease from the normal production. The results of the aerodynamic analysis indicated that a general decrease in air flow rate, intraoral air pressure, and duration of pressure was evident for the majority of compensated consonant productions. In addition a number of distinct aerodynamic patterns were evident for speaker DB. These results are discussed in relation to examining compensatory articulation in various speech disorderd populations. Additional discussion is related to the articulator variability for different phoneme classes during compensatory articulation. Accepted by the faculty of the Department of Audiology and Speech Sciences, College of Communication Arts, Michigan State University, East Lansing, Michigan. August 7, 1974 W J Daniel S. Beasley, Ph.D. Chairman utchinson, Ph.D. Leo V. Deal, Ph.D. THE ACOUSTIC AND AERODYNAMIC CHARACTERISTICS OF COMPENSATED ENGLISH CONSONANTS BY Howard David Schwartz A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Audiology and Speech Sciences 1974 ACKNOWLEDGMENTS I would like to offer my thanks to my three committee members, Dr. Daniel Beasley, Dr. John Hutchinson, and Dr. Leo Deal, for their able direction and assistance when it was so badly needed. Specific thanks are directed to my advisor and thesis director, Dr. Daniel Beasley, for providing me with a true learning experience that will never be forgotten. I would also like to thank Dr. John Hutchinson who provided words of encouragement in the worst of times. Finally I wish to thank my wife Reggie, who endured all the aggravation, grief and hostility I presented her with, and still provided encouragement and support. ii LIST OF LIST OF Chapter I. II. III. TABLE OF CONTENTS TABLES O O O O O O O O O FIGURES . . . . . . . . . INTRODUCTION . . . . . . . . Acoustic Characteristics of Speech Production Aerodynamic Characteristics of Speech Production . . . . . . . . Compensated Forms of Articulation . Statement of the Problem . . . . PROCEDURES . . . . . . . . Experiment I . . . . . . . . Subjects . . . . . . . . Speech Stimuli . . . . . . Test Environment . . . . . . Method . . . . . . . . . Analysis . . . . . . . . Experiment II . . . . . . . Subjects . . . . . . . . Speech Stimuli . . . . . . Test Environment . . . . . . Method . . . . . . . . . Analysis . . . . . . . . RESULTS 0 O O C C O O O 0 Acoustic Analysis . . . . . . Plosive Closure Duration . . . Fricative Constriction Duration . Plosive Transition Duration . . Fricative Transition Duration . Summary of the Acoustic Analysis Aerodynamic Analysis . . . . . Plosive Air Flow Rate . . . . Fricative Air Flow Rate . . . iii Page vi comm H l-‘ 10 10 10 10 12 12 13 18 18 18 18 20 20 25 25 25 27 28 30 30 31 31 32 Chapter Page Plosive Intraoral Air Pressure . . . . . 33 Fricative Intraoral Air Pressure . . . . 33 Distinct Aerodynamic Patterns . . . . . 34 Intraoral Air Pressure Duration . . . . . 45 Summary of the Aerodynamic Analysis . . . 46 Integration of Acoustic and Aerodynamic Characteristics . . . . . . . . . . 46 IV. DISCUSSION . . . . . . . . . . . . 48 Compensatory Articulation and Closure/ Constriction Duration . . . . . . . . 48 Compensatory Articulation and Transition Duration . . . . . . . . . . . . 49 Compensatory Articulation and Air Flow Rate . 50 Compensatory Articulation and Intraoral Air Pressure: Peak and Duration . . . . Compensatory Articulation and Qualitative Observations . . . . . . . . . . . 52 General Discussion . . . . . . . . . . 53 . . 52 V. SUMMARY . . . . . . . . . . . . . 57 LIST OF REFERENCES . . . . . . . . . . . . 59 APPENDICES . . . . . . . . . . . . . . . 62 A. Health Questionnaire . . . . . . . . . 63 B. Score Sheet . . . . . . . . . . . . 65 C. Spectrogram Reliability . . . . . . . . 67 D. Raw Data: JH . . . . . . . . . . . 70 E. Raw Data: DB . . . . . . . . . . . 74 iv 10. ll. 12. 13. LIST OF TABLES Page The list of speech stimuli representing normal and compensated consonant productions . 11 Mean plosive closure duration (in msec) for' ,speakers JH and DB . . . . . . . . . 25 The mean plosive closure durations (in msec) and ranges across the voiced-voiceless (V-VL) distinction . . . . . . . . . 26 Mean fricative constriction durations (in msec) for speakers JH and DB . . . . . . . . 27 Mean fricative constriction durations and ranges (in msec) for both speakers JH and DB . 28 Initial and final plosive transition durations . 28 Initial and final plosive transition durations (in msec) voiced—voiceless distinctions . . 29 Initial and final fricative transition durations (in msec): voiced-voiceless distinction . . . . . . . . . . . . 30 The mean plosive air flow rates (in cc/sec) and range of values for both speakers . . . . 32 The fricative air flow rate (in cc/sec) and range of values for speakers JH and DB . . . 32 The mean plosive intraoral air pressures (in cm/HZO) and range of values for both speakers JH and DB . . . . . . . . . 34 Mean fricative intraoral air pressures (in cm/H20) and range of values for both speakers JH and DB . . . . . . . . . . . . 35 Intraoral air pressure durations of plosives and fricatives (in msec) for speakers JH and DB 0 O O O O O O O O O O O O 45 Figure la. lb. 5a. 5b. 6a. 6b. 7a. 7b. LIST OF FIGURES Schematic representation of spectrographic analysis where determination of values are made from the first formant . . . . . Schematic representation of spectrographic analysis where determination of values are made from the second formant . . . . . Instrumental array used for recording aerodynamic data . . . . . . . . Pictorial representation of speaker DB and equipment used for aerodynamic measurement Schematic representation of quantitative differences between the normal and compen- sated consonants . . . . . . . . . Schematic representation of the normal pro- duction of /ada/ by speaker DB . . . . Schematic representation of the inhalatory pattern exhibited by DB during the produc- tion of the compensated /ada/ . . . . Schematic representation of the normal production of /upu/ by DB . . . . . . Schematic representation of a second inhalatory pattern produced by speaker DB Schematic representation of the normal production of /iIi/ by speaker DB . . . Schematic representation of /iIi/ compensated for DB . . . . . . . . . . . . vi Page 15 15 21 22 36 39 39 41 41 43 43 CHAPTER I INTRODUCTION There has been increasing interest in the acoustic and physiologic characteristics of speech and language. This interest has been related to both normal and abnormal functioning of the speech mechanism. For example, it has been suggested that the training of compensatory articula- tion may be useful for a therapeutic program involving stuttering, cleft palate, or other oral structural deformi- ties (Eisenson 1958, Wells 1971, Bloomer 1971). However to date there is a paucity of research that delineates the acoustic and aerodynamic consequences of such training. The examination of these compensatory movements may reveal therapeutic strategies that can be directed toward specific speech disorders. Before an analysis of the suggested com— pensations can be initiated, it is necessary to examine the acoustic and physiological phenomena related to normal consonant production. Acoustic Characteristics of Speech Production Speech production involves the generation of more or less noisy sounds somewhere within the vocal tract and the selective modification of those sounds by the resonance characteristics of the vocal tract (Minifie 1973). The major resonances produced within the vocal tract are termed for- mants (Denes and Pinson 1963). When a stop or fricative consonant is combined with a vowel, rapid articulatory move- ments occur that result in modifications of the formant frequencies. The rapid changes in the vowel formants that occur as the speaker moves his articulators from a vowel to a consonant or a consonant to a vowel, have been termed transitions (Halle, Hughes, and Radley 1957). Minifie (1973) reported that the duration of vocalic transitions are very rapid when they are adjacent to stop sounds because the movements toward and away from stop sound articulations are very rapid. However the vocalic transitions from one vowel sound to another are relatively slow. Many researchers have examined vocalic transitions to determine whether they can serve as cues for perception of consonants. Liberman, Delattre, Gerstman, and Cooper (1956) examined the rate of vocalic transitions and determined that the "direction and extent of the second formant transition enabled listeners to distinguish speech sounds within each of the three classes, voiceless stops, voiced stops, and nasal consonants." Transitions have been examined to determine whether they can serve as cues for the identification of manner of articula- tion and place of production for consonants. Delattre, Liberman, and Cooper (1955) examined transitional cues of selected stop consonants and concluded that the second formant transition is a cue for three places of articulatory production: bilabial, alveolar, and velar. Examining the acoustic properties of stop consonants, Halle et a1. (1957) determined that transitions were influenced by factors such as adjacent vowels, the steady state portion of the vowel, and the feature described as tense-lax (aspiration and degree of plosion). It thus appears that the formant transition is an acoustic variable that is affected by changes within the vocal mechanism. The duration of consonant closure has been identified as an acoustic variable that is affected by changes within the vocal tract. Sharf (1962) has shown that the closure duration of voiceless stops are typically greater than their voiced cognates. Lisker (1957) examined stop consonants appearing in the intervocalic position and determined that duration of consonant closure served as a cue for distin- guishing voiced from voiceless consonants. Rapid changes or modifications of articulatory movements may alter and influence the closure duration of specific consonants. Examination of the consonant closure duration will help to determine the effects of modified articulatory movements. Aerodynamic Characteristics of Speech Production The rapid articulatory movements that accompany consonant production are associated with modifications in respiratory dynamics. Fricative production is characterized by the creation of a constriction within the oral cavity that results in a turbulent air flow. The production of a plosive is dependent upon the impounding of air within the vocal tract and then releasing the air in the form of an "explosion." Several physiological variables have been identified as sensitive measures of vocal tract aerodynamics. Isshiki and Ringel (1964) examined the air flow rate of a variety of English consonants. The air flow rate was defined in terms of volume velocity and was measured in cubic centimeters per second. The investigators presented four factors that influenced air flow rate: 1. the opening characteristics of the oral and velopharyngeal stricture. 2. initial pressure within the vocal cavity behind the point of vocal closure. 3. the volume of the cavity in which the pressure is built up. 4. the pulmonary air supply during the period of explosion (p. 241). The authors determined that volume velocity was related to voicing, manner of articulation, and vowel context. In an attempt to determine the effect of context on air flow rate, Emanuel and Counihan (1970) examined a variety of stop consonants in different vowel environments. The examiners concluded that differences in volume velocities occurred as a result of different vowel environments. Isshiki (1965) attempted to determine whether varying the intensity and pitch would affect the airflow rate of selected speakers. Isshiki concluded that variations did occur as a function of vocal intensity and this occurred more at high pitch levels than at low pitches. Thus the research to date suggests that the measurement of air flow rate appears to be a sensi- tive measure of changes in respiratory dynamics. One of the determinants of the rate of air flow is the air pressure that is impounded behind a stricture within the vocal tract (Isshiki and Ringel 1964, Subtelny, Worth, and Sakuda 1966). Examination of consonant production has been conducted to determine the intraoral air pressure characteristics that accompany consonant production and the variables such as phonetic context and manner of production that are associated with the modification of that pressure. Subtelny gt_a1. (1966) and Arkebauer, Hixon, and Hardy (1967) examined the intraoral air pressure associated with a variety of English consonants. The manner of production appeared to be the significant variable in determining the intraoral air pressure. The voiceless consonants tended to have greater peak amplitudes and longer durations of intra- oral air pressure than their voiced cognates. The plosive group was characterized by the highest intraoral air pressure, whereas the fricatives exhibited the longest durations. Additional variations occurred as a result of phonetic context, subject grouping by age and sex, and increased vocal intensity. Malecot (1968) examined the intraoral air pressure of selected consonants as a function of position and stressing. He concluded that consonants occurring before a stressed vowel or in the initial position demonstrated higher intraoral air pressures. In addition he concluded that the duration of pressure values increased as a function of a stressed vowel. The measurement of air flow rate and intraoral air pressure appears to reflect the modifications that occur in respiratory dynamics. It appears that both measures would be sensitive to modifications that might occur in the pro- duction of compensated speech forms. Compensated Forms of Articulation A number of investigators have suggested the idea of teaching a compensated form of articulation as a result of structural and functional deficits. Bloomer (1971) has stated that the oral mechanism is capable of many compensa- tory movements. He explained that when a speech pathologist is confronted with an oral structural deviation, it is his ". . . responsibility for determining to what extent adaptation is not merely possible but feasible" (p. 735). A number of therapeutic approaches directed toward stutter- ing advocated the teaching of compensatory movements in the form of "loose contacts." As Eisenson (1958) reports: Some therapists teach the stutterer new patterns of articulation that call for less articulatory tension, for less "sticking" on some sounds that most of us employ in our speaking. Perhaps what these therapists are helping their cases to accomplish is a manner of articulation that does not require a firm motor set (p. 263). Van Riper (1958) proposed that the stutterer should "fill much of his speech with voluntary loose movements of the tongue, jaw, and lips.‘ In addition, as part of the stutterers approach to "fluent nonabnormal stuttering," Van Riper suggested the use of loose contacts. Wells (1971) has proposed the teaching of light contacts for alveolar sounds in the remediation of cleft palate speech. Van Hattum (1974) has suggested the follow- ing approach to speech therapy with cleft palate clients: Action of the articulators should be precise and rapid as possible . . . . Loose articulatory contacts similar to those advocated in some stuttering activities may result in relaxing closure or stricture, thus reducing the amount of air forced through the nasal cavity and lowering the air need, so that there is less likelihood of excessive air use. The articulators are kept in motion, and words are easier to produce (p. 329). Shelton, Hahn, and Morris (1968) explained that a person demonstrating veolpharyngeal inadequacy cannot be expected to have the necessary air flow associated with accurate fricative and plosive production. The authors prOposed the teaching of a light, quick, constriction for the production of fricatives. It is explained that any prolongation of a fricative will increase the amount of nasal emissions. When teaching the production of plosives, the authors suggested the use of light, quick, articulatory contacts. They explained: Sustained tense contacts may cause enough breath to escape through the nasal passages that the result is an undesirable facial grimace which comes about from an attempt to prevent nasal escape of air (pp. 251-252). The preceding recommendations reveal the fact that compensatory articulation is related to the aerodynamic mechanism. Any articulatory modifications that occur will directly affect the aerodynamic characteristics of the phonemes produced. Statement of the Problem Whereas various authors have suggested the teaching of a compensated form of articulation for various speech disorders (Bloomer 1971, Van Hattum 1974, Van Riper 1958), a lack of empirical data exists to support the training of the compensatory movements. Thus it appears necessary to examine the specific variables that will reflect the acoustic and aerodynamic changes that might occur during the production of compensated articulation. Acoustic variables that will reflect these changes include the duration of the closure or constriction phase of the plosive and fricative consonants and the duration of the VC and CV formant transitions. Aerodynamic measures that will indicate changes in the respiratory dynamics include the volume velocity and the intraoral air pressure. In summary the purpose of this investigation was to examine two acoustic and three aerodynamic characteristics of a variety of compensated English consonants. The following specific questions were asked: 1. Will the proposed compensations exhibit the same consonant closure or constriction duration and the same VC and CV transition durations as a normal production of the consonant? 2. Will the proposed compensations exhibit the same air flow rate, intraoral air pressure, and duration of intraoral air pressure as a normal production of the same consonant? CHAPTER II PROCEDURES Experiment I Subjects The subjects of the present study were two adult males with a mean age of twenty-eight (28) years. Each subject was a trained speech scientist and speaker of the General American Dialect. As reported in a health question- naire (Appendix A), each subject was reported to be in good physical health with no apparent physical disorders. Speech Stimuli The speech stimuli consisted of fourteen (14) consonants (p,t,k,b,d,g,s,z,f,v,o,0,;,3) appearing in the intervocalic position of a monosyllabic unit (VCV). The vowels that preceeded and succeeded each consonant include /i,a,u/, representing the range of tongue positions. A vowel-consonant-vowel (VCV) combination was constructed where the initial and final vowel remained the same (e.g., /ipi/, /ata/). The list of stimuli (Table 1) represented eighty—four (84) VCV combinations, whereby a normal produc- tion of the consonant was followed by the compensated form of the same consonant. The order of presentation for each 10 11 TABLE l.--The list of speech stimuli representing normal and compensated consonant productions. 1. apa *22. iei -43. ugu *64. idi *2. apa 23. udu *44. ugu 65. ata 3. ibi *24. udu 45. aza *66. ata *4. ibi 25. non *46. aza 67. ugu 5. aba *26. u0u 47. ufu *68. usu *6 aba 27. aIa *48. ufu 69. ivi 7. iti *28. afa 49. ifi *70. ivi *8. iti 29. iki *50. ifi 71. izi 9. uzu *30. iki 51. isi *72. izi *10. uzu 31. qu *52. isi 73. ipi 11. uku *32. qu 53. upu *74. ipi *12. uku 33. utu *54. upu 75. afa 13. ada *34. utu 55. igi *76. afa *14. ada 35. aba *56. igi 77. aka 15. usu *36. aba 57. ibi *78. aka *16. usu 37. ubu *58. ibi 79. uvu 17. ubu *38. ubu 59. isi *80. uvu *18. ubu 39. asa *60. isi 81. aOa l9. aga *40. asa 61. asa *82. aoa *20. aga 41. iIi *62. asa 83. ava 21. ioi *42. iIi 63. idi *84. ava *Compensated production. 12 normal compensated pair was determined by a random selection procedure. Test Environment The recording of the wxxmtn: signal was accomplished using a high quality reel to reel tape recorder (Ampex 64403) and a stationary microphone (Electrovoice 635A) located within a double walled prefabricated booth (I.A.C. series 1600). Each subject was able to monitor the intensity of each production by observing a VU meter (Calectro DI-930) located within the booth. External monitoring of the acoustic signal was facilitated by amplifying the acoustic signal (Ampex 440B) through an external speaker (Ampex AA620). Method The subject was seated within an I.A.C. booth, ten centimeters from a stationary microphone. Each subject was presented with a list of eighty—four (84) VCV combinations (normal/compensated) listed in a random order. Verbal instructions for the production of the compensated articula- tions were constructed according to the recommendations of previous investigators (Shelton et_al. 1968, Wells 1971, Van Hattum 1974). The instructions were as follows: 1. For the plosive compensations you should approximate the place of production of the normal phoneme, but you should have a lighter lip or tongue contact and a more rapid closure. l3 2. For the production of the fricative you should attempt a more rapid constriction with less frication. 3. For the production of alveolar sounds you should attempt to make a type of flapping movement with your tongue to and away from the place of production. 4. For the production of velar sounds you should attempt to have a less firm attack and more aspiration. Each subject was instructed to say the normal VCV combina- tion and compensated VCV combination five times. The subject was also asked to speak in a monotone with equal stress on each syllable. The speaker was advised to monitor his productions on a VU meter to attempt to maintain consistency of productions. Two judges, the author, a non-participating subject, and the speaker himself determined the acceptability of each production. This determination was made on a score sheet (Appendix B) where each production was rated on a three-point scale, representing good, better and best production. Final acceptance of one production was determined by the highest average score of the five productions. Analysis Wide band spectrograms were made of each production to provide the means for measuring the acoustic variables of VC and CV transition rate and the constriction or Closure duration of each consonant. 14 After evaluating the possible methods of transition measurement, it was determined that the duration of the transition would best reflect the articulatory dynamics occurring during normal and compensated consonant production. The duration of the formant transitions were measured for both the VC and CV segments. As indicated in Figure l, the initial measurement of the VC segment began at the point where rapid frequency changes began to occur in the structure of the first formant. If the first formant did not reveal a distinct point for measurement, determination of the point of change was made from the second and third formants (Figure 1). This rule was adopted for all transition duration measurements. The terminal measurement of the VC segment concluded at the downward frequency shift of the first formant and the beginning of the consonant closure. The initial measurement of the CV segment began at the point of an initial spike for a voiceless plosive or at the point of a rapid frequency shift in the first formant for the remainder of the consonants. The terminal measurement was made at the point where the frequency began to stabilize in the first formant. Vertical lines were drawn at each point of measurement to facilitate the analysis of the results. The procedures described by Dhman (1966) were used: An extrapolation rule was adopted that stated that the formant was assumed to move into the stop gap without a change in slope from the nearest observable value in the vowel (p. 152). 15 Figure la.—-Schematic representation of spectrographic analysis where determination of values are made from the first formant. The first formant exhibits a distinct transition as indicated at point a-b. The points b-c represent the closure duration of the consonant and the point c-d indicates the final transition duration. Figure lb.--Schematic representation of spectrographic analysisxdmme determination of values are made from the second formant. Because the initial formant appears straight, the second formant is examined to determine the point of transition. Point a-b represents the direction of the initial transition, point b-c closure duration, and point c-d the final transition. FREQUENCY (HZ) FREQUENCY (HZ) 16 TIME (MSEC) f--1 /AGA/ 75 MSEC 20. § 15” E3 5.. E5 <3 F———-vl TIMEUéflgEC) 75 Msec 17 The duration of the constriction or closure duration was measured by drawing a vertical line at the initiation of closure as indicated by the downward fre- quency shift of the first formant for the initial vowel. The termination of the constriction or closure duration was marked by a spike for the voiceless plosives or a rapid frequency shift in the first formant for the remainder of the consonants. A reliability check was performed on every fifth spectrogram. These measurements were performed by a Ph.D. student in Speech Science who was trained in spectrographic analysis. In addition, the student was instructed by the experimenter following the procedures previously presented. The reported reliability values for closure/constriction duration were a mean value of 14.6, standard deviation 5.99, and range 4-29. Original results reported were a mean value of 14.4, standard deviation of 6.02, and range of 5-28. The reliability measurements for transition durations were a mean value of 8.6, standard deviation of 3.42, and range of 3-22. The original measurements were a mean value of 11.4, srtandard deviation of 2.79, and range of 7-23. The table of iiiese values is presented in Appendix C. All of the duration measurements were based on the fornmflae 31.8 cm equals 2.4 seconds, or 1 mm equals 7.5 mSEC. 18 Experiment II Subjects The subjects used in this experiment were the same as those used in Experiment I. Speech Stimuli The Speech stimuli were the same as reported in Experiment I. Test Environment The procedure used was similar to those described by Hutchinson (1973). A catheter (#12 French) was utilized to obtain measurements of intraoral air pressure. The catheter was inserted through the nasal passage until it was visible in the oropharynx as determined by the experimenter. The opening of the catheter was perpendicular to the air flow to prevent spuriously high air pressure readings that can occur when the air flow directly impinges on the orifice of the tube (Hardy 1965). The catheter was attached to a pressure transducer (Statham l3lTC). The signal from the transducer was amplified (Accudata 113 Bridge Amplifier) (and.recorded on one channel of an optical oscillograph VVisicorder 1508B). Prior to the initiation of the exper-‘ iment, a static calibration was accomplished using a U-tube ‘Water manometer. The procedure enabled the experimenter to 4f the consonant. As a result, consistent measurements of 1:he air flow peaks were made following the consonant release. The duration of the intraoral air pressure was nneasured from the point where the pressure curve began to Irise from the base pressure and then terminated at the point there the pressure curve joined the vowel trace at the base- laine. These measurement procedures are similar to those rfiported by Subtelny et a1. (1966) and Malécot (1968). The quantitative data were collected for three trials for each subject. The data reported reflects the means of the three experimental trials. The qualitative analysis consisted of a search for disstinct new aerodynamic patterns that occurred as a result of’ the compensatory articulations. CHAPTER III RESULTS The results of the present investigation will be presented in two sections. The first section will deal with the acoustic analysis of normal and compensated consonant productions followed by an evaluation of the aerodynamic experimentation. A third section of this chapter will (attempt to integrate the measurements of the first two asections in an effort to examine the interrelationships 1:hat may exist between the acoustic and aerodynamic charact- eeristics as they relate to compensatory articulation. Acoustic Analysis I?1osive Closure Duration The duration of the plosive closures are presented in Table 2 . TABLE 2.--Mean plosive closure duration (in msec.) for speakers JH and DB. JH DB Normal Compensated Normal Compensated aCa 108 75 aCa 79 82 uCu 102 107 uCu 84 116 iCi 120 103 iCi 76 119 Means 110 95 80 105 25 26 An examination of the means for this table indicated that the closure duration of the compensated plosives decreased for JH while for DE, the duration of closure for the com- pensated plosives increased. A further breakdown of this category as exhibited in Table 2, revealed a number of inconsistencies within the voiceless plosive class, thus minimizing the drawing of any significant trends from the results associated with this class of phonemes. TABLE 3.--The mean plosive closure durations (in msec) and ranges across the voiced-voiceless (V-VL) distinction. VL V norm range comp. range norm range comp. range aCa 116 101-131 86 56-128 99 98-101 64 23- 86 uCU 103 86-124 115 60-173 80 86-113 100 34-161 iCi 118 105-128 115 45-180 123 113-135 91 49-120 IDB aCa 103 71-101 115 53-135 71 60- 83 63 41- 94 uCa 91 71-113 165 109-218 76 68- 86 66 45-101 iCa 83 60-105 190 173-218 70 45- 9O 48 34- 68 Ehxamination of the closure duration of the normal and com- Fkansated voiced plosives revealed that eighty-three percent (533%) of these plosives decreased in duration during the CCJmpensated production of the consonant. It was only within tile voiced /uCu/ contest for speaker JH, that the compen- SElted plosive production increased in duration. 27 Fricative Constriction Duration The fricative constriction durations for the normal and compensated fricative productions are presented in Table 4. TABLE 4.--Mean fricative constriction durations (in msec) for speakers JH and DB. JH DB Normal Compensated Normal Compensated aCa 148 125 aCa 106 83 uCu 155 131 uCu 115 94 iCi 136 109 iCi 88 88 Means 146 122 103 88 IExamination of the values presented for both speakers :indicated that the compensated form of the fricative pro- ciuctions decreased or remained the same in duration within 1:he three vowel contexts. When the fricative constriction Clurations were further broken down according to the voiced- \roiceless distinction (Table 5), it can be seen that ninety- tihree percent (93%) of the compensated productions decreased ii) duration. Only one context, /iCi/ for a single speaker, Slmowed no change between the normal and compensated pro- duction. 28 TABLE 5.--Mean fricative conStriction durations and ranges (in msec) for both speakers JH and DB. VL V norm range comp. range norm range comp. range g5 aCa 160 150-173 127 105-146 135 127-150 124 86-180 uCu 183 165-210 154 109-173 126 113-135 109 53-150 iCi 188 169-203 158 130-169 84 71- 94 60 38- 83 DB aCa 144 116-128 112 94-124 68 64- 75 54 45- 75 uCu 159 190-210 132 53-154 70 45-113 55 38- 71 iCi 118 60—165 120 94-158 56 53- 75 56 34- 94 Plosive Transition Duration The initial and final plosive are shown in Table 6. transition durations STABLE 6.--Initial and final plosive transition durations. JH DB initial final initial final norm comp. norm comp. norm comp. norm comp. EiCa. 81 81 91 95 94 74 108 89 uCu 74 98 132 112 89 74 109 132 iCi 69 71 121 97 71 74 101 100 Ddean 75 83 115 101 84 74 106 107 29 Examination of the results for the normal and compensated productions revealed inconsistencies that prevented the drawing of any significant conclusions. A further analysis of the plosive transitions according to the voiced—voiceless distinction can be found in Table 7. The results presented in Table 7 further support the inconsistencies reported for the plosive transitions. Because only fifty—eight percent (58%) of the initial and final transitions increased in duration from the normal to compensated productions, it was not possible to draw any significant conclusions. TABLE 7.--Initial and final plosive transition durations (in msec) voiced-voiceless distinctions. initial final mean aCa uCu iCi aCa uCu iCi £5 VLN 78 74 69 96 181 163 110 VLC 75 91 76 85 143 116 98 VDN 85 74 70 86 83 80 80 VDC 86 104 66 105 75 78 85 22 VLN 86 65 81 114 105 115 94 VLC 79 83 71 105 138 120 96 VDN 101 113 60 103 113 88 96 VDC 70 65 76 74 125 81 81 VLN = voiceless normal VLC = voiceless compensated VDN = voiced normal VDC = voiced compensated 30 Fricative Transition Duration As viewed in Table 8, the mean compensated duration values for the fricative transitions decreased for both speakers in the voiced and voiceless categories. TABLE 8.--Initial and final fricative transition durations (in msec): voiced-voiceless distinction. initial final mean aCa uCu iCi aCa uCu iCi 9.11 VLN 90 98 98 72 102 87 91 VLC 95 72 68 90 75 79 80 VDN 89 105 83 85 96 98 93 VDC 88 99 107 82 86 83 91 213 VLN 77 63 61 88 83 109 80 VLC 72 53 59 83 91 66 71 VDN 58 89 68 72 82 95 77 VDC 53 70 55 72 68 68 65 It must be noted however that for JH a number of instances occurred where the compensated transition duration exceeded the normal production of the consonant. In addition, in one context, speaker DB exhibited an increase in the transition duration during the production of the voiceless fricatives appearing in the /u/ context. Summary of the Acoustic Analysis In summarizing the previously reported data, the fricative results appeared to be more consistent between 31‘ speakers than the plosive results. The results presented in the previous experiment indicated that in the case of voiced plosive closure duration, fricative constriction duration, the fricative transition duration, and compensated forms of the consonants exhibited a decrease from the normal production. Aerodynamic Analysis The presentation of these date will be in four sections. These sections will include: the quantitative analysis of air flow rate, the quantitative analysis of intraoral air pressure, the qualitative characteristics of air flow rate and intraoral air pressure during the com— pensatory productions, and, finally, the durational measure- ment of the intraoral air pressure. Plosive Air Flow Rate The mean air flow rates for voiced and voiceless plosives are presented in Table 9. In the voiceless plosive category, both speakers exhibited a lower air flow rate for the compensated production when compared to the normal production of the consonant. It can also be noted that the voiced plosives exhibited a decrement in air flow during the production of the compensatory consonants. Although the air flow rates obtained in this study were higher than those previously reported in the literature, the normal productions followed the same pattern as in 32 TABLE 9.--The mean plosive air flow rates (in cc/sec) and range of values for both speakers. VT. \7 norm range comp. range norm range comp. range .JH aCa 2169 1666-2446 606 168-1712 632 240- 963 192 43- 382 uCu 2192 1269-2461 904 260-1682 751 459-1177 329 183- 566 iCi 1852 596-2446 482 275- 780 557 397- 657 226 168- 795 DB aCB 2224 1980-2400 1235 431-1896 516 296- 795 516 46- 996 uCu 2041 1590-2446 887 275-1942 841 431-1269 841 108- 612 iCi 1965 1387-2416 430 135- 876 743 611- 889 743 40-1758 previous studies, whereby, the voiced plosives exhibited a lower air flow rate than their voiceless cognates. Fricative Air Flow Rate The fricative air flow data is presented in Table 10. TABLE 10.--The fricative air flow rate (in cc/sec) and range of values for speaker JH and DB. ‘VL 'V norm range comp. range norm range comp. range JH aCa 1062 612-1773 1099 168-1620 620 275- 840 380 214-1207 uCu 1158 657-1544 1271 573-1788 745 489- 948 552 76-1193 iCi 997 627-1559 1010 214-1223 611 344- 826 360 275- 519 DB aCa 1411 1037-1941 1301 81-2447 751 275-1253 631 81-2128 uCu 1490 1978-1987 837 229-1712 1020 581-1482 399 153-1100 iCi 1308 963-1636 1040 121-1437 982 672-1559 592 306-1040 33 The decrement in air flow rate that was previously reported in the compensated plosive productions is also evident in the compensated fricative productions of speaker DB. However, JH has demonstrated some variability between the production of the normal and compensated consonants. Specifically, within the voiceless class of fricatives, JH has exhibited an increase in the air flow rate during the compensated productions. In addition, although not repre- sented by this table of mean values, speaker DB also exhibited a number of increased air flow rates for compen- sated productions of /asa, aea, ava, usu, iei/. For both speakers the voiced fricative class of phonemes appeared to be more stable with a majority of the voiced fricatives compensations exhibiting a decrement in air flow rate. Plosive Intraoral Air Pressure The mean plosive intraoral air pressure data for both speakers are presented in Table 11. A decrement in the intraoral air pressure was exhibited by both speakers during the production of all voiced and voiceless productions of the compensated plosives. The normative results presented in Table 11 are similar to those reported by Malécot (1968). Fricative Intraoral Air Pressure As in the case of the plosive intraoral air pressures, the fricative intraoral air pressures as produced by both speakers exhibited a decrease in value during the production 34 TABLE 11.--The mean plosive intraoral air pressures (in cm/ H20) and range of values for both speakers JH and DB. VL V norm range comp. range norm range comp. range .JH aCa 8.9 4.2-10.1 5.1 3.5-6.6 5.6 3.2-8.0 2.6 1.2—3.4 uCu 9.1 6.8-10.2 5.5 3.0-7.5 7.1 4.0—8.5 3.2 .4—5.0 .iCi 8.8 6.8-10.3 4.7 2.0-6.1 5.9 4.5-7.0 2.7 1.0-4.4 Egg aCa 8.5 6.5— 9.8 2 4 .7-5.5 4.5 2.7-5.4 .7 .3-1.4 uCu 9.6 7.6-11.0 1.4 .3-2.6 5.4 4 1-7.2 1.3 .5—2.9 iCi 9.5 8.6-10.3 1.4 .2-2.6 5.6 2.8-6.4 1.7 .4-3.2 of the compensated productions of the consonants. Consis- tent with previous reported investigations (Subtelny et_al. 1966, Arkebauer et_al. 1967), the voiced fricatives all displayed lower intraoral air pressures when compared with their voiceless cognates. The peak fricative pressures are reported in Table 12. Distinct Aerodynamic Patterns A number of distinct aerodynamic patterns were revealed from the oscillographic traces. As presented in Figure 4, this pattern occurs in a majority of compensated productions produced by speaker JH. In this type of pro- duction, the air flow pattern of the normal production was characterized by a relatively low continuous air flow occurring during the production of the vowel (point a-b). 35 TABLE 12.--Mean fricative intraoral air pressures (in cm/H O) and range of values for both speakers JH and DB. VL V norm range comp. range norm range comp. range .JH aca 6.7 3.7- 8.4 3.1 1-5.4 4 3 1.6- 7 3 1.7 .4-3.9 ucu 7.6 4.5- 8.8 3.1 6-4.6 5 5 3.1- 7 0 2.6 1.1-3.9 101 7.4 4.1-10.0 3.6 6-4.7 5 6 3.9- 7 6 2.1 1.1-3.8 21.3. aCa 8.8 6.0-10.9 1.6 .5-2.5 6.1 2.1- 8.6 1.7 .2-8.6 ucu 9.2 7.9-10.6 1.3 .2-3.5 7.6 2.9-11.3 1.6 .8-2.6 iCi 9.6 8.2-10.6 1.2 5-2.2 7.6 5.9- 9.2 2.1 .3-5.5 Following the vowel the air flow drops off rapidly (b-c) while the intraoral air pressure is impounded and elevated (1-2). Following the release of the closure (2-3), the air flow rate increases rapidly (c-d) and then returns quickly to the baseline for the production of the final vowel. The compensated form of the consonant appears to demonstrate the same configuration as reported for the normal production although it is evident that the compensated production dis— played a marked decrease in both the air flow rate and intraoral air pressure. This aerodynamic pattern is a graphic representation of the quantitative differences previously reported for the normal and compensated produc- tion. 36 Figure 4.--Schematic representation of quantitative differences between the normal and compensated consonants. The top trace represents the acoustic envelope; the second, intraoral air pressure; and the bottom air flow rate. Baselines are drawn through the traces. (Hereafter similar figures will follow the same organization.) 37 ii i MM 2 cm .301 m 3 pressure 4 T 306 41./‘73 (r\ cc/sec air flow— C H 200 msec . t . (Bl-és 1 L pressure 2cm H201 ) i W“ T306 . cc/sec air flow V H 2OO msec 38 A number of qualitative aerodynamic patterns were displayed by speaker DB. In the compensated production displayed in Figure 5, DB has produced the consonant on an inhalatory pattern. Examination of the air flow trace reveals an increased air flow rate prior to consonant release (point a). This exhalation of air is followed by an inhalation (b), and the production of the consonant. In this type of production there was a minimal rise in intra- oral air pressure from the initiation to termination of the consonant. A second distinct qualitative aerodynamic pattern is exhibited in Figure 6. This pattern was characterized by a rapid increase in air flow rate and intraoral air pressure at the initiation of the compensated consonant. These initial peaks quickly return to the baselines. The air flow pattern reveals an inhalatory pattern that is represented by the valley occurring below the baseline. While this inhalatory production is occurring, the intraoral air pressure remains minimal for about 100 msec. until the initiation of the vowel. A third qualitative aerodynamic pattern occurred for many of the compensated fricative productions for DB. As seen in Figure 7, the normal production of many fricatives are characterized by dual air flow peaks and a large intra- oral air pressure peak. The compensated productions are characterized by an increase in the initial air flow peak (a) 39 Figure 5a.--Schematic representation of the normal production of /ada/ by speaker DB. Figure 5b.--Schematic representation of the inhalatory pattern exhibited by DB during the production of the compensated /ada/. In this production there is minimal intraoral air pressure. ada W ‘ essure 2cm H201 V pr Min T306 cc/sec air flow. ' I—' 200 msec 21931. W 2 O ~fi-e pressure H cm 2 i I I | ‘ I l A l I N [306 . cc/sec emrfkwv I 817’ I——I 200 msec 41 Figure 6a.--Schematic representation of the normal production of /upu/ by DB. Figure 6b.--Schematic representation of a second inhalatory pattern produced by speaker DB. During this production the intraoral air pressure rises and falls rapidly and remains minimal for about 100 msec. 42 upu ‘1 pressure 2cm HzOl [J U [A I\ T306 . cc/sec aIr flow , 7 _ . , HZOOmsec _ ALP. U 2cm ‘01 WT? pressure H2 II | l | 306 lhi CC/sec air flow I ' I V -I—I ZOOmsec 43 Figure 7a.--Schematic representation of the normal production of /iLL/ by speaker DB. Figure 7b.--Schematic representation of /iLL/compensated for speaker DB. This pattern is representative of many of the compensated fricative productions. 1 g 8 i Al \I'" ‘ pressure 2cm H2O I I I ' I306 - [J I,\ CC/sec airflow - _ I--I ZOO msec 1 A8 i V, ‘ I. “VII“? pressure 2cm H201 1 l AI I . l l [a ‘II I306 [W I /, cclsec air flow , J I—I ZOOmsec C 45 and a decrease in the second peak (b). In addition, while DB has increased the initial air flow peak he has produced the consonant with a decrease in intraoral air pressure (1). Following the decreased second peak, DB produced an inhalatory pattern (c) for the termination of the consonant and initiation of the vowel. Intraoral Air Pressure Duration The results of the analysis of intraoral air pressure duration for plosives and fricatives are presented in Table 13. TABLE 13.--Intraoral air pressure durations of plosives . and fricatives (in msec) for speakers JH and DB. JH DB normal compensated normal compensated Plosives aCa 160 148 190 147 uCu 179 154 208 140 iCi 167 148 210 138 means 168 150 202 142 Fricatives aCa 187 143 175 150 uCu 214 159 202 138 .iCi 216 148 203 126 means 206 150 193 126 46 The results for the normative productions are similar to those reported by Hutchinson (1973) and Malécot (1968). The results appeared to indicate that for all compensated productions, the duration of the intraoral air pressure decreased. It must be noted, however, that for a number of compensated productions in which the inhalatory aerodynamic configuration appeared, it was impossible to obtain any distinct duration measurements. Summary of the Aerodynamic Analysis Upon completion of the aerodynamic analysis, it is possible to draw a number of conclusions regarding the normal and compensated productions of plosives and frica- tives. 1. The airflow rate for compensated plosives decreased in all productions. 2. The majority of the voiced fricative compen- sations exhibited a decrement in air flow rate. 3. Intraoral air pressure decreased for fricative and plosives in all compensated productions. 4. A number of distinct aerodynamic patterns occurred for both speakers upon examination of the oscillographic aerodynamic traces. 5. Intraoral air pressure durations decreased for both speakers in all plosive and fricative compensations. Integration of Acoustic and Aerodynamic Characteristics To the extent that comparison is possible this section will attempt to bring together some relative 47 information provided by the acoustic and aerodynamic investigation of compensatory articulation. It is evident from the results obtained, that compensatory articulation, whether measured acoustically or aerodynamically, decreased in the time required to produce the consonants. In only one instance for one speaker did the duration of the compen- sated productions increase. A second point worth noting was the fact that although the fricative compensations produced the more stable results during the acoustic experimentation, they appeared to be more variable than the plosive productions during the aerodynamic analysis. This variability may be related to the specific characteristics examined for each consonant, or possibly some physiological factor was put into play during the acoustic experimentation, or finally either acoustic or aerodynamic measurement was a better tool for examining compensated consonant production. CHAPTER IV DISCUSSION The results of Experiment 1 generally supported the hypothesis that the degree of closure/constriction of the compensated consonants decreased in duration when compared with the normal production of the consonants. In addition, the data reported for the fricative transition durations further supported the contention that the compensatory articulations decreased in duration when compared with the normal production. Compensatory Articulation and Closure/Constriction Duration The majority of the closure/constriction durations decreased during the production of the compensated conso- nants. The voiceless plosives appeared to be the only class of consonants that exhibited an increase in duration of closure during the compensated productions. The data reported for the increased closure durations revealed that speaker DB produced the compensatory voiceless plosives in 165 and 190 msec in the /u/ and /i/ context respectively. The normal production of the voiceless plosives average 100 msec in duration as reported in Table 3. 48 49 However, if the voiceless fricative constriction durations are examined, it is evident that the results reported for the increased plosive durations were similar in duration to the fricative productions. This fact is supported by an examination of the aerodynamic patterns for many of the compensated plosives produced by DB. The compensated consonant is characterized by the initiation of air flow rate and consonant production prior to the release of intraoral air pressure. This pattern is typically found in the production of normal fricatives. It appears that DB has produced a modified plosive that is similar to a fricative in duration. In this case the aerodynamic measurements helped to confirm the acoustic data. The fricatives within this closure/constriction classification were the more consistent group of phonemes when the normal and compensated productions were examined. Only one context /uCu/ evidenced an increase in duration during the compensated productions. Compensatory Articulation and Transition Duration The initial and final transition durations were examined to provide an indication of the modifications that occurred in the vocal mechanism during the production of the normal and compensated consonants. The fricative data appeared to exhibit an eighty-three percent (83%) decrease in duration for initial transitions and a 50 ninety-two percent (92%) decrease in duration for final transitions when comparing the normal and compensated pro- ductions. The plosives, however, provided inconsistent results between the normal and compensated productions. Because only fifty-eight percent (58%) of the initial and final transition durations decreased during the production of the compensated consonants, no significant patterns could be drawn. As indicated in the previous section, the fricatives appeared to be the more consistent group of phonemes when comparing the normal and compensated consonant productions. Perhaps a reason for the variability that occurred within the plosive group came as a result of various attempts by both speakers to produce acceptable compensations within the framework of the instructions presented and still maintain an acceptable plosive. The results of Experiment 2 indicated that a general decrease in air flow rate, intraoral air pressure, and duration of intraoral air pressure was evident in the majority of compensatory articulations when compared with the normal production. Compensatory Articulation and Air Flow Rate When comparing the normal production of the plosives to the compensated form, it was evident that the peak air flow rate decreased in all productions of the compensated 51 plosives. The normative data reported is higher than those previously reported in the literature (Isshiki and Ringel 1964, Hutchinson 1973). However, a possible explana- tion may be the fact that both speakers have had considerable training in public address and this may account for increased physiological intensity that would result in increased volume velocity. It should be noted that the voiced pro- duction of the plosives were of smaller air flow values than the voiceless plosives, thus supporting previous investigations (Isshiki and Ringel 1964). As previously reported, a majority of the voiced fricatives appeared to exhibit a decrease in peak air flow rate. Although DB exhibited a decrease in peak air flow rate for the voiceless compensated fricatives, JH exhibited an increase of all voiceless fricative compensations. The possibility exists that the laryngeal mechanisms helps to direct and control the air flow during the production of the voiced phonemes. During the production of the voiceless fricatives, the increase in air flow rate was related to manner of production. A normal production of a fricative requires a constriction within the vocal tract, resulting in a turbulent air flow. When the compensations were produced it is possible that the rapid articulatory movements and rapid constrictions yielded a constriction that was more relaxed than the normal productionof the consonant. During this production JH attempted to increase his volume velocity 52 in order to produce the same amount of turbulent air flow to produce an acceptable consonant. Compensatory Articulation and Intraoral Air Pressure: Peak and Duration The peak intraoral air pressures reported for the plosives and fricatives examined, exhibited a decrease in value from the normal to the compensated production. In addition, the duration of the consonant pressures decreased from the normal to compensated production. This decrement in duration was reflective of both speakers following the instructions for more rapid closure/constriction durations during the production of the compensated consonants. The decrease in duration was in general agreement with the reported duration of the acoustic analysis. Compensatory Articulation and Qualitative Observations Examination of the oscillograph traces for the consonants produced, provided some indication of how the compensated productions differed from the normal productions. Although both speakers exhibited a decreased air flow rate and intraoral air pressure for a majority of the compen- sated productions, the nature of the compensatory movements appeared quite different. It appeared that JH produced many of the compensatory articulations in similar manner as the normal production. This is supported by the high frequency of occurrence of the pattern exhibited in Figure 4. 53 Speaker DB on the other hand modified his method of pro- duction as evidenced by the occurrence of the inhalatory patterns. The evidence reported points to the fact that the human vocal tract is capable of many types of articu- latory compensations that can yield decreased air flow rate and intraoral air pressure. However, the question arises as to whether or not unwanted articulatory compensations may be produced as a result of recommended articulatory postures. Although speaker DB was able to produce the necessary compensatory movements, one cannot rule out the possibility that the inhalation type production may be an unwanted compensatory movement. Perhaps a direct training procedure for a length of time may be sufficient to teach adequate methods of compensatory movements. General Discussion As previously reported, various investigators have suggested the use of light articulatory contacts and rapid lip and jaw movements as a method of decreasing the intra- oral air pressure and air flow rate during consonant pro- duction. This proposed decrement in aerodynamic variables arises from the fact that a person who exhibits velo- pharyngeal inadequacy will usually have trouble impounding the necessary oral air pressure required for plosives or have difficulty maintaining adequate air flow for the pro- duction of fricatives. By decreasing these aerodynamic 54 variables, investigators such as Shelton et_§l. (1968) and Van Hattum (1974) have indicated that adequate consonant production can be accomplished. As indicated in the present investigation, an individual has the ability to decrease the intraoral air pressure and air flow rate provided he has been instructed as to how to produce the compensatory articulations, and he is capable of relatively normal articulatory movements. These facts would indicate the need to examine the teaching of compensatory movements to individuals demonstrating velopharyngeal inadequacy. In addition the amount of articulatory training should be examined to determine the effects of intense articulatory training upon compensatory movements. As noted previously a short instruction period for the production of compensa- tory articulation may result in unwanted compensations such as exhibited by speaker DB in the inhalatory productions. Examination of the consonant data throughout the present investigation indicated that the fricatives dis- played the more consistent results during the compensated productions when compared to the plosive productions. The place of production appeared to be the significant factor affecting this variability. During the production of a compensatory plosive, the speaker is not able to vary his place of production beyond a limited range. A bilabial stop can only be produced at the juncture of the lips. Any modification of this place of production will result in a 55 distorted consonant. The fricatives, however, appeared to have displayed a wider variability of articulator placement during the compensatory articulations and they still were able to exhibit more consistent results. This hypothesis suggests the possibility of examining lip, jaw, and tongue movements during the normal and compensated articulations to examine the extent that variability is possible for both fricatives and plosives. Although it appears that light contacts will be effective as a therapeutic approach to cleft palate speech, the need exists to examine whether the perception of the compensated consonants will remain the same as the percep- tion of the normal consonant. From the data presented relating to fricative production, the possibility exists that perception of the fricative compensations will be more accurate as a result of their wider range articulatory placement. This investigation should confirm the consis- tency of the fricative results. The use of "loose contacts" have also been suggested as a therapeutic tool in the remediation of stuttering (Van Riper 1958). In a later publication Van Riper (1971) has suggested three requirements that are necessary for initiation of a stuttering tremor. These include "(1) a localized area of hypertension, (2) a postural fixation of the muscle groups involved, a posture different from that normally used for the sound, and (3) a sudden ballistic 56 movement, or surge of tension or pressure." The omission of any one of the three factors seems to prevent tremor. The present investigation has revealed that an individual is capable of reducing both intraoral air pressure and air flow rate using compensated consonants. This would lead us to believe that a stutterer could reduce the "ballistic movement" and subsequently reduce or remove the tremor behavior if he was taught a form of compensatory articula- tion. Examination of compensatory articulation in stutterers should provide the results. CHAPTER V SUMMARY The purpose of this investigation was to examine the acoustic and aerodynamic characteristics of compensated English consonants. The results provided normative data from which further investigation of compensated consonants can be initiated. It was determined from the acoustic experimentation that the voiced plosive closure duration, fricative con- striction duration, and fricative transition duration, all decreased from the normal to compensated production of the consonants. In addition it appeared that the fricative results appeared to be more consistent between speakers than the plosive results. The results of the aerodynamic analysis indicated that the compensated productions of the consonants exhibited a general decrease in value for air flow rate, intraoral air pressure, and duration of the pressure measurement. Qualitative differences were also evident between speakers upon examination of the oscillographic configurations. The effects of compensatory productions were dis- cussed and compared with the normal productions of the 57 58 consonants. Practical applications were discussed and con- sideration for future research presented. These included the need to examine compensatory articulation in those .individuals with velopharyngeal insufficiency and stutterers. The period of training was also presented as a factor to be examined. The variability of place of pro- duction was offered as an explanation for consistency within phoneme classes and finally, the perception of com- pensatory consonants was offered to document the variability of production and consistency of results. LIST OF REFERENCES 59 LIST OF REFERENCES Arkebauer, H. J., Hixon, T. J., and Hardy, J. C. Peak intraoral air pressure during speech. J. Speech Hearing Res., 10 (1967)- Bloomer, H. Speech defects associated with dental malocclu- sions and related abnormalities; in Travis, Handbook of Speech Pathology and Audiology. New York: Appleton-Century-Crofts (1971). Delattre, P. C., Liberman, A. M., and Cooper. F. S. Acoustic loci and transitional cues for consonants. J. Acoustical Society of America, 27, 765-773 (1955). Denes, P. B., and Pinson, E. N. The Speech Chain. New York: Anchor Press, Doubleday 71973). Eisenson, J. A perseverative theory of stuttering; in Eisenson, Stuttering A Symposium. New York: Harper and Row (1958). Emanuel, F. W., and Counihan, D. T. Some characteristics of oral and nasal air flow during plosive consonant production. Cleft Palate J., 7, 249-260 (1970). Halle, M., Hughes, G. W., and Radley J., P.A. Acoustic properties of stop consonants. J. Acoustical Society of America, 29, 107-116 (1957). Hardy, J. C. Air flow and air pressure studies; in communicative problems in cleft palate. ASHA reports, number 5 (1965). Hutchinson, J. M. The effect of oral sensory deprivation on stuttering behavior. Ph.D. dissertation, Purdue University (1973). Isshiki, N. Vocal intensity and air flow rate. Folia Phoniatrica, 17, 92-104 (1965). Isshiki, N., and Ringel, R. L. Air flow during the produc- tion of selected consonants. J. Speech Hearing Res., 7, 233—244 (1964). 6O 61 Liberman, Delattre, Gerstman, and Cooper. Tempo of fre- quency change as a cue for distinguishing classes of speech sounds. J. of Experimental Psychology, 27, 765-768 (1955). Lisker, L. Closure duration and the intervocal voiced- voiceless distinction in English. Language, 33, 42-49 (1957). Malécot, A. The force of articulation of American stops and fricatives as a function of position. Phonetica, 18, 95-102 (1968). Minifie, F. D. Speech acoustics; in Minifie, Hixon and Williams, Normal Aspects of Speech, Hearing, and Language. New Jersey: Prentice Hall Inc. (1973). Ohman, S. E. G. Coarticulation in VCV utterances: spectrographic measurements. J. Acoustical Society of America, 39, 151—168 (1966). Sharf, D. J. Duration of post stress intervocalic stops preceeding vowels. Language and Speech, 5, 26-30 (1962). Shelton, Hahn, and Morris. Diagnosis and therapY; in Spriestersbach and Sherman, Cleft Palate and Communication. New York: AcademIc Press (1968). Subtelny, J. D., Worth, J. H., and Sakuda, M. Intraoral air pressure and rate of flow during speech. J. Speech Hearing Res., 9, 498-518 (1966). Wells, C. Cleft Palate and Its Associated Disorders. New York: McGraw-Hill (1971). Van Hattum, R. J. Communicative therapy for problems associated with cleft palate; in Dickson, Communication Disorders. Glenview, Illinois: Scott, Foresman and Co. (1974). \kni Riper, C. Experiment in stuttering therapy; in Eisenson, Stuttering A Symposium. New York: Harper and Row (1958). Van.IRiper, C. The Nature of Stuttering. New Jersey: Prentice Hall, Inc. (1971). APPENDICES 62 APPENDIX A HEALTH QUESTIONNAI RE 63 Name Date of Birth Chronological Age Sex Have you had your hearing tested within the last year? Any known hearing loss at the present time? If yes, to what extent? Do you ever experience any difficulty when speaking? To what extent? Do you demonstrate any known articulation errors? What is the nature of your general health? Please describe any physical limitations you might have that effect your speech production. 64 APPENDIX B SCORE SHEET 65 APPENDIX B.--Score Sheet. 1. apa 123 123 123 123 123 43. ugu 123 123 123 123 123 *2. apa 123 123 123 123 123 *44. ugu 123 123 123 123 123 3. ibi 123 123 123 123 123 45. aza 123 123 123 123 123 *4. ibi 123 123 123 123 123 *46. aza 123 123 123 123 123 5. aba 123 123 123 123 123 47. ufu 123 123 123 123 123 *6. aba 123 123 123 123 123 *48. ufu 123 123 123 123 123 7. iti 123 123 123 123 123 49. ifi 123 123 123 123 123 *8. iti 123 123 123 123 123 *50. ifi 123 123 123 123 123 9. uzu 123 123 123 123 123 51. isi 123 123 123 123 123 *10. uzu 123 123 123 123 123 *52. isi 123 123 123 123 123 11. uku 123 123 123 123 123 53. upu 123 123 123 123 123 *12. uku 123 123 123 123 123 *54. upu 123 123 123 123 123 13. ada 123 123 123 123 123 55. igi 123 123 123 123 123 *14. ada 123 123 123 123 123 *56. igi 123 123 123 123 123 15. usu 123 .123 123 123 123 57. ibi 123 123 123 123 123 *16. usu 123 123 123 123 123 *58. ibi 123 123 123 123 123 17. ubu 123 123 123 123 123 59. isi 123 123 123 123 123 *18. ubu 123 123 123 123 123 *60. isi 123 123 123 123 123 19. aga 123 123 123 123 123 61. asa 123 123 123 123 123 *20. aga 123 123 123 123 123 *62. asa 123 123 123 123 123 21. iei 123 123 123 123 123 63. idi 123 123 123 123 123 *22. iei 123 123 123 123 123 *64. idi 123 123 123 123 123 23. udu 123 123 123 123 123 65. ata 123 123 123 123 123 *24. udu 123 123 123 123 123 *66. ata 123 123 123 123 123 25. ueu 123 123 123 123 123 67. usu 123 123 123 123 123 *26. non 123 123 123 123 123 *68. usu 123 123 123 123 123 27. aIa 123 123 123 123 123 69. ivi 123 123 123 123 123 *28. aIa 123 123 123 123 123 *70. ivi 123 123 123 123 123 29. iki 123 123 123 123 123 71. izi 123 123 123 123 123 *30. iki 123 123 123 123 123 *72. izi 123 123 123 123 123 31. qu 123 123 123 123 123 73. ipi 123 123 123 123 123 *32. qu 123 123 123 123 123 *74. ipi 123 123 123 123 123 33. utu 123 123 123 123 123 75.- afa 123 123 123 123 123 *34. utu 123 123 123 123 123 *76. afa 123 123 123 123 123 35. aba 123 123 123 123 123 77. aka 123 123 123 123 123 *36. aba 123 123 123 123 123 *78. aka 123 123 123 123 123 37. u0u 123 123 123 123 123 79. uvu 123 123 123 123 123 *38. ubu 123 123 123 123 123 *80. uvu 123 123 123 123 123 39. asa 123 123 123 123 123 81. aea 123 123 123 123 123 *40. asa 123 123 123 123 123 *82. aea 123 123 123 123 123 41. 1:1 123 123 123 123 123 83. ava 123 123 123 123 123 *42. iIi 123 123 123 123 123 *84. ava 123 123 123 123 123 '*==GamxxsabaiPromxxion 66 APPENDIX C SPECTROGRAM RELIABILITY 67 APPENDIX C.--Spectrogram Reliability (1 mm - 7.5 sec.) Closure Duration (in mm) Transition Duration (in mm) Initial Final JM HS JM HS JM HS JH Production afa-n 21 20 7 10 8 7 ugu-n 16 14 8 11 8 ll aba-n 14 13 7 9 7 9 aza-n 12 17 14 15 11 13 wew-n 24 22 12 14 13 15 uvu-n 20 20 7 10 6 8 utu-n 15 13 14 10 10 22 usu-n 27 28 12 13 5 14 igi-n 16 18 9 8 7 9 udu-c 4 5 10 12 8 15 aga-c 10 ll 15 12 12 16 uzu-c 17 7 12 23 10 17 ivi-c 9 6 13 12 10 12 ifi-c 21 21 8 11 8 8 izi-c 14 10 14 15 ll 11 iki-c 25 24 6 12 6 12 ata-c 10 10 8 12 13 13 usu-c 15 14 9 9 13 12 utu-c 16 16 13 13 5 18 DB ifi-c 18 17 7 5 7 11 uzu-c 7 9 7 8 3 8 usu-n 29 28 5 7 ll 15 izi-c 6 5 12 8 5 13 ivi-c 8 8 6 6 10 13 aba-n 9 ll 11 7 5 10 ueu—n 18 18 7 7 4 13 ifi-c 16 17 5 7 5 7 aza-n 8 10 8 7 10 9 asa-c 15 16 7 9 4 10 aba-n 8 9 9 9 7 9 afa-n 16 17 7 11 5 10 ata-n 9 12 10 l4 l3 l9 iki-c 22 23 7 ll 22 19 aga-c 12 13 5 9 10 13 aea-n 16 16 4 ll 4 9 uvu-c 6 10 7 9 6 9 68 69 APPENDIX C.--Continued. Closure Duration (in mm) Transition Duration (innmfi Initial Final JM HS JM HS JM HS DB (Cont.) afa-n 16 17 9 11 15 10 igi-n 10 10 4 11 5 18 aba-n ll 11 5 7 4 10 ata-n 16 12 7 14 17 19 Jfl 115. 11! FE. Range 4-29 5-28 3-22 7-23 Mean 14.6 14.4 8.6 11.4 S.D. 5.99 6.02 3.42 3.79 APPENDIX D RAW DATA: JH 70 71 oo I ooo o. o.o o mom omm oom oo omo om 4.8 omo ooo ooo o.m m.m o.m omm ooo ooo omo on nmo mom ooo ooo ooo o.m m.o o.m mom omo omm mm mm omo omm ooo ooo ooo o.o m.o o.o omo ooo omo omo oo ooo Mam ooo omo ooo o.o o.m o. omo ooo oom om oo ooo 6am ooo ooo ooo m.o m.o o.o omo ooo ooo om ooo omo 656 omo ooo I o. o. I oom oom I oo om om m>m ooo oom ooo o.o m.m o.o moo ooo oom oo oo ooo ooo ooo ooo omo o.o o.m o.m moo momo omoo oo ooo moo omm omm oom o.o o.o o.o ooo ooo ooo oo oo ooo ooo ooo ooo oom m.o o.o o. omoo ommo omoo ooo om ooo oom omm ooo oom o.o o.o o.o mmmo omoo mooo ooo oo moo oom ooo ooo ooo o.o m.o o.o oomo oooo moo oo oo ooo mom omm oom oom o.m o o.o moo oooo moo oo ooo ooo omm omo ooo omo o.o o.o o.m oomo ooo ooo om ooo moo oom ooo omm oom o.o o.o o.o moo oooo moo oo om ooo oom ooo ooo ooo o.o o.m o.o mom omm oom ooo oo mo mom ooo ooo ooo m.o o.o m.o mmo mom oom ooo oo oo oom oo oo ooo o.m m.m o.o ooo oo om ooo moo mm mom om ooo oom m.o o.o o.o moo ooo ooo mo oo oo moo oo oo ooo o.o o.m m.o ooo I ooo oo oo oo mam ooo omo ooo o.o o.o m.o moo oom oom oo oo ooo mam omm omm ooo o.o o.m o.o mom mom ooo oo oo oo mom oom oom omm m.o o.o o.o ooom ooom oooo omo mo ooo mom ooo ooo ooo o.o o.o m.o moo mooo omoo oo oo om mam oom omm ooo o.o m.oo m.o ooom ooom momm om oo ooo mom ooo ooo ooo o.o m.o o.o omm ooo moo moo om omo mam oom omm omm o.o m.o o.o oooo ooom oooo oo oo ooo mam m m o m m o m m o omaoe omoooeo Hmong. Hmong. among. 309: 8.3 3me G3 0.53on AOmvSoUV 8:68.on ooom\oov some ad cofloooumcoo m0 258.830 H2 .298“..th 3mm 30am H2 soflmoso souflmmsmoe \omosmoau OBEUOHQ oflmsoo< Eh "puma 3mmll.Q XHQmemfi 72 ooo omo omo o.m m.m o.o moo moo oom om mo om so: ooo omo omo o.o m.m m.o oom mom mom om omo moo so: ooo ooo omo o.m o.m o.m mooo moo mom mo om ooo so: oom oom omm o.o o.m o.o omo mmo omo oo oo mmo so: ooo omo omo o.m o.m m.m moo omm on omo ooo mm 5m: oom oom omm m.o o.o o.m oom moo mom omo omo omo 5N: omo ooo omo o.o o.m m.o omm mom moo om mo omo s>s oom oom ooo m.m o.m o.m moo oom ooo oo oo omo s>s omo omo ooo o.o o. o. mono mmoo omo mo mm moo so: oom oom oom o.o m.o m.m ommo oomo ommo ooo moo moo so: oom omo omm m.o o.m o. mmoo mmoo mom om om moo so: omm omm oom m.o m.o m.o oooo mmmo mmoo moo omo omo so: oom ooo omm o.m m.o o.m oooo mono omo moo om moo 5mm oom omm oom o.m m.o o.m ammo mmmo mmo moo om oom 5m: omo omo ooo m.o o.m o.o mmoo mom mom om mm mm 5w: oom omm omm m.o o.o m.o moo omm omo mm on ooo 5w: omo ooo ooo m.o o.m o.m omm moo mom om omo ooo soc oom omo omo m.o o.m m.o omm moo omo on on moo so: om om omo N.m o.m o. oom ooo mom moo om om so: omo omo omo m.o o.o o.o oooo oooo omm ooo om moo nos ooo ooo omo m.m m.m o.m omm ooo mom mo ooo moo so: ooo ooo ooo o.o o o.m omoo omm moo mm mm mm on: omo omo oom o.o m.o o.m mmm moo mmm mmm oo oo 5x: oom omm omm o.m o.m o.m omom oomm momo mmm mm mm 5x: omo omo ooo m.o o.m o.m mmoo mmmo omo omo om moo sup omm omm omm o.m m.o m.o ooom ooom mmom ooo mm mm on: omo omo oom m.m m.o o.m mom oom omm on omo moo pm: omm oom oom m.oo o.m m.o ooom ooom oomo omo om omo an: m m o m m o m m o ooona omooe omooa oesom omouoco oooms coo Emma 53 0.5%on Smooég 0.53on oommoov Emma coo soouoomflmcoo mo coauMHSQ uo¢_omnomouco mumm soon oom cooumoso soouomsmna \mosmoou ooemnmoonm ooum500¢ .ooscoucooII.o xoozmmmm 73 on ooo ooo N.~ m.o o.o omm Nmm omm mm om mm ooo omo oom omo o.m o.m o.m mom omm omo moo om om ooo omo ooo omo o.m o.o m.m mmm mmm mmm mm moo ms omo oom omm omm o.m m.m o.m omm oom mmm om mm on omo omo omo ooo o.m o.m o.m omo mom omm mm omo mm omo oom oom omm o.m m.o m.o ooo oom omo moo mm on omo ooo omo I m.o o.o I omm mom I mm om mo o>o omo omm omo o.m m.o m.o vow omm oom mm mm om o>o omo ooo ooo m.m o.m m.o mooo omoo oom ooo mo omo ooo omm omm omo o.o o.m o.m omm mmm mom on moo omo ooo omm omo omo o.o o.m m.m mmmo oooo moo mm om omo ooo omm oom omm o.o m.o o.m ommo mmmo ommo moo mo omo ooo oom omo oom N.m m.o m.o mmm mmoo mom mm mm omo omo omm oom omm o.m oo m.o oooo mom ooh mm om mom omo ooo ooo omo o.m o.o m. oom mmoo oom om mm omo ooo omm omm oom m.o m.o o.m oom moo omm mo moo omo owo omo omo omo N.m o.m o.o omo oom omm om om omo omo oom omo oom m.o o.m N.m moo omo omo om mm mmo omo om om om o.m o.o o.m mmm mmm mom mm no mo ooo omo omo omo o.m o.m m.o moo mom omo mm on omo ooo om om ooo m.o o.m N.m omo oom omm mm mm moo ooo omo omo omo m.o m.m m.o omm mmm omo ooo mm moo ooo ooo oom omo o.m m.m N.m omm omm moo om mm omo oxo omm omm omm m.oo m.o m.o momo momo mom omo mm omo oxo omo omo omo o.m m.o o.m omo omm mom oom mm mo ooo omm oom oom o.m o.m o.o momm oomo mmoo oom om omo ooo ooo omo ooo m o.o o.m mmm mmm mmm oo oo omo omo omm omm omm m o.o o.m moom moon mmmo omo mm moo omo m m o m N o m N o omoaa oars omoo. one: oooaoao 68a ad AoomE_aov.mH5mmmom AOmm\EUV mosmmmom Aomm\oov “Dome coo soouoooumcou mo cooumoso Ho¢.omoomnuso mumm zoom oom. coopmoso coouomsmoe \oosmoou poemsmooomm ooumsoom .UODCHUGOUII.Q XHszmm¢ APPENDIX E RAW DATA: DB 74 75 ooo I omo o.m m. m. mmom mom omm oo mo mm mam ooo omo ooo o.m o.m o.o mom omm omm mo mo oo mam om ooo omo m.m o.m m.m moo mom oooo moo oo oo mom ooo ooo oom o.o o.m o.m oooo momo oooo omo mm om omm ooo I ooo m.o m. o.o om oom omm mm oo oo omm omo ooo omo o.m m.o m.o mmm moo omoo mo mm mm omm ooo omo ooo m. o.o o.m omoo moo mooo mo oo oo mam ooo omo ooo o.o m.o o.m oom mom ooo om om oo mam omo I omo m. m.o o.m omoo mmo momm mm om moo mom oom omm ooo m.o m.o o.m momo omoo omoo mo om ooo mom omo ooo omo m.o o.m m.o oo oom ooo om om om mom oom omm omm m.oo o.o o.oo omoo oomo momo om oo omo mom omo omo omo m.m o.o m.o ooo ooom ooom om oo ooo omm omm oom oom o.m o.m m.o omoo oooo oooo omo mm ooo omm ooo omo omo o. m.m m.m om I moo oo oo omo mom ooo ooo oom o.oo o.oo o.o oooo omoo omoo om mm omo mom I I I o.o o. m. I om oo oo oo oo mom ooo omo omo m.o o.m m.o oom omo omo omo ooo om mom o I I m.o o. m. ooo omm mom mm mm oo oom ooo ooo om o.m o.m o.m moo omo omm moo oo oo mom omo omo om o.o o. m. oom moo oo oo oo mm mam omo omo omo m.o o.m m.m omm omo mom mm mm mm 6am oom omo omo o.o o.m o.m oooo omoo omoo ooo ooo ooo mom omo omm omm m.o m.o o.o ooom ooom ooom ooo ooo om mom om ooo omo m.o m. o.m omo oooo omoo moo om mo . mom ooo ooo ooo o.o o.m o.m mmom momm ooom omo ooo om mum omo o oom o.m m.o m. ooo omo omm oo oo omo mam ooo oom oom o.m o.o m.o omoo ooom ooom mm mo ooo mom m m o m m o m m o Homo. oars. omoo. omaoro omfloao 18mg :3 Emma ad m.o:mmmom AOmmBQ mammmom ooom\00o ooomE ed cofloooumcoo mo £0.39on Hod omoomfioso mpmm 30o...Io Hod soopmpdo coflomcmoe \wHSmOoO 0% pounded.“ .mQ "Mn-mo 3mmll.m XHQmemd 76 I! I III Il|I . - ooo ooo ooo o.N m. m.o omm mom on mm mm mm so: omo ooo oNo o.o o.m m.N homo omm omm mm on mm so: I I I m.o o.o m. omm omm oooo mm mm mo so: ooo omo ooN o.m o.m o.m Nmoo omoo ommo oNo moo mo 5m: oNo om ooo o.N N.N m.o omo mNo moo mm mm. mm 5N: omN ooN omN o.o N.m m.oo omoo omoo mmNo on mNo moo 5N: oNo om omo m.o m.N m.o ooo mom mmo om om on 5>5 omN ooo ooo m.m o.m m.m omm omo Nom mm mm on 5>5 ooo ooo omo o.o N. o.N Nmm mmm mmm mm mm omo so: ooo omo ooN o.oo m.m o.oo Nooo oomo momo mm om mmo so: ooo oNN omo m.o o.N m.m omm mooo momo mo om moo so: ooN omN omN o.m m.m m.o mooo Nmmo omoo mo mm moo so: ooo ooo ooo m. o.o o.o on Nooo oooo om mo mm on: oNN ooN ooN o.m m.o o.m omoo memo mmNo moo mm ooN 5m: omo omo ooo m. m. m. oNN moo omm om mm omo so: omN omN omN N.m m.m o.o mNNo momo oomo om mo oNo so: omo ooo ooo m. m.o N.o I mom Nmm om om mm so: omo ooo omo o.m o.m m.m mom oom Nom moo om mo so: om oNo ooo o.o m.N o.N ooo I omo omo mo mo so: ooo ooo ooo m.m m.m N.o Now mom moNo om om mm on: ooo ooo ooo m. o o moo Nom moo omo mm ooo so: omo ooo omo m.m o.m o.o omo mom ooNo omo omo mo so: omo ooo omo m.N m.N o.o I mNo mmN ooN om mmo 5x: omo ooN omN N.m o.o N.o omoN ommo omoo moo mm on 5x: omo omo omo m. o.o m.o mom Nooo Nooo ooo om moN so: ONN oNN ooN m.m m.oo o.oo mmoN mooN oooN ooo mo moo sun ooo ooo ooN o. o.N m. oom Nmm mmo on om ooo on: ooN oNN oON o.oo o.m m.m ooNN oomN omoo om oo oo an: m N o m N o m N o I . omooe omoue omooe omoom omooocH 309: ed 053on 53322 0868.8 A00m\00v 30qu :3 soouoouumsou m0 cooumoso oom omuomuueo oumm zoom oom sooumoso soouomcmne \0H5m0oo ooEmcmooumm UHUmsooa .UmscouGOUIl.m XHOmem< 77 ooo I I o.o m. m. I mmo mom mm om oo ooo oNo ooo ooo m.m o.m m.m ooh mom Nom moo mm mo ooo ooo I No o.o m. N omo oom omo om om om omo omo ooN ONN N.m oo o.m oooo mNmo mmmo mm mm mo omo om om ow m.m o.o o.N oom omm mom mm mm om omo ooo ooN ooN N.m N.m m.o oooo oNo oooo ooo mm mm omo oNo ooo ooo m. m.o m.o mmo mom omm om mo om o>o omo ooo ooo m.m o.m m.m omo omo mooo mm mm mm o>o omo ooo ooo o.o m. o. ommo momo ommo mm mm mmo ooo ooN ooN omN N.m m.m m.m Nmoo mom mooo mmo on om ooo ooo omo omo N.N N.o o.o mom Nom omm mo om om ooo ooN omN ooN N.m N.m m.m oNoo oooo mmNo mm mo omo ooo omo omo ooo o.N o.o o ooo omo omm mm om ooo omo ooN omN omN m.oo o.m m.oo omoo omoo omoo om mo moo omo ooo omo omo m. o o . oNo omoo omNo mm mm oNo omo I oNN ooN m.oo m.oo N.m oNoo. omoo omoo mm mm moo omo omo ooo omo o.m m.o o.N oom mNN omo moo mm mm omo ooo ooo omo o.m m.o o.m mmo omo omm mmo mm mm omo ooo omo ooo m. o.o o.N moo omoo omoo on om oo ooo ooo omo omo m.m o.m o.m omm omo oom mm mo mo ooo oNo I omo N.m o. m. oo moN moN om mm om ooo ooo omo ooo .o.m o.N .o.m moo oom ooo oo oo oo ooo I ooo omo m. o.N o.N mmo Nmo mom moo mo mmo oxo omo ooN ooN m.oo m.m N.m momo omoo oNoo oNo on mm oxo omo omo omo o.o m. N. oNo mom Nom om mm ooN ooo oNN omo omo m.m N.m N.m omoo mmmN mooN moo mm om ooo omo ooo oNo m.o m.o o.o mom moN Nmm oNo mm omo omo I ooN oON oo m.m o.m mmoN mmoN mmoN moo om moo omo m N o m N o m N o Homo. omoyo. owns omaE omfloao 886 as A009: :3 mammum AoNgbv mosmwoum ooom\00v 3me ed :ofloooumsoo mo sooumosn Ho¢_omuomouso oumm zoom oom sooumoso economcmye \oosmOoo oosmsmpoomm ooumsoofi .GODGHuGOUII.m XHQmem¢ LII‘EIATE \JP‘WERSIIY .JBRIAP ES II I III “III III II 1293 031695