{ ;I:I III‘I “"V? I p'l‘ir II 5%,” q \ wkIQ‘“ Vgfi - :4“ ‘44} f {35“ fig fi 7"\ ‘u'l- ‘ 4II ' 53$}? 4 . 4.44 *4“ #5.; 4 “if 444444;” _ I'- u I. . . 44. . ;4:I.5.‘,I $44 I&gé§"§.a5= ,4. . . . -44 _ 4 4_ l . , ,4 v " 4:4 I 74.19:} 44.4445. \4,, ‘.'_I‘.Ii.44:44:‘- I?» 4: a m K444: ‘I II. '. . ‘“.I“ nu ”'1‘“? :«IJi‘u ‘:“3:’(" " ‘944‘ >353..- LJ‘I‘V‘A' “ . I ‘I,' ' égl'. I'I’I, 4.4\.\44~4 . ‘1 .“ . ‘.INII| 3". .4 ~~§>§1.°:_Ef::.$:n:n::: "" ‘ r “ 321:4.“ 3’11"“:4: “$354 ..-1 fi’ ‘nk~w"a-;}x“.zt$:1.u.;rmnfih | 4 ‘ II . I fififi}§‘n‘~' “ “15' €':\&1£A5~$ ,I 143.4 4 4' - .0444.I4 :4. :‘I I.4\1,f4:.,:$’{};\:'?-4-:~4‘{ 4%,..me |44 . .4? ‘h 4.4!“ 444 Q4‘0‘:};‘I::§;:: 13);; a?“ . - . 1.54 - "I“ 142.)“ I 4' ‘ 23:. 'V-‘rI-si ‘ . I. » I I . gfi-‘ifi‘a‘pn I III. I ‘ ‘ 4 313:. 2 L‘, {h ‘4}: ‘ . 44/ . '43“ 43' A wink?) 4, "8 44:14:44 .4 I ‘. .. III“<'~'- Ix‘t. [III/I' t ‘ 4.4 .4 “44,4 I) I}. I' LAWN-4‘ ‘ SLM‘: ‘44444 4 Livy . 4 . . ‘1 4 .4 x . . I344 32%;)“; m4"? A“ l‘ “ 3‘ \l ‘ . 4' 4 4 44 J ”44.4844 . ’ ‘I _ ‘ : h' ‘, \V: “i ‘2 .4 5,0?" . "v.34; 4 .‘-}:‘.~?.I .‘I.u{444 ‘ I :- ; "r "J". . 4 4 , "" ..w.ii.vu»u|"o ., %J«= ‘I I I. ‘ II‘IQ' '2 ‘ I ‘l'i‘é , 4 1 -l‘ U31‘IL“' ‘."‘4 I . 4 ,. '3’: . .444. ' 4 .4 4 | 4 I ”kt“. I “t :4 l‘v"<‘zl." V = | ‘ n D I '. ' ' I‘| | ‘4' I "N" (:v :‘és- . - . r, : ' A. 'I fivfl”+£%5 1 4| 4! 4 4Iv4w||\: 4.4K '*:"‘:.454II- .. ' . a ..-,. ‘I‘ ‘ 4‘ I‘ .' 'I 4 .4 . X", I‘I'I‘.'I .V' . u?‘ 21'§':‘\:“4';11 . .4I II. f._II I”... 3| ' 4\\' 'I 4 j."'-- III .‘IHIgfi . ' ‘ ‘ I ‘ ‘ ' ':I:‘H " h '4 J:|' 35”.; I. ' , I . “f 4!." I-:'4‘,II“;4I I?“ '\‘I"“-"‘ I ‘ ‘. .I I I. . ‘I ‘.I A? V'- A 414 I | .4 - ‘I' I ‘ I'- '4- . 4 4 | 4 ‘ .". I“ ' I‘II4‘4 ‘IIV 0| ‘1." ‘ "it“ ‘ “'\"I‘..‘ T'I‘I,‘ " :I ‘ ‘ ‘ I . “I. I I ‘ ' pr" ("1‘ I . :‘4. Q: ‘5"‘;‘ ‘ I | “‘ 4‘ I..4I.|‘ .JY4‘I4I4.:" “4'4 ’.' 4 ‘ 4 .‘ | 4 4 4 4 'I “‘r 4 IL”. II ‘444l14'4-._'I‘ Sn". " 'I-IID , . . | ‘ 4‘ 444. 4‘44‘4,.'._‘, 4.444?” I1,.II4‘I4. I. 414.4.4.1.44IH444 I' 44442444: 44444, “4'4..444t.44; 4.. . ._ . . ‘ :I ' I 2‘ . II II‘L'r SE.‘ ' J"" I II' .. I 'IIhru‘tlI‘“ | If 4 '44 4 . 4,44 ”‘44 l‘4 I _ >414 [I‘ll 4| 4.4 44A|NII “‘4" I 4 (3; III 4': 4",}: “I ‘1}‘(7‘Ng 0144;474:444; 45?; it? -‘ " .‘ . I. I, I‘ “ I‘If ‘ "' “|‘ ‘||“I“““ “"‘I1I‘.I'I"“ it»? I‘-'" ”4'1 -“‘-'\"' ‘ I'. 4 4. . H 44 4| 4. 4 .44I 44 444444 4. ‘4 I4 4-4 I I 444144 4 '4 44'44 ('18-?! ”L4. "Lx: 4'14 I 4 ,7 .- I “ “ |‘ ' ‘ “ 4.‘ ||‘ Hf.“ '4|"‘:r ““ I 4 I“: '_-I] .44I41. m ‘4! I I! “E"IIHLIU'L _ 4.4 [”440 4 .4444” .. 4. I' -v 4 I I 1II‘HI'I 44* '4424‘ I’ .II. I" I “ . ‘| ‘I " ‘ . - I II ".""I.',"‘ I'Ikv‘ ‘ / ‘ ‘ 4' 4' ‘ ‘ 1 I"'I 14'1“ : .4] :" :5"“4'l4l4‘~u I I 4 4 . 4I “"4 _ ‘ “my .J. I. I 'u'- ' ., I ‘ ‘WLIN .: IH’wat II*I;| .4 4 . I4 .44 . .44 4 I134 41‘ '4‘” "”140 ‘I “If ‘q‘I ‘ _ II | “ C'II ‘II I“. ‘ nun: “II. ‘I ‘.':| I ll|‘ I IIIIIIII IIIIIIIIIIII IIIIIIIIII II 3 129 e991IIII2IIIII' LIBRARY . Mich gun Scans:- University This is to certify that the thesis entitled An Analysis of Speech Dystnctions in Multiple Sclerosis Patients presented by Elizabeth T. I. Akpati has been accepted towards fulfillment of the requirements for Ph.D. degree in Audiglggx & Speech Sciences Date 5-18-78 0-7 639 é‘ffiofiu— . m 100'1'53. ./ \E wagon It APRflafl‘ ‘ fl ‘3; . 5 IEIN‘ u. .12, ' "1. < " A DISSERTATIQE attain-inn : ty ”W“ m “partial fulfil: “LEI: \} #935“ it}: twmu‘yvn‘ m¢M r. .‘- . \flv ‘ :i; r':..'.;.:. --‘ AN ANALYSIS OF SPEECH DYSFUNCTIONS IN MULTIPLE SCLEROSIS PATIENTS By Elizabeth T. I . Akpati A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Audiology and Speech Sciences 1978 6 ”aéya ‘ --' ABSTRACT AN ANALYSIS OF SPEECH DYSFUNCTIONS IN MULTIPLE SCLEROSIS PATIENTS By Elizabeth T.I. Akpati Although it has been documented since 1877 that dysarthria is one of the neurologic signs of multiple sclerosis, little recognition has been given to the disease in the speech pathology literature. The apparent neglect, perhaps, may be due to the fact that the symptoms as well as the neurologic signs may be evanescent. Nbre empirical data are necessary to the knowledge of speech dysfunctions in multiple sclerosis. Thus, the primary purposes of this investigation were: 1. to deternune the effects of imitative and Spontaneous methods of response elicitation on the speech production of patients with multiple sclerosis, 2. to determine the effects of the position of a phoneme in the articulatory responses of patients with multiple sclerosis. A purposive sample of multiple sclerosis patients with speech dysfunctions was selected for the study. Each subject had to meet the following criteria to be included in the study: normal hearing sensi— tivity as measured by audiometric pure tone screening test and speech impairment as deternuned by two or more misarticulations on an articula- tion screening test. The sample consisted of 11 females and 5 males Elizabeth T.I. Akpati and the subjects had varying degrees of severity and duration of multiple sclerosis. None of the subjects had visual—field defects and they all spoke General American English. Errors of articulation were assessed by employing a list of 64 meaningful monosyllabic CVC words constructed from 16 singleton conson- ants and two vowels. TWO types of verbal tasks (imitative, spontaneous) utilized 32 monosyllabic CVC words. Each target phonene was represented two times in the initial position (imitative initial, spontaneous initial) and two times in the final position (imitative final, spontaneous final). Responses were elicited in the following manner: 1. subjects repeated recorded monosyllabic CVC words under ear- phones (imitative task) 2. subjects read aloud monosyllabic CVC words printed in one-half inch letters on a 4" x 6" white card (spontaneous task). Responses for each subject were tape-recorded and scored by six trained graduate speech pathology students for correctness or incorrect- ness of articulation, employing conventional error categories of substi- tutions, omissions, and distortions. A phonetic and a distinctive feature inventory were employed to analyze the misarticulations. Descriptive statistical analyses were employed to 1. determine those consonants that were more susceptible to articulatory errors of substitutions, omissions, and distortions. 2. determine the distribution of errors among the five distinctive features of voicing, duration, affrication, place, and nasality (Miller and Nicely, 1955). 3. determine the predominant type of articulation error in the speech productions of multiple sclerosis subjects. Elizabeth T.I. Akpati 4. determine the relationship of misarticulations to the acquisition hierarchy of distinctive features. Inferential statistical analyses involved utilization of l. a two-way fixed effects analysis of variance in order to a. determine the differences in the misarticulations of the multiple sclerosis group as a fUnction of type of verbal task. b. deterndne the differences in the misarticulations as a function of phoneme position. 2. two one-way fixed effects analyses of variance in order to determine the relationship of misarticulations to the developmental hierarchy of phoneme emergence. The following are the results on the analyses of variance: 1. There was a main effect for task with imitative being poorer than spontaneous task. 2. There was a main effect for position with final—word position being poorer than initial—word position. 3. No interaction was found between task and position. 4. No convincing evidence was found to the effect that the break- down in the articulatory productions of the multiple sclerosis group was related to the developmental hierarchy of phoneme emergence. Based on the findings of this study and related investigations, suggestions were given for further research on speech dysfunctions in multiple sclerosis. 9 :5 ‘:‘ - ~ . . .. 3. ~ , , < v 35’"). .. f\. , - 1} ~‘ 11 j...;". ’9 I. tuft/3.1. 3.9‘ ~ . .11. . .' - . -. ’_‘2 ‘1 .. oxrrr as. .~..1- '. r ., 1.: Is . ~x. .tru .. 2."; we . .. . . - l . ' ‘ .. . },-)*-'.- I”: 1 1 i . ‘ ‘1' . z ‘u ‘YJJ --',1 I1? . $3-1. ' “9'. ' . ‘ ' 1' ' 7”": I. “ll-L-‘LV 54:97:" 1;, 171?: y); : -'..'t, v| . g > .|' Is ‘f...~_ v-j;. .‘_> ‘.. ._l I.‘ ., ., . . .. . - 1.-.,..- ..... 5’ a.» , pitym , , »‘ i" ‘ 'r;1‘ ‘.'.;,'Z‘I‘V: 313311-413 to fill finer. 00:5 1 .2 ' t 'r : m I‘ - arr we: 3m! rents: for their limitless sacrifice and funds 6 lltendless smly’of love audieani-‘z 1112. Snows NWbrothenf-Ben: my smelot‘Mpi’uthn ma 1: hr. . a great believer in me. . {eta-re in l-‘l‘:1'.'.'rl .oimutc. 1 71:19:11-: :nd jut-'11 #3:: i3 .111.- (Ls-13:5: it} additimx, I T'13::‘ 1'12; Judy Ear. ' so: ":11 .IT. Raw-Jar. Freeware) "7' - "iv-:iheirfi Stat; 11:31 33:1:21n-tv .qurxw; a .9; '.-"“ rats tn: the six 'Tt' i5": student . lf‘ Specch ”.17.". 3149.5: who hull; «(ems-d 71': ~. Wluflge‘ ‘ .‘ Wig-iii .qd-ryed :21 this. study. :1! f 22151;: Else Ems to. I ‘ l f Li. nu 2 int: lures Jul nm'iuu 1.1.“: i9! 14mm {i is? _ rs (Wally to the rur‘-i;>11.c.tro¢is {an} r0 .11 ant «Wri‘ - ‘x. Lid not have been possible. .. '2imuly,l Ithank Julie Alexander «5:»: 1'32??? sift. 13" It meet the deadline:, a 1d bark Gunman '1’". ACKNOWLEDGEMENTS I have depended upon many people, directly or indirectly, in many ways over the months of preparing this dissertation. Sincere appreciation goes to my academic advisers Dr. Leo Deal and Dr. Herbert Oyer fbr their invaluable support, encouragement, and guidance throughout my academic program at Michigan State University and during the period of this study. I extend my sincere gratitude to the other members of my dissertation committee — Dr. Jane Oyer and Dr. Linda Gillum for their attention and professional guidance. Special thanks are extended to Larry Johnson for his untiring assistance in writing my computer program and interpreting the output data. In addition, I thank Dr. Judy Frankmann and Dr. Raymond Frankmann for their statistical assistance. Appreciation also goes to the six graduate students in Speech Pathology who kindly accepted to evaluate the recorded stimuli employed in this study. My gratitude also goes to the employees of the nursing homes and medical care facilities and especially to the multiple sclerosis patients without whom this study would not have been possible. Finally, I thank Julie Alexander who typed this dissertation and helped me meet the deadlines, and Mark Greenwald for his assistance in preparing the charts. ii TABLE OF CONTENTS Page LIST OF TABLES ........................ v LIST OF FIGURES ....................... viii Chapter I. STATEMENT OF THE PROBLEM . ............... 1 Introduction .................... 1 Rationale for the Study ............... 4 Purpose of the Study ................ 5 Definition of Terms ................. 6 Organization of the Study .............. 7 II. REVIEW OF THE LITERATURE ................ 9 Epidemiology .................... 9 Classification ................... 13 Diagnostic Category of Miltiple Sclerosis ...... l4 Etiological Considerations ............. 15 Physical Symptomatology .............. 16 Age of Onset .................... 18 Sex Distribution .................. 19 Psychological Factors ................ 19 Hearing Deficits .................. 21 Phonatory Dysfunctions ............... 25 Speech Characteristics ............... 25 Distinctive Features and Conrmmication Disorders . . 31 III. SUBJECTS, INSTRUMENTATION, MATERIALS AND PROCEDURES . . 36 Subjects ...................... 36 2'" Criteria for Selection ............... 37 Instrumentation ................... 39 Speech Stimuli ................... 40 Procedures ..................... 41 Stimulus Generation ................. 41 Method of Stimulus Presentation (Imitative Task) . . 42 Method of Stimulus Presentation (Spontaneous Task) . 43 Assessment of Articulation Errors .......... 43 Chapter APPENDICES Appendix A. Case History Information Case History Information Fonn ............ B. Articulation Screening Test The Fisher-Iogemann Test of Articulation Competence Sentence Articulation Test .............. C. Monosyllabic CVC WOrd Lists for the TWO Types of verbal Tasks 32 Monosyllabic CVC Word List for the Imitative Task 32 anosyllabic CVC Word List for the Spontaneous Task. D. Score Fonm for anosyllabic CVC Word List ........ E. Instructions Used for the Imitative and Spontaneous Tasks F. Confusion Nhtrices of Responses On 16 Singleton Consonants Made by Multiple Sclerosis Subjects iv Page 48 105 105 112 116 128 129 130 131 132 133 134 LIST OF TABLES Table Page 1. Background Information on 16 Multiple Sclerosis Subjects . 38 2. Frequency of Phonemes Mdsarticulated by 16 Multiple Sclerosis Subjects Tabulated According to Conventional Error Categories ..................... 50 3. Frequency of Correct Production and Percentage of Error on 16 Phonemes Obtained from 16 blfltiple Sclerosis Subjects 51 4. Frequency Count of Phoneme Productions by 16 Multiple Sclerosis Subjects ................... 52 5. Frequency of Error Productions and Percentage of Error Obtained from 16 Multiple Sclerosis Subjects ....... 54 6. Classification of Phonenes by Place of Articulation, Nhnner of Production, and Vbicing Features ........ 57 7. Frequency of Substitution of the Feature —-Vbice for + Vbice Based on 64 Observations for Each Phoneme . . . . 67 8. Frequency of Substitution of the Feature + Vbice fbr -Voice Based on 64 Observations for Each Phoneme . . . . 67 9. Frequency of Substitution of the Feature —-Duration for + Duration Based on 64 Observations for Each Phoneme . . . 69 10. Frequency of Substitution of the Feature + Duration for ‘— Duration Based on 64 Observations for Each Phoneme . . . 69 11. Frequency of Substitution of the Feature —-Affrication for + Affrication Based on 64 Observations for Each Phoneme 71 12. Frequency of Substitution of the Feature + Affrication for -Affrication Based on 64 Observations for Each Phoneme 71 13. Frequency of Substitution of the Feature Place of Articulation Based on 64 Observations for Each Phoneme . . 72 14. Frequency of Substitution of the Feature —-Nasality for + Nasality Based on 64 Observations for Each Phoneme . . . 75 Table 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. Frequency of Substitution of the Feature + Nasality for —-Nasa1ity Based on 64 Observations for Each Phoneme . Overall Distinctive Feature Errors Made by 16 Nhltiple Sclerosis Subjects (Task and Position Combined) ..... Cumulative Errors and Percentage Distribution of the Categories of Errors Made on 16 Consonants ..... The Distribution of the Categories of Articulation Errors Nbde on 16 Consonants Based on the Type of verbal Task .......................... The Distribution of the Categories of Articulation Errors Made on 16 Consonants Based on the Position of a Sound in Context ..................... A Comparison of the Performance on Two Types of verbal Tasks by 16 hlfltiple Sclerosis Subjects ......... Types of Misarticulations as a Function of Imitative Task Based on 32 Observations for Each Phoneme ........ Types of Misarticulations as a Function of Spontaneous Task Based on 32 Observations for Each Phoneme ...... Distinctive Feature Errors Made as a Function of Type of verbal Task .................... Pooled Error by Position for 16 Multiple Sclerosis Subjects Types of Misarticulations as a Function of Initial-WOrd Position Based on 32 Observations for Each Phoneme . . . . Types of Misarticulations as a Function of Final-word Position Based on 32 Observations for Each Phoneme . . . . Distinctive Feature Errors Made as a Function of Phoneme Position ......................... TWO-way Fixed Effects (Task x Position) Analysis of variance with Repeated Measures for Response Elicitation by One Group of Multiple Sclerosis Subjects ....... Distribution of Errors in Relation to the Developmental Hierarchy of Phoneme Emergence Based on Responses by 16 Miltiple Sclerosis Subjects ............... Distribution of Errors in Relation to the Ibvelopmental Hierarchy of Phoneme Emergence Based on Responses by 16 Multiple Sclerosis Subjects ............... vi Page 75 77 80 82 82 85 87 88 90 92 93 94 96 98 100 100 Tab 1e 31. One-Way Fixed Effects (Age of Acquisition) Analysis of Variance with Repeated Measures for Response Elicitation by One Group of Multiple Sclerosis Subjects ....... 32. One-Way Fixed Effects (Age of Acquisition) Analysis of Variance With Repeated Measures for Response Elicitation by (he Group of Miltiple Sclerosis Subjects 33. Distribution of Errors in Relation to the Acquisition Hierarchy of Distinctive Features Based on Responses by 16 Miltiple Sclerosis Subjects ............ vii Page 101 LIST OF FIGURES Figure Page 1. A Madified Miller and Nicely (1955) Distinctive Feature System ...................... 46 2. Matrix of 16 English Singleton Consonants ......... 56 3. Percentage Error Rates According to Place Features . . . . 58 4. Percentage Error Rates According to Manner Features . . . . 6O 5. Percentage Error Rates According to Voicing Features . . . 61 6. Subphonemic Feature Distance of 204 Errors of Substitution Made by 16 Miltiple Sclerosis Subjects .......... 65 7. Mean of Errors for Task by Position ........... 99 viii CHAPTER I STATEMENT OF THE PROBLEM Introduction Multiple sclerosis, commonly referred to as MS, is one of the most widespread degenerative diseases of the Central Nervous System, often leading to paralysis and loss of coordination over affected parts. It is "a disease of obscure etiology, characterized clinically by symptoms indicating the presence of multiple lesions in the whiteinatter of the brain and spinal cord" (Harrison et a1., 1966). It remains today, a century and a third after it was first described, "a demyelinating disease with an unknown cause, an unexplained geographic distribution, an unpredictable course, an undiscovered cure and without a simple labor- atory test to confirm its diagnosis" (National Multiple Sclerosis Society, 1969b). Accordingly, Miller (1964) notes that "diagnosis can hardly ever be regarded as absolutely certain short of necropsy findings." Multiple sclerosis has an unpredictable course, varying widely in its mode of progression. The disease is generally dominated by irregular, fluctuating episodes of exacerbations ("attacks" or "bouts") and remis- sions. Because of the dramatic nature of the course of the disease, Bailey et a1. (1956) made the following comment: "The only predictable thing about the course of the illness is that it is unpredictable and may occur as a steady downhill course over many years, or one characterized by remissions and exacerbations lasting from weeks to years." In certain patients, multiple sclerosis advances characteristically by wave upon wave 1 2 of attacks. An attack is those episodic exacerbations which may alter the patient's condition generally or locally in the form of deficit symptoms. This definition also includes the concept that there is a remission. Sometimes, symptoms persist with no improvement. The out- look for recovery is very unfavorable although there may be long re- missions amounting to practical recovery. Those cases in which there are sudden remissions and relapses have little better prognosis than those of gradual onset which are of long standing (McClure, 1936). The disease produces a structural change in the nervous system. Patches of scar tissue develop throughout the brain stem and spinal cord. These areas vary in size from a pinpoint to no more than 1 cm in diameter (Chusid, 1973). The lesions are numerous in the brain and are especially well marked in the pens, medulla and cerebellar peduncles. The term "disseminated sclerosis” has thus been applied to the illness because of the scattered location of the individual lesions. When the involved area is large, the condition is associated with more subsequent gliosis or scarring. It is the last process which gives the nane "sclerosis" to the illness since at autopsy the lesions appear as hard areas (because of the scarring). Depending upon the site of the patches of demyelination, various clinical symptoms are produced; and when they occur within the cerebellar system, the signs and symptoms of cerebellar disease become prominent. Disturbances of the cerebellar function will result in defects of coordination and balance. Sone of the cerebellar signs which may be seen include nystagmus, clumsiness of the hands with intention tremor (i.e., tremor on movement), and ataxia of gait. Speech becomes slurred, typically with a scanning or staccato type of dysarthria. Other signs include difficulty performing rapidly repetitive and alternating move- nents (disdiadochokinesis) (Espir and Rose, 1970). The muscles of articulation may also be affected by cerebellar incoordination, re— sulting in ataxic dysarthria. In mild forms this begins as slurring of speech, and the articulation of consonants is particularly difficult. The ataxic patient frequently exhibits a head tremor which is communi- cated to the whole vocal tract, thus causing a tremulous voice. In other cases, the respiratory and phonatory musculatures may be tremulous themselves, with the tremor triggered by the speech attempt. The patient may experience sudden hypertonic laryngeal spasms during speech (Canter, 1967). Because of poor motor control of rhythm, there is difficulty in blending sound sequences fer connected speech. There is a tendency to produce scanning speech, wherein almost every syllable is pronounced separately and emphasis is put on the wrong syllables, some being pro- duced too loudly and others too softly (staccato speech). Dysrhythmic articulation may result from dysmetria, or the inability to stop articu- latoryxmnmments smoothly at the desired point. The resulting articula- tory contacts are made abnormally tightly, so that sound prolongations and sometimes complete blockages of speech occur. This behavior may accompany other types of dysfluency in the patient's speech, and a neurologic form of stuttering may occur (Canter, 1967). There is a controversy over the terminology about articulatory problems observed in patients with multiple sclerosis. However, it is generally agreed upon that these disturbances are distinct from those based on disturbance of higher centers of the brain related to faulty programming of movements and sequences of movements (apraxia of speech) and the inefficient processing of linguistic units (aphasia). Because speech changes seldom occur until multiple sclerosis is well advanced, the speech pathologist rarely participates in rehabili- tation programs of multiple sclerosis patients. Up to the present time little recognition has been given to multiple sclerosis in the speech pathology literature. Perhaps this neglect stems from the presumption that the disease is progressive and a program of speech retraining would not likely keep up with its "progression" (Guilford, 1956) and from lack of knowledge concerning the etiology and treatments for multiple sclerosis. west et a1. (1968) point out that the disease is an active cause of dysarthria. These authors report that because of the widely scattered lesions throughout the central nervous system, the effect on speech is unpredictable. Manifestations of multiple sclerosis affecting other parts of the nervous system include visual and sensory symptoms, vertigo, loss of bladder control, and upper motor neuron signs due to plaques in the spinal cord (Espir and Rose, 1970). Rationale for the Study There is absence of empirical data in much of the speech pathology literature on the dysarthric speech of patients with multiple sclerosis. Despite the documented importance of dysarthria as a neurologic sign of the disease, few authors have described the speech of these patients. hhny of the conclusions (usually in the medical literature) concerning locus and type of the lesion producing dysarthric speech have been based on gross clinical evaluations of the speech signals. Studies by Scripture (1916), Jenson (1960), Zemlin (1962), and Darley et a1. (1972) employed objective evaluations and qualitative measures to varying 5 degrees in investigations of speech problems in patients with multiple sclerosis. Much more research is needed to pinpoint subtle deficits in the speech and/or language behavior in patients with this neurolog- ical disease. Such an endeavor will lead to a better understanding of the speech characteristics of this population and will provide insight into the commonality between error phonemes in their production viola- tion. These data are of current importance in relation to developing clinical intervention procedures that could help the patient maintain a high level of functioning. The present research could contribute to the small amount of precise information available. Purpose of the Study The primary purpose of the present investigation was to describe . articulation errors made by a selected group of patients with multiple sclerosis. Specifically, the aim was to investigate the effects of imitative and spontaneous methods of response elicitation and the effects of position of sounds on articulation for a group of multiple ‘sclerosis patients. To this end, both a distinctive feature analysis and the conventional phonetic analysis were used. The following ques- tions were addressed in this investigation: 1. Which speech sounds are predominantly susceptible to articula- tion errors of substitutions, omissions, and distortions? 2. Which distinctive features account for the misarticulations that occur in the speech productions of multiple sclerosis patients? 3. Which is the predominant type of error (i.e., substitutions, omissions, and distortions) made by multiple sclerosis patients? 6 4. Is there a significant difference in the misarticulations that occur as a function of the type of verbal task? 5. Is there a significant difference in the misarticulations that occur as a function of the position of a sound in con- text? 6. Is the distribution of misarticulations related to the devel- opmental hierarchy of phoneme emergence? 7. Is the distribution of misarticulations related to the acquisi- tion hierarchy of distinctive features? Definition of Terms Degenerative Disease. A disease in which an essential organ pre— maturely ages or involutes. Disseminated Sclerosis. Lesions appearing in the brain stem and spinal cord. Distortion. The substitution of a standard speech sound by the one which is not normally used in the language. Distinctive Features. Those attributes that distinguish or con- trast one phoneme from others. They are "the physical (articulatory or acoustic) and psychological (perceptual) realities of the phoneme” (Singh, 1976). Exacerbation or Relapse. Aggravation of earlier existing symptoms and signs at least three months after the disease had become static (Oftedal, 1965). An increase in the severity of any symptoms or disease. Incidence. The nunber of new cases of a disease occurring in a given time period per unit population. In this study, the incidence rate of multiple sclerosis is expressed as the number of new cases per 100,000 population per year. Incidence rate reflects the risk of de- veloping the disease in the population. Omission. The replacement of a standard speech sound by a slight pause equal in duration to the omitted sound (Van Riper, 1972). Onset of the Illness. The appearance of the first synptoms or signs which later could be suggestive of multiple sclerosis. Prevalence Rate. It is the product of the incidence and the dura- tion of the illness. It is a reflection not only of the risk of devel— oping the disease in the population but also of patient survival which depends on the quality and availability of medical treatment. In this study, prevalence refers to the number of cases per 100,000 inhabitants in a comnunity at a given time. Remission. Earlier symptoms or signs are no longer present, or the patient is in a position to do what could have been impossible if earlier symptoms and signs had persisted. Substitution. An articulation error in which one speech sound is replaced by another. Organization of the Study Chapter I provides a description of multiple sclerosis, some of the physical and speech manifestations of the disease. The rationale and purpose of the study are addressed in this chapter. Pertinent terms used in the study are defined. Chapter II highlights the relevant literature that has emerged in the area of speech characteristics of multiple sclerosis. The epidemio- logy, etiological consideration, clinical types, physical synptomatology, and psychological effects of the disease are discussed. This chapter also focuses on investigations that have applied distinctive feature analysis as a tool for evaluative and corrective purposes when problems of sound production and/or perception are involved. Chapter III discusses the selection of subjects, the criteria for selection, the materials and the means used to obtain the data. Chapter IV provides the data analyses and results. Pertinent charts and tables are presented to support the data. Chapter V summarizes the study and provides conclusions drawn from the research. Recommendations for future investigations are given. The Appendices consist of the raw data employed in the study. A list of references is also included. CHAPTER II REVIEW OF THE LITERATURE Epidemiology The first comprehensive and integrated description of the clinical and pathologic aspects of multiple sclerosis was made by Charcot (1877). Charcot considered that multiple sclerosis was characterized by spastic paraplegia, ataxia, intention tremor, disturbance of speech, ocular ab- normalities, and nystagmus. Subsequent literature came to consider "Charcot's triad" of scanning speech, nystagmus and intention tremor as being characteristic of multiple sclerosis. Epidemiological investigations on the distribution of multiple sclerosis have uncovered many features of the disease. It has been possible, on the basis of these investigations, to calculate a possible critical starting point for the process (Kurtzke, 1968a). 0n the other hand, the epidemiological investigations have built up a framework whidi is necessary for studying possible etiological factors (Kurland, 1970). In Scandinavian countries, reports on the prevalence of multiple sclerosis have been published since the 1930's. These studies show that the south-west of Finland formed a region with a high prevalence rate of the disease (Parnelius, 1965). The first investigations into the distribution of multiple sclerosis were made by comparing the number of patients who had multiple sclerosis 10 with the total number of patients in various neurological clinics. These investigations, made mostly at the beginning of this century, in— dicated that multiple sclerosis is rather common, for example, in the western and central parts of Europe but uncommon in southern Europe (Hyllested, 1956). Behrend (1969a) collected reports on multiple sclerosis in Europe. He noted that there was an increase from North Germany southward to northern.Switzerland. This increase produced a high prevalence belt in the Rhine River basin. A survey of the fre- quency of multiple sclerosis by the French Multiple Sclerosis Society (1967) disclosed a frequency of 62.8 per 100,000 population in Haute- Garonne (44 degrees N) and 40.6 per 100,000 in Bas—Rhin (49 degrees N). A comparative study was carried out in two communities by Behrend (1966). The prevalence per 100,000 population was 72.7 in Hamburg, the northern city, and 20.7 per 100,000 in Marseille, the southern city. In Scotland, Sutherland (1956) reported a rate of multiple sclerosis of 67 per 100,000 population. A study was carried out by Poskanzer et al. (1963b) in Northumberland and Durham. Northumberland is the most northerly English county, whereas Durham lies just south of Northumberland. The prevalence rate in the two counties was 50 per 100,000 population. In Cornwall in southern England, Hargreaves (1961) noted an overall pre- valence of 63 per 100,000 inhabitants. Millar (1971) put the prevalence figure for northern Ireland at 80 per 100,000. The first estimate of the prevalence of multiple sclerosis in a restricted population was based on United States veterans. The frequency was 14.5 cases per 100,000, and it was greatest among veterans from the north-eastern part of the United States (Davenport, 1922). This characteristic distribution was confirmed in other studies (Acheson and Bachrach, 1960; Beebe et a1., 1967). There are only a few studies 11 describing multiple sclerosis among blacks in the United States. Kolb (1942) presented five cases from Baltimore which had rather typical features; and Alter et al. (1960) reported several cases among blacks in Charleston. Breland and Currier (1967) found cases in Mississippi and Alabama. Kurland and Westlund (1954) noted one case of autopsy proven multiple sclerosis in a black from Boston; and Acheson et a1. (1960) mentioned three other cases among United States military personnel. One salient fact has emerged from these studies. In general, blacks in the United States have a slightly lower multiple sclerosis frequency than whites in the same area. In general, blacks have a lower socio-economic status than whites and thus may not share identical environments, an important etiological factor (Leibowitz and Alter, 1973). With better facilities fer diagnosis and more accurate statistics available, it has been possible to evaluate the distribution of multiple sclerosis by comparing the mortality rate for this disease in different countries. According to these figures, multiple sclerosis is either a rare or non-existent disease in tropical and subtropical zones of the world but rather common in the temperate zone, especially above 38 de- grees in North America and Australia and above 45 degrees in Europe (Limburg, 1950; Acheson, 1965; Kurland et a1., 1965). In the artic zone, the frequency is not higher than in the temperate zone (Alter, 1968). Studies on multiple sclerosis in tropical regions of the world are of special interest in view of the widespread impression that multiple sclerosis is practically non-existent in the tropics. In Jamaica, an island with 1,700,000 inhabitants, Cruishank et a1. (1961) found seven cases of multiple sclerosis in approximately a decade. Three of (II ”T1 (D CLO (I) 12 these were Europeans; and in two of the Europeans, the disease commenced prior to their leaving Europe (Leibowitz and Alter, 1973). Multiple sclerosis is said to be rare in tropical Africa (Georgi and Hall, 1960). Foster and Harries (1970) found two cases in Kenya. South Africa has been of special interest with respect to multiple sclerosis frequency because of its unusual population. Several distinct ethnic groups live in the country including English, Afrikaners, Bantu, Asians, and colored inhabitants. Thus, South Africa has many character— istics similar to those found in Israel. Dean (1967) reported that in inndgrants to South Africa from the United Kingdom and Europe, the rate of multiple sclerosis per 100,000 population was 46.1; English- speaking native-born South Africans had a rate of 10.9; and Afrikaans- speaking native-born South Africans had a rate of 3.1 per 100,000 in- habitants. Among the colored, which include mixed black and white, and among the Bantu, no autopsy confirmed cases have yet been recognized. Inndgrants to South Africa from Europe had high rates oflniltiple sclerosis, like European immigrants to Israel where extensive studies of the two major ethnic groups have been conducted. The prevalence of multiple sclerosis among the European immigrants (largely Ashkenazim) was about five times more cmmmnlthan among the Afro-Asian immigrants (largely Sephardim). A difference in the frequency of multiple sclerosis between immigrant groups (of different ethnic stock) in Israel is com- patible with either an environmental or genetic etiology in multiple sclerosis (Goldschmidt, 1963; Brunner and Lobl, 1958; Kallner, 1958). On the basis of available information on the distribution of multiple sclerosis, the world has been divided into three frequency bands 13 or risk zones (Kurtzke, 1964; Acheson, 1965). The "high risk zone" (prevalence rate 30-60 per 100,000) includes the northern parts of North America (i.e., northern United States and southern Canada) and western Europe north of Switzerland, especially the countries bordering the North Sea. The ”medium risk zone” (5-15 per 100,000) includes southern Europe, the United States, and southern Australia; and the "low risk zone" (0-4 per 100,000) includes Asia and most of Africa, with nost reported values being 1 per 100,000 population. Classification It has long been the practice to set apart a group of diseases in which demyelination is a prominent feature. They are believed to possess characteristics that point to a unique etiology and pathogenesis, as yet unknown. The commonly accepted criteria of a demyelinative disease are (1) destruction of the myelin sheaths of the nerve fibers; (2) relative sparing of the other elements of nervous tissue, i.e., axis cylinders, nerve cells and supporting structures; and (3) a dis- tribution of lesions either in multiple, disseminated foci throughout the brain and spinal cord or in single foci spreading from one or'nore centers. This last attribute, which is not explicit in most definitions, is nonetheless shared by all the generally accepted members of this group of diseases. Two large types of demyelinating diseases have been described (Chusid, 1973): the multiple sclerosis type and the diffuse sclerosis type. The multiple sclerosis type includes (1) classic multiple sclero- sis, which may be acute or chronic, is usually slowly progressive and has its onset in early adult life; (2) acute encephalomyelitis, with a rapid and often fatal course, is characterized by acute onset of neurologic -._..-.. __'-_ 14 signs and symptoms as a result of demyelination of the CNS; and (3) neuromyelitis optica, which is characterized by demyelinating lesions in the optic nerves, brain and spinal cord. The diffuse sclerosis type shows degeneration of the white matter of the brain of a diffuse type and is probably related to genetically determined metabolic disturbances. TWO large subgroups of the diffuse type have been noted: (1) myelino- clastic (Schilder's disease), in which there is destruction of normally formed myelin of the cerebral hemispheres, and (2) leukodystrophies, in which there is a defect in the fermation of myelin, usually associated with pigment deposition in the degenerated areas and occasionally also in nerves and other organs of the body. The onset of the synptoms is common in infancy or early childhood, and the disease is steadily pro- gressive, with death occuring within a few months or years after onset. Diagpostic Category of Multiple Sclerosis Patients with multiple sclerosis are often divided into four diag- nostic categories, based on Allison and Millar's (1954) diagnostic cri- teria: (1) Early Probable or Latent Multiple Sclerosis: This includes patients showing as yet slight or no physical disability and few physical signs but in whom there is a recent history of remitting symptoms of the kind that are commonly associated with onset of the disease, for ex- ample, transient blindness, double vision, vertigo, ataxia, numbness, difficulties with bladder control, weakness in one or more limbs. (2) Probable Multiple Sclerosis: All the patients in this cate- gory are physically handicapped in general, and there is no reasonable clinical doubt as to the diagnosis. In general, the case history 15 contains notes on remissions. (3) Possible Nhltiple Sclerosis: This group is composed of patients showing physical handicap and presenting definite physical signs indicative of CNS disease, clinically suggestive of multiple sclerosis. Reasons for exclusion from the probable group are the lack of sufficient evidence of multiple lesions at various levels and the chronic progressive rather than remitting course of the disease. Hew- ever, in spite of careful examinations, no other cause for the symptoms and signs can be established in these patients. (4) Unaccepted or Discarded Patients: The diagnosis for the dis- carded patients include cases in which investigation reveals some other neurological disease which could anatomically and pathologically explain the symptoms and physical signs, despite their apparent similarity to multiple sclerosis. Etiological Considerations Thny theories and opinions have been advanced regarding the etiology and the mechanisms by which multiple sclerosis is produced (McAlpine et a1., 1955, 1965). The assumption has been made that multiple sclerosis is probably multifactorial in origin. Theories are based on the similarity of this disease to other diseases, on the response of individuals to certain environmental factors, on the examination of blood, and on tissue and,cells revealed at autopsies. There is a growing body of knowledge concerning the theory that this disease is due to an autounmunological process or that it is a "slow virus" infection with a long latent phase (Tourtellotte and Parker, 1968). This viral theory suggests that a specific virus latent within the nervous system becomes activated through trauma or infection 16 and in turn generates the process of demyelination. Autoimmune diseases are thought to be caused by something mis- leading the normally protective inmune mechanism of the body into producing antibodies against some of its own tissue. Other theories that have been advanced include the following: 1. Dbtabolic disturbance 2. loss or inactivation of enzymes necessary to the formation of myelin or their replacement Thrombophlebitis, venule spasm, or some imbalance in the blood Allergens Intoxications Nutritional deficiency states \IO‘UI-hbl Some unidentified poisonous agent None of these theories has yet been validated. Physical Symptomatology The early symptoms of multiple sclerosis are so varied in character, intensity, and duration that any classified description must fail to convey the diverse and often subtle manner in which the disease may first disclose itself OWcAlpine et a1., 1965). Although the diagnosis is made at an earlier stage than was the custom at the end of the 19th century, several years usually elapse before the patient comes under treatment. The individualized symptoms are similar to those caused by any localized lesion of the CNS. It is, however, the pattern of their be- havior which renders multiple sclerosis unique among the organic diseases of the nervous system. Certain common characteristics include the fact 17 that signs and symptoms may be "transient” (Poser et a1., 1966), thus causing difficulty in clinical diagnosis. The signs and symptoms include disturbances (such as nystagmus, diplopia, blurred vision, dimunition of visual acuity, visual field defects, etc.), muscle weak- ness, gait ataxia, nonequilibratory disturbances (intention tremor, dysdiadochokinesia, incoordination of fine movements, etc.), dysarthria (neither cortical in origin nor due to local conditions such as vocal cord paralysis), urinary disturbances, parasthesia (any spontaneous sub- jective disturbance of sensation), emotionalism and often euphoria. A positive Babinski sign, ankle clonus, and increased tendon reflexes are all indicative of a lesion of the pyramidal tract. "The real problem seems to lie with the fact that symptoms, as well as neurologic signs, may be evanescent and mild enough so that medical advice may not be sought by the patient unless some moderately severe degree of functional disability occurs" (Poser et a1., 1966). One of the problems which arise from the unclear and shifting symp- tom picture centers around the issue of credibility of the symptoms to the patient, his family and, at times, health workers. This aspect of the illness is documented by summons, Associate Director of the National Mu1tiple Sclerosis Society, with the following sensitive obser- vation: "Unless these vacillating features of the disease are known, the one caring for the patient is likely to be either bewildered by the rapidly changing character of the patient's complaints or suspicious of the validity of them" (cited in Parnelius, 1969). The cerebellum and its connections control coordination of movements and influence muscle tone. There are important connections with the vestibular mechanism and cranial nerve nuclei concerned with movements 18 of the eyes and neck. Because of this, the maintenance of balance and coordination of movements are dependent upon the integrity of the cere- bellun and its connections. Lesions in the cerebellar peduncles cause cerebellar incoordination evidenced by nystagmus, intention tremor, and scanning speech (Kraft and wessman, 1974). Diplopia, strabismus and ptosis are caused by patches of sclerosis in the midbrain (Boyd, 1970). Age of Onset It has been hypothesized that the process which later manifests itself in clinical symptoms of multiple sclerosis actually begins considerably early in life (McAlpine et a1., 1965). Based on the data from the Israeli studies (Alter et a1., 1962, Alter et a1., 1966), it was calculated that the incubation period for the European.inndgrants to Israel was at least nine years. Schapira et a1. (1963) also calcu- lated a mean incubation period of 20.9 years and a "critical exposure period" and came to the conclusion that multiple sclerosis is acquired at about the age of fourteen. Poskanzer et al. (1963b) and Poskanzer (1968b) report that the disease is rare in childhood. The risk of get- ting multiple sclerosis rises sharply from 15 to 30 and falls even more sharply after that. The mean age at onset for "certain" or "probable" cases is about 30 (Acheson, 1965, Oftedal, 1965, Gudmunsson and Gudmunsson, 1962). It is rather uncommon to find the disease in a patient under 15 years old. In information about 4,000 cases, only 40 occurred under 15 (Gall et a1., 1958). On the other hand, the onset of multiple sclerosis after 50 is very rare too. Of great importance from the point of view of pathogenesis is the established fact that children are especially susceptible to the acute and sometimes rapidly fatal forms r—f . of the disease, for example, neuromyelitis optica and diffuse multiple 19 sclerosis (Chusid, 1973). Sex Distribution Although the literature reports as many males as females with multiple sclerosis, sone studies have found it to have a greater pre- ference for females (Allison and Millar, 1954; Hyllested, 1961; McAlpine and Compston, 1952; Kurland and westlund, 1954; Miller et a1., 1960), McCall et a1. (1968) reported a high female to male ratio in Australia. Kurland (1952b) and Stazio and Kurland (1962) suggested that the apparent higher prevalence among females might be due to their seeking medical advice earlier. Espir and Rose (1970) found a ratio of 3:2 against females. Psychological Factors There has been little agreement of opinion as to the incidence, severity and nature of the changes seen in the mental state of patients with multiple sclerosis. There are reports of the presence of emotional lability and psychological problems. Early investigators like Charcot (1877) regarded intellectual deficits as the main disturbance and noted that enotional lability was not infrequently present. vu1pian (1886) was the first to record the occurrence of "morbid optimism" in patients with multiple sclerosis. Towards the end of the nineteenth century, there were numerous reports of acute psychosis resembling what is today called schizophrenia, occurring in cases of multiple sclerosis. Sachs and Friedman (1922) noted that psychic abnormalities occurred in 15.6% of 141 patients. By contrast Brown and Davis (1922) spoke of ‘r———_—‘. mental alterations in about 90% of patients, of whom 70% were euphoric. The literature seems to maintain that the most prominent change is euphoria. Cotterell and Wilson (1926) in a penetrating analysis of 100 patients, found 63% to be euphoric, 10% depressed, and 84% to show abnormal optimism. They also found general intellectual deterioration in only 29%. The investigators distinguished a eutonia as well as a euphoria -- a prevailing sense of well-being as well as emphasis on the pleasant versus the unpleasant side of every incident. They concluded that the cardinal symptoms of multiple sclerosis are not neurological but are emotional, affective, and visual in nature. Hallucinosis, delusions, excitement, and dementia are rare. However, in progressive forms of multiple sclerosis, lack of responsibility, poor judgment, and poor memory have been noted (Denny-Brown, 1952). Brain (1930) reported a frequency of hysterical symptoms in the disease. The effects of frontal lobotomy have helped to delineate the psychiatric disorders in patients with this disease. Braceland and Griffin (1950) relate the case of a young man with severe obsessive compulsion neurosis which vanished with the onset of multiple sclerosis. Failures in judgment are common. This is evident in such obvious features as concern over minor parasthesias, while disregarding disabling symptoms. Memory defect is a symptom only of gross disorder. In some patients hysterical symptoms may characterize the early stages of the disease (Brain, 1930; Langworthy, 1941). Surridge (1969) described the psychic states of 108 patients with multiple sclerosis selected from records of the United Oxford Hospitals in Kingston, Ontario, Canada. He found intellectual deterioration to be present in almost two-thirds of the patients, depression in about r——————_!- 21 25%, and euphoria to be strongly associated with intellectual deteriora- tion and denial of disability. In general, the literature has maintained that a sizeable propor- tion of patients exhibit lability in which exaggerated laughing or crying behaviors are emitted at inappropriate times. The presence of these emotional behaviors which are elicited by even minimal stimulation of the patient, frequently are involuntary and beyond the control of the patient (Grinker et a1., 1950; Brain, 1962). Hearing Deficits A thorough search of the literature did not reveal current studies on auditory impairment in multiple sclerosis. The research available reveals a diverse and confusing picture of auditory behavior associated with the disease. There are repgrts of both conductive and sensgIi;pegra1_types_of_hearing_loss,_as well as no hearing loss related specifically to multiple sclerosis. This diversity in the literature is not unexpected when one recalls that multiple sclerosis is a fluctua- tipg, although progressive, disease of the central nervous system. Possibly, the best justified conclusion fer the time being on the status of hearing in patients with multiple sclerosis is that advanced by IeZak and Selhub (1966): "there is a reason to conclude this population as a whole demonstrates hearing similar to that found in the general popula- tion of the same age." However, reviews of the research available have shown that auditory aberrations in sone persons with multiple sclerosis mimics that of cases with VIIIth nerve tumor. One of the earliest descriptions of a hearing problem in multiple sclerosis was made by Hess (1888) on a single case of sudden bilateral 22 hearing loss. One ear was permanently involved and hearing returned in the other ear after twenty—four hours. In England, Dundas-Grant (1922) reported a case of multiple sclerosis in which unilateral "nerve deafness" occurred. His study investigated the effect of multiple sclerosis on the auditory mechanism of 92 patients with the disease. Average age in 78 of these (85%) was less than 40 years. The study clearly indicated that defects of the auditory field, comparable to defects in the perimetric field, occur in patients with multiple sclero- sis. The study further demonstrated the destructive effects of the disease on the VIIIth cranial nerve and the associated nuclei. Von Ieden and Horton (1948) tested 92 patients with clinical multiple sclerosis and found that 39 (43%) had some degree of hearing loss (more than 25 dB) in one or both ears. He further stated that these findings had a pathologic basis. Kentner (1954) noted a dome-shaped audiogram, (convex upwards) in about 50% of the patients in his study. He suggests that this is indicative of brain-stem disease. Simpkins (1961) looked at the audiometric profile in patients with multiple sclerosis. He was concerned with whether there is a characteristic audiometric curve in this group of patients. TWenty-eight hospitalized patients with multiple sclerosis were routinely tested by pure-tone audiometry. Their audio- metric curves revealed a tendency to form a specific configuration of depressed acuity for 2000 Hz and 125 Hz. To deterndne whether these findings were due to chance or whether a curve that tended downward from high to low frequencies was characteristic of the air-conduction thresh- olds in patients with multiple sclerosis, an audiometric survey was made of 78 multiple sclerosis hospitalized patients over a three-year period. About 68% of these demonstrated the curve, 14% did not, and T'—"————_——." 23 another 14% were borderline patients. Philips (1952) claimed that vertigo, headache, tinnitus, and hearing loss were not infrequent otologic symptoms in multiple sclerosis. Rose—and Daly (1964) did an eighteen- nonth evaluation of 20 patients with various VIIIth nerve and brain-stem lesions. They discovered definite, reversible temporary threshold shift in two of the ten patients with multiple sclerosis. It has been demonstrated that tone decay from retrocochlear lesions first becomes manifest in the higher frequencies. A greater degree of damage is necessary before the lower frequency decay occurs. Rose and Daly (1964) inferred that the auditory damage in multiple sclerosis occurred in the VIIIth nerve between the Spinal ganglia and the insertion of the nerve trunk into the cochlear nucleus. The authors made this inference on the basis of marked tone decay and on the demonstrable reversibility of such decay and of the pure-tone loss. They suggested that the re- missive nature of the symptoms was attributable to either a reversible demyelination or a perimyelinic edema, i.e., an edemic pressure on the nerve which occurred during the active stages of the disease. The two patients in this study had almost total tone decay despite a normal pure-tone audiogram. Nhny of the investigations report audiometric patterns not unlike those associated with presbycusis. Dix (1968) reported severe loss in the right ear (of a multiple sclerosis patient) characterized by a drop- off to no hearing by 4000 Hz. The left ear had only a slight high fre- quency hearing loss. Hallpike (1967) claimed that "the lesions of multiple sclerosis are confined to the central nervous system. They may occur in the cranial nerves but are then restricted to the zone of the so-called glial protrusion which is correctly regarded as an extension 24 of the central nervous system" (p. 492). Sakamoto and Ichiro (1968) reported excessive adaptation from two multiple sclerosis patients during Bekesy audiometry, thus indicating the possibility of VIIIth nerve lesion. Parker et a1. (1962) speculated that hearing impairment due to multiple sclerosis results frcm damage to "second order neurons." Instances of poor speech discrimination in spite of pure-tone sensitivity have also been observed in cases of multiple sclerosis (Antonelli and DeMitri, 1963). It has long been recognized that the vestibular system may be highly susceptible to multiple sclerosis. Patients with the disease may report vertigo, dizziness, and tinnitus as some of the presenting symptoms. Unfortunately, sometimes, these symptoms have resulted in erroneous diagnosis of Meniere's disease or acoustic neurinona. The fact renains that varieties of nystagmus have been reported with an incidence of S to 57% (Rose and Daly, 1964). Dissociated and ataxic nystagmus are also not uncomon in some patients with the disease (Parker et a1., 1962). Dayal et a1. (1966) reported spontaneous and lateral gaze nystagmus. These varieties of nystagmts, as well as the dizziness and unsteadiness often described by patients with nultiple sclerosis, are probably due to foci of demyelination which may occur any- where from the cerebellum and root of the VIIIth nerve up through the remainder of the central vestibular system (Ward et a1., 1965). The implications of the foregoing review makes clear the fact that multiple sclerosis lesions vary from case to case and in a single individual from time to time. It also highlights the heterogeneity of the auditory and vestibular behaviors associated with multiple sclerosis. mny different suggestions about the type and site of lesion have been 25 offered. Such diversity is due partly to the episodic nature of the disease and partly to differences in the sites and extents of the lesions; and finally it is due partly to the variety of test procedures used by investigators. PhonatoyyrDysfunctions It is said to be characteristic of multiple sclerosis that the speaking voice is strikingly nonotonous and scanning. Collet (1946) found that the vocal range may be reduced by about three tones -- one cause of vocal monotony. The other factor may be explained by reduced tension of the vocal folds. According to Collet (1946), the vocal folds close well in such patients. What is peculiar, however, is the alter- nation between a tense and flaccid appearance of the vocal folds. In contrast to the frequent and typical finding of vocal monotony, other observers have described a chanting perserveration of vocal melody within certain specific intervals. Barth (1911) noted that multiple sclerosis may begin with vocal signs of spastic dysphonia. Leutenegger (1975) reported that the ”odd” sounding quality seen in patients with multiple sclerosis was due to aperiodic vibrational patterns, nasal voice quality (due to disturbed resonance characteristics), reduced pitch variability (due to monotone and severely restricted pitch range), weak'ypice (related to respiration), and trouble initiating vocal fold vibration (due to glottal spasticity). Speech Characteristics Dysarthria has always been recognized as a prominent neurologic feature of multiple sclerosis. The neurological causes of dysarthria (__. 26 are classified according to which part of the neuromusclar system is affected. The disorders may involve (l) mpsples, (2) lower motor neurons, (3) upper motor neurons, (4) extrapyramidal system, (5) cerebellum and its connections, (6) cerebral cortex (motor speech area). Dysarthria may affect the processes of respiration, phgnatigp, artigulation, resgpance, and pEosody. In the field of communication disorders, dysarthria implies any impairment of articulation caused by damage to the nerve centers or tracts (other than those of the language areas of the cerebral cortex) immediately involved in direct control of the musculature used in the enunciation and pronunciation of vowels and consonants (west et a1., 1968). Its basis is some type of abnormality of the central or peripheral nervous system controlling the speech mechanism. Darley et al. (1969a, 1969b) have greatly clarified this speech problem as it occurs in adult patients with neurological disorders. The clinical neurological diseases include pseudobulbar palsy, amyo- trophic lateral sclerosis, bulbar palsy, cerebellar ataxia, dystonia, and choreoathetosis. Each of these diseases affects specific parts of the motor nerve tracts from the cerebellum to the spinal cord. In these studies, Darley et al. described seven speechytypglggy which are useful in differential diagnosis of dysarthria. They are ataxic dys- arthria due to cerebellar ataxia, spastic dysarthria due to pseudobulbar palsy, flaccid dysarthria due to bulbar palsy, hyperkinetic dysarthria due to dystonia, hypokinetic dysarthria due to choreoathetosis and combined spastic and flaccid dysarthria due to amyotrophic lateral sclerosis. lesions in the cerebellum affect its regulatory functions such as timing, range, force, and direction of peripheral movements. A ‘1‘ y“ a <9.) 27 type of dysarthria results producing a coarse, forcibly strained voice. The speaking rate may be too slow or too rapid. Acceleration of the speaking rate results from the loss of inhibition, which is typical for all cerebellar diseases. The slow rate arises from poor coordina- tion among the articulators, and between articulatory, phonatory, and respiratory systems (Leutenegger, 1975). The patient is unable to move his tongue with the rapidity and precision needed for many phonetic adjustments necessary for intelligible speech. A functional disuse atrophy, if present, will also contribute to greater speech dysfunction (Farmakides and Boone, 1960). Because the lesions of multiple sclerosis are primarily of the white matter of the brain and spinal cord and less frequently encroach upon the grey matter, there seldom are symbolic language deficits and gross intellectual deterioration. The term "scanning" or "staccato" speech is often used to describe the speech of patients with multiple sclerosis. It implies that the individual is talking as if reciting poetry, often pronouncing every syllable of every word separately (Canter, 1967). One of the earliest research efferts aimed at describing speech changes as a function of multiple sclerosis was done by Charcot, a French neurologist. In 1877, in his classic description of the sympto- matology of multiple sclerosis, he listed scanning as one of the neurol- ogical symptoms. Charcot felt that speech difficulties which result from the onset and progression of the disease were an important aid for diagnosis. He described what he considered to be the characteristic triad of signs of multiple sclerosis -- intention tremor, nystagmus, and scanning speech. Since Charcot's day, it has been commonly accepted 28 that scanning speech is typical of multiple sclerosis. This view is reflected in many contemporary neurology textbooks. "In multiple sclerosis the speech is characteristicall scannin in type; with slowness, stumbling, halting, slurring, and ataxia of a cerebellar ’———fl] typejfl The spacing of the words with perceptible pauses between words and irregular accenting of the syllables give the sing-song or scanning character which has been described as pathognomonic of the disorder" (DeJong, 1967) . /" Further evidence concerning the dysarthric speech of multiple sclerosis was reported by Merritt (1973): "The characteristic scanning speech of multiple sclerosis is the result of cerebellar incoordination of the palatal and labial muscles combined with dysarthria of cortico- bulbar origin." west et a1. (1968), alluding to the dysarthria associated with multiple sclerosis, state: "The most frequent type of dysarthira that results from multiple sclerosis is a drawling, labored articulation, classically described as scanning.” The authors consider multiple sclerosis to be the only progressive neurologic disease which ”progresses slowly enough to warrant a program of speech rehabilitation." Some neurology textbooks and studies are not in agreement that scanning speech is characteristic of multiple sclerosis. Scripture (1916) did an instrumental analysis, using the phonautograph method, of the vocal changes in multiple sclerosis. Analysis of recordings of sustained vowels demonstrated that there is a presence of peculiar vibrations in recorded speech of patients with disseminated sclerosis "regardless of whether any speech defect could be detected by the ear or not." Scripture attributed the peculiar vibrations to laryngeal ataxia. He decried the use of the tenm "scanning" speech often attri- buted to multiple sclerosis. The tenm ”scanning" he emphasized "is 29 applied to prosody to the marking off of the long and short or the loud and weak syllables, that is, to indicate the maxima and minima. In scan- ning a line or verse with the voice, the speaker exaggerates the differ- ences between the two kinds of syllables, making the emphatic syllables more emphatic (longer and louder) and the unemphatic ones less marked (shorter or weaker). This is exactly what the patient with multiple sclerosis does not do.... The speech is thus neither scanning, nor anti-scanning, nor staccato nor rhythmic.” Rather, Scripture concluded that the subjects used in his study displayed irregular timing, faulty emphasis and variable articulatory errors. Grinker and Sahs (1966) also report that scanning, along with nystag- mus and intention tremor (the Charcot triad), is rarely seen in the early stages but is symptomatic of ed multi 1e sclerosis. Janvrin and worster-Drought (1932) extended Scripture's research strategy, using both smoked-paper tracing technique and film sound tracks. They also found irregularities indicative of laryngeal ataxia. According to Brain (1962), "Dysarthria may be due either to spastic weakness or to ataxia of the muscles of articulation or to a combination of these factors. In the early stages, articulation may be slurred, later it may become ex- plosive and almost unintelligible. The 'syllabic' or 'scanning' speech, sometimes regarded as typical is exceptional." Farmakides and Boone (1960) reviewed the case histories of 82 patients with multiple sclerosis and found five characteristics generally contributing to dysarthria: nasal voice quality, weak phonation of voice and poor respiration cycle, changes in pitch, slow rate, and intellectual de- terioration coupled with enotional lability. They reported that among the 68 of the 82 who received speech retraining, 85% (58 patients) 30 demonstrated improvement, especially in terms of increased rate of speech and louder phonation. TWO studies, Schumacher (1950) and Gordon (1951) show that neural impairments to the muscles used for speech and feeding (respiratory, laryngeal, pharyngeal, and oral) produce weakness and incoordination, reducing the intelligibility of speech. Along with this organic im- pairment, Gordon (1951) implies that there is usually a fnnctional over- lay Of disuse atrophy contributing to the overall involvement. The implication is that this inactivity results in disuse atrophy of the muscles used for speech and feeding, so that there is a greater speech dysfunction than can be attributed only tO organic impairment. This disuse atrophy which produces a more severe picture of dysarthria and prevents the patient from using his residual capacities efficiently is highly vulnerable to remedial speech training. Darley et al. (1969a) described a group of unique speech deviations characteristic of certain neurological disorders. Of these, they noted that ataxic dysarthria, due to cerebellar dysfunction, is characterized by imprecise consonants, stress disturbances, irregular articulation breakdown, distorted vowel, and harsh voice. The research of Zemlin (1962) addressed the acoustic aspects of multiple sclerosis. Zemlin did a spectrographic and motion picture sound track analysis Of both contextual speech and prolonged vowels. The results demonstrated that 14 of the 33 subjects manifested no wave pattern that differentiated them from normal subjects. Others had vibration patterns showing extreme variability in period functions. Also feund were gross changes in energy distribution in vowel produc- tion by patients as contrasted with normal subjects. Jenson (1960) studied the motor speech of 50 patients with multiple 31 sclerosis. He feund that 38% made errors (mean 1.72 errors) on an articulation test and 35% made articulation errors on a contextual speech task. Darley et al. (1972) observed, over a 38-month period, speech problems in 168 patients with confirmed diagnosis of multiple sclerosis. In contrast to early reports, the authors observed that dysarthria does not contribute to an "almost constant" part of the symptom picture of multiple sclerosis. Fifty-nine percent of their subjects showed overall normal speech adequacy. Furthermore, they feund that the most frequent speech deviations were impaired loudness control and harshness; and less frequently occurring deviations were defective articulation, restricted use of vocal variations for emphasis, poor pitch control, hypernasality, inappropriate pitch level, and breathiness. Accordingly, the so-called scanning speech was not a prominent characteristic of multiple sclerosis. They added that speech deviations were solely attributable to cerebellar involvement, becoming nore marked as additional motor systems were implicated. Effects of Drug Therapy on Speech various treatments based upon the etiologic and pathologic theories have been tried with disappointing results (Chusid, 1973). Drug therapy in multiple sclerosis is directed toward the control and amelioration of symptoms. This symptomatic therapy is used to control spasticity and reflex spasms, to relieve bladder dysfunctions, to ameliorate visual disturbances, and to help the patient cope with emotional distress of depression and apprehension. One of the more common treatment methods which has been employed is isoniazid. In a preliminary report on the effects of this drug in the treatment of multiple sclerosis, Kurtzke 32 and Berlin (1954) reported evidence which indicated that the drug pro- duced beneficial effects. Matthews et al. (1960) reported the effects of isoniazid on the speech of 12 experimental subjects who received dosages of this drug over a period of at least 3 months and 10 control subjects who received dosages of a placebo over the same length of time. The speech samples were rated by 42 listeners. The results in- dicated that there was no significant difference between the two groups as far as changes in speech behavior were concerned. It was concluded that the use of isoniazid did not improve the speech of the experimental group. This finding was consistent with that reported by Nagler et a1. (1957) and by Kurtzke and Berlin (1957). Distinctive Features and Communication Disorders Traditionally, research into misarticulation of sounds has been concerned with discovering which speech sounds are most vulnerable to omissions, distortions, substitutions and additions. In recent years, linguists have argued that the speech sound or phoneme should not be regarded as the primary unit of linguistic analysis. Rather, each phoneme should be described as a bundle of phonetic features, each of which is given a value of either plus or minus (Chomsky and Halle, 1968). The phonetic feature, then, is considered to be the primary unit of analy- sis. This approach has been used to advantage by a number of investiga- tors in the field of communication disorders who have sought to apply the linguistic concept Of distinctive features to various problems Of speech acquisition and speech therapy. Snow (1964) and Singh and Frank (1972) used this approach in studying the articulation of individuals with normal speech, whereas NtReynolds and Huston (1971) and anyuk (1968) 33 applied distinctive features to pathologies. Compton (1970) used phonetic features as well as phonologic rules in the analysis of mis- articulations. Some of the most influential exponents of distinctive feature theory have been Pole Jan de Courtenay of the late 19th century; Jakobson, Fant and Halle (1951); Halle (1964); and Chomsky and Halle (1968). Several reasons have been offered for analyzing correctness or incorrectness of phoneme production by means of a distinctive feature system: (1) an identification of the nature of misarticulations of a phoneme permits therapy to be directed specifically to the one violated characteristic of that phoneme, (2) an identification of feature violations common to several phonemes facilitates speech therapy, (3) the system provides greater descriptive economy than the traditional method of enumerating misarticulated phonemes, (4) the system permits clarity and completeness of a phoneme sound, (5) an analysis of the feature changes which affect substitutions provides insight into the aspects of speech sounds which are vulnerable to changes. Distinctive features provide the basis for specifying the relevant attributes of the phonemes in a language. Each attribute consists of two or more discrete and mutually exclusive values. For example, a binary feature such as voicing has just two values, and every phoneme is either voiced or unvoiced. A complete feature system distinguishes all phonemes from each other. various systems of distinctive feature analysis have been developed to describe the phonemic systems of world languages. These systems vary 34 in the number of features used to distinguish phonemes. They include Jakobson, Fant and Halle (1951); Nfiller and.Nicely (1955); Halle (1964); Singh and Black (1966); WiCkelgren (1966); Peterson and Shoup (1966); and Chomsky and Halle (1968). The more complex systems of distinctive feature analysis are the acoustic-based system of Jakobson, Fant and Halle (1951) and the articulatory-based system.of Peterson and Shoup (1966). Their complexity makes them less suitable fer routine articu- latory testing. Jakobson, Pant and Halle (1951) suggest that phonemic systems of all languages can be described by the use Of 12 features, of which only 8 apply to 21 English consonants described by them, Halle (1964) and Wickelgren (1966) use 8 of the 12 features to describe 23 consonants of English. ‘Wickelgren's system attempts to predict short-term.memory errors among 23 English consonants. Wickelgren posited five values for the place feature, one ternary feature (openness), and two binary features (voicing and nasality). The Miller and Nicely (1955) system predicts confusion among 16 English consonants. Singh and Black (1966) extended the number of Miller and Nicely features by adding liquid (to distinguish /d/ from /r,l/), glides (to distinguish /b/ from./w/ and /g/ from /j/), and retroflexion (to distinguish /r/ from /1/); and added a fOurth value to the place feature (to distinguish /tf/ from /h/). The resulting system (Singh and Black, 1966; Singh, 1968, 1970) distinguishes all phonemes from each other. Chomsky and Halle (1968) expanded the list of distinctive features to more than 30, with 13 features listed as applicable to English. Their system can be inter- preted at two levels. As binary features, the system specifies the phonemes as they are used in lexical entries. The entries can be quite abstract since complex phonological rules are required to convert theml 35 into phonetic representations. .As non-binary parameters, the features "provide a representation of an utterance which can be interpreted as a set of instructions to the physical articulatory system or as a refined level of perceptual representation” (Chomsky and Halle, 1968, p. 65). In a recent article (1972), Stevens and HOuse proposed.be- tween 25 and 30 features. From.the studies cited above, it is apparent that there is a theoretical disagreement among linguists regarding how many features are needed fer coding all phonemes in all languages of the world. walsh (1974) is not in agreement that the use of existing distinc- tive feature systems is economical when applied to clinical speech diagnosis and therapy. walsh contends that distinctive features are abstract, often far removed from the physical surface realities Of human speech. Forthermore, such a system.does not serve an advantage for explaining speech sounds that deviate grossly from normal speech patterns. walsh advocates a classificatory system.of feature analysis based.on speech production rather than the existing systeerhich is motivated by a concern fer Optimum notational economy. Hewever, Pollack and Rees (1972) made strong claims for the use of distinctive features. They suggest that the "distinctive feature concept may be applied constructively at every stage of clinical management of a child with an articulatory disorder." IMcReynolds and Huston (1971) and McReynolds and Bennet (1972) cite descriptive economy as one of the virtues of distinctive features. CHAPTER III SUBJECTS, INSTRUMENTATION, MATERIALS, AND PROCEDURES This chapter discusses the selection of sUbjects, the criteria for selection, the instrumentation, the procedures used to construct the speech stimuli, and the means used to Obtain the data. Subjects A.tota1 sample Of 16 subjects with a confirmed diagnosis of multiple sclerosis participated in the present study. There were 11 females and 5 males selected from a total population of 51 multiple sclerosis patients who were screened. The remainder of the sUbjects were re- jected on the basis of failing to meet the stated criteria fer the study such as lack of demonstrable speech problems, impaired vision which did not allow them to perform.the tasks, and hearing impairments as determined by a pure-tone hearing screening test. Nine of the 16 subjects who met the criteria were residents in nursing homes, medical care facilities, and seven resided in private homes in the midéMichigan area. The subjects ranged in age from 23 to 66 years with a mean age of 49 years. Seven patients in the group had.had the symptoms of the disease fer at least ten years, and three of the patients had had the symptoms for twenty years or more. All of the subjects in the sample were mentally alert as determined by the investigator's 36 37 conversation with each subject and by simple structured questions de- signed to elicit Specific answers. An example of the fOnm used to obtain background infOrmation from.the subjects is given in Appendix A. Current medical infermation (which included.medication taken, psychological problems, and the presence of neurological disease other than multiple sclerosis) was obtained from the attending director of nurses at the nursing homes and medical care facilities. FOr subjects who resided in private homes current medical information was determined by the subject's verbal account since no other means was available for obtaining this information. Other background infOrmation such as the onset and duration of the illness was Obtained directly from the patients. Thirteen patients were receiving medications fer the symptomatological control of multiple sclerosis at the time of the testing. Table 1 gives the summary of pertinent background information on each patient. Criteria for Selection All sobjects included in the sample were given a bilateral audiometric pure-tone screening test at 20 dB HL at octave intervals of 1000 Hz, 2000 Hz, and 4000 Hz (re ANSI - 1969). The pure-tones were presented via TDH-39 earphones mounted in MX 41/Ar cushions. .A calibrated pure-tone audiometer (Beltone Nbdel 10C) was used. Accord- ing to these criteria, all subjects in the present study had hearing thresholds within normal limits. The criterion fer failure was lack of response at any of the frequencies in either ear. Furthermore, all subjects included in the research sample were given an articulation screening test. All subjects demonstrating defective speech as 38 TABLE 1 BACKGROUND INFORMATION ON 16 MILTIPLE SCLEROSIS SUBJECTS Subject Sex Age Duration of Illness B.M. F 58 5 years D.H. M 30 11 P.G. F 57 25 D.M. F 63 18 C.M. F 45 13 G.T. M 27 7 K.B. F 27 7 J. I F 49 20 J.J. F 63 14 J.S. -M 56 4 E.P. F 53 11 J.E. M 66 17 V.M. F 55 12 S.L. M 23 4 C.B. F 43 14 D.M. F 64 25 39 determined by the case history reports as well as contributing two or more errors on the articulation screening test were considered to meet the criteria for selection. The Fisher-Logemann Test of Articulation Competence Sentence Articulation Test (1971) was the instrument of measm‘ement. None Of the subjects in the sample demonstrated any visual-field defects nor did the medical report on any of the patients indicate any premorbid psychological problems or senility. All subjects were native speakers of General American English, and all were from Michigan. Thus, the confounding variable of dialectal differences was eliminated. No effort was made to classify subjects according to type and severity of multiple sclerosis since the purpose of this study was not to compare phoneme production between patients with different types and degrees of severity of multiple sclerosis but rather to analyze overall phoneme production in multiple sclerosis. Instrumentation The equipment used to collect the data included the following: 1. Reel to Reel Tape Recorder (Crown Transport, Model 700) 2. Electrovoice Microphone (Model 545) LN Nakamichi Cassette Tape Recorder (hbdel 350) .b Portable Cassette Tape Recorder (Wollensak 2520) Portable Cassette Tape Recorder (Panasonic) 6. Portable Audiometer (Beltone Model 10C) 7. Sound Level Meter (Bruel and Kjaer, Type 2240) 40 Speech Stimuli .A list of 64 monosyllabic (CVC) English words was used to assess articulation errors made by multiple sclerosis patients on both imita- tive and spontaneous tasks. Thus each task consisted of 32 words. Since certain consonants cannot occur at the beginning and end of a syllable, a workable list of 16 true singleton consonants was con- structed on the basis of their ability to form cognates and to occur in the initial and final positions. The following exceptions were made: 43/ and /3/ do not occur in the initial position in English; /w/ and /h/ are not terminal in English; and /f/ was excluded from the list of true consonants used in this study because its voiced cognate /3/ was also excluded. According to Singh and Frank (1972), /h/, /w/, /j/, /l/, and /r/ are non-true consonants and since they have a vowel-like quality, they were excluded from the construction of the word list. An attempt was made to control not only for the phonetic environment of the vowels used.but also the type used. To this end, only two types of vowels were used in generating the CVC word list: /e/, a mid-front vowel, and /o/, a mid-back vowel. It was felt that these two vowels are less resistant to dialectal variations than other pure vowels in American English; thus, their interpolation would.mini- mdze slight differences in enunciation. The probabilities of some sounds following others in English are unequal. It was impossible to generate a meaningful monosyllabic CVC word for /9/ initial using either /e/ or /o/ as the interpolating vowel. Furthermore, no meaningful monosyllabic CVC words could be generated fer /37 initial and /z/ initial in the spontaneous task using either /e/ or /o/. Fer this reason, other vowel types were used: /3/ "thought" 41 in the imitative task; //\/ "thud", /E/ "them", and /1/ "zip" in the spontaneous task. The 64 words in which the 16 singleton consonants were tested are presented in Appendix C. In summary, the list of 64 monosyllabic (CVC) words consisted of 16 initial and 16 final consonants for the imitative task; 16 initial and 16 final consonants for the spontaneous task. Procedures StiJmilus Generation The 32 stimuli used in the imitative task were first recorded on a reel-to-reel tape recorder (Crown Transport, Model 700) by a Caucasian male who spoke General American dialect. The recording was done in a sound-treated recording booth. An Electrovoice microphone (Nbdel 545) was used. The 32 stimuli were then put in random order by a program run on the PDP-ll/40 computer system. The PDP-ll was able to play back the 32 stimuli automatically for the imitative task in the random order chosen by the randomizing program. Each stimulus used in the spontaneous task was then matched with the corresponding stimulus used in the imitative task with respect to the type of phoneme and the position of the phoneme in a word. Thus the order of presentation of the stimuli for the two types of verbal tasks was randomized for each subject. In Appendix D a sample of the randomized stimuli is illustrated. An interstimulus-interval of 8 seconds was inserted to allow for listener response and to eradicate the effect of articulatory transition from one stimulus to another. From the Crown Transport reel-to-reel tape recorder the 32 stimuli 42 used in the imitative task were transferred to a Nakamichi cassette tape recorder (hbdel 350). Speech production of the 16 subjects in the present study was assessed in the imitative and spontaneous conditions. The subjects were tested individually in a nursing home, medical care facility, or in a private home. All testing was administered by the same examiner. Each singleton consonant was tested two times in the initial position and two times in the final position. Four different stimulus words were used to test four representations (two initial and two final) of a given phoneme. Each phoneme included in the study was tested within a single stimulus item. The 64 test words were randomly assigned to the two treatment conditions as explained in the section titled "Stimulus Generation." Thus, each subject was given an opportunity to produce each of the 64 test words once in the imitative task and once in the spontaneous task. All subjects were tested in the imitative condition first and in the spontaneous condition last. Method of Stimulus Presentation (Imitative Task) All testing was done in a quiet room at the various testing locations. Each subject was seated facing the microphone which was placed on a table with the tape recorder. The examiner sat facing the subject. A practice session was provided prior to the presenta- tion of the test words. The volume of the tape recorder was set at a comfortable listening level for each subject. For each subject a set of instructions (See Appendix E) was given orally by the investigator. Formal testing began after the subjects demonstrated comprehension of the task. The subjects were asked to repeat with a single response after each stimulus word. The 32 stimulus words were presented via 43 earphones using a Wollensak (Model 2520) cassette tape recorder. The responses given by the subjects were tape-recorded on thell C60 cassettes using a Panasonic portable tape recorder. Method of Stimulus Presentation (Spontaneous Task) The testing environment and recording procedures were the same as for the imitative task. There was an interval of two minutes be~ tween the presentation of the two types of verbal tasks. In an effort to reduce the degree of visual complexity of the words employed in the spontaneous task, letters were printed in block form. Each stimulus word appeared alone in one-half inch letters on a 4" by 6" white card. The cards were presented one at a time to the subjects who were instructed to read the word aloud once. A practice session was also provided prior to formal testing. Assessment of Articulation Errors The responses for both the imitative and spontaneous tasks were evaluated and scored by the investigator at the time of the testing. All responses for the two types of verbal tasks were treated in the same manner. If the response was correct, a check was placed in the appropriate section of the response sheet. If the subjects substituted a sound for the one which would have been correct, the phonetic symbol for the substitution was used. If the target sound was omitted, a dash (--) was used to indicate the error of omission; and if the subjects distorted a sound, a phonetic symbol and diacritical marks and written explanations were used to indicate the nature of distortion. Any peculiar articulatory or phonatory behaviors demonstrated by a subject were noted. The tape-recorded words were subsequently scored independently 44 for type of error by six graduate students in Speech Pathology at Nfichigan State University. Each judge was required to pass a bilateral hearing screening test. As part of her educational training, each judge had completed at least one course in phonetics and one course in articulation disorders before being trained as a judge for the present study. The six judges listened to the tapes individually at two different sessions. They were required to indicate whether a response was correct, substituted, omitted, or distorted. .A response recording sheet was provided for each judge. Symbols of the International Phonetic Alphabet were used to denote the consonant errors. In order to reduce the effect of order-of-tape and familiarity effect with the stimulus words, the tapes were randomized for listening during the second session. The second listening session fOr each judge occurred not less than one month after the first listening session. NOne of the judges was allowed to view the score sheet of the first listening session. The six judges assessed the articulation errors by applying the response technique used by the investigator at the time of the testing. An agreement by four of the six judges constituted the criterion of acceptance of the correctness or incorrectness of a phoneme. The scored utterances for each subject used in the study were tabu- lated as confusion matrices. Only the errors of substitution were submitted to confhsion.matrices. For a given subject, a separate con- fusion matrix was tabulated for substitution errors in the initial position, and a separate one was done fer sounds in the final position (regardless of the type of verbal task). In other words, all substi- tution errors of sounds in the initial position fer both the umitative and spontaneous tasks were pooled and tabulated in a single confhsion 45 matrix. The same procedure Obtained for all substitution errors of sounds in the final position for the two types of verbal tasks. Group confusion.matrices were also tabulated for responses in the initial- word position, final-word position, initial- and finaléword.positions combined, imitative task, and spontaneous task. Each confusion matrix for each subject summarizes 256 potential confusions for the initial position and the same number fer the final position, thus making a total of 512 potential confusions per subject. The matrix shows the frequency with which each phoneme was correctly produced or substituted by another phoneme. Each row of the matrix indicates the phonemes presented, and each column of the matrix shows the response of the subject. Each cell of the matrix represents one of 256 possible phoneme response pairs, and the number of correct responses was obtained by totaling the frequencies along the main diagonal. .All errors of production contribute to the off-diagonal pattern. From the other cells to the right of the confusion matrix, the degree to which misarticulated sounds fall into categories of articulation errors is presented. Details of the confusion matrices can be fOund in Appendix F. As part of the organization of the raw data fer this investigation, a distinctive feature chart was constructed for each sUbject. The Nfiller'and Nicely (1955) five-feature system.was used with slight modifications. The features are veicing, Duration,.Affrication, Place, and.Nasality. The phonemes /f/ and./3/ used by Miller and Nicely were not included in the present study. It will be recalled that these two phonemes cannot fOrm.a pair since /3/ does not occur in.the initial position in English. Thus, in the present study, the phonemes /tf/ and 46 /d3/ replaced the Miller and Nicely /f/ and /3/. Similar classifications were retained for the affricates /tf/ and /d3/ as those used by Miller and Nicely for /f/ and /3/ except with respect to the feature of Duration. Whereas Nfiller and Nicely treated /f/ and /3/ as having the feature Duration, the present investigator treated /tf/ and /d3/ as not having the feature of Duration. The Miller and Nicely distinctive feature system is a combination of binary and ternary features. The authors consider the features veicing, Duration, Affrication, and Nasality as strictly binary; and the Place feature is considered as a three-category feature: front, middle, and.back. Thus, when the binary feature was used, a designation of zero (0) indicates the absence of a given feature and a designation of one (1) indicates the presence of a feature. HOwever, fer the Place feature the ternary system is used where zero (0) indicates the Place feature front of the oral cavity, (1) indicates the Place feature middle of the oral cavity, and (2) designates the Place feature back of the oral cavity. A.modified Miller and Nicely distinctive feature system.is presented in Figure 1. Data Analyses Several analyses were used to provide answers to the various re- search questions addressed in this study. Descriptive statistical analy- ses, a tweeway fixed effects analysis of variance with repeated.measures, two one-way fixed effects analyses of variance with repeated.measures were used to analyze the data. When a significant difference was feund be- tween two related.measures, the critical difference was used to determine the magnitude of the differences. Since the results from the various criterion.measures were not directly comparable, all raw scores were converted into percentages fer statistical analyses. Forthermore, all 47 Consonant Specification Feature p t k f 9 s tf b d g v 3 2 (is m n VOICING O 0 0 0 0 0 0 1 1 l 1 1 1 1 1 1 DURATION 0 O 0 0 0 l 0 0 0 0 0 0 l O 0 0 AFFRICATION O 0 0 l l 1 1 O 0 0 l 1 l 1 0 0 PLACE O 1 2 0 l l 2 0 1 2 0 l l 2 O 1 NASALITY 0 0 0 0 O 0 0 O O 0 0 0 O 0 1 1 Figure 1. A Modified Miller and Nicely (1955) Distinctive Feature System raw data were submitted to a computer program using the Statistical Pack- age fOr the Social Sciences (SPSS). The CDC 6500 computer was used to perform.the statistical test fOr significant differences. Chapter Iv discusses the statistical and descriptive analyses and the results of the study. Summagy The speech output of 16 patients with multiple sclerosis was tape- recorded. Each subject participated in two types of verbal tasks and was required to produce a total of 64 words which were constructed on the basis of monosyllabic CVC combination. Sixteen singleton consonants were tested in the initial and final positions only. The test session lasted approximately fifty minutes during which each sUbject individually received a hearing screening test, an articulation screening test, and the two types of verbal tasks -- imitative and spontaneous. The responses of the subjects were tape-recorded and subsequently scored by six gradu- ate students in Speech Pathology fOr correctness or incorrectness Of articulation. Each subject's productions were then tabulated as confu- sion matrices and a modified.Mfiller and Nicely (1955) distinctive feature system was used to summarize the data. CHAPTER IV RESULTS AND DISCUSSION This chapter presents the findings Obtained from the 16 multiple sclerosis subjects on two types of verbal tasks. Wherever applicable, descriptive and inferential statistical analyses which were perfOrmed to provide answers to the various research questions asked in the study are presented. A visual representation of pertinent data is also pro- vided by the use of tables and figures. Interpretation of the results is discussed with reference to certain limitations of the study. Results The following questions were addressed in the present investigation: 1. Which speech sounds are predominantly susceptible to articulation errors of sUbstitutions, omissions, and distortions? The purpose of the singleton consonant analysis was to determine whether certain consonants were much more impaired than others and.what particular difficulty they posed in their production. Difficulty of pro- duction was determined by the number of errors of substitutions, omissions, and distortions. Of the total 1024 singleton consonant observations, 238 (23.2%) were produced in error. Table 2 shows the distribution of errors made on eaCh of the 16 consonants. Table 2 reveals that certain conson- ants contributed to a greater incidence of errors in their production than others. These consonants have been ranked in descending 48 49 order from the most frequently misarticulated to the least frequently misarticulated. The values obtained for each consonant denote the frequency of misarticulations (out of 64 opportunities fOr error) made on each error category, i.e., substitutions, omissions, and distortions. The column classified as "No Response" refers to instances when a sub- ject failed to respond to a given stimulus item in the series. The results showed that no phoneme in the inventory is immune to substitu- tion errors. .A closer look at the error frequencies for the various phonemes shows the phoneme classes that were most impaired.by multiple sclerosis. A glaringly Obvious statement that can be made about Table 2 is that the phonemes /37, /z/, and /v/ accounted fOr the high incidence of errors of the substitution type, whereas the phonemes /t/. /k/, and /n/ were the easiest to produce for the majority of multiple sclerosis subjects in the present study. NOne of the 16 consonants contributed in a significant way to the misarticulations that occurred in the error category of omission. Table 2 further illustrates that no errors attribu- table to distortions appeared in the data. Details of misarticulations with respect to the three conventional error categories will be discussed in question 3 of the present investigation. Table 3 shows the frequency of correct production fer each of the 16 consonants. It will be recalled that each consonant had a total of 64 opportunities for error. The percentage of error on each conson- ant, regardless of the type of verbal task and the position of a sound in context, is also given in Table 3. 50 TABLE 2 FREQUENCY OF PHONEMES MHSARTICULATED BY 16 MULTIPLE SCLEROSIS SUBJECTS TABULATED ACCORDING TO CONVENTIONAL ERROR CATEGORIES Phonemes Substitutions Omissions No Response Total Errors 3 27 1 o 28 z 25 O 0 25 v 20 2 0 22 b 15 4 0 19 g 13 5 0 18 9 13 2 1 16 p 12 2 l 15 d 12 2 0 14 f 12 3 O 15 m 12 0 1 13 d3 12 O O 12 tf ll 0 O 11 s 8 0 0 8 k 5 1 1 n S 1 O 6 51 TABLE 3 FREQUENCY OF CORRECT PRODUCTION AND PERCENTAGE OF ERROR ON 16 PHONEMES OBTAINED FROM 16 MULTIPLE SCLEROSIS SUBJECTS Phonemes Frequency of Correct Percentage Error Production n 58 9.3% k 57 10.9 t 55 14.0 s 56 12.5 tf 53 17.1 d3 52 18.7 m 51 20.3 d 50 21.8 p 49 23.4 f 49 23.4 9 48 25.0 g 46 28.1 b 45 29.6 v 42 34.3 2 39 39.0 3 36 43.7 52 The percentage correct production for the 16 phonemes regardless of the type of task and the position of the phoneme in context was 76.8%. In Table 3, the frequency of correct production of 64 observations ranged between 58 for /n/ and 36 for /j'/. Table 4 summarizes the perfOrmance of the 16 multiple sclerosis subjects on a total of 1024 productions. TABLE 4 FREQUENCY COUNT OF PHONEME PRODUCTIONS BY 16 MULTIPLE SCLEROSIS SUBJECTS Type of Production Frequency Percentage Correct 786 76.8% Substitution 204 19.9 Omission 28 2.7 NO Response 6 .6 The present investigator was interested in determining whether the subjects in this study demonstrated a wide range of difficulty in their ability to produce singleton consonants. The type of verbal task and the position of a phoneme in context were disregarded in this analysis. Table 5 shows that there was a great deal of variability in.the group's error productions. This table presents the frequency error count attri- butable to each subject elicited from 16 singleton consonants. The con- ventional categories of errors (i.e., substitutions, omissions) and no response constituted the data used to quantify the types of errors that occurred. The values to the extreme right in Table 5 indicate the percentage errors attributable to each subject. The percentages have 53 been extrapolated from each subject's confusion matrix. Inspection of Table 5 reveals that sUbject G.T. had the most severe articulatory im- pairment of all subjects, contributing a total of 32 errors. Subject C.B., on the other hand, was the least affected. The remaining subjects fell between these two. G.T. showed serious problems of articulation which were consistent with the degree of severity of weakness of his speech musculature. This is in accord with the findings of Darley et a1. (1972) that severity of dysarthria in multiple sclerosis is positively related to severity of neurologic involvement. Analysis of Subphonemic Features It was felt that a gross phonetic feature analysis of the data would highlight the relationship of the misarticulated phonemes and their articulatory features. To this end, a simplifying three-parameter system of classification was employed; and it consists of place of articulation, manner of production, and voicing characteristics. The term subphonemic feature used in the present analysis applies to the particular speech gestures in conventional phonetic terms. The three subphonemic feature system are the articulatory features of the phonemes of a language. They employ a binary system fOr manner of production and voicing features, and a nonrbinary system fer the place of articulation. .A.binary system.has only two specifications, one antithetic in nature to the other. For example, a phoneme is either +voice or -voice, +nasal or -nasal. In a nonebinary system the use of the binary specification (+) or (-) is inconceivable fer the place feature. In the present analysis, the binary system was employed fOr the manner and voicing features, whereas a non-binary system was applied to the 54 TABLE 5 FREQUENCY OF ERROR PRODUCTIONS AND PERCENTAGE OF ERROR OBTAINED FROM 16 MULTIPLE SCLEROSIS SUBJECTS subjects Substitutions Omissions NO Response Cumulative Percentage Error Error G.T. 28 4 0 32 50.0% D.M. 20 2 2 24 37.5 DLM. l8 6 0 24 37.5 J.hL 20 2 1 23 35.9 J.S. 18 0 2 20 31.2 K.B. 16 3 O 19 29.6 B4M. 12 5 0 17 26.5 VTM. 14 l 1 16 25.0 S.L. 7 5 0 12 18.7 J.J. 10 0 O 10 15.6 E.P. 10 0 0 10 15.6 J.E. 10 0 0 10 15.6 D.H. 7 0 0 7 10.6 P.G. 5 O 0 5 7.8 C.M. 5 O 0 5 7.8 C.B. 4 O 0 4 6.2 55 place feature. Only errors of substitution were analyzable in terms of subphonemic features. Percentage scores were obtained for five phonetic categories Of place of articulation, four phonetic cate- gories of manner of production, and two phonetic categories of voicing. The place feature described in the present analysis has five specifications according to the integral parts of the speech mechanism involved in the production of the 16 phonemes. The place feature con- sists of bilabial, labiodental, interdental, alveolar, and velar. Emphasis is placed on the anatomy. Thus, a bilabial involves the use of both lips; labiodental - between the lower lip and upper teeth; inter- dental - tongue between teeth; alveolar - tongue tip articulating against the alveolar ridge; and velar - articulated with the velum, or soft palate. The factor manner of production represents a second criterion of classification. It consists of fOur specifications such as plosive, affricate, fricative, and nasal. IManner of production infers the kinesio- logical behavior which determines the movements of the machinery to accomplish the placement of articulators. Emphasis then, is on the neuromuscular apparatus. Such terms as plosive, affricate, fricative, and nasal suggest movement rather than position or auditory sensation. The voicing feature is a third means of classifying consonants. Each consonant is either voiced or voiceless. A voiced phoneme is pro- duced with the vocal fOlds vibrating, and an unvoiced phoneme is pro- duced with the vocal fOlds abducted. In this analysis, the voiced - voiceless dichotomy was employed. Figure 2 illustrates the matrix of 16 English phonemes included fOr analysis in this study with respect to place of articulation and.manner of production features. The phonemic system outlined in this matrix consists of two axes. 56 Manner of Production PLACE OF ARTICULATION Labio- Inter- Percentage Bilabial dental dental Alveolar velar of Error Plosive p b t d k g 15.4% Affricate tf d3 18.0% Fricative f v 9 J 5 2 27.3% Nasal m. n 13.3% Percent- age of 20.3% 25.0% 31.3% 16.7% 14.1% Error Figure 2. IMatrix of 16 English Singleton Consonants The x-axis represents the place of articulation feature and the y-axis represents the manner of production feature. A third coordinate which is the voicing feature could.not be represented in this matrix, but its relationship to the other two features is clearly outlined in Table 6. Of the total 64 errors of substitution made with respect to the place of articulation feature, the ordering revealed that velar place was the least easily disturbed and that interdental place component showed greater deterioration. Labiodental place was second in the fre- quency of features in error. Figure 3 shows the percentage of errors made on the five place specifications. It is encouraging to note that the place feature alveolar was second to last in the frequency of fea- tures in error. This represents a less serious articulation prOblem be- cause many English consonants are alveolar. 57 TABLE 6 CLASSIFICATION OF PHONEMES BY PLACE OF ARTICULATION, MANNER OF PRO- DUCTION, AND VOICING FEATURES. Phonemes Place Namer Voicing p bilabial plosive — b bilabial plosive + m bilabial nasal + f labiodental fricative — v labiodental fricative + 9 interdental fricative — é interdental fricative + t alveolar plosive — d alveolar plosive + tf alveolar affricate — d3 alveolar affricate + s alveolar fricative — z alveolar fricative + n alveolar nasal + k velar plosive - g velar plosive + Note: The designation of a minus (-) indicates the absence of vocal fold vibration and plus (+) indicates the presence of vocal fold vibration. Percent of Errors ")0 30 20 10 S8 3L3 2&0 20.3 '6" 14.1 Subphonemic Place Feature Figure 3. Percentage Error Rates According to Place Features Interdental H II Labiodental L" ll Bilabial if? Alveolar v: velar 59 Reviewing the results according to substitution made with respect to the manner of production, it was found that fricatives pOsed the greatest difficulty in their production. The error percentage for manner, as shown in Figure 4, ranged from 27.3% for fricatives to 13.3% fer nasals. .Affricates were the next highest in percentage error. It is not unreasonable to expect a high error rate for fricatives and affricates. These two classes of speech sounds are among the most im- paired class of sounds and are among the most difficult to perceive correctly. The manner of production of fricatives and affricates makes them much more susceptible to articulation errors. Fricatives are pro- duced by constriction of the articulators to allow fOr a narrow opening through which air is forced under pressure. The result is a friction- like noise. Sounds in this class require the use of more muscles and closer control of the amount and timing of movement than fOr any classes of speech sounds (Shankweiler and Harris, 1966). Affricates combine a plosive and fricative noise. Of the sounds in the fricative class, the phonemes Ad/ and /2/ were the most impaired. These two fricatives, as most speech pathologists are well aware, are frequently disturbed by articulation disorders of peripheral or central origin. Forthermore, they are among the last to be added to a child's repertoire of speech sounds (Templin, 1957). The high incidence of fricative errors is consistent with the findings of Shankweiler and Harris (1966) and Luchsinger and Arnold (1965) that fricative errors represent a consider- able number of speech disorders. Of the total 97 voicing errors made, 80 (13.9%) involved the sub- stitution of voiceless fOr voiced sounds and 17 (3.8%) involved the substitution of voiced for voiceless sounds. Figure 5 indicates that there were more errors of the voiceless for voiced constrast than vice versa. Percent of Errors 60 ”"1 30 27.3 2" 1&0 ”L4 13.3 '0 F A P N Subphonemic Manner Feature Figure 4. Percentage Error Rates According to Manner Features F= Fricative if Affricate "'U ll Plosive Z I! Nasal Percent of Errors 61 ”T x 30 - 20 - 13.9 10 3.8 .. 1 + Voice — Voice Subphonemic veicing Feature Figure 5. Percentage Error Rates According to VOicing Features + veice: Substitution of the negative value (—) for the positive value (+) —-VOice: substitution of the positive value (+) fOr the negative value (—) 62 The phonemes /3/ and /z/ again, contributed to the high incidence of errors made on voicing feature. This finding is in agreement with many studies in the areas of phonological acquisition and articulation prob- lems which show that when there is a voicing error, there is a greater probability that - voice will replace + voice rather than vice versa (Singh and Frank, 1972). In addition, developmental data show that + voice feature is generally acquired later than —-voice feature (Templin, 1957). Some measure of the degree of similarity between phonemes was of interest to the present investigator. Thus, further analysis of misarticulations was done to determine the extent to which the response phonemes approximated the target phonemes. Specifically, the relation- ship of misarticulated phonemes and their targets was closely inspected with respect to the physiological, articulatory movements involved in speech production. The errors were classified according to the proximity of the response phonemes to their targets with respect to place of articulation, manner of production, and voicing characteristics. Only errors of substitution were considered in this analysis. ‘With the sub- phonemic feature system it was possible to examine the distinctive fea- ture spread between phonemes. For example, the substitution of /t/ fer /p/ is an error of one feature (place) only; /d/ fOr /ds/ is an error of one feature (manner) only; /v/ fOr /f/ is an error of one feature (voice) only; /v/ for /b/ is a two-feature error (place and manner); /t/ for /g/ is a two-feature error (place and voice); /5/ fOr /d/ differs by two features (manner and voice); and finally, /b/ fOr /f/ is a three-feature error (place, manner, and voice). 63 Of the total 204 substitution errors noted, 125 (61.2%) involved the change of one-feature value, 44 (21.5%) involved the change of two features, 5 (2.4%) involved the change of three features; and 30 (14.7%) involved other substitution errors such as /st/ for /z/; /3/ fOr /db/. Thus, it appears that in the great majority of cases, one-feature errors were the most common. Substitutions were feund to be in close proximity to the target phonemes with a minimum of feature changes. The apparent relatedness of many of the substituted sounds to their targets, together ‘with the consistency of the substitutions, is consistent with the pre- dictability of articulation defects in adult dysarthrics (JOhns and Darley, 1970). Figure 6 shows the percentage error made on the sub- phonemic feature distance. The group of multiple sclerosis subjects, as a whole, were rarely Off-target as to produce three-feature errors. On the basis of these findings, conclusions can be drawn that the major- ity of multiple sclerosis subjects give responses in close approximation to target phonemes. One-feature substitutions can be considered a smaller magnitude of articulation problem than two-feature or three- feature substitutions. 2. WhiCh distinctive features account fer the misarticulations that occur in the speech production of multiple sclerosis subjects? The primary purpose of the distinctive feature analysis was to determine the distribution of the misarticulated consonants among dis- tinctive features. Distinctive features are those attributes that dis- tinguish one phoneme from.another or discriminate a large number of phonemes. Distinctive features provide an economic analysis of errors and.make possible the number of generalizations regarding articulatory behaviors of multiple sclerosis subjects used.in the present investiga- tion that would not be possible in the framework of phonetic analysis. 64 Given that multiple sclerosis has differential effects on in- dividuals, one would expect certain features to be affected more than others. The distinctive feature system employed in the present analysis was modification of the Miller and Nicely (1955) five-feature system. The distinctive features consist of voicing, duration, affrication, place of articulation, and nasality. .All of these, except place of articulation, are binary. Miller and Nicely (1955) proposed a ternary feature fOr place of articulation. Four of their distinctive features are described in terms of articulatory features and one in terms of acoustic feature. The articulatory features are voicing, affrication, place, and nasality; the acoustic feature is duration. The distinctive feature system used in the present investigation is by no means an exhaustive and precise list of the features necessary to describe completely all of the phonemes in the English language. Rather, Miller and Nicely (1955) designed this system to aid in de- scribing the role of distinctive features in the act of production and/or perception of selected English consonants. The rationale fOr selecting this feature system.over other feature systems was purely arbitrary. The present investigator felt that the Miller and Nicely (1955) distinctive feature system is explicit in the description of articulatory behaviors in the most simple and straight- fOrward.manner. The nomenclature used in other distinctive feature systems are often subjective and imprecise. In employing the Miller and Nicely (1955) feature system, the present investigator accepted the judgments of the authors who devised it in determining the presence or absence of the features. Percent of Errors 100 In- 70 60 50 40 30 20 10 65 61.2 P- 2L5 ,. 2¢4 .IL_ l 2 3 Subphonemic Feature Distance from Target Phonemes Figure 6. Subphonemic feature distance of 204 errors Of substitution made by 16 multiple sclerosis subjects 1 - One-feature distance 2 - TWo-feature distance 3 - Three-feature distance 66 Analysis of the veicingyFeature The feature voicing is an articulatory feature and implies the action of the vocal folds in production. The distribution of articula- tory errors was obtained from.the pooled errors on voicing regardless of the type of verbal task and the position of the phoneme in context. It must be recalled that the voicing feature is described in terms of a binary system of plus (+) voice, and minus (-) voice. Table 7 shows the distribution of substitution errors of -voice fOr +voice. This table illustrates instances where a phoneme should have been but was not produced with vocal fo1d vibration. The percentage errors contributed by each phoneme is presented to the right of the frequency table. Table 8 shows instances of the substitution of +voice for -voice. In other words, the phonemes presented in this table should not have been but were produced with vocal fOld vibration. The percentage of errors for each phoneme is also presented. When Tables 7 and 8 are compared, it can.be seen that a greater number of errors occurred on the substitution of ~voice for +voice than the reverse. These results would be expected because the substitution of + voice fOr -voice would involve additional effOrt than vice versa. In the literature, analysis of children's articu- lation errors rarely reveals substitution of +voice for -voice. .A close look at Table 7 reveals that the majority of the substitution errors occurred during the production of fricatives /z/ and A3/ and affricate /d3/. Furthermore, this result is consistent with the notion of the markedness theory (Chomsky and Halle, 1968; Cairns, 1969). It is gen- erally believed that the plus (+) specification of a phoneme is a marked feature specification. Presence of'markedness implies greater articulatory complexities. Thus, in Table 7, it can be seen that the 67 TABLE 7 FREQUENCY OF SUBSTITUTION OF THE FEATURE -VOICE FOR +VOICE BASED ON 64 OBSERVATIONS FOR EACH PHONEME Substitution Frequency Percentage - v for + v Error Error 2 23 35.9% 3 16 25.0 d; 10 15.6 g 9 14.1 d 8 12.5 b 6 9.4 v 6 9.4 m 1 1.6 n ___1__ 1.6 80 TABLE 8 FREQJENCY OF SUBSTITUTION OF THE FEATURE +VOICE FOR -VOICE BASED ON 64 OBSERVATIONS FOR EACH PHONEME Substitution Frequency Percentage + v for - v Error Error f 4 6.3% 9 4 6.3 k 3 4 . 7 p 2 3.1 t 2 3.1 tf l 1.6 s __i_ 1.6 17 68 voiced phonemes /z/, /3V, and /d3/ were replaced more readily by -voice. The highest error rates occurred for /2/ which is marked for voicing and.was substituted by its unmarked counterpart /S/ in 20 of 64 (31.3%) observations; /37 is marked for voicing and.was substituted by its un- marked cognate /9/ in 14 of 64 (21.8%) productions. It is clear in the analysis of the voicing feature that the minus (-) specification of the feature consistently showed better articulatory performance than the plus (+) Specification. The greater magnitude of the substitution errors of voiced sounds lends support to the concept of the markedness theory. Analysis of Duration Feature The feature duration is primarily an acoustic feature of some fricative consonants and it implies the ability to prolong certain con- sonants more than others. Of the 16 consonants used in the present study, only two: /s/ and /2/ have a greater amount of duration than other consonants. The binary system of plus and minus was used in the analysis of duration feature. The data presented in Tables 9 and 10 Show instances of the substitution of - duration fOr + duration and the substitution of + duration for - duration respectively. .All other consonants that were not produced with the feature :_duration'were excluded from Tables 9 and 10. Each frequency value indicates the number of times (out of 64 opportunities fer error) a particular consonant was produced in error with respect to the feature duration. Of 22 total errors of duration, 7 involved the substitution of - duration fOr + duration, and 15 involved the substitution of + duration fOr - duration. 69 TABLE 9 FREQUENCY OF SUBSTITUTION OF THE FEATURE —-DURATION FOR + DURATION BASED ON 64 OBSERVATIONS FOR EACH PHONEME substitution Frequency Percentage -D for + D Error Error 5 4 6.3% 2 _;3_ 4.7 7 TABLE 10 FREQUENCY OF SUBSTITUTION OF THE FEATURE + DURATION FOR —-DURATION BASED ON 64 OBSERVATIONS FOR EACH PHONEME Substitution Frequency Percentage + D for —-D Error Error tf 8 12.5% d 2 3.1 9 2 3.1 b 1 1.6 do 1 1.6 f 1 1.6 15 70 Analysis of Affrication Feature The feature affrication is an articulatory feature which involves a partial closure between a given articulator and the point of articu- lation. Because the closure is incomplete, a turbulence is heard during the production of consonants designated as having the feature affrica- tion. Eight consonants /£/, /v/, /9/, /s/, /z/, /3/, /tf/, and /d;;/ are produced with affrication. Tables 11 and 12 Show how the errors of affrication distributed themselves among certain consonants. It can be observed that there was a greater frequency of the replacement of —-affrication fOr + affrication than + affrication for -affrication. substitutions of —-affrication fOr + affrication occurred because the breathstream was completely impeded during the production of sounds having the feature + affrication. Conversely, substitutions of + affri- cation for — affrication occurred because the breathstream was not com- pletely impeded during production of sounds with the minus specification of affrication. Close inspection of the data reveals that the majority of the sUbstitution of -affrication fer + affrication occurred in /v/, /J/. and /9/. Analysis of Place of Articulation Feature The place of articulation refers to the point or area of the oral cavity where constriction is made during production of sounds. As has been mentioned before, Miller and Nicely (1955) proposed a ternary feature fOr place of articulation. The authors designated a score of zero (0) if the constriction.was made in the front of mouth, such as in the production of /p/, /b/, /f/, /v/, and /m/; a score of one (1) if the constriction was in the middle of the oral cavity such as in /t/, 71 TABLE 11 FREQUENCY OF SUBSTITUTION OF THE FEATURE — AFFRICATION FOR + AFFRI CATI ON BASED ON 64 OBSERVATIONS FOR BAG-I PHONEME Substitution Frequency Percentage — A for + A Error Error v 15 23.4% d 9 14.1 9 8 12.5 f 5 7.8 2 3 4.7 tf 2 3.1 d; l 1.6 s _l_ 1.6 44 TABLE 12 FREQUENCY OF SUBSTITUTION OF THE FEATURE + AFFRICATION FOR — AFFRICA- TION BASED ON 64 OBSERVATIONS FOR EACH PHONEME Substitution Frequency Percentage + A for — A Error Error g 6 9.4% d 3 4 . 7 b 2 3.1 t 1 1.6 k l 1.6 72 TABLE 13 FREQUENCY OF SUBSTITUTION OF THE FEATURE PLACE OF ARTICULATION BASED ON 64 OBSERVATIONS FOR EACH PHONEME Feiifggs Phonemes Error Rate for 6 Place Feature Specifications" 1_/9 212 0/_1 2/_1 1/2 12. p l 10 - - - - b 2 3 - - - - 55222 f 6 - - - - - v - - _ - - - m 8 - - - - - t - - - l - - d - - 2 l - - 9 - - l 2 - - Middle Place 3' - - 4 - - - s - - 1 6 - - z - - - 2 - - n - - 3 2 - - k — - - - - 1 Back g ' ' - ' 4 1 Place t f _ _ _ _ 2 _ d5 _'__ _;_ ; _:_ L _‘_ 17 13 ll 14 7 2 * 1/0: Middle for Front (5.3%) 2/1: Back for Middle (3.1%) 2/0: Back for Front (4.1%) 1/2: Middle for Back (2.7%) 0/1: Front for Middle (2.5%) 0/2: Front for Back ( .8%) 73 /d/, /9/, /37, /s/, /z/, and /n/; and a score of two (2) fOr sounds made with the constriction in the back of the oral cavity such as in /k/, /g/, /tf/, and /d3/. In Table 13, the phonemes have been examined in terms of the area of the oral cavity constricted during production. In this analysis, six combinations of substitution errors were possible: substitution of middle for front of the oral cavity (l/O); back for front (2/0); front for middle (O/l); back fOr middle (2/1); middle for back (1/2); and front fOr back (0/2). Table 13 shows the frequency of substitution errors made in each place Specification. The data revealed that there was a greater tendency for the middle place specification to substitute the front place (5.3%) than front fOr middle (2.5%). .Also, there was a greater incidence of the substitution of the back place for the front place (4.1%) than the front place fer the back place (.8%) and a greater incidence of the substitution of the back place fer the middle place (3.1%) than the middle place for the back place (2.7%). Although the place feature substitutions showed a non-unidirectional tendency, there was a greater incidence of the substitution of the less fronted place for the more fronted place in the same manner series. For example, /k/ is produced with constriction in the back of the oral cavity and replaced the fronted /p/ 9 of 64 opportunities (14.1%) for error. There was hardly any substitution of /p/ fOr /k/ in the data. This strange phenomenon of the substitution of the back /k/ for the front /p/ may be attributable to the distinctive feature similarities between /k/ and /p/. .A phoneme is usually replaced by another it is similar to in terms of distinctive features. The phonemes /k/ and /p/ differ only by one feature: place of articulation. Both phonemes involve considerable aspiration in their production, and thus the 74 acoustic nature of these two phonemes may be reflected in the diffi- culty in differentiating them. The substitution of /k/ fOr /p/ was highly unidirectional and occurred in the imitative task only. (See confusion matrix, Figure 11, Appendix F). The place specification middle for front (as exemplified by the substitution of /G/ for /f/ and /n/ for /m/) was another instance in which there was a greater sub- stitution Of the less fronted place for the more fronted place. The acoustic difference between /9/ and /f/ are among the most difficult to distinguish (Nfiller and Nicely, 1955), and it is not surprising that they were confused with regard to the place feature. This finding of the greater incidence of the substitution of the less fronted place fer the more fronted place is not in agreement with Observations of Singh and Frank (1972) whose analysis of consonant articulation problems in children showed a tendency for the substitution of the more fronted place for the less fronted place. For example, a substitution of labial fOr alveolar, alveolar fOr back, interdental fer alveolar. Analysis of Nasalitnyeature TwO consonants /m/ and /n/ are produced by nasal resonance. Nasality, as an articulatory feature, employs the binary system.of plus and.minus. Neuromuscular disturbances, identified as the dysarthrias of the speech.mechanism, can affect the velopharyngeal closure through paresis or paralysis of the musculature or a lack of coordination with other speech movements. Nasality results from this reduced or lack of velopharyngeal control. The data presented in Tables 14 and 15 represent the frequency with which nasal consonants were produced without nasal resonance and non-nasal consonants were produced with nasal resonance, 75 TABLE 14 FREQUENCY OF SUBSTITUTION OF THE FEATURE -NASALITY FOR + NASALITY BASED ON 64 OBSERVATIONS FOR EACH PHONEME substitution Frequency Percentage —-N for + N Error Error m 4 6.3% n _iy_ 1.5 5 TABLE 15 FREQUENCY OF SUBSTITUTION OF THE FEATURE + NASALITY FOR -NASALITY BASED ON 64 OBSERVATIONS FOR EACH PHONEME Substitution Frequency Percentage + N fOr —-N Error Error b 3 4.7% v 2 3.1 d 2 3.1 76 respectively. Errors on the nasal consonants occurred because the velopharyngeal port was occluded during production of /m/ and /n/, -whereas errors on non-nasal consonants occurred because the velopharyn- geal port was open during their production. Tables 14 and 15 Show that there was slightly more substitution of + nasality for — nasality than vice versa. This finding is consistent with the notion that dysarthria affects the production of non-nasal sounds because adequate velopharyn- geal closure is compromised by muscle weakness or incoordination. In addition, the small error rate fOr nasality is consistent with the findings of Darley et a1. (1972) that resonatory prOblems are low on the list of prOblems affecting speech processes in multiple sclerosis. Summary of Distinctive Feature Analysis In general, the results of the distinctive feature analysis showed that the substitution errors that occurred for the features voicing and affrication would be predicted from the markedness principle. It will be recalled that the unmarked specification of a feature involves less complex articulatory gestures than its marked counterpart. Thus, for the feature voicing, it is not surprising that the feature —-voice (un- marked) replaced + voice (marked) more freqUently than vice versa. The marked feature for voicing calls fOr an additional phonetic gesture, i.e., vocal fOld vibration. The markedness principle also obtained fer the articulatory feature affrication, The feature nasality showed slightly more errors of the substitution of + nasality for —-nasality. The acoustic feature duration also showed more errors of the substitution of the + specification fer the -specification. ‘With regard to the place feature specifications, the markedness principle did not hold true. 77 TABLE 16 OVERALL DISTINCTTVE FEATURE ERRORS MADE BY 16 MULTIPLE SCLEROSIS SUBJECTS (TASK AND POSITION COMBINED) Features veicing —-vo1ce + veice Duration -Duration + Duration Affrication —-Affrication + Affrication Place of Articulation l fOr 0 2 fOr 0 0 for l 2 fOr l l for 2 0 for 2 Nasality -Nasality + Nasality Number Presented 448 576 896 128 512 512 320 448 256 896 128 Error Rate 17 80 15 13 44 17 13 11 14 Percentage Error LN (I) o\° 13.9 78 There is evidence, both in language developmental data and in the patterns of articulation errors of children, that the place feature front of the oral cavity which involves fewer maneuvers of the articula- tors are less frequently produced in error than the place feature fOr which there is a complex articulatory maneuver. Singh and Frank (1972) observed that place 3 (back of the oral cavity) is substituted more often by place 2 (middle of the oral cavity) than vice versa and that place 2 is substituted more often by place 1 (front of the oral cavity) than vice versa. Thus, in terms of the markedness theory, place 1 is unmarked and place 2 may be considered unmarked when compared to place 3 which is considered marked. Similar statements cannot be made regarding the findings of the place feature specifications in the present analysis and the markedness theory. Table 16 presents the overall distinctive feature errors made by the group. 3. Which is the predominant type of error (i.e., substitutions, omissions, and distortions) made by multiple sclerosis patients? The present research question examines the distribution of the mis- articulated consonants among the conventional error categories. Gen- erally, phonemic errors are described as being of four types: substi- tutions, omissions, distortions, and additions. The error category addition was excluded for consideration in the present study because of its rarity of occurrence as an articulation error (Byrne and Shervanian, 1977). Addition involves adding another phoneme to the target phoneme(s). According to van Riper and Irwin (1958), addition, as an articulatory error, is much more common in aphasics and foreign language speakers. Furthermore, since the errors of additions were non-existent in the present data, its elimination was justified. substitutions are the result of a process whereby one phoneme from the store of phonemes in 79 the language is substituted for the target phoneme. Comparisons of the categories of articulation errors on the two types of verbal tasks and the two positions of a sound in context constituted the data in the present analysis. It was felt that pooling the data from.the two tasks and the two positions might overlook real differences in the misarticulations that occurred as a function of task and/or position. Of the total 1024 phoneme observations, 238 (23.2%) were misarticu- lated by the group of multiple sclerosis subjects. Contrary to expec- tations, the errors did not distribute themselves among the three con- ventional error categories of substitutions, omissions, and distortions. Rather, substitution category accounted for a significant number (19.9%) of errors made by the multiple sclerosis group. Articulation errors of the omission type (2.7% error) were rarely the basis fOr articulation errors in the present study. The surprising finding in the data was the total absence of distortion errors. The substitution of target phonemes by other phonemes which were not selected fOr Observation in the present investigation constituted the category classified as "Other substitutions" in the confusion.matrices (Appendix F). Errors due to other substitutions (a total of 30 out of 1024 Observations) were combined with the conventional error of substitutions to yield a total of 204 substitution errors made by the multiple sclerosis subjects. An example of "Other Substitution" errors was the replacement of /f/ fOr /s/. It will be recalled that the phoneme /f/ was excluded from the present study because its voiced cognate /3/ does not occur in the initial-word.position in English. Errors due to "No Response" involved instances when a subject failed to respond to a given stimulus and this 8O constituted .6% of misarticulations. Table 17 shows the cumulative errors and.percentage distribution of errors made on the 16 consonants. The percentages are based on the total of 238 errors made. Of interest to the present investigator was the distribution of the categories of misarticulations with respect to the type of verbal task and the position of a sound in context. Table 18 shows how the articulation errors distributed themselves among the 16 consonants as a fonction of task. It is clear that there was a greater incidence of substitution errors in the imitative task (23.6%) than in the spon- taneous task (16.2%). Table 19 shows the distribution of the types of articulation errors made on the 16 consonants as a function of the position of a sound in context. It is evident that substitution errors accounted for more (22.5%) of the total errors made in the final-word position than did the initialeword position (17.4%). Furthermore, more misarticulations for the omission type occurred in the final-word position (5.1%) as compared to omission errors in the initial—word position (.4%). TABLE 17 (IhflHAflTVE ERRORS.AND PERCENTAGE DISTRIBUTION OF THE CATEGORIES OF ERRORS MADE ON 16 CONSONANTS Error Category CUmulative Error Percentage Error substitution 204 19.9% Omission 28 2.7 No Response 6 .6 81 The most striking result to emerge from the present analysis was the significantly greater incidence of errors of the substitution type. Since multiple sclerosis, as a disease process, affects the articulators, the results might be interpreted to mean that sUbstitution errors are more representative of the basic physiological impairment in the group of multiple sclerosis subjects used in the present investigation. 'This result is not in agreement with the observations of Johns and Darley (1970) that distortion of consonants is most characteristic of the speech of dysarthric subjects. It is interesting to note that the clustering of errors in the substitution category was attributable to only a few subjects. The range was between a total of 28 substitution errors (sUbject G.T.) and 4 (subject C.B.). Table 5, page 54, shows that three subjects accounted for the high error rate of substitutions, making 20 or more errors of the sUbstitution type. There are certain considerations in examining the patterns of articulation errors that occurred- The basis fOr the differing re- sults may be attributable to the different tenminologies involved in the present study and other investigations. A.possib1e explanation as to why the distribution of articulation errors was highly skewed toward the substitution type could be due to the concept of distortion in the speech pathology literature. The definition of distortions is arbitrary depending on the experimenter. Imprecise production of con— sonants constitute distortion in certain research investigations (Darley et a1., 1969; Johns and Darley, 1970). This arbitrary classi- ficatory ternloften leads to different concepts of distortion. In the present investigation, the term.distortion was defined in the strict sense of the replacement of a standard speech sound by one not normally 82 TABLE 18 THE DISTRIBUTION OF THE CATEGORIES OF ARTICULATION ERRORS MADE ON 16 CONSONANTS BASED ON THE TYPE OF VERBAL TASK* Error Category Imitative Task Spontaneous Task Substitution Omission No Response 121 (23.6%) 14 c 2.7%) __g ( 1.2%) 141 83 (16.2%) 14 ( 2.7%) _9_( 0.0%) 97 * Each type of verbal task involved a total of 512 Observations. TABLE 19 THE DISTRIBUTION OF THE CATEGORIES OF ARTICULATION! ERRORS MADE ON 16 CONSONANTS BASED ON THE POSITION OF.A SOUND IN CONTEXT* Error Category Initial Position Final Position substitution Omission No Response 89 (17.4%) 2 ( .4%) _g_( .4%) 93 115 (22.5%) 26 ( 5.1%) 4 ( .8%) 145 * Each type of position involved a total of 512 observations. 83 used in a given language. Furthermore, substitution, as operationally defined in this study, had to be the replacement of the target phoneme by another well-articulated standard English phoneme. Thus, the sub- stitution of an implosive /6/ for /b/ would be considered a distorted /b/ sound since implosive /Q/ is non-existent as a standard sound in English. .A second possible reason for the absence of distortion errors in the data might rest with the criterion of acceptance of judging a re- sponse phoneme as correct or incorrect. Although there were isolated cases in which a judge or two scored certain phonemes as distorted (for a total of 11 distortions), these scores were disregarded because there was no agreement by 4 of the 6 judges with respect to a given phoneme. Another possible reason for the patterns of articulation errors that occurred.might be the interaction of task and/or position of a sound in context. This would also lend itself to an explanation as to why there were more substitution errors in the final-word versus initial-word position and in the imitative task versus the spontaneous task. It must be pointed out that only a selected group of consonants were tested in this study. The total absence of distortion errors in the data could be that distortions mdght not be reflected in a monosyllabic CVC context. The presentation of consonants in mono- syllabic initialsword and final-word positions is rather contrived such that the majority of the multiple sclerosis subjects produced well- articulated substitutions of other standard sounds fOr the target con- 5011311125 . 84 One interesting finding to emerge from these data was the demonstration by one subject (E.P.) of the articulatory behavior somewhat similar to what Charcot (1877) described as "scanning speech." Scanning involves increased stress on usually unstressed words or syllables. Darley et a1. (1972) refer to this articulatory speech dimension as impaired emphasis. Whether this subject (E.P.) displayed this articulatory behavior is questionable. The present investigator is of the opinion that the ”exaggerated” precise consonant articulation by this particular subject was perhaps a compensatory technique to allow for acceptable acoustic impressions. A possible therapeutic implication of this articulatory behavior is that scanning could be used to improve the speech of multiple sclerosis patients whose speech is "unscanning" and thus unintelligible. 4. Is there a significant difference in the mdsarticulations that occur as a function of the type of verbal task? The analysis of the misarticulations as a function of verbal task was designed to determine the effects of imitative (repetition of words) and spontaneous (reading of words) methods of response elicitation on the patterns of errors that occurred for the 16 multiple sclerosis subjects. In other words, under what type of treatment condition were subjects more apt to make more articulation errors? Table 20 compares the subjects' performances on the two types of verbal tasks. The results revealed that the group performed better in the spontaneous task than in the imitative task. This held for both initial- and.final-word.positions. The results also revealed that the type of verbal task was a factor in the categories of articulation errors that occurred. Fer instance, there was a greater incidence of 85 TABLE 20 A (DMPARISON OF THE PERFORMANCE ON TWO TYPES OF VERBAL TASKS BY 16 MILTIPLE SCLEROSIS SUBJECTS Subjects Imitative Errors Spontaneous Errors G.T. 17 15 D.M. 15 9 D.M. l6 8 J.M. 12 ll J.S. l3 7 K.B. 8 ll B.M. 8 9 V.M. 13 3 S.L. 7 5 J.J. 5 5 E.P. 7 3 J.E. 7 3 D.H. 5 2 P.G. 4 1 C.M. 1 4 C.B. 3 l 141 97 86 errors of the sUbstitution type in the imitative task, whereas errors of omission showed no difference as a fUnction of task. Tables 21 and 22 show the breakdown of the categories of articu- lation errors made on 16 consonants as a function of task. Although there is no general agreement in the literature concerning the most efficient method of response elicitation, the finding in the present analysis is not in agreement with the popular notion that response elicitation is easier in the imitative than the spontaneous method of presentation. Templin (1947), in an investigation of the influence of imitative and spontaneous methods of stimulus presentation on the articulation of 100 pre-school children, concluded that neither the imitative nor the spontaneous testing method.was superior. Siegel et al. (1963) found the imitative method to result in better articu- latory perfOrmance on at least 8 of the 40 sounds presented to 100 kindergarten children. Kresheck and Socolofsky (1972) found articula- tory responses to be superior in the imitative method. Paynter and Bumpas (1977) investigated the effects of the two methods of stimulus presentation on the articulatory responses of 3 and 3% year old children. They found no significant differences in articulatory re- sponses as a function of type of stimulus presentation. To the knowledge of the present investigator, no study has been done regarding the effects of the two methods of stimulus presentations on the articulatory responses of adults with articulation prOblems. Further analysis revealed that the group's articulation errors on individual phonemes kept essentially the same order of difficulty regardless of the type of verbal task. .A comparison was also made of the groups' responses in the initial and final positions fOr the 87 TABLE 21 TYPES OF MISARTICULATIONS AS A FUNCTION OF IMITATIVE TASK BASED ON 32 OBSERVATIONS FOR EACH PHONEME Phonemes Substitutions Onissions No Response Total Error z 15 - - 15 3' 14 — - 14 p 11 l 1 13 b 10 2 - 12 g 9 2 - 11 d 8 - - 8 d3 8 - - 8 v 8 l - 9 f 7 2 - 9 9 7 2 1 10 tf 6 - - 6 m 5 - l 6 n 5 1 - 6 s 4 - - 4 k 3 1 1 5 t l 2 2 S 121 14 6 141 88 TABLE 22 TYPES OF MISARTICULATIONS AS A FUNCTION OF SPONTANEOUS TASK BASED ON 32 OBSERVATIONS FOR EACH PHONEME Phonemes Substitutions Omissions No Response Total Error 3 13 1 - 14 v 12 l - 13 z 10 - - 10 m 7 - - 7 9 6 - - 6 b S 2 - 7 tf 5 - - S f 5 l - 6 d 4 2 - 6 g 4 3 - 7 d3 4 - - 4 s 4 - - 4 k 2 - - 2 p l l - 2 t l 3 - 4 n - _ _ - 83 14 - 97 89 imitative and spontaneous tasks. The results showed that the sub- jects performed better in the initial position fOr both tasks than in the final position. Only three subjects (G.T., J.J., and C.B.) deviated slightly from the above pattern, their responses being slightly better in the final position than in the initial. In terms of distinctive features, the subjects' responses were essentially the same for both the imitative and spontaneous tasks. Table 23 illustrates the misarticulations that occurred with respect to the five distinctive features and.mode of stimulus presentation. The number of observations, the cumulative error, and the percentage of error for each feature specification of a given distinctive feature is presented. Table 23 shows that the feature place of articulation deviated slightly from the patterns of errors that occurred as a function of verbal task. Subjects' responses on the imitative place feature were poorer (with a total of 48 place errors) than on the spontaneous place feature (with a total of 16 place errors). The diagnostic and therapeutic implications of the findings of the present analysis are that the multiple sclerosis subjects in this study might benefit more from visual (spontaneous) model than pure auditory (imitative) model. 5. Is there a significant difference in the misarticulations that occur as a function of the position of a sound in context? The analysis of the misarticulations as a function of phoneme position was designed to compare the errors made by the 16 multiple sclerosis subjects when a consonant appeared in the initial versus the final position of a word. Table 24 shows the frequency of errors made by the group as a function of phoneme position. The errors made in the initial-word position for the two types of verbal tasks 90 TABLE 23 DISTINCTIVE FEATURE ERRORS MADE AS A FUNCTION OF TYPE OF VERBAL TASK Number Imitative Percentage Spontaneous Percentage Features Presented Error Rate Error Error Rate Error VOicing -Vbice 224 6 2.7% 11 4.9% + Vbice 288 44 15.3% 36 12.5% .Duration —-Duration 448 11 2.5% 4 9% + Duration 64 5 7.8% 2 3.1% Affrication -Affric. 256 9 3.5% 4 1.6% +.Affric. 256 26 10.2% 18 7.0% Place of Articulation 9. O 1 fOr 0 160 12 7.50 5 3.1% 2 fer 0 13 8.1% 0 0.0% 0 for l 224 6 2 7 5 2 2% 2 for 1 10 4.5% 4 1.8% 0, 9 1 for 2 128 5 3 90 2 1 60 0 for 2 2 1.6% 0 O 0% Nasality -Nasality 448 3 .7% 4 .9% o\° N (N H o\° + Nasality 64 3 4.7 91 (imitative and spontaneous) were pooled for this analysis as were the errors in the final-word position. The data in Table 24 show there were more errors in the final-word position (145) than in the initial-word.position (93). This held fer errors of substitutions and omissions (Tables 25 and 26). The percentage error for the 16 consonants in the initial position was 18.5%, and the percentage error for the final consonants was 28.3%. There were 512 opportunities for error in each type of position. Several points should be stressed here. The first was that there was a high degree of similarity in the misarticulations as a function of position in the two types of verbal tasks. This finding would appear to indicate that the same or similar processes were in operation in the Observed misarticulations. The second point was the surprising finding that errors tended to be greater in the final-word position than initial. The findings in the present analysis are consistent with the linguistic theory of'regression.hypothesis of Hughlings Jackson (Ed. Taylor, 1958), Jakobson and Halle (1956) and of wepman and Jones (1964) that final consonants, appearing later than initial consonants in a child's repertoire, are more difficult. In addition, the findings lend support to reports by Templin (1957) that more errors occur in the finaliword.position. .Although the data in Table 24 showed some consistencies existing across subjects of the greater vulnerability of final-word position, there were some differences. TWO subjects (J.J. and C.B.) deviated from this pattern, contributing to more phoneme errors on the initial-word.position than on the final-word.position. TWO other subjects (G.T. and J.E.) showed essentially no difference in their 92 TABLE 24 POOLED ERROR BY POSITION FOR 16 MULTIPLE SCLEROSIS SUBJECTS Subjects Initial Position Final Position G.T. 16 16 D.M. 10 14 D.M. 10 14 J.M. 7 l6 J.S. 8 12 K.B. 8 ll B.M. 1 l6 V.M. 5 11 S.L. 3 9 J.J. 7 3 E.P. 4 6 J.E. 5 5 D.H. 2 S P.G. 2 3 C.M. 2 3 C.B. 3 l 1 93 145 93 TABLE 25 TYPES OF MISARTICULATIONS.AS.A FUNCTION OF INITIAL-WORD POSITION BASED ON 32 OBSERVATIONS FOR EACH PHONEME Phonemes Substitutions Omissions No Response Total Error 43' 11 - - 11 f 11 - - 11 v 10 - - 10 d 8 - - 8 z 8 - - 8 tf 7 - — 7 p 6 - - 6 g 5 - - 5 9 5 1 l 7 d3 4 - - 4 m 4 - 1 5 k 3 - - 3 s 3 - - 3 b 2 l - 3 n 2 - - 2 t _ - - - 89 2 2 93 94 TABLE 26 TYPES OF MISARTICUATIONS AS.A FUNCTION OF FINAL-WORD POSITION BASED ON 32 OBSERVATIONS FOR EACH PHONEME Phonemes Substitutions Omissions No Response Total Error 2 l7 - - 17 g 16 1 - 17 b 13 3 - 16 v 10 2 - 12 g 8 5 - 13 d3 8 - - 8 9 8 1 - 9 m 8 - - 8 p 6 2 l s 5 - - 5 d 4 2 — 6 tf 4 - - 4 n 3 l - 4 t 2 5 2 9 k 2 1 l 4 f 1 _3_ - .4. 115 26 4 145 95 responses in both positions. Individual consonants were also examined with regard to the difficulty they posed in their production in the initial- and final- word.positions. .A comparison of Tables 25 and 26 demonstrates that the majority of consonants were misarticulated in the final than in the initial-word position. The consonants /b/, /g/, A37, and /z/ ‘were the most impaired in the final-word.position. The consonants /tf/, /d/, and /f/ showed shifts in the reverse of the overall pattern. In other words, there were more errors on these consonants in the initial-word position than in the final. Further analysis was done to determine the distribution of errors among the five distinctive features with respect to the position of a sound in context. Table 27 shows the distinctive feature errors made as a fUnction of phoneme position. Overall subjects' responses showed that the distinctive features were more vulnerable in the final-word position than initial. Each distinctive feature is ex- amined in terms of the number of observations, the cumulative error, and the percentage of error made on each feature specification as a fUnction of phoneme position. The implications of the findings of the present research question are that, as a group, the multiple sclerosis subjects in this study have more difficulty with sounds in the finaleword position. This could be interpreted to mean that multiple sclerosis subjects have less difficulty with sounds in the initial position, perhaps, because they have more time and energy to set the articulators in a suitable position.prior to production. The difficulty experienced by this group of multiple sclerosis subjects with sounds in the final position 96 TABLE 27 DISTINCTIVE FEATURE ERRORS MADE AS A FUNCTION OF PHONEME POSITION Number Initial Percentage Final Percentage Features Presented Error Rate Error Error Rate Error VOicing -Voice 224 10 4.5% 7 3.1% + Vbice 288 28 9.7% 52 22.8% Duration —-Duration 448 9 2.0% 6 1.3% + Duration 64 4 6.3% 3 4.7% Affrication —-Affric. 256 3 1.2% 10 3.9% + Affric. 256 25 9.8% 19 7.4% Place of.Articulation 9 1 for 0 160 10 6.3% 7 4.40 2 for 0 5 3.1% 8 5.0% 0 for l 224 6 2 7% 5 2 2% 2 fOr l 5 2.2% 9 4.0% l for 2 128 3 2.3% 4 3 1% 0 for 2 0 0.0% 2 1.6% Nasality -Nasa1ity 448 3 .7% 4 .9% + Nasality 64 2 3.1% 3 4.7% 97 might be attributable to the difficult transition they have to make from one articulatory gesture to another. Analysis of variance of subjects' Responses as a Function of Type of verbal Task and Phoneme Position Further analysis was done to determine whether the results were clear-cut enough to demonstrate definite significant differences in the misarticulations as a function of the type of verbal task and the position of a sound in context. .A twosway fixed effects (2 x 2) analysis of variance with repeated measures was computed to determine the differences that existed in the responses of the 16 multiple sclerosis subjects as a fUnction of task (imitative, spontaneous) and position (initial, final). The results (Table 28) indicate that there was a significant main effect fOr task at the .01 level of confidence, and a significant main effect for position at the .05 level of confidence. The mean of errors for the effect of task and the mean of errors fer the effect of position were computed to deterrine the direction of the difference between tasks and the direction of the difference between positions. Figure 7 shows that the group's performance was poorer in the imitative task than in the spontaneous task. Furthermore, Figure 7 shows that the group's per- formance was poorer in the final position than in the initial position. No interaction was feund between task and position. 6. Is the distribution of misarticulations related to the de- velopmental hierarchy of phoneme emergence? The present analysis attempts to determine the order in the break- down of articulation in adult multiple sclerosis subjects with respect to the acquisition hierarchy of sounds in children. The 16 phonemes were tabulated (Table 29) on the basis of the age at which they are acquired. 98 TABLE 28 TWO-WAY FIXED EFFECTS (TASK x POSITION) ANALYSIS OF VARIANCE WITH REPEATED MEASURES FOR RESPONSE ELICITATION BY ONE GRCIIP OF MJLTIPLE SCLEROSIS SUBJECTS Source of Degrees Sums of IMean F-Ratio F-Probability variation of Freedom Squares Square T (Task) 1 30.250 30.250 9.213 .008* Error (TS) 15 49.250 3.283 P (Position) 1 42.250 42.250 8.422 .011** Error (PS) 15 75.250 5.017 Interaction (TP) 1 .063 .063 .033 .858 Error (TPS) 15 28.438 1.896 8 (Subjects) 15 257.438 17.163 Total 63 482.938 * Significant at the .01 level of confidence. ** Significant at the .05 level of confidence. Notes: TS = Task x Subjects PS = Position x subjects TPS = Task x Pesition x Subjects 99 O a I Mean of Errors .1? :40 2!! '31 '22 Figure 7. .Mean of Errors for Task by Position. Notes: 1 Position 1: Initial P P2 Position 2: ~Final T Task 1: Imitative (—) 1 T2 = Task 2: Spontaneous (---) PlTl : 57 Ple : 84 Psz : 36 P T : 61 2 2 100 TABLE 29 DISTRIBUTION OF ERRORS IN RELATION TO THE DEVELOPMENTAL HIERARCHY OF PHONEME EMERGENCE BASED ON RESPONSES BY 16 MULTIPLE SCLEROSIS SUBJECTS Period of Age of Phoneme Error Average Acquisition Acquisition Category Rate Error (years) Early 3-4 /m, n, f. p. 107 13,4 k. b, g. d/ Late 4-7 /S, U» ’0» 9: 131 16.4 m3, 2, ds/ TABLE 30 DISTRIBUTION OF ERRORS IN RELATION TO THE DEVELOPMENTAL HIERARCHY OF PHONEME EMERGENCE BASED ON RESPONSES BY 16 MULTIPLE SCLEROSIS SUBJECTS Period of Age of Phoneme Error Average Acquisition Acquisition Category Rate Error (years) Early 3-4 /m, n, f, p/ 49 12.3 Late 4-7 /v, 5, z, d5/ 87 21.6 101 TABLE 31 ONE-WAY FIXED EFFECTS (AGE OF ACQUISITION) ANALYSIS OF VARIANCE WITH REPEATED MEASURES FOR RESPONSE ELICITATION BY ONE GROUP OF MULTIPLE SCLEROSIS SUBJECTS Source of Degrees Sum of Mean variation of Freedom Squares Square F-Ratio F-Probability A.(Age) 1 18.00 18.00 1.280 .276 Error (AS) 15 211.00 14.07 S (Subjects) 15 514.88 34.33 Total 31 743.88 NOte: AS - Age of Acquisition x Subjects. TABLE 32 ONE-WAY FIXED EFFECTS (AGE OF ACQUISITION) ANALYSIS OF VARIANCE WITH REPEATED MEASURES FOR RESPONSE ELICITATION BY ONE GROUP OF MJLTIPLE SCLEROSIS SUBJECTS Source of Degrees Sum of Mean variation of Freedom. Squares Square F-Ratio F-Probability A.(Age) l 45.13 45.13 5.019 .041* Error (AS) 15 134.88 8.99 8 (Subjects) 15 136.00 9.067 Tbtal 31 316.00 * Significant at the .05 level of confidence. the: .AS - Age of Acquisition x subjects. 102 The error rate and average error for the 8 phonemes in each developmen- tal period are also presented. .A one‘way fixed effects analysis of vari- ance with repeated measures was computed to determine if there were any differences in the errors that occurred with respect to the developmental period. Table 31 indicates no significant difference between the 8 phonemes acquired early and the 8 phonemes acquired late. Hewever, the average errors for the first 4 of the phonemes acquired early and the last 4 of the phonemes acquired late were computed (Table 30). .A one- way fixed effects analysis of variance (Table 32) with repeated measures revealed a significant difference at the .05 level of con- fidence. 7. Is the distribution of misarticulations related to the acquisition hierarchy of distinctive features? Table 33 represents age in years, acquired phonemes, and the present investigator's analysis of the acquisition hierarchy of distinc- tive features. It is well to keep in mind that only 16 consonants were selected fOr Observation in the present study. Thus, the fricative /h/, all glides, and all semivowels not intended fOr distinctive feature analysis have been excluded from the present research question. Any interpretation of distinctive feature acquisitional patterns is limited since only a selected number of English consonants have been considered. Table 33 was based on a combined Templin (1957) distinctive feature system and the Miller and Nicely (1955) distinctive feature system. According to the Templin (1957) phonemic acquisition data, the earliest age at which a phoneme could be considered acquired was three years. However, it is well established that "most phonological learning occurs in the first three years of life" (Berko and Brown, 1960, p. 526). It 103 is therefore assumed that by age three, most children have acquired the contrasts of phonemes that contribute to differences in.meaning. Table 33 shows that the phonemes /mM, /n/, /f/, /p/, /k/, /b/, /g/, and /d/ indicate the acquisition of the features nasality, labiality, voicing, and the feature front/back place of articulation. .At age 4-7, the feature affrication, represented by /s/, /z/, /tf/, /d3/, /9/, /37, and /v/, is added to a child's articulatory repertoire. In addition, the features duration (represented by /s/ and /z/), the feature middle/back place contrast (represented by /t/, /d/, /tf/, and /d3/), and the voiced/voiceless contrast (represented by /s/, /z/, /t/. /d/, /9/. /J/, /t//, and /d3/) are acquired. TABLE 33 DISTRIBUTION OF ERRORS IN RELATION TO THE.ACQUISITION HIERARCHY OF DISTINCTIVE FEATURES BASED ON RESPONSES BY 16 MULTIPLE SCLEROSIS SUBJECTS Period of Age of Phoneme Acquisition Acquisition Category Distinctive Features (years) Early 3-4 /m, n, f, p, Nasality, Labiality, VOicing, Front/Back k, b, g, d/ Place Late 4-7 /s, tf, t, 9, Affrication, Dura- tion, Middle/Back v,3’, 2, ds/ Place, and Voicing An overall examination of Table 33 reveals that the features nasality, front/back place contrast, and voicing are acquired earlier than the features affrication, duration, middle/back place contrast. Since the phonemes in the two periods of acquisition (early, late) 104 share certain features in common, it cannot be ascertained that there is a tendency fOr articulatory breakdown to occur in one period versus another. It is difficult to determine feature acquisition from phonemic acquisition data not originally intended fOr distinctive feature analysis. One conclusion that can be drawn from Table 33 is that no obvious trend could be seen in the breakdown of articulation as a fUnction of acquisition hierarchy of distinctive features. Further fUture research is recommended that might group phonemes with identical distinctive features with respect to acquisition hierarchy. CHAPTER V SUNMARY, CONCLUSIONS, AND RECOVNENDATIONS Sumnag Miltiple sclerosis is a neurological disease which leads to demye- lination of afferent neural fibers. It is usually a diffuse, chronic, slowly progressive neurological disorder which affects predominantly the white matter of the central nervous system but may also involve the gray matter. Demyelination or destruction of the myelin sheath surrounding the nerve fiber causes complete interruption of nerve inpulses, thus producing paralysis to parts of the body supplied by the nerve. The cause of the disease is unknown, as is the explanation for the patchy distribution of plaques throughout the brain and spinal cord. Miltiple sclerosis is dominated by episodes of "attacks." In certain attacks, the course can be of a very dramatic nature. The other side of the clinical picture, namely the quieter progression of the condition, usually escapes the attention of the patient and physician. The symptoms of multiple sclerosis are similar to those caused by any localized lesion of the central nervous system. It is, however, the pattern of its behavior which renders multiple sclerosis unique among the organic disease of the nervous system. One of the unique features of the disease and one of major importance in diagnosis is the manner in which early symptoms tend to clear partially or 105 106 completely, only to return on one or more occasions. The pattern of occurrence suggests that it is very likely caused by a viral infection early in life. But no one has ever isolated a virus that, when injected in animals produces multiple sclerosis. In the past five years much more direct evidence has been Obtained that associates the disease with a virus. Many investigators have fOund traces of the measles virus at different sites in the bodies of multiple sclerosis patients. Some of the results conflict with each other and some are controversial. NOnetheless, these findings promise that a firm identification of the causative agent of the disease may be made in the fOreseeable future. IMultiple sclerosis characteristically affects the young adults between the ages of 20 and 45 years but occasionally starts in the late teens and less frequently in middle age. A.pecu1iar feature of the disease is its predilection fOr persons living in temperate climates. The plaques involving the cerebellum and its brain-stem connections cause a variety of cerebellar signs. Among these are the so-called Charcot triad of symptoms: nystagmus, intention tremor, and scanning speech. The disease disrupts the normal processes of speech such as respiration, phonation, articulation, resonance, and.prosody. It reduces or interferes with the kinesthetic feedback information necessary for delicate motor adjustments such as those involved in phonation. Muscles not receiving proper neural innervation are impaired and eventually, because of disuse, become weaker than the organic impairment justifies. The disuse atrophy can affect the musculatures used for speech. 107 Speech intelligibility is adversely compromised as the progress of the disease affects neural impairment to the muscles used for talking. When the muscles of articulation are involved because of cerebellar incoordination, ataxic dysarthria manifests itself in distortions and substitutions of speech sounds. In mild forms ataxic dysarthria begins as slurring of speech and.the articulation of consonants become particularly difficult. A."scanning" or lalling type of speech has been described as typical of multiple sclerosis. The term "scanning" implies a tendency to accent every syllable of a 'word, giving a sing-song or scanning characteristic to speech. It is believed that scanning is attributable to poor control of rhythm, incoordination of palatal and labial muscles combined with dysarthria of central origin (west et a1., 1968). Literature on speech dysfunctions in multiple sclerosis reveals a dearth of clinically useful findings. Despite the fact that multiple sclerosis has been shown to have adverse effects on the processes of speech, little recognition has been given to multiple sclerosis in the speech pathology literature. More empirical data are needed to ascertain the specific deficits of speech in the disease. Thus, the primary purpose of the present investigation was: 1. to determine the effects of imitative and spontaneous methods of response elicitation on the speech production of patients with multiple sclerosis. 2. to determine thereffects of the position of a phoneme on the articulatory responses of patients with multiple sclerosis. A.purposive sample of 16 patients with multiple sclerosis was selected fOr the present study. The sample consisted of 11 females and 5 males selected on the basis of meeting the criteria fOr the study. 108 Each subject had to pass a bilateral hearing screening test, exhibit two or more articulation errors on an articulation screening test, and speak General American English. Some of the patients resided in nursing homes and medical care facilities. The sample included patients with varying degrees of severity and duration of multiple sclerosis. Since the primary purpose was to describe articulation.probleme in patients with multiple sclerosis at the phonological level, the present investigation was limited to analysis of the production of monosyllabic CVC words. Furthermore, since vowel differences were not one of the concerns of this study, it was assumed that slight differences in vowel articulation were not a significant source of variability in the data. Thus, a list of 64 meaninngl English words was constructed. The list consisted of 16 consonants and two vowels. The rationale fOr using two vowels /o/ and /e/ was to eradicate the phonetic variability of vowels between speakers of the same language. Each consonant phoneme occurred two times in the initial position and two times in the final position. Responses were elicited from.the subjects by employing two types of verbal tasks. In the imitative task the subjects were required to repeat recorded.monosyllabic CVC words presented to them under ear- phones. In the spontaneous task the subjects were instructed to read aloud monosyllabic CVC words printed in one-half letters on a 4" x 6" white card. The responses of each subject were tape-recorded at the time of the testing and were subsequently scored by six trained graduate speech pathology students at Michigan State University. There were 64 re- sponse events fOr each subject. Pooling the 16 subjects, a total of 1024 Observations was obtained for the two types of verbal tasks. The initial imitative, initial spontaneous, final imitative, and final spontaneous 109 utterances constituted the data fer the analyses. A.phonetic inventory and a distinctive feature inventory were employed to group the articu- lation errors in terms of some comnon attributes of each misarticulated sound. Descriptive and inferential statistical procedures were employed in the analyses of the data. The results presented here are general. The findings of this study are presented in detail in Chapter Four. The descriptive statistical analyses provided interpretation of research questions designed: 1. to determine the difficulty that certain consonants pose in their production. Of the 16 consonants observed in this study, some were more difficult than others in the articulatory responses of multiple sclerosis subjects. 2. to determine the misarticulations that occurred in terms of their subphonemic features of place of articulation, manner of produc- tion and.voicing. .Although subphonemic feature errors have been treated as entities, they are by no means exclusive. Speech is a dynamic process, and it is possible that the misarticulations Observed might have resulted from interrelationships among these subphonemic features. Gross analysis of substitution errors with respect to place, manner, and voicing characteristics showed errors to closely approximate target sounds. 3. to determine the patterns of distinctive feature errors. The findings revealed a tendency for the plus specification of a feature to be more vulnerable than the minus specification. This was obtained fer the voicing and.affrication features. 4. to determine the categories of articulation errors made by multiple sclerosis subjects. There was a predominance of errors of the 110 substitution type in the data. Interpretation of the two-way fixed effects analysis of variance revealed that articulation errors were affected by the type of verbal task and the position of a sound in context. Both task and position showed significant main effects. There was no interaction between them. Interpretation of two one-way fixed effects analyses of variance yielded the following results: the first one-way analysis of variance was used to determine the relationship between misarticulations and developmental hierarchy of phoneme emergence. .A comparison between 8 phonemes which appear early and another 8 which appear late in a child's articulatory repertoire showed a nonsignificant difference in misarticu- lations as a function of phoneme developmental.hierarchy. The second one— 'way fixed effects analysis of variance was used to determine the dif- ferences in errors that occurred in the first 4 phonemes which appear early and the last 4 phonemes which appear late in a child's articulatory repertoire. .A significant difference was found at the .05 level of confidence. There were more articulation errors in the late-appearing sounds. Finally, no clear-cut evidence was found to the effect that there is a relationship between breakdown in articulation by the multiple sclerosis group and acquisition hierarchy of distinctive features. Conclusions It is difficult to relate the findings of this study to the results of other studies primarily because of the dearth of litera- ture on the subject. NOnetheless, the results obtained from the various research questions in this study lead to the following conclusions: 111 1. On the basis of the subphonemic feature analysis the conson- ants which posed the greatest difficulty in their production share certain features in comnon: they are fricatives; they are produced by vocal fold.vibration; and they are produced in the anterior part of the oral cavity. 2. The 16 multiple sclerosis subjects did not demonstrate identical error tendencies. The wide range of difficulty in their error productions would.imply that multiple sclerosis, as a disease process, manifests itself differently in the speech productions of different individuals. 3. With the multiple sclerosis subjects in this study, analysis of errors according to the subphonemic features cannot be attributed solely to place of articulation, manner of production, or voicing but to interrelationships among these three subphonemic features. 4. Tabulation of substitution errors as confusion matrices indicate that substituting phonemes are close approxrmations of target phonemes. In general, substituting phonemes are simplifications of target phonemes. 5. In general, there is an orderly pattern in the misarticulations of multiple sclerosis subjects. This would appear to be in agreement with the orderly pattern in misarticulations made by dysarthrics. 6. On the basis of the distinctive feature analysis, the features voicing and affrication demonstrate clearly the tenets of the markedness theory. No convincing evidence was fOund to the effect that the features place and nasality are inconsistent with the markedness theory. 7. The group of multiple sclerosis subjects is homogeneous with respect to the general categories of articulation errors made. The predominance of substitution errors lend support to this conclusion. 112 Adthough a careful segment of multiple sclerosis sUbjects was in- vestigated in this study, the articulation errors Observed can be said to be more representative of the speech productions of the multiple sclerosis population from which the segment was derived. 8. Articulation errors appeared to be more influenced by the imitative (recorded) method of stimulus presentation than the spon- taneous method. The difficulty of responding to the imitative stimuli infers that recorded monosyllabic CVC words are difficult to discern in the absence of other contextual cues. FUrthermore, the propagating medium (tape recorder and/or room characteristics) may cause distortion not present in the speech stimuli. 9. .Although the position of a sound in context has different effects on different individuals, phonetic disintegration in.multiple sclerosis is influenced more by the final position of a phoneme than the initial. 10. No convincing evidence was found to the effect that misarticu- lations are related to the developmental hierarchy of phoneme emergence. 11. The breakdown in articulation could not be inferred from.the acquisition hierarchy of distinctive features. Recommendations On the basis of the findings of the present study, the investi- gator suggests the need for future research in the following areas: 1. An analysis of the articulatory behavior in spontaneous connected speech and isolated word responses of persons with multiple sclerosis to see whether there are large and real differences between words spoken in connected speech and the same words spoken as isolated 113 responses. Such a study would help determine the proportion of errors attributable to problems of coarticulation. The phonetic contexts whiCh are observed in the traditional, word-oriented, two-position test are not representative of the phonetic contexts which occur in connected speech. 2. Several studies have suggested that error production increases as the length of utterance increases. JOhns and Darley (1970), for example, fOund.that errors increased as the number of syllables increased. Research is recommended on phonetic and phonemic analysis that reflects progressive difficulty to see whether articulation difficulty is best reflected in phonemic and.phonetic complexity of words. For instance, the word list should reflect the fellowing phonemic construction: CVC as in "cat"; CVCV as in "today"; CCVCC as in "sponge"; CVCCV as in "dancer". 3. To ascertain the effects of speech defects in multiple sclerosis, it would be necessary to observe and obtain speech samples from multiple sclerosis subjects over time. Given the characteristics of the disease 'with symptoms coming and going, it would be necessary to measure speech problems during relapses and speech problems during remissions. 4. Research should determine to what extent vowel articulation is preserved by the disease process of multiple sclerosis. Thus, a comparison of the effects of the disease on vowel articulation and consonant articulation is necessary. As many vowels as possible should be utilized with patients who share geographic and social environment so that vowel quality is not the main source of variability between speakers of the same language. 5. There is a need for research that would yield data concerning 114 identifiable subgroups of patients with articulatory dysfhnctions. For instance, multiple sclerosis patients may be grouped into diagnos- tic categories according to the site of lesion in the central nervous system to see whether certain articulatory profiles emerge as a function of diagnostic categories. Such a study will have diagnostic and thera- peutic values. 6. The acoustic difference between phonemes, for example, /f, 0/, /v, d/, is the most difficult fOr listeners to hear particularly in recorded stimuli. It seems that in most natural situations, the verbal context and the visual observation of the speaker's lips play an important part in distinguishing consonant pairs. Future research should.employ the auditory-visual method of stimulus presentation (i.e., speech models from the visible experimenter) rather than the auditory stimuli (i.e., recorded speech stimuli). The use of live- modelling by the experimenter will mitigate the difficulties of de- termining phonemic productions. 7. The use of nonsense syllables or nonsense words as stimuli in a future research might rule out top-of-the-head effect (memory) due to meaningful syllables or words. 8. There is a need for a spectrographic analysis of the acoustic properties of speech sounds produced by multiple sclerosis subjects. This acoustic viewpoint should examine such properties as fOrmant structure, fUndamental frequency, duration, and intensity. Such a study may help the clinician to predict which changes in pitch, voice, intensity, and timing are needed to improve intelligibility. 9. Instrumental measurements of velopharyngeal closure competency or incompetency in multiple sclerosis should determine resonance de- viations in patients with the disease. 115 10. Given that multiple sclerosis affects each patient differently, it is necessary to determine whether there is any idiosyncratic produc- tion of particular sounds in clusters and in polysyllables. The analysis should examine the speech productions from a phonetic and a distinctive feature vieWpoint. 11. One of the limitations of this study relates to the small sample size. The use of only 16 subjects limits the conclusions that can be drawn. Further research should use a larger number of patients. 12. It would be of interest to explore potential articulatory com- pensatory techniques employed by multiple sclerosis subjects that could be integrated into therapy programs to increase speech intelligi- bility. 13. Research should determine how the incidence of phonetic errors varies according to age and sex of each patient. For example the categories of articulation errors (i.e., substitutions, omissions, and distortions) could be examined to see whether these errors occur in differential proportions as a fUnction of age and sex. 14. Further researCh should determine whether the breakdown in articulation could be inferred from the acquisition hierarchy of dis- tinctive features. The study might group phonemes with identical dis- tinctive features to see whether there is a definite pattern in the breakdown of articulation with respect to distinctive feature acquisi- tion hierarchy. 15. Given the odd audiogram.configurations of some multiple sclerosis patients reported in the literature, an analysis of phonetic errors based on such factors as those stemning from auditory processing problems is necessary. BIBLIOGRAPHY BIBLIOGRAPHY Acheson, E.D. The Epidemiology of Mu1tiple Sclerosis. In.MtAlpine, D., Lumsden, C.B., and Acheson, E.D. Multiple Sclerosis, A Reappraisal, pp. 3-58. Edinburgh and London (1965). Acheson, E.D., Bachrach, C.A., and Wright, PM. Some Comments on the Relationship of the Distribution of Multiple Sclerosis to Latitude, Solar Radiation, and other Variables. Acta Psychiat. Scand., 35 (suppl. 147), 132-147 (1960). Acheson, E.D., Bachrach, C.A. The Distribution of Multiple Sclerosis in United States Veterans by Birthplace. Amer. J. Hyg., 72, 88-99 (1960). Adams, R.D. and Kubrik, OS. The Morbid Anatomy of the Demyelinative Diseases. Amer. J. of Med., 12, pp. 510-546 (1952). Allison, R.S. and Miller, J.H.D. Prevalence and Familial Incidence of Disseminated Sclerosis. Ulster M. J. (suppl. 2) (1954). Alter, M. and Kurtzke, J.F. The Egdemiology of Miltijfle Sclerosis. Springfield, Illinois (196@. Alter, M., Halpern, L., Kurland, L.T., Bornstein, B., Leibowitz, U., and Silberstein, J. Miltiple Sclerosis in Israel. Prevalence among Immigrants and Native Inhabitants. Arch. Neurol. (Chicago) 7, 253-263 (1962). Alter, M., leibowitz, U., and Speer, J. Risk of Multiple Sclerosis Related to Age at Immigration to Israel. Arch. Neurol. (Minneapo- lis) 15, 234-237 (1966). Alter, M., Allison, R.S., Talbert, O.R., and Kurland, L.T. Geographic Distribution of Miltiple Sclerosis. A Comparison of Prevalence in Charleston County, South Carolina, USA, and Halifax County, Nova Scotia, Canada. World Neurol., 1, 55-68 (1960). American National Standards Institute. Guidelines for Identification Audiometry. ASHA, 17, p. 94-99 (1975). Antonelli, A.R. and DeMitri, T. The Audiometric Features of Disseminated Sclerosis. Sist. Nerv., 15, 138-145 (1963). Aring, G. Observations of Multiple Sclerosis and Conversion Hysteria. Brain, 88, 663-674 (1965). 116 117 Aronson, A., Brown, J., Litin, E., and Pearson, J. Spastic Dysphonia. J. Speech and HearinLDis., 33, 219-231 (1968). Bailey, P., thkay, R.P., and Millen, F.J. Mlltiple Sclerosis. Postgrad. Med., 20, 292-299 (1956). Barth, E. Physiologie, Pathologie und Hygiene der Menschlichen Stimme. Leipzieg: Thieme (1911). Beebe, G.W., Kurtzke, J.F., Kurland, L.T., Auth, T.L., and Nagler, B. Studies on the Natural History of Multiple Sclerosis: Epidemiologic Analysis of the Army Experiences in World War II. Neurology, 17, l-l7 (1967). Behrend, R.C. Prevalence of Miltiple Sclerosis in Hamburg and Marseille. Acta Neurol. Scand. (suppl. 19) 27-42 (1966). Behrend, R.C. Miltiple Sclerosis in Europe. European Neurol., 2, 129-145 (1969a). Boyd, W.A. Textbook of Pathology, Structure and Function in Disease. (8th Ed.) Philadelphia, Lea and Fibiger (1970). Braceland, F.J. and Griffin, M.E. The Mental Changes Associated with Multiple Sclerosis. Res. Nerv. and Ment. Dis., 28, 450-455 (1950). Brain, R. Diseases of the Nervous System. (6th Ed.) New York: Oxford University Press (1962). Brain, W.R. Critical Review: Disseminated Sclerosis. kart. J. Med., 23, 343-391 (1930). Breland, A.E. and Currier, R.D. Miltiple Sclerosis and Amyotrophic Lateral Sclerosis in Mississippi. Neurology, 17, 1011-1016 (1967). Brookshire, R. Control of "Involuntary" Crying Behavior Emitted by Multiple Sclerosis Patients. J. Commn. Disorders, 3 (3), 171-176 (1970). Brown, J.R., Darley, F.L. and Aronson, A.E. Ataxic Dysarthria. Interna- tional J. Neurology, 7, 302-318 (1970). Brown, S. and Davis, T.K. The Mental Symptoms of Multiple Sclerosis. Arch. of Neurology and Psychiatry, 7, 629-634 (1922). Brumlik, J. Disorders of Motion. Am. J. Physical Medicine, 46 (1), 536-543 (1967). Brunner, D. and Lobl, K. Serum Cholestral, Electrophoretic Lipid Pattern, Diet and Coronary Artery Disease: A Study in Coronary Patients and in Healthy Men of Different Origin and Occupation in Israel. Ann. Intern. Med., 49, 732-750 (1958). 118 Byrne, M.C. and Shervanian, C.C. Introduction to Commumicative Disorders. Harper and Row, Publishers, New York (1977). Cairns, C.B. Markedness, Neutralization, and Universal Redmdancy Rules. Langu_age, 45, 863-885 (1969). Cairns, C.E. and Williams, P. Analysis of the Substitution Errors of a Group of Standard-English-Speaking Children. J. Speech and Hearing Res., 15, 811-820 (1972). Cairns, H.S., Cairns, C.B. and Williams, F. Some Theoretical Consider- ations of Articulation Substitution Phonomena. Lang. Speech, 17, 160-173 (1974). Canter, G.J. Neuromotor Pathologies of Speech. Amer. J. of Physical Medicine, 46 (1), 659-666 (1967). Carrow, E. Deviant Speech Characteristics in Motor Neuron Disease. Arch Otolaryngol., 100 (3), 212-218 (1974). Charcot, J.M. Lectures on the Diseases of the Nervous System. London, England. The New SydenHam Society (1877). Chomsky, N. and Halle, M. The Sound Pattern of En lish. New York: Harper and Row, pp. 176-177 and 303-305 (1968;. Chusid, J .G. Correlative Neuroanatomy and Functional Neurology. Lange Medical Publications, (15th Ed.) (1973). Collet, J. Les Troubles de l'innervation pharyngo-laryngee et Oesgphgienne. Paris: msson (I946). Compston, H. and McApline, D. The Familial Incidence of Disseminated Sclerosis and its Significance. Brain, 74, 191-232 (1951). Compton, A.J. Generative Studies of Children's Phonological Disorders. J. Speech and Hearing Dis., 35, 315-339 (1970). Cotterell, 8.8. and Wilson, S.A. Affective Symptomatology of Disseminated Sclerosis: Study of 100 Cases. J. Neurol. Psychopathol., 7, 1-30 (1926). Cruishank, E.K., Montgomery, R.D. and Spillane, J.D. Obscure Neurologic Disorders in Jamaica. World Neurology, 2, 199-210 (1961). Darley, F.L., Aronson, A.E. and Brown, J.R. Motor-Speech Signs in Neurologic Disease. Med. Clin. in Amer., 52, 835-844 (1968). Darley, F.L., Aronson, A.E. and Brown, J.R. Clusters of Deviant Speech Dimensions in the Dysarthrias. J. Speech and Hearig Res., 12, 462-469 (1969a). Darley, F.L., Aronson, A.E. and Brown, J .R. Differential Diagnostic Patterns of Dysarthria. J. Speech and Heari_ng Res., 12, 246-269 (1969b). 119 Darley, F.L., Brown, J.R. and Goldstein, N.P. Dysarthria in Multiple Sclerosis. J. Speech and Hearing Res., 15, 229-245 (1972). Davenport, C.B. Multiple Sclerosis from the Standpoint of Geographic Distribution and Race. Arch. Neurol. Psychiat. (Chicago), 8, 51-58 (1922). Davis, M.Z. Livingwith Wltiple Sclerosis. Charles C. Thomas, Springfield, Illinois (1973). Dayal, v.S., Tarantino, L., and Swisher, L.P. Neurologic Studies in Multiple Sclerosis. Laryngoscog, 76, 1798-1809 (1966). Dean, G. Annual Incidence, Prevalence and Mortality of Multiple Sclerosis in White South-Africa-Born and in White Immigrants to South Africa. Brit. Med. J., 2, 724-730 (1967). DeJOng, R.N. The Neurologic Examination. (3rd Ed.) New York: Harper and Row (1967). Denny-Brown, D. Multiple Sclerosis: The Clinical Problem. Amer. J. Medicine, 12, 501-509 (1952). Dix, M.R. Loudness Recruitment and It's Measurement with Special Reference to the Loudness Discomfort Level Test and Its Values in Diagnosis. Ann. Otol. Rhinol. Laryngo_l., 77, 1131-1151 (1968). Dundas-Grant, J. Case of Unilateral "Nerve Deafness" in Disseminated Sclerosis, with Inmobility of the Opposite Vocal Cord. Proc. pr. Soc. Med., (Lect. Otolaryng.), 15, 42 (1922). Edwards, K. and Anderson, J. .A Factor-Analytic Study of the Articulation of Selected English Consonants. J. Speech and Hearing Res., 15, 720-728 (1972). Espir, M. and Rose, C. The Basic Neurology of Speech. F.A. Davis Company, Philadelphia, PA (1970). Faircloth, M.A. An Analysis of the Articulatory Behavior of a Selected Group of Speech-Defective Children in Spontaneous Connected Speech and In Isolated-Word Responses. (Dissertation, Florida State University) (1970). Farmakides, M.N. .and Boone, D.R. Speech Problems of Patients with Miltiple Sclerosis. J. Speech and Hearing Res., 25, 385-399 (1960). Fisher, H.B. and Logemann, J .A. The Fisher-Logemann Test of Articulation Competence. Boston: Houghton-Mifflin (1971). Fleming, K.J. Guidelines for Choosing Appropriate Phonetic Context for Speech-Sound Recognition and Production Practice. J. Speech and Hearing Dis., 36, 356-367 (1971). 120 Fog, T. and Linnemann, F. The Course of Multiple Sclerosis in Sweden: Geographic Pathology of Multiple Sclerosis. Acta Neur. Scand., 46, (suppl. 47), ll-30 (1970). Foster, R.M. and Harries, J.R. Multiple Sclerosis in the African. British Med. J., 3, 628 (1970). Gall, J.C., Hayles, A.B., Siekert, R.G. and Keith, H.M. Multiple Sclerosis in Children: A Clinical Study of 40 Cases with Onset in Childhood. Pediatrics, 21, 703-709 (1958). Georgi, F. and Hall, P. Studies on Multiple Sclerosis Frequency in Switzerland and East Africa. Acta Psychiat. Scand., 35, (suppl. 147) 75-84 (1960). Goldschmidt, E. (Ed.) The Genetics of Mi ant and Isolate Populations. Williams and Wilkins, Baltimore (19 3 . Gordon, E.E. Multiple Sclerosis, Application of Rehabilitation Tech- niques. New York: National Multiple Sclerosis Society (19$. Griffin, M.E. The Mental Changes Associated with Multiple Sclerosis. Res. Nerv. and Ment. Dis., 28, 450-455 (1950). Grinker, R.R. and Sahs, A.L. Neurology (6th Ed.) Springfield, Illinois: Charles C. Thomas (1966). Grinker, R.R., Ham, G.C. and Robins, F.P. Some Dynamic Factors in Multiple Sclerosis. Res. Nerv. and Ment. Dis., 28, 457-460 (1950). Gudmunsson, K.R. and Gudmunsson, G. Studies in Multiple Sclerosis, V, Multiple Sclerosis in Iceland. Acta Neurol. Scand., (suppl. 2) (1962). Guilford, J .P. Fundamental Statistics in Psychology and Education. New York: McGraw-Hill Book Co., Inc. (1956). Halle, M. On the Basis of Phonology. In Fodor, J.A. and Katz, J.J. (Eds.), The Structure of Language. Prentice Hall, Englewood Cliffs, NJ, 324-333 (1960. Hallpike, C.S. The Loudness Recruitment Phonomenon: A Clinical Con- tribution to the Neurology of Hearing. In Graham, A.B. (Ed.) Sensorineural Hearing Process and Disorders. Little, Brown, Boston, p. 493 (1967). Hargreaves, E.R. Epidemiological Studies in Cornwall. Proc. Roy. Soc. _M_ed., 54, 209-216 (1961). Harrison, T.R., Adams, R.D., Bennett, I.L., Resnick, W.H., Thorn, G.W., and Wintrobe, M.M. Principles of Internal Medicine (5th Ed.) New York, McGraw-Hill (1966). Hess, K. Ueber Einen Fall von Multipler Sklerose des Centralnerven- systems. Arch. F. Psychiat., 19, 64-81 (1888). 121 Hyllested, K. Disseminated Sclerosis in Denmark. Prevalence and Geographical Distribution. Dissertation. COpenhagen (1956). Hyllested, K. Lethality, Duration, and Mortality of Disseminated Sclerosis in Denmark. Acta Psychiat. Scand., 36, 553-564 (1961). Ince, L.P. and Rosenberg, D.N. Modification of Articulation in Dysar- thria. Arch. Phys. Med. Rehab., S4 (5), 233-236 (1973). Ivers, R.R. and Goldstein, N.P. Multiple Sclerosis: A Current Appraisal of Symptoms and Signs. Proc. Staff Meet. Mayo Clinic, 38, 457- 466 (1963). Jakobson, R., Fant, G., and Halle, M. Preliminaries to Speech AnalLsis: The Distinctive Features and Their Correlates. MIT Press, Cambridge, Massachusetts (I95D. Jakobson, R. and Halle, M. Fundamentals of Language. bibuton. The Hague (1956). Janvrin, F. and Worster-Drought, C. Diagnosis of Disseminated Sclerosis: By Graphic Representation of Film Tracks. Lancet, 2, 1384-1385 (1932). Jenson, J .R. A Study of Certain Motor-Speech Aspects of the Speech of Multiple Sclerosis Patients. (Doctoral Dissertation, University of Wisconsin) (1960). Johns, D.F. and Darley, F.L. Phonemic Variability in Apraxia of Speech. J. Speech and Hearing Res., 13, 556-583 (1970). Kallner, G. Epidemiology of Arteriosclerosis in Israel. Lancet, 1, 1155-1156 (1958). Kentner, W. Hearing Disorders in Multiple Sclerosis. H.N.O. Berlin, 4, 133 (1954). Kolb, L.C. Multiple Sclerosis in the American Negro. Arch. Neurol. Psychiat., 47, 413-421 (1942). Kraft, A. and Wessman, H. Pathology and Etiology of Multiple Sclerosis. Physical Therapy, 54 (7), 716-720 (1974). Kresheck, J .D. and Socolofsky, G. Imitative and Spontaneous Articula- tory Assessment of Four-Year-Old Children. J. Speech and Hearing D_i_s_., 15, 729-733 (1972). Kurland, L.T. Epidemiologic Characteristics of Multiple Sclerosis. Amer. J. Med., 12, 561-571 (1952b). Kurland, L.T. and Westlund, K.B. Epidemiologic Factors in the Etiology and Prognosis of Multiple Sclerosis. Ann. N.Y. Acad. Sci, 58, 682-701 (1954). 122 Kurland, L.T. Introduction for the Geomedical Conference in Copenhagen, 1959. Acta Psychiat. Scand., (suppl. 147), 13-17 (1970). Kurland, L.T., Stazio, A., and Reed, D. An Appraisal of Population Studies of Multiple Sclerosis. Ann. N.Y. Acad. Sci., 122, 520- 541 (1965). Kurtzke, J .F. and Berlin, L. The Effects of Isoniazid on Patients with Multiple Sclerosis. Amer. Rev. T‘uberc., 70, 577-592 (1954). Kurtzke, J .F. and Berlin, L. Isoniazid in the Treatment of Multiple Sclerosis. J. Amer. Med. Assoc., 163 (3), 168-174 (1957). Kurtzke, J .F. General Features of the Prevalence of Multiple Sclerosis. J. Indian Med. Prof., 11, 4896-4901 (1964). Kurtzke, J .F. Further Notes on Disability Evaluation in Multiple Sclerosis, with Scale Modification. Neurology, 15, 654-661 (1965). Kurtzke, J .F. Multiple Sclerosis and Infection from an Epidemiologic Aspect. Neurology, 18, 170-175 (1968a). Kurtzke, J.F., Beebe, G.W., Nagler, B., Kurland, L.T. and Auth, T.L. Multiple Sclerosis in United States Veterans. Preliminary Observations. In Alter, M. and Kurtzke, J.F. (Eds.) The Epidemiology of Multiple Sclerosis. Charles C. Thomas, Springfield, Illinois, 26-34 (1968) . Langworthy, O.R. , Kolb, L.C., and Androp, S. Disturbance of Behavior in Patients with Disseminated Sclerosis. Am. J. Psychiat., 98, 243-249 (1941). Leibowitz, U. and Alter, M. Multiple Sclerosis. American Elsevier Publishing Company, Inc., New Yfik, NY (1973). Leutenegger, R.R. Patient Care and Rehabilitation of Communication- Impaired Adults. Charles C. Thomas Publishers, Springfield, Illinois (1975). Lezak, R.J. and Selhub, 8. On Hearing in Multiple Sclerosis. App. Otol. Rhino. Lama” 75 (4), 1102-1110 (1966). Limburg, C.C. The Geographic Distribution of Multiple Sclerosis and its Estimated Prevalence in the United States. Res. Pub. Ass. Nerv. Ment. Dis., 28, 15-24 (1950). Luchsinger, R. and Arnold, G.E. Voice-Speech-Language. Clinical Communicology: Its Physiology and Pathology. Wadsworth Publishing Company, Inc., Belmont, California (1965). Matthews, J ., Everson, R., and Burgi, E.J. Effect of Isoniazid on the Speech of Multiple Sclerosis Patients. J. Speech and Hearing Dis. , 25, 38-42 (1960). 123 BkAipine, D. and Compston, N. Some Aspects of the Natural History of Disseminated Sclerosis, Incidence, Course, and Prognosis: Factors Affecting Onset and Course. Quart. J. Med., 21, (82), 135-167 (1952). McAlpine, D., Compston, N. and Lumsden, C.B. Multiple Sclerosis. London: E. and S. Livingston, Ltd. (1955). McAlpine, D., Lumsden, C.B. and Acheson, E.D. Multiple Sclerosis: A Reappraisal. Baltimore, The Williams and Wilkins Company U965). M:Call, M.G., Brereton, T.L.G., Dawson, A., Millingen, K., Sutherland, J .M. and Acheson, E.D. Frequency of Multiple Sclerosis in Three Australian Cities -- Perth, Newcastle, and Hobart. J. Neurol. Neurosurg. Psychiat., 31, 1-9 (1968). McClure, H.D. Clinical Neurology. Kikksville, Missouri (1936). McReynolds, L.V. and Huston, K. A Distinctive Feature Analysis of Children's Misarticulations. J. Speech and Hearing Dis. , 36, 155-166 (1971). McReynolds, L.V. and Bennett, 8. Distinctive Feature Generalization in Articulation Training. J. Speech and Hearing Dis. , 37, 461- 470 (1972). Merritt, H. A Textbook of Neurology. (5th Ed.) Philadelphia, Lea and Febiger (1973). Menyuk, P. The Role of Distinctive Feature in Children's Acquisition of Phonology. J. Speech and Hearing Res., 11, 138-145 (1968). Millar, J .H.D. Multi 1e Sclerosis: A Disease Acquired in Childhood. Charles C. IThomas, Springfield, IIlinois (I97I). Miller, G. and Nicely, P.E. An Analysis of Perceptual Confusions Among English Consonants. J. Acoust. Soc. Amer., 27, 338-352 (1955). Miller, H., Ridley, A.R. and Schapira, K. Multiple Sclerosis: A Note on the Social Incidence. Brit. Med. J., 2, 343-345 (1960). Miller, H. Trauma and Multiple Sclerosis. Lancet, 1, 848-850 (1964). Morley, M. Defects of Articulation. Folia Phoniat., 11, 65-124 (1959). Nagler, B., et a1., (Veterans Administration Multiple Sclerosis Study). Isoniazid. J. Amer. Med. Assoc., 163 (3), 168-174 (1957). National Multiple Sclerosis Society: Multiple Sclerosis. New York, National Multiple Sclerosis Society (1969b). Oftedal, S-I. Multiple Sclerosis in Vestfold County. Acta Neurol. Scand., (suppl. 16) (1965). 124 Oh, S.J. Multiple Sclerosis in the Negro. J. National Med. Assoc., 61 (S), 388-392 (1969). Papac, R. Multiple Sclerosis: A Clinical Review. Stanford Med. Bull., 15, 75-77 (1957). Parker, F. Distinctive Features in Speech Pathology, Phonology or Phonemics. J. Speech and Hearing Dis., 41, 23-39 (1976). Parker, W., Decker, R.L. and Gardner, W.H. Auditory Function and Intracranial Lesions. Arch. Otolaryngol., 76, 425-435 (1962). Parnelius, M. The Prevalence of Multiple Sclerosis and Some Related Neuro-Pathological Syndromes in South-Western Finland. Acta Neurol. Scan., 11 (suppl. 13), 653-659 (1965). Parnelius, M. Studies of Epidemiological, Clinical, and Etiological Aspects of Multiple Sclerosis. Munksgaard, Copenhagen (1969). Paynter, E.T. and Bumpas, T.C. Imitative and Spontaneous Articulatory Assessment of 3-Year-Old Children. J. Speech and Hearing Dis. , 42, 119-125 (1977). Penfield, W. and Roberts, L. Speech and Brain Mechanisms. Princeton, NJ: Princeton Univeristy Press (1969) . Peterson, G.E. and Shoup, J. A Physiological Theory of Phonetics. J. Speech and Hearing Res., 9, 5-67 (1966). Phillips, J .W. Otologic Symptoms in Multiple Sclerosis. Laryngoscope, 62, 271-276 (1952). Pollack, E. and Rees, N. Disorders of Articulation: Some Clinical Applications of Distinctive Feature Theory. J. Speech and Hearing Dis., 37, 451-461 (1972). Poser, C.M., Presthus, J. and Horsdal, 0. Clinical Characteristics of Autopsy-Proved Multiple Sclerosis. Neurology, 16 (18), 791- 798 (1966). Poskanzer, D.C., Schapira, K. and Miller, H. Epidemiology of Multiple Sclerosis in the Countries of Northumberland and Durham. J. Neurol. Neurosurg. Psychiat., 26, 368-376 (1963a). Poskanzer, D.C., Schapira, K. and Miller, H. Multiple Sclerosis and Poliomyelitis. Lancet, 2, 917-921 (1963b). Poskanzer, D.C. Etiology of Multiple Sclerosis; Analogy Suggesting Infection in Early Life. In Alter M. and Kurtzke, J .F. The Spidemiology of Multiple Sclerosis, pp. 62-82, Springfield,— llinois 1968 . Rose, R.M. and Daly, J.F. Reversible Temporary Threshold Shift in Multiple Sclerosis. Laryngoscope, 74, 424-432 (1964). 125 Sachs, B. and Friedman, E.D. General Symptomatology and Differential Diagnosis of Disseminated Sclerosis. Arch. Neurol. Psychiat., (Chicago), 7, 551-560 (1922). Sakamoto, S. and Ichiro, K. Cases of Multiple Sclerosis Accompained by Nerve Deafness. Otolapyngol. (Tbkyo) 40, 53 (1968). Sallstrom, T. Occurrence and Distribution of Multiple Sclerosis in Sweden. Geographic Pathology of Multiple Sclerosis. Acta Med. Scandinav., (suppl. 137), 1-141 (1942). Schapira, K., Poskanzer, D.C. and Miller, H. Familial and Conjugal Multiple Sclerosis. Brain, 86 (2), 315-332 (1963). Schumacher, G.A. Multiple Sclerosis. Incidence and Statistical Data. J. Amer. Med..Assoc., 143 (12), 1059-1065 (1950). Schumacher, G.A. Multiple Sclerosis. In Conn, H.F. (Ed.) Current Therapy, 645-650, Philadelphia, Saunders (1970). Scripture, E.W. Records of Speech in Disseminated Sclerosis. Brain, 39, 455-477 (1916). Shankweiler, D. and Harris, K. An Experimental Approach to the Problem of Articulation in Aphasia. Cortex, 2, 277-292 (1966). Siegel, G., Winitz, H. and Conkey, H. The Influence of Testing Instruments on Articulatory Responses of Speech Defective Children. J. Speech and Hearing Dis., 28, 67-76 (1963). Simpkins, W.T. An Audiometric Profile in Multiple Sclerosis. Arch. of Otolaryngol., 73, 557-564 (1961). Singh, S. A Distinctive Feature Analysis of Responses to a Multiple Choice Intelligibility Test. Int. Rev. Appl. Linguist., 6, 37- 53 (1968). Singh, S. Interrelationship of English Consonants. In Hala, B., Romportl, M., and Janota, P. (Eds.) Proceedings of the 6th Inter- national Congress of Phonetic Sciences. Publishing Heuse of the Czechoslavak Academy of'Sciences, Prague (1970b). Singh, S. Distinctive Features: Theory and validation. university Park Press, Baltimore (1976). Singh, S. and Black, J.W; Study of TWenty-Six Intervocalic Consonants as Spoken and Recognized by Feur Language Groups. J. Acoust. Soc. Amer., 39, 372-387 (1966). Singh, S. and Frank, D.C. A Distinctive Feature Analysis of the Consonantal Substitution Pattern. Lang, and Speech, 15, 209-218 (1972). Snow, K. .A comparative Study of Sound Substitutions by "Normal" First- Grade Children. Speech Monoqu 31 135-141 (1964). 126 Stazio, A. and Kurland, L.T. Multiple Sclerosis: Its Frequency and Distribution with Special Reference to Washington, D.C. Neurology, 12, 445-452 (1962). Stevens, K.N. and House, A. Speech Perception. In Tobias, J .V. (Ed.) Foundations of Mudern Auditory Theory. Academic Press, New York (I972) . Surridge, D. An Investigation into Some Psychiatric Aspects of Multiple Sclerosis. British J. Psyghiat., 115, 749-769 (1969). Sutherland, J .M. Observations on the Prevalence of Multiple Sclerosis in Northern Scotland. Brain, 79, 635-654 (1956). Taylor, J. (Ed.) Selected Writings of Hughlings Jackson. vo1umes I and II. Staple Press London (1958). Templin, M.C. Spontaneous Versus Imitated Verbalization in Testing Articulation in Pre-School Children. J. Speech and Hearing Dis., 12, 293-300 (1947). Templin, M.C. Certain Language Skills in Children, Their Development and Interrelationships. Institute Child Welf. Monogr., 26, Minneapolis: University of Minnesota Press (1957) . Tikofsky, R.S. Phonetic Characteristics of Dysarthria. The University of Michigan, Ann Arbor, ORA Project 06066 (1965). Tikofsky, R.S. and Tikofsky, R.P. Intelligibility Measures of Dysarthric Speech. J. Speech and Hearing Res., 7, 325-333 (1964). Tourtelotte, W.W. and Parker, J .A. Some Spaces and Barriers in Post- Mortem Multiple Sclerosis. In Lajtha, A. and Ford, D.H. Progress in Brain Research. Amsterdam (1968). Trost, J .E. and Canter, G. Apraxia of Speech in Patients with Broca's Aphasia: A Study of Phoneme Production Accuracy and Error Patterns. Brain and Lang., 1, 63- 79 (1974). Van Riper, C. and Irwin, J.V. Voice and Articulation. Prentice-Hall, Englewood Cliffs, NJ (1958). Van Riper, C. Speech Correction: Principles and Methods. Prentice- Hall, Inc., Englewood Cliffs, NJ (1972). Von Leden, H. and Horton, B. Auditory Nerve in Multiple Sclerosis. Arch. Otolagyngol., 48, 51-57 (1948). Vulpian, A. Maladies du Systeme Nerveux. Doin, Paris, 2, 707-708 (1886). Walsh, H. On Certain Practical Inadequacies of Distinctive Feature Systems. J. Speech and Hearing Dis., 39 (2), 32-43 (1974). Ward, P.H., Cannon, D. and Lindsay, J .R. The Vestibular System in Multiple Sclerosis. Lamgoscope, 75, 1031-1047 (1965). 127 Weber, J .L. Patterning of Deviant Articulation Behavior. J. Speech and Hearing Dis., 35 (2), 135-141 (1970). Wepman, J .M. and Jones, I.v. Five Aphasics: A Commentary on Aphasia as a Regressive Linguistic Phenomenon. In Rioch, D. and Weinstein, F. (Eds.) Disorders of Communication. Williams and Wilkins Company Baltimore (1964). West, R., Ansberry, M., and Carr, A. The Rehabilitation of Speech. (4th Ed.) Harper and Brothers, New York (1963 . Whitty, C.W.M. Cortical Dysarthria and Dysprosody of Speech. J. Neurol. Psychiat., 27, 507-510 (1964). Wickelgren, W.A. Distinctive Features and Errors in Short-Term Memory for English Consonants. J. Acoust. Soc. Amer., 39, 388-398 (1966). Williams, F. (Ed.), Cairns, H.S., Cairns, G.E. and Blosser, D.F. Analysis of Production Errors in the Phonetic Performance of School-Age Standard-English-Speaking Children. Project Report Center for Communication Research, University of Texas at Austin (1971b). Zemlin, W.R. A Comparison of the Periodic Function of Vocal Fold Vibration in Multiple Sclerosis and a Normal Population. (Doctoral Dissertation, University of Minnesota, 1962). APPENDICES APPENDIX A CASE HISTORY INFORMATION 10. CASE HISTORY INFORMATION Name Date of Birth Age Sex Address Phone Patient's Educational Level Date of Illness Medication Does the patient have any history of neurological disease other than multiple sclerosis? Does the patient have any premorbid psychological prOblems? Is the patient senile? Does the patient have any history of cerebrovascular disease? Interview notes 128 APPENDIX B ARTICULATION SCREENING TEST 10. ll. 12. 13. 14. 15. TIE FISIER-IOGEMANN TEST OF ARTICULATIGV COMPETENCE SENTENCE ARTICULATION TEST PETE'S JOB WAS TO KEEP TIE BABY HAPPY. TODAY DICK TOLD PATTY ABOUT IT. THE GIRLS WERE BAKING THE BIGGEST CAKE FOR MR. TAG. TIEIR BROTHER WOULDN'T BATIE BECAUSE IE THOUGHT A BATH WOULD MAKE HIS TOOTHACIE WORSE. IN HALF A DAY, IE REPAIRED FIVE TELEVISION SETS, TWO TELEPHONES, AND A VERY OLD STOVE. SUZIE SEWED ZIPPERS ON TWO NEW DRESSES AT BESSIE'S HOUSE. SIE USUALLY RUSHES TO PUSH TIE GARAGE DOOR CLOSED. GEORGE IS AT TIE CHURCH WATCHING A MAGIC SHOW. WE RODE WITH LUCY AROUND TIE TALL TOWER IN IER NEW YELLOW CAR. WHY HAVEN'T YOU LOOKED ANYWHERE BEHIND THE HOUSE OR BEYOND TIE HILL YET? NANCY FOUND SCME FINE HANGERS ANDNG TIE MANY THINGS AT TIE SALE. LET ME KEEP A LITTLE OF THIS WEDDING CAKE TO EAT LATER. FATIER ASKED HOW MUCH M)NEY TIM HAD SAVED TO BUY A BIRD CAGE. RUTH CAUGHT A COLD BECAUSE STE WOULDN'T WEAR HER NEW WARM WOOL COAT. I FOUND A HUGE TOY MUSIC BOX OJTSIDE ROY'S HOUSE. 129 APPENDIX C MONOSYLLABIC CVC WORD LISTS USED FOR TIE TWO TYPES OF VERBAL TASKS INITIAL POSITION 10. 11. 12. 13. 14. 15. 16. /p/ /b/ /t/ /d/ /k/ /g/ /tf/ /d3/ /f/ /V/ /9/ /.’5/ /5/ /2/ /m/ /n/ POKE BONE TALE DATE COPE GAZE CHOKE JADE FAKE VAGUE THOUGHT THOSE SOAK ZONE NEMOJ 130 32 MONOSYLLABIC CVC WORD LIST FOR THE IMITATIVE TASK FINAL POSITION MOPE ROBE BAIT FADE SOAK ROGUE COACH PAGE LOAF SHAVE BOTH BATHE CASE DOZE TONE INITIAL POSITION 32 MONOSYLLABIC CVC WORD LIST FOR TIE SPONTANEOUS TASK 1. 2. 10. 11. 12. 13. 14. 15. 16. /p/ /b/ /t/ /d/ /k/ /g/ /tf/ /ds/ /f/ /v/ /9/ 737 /s/ /z/ /m/ /n/ POLE BAI L TAKE 113MB CAPE GATE CHAIN JOKE FOAM VE IL ZIP NOSE FINAL POSITION SOAP GABE FATE LOAD VOGUE POACH CAGE CAVE FAITH LOATIE FACE HOSE CAIN APPENDIX D SCORE FORM FOR MONOSYLLABIC CVC WORD LIST IMITATIVE POKE DOZE TONE ZONE GAZE VAGUE MOPE COPE BOTH SCORE FORM FOR MONOSYLLABIC CVC WORD LIST POSITI INIT. FINAL FINAL INIT. INIT. INIT. FINAL INIT. FINAL INIT. INIT. INIT. FINAL INIT. INIT. FINAL FINAL FINAL FINAL FINAL INIT. INIT. FINAL INIT. FINAL FINAL INIT. FINAL INIT. INIT. FINAL FINAL NO ERROR E SUBST. OMISS. DISTORT OTHER 132 SPONTANEOUS POLE HOSE CAIN ZIP GATE VEIL SOAP CAPE FAITH DOME FOAM NOSE SAFE THEM TAKE GABE VOGUE FATE FACE LOATHE JOKE SAKE RAKE MANE CAGE DAME CHAIN LOAD THUD BAIL POACH CAVE APPENDIX E INSTRUCTIONS USED FOR TIE IMITATIVE AND SPONTANEOUS TASKS IMI TAT IVE TASK YOU SHALL IEAR SEVERAL WORDS SPOKEN BY A MAN. PLEASE REPEAT TIE WORDS THAT YOU HEAR. DO NOT TRY TO IMITATE HIS VOICE OR PRONUNCIA- TION: ONLY REPEAT TIE WORDS THAT IE SAYS. SPONTANEOUS TASK I AM GOING TO SHCW YOU A SERIES OF WORDS. PLEASE SAY TIE WORDS ALOUD ONEBYONEASISHCMTIEMTOYOU. 133 APPENDIX F CONFUSION MATRICES OF RESPONSES ON 16 SINGLETON CONSONANTS MADE BY MILTIPLE SCLEROSIS SUBJECTS .Gofiaao .863an we: Ea ESE: 88.38 $6828 Hmmcoamomv @8395 3528200 HmcHnH Home HmHuHcH oHaHqua 0H .3 355980 3 :o mammommoh mo xHhumE :onzmcoo .w 833m r1..me on em «NH OH N H mN m mH m m m OH x mH 0H NH HN NH Heupr o N H m mm N H : mH o H NH w Hm m H E N m o NN - mm ONII N N. 1 w v o e cm H H H H m N o H NN N cm H N N N m ...2 H m NH N N a. Higjrml. H m NN o N 8 N 3...-.- N4 m lem s IH N m OH um I e N 3 1 H N .H NH H o HH Nm OH H no IHIIN o m H mm N 3 NH 0 m N 3 N m H H m N H 17..N e I. m Nm H O. H N N OH N H1- H om e N e Ila o N N H H mm H a m a a m H a m H a a a mH o m NH H m H H me a 36...»:ch flee gem. a ....__ N m m .m s c... :6 E M. a. T. to a a paiuasaud siueuosuog {any pue IBDIUI 134 135 .HNHco 8386a HHHHHeHV 866.3% mHmopoHum Homcommmmv @8355 3:38:00 HmHuHHHH oHoHuHHHE 0H .3 352028 3 :o 398ng we KHNumE :onPHHHoo .m oEmHm . ma 2 H HN m N H w o N N m H H H N a H HH m H38 N o o N cm H H c 1 m o IHI H m NN H a N H o N HN m N N m N o H .. N H 8. HH o o HH N HN m H % , N H N H N N H H m OH O o S H NN N - a H > 1 HH N o a TI H m N HN I H N H H o o H N H we N m o N H N H a. Imi TIN o m NN H N all IN...) o o m m N 43“.)... N N Tami Ia... H HN m N e r o o o O NH H m o H N H - -T H N a e o o a m H .N a immnwoflflfimm 11.19.15 ism -mlmm. N m 1.1.-.. .¢ I.» N at 3 w x H. H 2 a paiuasaud siueuosuog return} .HNHco :oHuHm0i HmchV muoonnsm mHmopoHON oHaHuHSE 0H Np mucm:0m:oo couonch NH :0 Newcoam0H mo xHNme :onsmcou .OH opsmHm 136 pauuesaud siueuosuog Iaurg 1. NHH NH NN NNH N N N NH N HH N N N N H N N N NH N H86... H N H H NN H c N N N N N HN N 5 NH N N NH NH NH -- N N N N N NN H EH. 3; m N NH N H NH NH N N N I. N h N N H N N NN H N N {NH N N NH H .11. NN N l. N s .IH N N H H NN c N H N N HN N H 3 H N N H NN H 3 NH H N H 1 I- NH H H H H N IH H N H .. NN H N N N N H N H NN H N N N N N - H H NN N NH N N NH N H N NH N N N N N N H H H NN N IMWWWWIwmmWOIMmmmw .mzw m 419 N m mm m > w No «mwstwu x a u n a Homcommoxv woosuoai mucm:0mcou Hmch 137 HHNNN 0>HumuHEH0 m300nnsm mHmopoHom oHaHuHHHE 0H .3 3:38:00 NH :0 33033 HH0 0339: 5.33:8 .HH 09mg IHHH NN NN HNH H H H NH N N N N N N N NH NH NH N N HSNN N N H N NN N . H c N - N H N N NN N 5 NH N N NH NH HH N N H N N N NN H H N 4. HH N N HH H NH H H -H N- {IIIwI NH N N N H NN 11-12 N N N N H N H NN H N s N. N N N H H NN - N N H N N HN N H NW1 N N N H NN H 3 1.le i H N N HN H N .H... H NI -.N N N N N NN H N H N N N N N H .- HN N H NI N N H H H NN H NH o N OH H H H m H NN m N. NH N N HH H NI... H NH a H36... Em. INNS .sz : a WN N W 0 > .H :3 B Midlafliui N N b - x35 0:339: 75:33.5 08:00.5 3:88:00 (x591, amtueipuq) peuuesaud suueuosuog .HHNNN N300:mp:0am0 m300nnsm mHmopoHom 0HQHHH25 0H Ne mp:m:0m:00 NH :0 mom:00m0N mo xHHumE :0Hmsm:00 .NH OHSNHN 138 (xsel snoaueiuods) pauuaseud siueuosuog INN NH HH NN N N N NH N HH N N H H H H N N HH N H82. N N N N NN : N N N N N NN H H 2 NH H N N NN N N H N N N NN H H N HH N H! NH _..Hl NH NH H H m N H N N H N NN H H N NH N H NH H NH H N H > N N H N N NN H N N H N N H NN H S N N N N H NN H 3 N N N N NN H H N N H N H H NN H I- - z N N N H H NN N H N IH N N H 1.. ;H NN N N N N N N NN N N N N H H 1 H NN N HNHoN ammo Na: Asmfi..- a N -N M N > N 2. 3 N 0. .NH. H N N Hmmb m:00:mu:0mm flom:omm0m0 000300Nm mu:m:0m:00 139 .HNHNN :oHHHNoN HNHHHNHN .NN Hommnsm Na mu:m:0m:00 NH :0 mom:oammh mo xHuumE :onsm:00 .mH OHDNHm NH m o HH o o N o H o N o o Hmuoe o o o o N : o o o o N II E N o o N o N N N o o N N o o N o MW H N N H N H o o H H > o o o o N N H... N N N N IN... N NHN N -N o o o :3 .... N? .H. o H o N - N :N-l N N H .zllmllr o o N N o o o o a N o o o N o N o o N a .230... :mmmo -NHEC .NH—v. 1: E N W H N0 3 m. all Nounpopm mu:m:0m:00 HmHuH:H paiuesaud suueuosuog tepitul 140 .HNHNN :oHHHNoN HNNHNN .N.N 30033 :3 3:38:00 NH :0 3383: 00 0332. :onsm:00 .NH 33mm NH N H N H N N N N N N N N N N N N N HNHNN H H N N H ...c H N N H H H a N N N N N 1 N N N N N N N N N N N N N mm H N N H H N N N N N N > N N N N - N 1-- N N N N N N N NN N N N N N HN .1.-N N N N N N N N N N :jiui; N HI. N N N N 1 N H N H N - -.:- H H N N H H N N N N N N z N .erHNHmo Nae .aN.H_ E N N m HN N, N NN .H u- N N N fimUDUOHm mwfimflomfiou Hmfifim paiuasaud SJUBUOSUOD [EUTfl 141 .38 ESE: HNHNHNHN .zd pumnnsm kn mu:m:om:cu 0H :0 mam:o:mmh mo xHhumE :onsmcoo .mH mastm .I OH m H o o o o o o smack o o o o : N o H‘ H o E o o o o N o o o o m IHLLN .. N H I H % o o o o m 1 «I4 Neeo o o o N > r-;.H. ........ Hi. N N N o o o o .n N n: tilm. 2.- EN 0 o o 5 o J o o o m H o o H x H o o H 1 .Ibl o o o a WILHII o o H 4-: n H o o H a :38. mewb .mWHC :sz 4|E| m lufll . > n5 2 wousvoym mu:m:om:oo HwHuH:H penuesaxd snueuosuog IBIJIUI 142 .HNHHNN SHNHNNN :8an .20 5033 N3 3530200 0H :0 320:8: H0 £55: 85550 0H 0.53: «H N m m o o N o H o o H38. H o H o : N o o N o E N N N~ N N N N o o o o N m H N H N H w H o o H w H o o H H > o o o o u o o o o N n H0 o o o o N 3 N N o o m N N N N H o o o L o H0 N o H I H H H 0 cl: H W. n H N N H I... N HEE‘ Lamb .350 .05 E N v. M > 20 Z @005on 3:28:00 H05: pamasaxd slueuosuog) {any .HNHNN NNHNHNNN HNHNHNHN .zaN 3.003% 3 3:38:00 0H :0 momgmmmu m0 033:: :0Hmsm:0o .NH PSNHNH 143 pamasexd slueuosuog [Bung OH N H N. o o o o o H H H H o o o N bl H 0 H38. o o o o N : TI c o o o N E o o o o N . N o o o o N J m N H0 N N N H H _m H o H o H m N o o N o H .1 H > N H N H H:H N N o o o o N n H0 N H N H-.. H N .N H N N H H H N N N N N - N H o o o o N H0 o o o o N a o o o o N s N N N N .::T. a... , 1.4;. N N H38. .HmHNWo .m .Eo . aim. : E N m: M m > N NHN 3 a v— : a N. : NNNNNNNN mNNNNomNou HNHNHNH .HNHNN SHNHNNN HNNHNN .zd 300.33 .3 3:28:00 0H :0 338mg m0 339: :0Hmsm:00 .NH 0.8mm 144 peuuaseld smeuosuog tenm HNH o m m o o o N o N O o N o H o o o N o HNHOF o o o o N c o O o o N E N o o N o N N H o o H H H m N N N N N N m H o H o H ¢ N o H H J- o H > H o H o H ..H o to o o N nv o o o o... N - 3 N o H H I. o H m o o o o N v— o o o o N H0 H N N H 3!. H H N H N H N .. H N H N N H H H N H38. .Hmmmb .HS ism : Q N r. m c > .N. at Efami 0— .0 u H— a 80300.5 3:28:00 HgHm 145 .HNHNN :oHNHNoN HNHNHNHN .zqw 300Hnsm Na mu:m:0m:00 0H :0 momcommmh mo prumE :0Hm0m:0u .mH mysmHm wouzwoum m3:w:0w:0o HmHuH:H N N N N H N N N N N N N N H N N N HNNpN N N N N N N H N N H H s H N N H H H N.:. N N N N N N N N H0 N N h“ H N N H H N H N N H > H N N H H -1 0 N N NuyNunmu.Hn‘ N -- Nu N N N N N Na N N N N N N N N N N N x N N N N N - N y. N N N N :.N .auve. N N H N N H H N H N N H -5- w H H N .M H138. .Hmfimo .m 26 . Lam E 1W. H.M..Llufl. ¢ .H at 3 w M: H. u N.— penuesexd slueuosuog [egngul .HNHNN :oHNHmoN HNNHNN .26 pumHNSm Np mu:m:0m:00 0H :0 mmm:0:m0: m0 xHNumE :0Hmpw00u .ON musmHm \O H H N o o o N o H o o o o o H N o H:30h N : 146 IOOHOO'NONOONOQ N I } NOHNNNOONONOONO §c> :3 r4 r4 :3 c> c: c: c: c: c: c: c: c> c> V) n: c: c: N 4..) pa1uesaxd squeuosuoo [eugg TII'J .0: .Im E N , . . a ..m . grmc > N 5. 3 .2 H. N N N H r4 O H G. H88 L". gigca c: c> c> vi a: r4 :3 c> c: c: c: c> c: c> c: <- 1.... 0.2.. J :3 vii I E! l i 1 I 3 l_ II .C‘Il I II wmusvopm mu:m:0m:0o Hm:Hm .HNHNN NNHNHNNN HNHNHNHN .NH. 300anm Na mu:m:0m:00 NH :0 mmm:0mmoh m0 xHHumE :0Hmzw:0o .HN mhstm 147 wm0sv09m mu:m:0m:0o HmHuH:H N N N N N N N HEN... H N N H H N H N N H H e N N m N N N N N N m N N4 N N N w H N N H N N N N N > -4 N N N N N N N N N S N N N N 3 N N N N N N N N N V. H N N H H0 N N N N N N N N N N o o o o : NNNNN 2:. him: N1 pauuesaxd szueuosuog {egngul 148 38.3% NS 3:300:00 .HNHNN NNHNHNNN HNNHNN .N.N 0H :0 330933 m0 033:: 53350 .NN 0.83m 1- NH N N NH H N N N N N N N H HNNpN H N N Lr H H c N N N N N e N N N1 N N N H N N H H m N N N N N NW H N N H N H N N H H H > 1. H N N H H N H N N H H nN.J N N N N ...1. N EN H N N H .sNauu;1-.. N H N H N N N N N N .a:. N H N H N N H N N H H N N N N N N 11309 Nmmm 3.5 .35 E N IN: 1r IM > N -NH. 2 N— 083005 3:28:00 HN:HNH paws 591d 5111811051103 {8111:} 6:8 NNHNHmoN HNHNNNHN No. pumnnsm >0 mu:m:0m:0u 0H :0 mmm:0mm0H m0 xHHumE‘:0wm:m:0o .mN whamHm 149 peauasaxd snueuosuog {egzyul N N H N H N N N N N N H N H N N N N N N HSNN N N N N N N N N N N N a N N N N N N N H N N H H H m H N N H H H m N N N N N N H N N H H H > H N N H H H N I H N N H H H nN N N N N N 3 N N N N N N N N N N N v. N N N N N N N N N N i- - N N H N H N H N N N N N N N H38 5%: .30 .Néfi: s N m £0.11»... KL r.N :TN N N N N N wmuswohm mu:m:0m:0o HmHuH:H 150 .HNHNo NNNNNmoN HNNNNN .N.N 80.3% 3 3:33:00 3 :0 $303.09 m0 N359: 833:8 .3. 0.53: 00800.5 3930200 35: HH H N N N N N H H N N N H HNNNNy H H N N N N N N N N s N N N N N N H N N H H N N N N N N “N N N N N N m N N N N H > N N N N .w N N N N N 1 N N nN o o o o N 3 H N H N H N N N N N N N N N N N H N H N N H N N H -1 H N o o o .I o - i; N a :30... Nww—mo .mHHNc .Nsm E N n .m m EN 3 H. penuasaxd snueuosuog) tenm 151 .HNHNo NNNNNNNN HNNNHNHN .zN N823. 3 3:23:00 3 :0 3200mm: mo 5.5m: :0wmsm:0u .mm 0.53m H N N H N N N N N N N N H N N HNNNN N N N N N N N N N N N _._._i; N N N N N N N N N m H N N H H n N N N N N N N N N N > N N N N N : N N N N N N N N N N N N B N N N N N N N N N N :1 N N N N N N N N N N N N N N N N N N N N N N N N N N :38. :mmc .350 .Nzw IHIHINH... Ii > - .N N: 3 m N. .0 N 0 @0038: 3:28:00 HNHNHE penua S 81d 5111911051103 IBIJIUI .HNHNN NoNNNmoN HNNHNN .NNN 80.33 3 3530300 0H :0 30:83» 00 x059: 53.0050 .3 ohsmHNH 152 pauuaseld saueuosuog) {Bum NH N N N m N N N H H N N N H H H N N H N HNNNN N N N N N .:n N N N N N N N N N N N N N N N N N N N - m H N N H .s.1.. H H my N N N N N I N u N N H H N H > N N N N N N N H N H N H nN N N N N N NN N N H H N H N N N N N N N H o o H IIuH H .0 N N N N N N N N H H H N N N N N N ud N N H300. .H 0m 0.350 :30 : E N m M x > .H .6 3 m N H0 N a a 08:00.5 3:28:00 Hm:Hm 153 .030 83:8 HNNNNNHN .z.> 000.38 3 3:28:00 NH :0 98:08.8 “Ho 0339— 8st050 .HN 9305 N H H N N N N N H N N N H N N N HNNo.H N N N N N N N N N N N N N N N N N N N N N N N N N N N N m H N H N H N N N N N N > H N N H H H N H N N H H H NN N N N N N B N N N N :11 N N all) N N N N N H H N N I. z- N N N N N N N N N N N - I N N H N N H .. N #30:. Hummb .NHNC .st : E N u M .0 > N E. 3 m a a 000000NN mN:N:0m:00 HNHNH:H [3011.195 01d 5111811051103 1811111]: .HNHNN NoHNHmoN HNNNNN .z.> 000n000 x0 mu:0:0m:00 0H :0 000:0000N 00 xHNNNE :0Hmsm:00 .wm 0H=me 154 peuuaseld s1u3uosu03 taugd HH N Hi. NH H N N N N N H N N N H H N H N N .033 N N N N N N H N N. H H H N N N N N N N N N N Ni N N N N N N WIN... N H H m H N N H H H N N N N N N I. > N N N N N N N N N N... N nN N N N N N 3 H N N H H H N N N N N N N H N N H H H N H N H N H N H N N H , H H N H N N H it H H N HNNNNNEmN .NNNN .NNN N N‘ N , m .N N N NN B N N N N N N N00000:: NN:N:00:00 HN:HN .HNHNN NoNNNmoN HNNNNNHN :8 000.38 3 00:28:00 0H :0 08:088 mo £59: 580080 .3 050E 155 paguesaxd snueuosuog 19111111 N N N N N N N H N N N N N N N N N N N N HNNoN. N N N N N .l. N N N N N N N N N N N N N N N N N N N N N N N N N m N N N N N N N N N N N > H N N H H H N .N l N N N N N NN N N N N ll: N N... N N N N N N N N N N N N N N N N N N o o o I N N N N N N N N N N N N N N 230,—. hmfijc mmla .cHHV. : E .N. vllN M SQ > u 30 .3 a v— Hu u D Q 0005005 00:88:00 HNHNHE 156 .HNHNo :oHuHmoN HNNHNN .H.m pummnsm >3 mucmcomcou cH no momcommmu mo xflpums :owmamcou .om mHsmHm wmuzwopm muchOmcou Hmcflm N N m N N N H H N H N N N N N N HNHNN N N N N c N N N N N E H N N H H H N N N N N N m H N N H H H aw H N N H H H m [I.N N N N .l.l N > N N N N N N N N N N N nN N N N N N HH N N N N N N N N N N l.. H H NllWlTN H N N N N N - N u H. N N H Llll N N N N N ll. N N HwHoh Hummo .mNEo .::m E N m MW,.¢ >.l u Ncl Ha w v u pe1uesaxd szueuosuog IBUIQ 157 .HNHNN :oHpHmoN HNHNHNHN .N.N uuonnSm kn mucmcomcou 0H :0 mmmcommmh mo Kahuna :onpwaou .Hm mhswwm HNNNN N PrmeE: N n u— > <3 '0 N bug 4.: N CID OONHOOOOOHOOOHOB OOOOOOOOOHOOOOOI—l OOOOCOOOOOOOOOOO O NOONHOOOOOOOOOHOO HNHNN HEWN wwso ; I? I'll mfl.m. >. H aw H» lw x v a 3 a wouswopm mucwcomcou HmwuwcH panuaseld szueuosuog 19:1yu1 .35 83:8 HNNHN: HzN pummpzm kn munmcomcou 0H co momcoammu mo xfihume sawmsmcoo .Nm mhsmfim 158 wmusw0pm mucmcomcou Hmcwm m N N N N N N N N H N N N H N H N N N N H82. N N N N c N N N N N a N N N‘ N N o o o o m H N N H H H h N N N N N N N N N N l > N N N N l H H N N H H H 3 N N N N N 3 N N N N N N N N N N N N N N N N N N N N lN N N N N H N N H a H83 mums .35 .Nsm a m N NH. 3 penuesexd snueuosuog taugg 159 .38 8338 HNHNHNHN .N.m pummn3m xn mucmcomcou 0H no momcommou mo xfihums newmpwcou .mmnmnswwm N H N N N N N N N H H38. o o o o c N N N. N N N. H H N N H N N N N N N H N N H All H H h N N N N N l-N N N N > H N N H NH N N N N N N N H N N H H 3 N N N N N TlN N N N N. N N N N N N N N N N N N N N N N N N N N N Haas? mmmwo .mNEo .sz E N mmll Nu Ha c wousw0pm mucmcomcou HmwuwcH panuasald snueuosuog {egngul 160 .HNHNN :oHNHmNN HNNHNN .N.N 30.3% NS mu:m:om:ou NH :0 320:3: mo Nahum: :owmsmgo .3“ $5me N N N N sLfiN N N N N HNNo: N N N N N a N N N N N s N N N N N N N N N m N N N N N N" H N N H N H N N H > N N N N N N N N N N NN N N N Nll-7 N H: N N N N N N N N N N N N N N N N N N N N H N N H N H N N H N .H 38. wwwo .m :5 . azm : a M N v 3 @832: 3:23:00 HNENH panuesald snueuosuog {Bum 161 .35 :oHNHNoN HNHNHNHN .N.N NUNBHHN N3 mu:m:om:ou 3 :0 320989 mo NFSNE 533:8 .mm 933: 8960.5 3530300 Hmfiwfi l N N N N H N N N N H HEN? H N N H H N N N N N N a N N N N N N N N N N N N N N N N n N N N N N N N N N > H N N H N N N N N N NN N N N N N B T H H N N H N H N N H H N H H N N N N N N N N N N N N N o o o d ll Q 138. :mmmb .350 .35 E N m N... B u penuasaxd snueuosuog {9pm} 162 .HNHNN NNNNHNNN HNNHNN .N.N NUNNNEN >3 mp:N:om:oo NH :0 mam:o:mop mo xwhuNE :owmpm:oo .om Nasmfim NNuzwopm NN:N:om:ou HN:Nm N N N N N H N H H H N N N N N HNNNN N N N N N N N N N N N N N N N. N N N N N N N N N H N N H H H .. H N N H H H -N.lJ N N N N N > N N N N N N N N N N N NN N N N N N Hy N N N N - N N N N N N N H N N H H N N N N N N H N N H H N H N N H N fiHNNON ham 0 .NNEO .sz : E N m lmm .® > m N: Ne w penuasald snueuosuog {euyd 163 .35 8338 HNEHNHN .:.N 30.33 3 35.528 0H :0 momcommop hHo xHHumE :onacoo .3 0.53m panuasaxd slueuosuog [engul N o o N o o o o o o o o o o o o H o H o Hmaoh o o o o N c o o o o N 5:1; H o o H H H N o o o o N m o o o o N _w o o o o N m H o o H H H > o o o o N w L o o o o N at o o o o N .9 o o o p. N m o o o o N x o o o o -Lrs- :- N w o o o o N a o o o o N a TIN N N N N N 1 E8. (Hmmwo .mHEo . 22m = E N m M .m > .H N. m. 2 w x H. u L a wouswopm mucchmcou HmHuHcH 164 .HNHNN :oHHHmoN HNNHNN .m.N Homnnsm kn mucm:0m:cu OH no momnoammh mo prumE :onsmcoo .Nm ohstm panuasald snueuosuog [eutg 1 m :J N N m N N N N N N N N N N N N N H N N HNNNH N N N N N a N N N N N a N N N N N N N N N N N N m N N N N N MW N N N N N N N N N N N > N N N N N N H. N N N N N Nu r N N N N I. N HH N N N N N N H N N H H H H H o o H H H a N N N N N H H N N H H H N N N N N N N H .3 D.H. MMHHHHMD .wwzé .m.l:m : E N w M O > u n v 3 u v— 1 u n. 9 wmuzwogm mucmcomcou Hmch .HNHNN NNHNHNNN HNHNHNHN .N.N uuonnsm 3 358028 NH :0 Newcommmh mo N359: :onancoo .mm mhsmHN 16S N N N N N N N N N N N N N N N N N N N N HNNNN N N N N N N N N N N N N H N N H .:- H H N N N N N N m N N N N N MN H N N H ltx H H N N N N N N > N N N N N N N N N N N Na o O o o N In? —H N N N N N N o o o o N N— o o o o N Hg N N N N N N o o o o N N N N N N - N N H5£N%#.25 N5 N W N m M.N > N S D W N N N N N @8355 3:38:00 HmHuNHNH panuasaxd slueuosuoo [91.231111 166 .HNHNO :oHuHNON HmchV .u.N pummnsm kn mucwccmcou NH no mamaommmp mo prumE :onsmcoo .ov musmHm wousvohN mucmcomcoo HNNHN N N N N N N N N HSNN N N N N N N N N N s N N N N N N N N N N H N N H n N N N N m N N N N > N N N N N N N N N N N 3 N N -IN. N N J B N N N4 N N N N N N N N N N N N N 1 o o o o N N N N N N N N N N N IHWNNN » RE .35 .NNN .H on D N panuesald snueuosuog {euxd 167 .220 NNHNHmoN HNHNHNHN .2.N pommnsm >9 mucmcomcou NH :0 Newcommmp mo xHhumE :onsmcou .HN musmHm c o o o o o o o o o o o H o H o Hmuoh N C 1 T7 Y penuesexd saueuosuoo {egnrul OOOOHOOOHOOOOOON OOOOOOOOOOOOOOOO OOOOOOOOOOOOOOOOOI c> c> c: c: c: H4 c: c: c> F4 c: c: c: c> c> c> ca N n n. O O Hmapr 9": 'Z 5 (“mm—mo .N:m:ENNrm_¢>..HNH.3NVHHNNNN wmunwopm mucmcomcoo HmHuHcH 168 .HNHNo NNNNHNNN HNNNNN .ZNN pumnnsm xn Nucmcomcou NH co momaoamou mo xHhumE :onpwcou .Ne ohsmHm N N N N H N N N N N N HNNNN N N N N N N N N N N N N .N N N N N N N N N m H N N H H _w N N N N o H N N H H > N N N N N N N N N N N N NN N N N N N Hp N N N N N N N N N N N N N N N - N N N N N N H N N H - N N N N N N H38 ammo .NN .NNN a N W _ INi. NH. 3 N NmuSNONN NHNNNONNOU HNNHN panuasald snuuuosuog {anti 169 .mxAHflO flowpfimom HNfiHHGHV .m.U 30.3% 3 352098 NH no 320939 mo prumN 53350 .3 musmHN ooooooooooHNHHcHoHHfic... N : N E N N H H >CD*\’Om N h... 4.: panuesald snueuosuoj {Bung N *4wa OOOOOOOOHOHOOOOM OOOOOOOOOOOOOOOO OOOOOOOOCOOOOOOOC HOOOOOOOOHOHOOOOM N IOU-a H O g Hmugwmmmc .m. .NJHWN: N. N m rum m > .N NH. 3 u v. N u N a NmuflNPE 352898 HNHHNNH 170 .35 8338 35": .md ~03an 3 mpcmcomcou 3 :o mmmcommmh mo Kaunas gwmacou .3 magma IN C o N o o o o o o o o N38 o o lo. 9 N c o o o o N s o o o o N - o o o o N m o o o o N m g N o o N N w o o o o N > [T o o o o m , o o o o N 3 o o o o t .. N 3 o o o o u o o o o x o o o o .. w 1an o o o a o o p- o a o o o o a 1::lJamlj; . , .r.; :38. uwdmp Imlmmm~11¢3m : E v % WWI. -w Lu 3 n. @8895 3:88:00 TEE penuasaxd snueuosuog {any ‘WEflifllflflfi’ifwfiiiflflifltflmfiflfi“