Вы находитесь на странице: 1из 118

JAIISH, Vol.

27, 2008
(The Journal of All India Institute of Speech and Hearing) ISSN 0973-662X Editorial Chief Dr. Vijayalakshmi Basavaraj Board of Editors
Dr. S. R. Savithri Prof. of Speech Sciences AIISH, Mysore. Dr. Asha Yathiraj Prof. of Audiology AIISH, Mysore. Dr. S. Venkatesan Prof. of Clinical Psychology AIISH, Mysore. Dr. K. C. Shyamala Prof. of Language Pathology AIISH, Mysore. Dr. R. Manjula Prof. of Speech Pathology AIISH, Mysore. Dr. K.S. Prema Prof. of Language Pathology AIISH, Mysore. Dr. Y.V.Geetha Prof. of Speech Sciences AIISH, Mysore. Dr. H. Sundara Raju Reader in Otorhinolaryngology AIISH, Mysore. Mr. Ajish K. Abraham Reader in Electronics AIISH, Mysore.

Guest Editors
Dr. Geetha Mukundan Deputy Director(Technical), AYJNIHH, Mumbai. Dr. A. Ramesh Associate Professor, Dept. of ENT, St Johns Medical College Hospital, Bengaluru. Dr. C. S. Vanaja Professor of Audiology Bharati Vidyapeeth University School of Audiology & Speech Language Patholgy, Pune Dr. P. S. Pradeep Kumar Prof. of ENT Meenakshi ENT Speciality Centre,

Bengaluru.
Co-ordinator
Dr. N. Sreedevi Lecturer in Speech Sciences

The Journal of All India Institute of Speech and Hearing, an annual publication, carries articles pertaining to the normal and abnormal processes and disorders of Language, Speech and Hearing. Contribution may take the form of reports of experimental studies, theoretical papers review papers and case reports. Manuscript already published elsewhere will not be accepted for publication in the journal. Manuscript submitted will be reviewed by the Editorial Board on the recommendation of the Editorial Board, author/s may be asked to revise the articles. Articles not accepted will be returned to the author. The decision of the Editorial Board shall be final. All Editorial Correspondence should be addressed to the Editorial-chief, JAIISH, All India Institute of Speech and Hearing, Mysore-570006, India. Subscription: Individual Subscription Rs.250/- per year. Institutional subscription outside India $100. Individual subscription $40. Copy right: All India Institute of speech and Hearing, Mysore. Opinion expressed in the articles rest with the authors and publishers are not responsible for it. Dr. Vijayalakshmi Basavaraj Director & Editorial Chief

JAIISH
ISSN 0973-662X Vol. 27, 2008

Journal of All India Institute of Speech & Hearing


An Official Publication of

ALL INDIA INSTITUTE OF SPEECH AND HEARING


Manasagangothri, Mysore 570 006 Phone: 0821 2514449, 2515218, 2515905 Fax: 0821 2510515 Website: www.aiishmysore.com, e-mail: aiish_dir@yahoo.com

Editorial
Greetings from AIISH, Mysore! Journal of All India Institute of Speech and Hearing (JAIISH) was resumed as an indexed journal with the ISSN No.0973-662X in the year 2007 by bringing out its 26th volume. I am very happy that the institute is able to bring out the 27th volume, which is the second volume of the indexed journal on time in the year 2008. Vol. 26 of Journal of AIISH received excellent feedback and our subscriptions increased. This is very encouraging. I will look forward to continued feedback in the future as well. Vol. 27 of Journal of AIISH carries very interesting nineteen articles in the area of speech, language and hearing. I am happy to share that special educators, otolaryngologists, apart from audiologists, speech and language pathologists have contributed to this volume. Seven articles under the category of Speech cover topics in the area of speech sciences as well as in the area of speech pathology. Eight articles under Language focuses on issues related to semantic and pragmatic issues; phonological processing; autism spectrum disorders among others. The four articles under Hearing cover topics on neonatal hearing screening, regenerative Myringoplasty, sudden sensory neural hearing loss, auditory memory and sequencing. It is disappointing to note that not many articles in the area of hearing and hearing disorders were received. I look forward to receiving more articles in these areas for the next volume. The institute is aspiring to expand its multi-disciplinary team activities. Such an expansion will be supported if we receive papers from the members of the multidisciplinary team dealing with communication disorders. Apart from the designated ten editorial members of the journal, three guest editorial members contributed in reviewing the articles. Their contribution is sincerely acknowledged. My sincere acknowledgements are also due to all the members of the Editorial Board for meeting our deadlines in reviewing the articles. Special appreciation to Dr. N Sreedevi, Lecturer in Speech Sciences, Department of Speech Language Sciences for her efficient follow up work as well as for proof reading the final version. As I had mentioned in my previous editorial, AIISH, for the first time, published full length papers based on the dissertations of the years 2002-03 and 2003-04. I am happy to share with you that the 3rd volume covering the dissertation articles of the year 2004-05 was released on 9th August 2008. Since specialized masters programs in Audiology and Speech Language Pathology were introduced in the year 2003-04, the dissertation articles have been compiled separately for Audiology and Speech Language Pathology topics as Vol. III Part A and Vol. III Part B respectively. The JAIISH Volume will have a section on book reviews and a section on letters to the editor. I request you all to contribute to these sections. I look forward to your continued support in contributing your valuable research publications in the Journal of AIISH. You may please email your suggestions in improving the standard of the journal to aiish_dir@yahoo.com.

Dr.Vijayalakshmi Basavaraj Director & Editorial-Chief

Table of Contents
Speech
1. 2. 3. Effect of Vowels on Consonants in Nasalence Gopishankar R. & Pushpavathi M. Investigation into Voice Source of Monozygotic Twins using Formant based Inverse Filtering Jayakumar T. & Savithri S.R. Effect of Spectral Variation on Phoneme Identification Skills in 2 - 3 year old Typically Developing Children Powlin Arockia Catherine & Savithri S.R. Keratosis of the Larynx Rajasudhakar R. & Sundara Raju H. Nasalence Value for Rainbow Passage: Normative Data for Non- Native Speakers Sangeetha M. & Pushpavathi M. Rate of Speech/Reading in Dravidian Languages Savithri S.R. & Jayaram M. Speech Rhythm in Hearing-Impaired Children Savithri S.R., Ruchi Agarwal & Johnsi Rani R. 3-7 8-14 15-18

4. 5. 6. 7.

19-21 22-28 29-39 40-43

Language
8. 9. The Semantic Association in the Mental Lexicon Gopee Krishnan & Shivani Tiwari Semantic Pragmatic Attributes and Cognition in Acute and Chronic Schizophrenics: A case comparative study Mithila Poonacha, Shivani Tiwari & Rajashekhar Bellur Pragmatic Skills in Typically Developing Infants Shilpashri H.N. & Shyamala K. Chengappa Pragmatic Skills in Nonverbal Identical Twins with Autism Spectrum Disorders Shilpashri H.N. & Shyamala K. Chengappa Continuum of Developmental Language Disorders: Where Does PLI Fall? Shivani Tiwari, Ashwini Bhat & Rajashekar Bellur Analysis of Oral and Written Narratives of Children with Language Impaired Learning Disabilities Siddique Tehniat & Mukhopadhyay Sourav Phonological Processes in Typically Developing Kannada Speaking Children Sreedevi N. & Shilpashree H.N. A Profile of Aetiological Therapeutic Searches by Netizen Parents/Caregivers of Children on the Autism Spectrum Venkatesan S. & Purushotham K. 44-48 49-53

10. 11. 12. 13.

54-57 58-63 64-73 74-82

14. 15.

83-88 89-94

Hearing
16. 17. 18. Auditory Memory and Sequencing in Children aged 6 to 12 years Devi N., Sujitha Nair & Asha Yathiraj Regenerative Myringoplasty A Case Report Rajeshwari G. & Sundara Raju H. Guidelines to Establish a Hospital Based Neonatal Hearing Screening Program in the Indian Setting Ramesh A., Nagapoornima M., Srilakshmi V., Dominic M. & Swarnarekha Reversible Sudden Sensory Neural Hearing Loss A Case Report Sundara Raju H. & Rajeshwari G. 95-100 101-103 104-109

19.

110-113

JAIISH, Vol. 27, 2008

Effect of Vowels on Consonants in Nasalence

Effect of Vowels on Consonants in Nasalence


1

Gopi Sankar R. & 2Pushpavathi M.

Abstract
The present study investigated the mean nasalence value of three isolated vowels and explored the nasalence value across CV combinations based on various place of articulation of consonants. The subjects consisted of fifty (21 males, 29 females) normal young adults in the age range of 18 to 27 years. The subjects were instructed to repeat the isolated vowels and CV combinations. The mean nasalence value was calculated. Repeated measures of ANOVA were used to find the significant difference in within and across the condition (CV combination). The results indicated significant difference across vowels with the high nasalence value for the high front vowel / i / followed by / a/ and /u/. Unvoiced bilabial and retroflex stop consonants with / i / had high nasalence value followed by /u/ and /a/. This results support the finding that high front vowel have significantly higher nasalence value than other vowels. This result also aids the speech pathologists to develop the stimuli for assessing the Velopharyngeal closure.
Key words: Nasalence, Vowels, Consonant vowel combination, Vowel effects.

Speech is a fleeting event. Researchers and clinicians strive to capture the speech signals and to analyze the same using the sophisticated methods. There is considerable information available concerning the acoustic characteristics of abnormal and normal resonance, as well as clinical assessment and management of resonance impairments. Nasalence is intended to be a measure of the acoustic energy that occurs primarily on vowels, glides and liquids. Traditionally, clinicians have used long passages, such as the Zoo Passage, rainbow passage to assess nasalance with the Nasometer. Shorter stimuli have been proposed (MacKay and Kummer, 1994; Watterson T, Hinton J & McFarlane S 1996; Awan, 1998) to measure the nasalence. But short stimuli create the potential for vowel and consonant content to have a weighting effect on the nasalance value (Karnell, 1995; Watterson T, Lewis KE & Foley-Homan N 1999). Because the Nasometer is designed primarily to measure the acoustic energy in vowels, the vowel content of the short stimulus would be of particular concern (Fletcher SG, Adams LE, & McCutcheon MJ. 1989). Most of the studies in nasalence measurement are focused on measuring and comparing the nasalence for high pressure and low

pressure consonants. In recent years, growing evidence has evolved concerning the relation between nasalence measurement and velopharyngeal closure specifically on vowels. Variation in the nasalence during the nasal airflow is closely related to the velar height and velopharyngeal closure. Carney and Sherman (1971) studied the effects of three speech tasks upon the perception of nasality for 10 normal subjects and 10 subjects with cleft palate. The three speech tasks consisted of the production of five isolated vowels, same vowels in consonantvowel-consonant (CVC) syllables and same CVC in connected speech passage. The results indicated that for both groups, CVC syllables from a connected speech are judged to be less nasal than either isolated vowels or isolated CVC syllables. The variations in results were attributed to coarticulatory influences. Subjects with cleft palate are more nasal on high vowel than on low vowels, while subjects without cleft palate were more nasal on low vowels than on high vowels. MacKay and Kummer (1994) provided data that supported the contention that nasalance values from short stimuli may be markedly influenced by vowel content. For the Simplified Nasometric Assessment Procedures Test (SNAP Test), MacKay and Kummer (1994)

1 Research officer, Dept. of Clinical Services, All India Institute of Speech and Hearing, Manasagangothri, Mysore570006, email:sankaj2002@yahoo.co.in, 2Reader in Speech Pathology, All India Institute of Speech and Hearing, Manasagangothri, Mysore- 570 006, email: pushpa19@yahoo.co.in.

JAIISH, Vol. 27, 2008

Effect of Vowels on Consonants in Nasalence

provided mean nasalance data for normal subjects using a variety of stimuli. The syllable repetition subtest requires subjects to repeat a CV syllable 6 to 10 times (e.g., ti-ti-ti), and data were provided for CV stimuli that differ only with respect to the vowel. The data showed that nasalence values for stimuli with the high front vowel /i/ were markedly higher than nasalance values obtained from stimuli with the low back vowel /a/. According to the authors, individual consonant environments (i.e., voicing, manner, and place) exerted different influences from vowel to vowel, where voicing produced the greatest effects on nasal perception. Vowels in voiced environments, and fricative environments were found to be longer in duration, lower in fundamental frequency, and greater in intensity than vowels in voiceless or plosive environments. The perception of nasality increased when these acoustic correlates (i.e., longer duration, lower fundamental frequency, and higher intensity) accompanied the phonetic context. Results indicated that perception of nasality followed this progression from least to most: (a) voiceless plosive environments /p, t/, (b) voiceless fricative /s, f/ and voiced plosive environments /g, d/, and (c) voiced fricative environments /v, z/. Overall, tongue height and voicing were found to have the most significant influence on the perception of nasality (Lintz & Sherman, 1961). In another study, Watterson T, Lewis KE & FoleyHoman N (1999) compared nasalance values for 17-syllable passage, 6-syllable sentence, and 2syllable word from a standard 44- syllable passage. The results showed that the longer the stimulus, the stronger the association with the standard passage. The shortest stimulus (two-syllable word) had insufficient criterion validity to warrant its use in clinical applications; however, the authors expressed concern that the vowel content might unduly influence the nasalance value in such a short stimulus. Kerry, L, Watterson, T, & Terasa ,Q (2000) compared the nasalence values with nine different speech stimuli with vowel content controlled. The subjects were 19 normal children and 19 children with velopharyngeal dysfunction. The stimuli consisted of nine speech stimuli which included four vowels in isolation and five sentences which were loaded with high front, high back, low front and low back vowels and one sentence with a mixture of vowel types, five sentences and four sustained vowels. The result showed that high vowels were associated with significantly higher nasalence values than low vowels for both sentence and sustained vowels.
4

For the velopharyngeal dysfunction (VPD) group, nasalence values for high vowel sentences and mixed vowel sentences were significantly higher than the nasalence value for the low vowel sentences. In both groups, nasalence values for sustained vowels were significantly higher for the high front vowel /i/ than for the other vowel .Difference was evident among front / back vowel contrasts. Nandurkar (2002) studied the Nasalance measures in Marathi consonant-vowel-consonant syllables with pressure consonants produced by children with and without cleft lip and palate. The results indicated differences between groups. As the nasalence value may be markedly affected by the vowel, it is necessary to determine the specific influence of various vowels in CV combination. Hence the present study investigated the mean nasalence value of the three isolated vowels and explored the nasalence value across CV combinations.

Method
Subjects: Fifty (21 males, 29 females) normal young adults with age range of 18 to 27 years (mean 19) participated in the study. All participants were judged by the investigators to possess speech and hearing within normal limits and reportedly were free from upper respiratory infection. None of the participants had a history of craniofacial anomalies or velopharyngeal impairment. Instrumentation: Nasometer model 6400 (Kay Elemetrics, New Jersey) was used to measure resonance using a lightweight headset made up of a harness that holds a (oral/nasal) separation plate. The separation plate was firmly fitted against the area between the nose and the upper lip and had two directional microphones mounted on either side of it, which collected the separated acoustic signals. The signals were transmitted to the computer database where they were calculated and analyzed by the Nasometer software. The resultant acoustic values were a ratio of nasal to nasal-plusoral acoustic energy, which was multiplied by 100, and expressed as a nasalance. Prior to testing, the Nasometer was calibrated and disinfected in accordance with the procedures outlined in the instruction manual. Stimuli: The stimuli consisted of vowels /a/, /i/, and /u/, and CV syllables in which /p, t, k/ was paired.

JAIISH, Vol. 27, 2008

Effect of Vowels on Consonants in Nasalence

Procedure: Subjects were tested individually. They were seated comfortably in a chair. The Nasometer headset was positioned perpendicular to the facial plane and seated firmly against the upper lip. Subjects were instructed to sustain vowels and nasal consonants in isolation. For CV combination, the subjects were instructed to repeat a CV syllables 3 times (e.g., pa-pa-pa) at a normal speed. A single mean nasalance percentage or nasalance values for 3 repeated stimuli was computed. Repeated measures ANOVA were used to find the significant difference between conditions and CV combination.

Consonants Mean /k/ 21.67 /t/ 20.33 /p/ 18.67 Table 2: Mean values for consonants.

Results of repeated measures of ANOVA indicated significant difference between CV combinations. Table 3 shows F and p values for across the vowel and consonant combinations.
Across the vowel F value /pa/-/pi/ /pa/-/pu/ /pi/-/pu/ /ta/-/ti/ /ta/-/tu/ /ti/-/tu/ /ka/-/ki/ /ka/-/ku/ /ki/-/ku/ F(2,98)= 27.345 F(2,98)= 34.808 F(2,98)= 72 Across the consonants P F value value /pa/-/ta/ F(2,98) <0.00 >0.05 /pa/-/ka/ =2.481 1 /ta/-/ka/ /pi/-/ti/ F(2,98) <0.00 <0.00 /pi/-/ti/ =24.203 1 1 /ti/-/ki/ /pu/-/tu/ F(2,98) <0.00 >0.05 /pu/-/ku/ =3.125 1 /tu/-/ku/ P value

Results and Discussion


a. Mean nasalence value for vowels in isolation: Results showed that high front vowel /i/ had the highest nasalence followed by low mid vowel /a/ and high back vowel /u/. Table 1 depicts the mean and SD of nasalence. Mean S.D /a/ 25.56 14.17 / I / 36.84 20.25 /u / 19.70 17.62
Table 1: Mean and SD for vowels.

Table 3: F values on repeated measures of ANOVA.

Results of the repeated measures of ANOVA indicated significant difference between vowels. Vowel /a/ had significant lower nasalence scores compared to vowel /i/. (F (2, 98) =28.371, p<0.001). b. Nasalence value for the oral consonants across the vowels: results indicated higher nasalence value for /p/, /t/ and /k/ when followed by vowel /i/ compared to when followed by other vowels. Figure 1 shows the mean nasalence value for unvoiced stop consonants combined with vowels. Consonant /k/ had higher nasalence values compared to /t/ and /p/. That is, the nasalence value decreased as the place of articulation moved forward the oral tract.

The results indicated that front high vowel /i/ had significantly higher nasalence value compared to low mid vowel /a/ and high back vowel /u/. This supports the findings of Neumann and Dalston (2001), who reported the similar findings. This may be due to the articulatory postures assumed during the production of these vowels. The low mid vowel /a / is a open vowel which creates relatively little resistance to airflow out of the mouth. Therefore the maximum energy is transmitted through the oral cavity. But high vowels /i/ and /u / impose relatively high resistance to airflow. However, during the production of the /u/ the tongue is placed in close proximity to the velum. This placement may tend to dampen the velar oscillations and thereby reduce acoustic transfer. The results also support the findings of Moore and Sommers (1973) who reported the greater degree of nasality on high vowels as the high vowels make greater demand upon the valving function i.e higher points of posterior pharyngeal wall/ velar contacts, tighter velopharyngeal seals and greater velar excursion. Kendrick (2004) provided a physiological explanation for higher nasalence value on vowel /i/. He has suggested a strong effect of horizontal position of the tongue on the nasalance of vowels. Back vowels are reported to have lower nasalance values because some of the muscles that pull the body of the tongue back also pull the velum down
5

Figure 1: Mean scores for CV combination.

JAIISH, Vol. 27, 2008

Effect of Vowels on Consonants in Nasalence

securing a tight closure between the two structures. To keep the velum from lowering during vowel production, the muscles that elevate the velum may be more active during back vowel production than front vowel production to counteract the downward force of the muscles pulling the tongue back. The production of the higher vowel requires the positioning of the velum in high position making the tight velopharyngeal closure. This is a feature of normal speech production. Mc Donald and Baker (1951) suggested that the correlation might be due the speaker's efforts to maintain a "characteristic balance or ratio between oral and nasal resonance." This resonance ratio presumably depends on the relative sizes of the velopharyngeal port and the posterior opening into the oral tract. Hence, when the speaker intends to produce no audible nasal output, a lower velum is tolerated for an open vowel than for a close vowel. However, the results do not partially support the findings of Lintz and Sherman (1961) who found that the perception of nasality increased as tongue height decreased during sustained vowel production (i.e., low vowels were perceived as more nasal than high vowels for normal speakers). This may be due to methodological difference as they used perceptual judgment and the subjects were children whereas the present study used an objective evaluation and the subjects were adults. The present study is the first attempt to explore the co articulation effect based on nasalence measures in consonant- vowel context. Most of the studies which are cited in the literature are based on perceptual measurement are measuring only in sentences or words. Since coarticulation effects allied with perceptual phenomena operating both forward and backward in time are known to cause interactions between adjacent phones, it is possible that a similar interaction could also be observed in consonantvowel syllables. Bell-Berti, F., Baer, T, Harris, K. S and Niimi, S (1979) have shown that the effects of vowel height on velar height extend into adjacent consonants. Alternatively, as Ackerman (1935) have suggested, movements of the larynx and pharynx may determine velar position through connections provided by the palatopharyngeus muscles. From the results of this study it can be speculated that tongue position had the greatest influence on nasalance values during sustained vowel production. If the tongue was in an elevated and retracted position, as was on the vowel / u /,
6

the velum achieved increased velar elevation and tighter VP closure, resulting in lower nasalence values for the normal speaker. The palatoglossus muscle, which is involved in tongue and velar functions, is active in achieving a front tongue position and at the same time pulls downward on the velum. This would result in less velar elevation, loose VP closure, and in turn higher nasalence values. Previous research has demonstrated that tongue height during vowel production significantly influenced nasalence, and the results from this study were in agreement with the findings of MacKay & Kummer (1994), Kuehn & Moon (1998), Lintz & Sherman, (1961).The results of the present study indicated that unvoiced consonants do not influence the nasalence value. However, vowels play a major role in nasalence values.

Conclusions
The results of the study showed that nasalence values are vowel dependent. High front vowels had significantly higher nasalence value than other vowels. This data also helps the speech pathologists to develop the stimuli for assessing the Velopharyngeal closure which is very important for determining the nasalence value.

References
Ackerman, E. L. (1935). Action of the velum palatinum on the velar sounds /k/and /g/, Vox 31, 2-9. Cited in ArthurS, Abramson, Patrick W. Nye, Janette B. Henderson and CharlesW. Marshall (1981).Vowel height and the perception of consonantal nasality. Journal of the Acoustic Society of America, 70(2), 329-339. Aparna Nandurkar (2002).Nasalance measures in Marathi consonant-vowel-consonant syllables with pressure consonants produced by children with and without cleft lip and palate. The Cleft palate-Cranofacial Journal, 39(1), 59-65. Awan SN(1998). Analysis of nasalance: NasalView (the nasalance acquisition system). In Zigler W, Deger K, eds. Clinical Phonetics and Linguistics. London:Whurr, 519527.

Bell-Berti, F., Baer, T, Harris, K. S and Niimi, S (1979). Coarticulatery effects of vowel

JAIISH, Vol. 27, 2008

Effect of Vowels on Consonants in Nasalence

quality on velarfunction, phonetica, 36, 187193. Carney & Sherman (1971).sevearity of nasality in three selected speech tasks. Journal of Speech and Hearing Research, 14,396-407. Fletcher SG, Adams LE, & McCutcheon MJ. Cleft palate speech assessment through oral nasal acoustic measures. In: Bzoch KR, ed. CommunicativeDisorders Related to Cleft Lip and Palate. Boston: Little Brown; 1989:246257. Karnell M P(1995). Nasometric discrimination of hypernasality and turbulent nasal airflow. The Cleft palate-Craniofacial Journal, 32,145148. Kendrick K.R (2004). Nasalance Protocol Standardization. Unpublished Master Thesis Submitted to the Graduate Faculty of the Louisiana State University and Agricultural and Mechanical College. Kerry,L, Watterson ,T, & Terasa ,Q (2000). The Effect of vowels on nasalence Values. The Cleft palate-Cranofacial Journal, 37(6), 584-589. Kuehn & Moon, (1998) Velopharyngeal Closure Force and Levator Veli Palatini Activation Levels in Varying Phonetic Contexts,Journal of Speech Language Hearing Research, 41, 51-62.

Lintz LB & Sherman D (1961). Phonetic elements and perception of nasality. Journal of Speech and Hearing Resarch, 4,381396. MacKay IR & Kummer A.W (1994). Simplified Nasometric Assessment Procedures. Lincoln Park, Kay Elemetrics, NJ. Mc Donald & Baker (1951) nasal air flow and nasal sound pressure level. In jhon hajek(Eds)universals of sound change in nasalization,pp.127-129.Boston:Blackwell. Moore & Sommers (1973) .Phonetic contexts: their effect on perceived nasality in cleft palate speakers. The Cleft palateCraniofacial Journal, 10, 72-83. Neumann & Dalston (2001). Nasalence Values in noncleft individuals: Why not zero? The Cleft palate-Craniofacial Journal, 38(2), 106-117. Watterson T, Hinton J & McFarlane S(1996). Novel stimuli for obtaining nasalance measures from young children. The Cleft palate-Craniofacial Journal. 33 ,6773. Watterson T, Lewis KE & Foley-Homan N (1999). Effect of stimulus length on nasalance values. The Cleft palate-Craniofacial Journal, 36, 243247.

JAIISH, Vol. 27, 2008

Voice Source of Monozygotic Twins

Investigation into Voice Source of Monozygotic Twins using Formant based Inverse Filtering
1

Jayakumar T. & 2Savithri S.R.

Abstract
Studying glottal flow gives potential benefit in many disciplines. Several methods have been developed for the estimation of the glottal flow. Glottal flow can be estimated from microphone pressure signal, or Inverse filtered signal. Inverse filtering (IF) has been used widely for the understanding of phonation type, intensity, voice quality, emotions and vocal loading. However there is a dearth of information about the effectiveness or reliability of IF techniques especially in the evaluation of individuals having similar voice and speech characteristics like monozygotic twins. Also, voice source through inverse filtering has not been investigated in twins so far. In this context, the present study investigated similarity of voice source in monozygotic twins using inverse filtering and the consistency of inverse filtered parameters. Two groups of females participated in the study. Group I had 6 monozygotic twins and Group II had 6, age and gender matched unrelated pairs. None of them had any voice disorders. Subjects Phonated vowel /a / three times at least for 5 seconds in comfortable pitch and loudness. Samples were audio-recorded at a sampling rate of 48 kHz and phase linear recording. Samples were inverse filtered using Vag_physio module of VAGHMI software in formant based method. Results showed that IF parameters were reliable over the repeated trials in all individuals. Also, ANOVA showed no significant difference between groups on voice source characteristics. The open quotient (OQ) and speed quotient (SQ) was significantly different across groups. However further investigation on twin pairs selection based on perceptual similarity and confirmed genetic analysis is warranted.
Key words: Glottal flow, Genetic similarity, Reliability.

The study of the glottal flow gives insight into the voice signal, which is of potential benefit in many disciplines such as speech synthesis, study of vocal expression of emotions, and clinical diagnosis and treatment of the voice. Due to the location of the larynx, (surrounded by many sensitive and vital organs and arteries), glottal flow is difficult to measure directly. Hence, several methods have been developed for the estimation of the glottal flow. They typically use the fundamental assumptions of Fants source - filter theory. Although the source -filter theory was formally published in 1960 (Fant, 1960), Inverse filtering (IF) was already presented by Miller a year earlier (Miller, 1959). Using inverse filtering can be estimated the source of voiced speech and the glottal flow can be acquired by removing the effects of the estimated vocal tract and lip
1

radiation from a measured air-flow or pressure waveform (Airas, 2008). Two methods exist for the input signal in inverse filtering. Either a flow mask may be used to estimate the actual air-flow out of the mouth (Rothenberg, 1973) or microphone at a certain distance may be used to measure the speech pressure signal (Anathapadmanabha, 1984). If absolute flow value and measurement of the minimum flow are required, a calibrated flow mask has to be used. However, flow masks have poor frequency responses (linear only up to 1.6 kHz to 9 kHz), and positioning the mask tightly around the mouth and the nose poses restriction on natural production of speech (Rothenberg, 1977). In contrast, good low frequency response microphone placed at constant distance from the speaker may overcome disadvantages of mask.

Junior Research Fellow, Dept. of Speech Language Sciences, All India Institute of Speech and Hearing, Mysore570006, email: jayakumar82@gmail.com, 2 Professor, Dept. of Speech Language Sciences, All India Institute of Speech and Hearing, Mysore-570006, email: savithri_2k@yahoo.com.

JAIISH, Vol. 27, 2008

Voice Source of Monozygotic Twins

The amplitude and phase response characteristics of Condenser microphones are excellent and will not affect natural speech production. Due to these reasons, microphone recordings are widely used (Airas, 2008). Inverse filtering was used widely for different phenomena of voice production concentrating on issues like phonation type (Alku, Vilkman, 1996), intensity (Dromey, Stathopoulos, Sapienza, 1992), voice quality (Gobi, NiChasaide, 2003), emotions (Airas, Alku, 2006), pitch, (Price, 1989) and vocal loading (Vinnuri et al, 2001). In addition some studies have discussed inverse filtering from methodological point of view (Alku, Vilkman, Laukkanen, 1998). Given the prevalence of IF in the field of voice science, there is dearth of information about the effectiveness or reliability and sensitivity of the IF technique especially in the evaluation of individuals having similar voice and speech characteristics like monozygotic twins. Monozygotic twins resemble each other in many aspects like aptitude, habit, taste and style that constitute what we think of as human individuality (Gedda, Fiori & Bruno, 1960). It may be hypothesized that their voice also may sound similar at least to a certain degree. It is generally accepted that the physical characteristics of the laryngeal mechanism, such as vocal fold length and structure, size and shape of the supraglottic vocal tract, and phenotypic similarities elsewhere in the vocal mechanism are genetically determined (Sataloff, 1997). Several research groups have studied genetic similarities in monozygotic twins. Though voice is unique to individuals, studies involving listeners perception have showed the perceptive similarity in monozygotic twins (Decoster, Van Gysel, Vercammen & Debruyne, 2001). Also, several quantitative measures like fundamental frequency in phonation (Przbyla, Hori, & Crawford 1992; Decoster, Van Gysel, Vercammen, & Debruyne 2001; Kalaiselvi, Santhosh & Savithri 2005), speaking fundamental frequency (Debruyne, Decoster, Van Gysel, & Vercammen 2002), formants (Forrai, & Gordos 1983) and Dysphonia Severity Index (Van Lierde, Vinck, De Ley, Clement, & Van Cauwenberge 2005) show similarity in monozygotic twins. However, voice source through inverse filtering has not been investigated in twins so far. In this context, the present study investigated similarity of voice source in monozygotic twins using inverse filtering, and consistency of inverse filtered parameters.
9

Method
Participants: Two groups of females participated in the study. Group I had 6 monozygotic twins and Group II had 6 age and gender unrelated pairs. All the subjects were between 19 to 25 years of age. Criteria for selecting the monozygotic twins included; (a) they should be same in gender, (b) should have the same blood group, and (c) should have approximately similar height and weight. Criteria for selecting the monozygotic unrelated pairs were: (a) non siblings of the same gender and (b) height should be approximately similar. None of the participants had any unstable voice, voice disorders, speech disorders, neuro-motor disorders, endocrinal disorders and/or hearing disorders. Recordings: The recording was made in quiet room. Participants were instructed to phonate vowel /a / three times at least for 5 seconds at comfortable pitch and loudness. Before the actual recording the Speech pathologist demonstrated the phonation. All samples were audio-recorded using Sony portable mini disk recorder MZ-R3 (Sony Corporation, Tokyo, Japan) at a sampling rate of 48 kHz and phase linear recording. Recording was made using Alcom-unidirectional microphone (frequency range from 40 Hz to 12000Hz ( 2dB) placed at a distance of 10 cm from participants. IF Procedure: The acoustic pressure waveforms were inverse filtered using Vag _ phsio module of VAGHMI software (Voice and Speech System, Bangalore, India). This program has two ways to obtain the glottal flow signal using IF- LPC analysis and formant frequency analysis. IF using Formant analysis gives clear glottal flow wave with out any high frequency ripples (ripples-free) compared to LPC based IF (Anathapadmanabha, 2008). Hence, in the present study, formant based IF was used to obtain the glottal flow wave. The edited downsampled phonation samples were fed in to IF. This software also has semiautomatic marking of the glottal flow wave to get the IF parameters. If semiautomatic marking fails to make decision user can switch to manual mode. Parameterization: The glottal flow waveforms estimated by the formant based IF were parameterized based on temporal [Open quotient (OQ), Speed quotient (SQ), Leakage quotient (LQ) and Pitch Period (T0)] and spectral [Roll-off, First Harmonic (H0), Harmonic ratio (H0-H1), EI/EE, & Dynamic leakage (AR)] measurement. Figure 1 shows the modeled volume velocity glottal pulse

JAIISH, Vol. 27, 2008

Voice Source of Monozygotic Twins

and its derivative in the time domain and figure 2 shows the log spectrum of voice source.

Harmonic ratio (H1-H0) is the ratio of energy at first harmonic and fundamental frequency. EI/EEis the mean ratio value of positive area and the negative area in each derivate cycle. Dynamic leakage (AR) is the residual flow during the return phase, which occurs from the time of excitation to the time of complete closure. Analyses: Phonation signals were recorded at a sampling rate of 48 kHz. To make the signal compatible with VAGHMI software Program (Voice & Speech Systems, Bangalore, India), signals were downsampled to 16 kHz using Wavesurfer software. The middle 3 second of each phonation sample was subjected to IF analysis. Vag _ phsio module of VAGHMI software Program was used for IF analysis. Each parameter was extracted 180 times each for group I and group II (6 * 2subjects * 3 trails * Five times): All analyses were made using semi-automatic marking methods. Whenever software failed to make mark, manual mode was selected for marking the glottal cycles. Ten present of the samples were subjected to test-re test reliability, which showed 89 % reliability. Statistical analysis: SPSS 10 was used to make the statistical calculations. Pearson product correlation was used to find the relation between the three trails measured within the subject. One way ANOVA was used to find the difference between the twin and co-twin as well as in unrelated pairs. Also the Absolute difference between the twin and co-twin as well as between the two participants in the unrelated group was calculated. From these values the statistical difference were made using one way ANOVA. Similarly the over all parameter difference was found.

0 glottal Onset, P peak flow, E Epoch, C Closure, T0-Pitch period, TP-Opening Interval, TNClosing Interval, TL -Leakage Interval, TC-Closed Interval. Figure 1: Modeled glottal pulse and its derivative in time domain.

Open quotient is defined as the ratio between the duration of glottal opening and the fundamental period (OQ = (TP+TN+TL) / T0. Speed quotient is defined as the ratio between the duration of for opening and closing of the glottis (TP/TN). Leakage quotient is defined as the ratio between TL and T0, here TL is the time taken for the voice source signal to return from epoch (E) to the baseline.

Results
Reliability over repeated trails To check the reliability of repeated trails the mean value of three trails was correlated over group I and group II. Tables 1 and 2 shows the rvalues of groups. r values suggest that there was not much variability among trails, except in few parameters.

Figure 2: Log Spectrum of single voice source.

Spectral roll-off indicates the smoothness of the glottal closure or the change in the spectral level over an octave change in the frequency.

10

JAIISH, Vol. 27, 2008

Voice Source of Monozygotic Twins

Parameters T0 OQ SQ LQ EIEE AR H0 H0-H1 Roll-off

1&2 0.892** 0.924** 0.704* 0.768** 0.727** 0.874** 0.380 0.803** 0.979**

Trial 1&3 0.977** 0.891** 0.838** 0.716** 0.916** 0.785** 0.429 0.875** 0.865**

2&3 0.948** 0.850** 0.871** 0.938** 0.741** 0.798** 0.892** 0.751** 0.924**

Table 1: r- values of group I. (** p< 0.01, * p< 0.05).

Trial 1&2 1&3 2&3 T0 0.980** 0.973** 0.968** OQ 0.773** 0.831** 0.752** SQ 0.774* 0.637* 0.671* LQ 0.482 0.706* 0.438 EIEE 0.414 0.724* 0.747* AR 0.774** 0.725** 0.768** H0 0.732** 0.711* 0.705* H0-H1 0.933** 0.904** 0.836** Roll-off 0.879** 0.865** 0.784** Table 2: r- value of group II. (** p< 0.01, * p< 0.05).

Parameters

Comparison with in group I & group II Results of one-way ANOVA revealed no significant difference within twins in all pairs in
T0 OQ SQ 3.81(.01)**.75(.01) 1.31(.06) 4.08(.02) .74(.02) 1.35(.25) 4.02(.06)** 72(.03)**1.81(.24) 4.83(.02) .75(.01) 1.81(.17) 4.53(.03)**.89(.07)**1.29(.16) 4.48(.05) .97(.01) 1.26(.06) 4.54(.25)**.86(.04)**1.88(.32) 4.17(.06) .94(.06) 1.74(.09) 4.02(.03) .93(.03) * 1.57(.16)** 4.0(.02) .81(.01) 2.16(.10) 4.58(.08)**.63(.05)**1.41(.17)** 4.46(.09) .77(.04) 2.10(.34) LQ .25(.04)** .19(.06) .23(.03)** .19(.05) .22(.05) .24(.02) .11(.03)* .13(.03) .20(.03)** .16(.01) .03(.01) .05(.02)

various parameters. Table 3 show mean, standard deviation of group I and Table 4 show mean, and standard deviation in group II.

Pair 1 Pair 2 Pair 3 Pair 4 Pair 5 Pair 6

EI/EE AR H0 Ho-H1 Roll-off 7.03(1.36)* .67(.02) 1.70(.29)**110.1(1.5) 9.7(3.21) .65(.07) 1.96(.15) 108.6(4.6) 11.5(6.33) 5.62(.09) .61(.04)**2.21(.14)**108(1.4)** 15.5(6.28) 5.53(.16)* .57(.01) 1.70(.38) 102 (7) 13.8(4.44) 5.65(.09) .54(.02)**1.35(.17)* 108.8(1.6)**11.07(1.40)**14.33(1.3)** .61(.05) 1.26(.07) 101.3(.6) 13.55(2.55) 12.36(.63) .39(.04)**1.96(.29)**101.(1.5)** 5.6(.59)** 7.17(.35)** .58(.06) 1.71(.21) 107.(1) 9.0(.83) 12.47(2.50) .50(.02)**1.56(.15)**100.3(.53)**9.47(.48) 15.36(.82)** .45(.01) 2.33(.10) 98.3(.22) 3.85(.18) 15.92(.48) .61(.04) 1.86(.11) 101.5(.53) 3.6(.51) ** 6.54(.74) 6.30(.58) .56(.09) 1.92(.18) 103.1(3.1) 8.3(1.2)

Table 3: Mean and SD of 1-3 twin pairs. (** p< 0.01, * p< 0.05)
T0
UPair 1 UPair 2 UPair 3 UPair 4 UPair 5 UPair 6 4.21(.04)** 4.31(.06) 4.33(.10)** 4.43(.07) 5.18(.11)** 4.87(.06) 4.57(.04)** 4.23(.11) 4.42(.07)* 4.66(.03) 5.18(.11) 4.29(.04)

OQ
.84(.01)* .89(.07) .81(.04)* .86(.05) .63(.08)** .74(.04) .79(.10)** .92(.08) .94(.02)** .80(.05) .63(.08)** .84(.01)

SQ
1.95(.09)** 1.55(.39) 2.93(.72)** 1.82(.14) 2.31(.67) 2.24(.16) 1.81(.33) 1.63(.33) 2.13(.22)* 2.39(.26) 2.31(.67) 1.95(.09)

LQ
.14(.01)** .14(.06) .06(.03)** .10(.03) .09(.06)** .02(.01) .09(.04)** .19(.06) .16(.03)** .10(.01) .09(.06)** .14(.01)

EI/EE
.49(.03)* .55(.16) .49(.16) .50(.08) .37(.03)** .52(.05) .39(.05)** .64(.08) .35(.04) .35(.02) .37(.03)** .49(.03)

AR
1.46(.10)** 1.67(.44) 2.41(.71)* 1.85(.13) 2.55(.64)** 2.09(.13) 2.08(.53)** 1.45(.18) 1.96(.21)* 2.19(.14) 2.55(.64) 1.46(.10)

H0
111.9(.72)** 106.1(2.5) 101.3(4.5)** 105.8(1.5) 99.(2.04)** 103.(1.06) 101.4(1.53) 101.2(2.19) 105.5(.86) 102.3(1.0) 99.5(2.0) 111.9(.72

Ho-H1
7.52(.49)** 10.7(1.04) 4.87(1.57)** 7.66(.89) 1.75(1.4)** 6.06(.45) 6.55(1.80)** 13.15(2.34) 6.71(1.24) 6.0(.50) 1.75(1.4) 7.52(.49)

Roll-off
6.11(1.03)** 10.5(.71) 6.07(.74)** 9.23(3.22) 5.66(.44)** 6.24(.24) 9.47(2.36)** 9.88(.83) 7.23(1.23) 8.27(.89) 5.66(.44) 6.11(1.03)

Table 4: Mean and SD of 1-3 group II. (** p< 0.01, * p< 0.05)

Comparison of IF parameters between the groups Absolute difference between twin, and cotwin was used to perform one way ANOVA. Twin, co- twin difference is the arithmetic differences between the two members of twin members. Similarly, pair, co-pair differences values where calculated for unrelated members.
11

Appendix I shows mean and standard deviation of difference values of twin pairs and unrelated pairs. Using the difference of twin, co twin and pair, co-pair values group comparison was made between group I and group II. Results indicated significant differences between groups on OQ, SQ, LQ EI/EE, AR, and H0. Table 5 shows the mean and SD in both groups.

JAIISH, Vol. 27, 2008

Voice Source of Monozygotic Twins

T0 OQ SQ LQ EI/EE AR H0 Ho-H1 Roll-off

Unrelated 0.33 (.28) 0.12 (.09) 0.62 (.47) 0.06 (.05) 0.14 (.10) 0.60 (.52) 5.79 (3.6) 4.52 (2.35) 2.18 (1.95)

Twins .28 (.29) 0.08 (.06)** 0.36 (.29)** 0.05 (.04)* 0.08 (.06)** 0.38 (.28)** 4.88 (3.6)* 4.05 (3.27) 1.78 (2.14)

quotient and speed quotient are parameter which is more similar in monozygotic twins pair compares all other inverse filtering parameters. It shows the opening phase and speed of vocal fold movement is similar in monozygotic group. OQ & SQ value gives the overall morphology of the glottal wave. In conclusion, IF parameters was reliable over the repeated trials in all individuals. Monozygotic twins showed considerable similarity in voice source on inverse filtering. Also majority of the parameters, specially OQ and SQ were sensitive enough to differentiate monozygotic twins voice. Further investigation on twin pairs based on perceptual and genetic analysis is warranted.

Table 5: Mean and SD of difference values in two groups. (** p< 0.01, * p< 0.05)

Discussion
Very few studies have investigated the consistency of inverse filtering parameters either in subjects with normal voice or in pathological voice. Few studies investigated the voice source but none of the published work had been done on analysis of IF parameters in monozygotic twins. Current study investigated the inverse filtering parameters of six monozygotic twins comparing with age and gender matched unrelated pair. The coefficient value suggested that there was good consistency between trails of individuals in both groups, and a good consistency of IF parameters over repeated trials. Secondly, within each twin (Group I), no significant difference was found on several parameters rather voice source similarity was very few in group II. Speed quotient was more similar in monozygotic twins pair compares all other inverse filtering parameters. These results were in hand with Van Lierde et al (2005). They investigated voice quality of 45 monozygotic twins using qualitative and quantitative assessment. The results showed similarity in laryngeal, aerodynamic measurement. The voice source similarity in group I can be attributed to would be physical characteristics of the laryngeal mechanism, such as vocal fold length and structure, size and shape of the supraglottic and vocal tract. Since twins are similar in genetically they have high similarity on voice source characteristics (Sataloff, 1997). Variability that is seen in the voice source of twins group can be due to variation in genetic similarity. That is some pairs have more genetically similar component than others. Using the absolute difference value the group comparison made between twin pairs and unrelated pairs. OQ, SQ, LQ, EI/EL, AR, HO were significantly different between groups. Open
12

References
Airas, M., (2008). TKK Aparat: An environment for voice inverse filtering and parameterization. Logopedics Phoniatrics Vocology, 33(1), 49-64. Airas, M., & Alku, P. (2006). Emotions in vowel segments of continuous speech: analysis of the glottal flow using the normalized amplitude quotient. Phonetica,63,26-46. Alku, P., & Vilkman, E., (1996). A comparison of glottal voice source quantification parameters in breathy, normal and pressed phonation of female and male speakers. Folia Phoniatrica Logopaedica,48, 240254. Alku, P., Vilkman, E., & Laukkanen, AM. (1998). Parameterization of the voice source by combining spectral decay and amplitude features of the glottal flow. Journal of Speech Language Hearing Research, 41, 990-1002. Ananthapadmanaba, TV. (1984). Acoustic analysis of voice source dynamics. STL-QPSR, 1-4. Ananthapadmanabha, T.V., (2008). Voice source characterization by Inverse Filtering Proceedings of the national workshop on voice: Assessment and Management, 1-25. Debruyne, F., Decoster, W., Van Gysel, A., & Vercammen, J. (2002). Speaking fundamental frequency in monozygotic and dizygotic twins. Journal of Voice, 16, 466471. Decoster, W. A., Van Gysel, A., Vercammen, J., & Debruyne, F. (2001). Voice similarity in

JAIISH, Vol. 27, 2008

Voice Source of Monozygotic Twins

identical twins. Acta Oto laryngological belg, 55, 49-55.

Rhino

Dromey, C., Stathopoulos, E., & Sapienza, C. (1992). Glottal airflow and EGG measures of vocal function at multiple intensities. Journal of Voice, 16, 44-54. Fant, G. (1960). Acoustic theory of speech production. The Hague, Netherland: Mouton. Forrai, G., & Gordos, G. (1983). A new acoustic method for the discrimination of monozygotic and dizygotic twins. Acta paediatrica Hungarica, 24, 315-321. Gedda, L., Fiori, R., & Bruno, G. (1960). Viox chez les jumeaux monozygotiques. Folia phoniatrica, 12, 81-94. Gobi, C., & NiChasaide, A., (2003). The role of voice quality in communicating emotion, mood and attitude. Speech Communication, 40, 189-212. Kalaiselvi, P. S., Santhosh, M., & Savithri, S. R. (2005). Multi dimensional analysis of voice in monozygotic twins. Journal of the Acoustical Society of India, C, 25-28. Miller, RL. (1959). Nature of the vocal Cord wave. Journal of the Acoustical Society of America,31, 667-77. Price, PJ. (1989) Male and female voice source characteristics: inverse filtering results. Speech Communication, 8, 261-277. Przbyla, B., Horii, T., & Crawford, M. (1992). Vocal fundamental frequency in a twin sample: looking for a genetic effect. Journal of Voice, 6, 261-266.

Rothenberg, M. (1973). A new inverse-filtering technique for deriving the glottal air flow waveform during voicing, Journal of the Acoustical Society of America, 53,1632-45. Rothenberg, M. (1977). Measurement of air flow in speech. Journal of Speech Hearing Research, 20, I55-76. Santosh, M., & Savithri, S. R. (2005). Voice prints in monozygotic twins. Journal of the Acoustical Society of India, C, 1-7. Sataloff, R. (1997). Professional Voice, the Science and Art of Cliniccal Care. Singular Publishing Group, San Diego: CA. Van Lierde, K. M., Vinck, B., De Ley, S., Clement, G., & Van Cauwenberge, P. (2005). Genetics of vocal quality characteristics in monozygotic twins: a multiparameter approach. Journal of Voice, 19, 511-518. Vinnuri, J., Alku, P., Lauri, ER,, Sala, E., Sihvo, M., & Vilkrnan, E. (2001) Objective analysis of vocal warm-up with special reference to ergonomic factors. Journal of Voice,15,36-53.

Acknowledgements
Authors would like to extend their gratitude to Dr. Vijayalakshmi Basavaraj, Director, All India Institute of Speech and Hearing for allowing us to carryout this study and also the co-operation of the participants.

13

JAIISH, Vol. 27, 2008

Voice Source of Monozygotic Twins

Appendix 1 Mean and SD of absolute difference values in both groups.


(U-unrelated pair, Pair twin pair)
T0 OQ SQ LQ EI/EE AR H0 Ho-H1 Roll-off U 1 & Pair 1 .09(.05) .27(.03) .07(.04) .02(.01) .50(.26) .20(.21) .05(.03) .07(.04) .15(.10) .06(.05) .34(.36) .27(.20) 5.85(2.29) 3.49(1.61) 3.18(1) 4.72(6.03) 4.40(.98) 1.41(1.35) U 2 & Pair 2 .13(.10) .81(.07) .05(.03) .04(.02) 1.11(.72) .24(.15) .03(.03) .07(.04) .16(.14) .05(.03) .79(.38) .49(.18) 5.75(1.03) 6.87(6.58) 2.90(1.50) 6.14(4.15) 3.58(2.9) .16(.11) U 3 & Pair 3 .31(.14) .06(.05) .10(.06) .08(.06) .68(.38) .16(.09) .07(.07) .07(.03) .15(.04) .07(.04) .75(.38) .16(.12) 4.32(1.73) 7.54(2.09) 4.30(1.51) 2.63(1.75) .62(.13) 2.27(1.46) U 4 & Pair 4 .33(.12) .37(.23) .15(.14) .11(.04) .44(.24) .25(.19) .10(.09) .03(.02) .24(.10) .18(.07) .68(.53) .39(.29) 3.18(1.19) 6.39(1.30) 6.60(2.53) 3.42(1.07) 2.19(1.01) 5.31(2.39) U 5 & Pair 5 .23(.06) .03(.03) .14(.05) .12(.04) .32(.30) .59(.16) .05(.03) .04(.03) .04(.04) .05(.02) .26(.26) .79(.18) 3.19(1.70) 2.09(.70) 1.37(.93) 5.61(.51) 1.23(.68) .91(.75) U 6 & Pair 6 .88(.12) .13(.11) .21(.09) .14(.05) .67(.37) .71(.33) .08(.02) .02(.01) .12(.04) .09(.06) 1.5(.10) .17(.11) 12.44(2.21) 2.93(1.86) 5.76(1.46) 4.75(.96) 1.03(.77) .62(.52)

14

JAIISH, Vol. 27, 2008

Voice Source of Monozygotic Twins

15

JAIISH, Vol. 27, 2008

Effect of Spectral Variation on Phoneme Identification Skills

Effect of Spectral Variation on Phoneme Identification Skills in 2 - 3 year old Typically Developing Children
1

Powlin Arockia Catherine S. & 2Savithri S. R.

Abstract
The present study investigated the ability of 2 to 3 year old Kannada speaking typically developing children to identify synthetic phonemes varying in second formant frequency (F2). Two picturable (minimal pair) words with stop consonants contrasting in place of articulation (labial /p/ and retroflex /t/) in word initial position in Kannada were selected. These words as uttered by 21-year-old female native Kannada speaker were recorded and stored onto the computer memory. Place of articulation continuum was prepared using Acophon 1 programme of SSL Pro3V3 software and tokens were audio recorded onto a CD. The synthesized tokens were presented to thirty (15 boys and 15 girls) 2 to 3 year old children individually. Subjects were instructed to point to the pictures placed before them as they listen to the token and the investigator noted their responses on a scoring sheet. Percent identification scores were calculated. The results indicated that the 50% crossover from labial to retroflex cognate occurred at 1692 Hz in children. Also, phoneme boundary width was wider in girls compared to that in boys. The results of the current study can be used to compare phoneme identification skills in clinical population of same age.
Key words: Identification, Synthesis, F2 transition.

Speech perception is the decoding and interpretation of speech by the listener. Research on speech perception seeks to understand how human listeners recognize speech sounds and use this information to understand spoken language. Studies on infant speech perception postulates that the ability to perceive universal phoneme contrast is present at birth and with exposure infants loose this ability and could perceive only the native contrasts (Werker & Tees, 1984). Also cross language studies on adults have demonstrated language specific perception patterns (Abramson & Lisker, 1970). Modification of perceptual abilities takes place between infancy and adulthood. It is important to investigate and document the modification process in phoneme perception during language development period, as it would strengthen our understanding of perception-production relationship. Of all phonemes, stop consonants are widely studied. They are produced by occluding the oral cavity and releasing the articulator after sufficient
1

air pressure is built up. Temporal and acoustic parameters cue voicing and place of articulation in stop consonants. Cues for place of articulation include burst spectrum at consonantal release (Cooper, Delattre, Liberman, Borst & Gerstman, 1952; Winitz, Scheib & Reeds, 1972) and onset frequency of second and third formants (Delattre, Liberman, Cooper, 1955). Potter, Kopp & Green (1947) stressed the importance of second formant (F2) transition to cue place of articulation and described various pattern of transitions. Rising F2 transitions are perceived as bilabials (/b/ and /p/), F2 slight fall for front vowels and sharp fall for back vowels are perceived as alveolars (/d/ and /t/) and sharp F2 fall for front vowels and slight fall of F2 for back vowels are perceived as velars (/g/ and /k/) (Borden & Harris, 1980). In the past, several investigations (Liberman, Delattre, Cooper and Gerstman, 1954; Liberman, Harris, Hoffman & Griffith, 1957; Sussman, 1993) have reported the importance of F2 onset in identifying and discriminating place of articulation

Junior Research Fellow, Dept. of Speech-Language Sciences, All India Institute of Speech and Hearing, Mysore, email:arockia_cath@yahoo.com, 2Professor, Dept. of Speech-Language Sciences, All India Institute of Speech & Hearing, Manasagangothri, Mysore 570 006, email:savithri_2k@yahoo.com,

16

JAIISH, Vol. 27, 2008

Effect of Spectral Variation on Phoneme Identification Skills

of stop consonants in adults and children. However, the F2 varies depending on the place of articulation of the stop consonant in a language. Also, most of the studies are in English. India being a multilingual country offers great potential for research in this area. While this area has received some attention at the international level, the scene at the national level is dismal. Although the sources of data in non-Indian languages are useful, there is a pressing need for experimental evidence in various Indian languages. Both crosssectional and longitudinal studies are needed with groups of children of sufficient size to allow at least preliminary generalizations about normative development. In this context, the present study investigated phoneme identification skills in typically developing Kannada speaking children between the age range of 2 and 3 years by systematically varying the F2 onset frequency.

module of Acophon 1 programme in SSLPro3V3 software. The word /pa: ta/ was analyzed with a block duration of 40 ms and block shift of 10 ms. Linear predictive coefficient was kept at 18 and pre-emphasis factor was 1. Following analysis, the F2 continuum was synthesized using edit option by selecting glottal voice source pulse shape. F2 onset frequency of word-initial vowel was varied from 1300 Hz to 2400 Hz in steps of 100 Hz keeping the other formants constant in the initial vowel. The F2 was interpolated between onset and onset of vowel steady state and the word was synthesized. Figure 1 illustrates F2 at mid and end points of the continuum. A total of 12 synthetic words for a /p-t/ continuum were generated. The synthetic words were iterated thrice, randomized and recorded onto a CD. Thus a total of 36 tokens (12*3) formed the stimulus.

Method
Subjects: Thirty typically developing, Kannada speaking children from four play schools in Mysore participated in the study. The subjects included 15 boys and 15 girls in the age range of 2 to 3 years. All the children were from middle socio-economic status. The children were formally screened for speech, language and hearing abilities by the experimenter and those who passed the screening were included in the study. Stimuli: Two meaningful, picturable, bisyllable words with stop consonants in the initial position in Kannada were selected. The word initial stop consonants in the word pair contrasted in place of articulation (p- t; pa:ta ta:ta). These words as uttered thrice by a 21-year-old female, native Kannada speaker were recorded onto a computer using SSL Pro3V3 software (Voice and Speech Systems, Bangalore) and stored onto the computer memory. From the wide-band bar type spectrograms with LPC analysis, the onset frequency of first three formants, F2 transition duration and F2 frequency of steady state of vowel for the initial vowel of the word pair was obtained. For /pa/, onset frequency of F1, F2 and F3 were 785 Hz, 1300 Hz and 2960 Hz, respectively and F2 transition duration was 37 ms followed by F2 steady state of 130ms. Formant frequencies of vowel steady state were 890 Hz (F1), 1547 Hz (F2) and 3030 Hz (F3). F2 continuum was prepared using formant based analysis by synthesis (FBAS)
17

Figure1: Illustration of F2 at 1300 Hz (a), 1800 Hz (b) and 2400 Hz (c).

Procedure: Children were tested individually. Each child was familiarized with the pictures of the words used in the experiment on the previous day of the data collection. Experiment was carried out in a quiet room. The child was seated comfortably in a chair. The stimulus was audiopresented through two speakers placed at 45o azimuth at a comfortable loudness. The child was instructed to carefully listen to the stimuli and point to the respective picture out of two picture cards placed before him/her. The investigator noted the child's responses on a scoring sheet. Percent identification scores were calculated and identification curves were drawn. Fifty percent crossover, lower limit of phoneme boundary (LLPB), upper limit of phoneme boundary (ULPB) and phoneme boundary width (PBW) were obtained (Doughty, 1949). Fifty percent crossover is the point at which 50% of the subjects response corresponds to the labial (alveolar) category. Lower limit of phoneme boundary width is the point along the acoustic cue

JAIISH, Vol. 27, 2008

Effect of Spectral Variation on Phoneme Identification Skills

continuum where an individual identified labial (alveolar) stop 75% of the time and upper limit of phoneme boundary width defined as the corresponding point of the identification of the labial (alveolar) cognate 75% of the time. Phoneme boundary width was determined by subtracting the lower limit from upper limit of boundary width. Identification data obtained from one child (B9) was not considered for analysis as it didnt show crossover.

Results
Result indicated that 50% crossover occurred at 1680 Hz. The mean LLPB and ULPB were at 1520 Hz and 1920 Hz, respectively. The mean PBW was 400 Hz. In girls 50% crossover, LLPB and ULPB occurred earlier compared to those in boys. PBW was wider in girls compared to that in boys. Also, the range of all the measures was wider in girls compared to those in boys. Results of independent t-test revealed significant gender difference on LLPB [t (27) = -2.176; p<0.05]. No significant difference was found across gender for 50% crossover [t (27) = -1.980; p>0.05], ULPB [t (27) = -1.754; p>0.05] and PBW [t (27) = 0.462; p>0.05]. Table 1 shows the mean values of all parameters in children. Figures 2 and 3 shows mean percent identification in boys and girls.
Parameters Boys Girls Average 50%crossover 1820 1620 1680 (Range) (590) (630) (650) LLPB 1630 1485 1520 (Range) (610) (650) (650) ULPB 1995 1880 1920 (Range) (470) (610) (660) PBW 365 395 400 (Range) (240) (460) (430) Table 1: Mean and range values of measures of /p- t/ continuum (in Hz).

Figure 3: Mean percent identification scores in girls.

Discussion
The results revealed interesting findings. First, it was found that F2 onset frequency was an important cue to identify the place of articulation of stop consonants (/p/ and /t/) in Kannada. Lower F2 onset was perceived as bilabial /p/ and the percept changed to retroflex /t/ at higher F2 onset frequencies. This finding is in consonance with Liberman, Delattre, Cooper & Gertsman (1954) and Liberman, Harris, Hoffman & Griffith (1957) in adult English speakers. Second, children shifted their percept from bilabial /p/ to retroflex /t/ at 1680 Hz. In English speakers the shift from bilabial to alveolar place occurred at a much earlier frequency, 1199 Hz for 4 year old children, 1120 Hz for 5 6 year old children and 1144 Hz for adults (Sussman, 1993). This difference in shift of percept can be attributed to differences in place of articulation of stop consonants or methodological difference. In the present study the extreme frequencies of the continuum were between 1300 Hz and 2400 Hz but in Sussmans study they were 543 Hz and 1620 Hz. Also, stimulus in the current study was bisyllable meaningful Kannada words uttered by 21 year old female native Kannada speaker and F2 continuum was prepared by manipulating the F2 onset in initial vowel using FBAS module of Acophon 1 programme in SSLPro3V3 software. In Sussmans study the stimulus was synthetically generated using Klatt synthesis package using digital signal processing board (Data Translation model 2821). Third, mean phoneme boundary width was 400 Hz in children and it was wider in girls (395
18

Figure 2: Mean percent identification scores in boys.

JAIISH, Vol. 27, 2008

Effect of Spectral Variation on Phoneme Identification Skills

Hz) compared to boys (365 Hz). This can be attributed to more number of girls having wider PBW compared to that in boys. To conclude, typically developing Kannada speaking children in the age range of 2-3 years were able to shift their percept from bilabial /p/ to retroflex /t/ when F2 onset frequency increased. Future research on older age groups and adults to study the developmental pattern of speech perception in Kannada and other languages are warranted.

Delattre, P. C., Liberman, A. M., & Cooper, F.S. (1955). Acoustic loci and transitional cues for consonants. The Journal of the Acoustical Society of America, 27, 769-773. Doughty, J. (1949). In Zlatin, M., & Koenigsknecht, R. (1975). Development of the voicing contrast: Perception of stop consonants. Journal of Speech and Hearing Research, 18, 541-553. Liberman, A. M., Delattre, D. C., Cooper, F.S., & Gerstman, L.J. (1954). The role of consonant-vowel transitions in the perception of the stop and nasal consonants. Psychological monographs, 68, 1-13. Liberman, A. M., Harris, K. S., Hoffman, H. S., & Griffith, B. C. (1957). The discrimination of speech sounds within and across phoneme boundaries. Journal of Experimental Psychology, 54, 358-368. Potter, R. K., Kopp, G. A., & Green, H. (1947). Visible speech, NewYork: Van Nostran. Sussman, J.E. (1993). Perception of formant transition cues of place of articulation in children with language impairments. Journal of Speech and Hearing Research. 36, 1286-1299. Werker, J. H., & Tees, R.C. (1984). Crosslanguage speech perception. Evidence for perceptual reorganization during the first year of life. Infant behavior and development, 7, 49-63. Winitz, H., Scheib, M. E., & Reeds, J. A. (1972). Identification of stop and vowels for the burst portion of /p,t,k/ isolated from conversational speech. The Journal of the Acoustical Society of America, 51, 13091317.

Conclusions
The present study provides data on phoneme identification skills in 2-3 year old Kannada speaking typically developing children. The phoneme identification skills of normal children can be compared with clinical population like latetalking children, children with hearing impairment, mental retardation, seizure disorder and high-risk children. More specifically, late talking children between 2 to 3 years who exhibit language disorder in the absence of specific causes may be impaired in phoneme identification. Also, using the findings of this study as baseline, perception training program for children in this age group can be devised.

References
Abramson, A., & Lisker, L. (1970). Discriminability along the voicing continuum: Cross-language tests. Proceedings of the sixth International congress of phonetic sciences, Academia, Prague, pp 569-573. Borden, G. J., & Harris, K. S. (1980). Speech science primer- Physiology, acoustics and perception of speech. Williams & Williams, Baltimore, USA, 171-214. Cooper, F. S., Delattre, P.C., Liberman, A. M., Borst, J., & Gerstman, L. J. (1952). Some experiments on the perception of synthetic speech sounds. The Journal of the Acoustical Society of America, 24, 597-606.

Acknowledgements
This study is an outcome of the doctoral research of the corresponding author. The authors are thankful to Dr. Vijayalakshmi Basavaraj, Director, All India Institute of Speech and Hearing for granting permission to carry out this study.

19

JAIISH, Vol. 27, 2008

Keratosis of the Larynx

Keratosis of the Larynx


1

Rajasudhakar R. & 2Sundara Raju H.

Abstract
Laryngeal keratosis is a rare condition characterized by irregular areas of thickening and cornification of the laryngeal mucous membrane. Much controversy exists regarding its etiology, treatment and terminology. Some authors interchange the term keratosis laryngis with that of leukoplakia of the larynx. There are however, others who believe that these terms are not inter-related. This paper made an attempt to describe some of the characteristic features of the disease by describing a case report to emphasize the acoustic parameters of voice before and after the medication.
Key words: Keratosis, Leukoplakia, Medication. Case history, Acoustic parameters,

Hyperkeratosis of the larynx is a localized form of epithelial hyperplasia characterized by white leucoplakic raised patches on the vocal folds. It is considered a clinically premalignant lesion that develops into invasive carcinoma in about 7-20% of cases (Garca, Aranzbal, Salas, Olano & Guimera, 1996). The condition is rare and occurs more commonly in men. Although its causation is often unclear, sometimes it resembles and is associated with some chronic inflammations of the larynx. There is a hyperplastic change in the epithelium, leading to excessive cornification, together with extension of the papillae into the corium, the basement remaining intact. Etiology Little is known about the cause of this disease although many theories have been postulated. Some of them include: (a) use of tobacco, (b) excessive use of alcohol, (c) vocal abuse, (d) chronic postnasal discharge with laryngeal irritation, (e) Syphilis, (f) dietary deficiencies with low vitamin A and B intake, (g) virus infection, and (h) mycosis. Signs and symptoms Some of the clinical features include: (a) hoarseness which is gradual in onset and persistent in nature, (b) occasionally associated with cough, dyspnea and stridor, (c) white raised patches that appear on one or both vocal cords. The anterior and middle thirds are usually involved. These

patches may involve one side only, but more often they are irregularly bilateral in distribution. There is no ulceration but one may observe strict demarcation between normal mucosa and the borders of involved tissue. Mobility of the cords is not impaired. The condition is considered precancerouos and carcinoma in situ frequently supervenes. Histopathologically, one finds the laryngeal epithelium thickened, elevated, and keratinized but the basement membrane is unchanged. Treatment Both medical and surgical techniques are usually employed. Intensive vitamin A, B and C therapy is prescribed and cures have been described following their employment, in conjunction with laryngeal stripping procedures. Simpson, Robin, Ballantyne, & Groves (1967) reported that vitamin A control the rate of growth of epithelial structures. But, it tends to persist in spite of conservative treatment. Complete voice rest along with exclusion of alcohol, tobacco, and spicy food is considered useful. Studies on the voice parameters in laryngeal keratosis are limited. This study describes the acoustic characteristics of voice in an individual with hyperkeratosis of the larynx before and after medication. Case history A 52 years old male reported to AIISH clinic with the complaint of pain in throat for fifteen

1 Junior Research Fellow, Dept. of Speech-Language Sciences, All India Institute of Speech & Hearing, Manasagangothri, Mysore-570006, email: rajasudhakar82@yahoo.co.in, 2Reader in ENT, All India Institute of Speech & Hearing, Manasagangothri, Mysore-570006.

20

JAIISH, Vol. 27, 2008

Keratosis of the Larynx

days. He reported of pain while speaking and was unable to increase loudness. Excessive cough was also reported for fifteen days. The severity of the problem was reportedly more in morning. He had no difficulty in swallowing solids or liquids. Voice evaluation Quantitative analysis The quantitative analysis of voice was done by using Vaghmi software (VSS, Bangalore), MDVP (Kay Elemetrics, New Jersy) and Electroglottograph (EGG). The results indicated higher habitual frequency in phonation, reading and speaking, reduced frequency and intensity range. Abnormal extent and speed of fluctuations in intensity and frequency, reduced maximum phonation duration in condition 1 (before medication) compared to condition 2 (after medication). MDVP showed abnormality in frequency, intensity and perturbation related measures in condition 1 compared to condition 2. Table 1 shows the measures in both conditions. Figure 1 shows the MDVP results in condition 1 (a) and condition 2 (b).
Parameters Fundamental frequency (Fo) /a/ /i/ /u/ Reading Fo Speaking Fo Frequency range Intensity range Extent of Fluctuation in -Intensity -Frequency Speed of Fluctuation in -Intensity -Frequency Maximum phonation duration /a/ /i/ /u/ /s/ /z/ EGG measures -Open Quotient (OQ) -Closed Quotient (CQ) MDVP Condition 1 135 Hz 148 Hz 136 Hz 145 Hz 140 Hz 108-216 Hz 88-105 dB 2.52 dB 3.19 Hz 4.50 % 12.5 % 15 sec 12 sec 14 sec 12 sec 10 sec 59 % 41 % frequency, intensity and perturbation related measures affected Condition 2 124 Hz 128 Hz 127 Hz 130 Hz 126 Hz 102-228 Hz 70-110 dB 1.25 dB 2.51 Hz 2.31 % 3.87 % 17 sec 15 sec 13 sec 12 sec 12 sec 54.11 % 45.89 % Normal

Qualitative analysis The qualitative analysis revealed high pitch, soft severe hoarseness voice in condition 1. The diagnosis was clinically normal voice in condition 2. Laryngeal examination Endoscopy revealed the presence of white patches on both vocal folds. Figure 2 (a) shows the endoscopic image of Keratosis larynx. The white patches were

(a)

(b)

Figure (a) Results one-third region of condition seen in1:the anteriorof MDVP parameters inthe vocal 1; and in medication like antibiotics folds. Conservative condition 2; (b) (beyond green circle indicates abnormal). (Clamp, 625 mg for five days), anti-inflammatory

analgesic (Emanzen D, for seven days) and Cobadex CZS for fifteen days was prescribed along with voice rest.

(a)

(b)

Figure 2: a) view of the Keratosis larynx in condition 1; b) view of larynx in condition 2 (aftermedication).

After fifteen days of medication, the case came for re-evaluation. The quantitative and qualitative voice parameters were measured in condition 2 which revealed that the acoustic parameters improved. Also, laryngeal endoscopy was done where white patches were not seen on the vocal folds. Figure 2 (b) shows the endoscopic image of larynx after medication.

(Condition 1 = before medication; Condition 2 = after medication) 21

JAIISH, Vol. 27, 2008

Keratosis of the Larynx

Discussion
The results revealed several points of interest. First of all, Fo was high in condition 1 compared to condition 2. Owing to the keratosis in the anterior portion of the vocal folds, the patient may be using only the posterior part of the vocal folds for vibration resulting in high Fo in condition 1. Reduced MPD and increased open quotient indicate that the vocal folds were open for longer time in condition 1 compared to condition 2. That is probably, the vocal folds did not close completely or the closure was not symmetrical as indicated by abnormal perturbation values. All these can be attributed to the presence of keratosis. Grossman & Mathews (1976) quoted Reimann's theory that keratosis is "a disease of shedding of the superficial layers. These cells do not differentiate as do normal cells, but stick together, pile up and produce the islands called as keratinized because of an excess of keratin in the cell". In this case, the disease may be in the initial stage that may spread in the epithelial level of the vocal folds. Henceforth it was controlled by the drug before it progressed to the other layers of vocal folds. The composition of cobadex czs consisted of multivitamins and chromimum zinc. According to Simpson, Robin, Ballantyne, & Groves (1967), the rate of growth of epithelial structures are controlled by the vitamin A. It can be interpreted in this case that the disease might be caused due to dietary deficiencies with low vitamin A and B intake.

possibility of becoming malignant change. It is believed that laryngeal keratosis progressed to carcinoma in situ or invasive carcinoma through the stage of epithelial proliferation and epithelial dysplasia. The acoustic aspects of voice in keratosis larynx are abnormal as like other organic dyaphonias. The quantitative and qualitative measures of voice improve with medication unless the disease has progressed towards the other layers of vocal folds. Generalization of the results are uncertain because of single case study and the lesion size, shape, extent may vary.

References
Garca, R. L., Aranzbal, L. M. A., Salas, R. A., Olano, A. M. & Guimera, A. J. (1996). The study of the evolution of laryngeal premalignant lesions. Acta Otorrinolaringol Esp., 47(2), 129-33. Grossman, A. A. & Mathews, W. H. (1952). Keratosis of the larynx with progression to malignancy. Canad. M. A. J. Vol. 66. 39-41. Simpson, J. F., Robin, I. G., Ballantyne, J. C. & Groves, J. (1967). A synopsis of otolaryngology. John Wright & Sons Ltd, Bristol. 366-367.

Acknowledgements
We thank the Director, All India Institute of Speech & Hearing, Mysore, for allowing us to conduct the study. We extend our thanks to Savithri, S. R., Professor of Speech Sciences, AIISH, Mysore, for her guidance and discussion during the preparation of this paper.

Conclusions
Early diagnosis of laryngeal keratosis, a precancerous lesion, is important since it has the

22

JAIISH, Vol. 27, 2008

Nasalence Value for Rainbow Passage

Nasalence Value for Rainbow Passage: Normative Data for NonNative Speakers
1

Sangeetha Mahesh & 2Pushpavathi M.

Abstract
The development of normal speech is the most important goals of a clinician. The speech disorders associated with cleft lip and palate include abnormal consonant production, abnormal nasality, nasal air emission, nasal turbulence, and unintelligible speech. Nasality is measured subjectively and objectively. The ratio of acoustic energy output from the oral and nasal cavities of the speaker is called Nasalence. Most normative data available for assessing resonance through instrumentation have been on English speaking population. The nasalance is influenced by several parameters such as age, language, dialect, speech stimuli and gender. Hence the present study investigated nasalence values in Non- native English speakers using RAINBOW passage. Mean nasalence scores were obtained from 45 normal males and 70 normal females. The results indicated higher nasalence percent and variability in females compared to males. The mean nasalence value was 31.39 for females and 27.93 for males. A comparison of the nasalence values for Rainbow passage across various studies reveal significant differences except Hutchinson etal. (1978). this difference is due to the difference across subjects, age and the instrumentation.
Key words: Nasometer, Nasalence, Rainbow passage, Non-native English speaker.

Nasality is a common problem in subjects with repaired / unrepaired cleft palate, which affects the speech intelligibility. Nasal resonance is not only seen in disordered speech, it is also seen at certain extent in normal speech. Nasality can be assessed by subjective as well as objective methods. Judgment of nasality is done using various rating scales. Instrumentation provides explicit information with respect to certain ranges of nasal resonance that was particularly difficult for listeners to resolve. Earlier studies indicated that nasalance values vary across languages. (Anderson, 1996; Van Doorn and Purcell, 1998; Van Lierde, 2001; Whitehill, 2001; Van Lierde, Wuyts, De Bodt, and Van Cauwenberge, 2001: Van Lierde, Wuyts, Bodt, and Cauwenberge, 2003; Sweeney and ORegan, 2004: Mahesh and Pushpavathi, 2008) An initial step towards refining the use of nasometry as an objective measure of perceived nasal acoustic energy involves manipulating the speech sample used. Several speech samples and reading materials are included in the nasometry

package for use in assessment of resonance disorders. Three standard stimuli for data collection were recommended by Fletcher (1978) Rainbow passage (a passage in which the occurrence of phonemes is similar to the occurrence in English conversational speech), Zoo passage (which has only oral sounds) and a set of nasal sentences. Most of the studies used speech stimuli developed in their own languages and are comparable to Standard English passages. Nasalence data has been published for normal speakers (Hutchinson, Robinson and Nerbonne, 1978: Seaver, Dalston and Leeper, 1991: Leeper, Rochet and MacKay, 1992) as well as in clinical groups (Fletcher, 1978). Nasalence value also varies with reference to the gender. Gender related differences in nasalance value can possibly be related to basic structural and functional differences across gender. The resonance of voice is influenced by the size, shape and surface of infraglottal and supraglottal resonating structures and cavities. Previous studies found that female speakers have

1 Clinical Lecturer, Dept. of Clinical Services, All India Institute of Speech and Hearing, Manasagangothri, Mysore- 570 006, email: sangmahesh9@yahoo.co.in, 2Reader, Speech-Language Pathology, All India Institute of Speech and Hearing, Manasagangothri, Mysore- 570 006, email: pushpa19@yahoo.co.in

23

JAIISH, Vol. 27, 2008

Nasalence Value for Rainbow Passage

significantly higher nasalance values compared to male speakers on passage containing nasal consonants (Seaver, Dalston, and Leeper, 1991; Van Lierde, Wuyts, De Bodt and Van Cauwenberge, 2001; Fletcher, 1978; Hutchinson, Robinson, and Nerbonne, 1978). Seaver, Dalston and Leeper (1991) compared the nasalence values of 148 normal adult subjects speaking four dialects of American English using Rainbow passage, Zoo passage and a set of nasal sentences. The mid atlantic speakers were found to have significantly higher nasalence value in all three stimuli. The female subjects had significantly higher nasalence value on the nasal sentences. They also found significant difference across dialects. Post hoc comparison revealed significant difference among the values of Mid Atlantic and Mid Western speakers, Mid Atlantic and Ontario speakers, Mid Atlantic and southern speakers. Correlations of moderately high strength were found between nasalence values of Rainbow passage readings with Zoo passage readings probably due to predominance of oral consonants and vowels, even though it does contain some nasal consonants. Correlations of moderately high strength were also found between nasalence values of Rainbow passage readings with nasal passage readings probably due to both passages contain nasal consonants hence the instrument would respond accordingly. There is very limited data on nasometric values in non-native English speakers using rainbow passage. Normative data are available for English speakers, as most of the studies have been conducted in native English speakers. These data in turn indicate that not all native English speakers obtain the same nasalence values. Factors such as English dialect spoken and gender of the subject appear to affect nasalence value which suggest cross dialect differences. These results stress the importance of developing normative data for various subgroups in the general population using the standard rainbow passage. Speech pathology clinics in India are using the Nasometer to confirm the perceptual judgment of abnormal levels of speech nasality. In particular, it is being used to assess the velopharyngeal dysfunction and to evaluate its treatment in clients with cleft palate. Normative nasalence measures will provide the database for future investigation on clinical population in India. In this context, the present study developed normative data on nasalence for non-native English speakers.
24

Method
Subjects: Forty five males and seventy females in the age range of 18 to 30 years served as subjects in the present study. All the subjects had normal structure and function of the oral mechanism. The subjects considered were from different parts of India. Subject had learnt English as a second language. Table 1 shows the subject details considered in the present study.
Subjects Males (N=45) Age range 12 30 Language (Mother tongue) Kannada Malayalam Tamil Telugu Hindi 8 15 7 6 3 2 15 22

18-30 yrs Females 18-30 (N=70) yrs

Table 1: Details of the subjects.

Instrumentation and Material: The Nasometer Model 6400 (Kay Elemetrics, New Jersy) was used in the present study. The oral and nasal components of the subjects speech are sensed by microphones on either side of a sound separator that rests on the patients upper lip. Nasometer computes a ratio of the nasal to nasal plus- oral acoustic energy from the digitized signals. Nasalence is expressed as a percentage value computed from that ratio (nasalence= nasal/{oral + nasal}X100. Prior to data collection, the nasometer was calibrated as prescribed by the manufacturer. One of standard passage, Rainbow passage provided in the manufacturers manual was used as stimuli in this study. The Rainbow passage contains a mixture of oral and nasal consonants in the approximate proportion found in everyday speech (Fairbank, 1960). It contains about 11% of nasal phonemes and the nasal sentences are 35%. Procedure and analysis: Subjects were seated in a quiet setting with the Nasometer headgear adjusted so the separation plate rested comfortably but firmly on the subjects upper lip and perpendicular to the plane of the face. Each subject read the Rainbow passage displayed on the monitor. Once the subject completed the task, the mean nasalence value was computed using the software package. For each subjects production, data on mean nasalence value, standard deviation, maximum and minimum nasalence value were collected using nasometer software. Independent t tests was computed to determine significant differences in nasalence values across gender and to compare with other studies.

JAIISH, Vol. 27, 2008

Nasalence Value for Rainbow Passage

Results and Discussion


Mean Nasalence value across gender The nasalence value ranged from 16% to 35% and the mean was 27.93 (SD 4.17) in males. In females the range was 20 % to 56% and the mean was 31.39 (SD-7.31). Table 2 shows the mean nasalence value and standard deviation of nasalence.
Rainbow passage N Mean SD Male 45 27.93 4.17 Female 70 31.39 7.31 Total 115 30.04 6.48 Table 2: Mean nasalence value for Rainbow passage. Gender

produced more coarticulatory anticipation of nasal consonants and had greater degree of nasal air flow during production of nasal consonants than males. The nature of measurement procedure dictates that degree of nasalence in speech will be proportional to the acoustic energy of the signal as it exits from nasal and oral chambers. This proportion is controlled by the physical characteristics of the oral and nasal chambers, integrity of velopharygeal valve, postures of lips and tongue and by the phonetic demands of the sounds spoken. The present study does not support the findings of Fletcher (1978) who found that males had higher mean nasalence value than females for nasal sentences. The present study also does not support the notion that there is no significant difference in nasalence value across gender (Litzaw and Dalston, 1992; Kavanagh, Fee and Kalinowski, 1994). Comparison of nasalence value for Rainbow passage across studies The normative data for sentences using Rainbow Passage across various studies are presented in Table 2. Single sample t test was used to compare the present study with earlier studies using Rainbow passage and is presented in table 3. Table 4 shows the results of single sample t test.
Author Language (Year) Fletcher et al. American (1989) Seaver etal(1991) American North west American North west American Mid Atlantic Mid Atlantic N 117 140 Subjects Children Mean S.D 35.6 5.20 6.0 6.0 5.1 10.7 4.0 4.0 4.1 7.3

Results of Independent t test showed significant difference across gender (t= 2.85, p<0.01) with females having higher nasalence percent compared to males. Also, the standard deviation was higher in females compared to males indicating higher variability among females.This result supports the findings of Seaver et al., 1991; Van Lierde et al., 2001; Fletcher, 1978; Hutchinson et al, 1978, who reported that female speakers have significantly higher nasalance values than male speaker on passage containing nasal consonants. Gender related differences in nasalance value can possibly be related to basic structural and functional differences. The resonance of voice is influenced by the size, shape and surface of infraglottal and supraglottal resonating structures and cavities. Two subject variables could be associated with increased nasal flow rate in female speakers, increased respiratory effort and increased nasal cross sectional area. Since females have longer nasal cross sectional area than males such a difference could be seen (Liu,1990). Mc Kearns and Bzoch (1970) discovered different patterns of velopharyngeal closure for females as determined by cineradiographic analysis. They suggested that different velopharygeal muscle insertions occur across gender, which may arise from differences in the relationship of the skull and cervical column or differences in vocal tract dimensions. Thompson and Hixon (1979) studied 112 normal children and adults and found that females

Adult male 35.0 Adult female Geriatric male Geriatric female Adult male Adult female Adult male Adult female 36.0 23.5 32.0 36.0 37.0 27.9 31.3

Hutchinson etal (1978) Litzaw & Dalston (1992)

30 30 15 15 45 70

Present study Indian (2008)

Table 3: The normative data for Rainbow passage across various studies.

25

JAIISH, Vol. 27, 2008

Nasalence Value for Rainbow Passage

Sl Author No. (Year) Fletcher et 1 al. (1989) 2 Seaver etal(1991)

Subjects

t p value value <0.001

Interpretation Significant difference Significant difference Significant difference Significant difference No significant difference Significant difference Significant difference Significant difference

has indicated that nasal cross sectional areas is not affected by age after the age of 18. On theoretical grounds, one might imagine that Rainbow passage would be particularly useful in sampling the acoustic consequences of velopharyngeal behaviour since the frequency of occurrence of phonemes in this passage roughly mirrors that found in conversational speech. Eleven percent of the phonetic elements are nasal consonants. The effect of these nasal consonants is not limited to the moment of their utterance, however rather a coarticulatory spread of nasalization is found in which the nasal consonants are anticipated by opening of the velopharyngeal valve prior to the onset of the nasal element in the speech output (Fletcher 1989). On other hand , this passage is much longer than Zoo passage. Moreover, it is syntactically more complex and contains a number of words that are difficult for very young children (Dalston and Seaver, 1992). Fletcher, (1978) determined that Nasalence values for the rainbow passage were consistently higher than the Zoo passage. Litzaw and Dalston, (1992) measured nasalence on adults in the age range of above 18 years with mid atlantic dialect using Nasometer 6200 instrument. A nasalence mean value of 36.0 and 37.0 in adult males and adult females, respectively, was found in the study. Though the subjects taken up in both the studies were adults there was significant difference across the values. Seaver etal.(1991) also measured nasalence in the age range of 16 years to 63 years belonging to four geographic regions using Nasometer 6200 instrument. A nasalence mean value of 35.0 and 36.0 in males and females respectively was found in the study. There was significant difference across mean nasalence value in males and females (p<0.01) when compared to the present study. These differences could be attributed to significant cross dialectal (English) and cross linguistic differences in nasometric values (Seaver et al., 1991; Leeper, Rochet, and MacKay, 1992). Fletcher et al. (1989) measured nasalence in Children in the age range of 5 years to 12 years using Nasometer 6200 instrument. A nasalence mean value of 35.69 was found in the study. There was significant difference across mean nasalence value in children (p<0.01) when compared to the present study. These differences could be attributed to subjects taken up in both the studies, as the present study included only adults.
26

Children 9.25 Males

11.23 <0.001 <0.001 <0.001 >0.05 <0.001 <0.001 <0.001

Females 5.23 Males 7.04

Hutchinson etal (1978)

Females 0.69 Males 12.82

Litzaw & Dalston (1992)

Females 6.36 Total 9.76

Table 4: Results of single sample t test.

The above table depicts the mean nasalence value across studies and across age. The mean value ranges from 23 % to 37%. This difference is due to the difference across subjects, age and the instrumentation. A comparison of the nasalence values for Rainbow passage across various studies reveal significant differences except Hutchinson etal. (1978). They measured nasalence on elderly subjects in the age range of 50-80 years using TONAR II instrument. A nasalence mean value of 23.5 and 32.0 in geriatric males and geriatric females, respectively was found in the study. There was significant difference across mean nasalence value in males (p<0.01) when compared to the present study. The probable difference could be attributed to the age and the instrument used. Aging is accompanied by degeneration of receptor cells, decline in number of nerve fibres in associated neural tracts, loss of brain cells in corresponding projection areas, decrease in muscular strength, slowness, lack of fine coordination of movement, cognitive slowing and deterioration in neural density and general delay in synaptic transmission (Corso, 1975; Botwinick, 1973; Crossman and Szafran 1956; Griew, 1963). The present results permit the general conclusion that where relatively continuous demands for velopharyngeal closure are required, older subjects exhibit notably less competence than normal young adults. There was no significant difference across mean nasalence value in females when compared to the present study. Seaver etal. (1991) reported that nasometer performance was not significantly influenced by age. Warren and collegues, (1990)

JAIISH, Vol. 27, 2008

Nasalence Value for Rainbow Passage

Differences in these values are difficult to interpret because there were methodological differences in the studies. However, the extent of the differences may indicate that there is need to control for dialect, age, and gender before meaningful across study comparisons can be made. Research has reported significant cross dialectal and cross linguistic differences in nasometric values (Seaver et al., 1991; Leeper et al., 1992). Normative data for nasalence values and clinically determined cutoff values have both been found to be sensitive to dialectal differences in different regions of North America. Differences in mean Nasalence values across languages may be explained by different use of vowels, oral and nasal consonants across languages (Leeper et al., 1992; Anderson, 1996). Furthermore, even in bilingual speakers, differences in nasometric values across languages are significant (Leeper et al., 1992). Thus it is essential that for establishing normative data for nasometer, issues pertaining to dialect and language need to be considered. Mean nasometric values obtained for a specific linguistic group may not be valid for use with other groups, even though they may speak the same language. Hence the normative data for different dialectal and linguistic populations are necessary.

opportunity to probe the modifiability of nasalence systematically. However, study by Dalston and Seaver, 1992 reported, that Rainbow passage contains a number of words that are difficult to produce for very young children to pronounce. Establishing the cut off values for clinically significant abnormalities is important in many areas of medical epidemiology. It can be approached from a clinical or statistical perspective (Barker and Rose, 1984). For the Nasometer, the issue of determining cut off nasalance value for clinical populations has been approached from both statistical and clinical perspective. Clinically the normative data reported in the present study may help identifying adults with resonance disorders. Nasalence may also be a sensitive indicator of the presence and progress of neuromuscular disease (Fletcher and Bishop, 1970). Speech pathologists, Otolaryngologists, and Plastic surgeons can use the data to help objectify and supplement their diagnostic, follow-up testing, and treatment protocols.

References
Anderson, R. T. (1996). Nasometric values for normal Spanish-Speaking females: A preliminary report. Cleft Palate Craniofacial Journal, 33, 333-336. Barker, D. J. P., & Rose, G. (1984). Epidemiology in medical practice. 3rd ed. Edinburgh: Churchill Livingstone. Botwinick, J. (1973). Aging and behavior. New York : Springer Publishing Co. Corso. J. M. (1975). Sensory processes in man during maturity and senescence. In J. M. Ordy & K.R. Brizzee (eds.), Neurobiology of aging. New York: Plenum press. Crossman, E. R. F. W., & Szafran, J.(1956) Changes with age in the speed of information intake and discrimination . Experiantia ( Suppl.). Dalston, R. M., Warren, D., & Dalston, E. (1991a). A primary investigation concerning use of Nasometery in identifying patients with hyponasality and/or nasal airway impairment. Journal of Speech and Hearing Research, 34, 11-18. Dalston, R.M., & Seaver, E.J. (1992). Relative value of various standardized passage in the
27

Conclusions
Very few Indian studies have been done on developing a normative data in Indian context using Rainbow passage as the standard stimuli using Nasometer 6400. The primary purpose of this present study has been to provide speech and language pathologists with instrumental verification of their perceptual judgements concerning the diagnostic evaluations. There are no Indian studies reported on nasalence value using the Nasometer II 6400. Most of the earlier studies used TONAR system and Nasometer 6200 to measure nasalence. The nasometer instrumentation differs substantially from the earlier Tonar system. The reported normative nasalance data provide important reference information for the assessment of nasality disorders in adults on using Nasometer II 6400 with Rainbow passage as the speech stimuli. The performance on Rainbow passage would provide information over and above that available from the values of Nasal and Zoo passage. The provision for rapid, accurate, biometric feedback provides

JAIISH, Vol. 27, 2008

Nasalence Value for Rainbow Passage

nasometric assessment of patients with velopharyngeal impairment. Cleft Palate Craniofacial Journal, 29, 17-21. Fairbank, G (1960). Cited in Hutchinson, J. M., Robinson, K. L., & Nerbonne, M. A. (1978). Pattern of nasalance in a sample of normal gerontologic subjects. Journal of Communication Disorders, 11, 469 -481. Fletcher, S.G. (1978). Diagnosing speech disorder from the cleft palate. New York: Grune & Stratton. Fletcher, S.G., & Bishop M.E (1970). Measurement of nasality in Tonar. Cleft palate journal. 7;610-621. Griew, S . (1963). Information transmission and age. In R. H. Williams, C. Tibbitts, and W. Donahue ( eds.),Processes of Aging. New York: Atherton Press. Hardin, M. A., Van Demark, D. R., Morris, H. L., & Payne, M. M. (1992). Correspondence between nasalance value and listener judgments of hypernasality. Cleft Palate Craniofacial Journal, 29, 349-351. Hutchinson, J. M., Robinson, K. L., & Nerbonne, M. A. (1978). Pattern of nasalance in a sample of normal gerontologic subjects. Journal of Communication Disorders, 11,469 -481. Kavanagh, M.L., Fee, E.J., Kalinowski, J. (1994). Nasometric values for three dialectal groups within the Atlantic provinces of Canada. Journal of Speech Language Pathology Audiology, 18, 7-13. Leeper, H. A., Rochet, A. P., & MacKay, I. R. A. (1992). Characterisitics of nasalance in Canadian speaker of English and French. International conferences on spoken language processing (Abstract), Banff, Alberta, October, 49-52. Litzaw, L. L & Dalston, R. M. (1992). The effect of gender upon nasalence values among normal adult speakers. Journal of Communication Disorders, 25,55 -64. Liu, H. (1990). Cited In Litzaw, L. L & Dalston, R. M. (1992). The effect of gender upon nasalence values among normal adult speakers. Journal of Communication Disorders, 25,55 -64.
28

Mahesh, S & Pushpavathi, M ( 2008 ) . Nasometer values for zoo paasage in non native speakers: A preliminary study. Proceedings of International Symposium on Frontiers of Research on Speech and Music (FRSM)2008), 115-119. McKearns, D., Bzoch, K. R. (1970). Variations in velopharyngeal valving: the factor of sex. Cleft Palate Craniofacial Journal. 652-662. Nichols, A.C. (1999). Nasalance statistics for two Mexican population. Cleft Palate Craniofacial Journal, 36, 57-63. Seaver, E.J., Dalston, R.M., & Leeper, H. A. (1991). A study of nasometric value for normal nasal resonance. Journal of Speech and Hearing Research, 34, 715-721. Sweeney, T., Sell, D., & ORegan, M. (2004). Nasalance values for normal speaking Irish children. Cleft Palate Craniofacial Journal, 41(2), 168-174. Thompson, A. E., & Hixon, T. J (1979) Nasal air flow during normal speech production. Cleft Palate Craniofacial Journal 16: 412-420. Van Doorn, J., & Purcell, A. (1998). Nasalance level in speech of normal Australian children. Cleft Palate Craniofacial Journal, 35, 287-292. Van Lierde, K. M., Wuyts, F. L., Bodt, M. D., & Cauwenberge, P (2003). Age related pattern of nasal resonance in normal Flemish children and Young adults. Scandinavian Journal of Plastic Reconstructive Surgery, 37,344-350. Van Lierde, K. M., Wuyts, F. L., De Bodt, M., & Van Cauwenberge, P. (2001). Normative value for normal nasal resonance in the speech of young Flemish adults. Cleft Palate Craniofacial Journal, 38, 112-118. Warren, D.W., Hairfield, W. M., & Dalston, E. T. (1990). Effect of age on nasal crosssectional area and respiratory mode in children. Laryngoscope. 100:89-93. Watterson, T., McFarlane, S., & Wright, D.S. (1993). The relationship between nasalance and nasality in children with cleft palate. Journal of Communication Disorders, 26, 13-28.

JAIISH, Vol. 27, 2008

Nasalence Value for Rainbow Passage

Whitehill, T.L. (2001). Nasalance value in Cantonese speaking women. Cleft Palate Craniofacial Journal, 38, 119-125.

Acknowledgements
The authors wish to thank Dr. Vijayalakshmi Basavaraj, Director, AIISH for her encouragement and all the support provided to conduct this study. The authors also wish to thank all the subjects who volunteered to serve as subjects in this study.

29

JAIISH, Vol. 27, 2008

Rate of Speech

Rate of Speech/Reading in Dravidian Languages


1

Savithri S.R. & 2Jayaram M.

Abstract
Rate of speech is an important variable in the evaluation and treatment of fluency disorders. The present study investigated rate of speech/ reading in 4 Dravidian languages and established normative data on rate of speech in Kannada, Tamil, Telugu, and Malayalam in subjects ranging in 10 decades of age. Four hundred and one (Kannada = 136, Telugu = 69, Tamil = 103, Malayalam = 93) normal subjects participated in the study. All subjects were literates and were from urban population. Informed consent was obtained from subjects prior to collection of speech samples. Subjects in the age range of 4-10 years described cartoons and narrated stories and subjects in the age range of 11-100 years read passages which were developed by the investigators. All these were audio-recorded and analyzed to obtain syllables per second, syllables per minute and words per minute. Cool Edit software was used to eliminate pause and calculate duration. Results indicated significant difference between age groups, and languages. Rate of speech increased with increase in age till about 40 years and declined after 40 years (except Tamil). Malayalam had the highest SS and SPM compared to other languages. No significant difference between genders was observed. The results provide normative data for clinical purposes and to set rate of speech in persons with stuttering and cluttering.
Key words: Rate of speech, Words per minute, Syllables per minute, Dravidian languages.

Fluency is the effortless production of long, continuous utterances at a rapid rate (Starkweather, 1981). The rate at which speech is produced is an important aspect of fluency. Writers attempting a description of stuttering have most often not included material on normal fluency, even though a substantial scientific literature on the rate, rhythm and timing of normal adult speech has long existed. Descriptions of children's fluency development have referred only to the frequency and type of discontinuities. But it is not just the continuity of speech that signals fluency. The rate of speech, the length of utterances, consistency in the duration of elements, and the overall quantity of speech are also signs of the facility with which speech is produced. With increased age, these signs show developmental change, signaling development of fluency. Rate of speech is an important variable in the evaluation and treatment of fluency disorders. It is well known that the rate of speech correlates negatively with the severity of stuttering (Bloodstein, 1944; Sander, 1961) because frequent
1

and/or long duration stutters result in reduced speech output. However, a deliberate reduction in the rate of speech has a beneficial effect on the frequency of stutters (Adams, Lewis & Besozzi, 1973; Johnson & Rosen, 1937) and, therefore, nearly all stuttering treatment approaches include rate reduction as one of their goals. Rate control is also a treatment target in cluttering (Daly, 1986). In addition, rate of speech is an important factor in the perceptual evaluation of normalcy of speech following treatment of stuttering (Ingham & Packman, 1978). Extremely slow rate of speech even if completely devoid of dysfluencies, is perceived as unnatural by listeners. There is some preliminary evidence that the rate of speech of parents may be an important factor in incipient stuttering (Meyers & Freeman, 1985; StephensonOpsai & Bernstein Ratner, 1988). The rate of speech primarily depends on the speed of articulatory movement and the degree of coarticulatory overlap. It also depends on the linguistic structure and culture. Rate of speech can be measured in two ways. One is a measure of the

Professor, Dept. of Speech-Language Sciences, All India Institute of Speech & Hearing, Manasagangothri, Mysore 570 006, email:Savithri_2k@yahoo.com, 2Professor, Dept. of Speech Pathology and Audiology, NIMHANS, Bangalore 560 029,

30

JAIISH, Vol. 27, 2008

Rate of Speech

number of syllables per unit time. The second is words per minute. The word is more of an information unit than a speech production unit. Consequently, words per minute are a measure of the amount of information a speaker is producing. It is related to but not the same as the rate at which syllables are produced. The more syllables a word contains, the more rapidly each syllable in the word is produced (Klatt, 1973). Word duration and utterance duration seem to depend on the amount of information contained in the utterance, but syllables per second seems independent of content, as long as the sample is to contain a large variety of syllables. It is important for the practicing clinician to know about these changes, for the assessment of stuttering and cluttering should be made by comparing the client's fluency with the level of fluency that would be expected for a normal person of the same age. Surprisingly, however, there are few empirically derived guidelines available for clinical measurement of rate and for setting goals for rate (Ingham & Cordes, 1997). Adult speakers of English speak at an average rate of 5 to 6 syllables per second (Walker & Black, 1950). When rate is measured in words per minute, most of the variation is attributable to the duration and frequency of pauses. When these pauses are excluded, the variability of speech rate is much reduced (Goldman-Eisler, 1968). The normal rate of speech is 80-180 words per minute. However, one can consider a rate of up to 280 words per minute as normal provided the intelligibility is not affected. This is supported by reports that speech could be compressed up to 275 wpm mechanically with little loss of comprehension, but comprehension declines rapidly at higher speeds. Rathna, Subba Rao & Bharadwaj (1979) reported 361 syllables per minute and 104 words per minute in spontaneous speech and 427 syllables per minute and 94 words per minute in reading in Kannada. Venkatesh, Purushothama & Poornima (1983) investigated rate of speech in 64 Kannada speakers in the age range of 17-66 years. They reported 282 syllables per minute in adult Kannada speakers. Rathna et. al. took 1-minute sample and Venkatesh et. al. took males, females, urban, rural, literate and illiterate population and recorded conversation. Samples in both these studies are not adequate and do not represent any specific population. Also, pauses were not eliminated in these studies.
31

Some studies have investigated differences between genders. Johnson (1961) reported higher range and docile values for adult females than for adult males in two spontaneous speech tasks and one reading task. In contrast, Lutz & Mallard (1986) found that adult male subjects talked and read at a faster rate than adult female subjects in their study. However, statistical tests were not performed in either study to determine whether the differences in rate between males and females were significant. In a study of rate of speech in cildren, Haselager, Slis & Rietveld (1991) reported that the boys did not significantly differ from girls in the rate of production of spontaneous speech. It is not known whether the findings of Haselager et. al. would also apply to young adults. It is important, for clinical purposes, to determine whether statistically significant differences exist between adult males and females in the rate of reading and discourse. Values of approximately 200 SPM or 150 SPM are frequently used in setting goals for rate of speech (Perkins, 1973; Boberg & Kully, 1985) because mean rates of adult discourse tend to converge around these values (Luchsinger, 1965). Answers to basic questions such as the size of the sample necessary to obtain a reliable measure of rate and the variability or stability of rate in different tasks typically used to measure rate in clinical settings are not known. Johnson, Darley & Spriesterbach (1963) recommend that a 3-minute sample of spontaneous speech and a 300-word reading passage be used to determine rate. Castello & Ingham (1984) suggest that a 2-minute sample of uninterrupted speech be used to determine the rate during the evaluation of stuttering. However, there is little objective data in support of these recommendations. There is also an uncertainty concerning the unit of speech appropriate for the computation of rate. Expression of rate in syllables per minute (SPM) appears to be generally favored over the computation involving words per minute (WPM) because the length of syllables, whether measured in phonemes or in units of time tends to be less variable than length of words (Umeda & Quinn, 1980; Ingham, 1984; Costello & Ingham, 1984). However, there is little empirical evidence to support the view that SPM is a more valid measure of speech rate than WPM, particularly in clinical situations where a certain amount of variability in speech rate is expected and accepted.

JAIISH, Vol. 27, 2008

Rate of Speech

Though the rate of speech is an important parameter it has not been studied extensively in the Indian context. As the linguistic structure of Dravidian and Indo-European languages differ, it is probable that the rate of speech also differs. Also, most rate control therapies donot consider age as criteria to set goals for rate of speech. It is hypothesized that age and language have significant effect on the rate of speech. In this context, the present study established normative data on rate of speech in Kannada, Tamil, Telugu, and Malayalam in subjects ranging in 10 decades of age. The end results brought out from this project can be utilized as normative data.
Age range/ Language Age 3-3.11 4-4.11 5-5.11 6-6.11 7-10 11-15 16-20 21-30 31-40 41-50 51-60 61-70 71-80 81-90 91-100 Sub-total Total Kannada M 5 5 5 5 5 5 5 5 5 5 5 5 5 5 0 70 F 5 5 5 5 5 5 5 5 5 5 5 5 5 1 0 66

This can be used as a reference or standard to measure rate of speech in clients with fluency disorders.

Method
Subjects: Four hundred and one normal subjects in four languages - Kannada, Tamil, Telugu, and Malayalam - participated in the study. All subjects were literates and were from urban population. Informed consent was obtained from the subjects prior to collection of speech samples. Table 1 shows subject details.

Telugu Tamil Malayalam M F M F M F

2 2 3 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 3 3 5 5 1 0 5 5 2 0 0 1 0 0 0 0 36 33 50 53 401 Table 1: Details of subjects.

5 5 5 5 5 5 5 5 5 1 46

5 5 5 5 5 5 5 5 5 2 47

Material: Cartoons (4-6 years), pictures depicting Panchatantra stories (7-10 years), and standardized reading passages were used to elicit spontaneous speech/reading samples. Pictures of cartoons and Panchatantra stories were taken from Indu (1992) Yamini (1992) and Rajendra Swamy (1995). Passages in four languages were developed by the experimenters. There were 304, 306, 414, and 307 words in Kannada, Telugu, Tamil, and Malayalam passages, respectively. Procedure: Children in the age range of 4-10 years were instructed to describe the cartoons and story, and adults read the passage at a comfortable pitch and loudness. All samples were audiorecorded and were digitized at 16,000 Hz sampling frequency. Pauses, if any, were removed from the waveform using Cool Edit software. Each syllable
32

and word was highlighted using the waveform and the duration was measured using the software. The number of syllables per second (SS), syllables per minute (SPM) and words per minute (WPM) were calculated by using the following formule:
SS = Total number of syllables / Total time taken (seconds). SPM = Total number of syllables /Total time taken (minute). WPM = Total number of words / Total time taken (minute).

Statistical analysis: ONE-WAY ANOVA was done to compare the rate of speech across age, gender, and language, and MANOVA was used to examine the interaction effects.

JAIISH, Vol. 27, 2008

Rate of Speech

Results
Kannada The results indicated an increase in SS, SPM and WPM from 3 years to 40 years and decrease in SS, SPM and WPM in the age group of 41-90 years. Table 2 shows the mean and range of syllables per second (SS), syllables per minute (SPM) and words per minute (WPM) from 3 to 90 years in Kannada speaking subjects.
M/SD SS SPM WPM Mean 4.9 291 129 SD 1.3 78 34 4-4.11 Mean 4.2 252 91 SD .79 48 17 5-5.11 Mean 4.2 254 120 SD 1.3 79 37 6-6.11 Mean 4.3 261 92 SD 0.84 50 18 7-10 Mean 4.1 250 85 SD 0.65 41 14 11-15 Mean 5.7 343 104 SD .83 48 14 16-20 Mean 7.0 425 127 SD .96 58 17 21-30 Mean 6.5 386 116 SD .96 70 21 31-40 Mean 7.2 434 131 SD .31 19 6 41-50 Mean 6.8 410 124 SD .87 53 16 51-60 Mean 6.9 415 124 SD .62 37 11 61-70 Mean 6.9 404 125 SD .73 52 12 71-80 Mean 6.4 390 118 SD .85 51 15 81-90 Mean 5.6 337 102 SD .97 58 17 Mean 5.3 318 104 SD 2.20 133 40 Table 2: Mean and SD of rate of speech in Kannada speakers. Age 3-3.11

Age Group 3-3.11 4-4.11 5-5.11 6-6.11 7-10 11-15 16-20 21-30 31-40 41-50 51-60 61-70 71-80 81-90

1 4.2358 4.2031 4.2358

5.7000

5.7000 7.0300 6.4900 7.1800 6.7900 6.8640 6.9540 6.4400

6.4400 5.6200

Table 3: Results of Duncans post-hoc test for SS (Kannada).


Age Group 3-3.11 4-4.11 5-5.11 6-6.11 7-10 11-15 16-20 21-30 31-40 41-50 51-60 61-70 71-80 81-90 1 290.9821 252.1879 254.1461 2 3 4 5

343.10 343.10 425.10 385.50 385.50 385.50 434.20 410.41 415.10 403.60 389.80 389.80 337.3400 337.34

Table 4: Results of Duncans post-hoc test for SPM (Kannada).


Age Group 3-3.11 4-4.11 5-5.11 6-6.11 7-10 11-15 16-20 21-30 31-40 41-50 51-60 61-70 71-80 81-90 1 91.3535 119.7360 92.3000 84.5000 103.6844 119.7360 2 3 128.6358

103.6844 116.2539 127.3501 116.2539 131.0137 123.6824 124.1360 124.8651 117.7516

ONE-WAY ANOVA indicated significant difference between age groups {SS = [F (14, 121) = 16.70, p < 0.01], SPM = [F (14, 121) = 15.665, p < 0.01], WPM = [F (14, 121) = 5.718, p < 0.01]}. Tables 3 to 5 show results of the Duncans posthoc test. Values in the same column are not significantly different.

101.7867

117.7516 101.7867

Table 5: Results of Duncans post-hoc test for WPM (Kannada).

Telugu SS, SPM, and WPM increased till 30 years and declined thereafter. Table 6 shows mean and SD of SS, SPM, and WPM.

33

JAIISH, Vol. 27, 2008

Rate of Speech

Age 11-15 16-20 21-30 31-40 41-50 51-60 61-70 71-80 81-90

M/SD Mean SD Mean SD Mean SD Mean SD Mean SD Mean SD Mean SD Mean SD Mean SD Mean SD

SS 7 1.25 7 0.61 8 0.58 6 0.95 7 1.04 6 1.04 6 0.99 5 0.19 5.14 0.19 6.4 0.76

SPM 431 132 439 36 466 36 384 54 389 75 392 69 336 60 309 11 309 11.31 384 53.81

WPM 123 37 125 10 133 10 116 23 117 20 110 18 96 17 89 4 88.82 3.85 111 15.87

Age 1 2 3 Group 11-15 126.0180 16-20 120.3800 120.3800 21-30 125.2300 125.2300 31-40 133.1500 41-50 116.0610 116.0610 116.0610 51-60 116.6800 116.6800 116.6800 61-70 110.1630 110.1630 110.1630 71-80 96.4930 96.4930 81-90 88.8200 Table 9: Results of Duncans post-hoc test for WPM (Telugu).

Table 6: Mean and SD of rate of speech in Telugu speakers.

Tamil SS, SPM, and WPM increased from 5 years to 80 years. However, there was no consistent linear increase in the rate of speech. Table 10 shows mean and SD of SS, SPM, and WPM.
Age 5-5.11 M/SD SS SPM WPM

ONE-WAY ANOVA indicated significant difference between age groups {SS = [F (8, 73) = 4.832, p < 0.01], SPM = [F (8, 731) = 2.495, p < 0.05], WPM = [F (10, 91) = 2.285, p < 0.05]}. Tables 7 to 9 show results of the Duncans posthoc test. Values in the same column are not significantly different.
Age 1 Group 11-15 16-20 21-30 31-40 41-50 51-60 61-70 71-80 5.5990 81-90 5.1400 Table 7: Results of (Telugu). 2 6.2460 3 6.2460 6.9700 7.4060 6.3750 6.8970 6.4230 4 6.9700 7.4060 7.7100 6.8970

6.3750 6.4230 5.5990

Duncans post-hoc test for SS

Age 1 2 3 Group 11-15 441.1000 16-20 421.0000 421.0000 21-30 438.7000 31-40 466.1000 41-50 384.3000 384.3000 384.3000 51-60 389.4000 389.4000 389.4000 61-70 391.9000 391.9000 391.9000 71-80 337.3000 337.3000 81-90 309.0000 Table 8: Results of Duncans post-hoc test for SPM (Telugu).

Mean 4.9950 SD 1.0360 6-6.11 Mean 4.6484 SD 1.0709 7-10 Mean 6.6920 SD 1.6674 11-15 Mean 4.5113 SD .3627 16-20 Mean 6.5238 SD .7114 21-30 Mean 6.5044 SD 1.2267 31-40 Mean 5.4654 SD 1.1116 41-50 Mean 5.9058 SD .8357 51-60 Mean 5.4330 SD .8551 61-70 Mean 5.6564 SD 1.0284 71-80 Mean 5.9033 SD .9218 Average Mean 5.7169 SD 1.2058 Table 10: Mean and SD of speakers.

299.4675 113.9975 62.3379 23.7109 278.9059 86.0060 64.2550 19.8293 401.5178 123.8661 100.0455 30.8635 271.7749 94.6891 21.8501 7.7201 383.5347 133.0958 53.0329 18.4800 390.8467 135.6623 73.5010 25.4697 326.3887 113.5195 67.7142 23.5099 352.6335 122.4497 49.1482 17.2613 326.3735 114.8765 51.5573 20.3268 339.9639 118.0024 61.8869 21.4296 353.8821 123.3533 54.2288 18.5347 342.1399 117.0508 72.3806 24.6119 rate of speech in Tamil

ONE-WAY ANOVA indicated significant difference between age groups {SS = [F (10, 91) = 4.719, p < 0.01], SPM = [F (10, 91) = 4.398, p < 0.01], WPM = [F (10, 91) = 4.459, p < 0.01]}. Tables 11 to 13 show results of the Duncans posthoc test. Values in the same column are not significantly different.
34

JAIISH, Vol. 27, 2008

Rate of Speech

Age 1 2 3 4 5 Group 5-5.11 4.9950 4.9950 4.9950 6-6.11 4.6484 4.6484 7-10 6.6920 11-15 4.5113 16-20 6.5238 6.5238 21-30 6.5044 6.5044 31-40 5.4654 5.4654 5.4654 5.4654 41-50 5.9058 5.9058 5.9058 51-60 5.4330 5.4330 5.4330 5.4330 61-70 5.6564 5.6564 5.6564 5.6564 71-80 5.9033 5.9033 5.9033 Table 11: Results of Duncans post-hoc test for SS (Tamil).
Age Group 5-5.11 6-6.11 7-10 11-15 16-20 21-30 31-40 41-50 51-60 61-70 71-80 1 299.4675 278.9059 271.7749 383.5347 390.8467 326.3887 352.6335 326.3735 339.9639 353.8821 383.5347 390.8467 352.6335 339.9639 353.8821 2 299.4675 278.9059 3 299.4675 401.5178 4 5

Age 11-15 16-20 21-30 31-40 41-50 51-60 61-70 71-80 81-90 91-100

M/SD Mean SD Mean SD Mean SD Mean SD Mean SD Mean SD Mean SD Mean SD Mean SD Mean SD Mean SD

SS 7.9082 .8486 8.8040 1.2238 9.2600 .9902 8.2000 .6055 7.9000 1.2275 7.9390 1.0498 7.8170 .5858 7.0350 1.6478 6.6160 .8643 4.2800 .7302 7.8240 1.4037

SPM 474.4926 50.9189 528.9257 73.3538 557.9370 61.7733 492.1712 35.5242 476.6454 80.9450 482.6940 66.8788 447.6767 66.2682 423.1902 99.4643 397.8133 51.8349 257.4613 43.9912 468.6865 87.6557

WPM 115.2427 12.3654 128.2894 17.7127 135.4231 14.9202 119.4511 8.5534 115.6346 19.7229 117.1112 16.1906 109.9117 12.4797 102.3253 23.9974 96.4343 12.7868 62.6108 10.7018 113.8287 21.0098

326.3887 326.3735

326.3887 326.3735 339.9639

326.3887 352.6335 326.3735 339.9639 353.8821

Table 14: Mean and SD of rate of speech in Malayalam speakers.

Table 12: Results of Duncans post-hoc test for SPM (Tamil).

Age Group 5-5.11 6-6.11 7-10 11-15 16-20 21-30 31-40 41-50 51-60 61-70 71-80

1 86.0060

2 113.9975

3 113.9975 123.8661

94.6891

94.6891 133.0958 135.6623 113.5195 122.4497 114.8765 118.0024 123.3533

ONE-WAY ANOVA indicated significant difference between age groups conditions {SS = [F (9, 83) = 9.173, p < 0.01], SPM = [F (10, 91) = 8.241, p < 0.01], WPM = [F (10, 91) = 8.606, p < 0.01]}. Tables 15 to 17 show results of Duncans post-hoc test. Values in the same column are not significantly different.
Age Group 11-15 16-20 21-30 31-40 41-50 51-60 61-70 71-80 81-90 91-100 1 2 3 7.9082 4 7.9082 8.8040 8.2000 7.9000 7.9390 7.8170 7.0350 5 8.8040 9.2600 8.2000 7.9000 7.9390 7.8170 7.0350

113.5195 114.8765

8.2000

Table 13: Results of Duncans post-hoc test for WPM (Tamil).

Malayalam SS, SPM, and WPM increased from 11 years to 30 years and decreased steadily from 31 years to 100 years. Table 14 shows mean and SD of SS, SPM, and WPM.

6.6160 4.2800

Table 15: Results of Duncans post-hoc test for SS (Malayalam).

35

JAIISH, Vol. 27, 2008

Rate of Speech

Age 1 2 3 Group 11-15 474.4926 16-20 21-30 31-40 492.1712 41-50 476.6454 51-60 482.6940 61-70 447.6767 447.6767 71-80 423.1902 423.1902 81-90 397.8133 91257.4613 100

Discussion
The results indicated significant differences in rate of speech/ reading across Dravidian languages. On an average the rate was 6, 6, 6, and 8 syllables per second in Kannada, Telugu, Tamil, and Malayalam, respectively; the syllables per minute was 346, 384, 342, and 469; words per minute was 113, 111, 117, and 114 in the four languages, respectively. The average rate of reading in Kannada, Telugu, and Tamil are in consonance with the earlier studies by Walker & Black (1950), Rathna et. al. (1979), and Venkatesh et. al (1983). However, Malayalam seems to be an exception with higher rates of SS and SPM. The differences between languages can be attributed to differences in syllable structure. Table 19 shows the syllable types and percent occurrence of such syllables in four Dravidian languages. The data is extracted from the reading passages used in this study. Unlike in Kannada, consonants occur in word-final position in Telugu, Tamil and Malayalam. Higher occurrence of V type of syllables and lower occurrence of CCV type of syllables in Malayalam compared to Kannada and Tamil seems to contribute to higher syllables per second in Malayalam.
Syllable type V CV CCV CVC CCVC CCCV VCC VC CVCC Total Kannada 6.3 73 21 Telugu 8.2 67 22 1.6 0.5 0.2 Tamil 4.8 54 0.4 38 Malayalam 7.3 74 3.6 13.4 0.2 0.2 1.1 0.2 100

474.4926 528.9257 528.9257 557.9370 492.1712 492.1712 476.6454 482.6940

Table 16: Results of Duncans post-hoc test for SPM (Malayalam).


Age Group 11-15 16-20 21-30 31-40 41-50 51-60 61-70 71-80 81-90 91100 1 2 3 115.2427 4 115.2427 128.2894 119.4511 115.6346 117.1112 5 128.2894 135.4231 119.4511

109.9117 102.3253 96.4343 62.6108

119.4511 115.6346 117.1112 109.9117 102.3253

Table 17: Results of Duncans post-hoc test for WPM (Malayalam).

Comparison of languages MANOVA showed significant difference between languages {S [F (3,325) = 85.323, p < 0.001], SPM - [F (3,325) =64.822, p < 0.001]} on syllables per second and syllables per minute. No significant differences between languages on WPM were evident. Also, no significant gender difference was observed. Table 18 shows results of post-hoc Duncans test for significant difference between languages. Results indicate no significant difference between Tamil and Kannada and significant difference between Telugu and other languages, and Malayalam and other languages (SS and SPM). Values in the same column are not significantly different.
N Language Tamil Kannada Telugu Malayalam Tamil Kannada Telugu Malayalam Subset SS 103 136 82 93 103 136 82 93 1 5.6720 5.7771 2 3

100

100

0.1 2.7 0.3 100

Table 19: Percent syllable type in four Dravidian

languages.

6.6651 7.8240 SPM 339.4596 346.7957 406.2927 468.6865

Table 18: Results of Duncans test on significant difference between languages.

One of the criteria for successful treatment outcome in fluency disorders is a speech rate within normal limits. This is to ensure that a reduction in stuttering is not achieved by abnormally slowing down speech rate that might adversely affect speech naturalness. A major basis of this study was that speech rate data available in the literature is not appropriate for formulating target rates in rate control therapies because rate of speech/ reading depends on age and language. A rate at the lower boundary of 95% confidence interval for mean may be appropriate for setting goals in rate control therapies. That is the rate as
36

JAIISH, Vol. 27, 2008

Rate of Speech

prescribed in Appendix I may be set as a goal in rate control therapy depending upon the age and language of the subjects.

Summary and Conclusions


Rate of speech is an important variable in the evaluation and treatment of fluency disorders. It is well known that the rate of speech correlates negatively with the severity of stuttering because frequent and/or long duration stutters result in reduced speech output. However, a deliberate reduction in the rate of speech has a beneficial effect on the frequency of stutters and, therefore, nearly all stuttering treatment approaches include rate reduction as one of their goals. Rate control is also a treatment target in cluttering. In addition, rate of speech is an important factor in the perceptual evaluation of normalcy of speech following treatment of stuttering. Extremely slow rate of speech, even if completely devoid of dysfluencies, is perceived as unnatural by listeners. It is important for the practicing clinician to know about these changes, for the assessment of stuttering and cluttering should be made by comparing the client's fluency with the level of fluency that would be expected for a normal person of the same age. Surprisingly, however, there are few empirically derived guidelines available for clinical measurement of rate and for setting goals for rate. Values of approximately 200 SPM or 150 SPM are frequently used in setting goals for rate of speech (Perkins, 1973; Boberg & Kully, 1985) because mean rates of adult discourse tend to converge around these values (Luchsinger, 1965). However, rate of speech depends on age, gender and language and therefore a common rate cant be set as a goal in rate control therapy for various age groups and languages. But, there is little objective data in support of these recommendations. In this context, the present study investigated rate of speech/ reading in 4 Dravidian languages and established normative data on rate of speech in Kannada, Tamil, Telugu, and Malayalam in subjects ranging in 10 decades. Four hundred and one (Kannada = 136, Telugu = 69, Tamil = 103, Malayalam = 93) normal subjects participated in the study. All subjects were literates and were from urban population. Informed consent was obtained from subjects prior to collection of speech samples. Subjects in the age range of 4-10 years described
37

cartoons and narrated stories and subjects in the age range of 11-100 years read passages which were developed by the investigators. All these were audio-recorded and analyzed to obtain syllables per second, syllables per minute and words per minute. Cool Edit software was used to eliminate pause and calculate duration. Results indicated significant difference between age groups, and languages. Rate of speech increased with increase in age till about 40 years and declined after 40 years (except Tamil). Malayalam had the highest SS and SPM compared to other languages. No significant difference between genders was observed. The results provide normative data for clinical purposes and to set rate of speech in persons with stuttering and cluttering.

References
Adams, M., Lewis, J., & Besozzi, T. (1973). The effect of reduced reading rate on stuttering fluency. Journal of Speech and Hearing Research, 16, 671-675. Bloodstein, O. (1944). Studies in the psychology of stuttering: XIX. The relationship between oral reading rate and severity of stuttering. Journal of Speech Disorders, 9, 161. Boberg, E., & Kully, D. (1985). Comprehensive stuttering program: Clinical Manual. San Diego, CA: College-Hill. Costello, J. M., & Ingham, R. J (1984). Assessment strategies in stuttering: In R. F. Curlee & W. H. Perkins (Eds.). Nature and treatment of stuttering: New directions. San Diego, Ca: College-Hill Press. Daly, D. A. (1986). The clutterer. In K. O. St.Louis (Ed.). The atypical stutterer. Orlando, Fl: Academic Press. Goldman-Eisler, F. (1968). Psycholinguistics: Experiments in spontaneous speech. New York: Academic Press. Haselagar, G. I. T., Slis, I. H., & Rietveld, A. C. M. (1991). An alternative method of studying the development of rate. Clinical Linguistics and Phonetics, 5, 53-63. Ingham, R. (1984). Stuttering and behavior therapy: current status and experimental foundations. San Dieogo, Ca: College-Hill Press.

JAIISH, Vol. 27, 2008

Rate of Speech

Ingham, R. J., & Cordes, A. K. (1997). Selfmeasurement and evaluating stuttering treatment efficacy. In R. F. Curlee & G. M. Siegel (Eds.). Nature and treatment of stuttering. Boston: Allyn and Bacon. Ingham, W., & Packman, A. C. (1978). Perceptual assessment and speaking rate and dysfluency of adult male and female stutteres and nonstutterers. Journal of Speech and Hearing Research, 21, 63-73. Johnson, W (1961). Measurements of oral reading and speaking rate and dysfluency of adult male and female stutterers and non stutterers. Journal of speech and Hearing Disorders (Monograph Supplement), 7, 120. Johnson, W., & Rosen, L. (1937). Studies in psychology of stuttering: VII. Effect of certain changes in speech pattern upon frequency of stuttering. Journal of Speech Disorders, 2, 105-109. Johnson, W., Darley, F. L., & Spriesterbach, D. C. (1963). Diagnostic methods in speech pathology. New York: Harper & Row. Klatt, D. (1973). Interaction between two factors that influence o\vowel duration. Journal of the Acoustical Society of America, 54, 11021104. Luchsinger, R. (1965). Voice Speech Language. Belmont, CA: Wadsworth Publishing Company. Lutz, K. C., & Mallard, A. R. (1986). Disfluencies and rate of speech in young adult nonstutterers. Journal of Fluency Disorders, 11, 307-316.

Meyers, S. C., & Freeman, F. J. (1985). Mother and child speech rates as a variable in stuttering and dysfluency. Journal of Speech and Hearing Research, 28, 436-444. Perkins, R. W. (1973). Replacement of stuttering with normal speech: II Clinical procedures. Journal of Speech and Hearing Disorders, 38, 283-294. Rathna, N., Subba Rao, T. A., & Bharadwaj, A. K. (1979). Rate of speech in Kannada. Cit. in C. S Venkatesh, G. Purushothama, & M. S. Poornima (1983). Normal rate of speech in Kananda. Journal of the All India Institute of Speech and Hearing, 15, 7-12. Sander, E. (1961). Reliability of the Iowa speech dysfluency test. Journal of Speech and Hearing Disorders (Monograph Supplement), 7, 21-30. Stephenson-Opsal, D., & Bernstein Ratner, N. (1988). Maternal speech rate modification and childhood stuttering. Journal of Fluency Disorders, 13, 49-56. Umeda, N., & Quinn, M. S. (1980). Some notes on reading: Talkers, material and reading rate. Journal of Speech and Hearing Research, 23, 56-72. Venkatesh, C. S., Purushothama, G., & Poornima, M. S. (1983). Normal rate of speech in Kannada. Journal of the All India Institute of Speech and Hearing. 15, 7-12. Walker, C., & Black, J. (1950). The intrinsic intensity of oral phrases (Joint Project Report No. 2). Pensacola, Fla,: Naval Air Station, U. S. Naval School of Aviation Medicine.

38

JAIISH, Vol. 27, 2008

Rate of Speech

Appendix I Normative data on rate of speech/reading in 4 Dravidian languages


Age 3-3.11 4-4.11 5-5.11 6-6.11 7-10 11-15 16-20 21-30 31-40 41-50 51-60 61-70 71-80 81-90 91-100 A M SD M SD M SD M SD M SD M SD M SD M SD M SD M SD M SD M SD M SD M SD M SD M SD SS 4.9 1.3 4.2 .79 4.2 1.3 4.3 0.84 4.1 0.65 5.7 .83 7.0 .96 6.5 .96 7.2 .31 6.8 .87 6.9 .62 6.9 .73 6.4 .85 5.6 .97 Kannada SPM WPM 291 78 252 48 254 79 261 50 250 41 343 48 425 58 386 70 434 19 410 53 415 37 404 52 390 51 337 58 129 34 91 17 120 37 92 18 85 14 104 14 127 17 116 21 131 6. 124 16 124 11 125 12 118 15 102 17 M = Mean, A= Average Telugu SS SPM WPM SS Tamil SPM WPM SS Malayalam SPM WPM

7 1.25 7 0.61 8 0.58 6 0.95 7 1.04 6 1.04 6 0.99 5 0.19 5.14 0.19

431 132 439 36 466 36 384 54 389 75 392 69 336 60 309 11 309 11.3

123 37 125 10 133 10 116 23 117 20 110 18 96 17 89 4 88.8 3.85

5 1 4.65 1 6.7 1.67 4.5 .4 6.5 .7 6.5 1.2 5.5 1.1 5.9 .8 5.4 .9 5.7 1 5.9 .9

299 62 278 64 402 100 272 22 384 53 391 74 326 68 353 49 326 52 340 62 354 54

114 24 86 20 130 31 95 8 133 18 136 25 114 24 122 17 115 20 118 21 123 19

5.3 2.20

318 133

104 40

6.4 0.76

384 53.8

111 15.9

5.7 1.2

34 72

117 25

7.9 .8 8.8 1.2 9.3 .9 8.2 .6 7.9 1.2 7.9 1 7.8 .6 7 1.6 6.6 .86 4.3 .7 7.8 1.4

474 51 529 73 558 62 492 36 477 81 483 67 448 66 423 99 398 52 257 44 469 88

115 12 128 18 135 15 119 9 116 20 117 16 110 12 102 24 96 13 63 11 114 21

39

JAIISH, Vol. 27, 2008

Rate of Speech

95% confidence intervals for mean LB: Lower boundary, UB: Upper boundary Kannada Age SS SPM WPM LB UB LB UB LB UB 3-3.11 3.9 5.8 235 347 104 153 4-4.11 3.6 4.8 218 286 79 104 5-5.11 3.3 5.2 198 310 93 146 6-6.11 3.7 4.9 225 296 80 105 7-10 3.7 4.6 220 279 75 94 11-15 5.1 6.3 309 377 94 114 16-20 6.3 7.7 383 466 115 139 21-30 5.8 7.2 336 435 101 131 31-40 7.0 7.4 421 448 127 135 41-50 6.2 7.4 372 448 112 135 51-60 6.4 7.3 388 442 116 132 61-70 6.4 7.5 366 441 117 133 71-80 5.8 7.1 353 426 107 129 81-90 4.6 6.6 276 399 83 120

Age 5-5.11 6-6.11 7-10 11-15 16-20 21-30 31-40 41-50 51-60 61-70 71-80

SS LB UB 3.3 6.6 3.7 5.5 5.5 7.9 4.3 4.8 6.0 7.0 5.6 7.4 4.7 6.2 5.3 6.5 4.8 6.0 4.9 6.4 5.2 6.6

Tamil SPM LB UB 200 399 225 333 330 473 256 287 346 421 338 443 278 374 317 388 289 364 296 384 315 393

WPM LB UB 76 152 69 103 101 146 89 100 120 146 117 154 97 130 110 135 100 129 103 133 110 137

Age 11-15 16-20 21-30 31-40 41-50 51-60 61-70 71-80 81-90

SS LB UB 5.4 7.1 6.1 7.9 7.0 7.8 7.3 8.1 5.7 7.1 6.2 7.6 5.7 7.2 4.9 6.2 3.4 6.9

Telugu SPM LB UB 315 567 367 475 413 464 441 492 346 422 336 443 342 441 297 377 207 411

Age WPM LB UB 90 162 105 136 118 132 126 141 100 132 102 131 97 123 85 108 54 123 11-15 16-20 21-30 31-40 41-50 51-60 61-70 71-80 81-90 91-100

Malayalam SS SPM LB UB LB UB 7.3 8.5 438 511 7.9 9.7 476 581 8.6 10 514 602 7.7 8.6 467 518 7.0 8.8 419 535 7.2 8.7 435 531 7.4 8.2 400 495 5.9 8.2 352 494 6.0 7.2 360 435 2.5 6.0 148 367

WPM LB UB 106 124 116 141 125 146 113 126 101 130 106 129 101 119 85 119 87 106 36 89

40

JAIISH, Vol. 27, 2008

Speech Rhythm in Hearing-Impaired Children

Speech Rhythm in Hearing-Impaired Children


1

Savithri S.R., 2Ruchi Agarwal & 3 Johnsi Rani R.

Abstract
Rhythm, a prosodic feature, refers to an event repeated regularly over a period of time. Research on speech rhythm in the last few years has been largely concerned with attempting to search for the acoustic correlates of rhythmic classification. Speech rhythm in pathological cases has not been studied extensively. In this context, the present study investigated speech rhythm in normal and hearing-impaired children using Pairwise Variability Index (PVI) index. The PVI is one measure used to calculate the extent of unit-to-unit variation in speech. Twenty normal hearing and twenty hearing impaired Kannada speaking children in the age range of 5-10 years participated in the study. The subject's spontaneous speech and narrations were audio recorded and stored onto computer. The vocalic and intervocalic durations were measured using Cool Edit Pro software. The results revealed longer vocalic and intervocalic durations in hearingimpaired children in comparison with normals. It is concluded that both the hearingimpaired and normal children have unclassified rhythmic patterns.
Key words: Rhythm, Pair wise variability index, Rhythmic patterns.

Rhythm, a prosodicfeature, refers to an event repeated regularly overa period of time. In stresstimed languages, intervals between stresses or rhythm are said to be near equal, whereas in syllable-timed languages successive syllables are said to be of near-equal length. A third type of rhythm, mora timing, was proposed by Bloch (1950), Han (1962), and Ladefoged (1975) where successive morae are said to be near equal in duration. Mora- timing was exemplified by Japanese. The Pairwise Variability Index (PVI) is a quantitative measure of acoustic correlates of speech rhythm, which calculates the pattering of successive vocalic and intervocalic (or consonantal) intervals showing how one linguistic unit differs from its neighbour (Low, 1998). Grabe & Low (2000) developed normalized Pairwise Variability Index (nPVI) for rhythmic analysis of vocalic durations. The raw Pairwise Variability Index" (rPVI) is used for rhythmic analysis of intervocalic durations. Table 1 summarizes the basic characteristics of each language class regarding relative values of vocalic nPVI and intervocalic rPVI.
1

Language Intervocalic Vocalic Languages Class rPVI nPVI StressEnglish, High High timed Germany SyllableFrench, High Low timed Spanish Mora-timed Japanese Low Low Table 1: Summary of basic characteristics of each language class regarding relative values of vocalic nPVI and intervocalic rPVI.

Few investigators have reported rhythm in pathological population. Dankovicova, Gurd, Marshall, Macmohan, Stuart-Smith, Coleman, & Slater (2001) reported that speech of English speaker individual with foreign accent Syndrome and ataxic dysarthria is more syllable-timed than that of normal controls. Using an early forerunner of the PVI, Ackerman & Hertrich (1994) and Kent, Rosenbek, Vorperian, & Weismer (1997) found little evidence of syllable timing for ataxic population. As there is dearth of studies in rhythm for hearing impaired, the present study investigated the rhythm in normal and hearingimpaired children.

Professor, Dept. of Speech-Language Sciences, All India Institute of Speech & Hearing, Manasagangothri, Mysore570006, email:savithri_2k@yahoo.com. 2Research Officer, Dept. of Speech-Language Sciences, All India Institute of Speech and Hearing, Manasagangothri, Mysore-6, email:ruchislp@rediffmail.com, 3Research Officer, Dept. of SpeechLanguage Sciences, All India Institute of Speech and Hearing, Manasagangothri, Mysore-570006, email: rjohnsirani@yahoo.com.

41

JAIISH, Vol. 27, 2008

Speech Rhythm in Hearing-Impaired Children

Method
Subjects: The experimental group included 20 hearing impaired Kannada speaking children in the age range of 5-10 years. All subjects had bilateral severe sensori-neural hearing loss, normal oral structures, no other associated psychological or neurological problems, language age of atleast 3 years on REELS and all subjects attended regular speech therapy or normal school. The control group consisted of age, gender and language age matched Kannada speaking 20 normal children. Material: Pictures developed by Indu (1991), Nagapoornima(1991), Yamini (1991) and Rajendra Swamy (1992) were used to elicit speech. In addition story telling was also used. Procedure: Subjects were instructed to describe the pictures, narrate the story, which were audiorecorded using MZ-R30 digital Sony recorder and stored onto the computer. Waveform display obtained from Cool Edit Pro software was used to measure vocalic (V) and intervocalic (IV) interval. The vocalic measure (nPVI) refers to the duration of vowel, which was measured as the time duration from the onset of voicing to the offset of voicing for the vowels. Intervocalic measure (rPVI) refers to the duration between two vocalic segments. It was measured as the time duration between the offset of the previous vocalic segment to the onset of subsequent vocalic segment. A program in C language was developed (Vasanthalakshmi, 2005) to compute nPVI and rPVI. The raw Pairwise Variability Index (rPVI) is defined as follows:

Pairwise Variability Index (nPVI) is defined as follows:

Where, m is the number of intervals and dk is the duration of the kth interval. The duration difference between the first and second, the second and third vocalic segment and so on was averaged to get nPVI. The same procedure was used to obtain averaged intervocalic durations. Pauses between intonation phrases, as well as hesitations, were excluded from the analysis. Statistical analysis: The mean rPVI and mean nPVI values were calculated for both normal and hearing impaired children. Repeated measure ANOVA was used to obtain significant differences, if any, between groups.

Results and Discussion


Independent t-test indicated a significant difference between groups on rPVI (intervocalic) [t (38) = 2.54, p<0.05] and nPVI (vocalic) [t (38) = 2.30, p<0.05] values. Both rPVI and nPVI values were higher in hearing-impaired children compared to normal children. Within the normal group, no statistically significant difference was obtained between genders for rPVI [t (18) = 0.217, p> 0.05] and nPVI [t (18) = 0.293, p>0.05]. Also, within the hearing impaired group, no statistically significant difference was obtained between genders for rPVI [t (18) = 0.365, p> 0.05] and nPVI [t (18) = 0.685, p>0.05]. Table 2 shows the rPVI and nPVI values of normal and hearing impaired children. Mean and Standard deviation of normal and hearing-impaired children is depicted in figure 1.

Where, m is the number of intervals and dk is the duration of the kth interval. The normalized

42

JAIISH, Vol. 27, 2008

Speech Rhythm in Hearing-Impaired Children

Normals Hearing-Impaired rPVI nPVI rPVI nPVI 1 13.80 64.16 18.14 71.01 2 14.15 67.15 31.40 73.02 3 12.99 60.66 31.51 62.81 4 16.47 64.04 19.50 58.05 5 16.47 62.37 15.54 79.55 6 17.68 60.90 13.28 77.11 7 15.75 63.21 17.86 83.91 8 18.71 65.83 10.74 57.57 9 17.79 62.08 27.23 76.70 10 13.23 64.85 10.09 70.72 11 19.63 59.26 26.21 61.88 12 13.86 63.55 30.10 78.66 13 15.52 62.30 10.30 60.93 14 16.92 63.60 15.78 58.06 15 17.20 68.27 15.55 59.88 16 12.69 61.85 32.50 58.66 17 15.56 54.81 21.40 61.56 18 21.01 56.99 11.63 62.08 19 14.02 62.94 18.58 63.26 20 10.63 61.17 33.65 67.57 Mean 15.70 62.49 20.54 67.14 Table 2: rPVI and nPVI values in normal and hearing impaired children.
80

Subject

timed, syllable-timed, mora-timed) since the nPVI values were found to be higher than values of rPVI in both groups which implies that the usage of vowels in their speech was more; subjects are still in the acquisition stage of rhythm. Comparison of the PVI values in adults and children showed that nPVI values were higher in children than the adults whereas rPVI values were higher in adults when compared to children. This may be due to the reason that the speech task in children is unpredictable unlike the reading task by adults and hence would have an influence on differences in PVI values. Figure 2 shows the PVIs in children and adults. adults.

70

Figure 2: Mean rPVI and nPVI values for normal, hearing-impaired children and adults.

60

50

40

Groups
30 Normals

20 10 nPVI rPVI

HI

PVIs

Also, the types of syllable used in children and adults were different. The percentage use of different types of syllable structures by normal and hearing impaired children and adults (Savithri et.al.2005) was calculated and tabulated. Results indicated that the type of syllable structure used by three groups differed. Children used V, CV and CCV syllables, and adults, in addition, used CCV syllables. The percentage use of V and CV syllables was more in HI children compared to normal children. Table 3 shows the percentage of various syllables used in three groups of subjects.
Groups V CV CCV CCCV Normal children 21.15 61.62 17.21 0 HI Children 23.27 68.82 7.95 0 Adults 6.86 80.63 12.38 0.11 Table 3: Percentage use of different syllable structures.

Figure 1: Mean and standard deviation in normals and hearing-impaired.

The results indicated high nPVI and low rPVI values in both groups and therefore the rhythmic pattern found in hearing impaired and normal children remains unclassified and cannot be placed in any of the rhythmic classes (stresstimed, syllable-timed, mora-timed). The results indicated several points of interest. Firstly, the results indicated that the rhythmic pattern found in hearing impaired and normal children remained unclassified and cannot be placed in any of the rhythmic classes (stress43

95% CI for Scores

The usage of more vocalic syllables and less CV syllables in children compared to adults might be a reason for low rPVI.

Conclusions
Rhythm has been defined as an effect involving the isochronous recurrence of some type of speech unit. Basically languages have

JAIISH, Vol. 27, 2008

Speech Rhythm in Hearing-Impaired Children

been organized under three types of rhythm i.e. stress-timed, syllable-timed and mora-timed. The present study investigated the rhythm in normal and hearing-impaired children. Pair-wise Variability Indices (PVIs) were used to find the vocalic and intervocalic durations. The vocalic measure (nPVI) refers to the duration of vowel, which was measured as the time duration from the onset of voicing to the offset of voicing for the vowels. Intervocalic measure (rPVI) refers to the duration between two vocalic segments. The results showed that the speech rhythm in normal and hearing impaired children remained unclassified. No gender differences were observed in any group. A high nPVI value and low rPVI value was obtained in both groups. Savithri, Jayaram, Kedarnath, & Goswami (2005) classified Kannada as a mora- timed language in normal adults. But, the results of the present study are not in consonance with the earlier study, which indicates that the acquisition of adult-like rhythm is not yet achieved in the children of the present study. The syllabic structure used by children also differed from that of the adults. Therefore the results of the present study reveals that the syllabic structure used by children is simpler than the adults and children are in acquisition stage of rhythmic patterns. Thus, it implies that there is a need to develop the normative data for the age at which children acquire the adult like rhythmic pattern.

hypothesis. In C. Gussenhoven & N. Warner (Eds.). Laboratory Phonology. 7, 515-546. Berlin: Mouton de Gruyter. Han, M. S. (1962). The feature of duration in Japanese. Onsei no kenkyuu, 10, 65-80. Indu (1990). Some aspects of fluency in children: 4-5 years. M. Jayaram, & S.R. Savithri (Eds.). Dissertation abstract: Research at AIISH, Vol.2, pp 171-173. Kent, R., Rosenbek, J., Vorperian, H., & Weismer, G. (1997). A speaking task analysis of the dysarthria in cerebellar disease. Folia Phoniatrica et Logopaedica, 49, 63-82. Ladefoged, P. (1975). A Course in Phonetics. New York: Harcourt Brace Jovanovich. Low, E. L. (1998). Prosodic prominence in Singapore English. Unpublished Ph.D. Thesis, University of Cambridge. Nagapoornima, M.N. (1990). Dysfluencies in children: 3-4 years. M. Jayaram, & S.R. Savithri (Eds.). Dissertation abstract: Research at AIISH, Vol.2, pp 171-173. Rajendra Swamy, (1992). Some aspects of fluency in children: 6-7 years. M. Jayaram, & S.R. Savithri (Eds.). Dissertation abstract: Research at AIISH, Vol.3, pp 6-7. Savithri, S.R., Jayaram, M., Kedarnath, D., Goswami, S. (2005). Rate of speech /Reading in Dravidian languages. Journal of the Acoustic Society of India, 33, 352355. Vasantalakshmi (2005). Development of C language program. AIISH, Mysore. Yamini, B.K. (1990). Dysfluencies in children: 56 years. M. Jayaram, & S.R. Savithri (Eds.). Dissertation abstract: Research at AIISH, Vol.2, 171-173.

References
Ackermann, H., & Hertrich, I. (1994). Speech rate and rhythm in cerebellar dysarthria: an acoustic analysis of syllable timing. Folia Phoniatrica, 46, 72-78. Bloch, B. (1950). Studies in colloquial Japanese IV: Phonemics. Language, 26, 86-125. Dankovicova, J., Gurd, J., Marshall, J., Macmohan, M., Stuart-Smith, J., Coleman, J., Slater, A. (2001). Aspects of non-native pronunciation in a case of altered accent following stroke (foreign accent syndrome). Clinical Linguistics and Phonetics, 15, 3, 195-218. Grabe, E. & Low, E. L. (2000). Durational variability in speech and rhythm class

Acknowledgements
The authors acknowledge the financial support of ARF. The authors would like to thank Dr.Vijayalakshmi Basavaraj, Director, AIISH, for permitting to publish the paper. They also thank all subjects for their participation in the study.

44

JAIISH, Vol. 27, 2008

Semantic Association

The Semantic Association in the Mental Lexicon


1

Gopee Krishnan & 2Shivani Tiwari

Abstract
The organization of mental lexicon has been extensively debated and discussed in the contemporary psycholinguistics. Specifically, this study investigated into the nature of organization of semantically related and unrelated concepts in the mental lexicon. A group of 19 participants (age range: 17-23 years; 11 females & 8 males) was required to judge the semantic association between the word pairs presented through reaction time software (DMDX). The participants judged the semantically associated word pairs faster compared to the semantically unrelated pairs. This finding could be explained by the Spreading activation theory of lexical processing (Collins & Loftus, 1975; Dell, 1986). The semantic features that are in common to the words of the stimulus pairs received double activation and this facilitated a faster judgment in the case of semantically associated word pairs compared to unassociated word pairs.
Key words: Spreading activation theory, Speech production, Architecture of the mental lexicon, Lexical processing.

The selection of words during speech production is an effortless act for a native speaker. However, the underlying processes in word production are far from the simplicity with which it is performed. For example, while naming a picture, the speaker performs a visual analysis to identify the features of the picture and activates the conceptual knowledge (lemma) associated with that picture. The activated lemma further activates the word form (lexeme) associated with it down the process. This is known as the lexical selection. Once the lexical item associated with the concept in question is selected, the phonological encoding takes place, where the speaker correctly selects the various phonemes necessary for the speech articulation and these selected phonemes are sent to the speech articulation circuit for their execution. Thus, a seemingly simple task such as picture naming involves various underlying processes such as visual analysis, semantic activation, lexical selection, phonological encoding and finally the speech articulation (Costa, Colom , & Caramazza, 2000). Spreading Activation Theory in Lexical Selection The concept of Spreading Activation Theory an idea originally introduced by Collins and
1

Loftus (1975) has received widespread acceptance in the contemporary cognitivelinguistic literature. The notion behind this theory is that each concept spreads a proportion of its activation to other representations with which it is linked. For example, when naming the picture of a dog, the concepts associated with that stimulus such as an animal, has a tail, has four legs, pet, faithful etc. are activated. However, some of these features are also applicable to other animals too; say cat. The spreading activation theory thus postulates that the presentation of the picture of a dog also partly activates the concept cat and other members that share the similar features (Caramazza, 1997; Collins & Loftus, 1975; Dell, 1986). In other words, the concept cat becomes a competitor while selecting the concept of dog (Semantic Interference Effect) (Glaser & Glaser, 1989; Roelofs, 1992; Starreveld & La Heij, 1995). However, under normal conditions, the speaker does not face such difficulties as s/he correctly picks up the right item (dog). This selection process could be damaged in aphasic subjects leading to, what is known as semantic paraphasias (Caramazza & Hillis, 1990). The partial activation received by the semantically related concepts has some important bearing on our understanding of the functional

Assistant Professor, Dept. of Speech and Hearing, Manipal College of Allied Health Sciences, Manipal University, Manipal-576 104, email:brain.language.krishnan@gmail.com , 2Lecturer, Dept. of Speech and Hearing, Manipal College of Allied Health Sciences, Manipal University, Manipal-576 104, email:shivanitiwari8@yahoo.co.in.

45

JAIISH, Vol. 27, 2008

Semantic Association

architecture of mental lexicon. In the previous example, the presentation of the stimulus dog activates other semantically related concepts such as cat and various other concepts in the mental lexicon that share some of the features (not all) of the target item. Though there is a lack of consensus on the amount of activation received by each of these related concepts in the mental lexicon, it is widely accepted that the amount of activation received by related concepts is proportional to the number of features shared by both the target and the related items. Therefore, it is possible to assume that the concept of cat receives higher activation compared to other members that do not share any features with the target item (for example, stone). In a way, we can assume that the concepts cat and dog are closely located in the mental lexicon compared to concepts stone and dog. This type of conceptualization about the organization of the items in the semantic storage has gained strong evidences from reaction time studies. A related and simple, yet interesting question is the robustness with which a word-pair is judged on its semantic association. Put it in a simpler way, are semantically associated word-pairs judged faster compared to unassociated pairs on their semantic association? Rubenstein, Lewis, and Rubenstein (1971) and Stanners and colleagues (1971) have reported that semantically associated word pairs are judged faster with compared to the unassociated word pairs. However, this needs to be tested empirically again to check the validity of the finding as well as a theoretical explanation should be put forth for the observed finding. We address this issue in this study. Objectives The objectives of the study were to replicate the findings of previous findings and more importantly, if similar findings were obtained, provide an explanatory hypothesis for the faster judgment time in the case of semantically associated word pairs compared to the unassociated word pairs.

as their medium of instruction starting at the age of 4-5 years. The mean age of the participants was 20 years (age range 17-23 years). Materials A pool of 110 items consisting of 56 semantically related and 54 unrelated items was initially selected. Five proficient English speakers rated these items for their semantic association. The raters task was to write either yes or no against each word-pair if the pair was semantically related or unrelated, respectively. Three raters did not agree on three semantically associated items and two semantically unassociated items. One rater did not agree on two semantically associated and two semantically unassociated items. However, the items the single rater did not agree were same as that of the other three raters, therefore, finally rejecting three semantically related and two semantically unrelated items from the test stimuli. Thus, the final version consisted of 53 semantically related and 52 unrelated word pairs. Among these 105 items, three semantically related and two semantically unrelated items were randomly selected for training purpose. Procedure The subjects were made to sit in a soundproof room and verbal instructions were given about the task. This was followed by the presentation of training items and the subjects were made familiar with the task and the response. The stimuli were presented through a computer using DMDX reaction time software (Foster & Foster, 2003). A semantically associated word-pair was indicated by m button press and unassociated pair by n button press on the keyboard. The subjects were instructed to rest their middle and index fingers on these buttons while performing the task in order to avoid time delay in reaching the button while responding. The stimulus words appeared as black capital letters in Times New Roman font in white background. The font size remained 26 across the stimuli. Before the presentation of each stimulus, a fixation point (+) appeared for 500 ms in the center of the screen on which the participants were instructed to fixate. This was followed by the first word of the word pair for a duration of 750 ms. This was further followed by a blank screen for 500 ms and the second word of the word pair. The second word remained on the screen for 2000 ms. The DMDXs clock was set on with the presentation of the second word. 100 word-pairs
46

Method
Subjects Nineteen subjects (11 females & 8 males) volunteered to participate in the present study. The subjects were the undergraduate students of Manipal University. All the subjects had English

JAIISH, Vol. 27, 2008

Semantic Association

were randomly categorized into five blocks of 20 each. At the end of each block, a rest period (1 minute) was given and for each subject, the testing was completed in a single sitting. The chronological sequence of the testing procedure is seen in Figure 1.

Figure 2: The reaction times (ms) on semantically associated versus unassociated conditions across the subjects.

Discussion
The findings of the present study supported that of similar studies done in the past (e.g., Rubenstein et al., 1971; Stanners et al., 1971). In the following section, we provide an explanatory hypothesis for the observed findings from the perspectives of spreading activation theory of lexical access (Collins & Loftus, 1975; Dell, 1986). As mentioned in the introduction, the members in the mental lexicon receive partial activation when a related item is activated. The activation strength is a function of the number of features shared by the target item with its distracters (Caramazza, 1997). Therefore, an item that shares a large number of features with the target item will be highly activated compared to the items that receive only minimal activation. These highly activated items lexical nodes could act as strong competitors to the target items at the lexical selection stage. Reaction time studies have added significantly to our existing knowledge on the semantic organization in the mental lexicon. In the current study, all the subjects required lesser time to judge a word pair as semantically associated compared to one that was semantically unassociated. According to the Spreading Activation Theory (Collins & Loftus, 1975; Dell, 1986), lesser amount of time for semantically related word pair could be interpreted as follows: upon seeing the first word, the subject activates its corresponding semantic concept from his/her mental lexicon. This partially activates the semantically related items (to the target) as well. The presentation of the second word of the word pair soon after the first word elicits an activation of its corresponding semantic representation. This
47

Figure 1: The chronological sequence of the testing procedure.

Results
The responses of the trial items were eliminated from the reaction time analysis. The remaining data was analyzed with SPSS.11 software for Windows. For the entire group of subjects, 73/1026 responses (7.11 %) in semantically associated condition and 72/874 responses (8.23%) were either wrong or no responses. For the statistical analysis, the reaction time (RT) from correct responses was used. The group mean for the semantically associated condition was 737 ms (SD = 134) whereas in the unassociated condition, the mean RT was 866 ms (SD = 171). The mean reaction times were submitted to Paired sample Student t-test to find out the differences between the two conditions, if any. The t-test results revealed a significant difference between the semantically associated and unassociated word pairs (t = -6.51, p < 0.001). The individual performance across the subjects is given in Figure 2. A closer look at the Figure 2 reveals that the reaction times were shorted for semantically associated word pairs on an individual basis.

JAIISH, Vol. 27, 2008

Semantic Association

in turn could activate some of the features of the first word that are just activated by the first word. Therefore, as depicted in Figure 3, the set of features shared by both words (for example, animal) are highly activated compared to other features that are not common to both the words of the stimulus pair. During the semantic association judgment, these highly activated semantic features could facilitate a faster yes response.

Figure 4: Absence of heightened activation (semantic association) in semantically unrelated word pair.

Conclusions
The current study supported the findings of previous similar studies on the representation of associated and unassociated words in the mental lexicon. More importantly, an explanatory hypothesis based on the spreading activation theory has been put forth to explain the observed findings. The mechanism behind faster judgment time in the case of semantically associated word pairs in contrast to the unassociated word pairs may hypothesized be due to the presence/absence of heightened activation (semantic association). That is, the presence of heightened activation terminates the search and a yes response is made whereas the absence of such activation demands continued search until all the features are searched, in order to make an accurate no response, at the expense of increased response time.

Figure 3: An increased activation of the semantic feature (animal) shared by both words of the stimulus pair.

In the case of a semantically unassociated word-pair (Figure 4), each word activates a set of semantic features corresponding to its concept. However, there is no facilitation of any semantic features resulting from the lack of overlap of features between the word pairs, unlike in semantically associated condition. Hence, the subject has to search for all the semantic features to ascertain the presence of any heightened activation (i.e., semantic association) before making an accurate no judgment. Logically, this process is more time consuming compared to semantically associated condition where the presence of a highly activated feature (semantically associated) ascertain the semantic association between the words of the stimulus pair. In simpler terms, in the absence of any such heightened activation, the subjects need to search the entire semantic features (of both words of the word pair) before making a correct response; at the expense of increased response time.

References
Caramazza, A., & Hillis, A. E. (1990). Where do semantic errors come from? Cortex, 26, 95 122. Caramazza, A. (1997). How many levels of processing are there in lexical access? Cognitive Neuropsychology, 14, 177 208. Collins, A. M., & Loftus, E. F. (1975). A spreading-activation theory of semantic processing. Psychological Review, 82 (6), 407 428. Costa, A. Colom , A., & Caramazza, A. (2000). Lexical Access in Speech Production: The
48

JAIISH, Vol. 27, 2008

Semantic Association

Bilingual Case. Psychologica, 21, 403 437. Dell, G. S. (1986). A spreading activation theory of retrieval in sentence production. Psychological Review, 93, 283 321. Forster, K. I., & Forster, J. C. (2003). DMDX: A Windows display program with millisecond accuracy. Behavior Research Methods, Instruments, & Computers, 35, 116124. Glaser, W. R., & Glaser, M. O. (1989). Context effect in Stroop-like word and picture processing. Journal of Experimental Psychology: General, 118, 13 42. Roelofs, A. (1992). A spreading-activation theory of lemma retrieval in speaking. Cognition, 42, 107 142. Rubenstein, H., Lewis, S., & Rubenstein, M. (1971). Evidence of phonemic recording in

visual word recognition. Journal of Verbal Learning and Verbal Behavior, 10, 645-657. Stanners, M. S., Peterson, A., & Waters, G. S. (1971). Reading without semantics. Quarterly Journal of Experimental Psychology, 23, 111-138. Starreveld, P. A., & La Heij, W. (1995). Semantic interference, orthographic facilitation and their interaction in naming tasks. Journal of Experimental Psychology: Learning, Memory, and Cognition, 21, 686 698.

Acknowledgments
We thank Ms. Vanessa and Ms. Elizabeth for their valuable help during the current study. We also thank Dr. B. Rajashekar, Head of the Dept. of Speech and Hearing, Manipal College of Allied Health Sciences, Manipal University, India, for permitting us to undertake this study.

49

JAIISH, Vol. 27, 2008

Semantic Pragmatic Attributes and Cognition in Schizophrenics

Semantic Pragmatic Attributes and Cognition in Acute and Chronic Schizophrenics: A Case Comparative Study
1

Mithila Poonacha, 2Shivani Tiwari & 3Rajashekhar Bellur

Abstract
Schizophrenia is a thought disorder, displaying unusual language and cognitive impairments. There exists a dearth of studies relating the language deficits to the onset of the disorder. This study profiled few aspects of semantic and pragmatic abilities in acute and chronic schizophrenics and compared with their cognitive abilities. One acute schizophrenic and one chronic schizophrenic patient participated in the study. Cognition was assessed using the Addenbrookes Cognitive Examination (ACE-R) Kannada version. Aspects of semantics (semantic storage, recall/access and the word association) and pragmatics were assessed using a test battery.Results showed distinct variations in both subjects in cognitive as well as linguistic aspects (semantic and pragmatic). Abnormalities were found both at single word level as well as discourse comparable to dysfunction of cognition, and onset of the disorder. The findings highlighted the differences in semantic-pragmatic and cognitive aspects in terms of onset of the disorder. Though, the study is a preliminary attempt and warrants further research for substantiation.
Key words: Word association, Discourse.

Language disorder has long been considered a diagnostic indicator of schizophrenic disorder (American Psychiatric Association, 1994). Various distinct hypotheses have been put forth by several researchers, regarding the root problem underlying language dysfunction. Many psychopathologists regard speech disturbances as reflective of an underlying disorder of thinking. While, content and form of schizophrenic speech has been described as deviant by other group of authors. The language disturbances in schizophrenics could be at individual levels or a combination of different levels. Semantics refers to the meaning of words. Several investigators have reported that patients with schizophrenia are slower and less accurate in words/ word pairs as members of conceptual categories (Chen, Wilkins, McKenna, 1994). Some other studies also suggest that schizophrenia maybe characterized by a disorganized semantic memory store. Pragmatics is the study of how language is used and how language is integrated in

the context. A number of researchers have concluded that the primary language impairment in schizophrenia is in the area of pragmatic performance. Crow, (1998) argued that the language disturbances in schizophrenia are a reflection of the way in which individuals with schizophrenia use language. Cognition refers to the mental processes used in the acquisition and use of language including sensations, perception, attention, learning, memory, language, visuospatial abilities, thinking, and reasoning. Schizophrenia is often associated with cognitive deficits, particularly executive function, attention, memory and language. Specific cognitive deficits have been linked to psychotic phenomena, including verbal hallucinations and disorganized speech. In addition, selective deficits have also been described in the pattern of retrieval from both semantic and episodic memory. Clinically, cognitive dysfunction is a direct predictor of poor social functioning. The existence of specific patterns of cognitive dysfunction

1 Post Graduate Student, Dept. of Speech and Hearing, Manipal College of Allied Health Sciences, Manipal University, Manipal-576 104, 2Lecturer, Dept. of Speech and Hearing, Manipal College of Allied Health Sciences, Manipal University, Manipal-576 104, email:shivanitiwari8@yahoo.co, 3Professor, Dept. of Speech and Hearing, Manipal College of Allied Health Sciences, Manipal University, Manipal-576 104.

50

JAIISH, Vol. 27, 2008

Semantic Pragmatic Attributes and Cognition in Schizophrenics

suggests several important avenues for future research. However, the nature of language disturbance in relation to the onset of the disorder is poorly understood. Thus, the study aimed at investigating the semantic and pragmatic skills in subjects with acute and chronic schizophrenia and comparing with cognition.
Details Acute Schizophrenic (X) 30 yrs, female Kannada B.Sc. Computer science (incomplete) c/o no problem 2 months Maintaining well Wandering behavior, anger outbursts, decreased personal care, decreased sleep, talking and smiling to self, increased & irrelevant speech, inappropriate dressing, belief that black magic has been done on her. Auditory hallucinations, delusion of love (?), sociooccupational dysfunction. Chronic Schizophrenic (Y) 24yrs, male Kannada S.S.L.C (fail) c/o hand pain since 4 days 2 years Well-adjusted personality Decreased interaction, increased suspiciousness and abusive behavior, withdrawal from family, smiling & muttering to self, irrelevant talk, increased religiosity.

Method
Subjects One acute schizophrenic and one chronic schizophrenic, diagnosed by the consultant psychiatrist based on DSM-IV (A) criteria, participated in the study (Table 1). The subjects were recruited from Kasturba Hospital, Manipal and from Government Hospital, Udupi, Karnataka. Materials and Procedure The subjects were tested for cognition, semantics and pragmatics: Cognition: Cognitive abilities of the subjects were assessed using Addenbrookes Cognitive Examination (ACE). The first adaptation of this test in Indian language (Malayalam) was provided by Mathuranath, Hodges, Mathew, Cherian, George, and Bak (2004). Further, this Malayalam version of ACE (M-ACE) was validated on 488 subjects of age 55 75 yrs (Mathuranath, Cherian, Mathew, George, Alexander, Sarma, 2007). Kannada version of ACE was developed and standardized on 68 subjects (age 40 74 yrs), by the Department of Neurology, Kasturba Hospital, Manipal, (2007). The test checks cognition under 5 sections of Attention and Concentration, Memory, Verbal Fluency, Language, and Visuospatial abilities. Table 2 provides the subdivision of the 5 sections of ACE and the split of scores, with the total score summing to 100. Instructions for the test were given verbally, except for the section of Language for which instructions were given in writing. The obtained scores were then compared to the normative (cut-off score <88). Semantics: Included three measures: a. Semantic storage: A spoken word-topicture matching task was given involving the presentation of 10 pictures from various lexical categories. Score of 1 was given for each correct response. b. Recall/access: Included three tasks: (i) Confrontation naming task: 5 nouns from various lexical categories and 5 verbs were presented and the subject named the picture. Score of 1 was given for every correct response. *(Pictures were black & white line diagrams taken from With a little bit of help, language training manual). (ii) Category fluency task: 2 lexical categories were given and subjects
51

Age/Gender Language Education Complaint Onset Pre-Morbid History Post-Morbid History

Diagnostic Criteria

Treatment

On antipsychotic treatment for a month

Auditory hallucinations, delusion of persecution, delusion of reference, delusion of grandiose, breaks in the train of thoughts, selfabsorbed attitude, sociooccupational dysfunction. On antipsychotic treatment along with electroconvulsive treatment (ECT) for more than a year

Table 1: Case details and demographics.

JAIISH, Vol. 27, 2008

Semantic Pragmatic Attributes and Cognition in Schizophrenics

were instructed to name as many items possible under each category for 1 min. Score of 1 was given for every correct response. (iii) Letter fluency task: 2 phonemes, commonly used in Kannada were given and were asked to generate as many words starting with the given phoneme in 1 min. Score of 1 was given for every correct response. c. Word association: A list of 10 words (abstract & concrete) was prepared and rated on familiarity and concreteness by 3 native Kannada speakers. Equal representations of abstract and concrete words (5 each) were taken in the list. Words from the list were presented orally and the subject had to give the most similar or associative word for the given stimuli. The experimenter recorded the responses and scored 1 for each of the most associative response.

Pragmatics: A pragmatic protocol by Prutting & Kirchner (1987) was adopted for profiling of pragmatics. Two conversations were recorded for each participant, first with a familiar and second with a non-familiar partner (15 mins each). The protocol consists of 30 parameters classified into verbal, paralinguistic, and nonverbal aspects. The experimenter rated the conversation samples on these parameters, either as appropriate or inappropriate.

Chronic schizophrenic (Y) Attention and Attention and concentration: [Orientation: concentration: [Orientation: 5/10, Registration: 3/3, 10/10, Registration: 3/3, Attention & Concentration: Attention & Concentration: 4/5] 5/5] Total: 12/18 Total: 18/18 Memory: [Recall: 2/10, Memory: [Recall: 3/10, Anterograde: 3/7, Anterograde:4/7, Retrograde: 1/4, Retrograde: 3/4, Recognition: 3/5] Recognition: 1/5] Total: 9/26 Total: 11/26 Verbal fluency: [Letter: 1/7, Verbal fluency: [Letter: 5/7, Categorical: 3/7] Categorical: 3/7] Total: 8/14 Total: 4/14 Language: Language: [Comprehension: 6/8, [Comprehension: 3/8, Writing: 0.5/1, Repetition: Writing: 1/1, Repetition: 4/4, Naming: 6/12, 4/4, Naming: 7/12, Reading: 1/1] Reading: 1/1] Total: 17/26 Total: 16/26 Visuo-spatial abilities: Visuo-spatial abilities: [Visuo-spatial abilities :4/8, [Visuo-spatial abilities :7/8, Perceptual abilities: 8/8] Perceptual abilities: 8/8] Total: 14/16 Total: 12/16 Overall ACE Score: 59/100 Overall ACE Score: 62/100 Table 2: Subjects performance on Cognition (ACE R, Kannada). Acute Schizophrenic (X)

Semantic measure assessment revealed poor performance by chronic schizophrenic (Y) on word association task. However, chronic schizophrenic (Y) performed better on recall/access task when compared to acute schizophrenic (X) (Refer Table 3).
Acute Schizophrenic (X) Chronic Schizophrenic (Y) Semantic storage: 10/10 Total: 10 Recall/access: - Confrontation naming task (Noun & Verb): 9/10 - Category fluency task: animals: 9, body parts: 15 - Letter fluency task: /k/6, /a/- 9 Total: 48 Word association: 3 Total: 3

Results
Performance on cognitive measure indicated that the acute schizophrenic had better attention and concentration when compared to the chronic schizophrenic, but was more impaired for memory and verbal fluency. While both subjects total score fell below the cut-off score (< 88), indicating cognitive dysfunction (Table 2).

Semantic storage: 10/10 Total: 10 Recall/access: - Confrontation naming task (Noun & Verb): 10/10 - Category fluency task: animals: 9, body parts: 8 - Letter fluency task: /k/- 5, /a/- 3 Total: 35 Word association: 8 Total: 8

Table 3: Performance on semantic measures.

Performance on pragmatic domain was considerably impaired in both the subjects, though on different parameters of verbal, paralinguistic and nonverbal aspects. Table 4 shows a few of the more significant parameters.
52

JAIISH, Vol. 27, 2008

Semantic Pragmatic Attributes and Cognition in Schizophrenics

Verbal Aspects X Y Paralinguistic Aspects X Y Nonverbal Aspects X Y

Variety of speech acts

Topic selection

Topic initiation

Topic maintenance

Pause time

Lexical selection


Intelligibility

+
Fluency

+
Facial expression

+
Prosody


Vocal quality


Eye gaze


Gestures Body posture

(Key: + indicates appropriate, - indicates inappropriate) Table 4: Performance on pragmatic measures.

Discussion
There is increasing evidence that cognitive deficits are not global and generalized, rather are specific and selective. Similarly the test of cognition in the present study revealed a difference in the two subjects in orientation task, memory task, verbal fluency, aspects of language and visuo-spatial abilities. Semantic memory has been conceptualized as an associative network. The pattern of recall depends upon both the strength and the number of associative links with other words in the network. Schizophrenic patients recall fewer words than controls in a retrieval task (Nester, 1998). More interestingly, in this study both subjects showed poor overall performance in word recall, suggesting a specific impairment in either the structure, or modulation of this associative network. On semantic tasks, difference in performance was observed for category and letter fluency. This could be owing to either problem at access/retrieval, and/or using semantic knowledge effectively, both being impaired in individuals with schizophrenia (Marcel, 1983). Further, Kuperberg & Caplan, (2003), reported that poor verbal fluency in patients with chronic schizophrenia may partly be attributable to reduction in semantic store. Word association task showed poor performance by both the subjects, wherein they tended to explain meaning of the given stimuli (word), rather than giving a similar and the most associative word. This finding also supported findings of Gordons (1982) study. Further, Johnson and Shean (1993), in their study found that some patients with negative symptoms were unable to put their idiosyncratic associations into meaningful sentences, and patients with
53

positive symptoms were unable to place common associations in meaningful sentences. For pragmatic task, performance varied for the subjects in relation to onset. The verbal aspects (topic selection, initiation, change, etc.) were affected in subject with acute schizophrenia. The increased pause time within responses can be correlated to recall deficits, supported by the findings of Alpert, Clarck and Pouget, (1994). On the other hand, subject with chronic schizophrenia had impairments more in terms of topic maintenance and specificity of the topic. The subject deviated much from the topic but would eventually connect them all and make it look meaningful. Also the variety of core speech act was limited in both the subjects. The paralinguistic aspects (intelligibility and fluency) were restricted in the acute schizophrenic subject owing to limited speech output and imprecise articulation. However, the subject with chronic schizophrenia exhibited inappropriate prosody (monotonic), intelligibility and vocal quality. Prutting and Kirchner, (1987) concurred that persons with schizophrenia show deficits in decoding basic emotional expressions. On nonverbal aspect of pragmatics, both subjects showed deficits (inappropriate eye gaze, facial expression and body posture). They also exhibited difficulties in performing and understanding appropriate gestures when using language in context.

Conclusions
Schizophrenia is a complex disorder demonstrating abnormalities in both language comprehension and output. The present study reports abnormalities at the level of single words (deficits in the structure and function of lexicosemantic memory) as well as in discourse

JAIISH, Vol. 27, 2008

Semantic Pragmatic Attributes and Cognition in Schizophrenics

(abnormal relationships between sentences) in relation to cognition and onset of symptoms in schizophrenics. However, further validation of the results is required to assert the findings with relation to the onset.

Marcel, A. G. (1983). Conscious and unconscious perception: An approach to the relations between phenomenal experience and perceptual processes. Cognitive Psychology, 15, 238-300. Mathuranath, P. S., Hodges, J. R., Mathew, R., Cherian, J. P., George, A., & Bak, T. H. (2004). Adaptation of the ACE for a Malayalam speaking population in southern India.International Journal of Geriatric Psychiatry, 19, 1188-1194. Mathuranath, P. S., Cherian, J. P., Mathew, R., George, A., Alexander, A., & Sarma, S. P. (2007). Mini Mental State Examination and the Adenbrooke's Cognitive Examination: Effect of Education and norms for a multicultural population. Neurology India, 55(2), 06-110. Nestor, P. G., Akdag, S. J., & O'Donnell, B. F. (1998). Word recall in schizophrenia: A connectionist model. American Journal of Psychiatry, 155, 1685-1690. Prutting, C., & Kirchner, D. (1987). A clinical appraisal of the pragmatic aspects of language. Journal of Speech and Hearing Disorders, 52, 105-119.

References
Alpert, M., Clarck, A., & Pouget, E. R. (1994). The syntactic role of pauses in the speech of schizophrenic patients with alogia. Journal of Abnormal Psychology, 4, 750-757. American Psychiatric Association. (1994). Diagnostic and Statistical Manual of Mental Disorders (4th ed.). Washington, DC. Chen, E. Y. H., Wilkins, A. J., & McKenna, P. J. (1994). Semantic memory is both impaired and anomalous in schizophrenia. Psychological Medicine, 24, 193-202. Crow, T. J. (1980). Molecular pathology of schizophrenia: more than one disease process. British Medical Journal, 280, 1-9. Gordon, R., Silverstein, M. L., & Harrow, M. (1982). Associative thinking in schizophrenia: A contextualist approach. Journal of Clinical Psychology, 38, 684696. Johnson, D. E., & Shean, G. D. (1993). Word associations and schizophrenic symptoms. Journal Psychiatry Research, 27, 69-77. Kuperberg, R. G., & Caplan, D. (2003). Language dysfunction in schizophrenia. In (pp. 444466).

Acknowledgments
We acknowledge our subjects and family members for their cooperation during the study. We also thank the doctors and post graduate students of Department of Psychiatry, Kasturba Hospital for their support and guidance during this study.

54

JAIISH, Vol. 27, 2008

Pragmatic Skills in Infants

Pragmatic Skills in Typically Developing Infants


1

Shilpashri H.N. & 2Shyamala K. Chengappa

Abstract
Communication refers to exchange of information between the speaker and listener. Among the various modes, language forms the primary means of human communication. The knowledge of language is viewed as an integration of content, form and its use. The use of language in social contexts refers to as pragmatics. Pragmatics forms a critical intersection for children's developing language competence and social interactions. As speech and motor milestones develop, pragmatic skills also develop during infancy. The present study was an attempt to understand the type of pragmatic skills acquired at the age of 6 months to 12 months of infancy with Kannada as their mother tongue. Eight typically developing infants (4 male and 4 female) were considered for the present study. One hour audio - video sample of mother-child interaction was recorded and analyzed for eight different pragmatic skills. The results of the present study are discussed in terms of acquisition of pragmatic skills during infancy.
Key words: Infants, Kannada speakers, Pragmatic skills.

The ongoing exchange of message is the act of communication. Communication is a key element in defining humans as social beings. Language is primary means by which human beings maintain interpersonal contact, socialize with others and regulate interactions. Effective communication requires not only linguistic knowledge but social knowledge as well. Efficiency in both linguistic and social abilities is therefore necessary for contextually appropriate, meaningful and effective interpersonal communication (Adams, 2005). Just as learning phonological, semantic (content) and syntactic (structural form) rules of language, a child must also master the rules that underlie how language is used for the purpose of communication (Hymes, 1971). The use of language for social communication is termed as pragmatics. One of the keywords of interest in past decades for speech-language pathologist has been the study of pragmatics. Focus on pragmatics has broadened our view of communication towards the social dimension. The term pragmatics has been introduced into the field of speech language pathology by Elizabeth Bates. Bates (1976) defined pragmatics as rules governing the use of

language in context. Childrens pragmatic language development can be observed at the infancy stage and is seen to rapidly increase and be more sophisticated during the preschool years. From the earliest stage of language development itself, childrens utterances reflect social acts more than linguistic achievements. Woolfolk and Lynch (1982) have reported that infants between the age of 2 and 10 months use eye contact and gaze exchange to regulate joint attention on an activity. Presence of eye contact, smiling and attention indicates that the child takes notice of someone or something. Pointing plus vocalization suggests demand for someone or something. Longitudinal studies that have concentrated on childrens earliest gestural and verbal communicative intents have demonstrated that children begin as early as 9 to 10 months of age to use their gestures and vocalizations for pragmatic functions as requesting, labeling, answering, greeting and protesting (Bates, Camaioni & Volterra, 1975; Dale, 1980). By 12 months typical infants routinely engage in coordinated joint attention with their caregivers (Bakeman & Adamson, 1984; Carpenter, Nagell

1 Junior Research Fellow, Dept. of Speech Language Pathology, All India Institute of Speech and Hearing, Mysore570006, email:shri_shilpa_15@yahoo.co.in, 2Professor, Dept. of Speech-Language Pathology, All India Institute of Speech and Hearing, Mysore-570006, email:shyamalakc@yahoo.com

55

JAIISH, Vol. 27, 2008

Pragmatic Skills in Infants

& Tomasello, 1998). Mathew (2004) has reported in her study that children acquire verbal indication of negation by 1 - 2 years of age. Studies have also focused on how caregivers regulate interaction with the baby by selectively responding to the babys gestures. These early interactions have been referred as protoconversations (Bateson, 1975). Snow (1977) has examined such conversations over several stages between mothers and infants aged from 0 to 18 months. Individuals who fail in using language appropriately to the context are usually put under the diagnosis of pragmatic disorder. In order to identify these deficits, it is important to understand the normative aspects. Hence, the present study aimed to identify type of pragmatic skills acquired by 6 months to 12 months aged typically developing infants.

Materials: The test materials included sound makers, toys and picture books. Procedure: An informed consent was obtained in writing from the mothers of all the children, prior to the study. An interview was conducted with the mothers to rule out history of speech, language and motor developmental delay and hearing problem if any. The procedure undertaken in the present study was audio - video recording of mother-child interaction, at home. Prior to video recording, mothers of all the infants were instructed to interact naturally and to play with the child using toys/materials given. They were also instructed to feel free and to focus on play activity and not to the camera. Semi-instructed method was used where mothers were demonstrated the method of using toys / materials given to elicit the target behavior to be studied. One hour audio-video sample of mother-child interaction was collected in 2 3 sittings for 20 30 minute duration each. Recording for all the infants were done with in a week at their homes. Scoring and analysis: Two point rating scale (0 and 1) was used for scoring of pragmatic skills. 0 indicating absence of the pragmatic skill and 1 indicating presence of the pragmatic skill. Three speech-language pathologists (undergraduates, currently working for their internship program) served as judges for the present study. All the three judges were trained for the use of terminologies and analyzing the pragmatic skills from the video clipping. After the training, judges were instructed to rate the pragmatic behavior using two point rating scale only if the child exhibit the behavior for minimum of three times in the whole recording.

Method
The present study aimed at investigating the type of pragmatic skills acquired by typically developing infants between the age ranges of 6 months to 12 months. The eight pragmatic skills studied were as follows. 1. 2. 3. 4. 5. 6. 7. 8. Smiling. Attention. Eye contact. Vocalization. Play behaviors. Non verbal turn taking. Giving on request. Non verbal indication of negation.

Subjects: 8 infants (4 male and 4 female) in the age range of 6 months to 12 months with Kannada as their mother tongue and their mothers (as mother spends most of the time with the child) were considered for the present study. These infants were screened for speech, language, motor development and hearing, to rule out any associated disorder.
Sl.No. 1 2 3 4 Age in Age in sex Sl. No. months months 06 F 5 07 09 F 6 08 09 F 7 08 11 F 8 09 Table 1: Demographic data. sex M M M M

Results and Discussion


The present study investigated the pragmatic skills in eight typically developing infants between the age ranges of 6 months 12 months. The infants were assessed for 8 pragmatic skills namely; smiling, attention, eye contact, vocalization, play behaviors, non verbal turn taking, giving on request, and non verbal indication of negation. The responses were scored by 3 judges according to the rating scale as mentioned above. All the 3 judges had similar ratings for all the behaviors analyzed. Table-2 and 3 gives the summary of pragmatic skills seen in all the infants.
56

JAIISH, Vol. 27, 2008

Pragmatic Skills in Infants

Pragmatic skills

Smiling Attention Eye contact Vocaliza -tion Play behaviors Non verbal turn taking Giving on request 0 0 0 1 Non verbal 0 0 0 1 indication of negation 0-Absent, 1-Present. Table-2: Pragmatic skills of 4 typically developing infants (females).
Pragmatic skills Males (age in months) M M M 308 M 409 107 208 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0

Females (age in months) F 106 F 209 F 309 F 4 11 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

Pragmatic skills namely, non verbal indication of negation, giving on request was seen only in one female participant aged 11 months. Woolflok & Lynch (1982) and Dheepa (2005) reported that giving on request is acquired by 1 2 years of age. Children acquire negation generally by 1 2 years (Dheepa, 2005) and more specifically by 1 2 years (Mathew, 2004). In the present study, as there was only one child in the higher age (11 months) with in the group who confirmed, it cannot be definitely concluded whether giving on request and non verbal indication of negation are acquired (or not) by one year of age. In the present study an attempt was made to highlight the performance of infants on few pragmatic skills. However, further research on large population and on various pragmatic skills is necessary to arrive at the normative values or for further generalization of the results.

Smiling Attention Eye contact Vocaliza -tion Play behaviors Non verbal turn taking Giving on request Non verbal indication of negation

Conclusions
Development of pragmatic skills starts during infancy. Understanding the normal aspects of pragmatic skills helps in identifying and planning therapeutic intervention for children with pragmatic disorder at an early age. The present study aimed at identifying the type of pragmatic skills acquired by eight 6 12 months aged typically developing infants. Only eight pragmatic skills were included in the present study. The findings of the study revealed no gender differences on any of the eight skills studied. The Pragmatic skills namely, smiling, attention, eye contact, vocalization, play behaviors, non verbal turn taking were found acquired by all the eight infants while giving on request and non verbal indication of negation were seen only in one participant of 11 months age (participant of highest age of the group studied). However, normative data needs to be developed on a larger sample in this regard.

0 Absent, 1 Present Table 3: Pragmatic skills of 4 typically developing infants (males).

As shown in the above table 2 & 3, pragmatic skills namely, smiling, attention, eye contact, vocalization, play behaviors, non verbal turn taking, were seen in all the participants belonging to both the gender groups. Hence, in the present study gender differences were not seen for acquisition of pragmatic skills. The above results obtained are in agreement with study conducted by Woolflok & Lynch (1982) wherein the pragmatic skills namely attention, eye contact, smiling, vocalization stabilizes by 2 10 months of age. Owens (1984) reports nonverbal turn taking begins by first 6 month of life. Dheepa (2005) studied pragmatic skill development in typically developing Tamil speaking children and she has reported that children acquire skills namely, smiling, attention and eye contact by one year of age. As 6 months is the lowest age considered in the present study, few of the pragmatic skills listed may also have been acquired before 6 months of age.
57

References
Adams, C. (2005). Social communication intervention for school-age children: Rationale and description. Seminars in Speech and Language, 26, 181-188. Bakeman, R., & Adamson, L. (1984). Coordinating attention to people and objects in mother-infant and peer-infant

JAIISH, Vol. 27, 2008

Pragmatic Skills in Infants

interaction. 1278-1289.

Child Development, 55,

Bates, E. (1976). In C. A. Prutting (1982). Pragmatics as social competence. Journal of Speech and Hearing Disorders, 47, 123 134. Bates, E., Camaioni, L., & Volterra, V. (1975). The acquisition of performatives prior to speech. Merrill Palmer Quarterly. Bateson, (1975). Pragmatic ability in children. In M.F. McTear & G. Conti, (1992). Pragmatic disability in children.London:Wuhrr Publications. Carpenter, M., Nagell, K., & Tomasello, M. (1998). Social cognition, joint attention, and communicative competence from 9-15 months. Monographs of the Society for Research in Child Development, 63. Dale, P. S. (1980). Is early pragmatic development measurable? Journal of Child Language, 8, 1-12. Dheepa, D. (2005). Developmental protocol for pragmatic skills Unpublished masters dissertation. University of Mysore, Mysore.

Hymes, D. H. (1971). In E. C. Woolfok & J. I. Lynch, (1982). An integrative approach to language disorders in children. New York: Grane and Stratton. Mathew, N. (2004). Pragmatic skills in very young children. Unpublished masters dissertation. University of Mangalore, Mangalore. Owens, R. E. Jr. (1984). Language development. An introduction Columbus, OH: Charles E. Merrill. Snow, R. (1977). Pragmatic ability in children. In M. F. Mctear & G. Conti, (1992). Pragmatic disability in children. London: Wuhrr Publications. Woolfolk, E. C. & Lynch, J. I. (1982). An integrative approach to language disorders in children. New York: Grane and Stratton.

Acknowledgements
The authors would like to thank Dr. Vijayalakshmi Basavaraj, Director, All India Institute of Speech and Hearing, Mysore, for granting permission to carry out the present study. The authors thank all the participants for

58

JAIISH, Vol. 27, 2008

Pragmatic skills in Autism Spectrum Disorders

Pragmatic Skills in Nonverbal Identical Twins with Autism Spectrum Disorders


1

Shilpashri H.N. & 2Shyamala K. Chengappa

Abstract
The present study investigated the use of six pragmatic language skills by a pair of five year old male identical twins with autism spectrum disorders during the course of a session of mother-child interaction. Frequency of each pragmatic language skill used by the mother along with type and frequency of pragmatic language skills used by the two subjects were analyzed. Results showed similarities in use of pragmatic skills in both. However, differences existed with respect to frequency of use.
Key words: Autism Spectrum Disorders, Identical Twins, Pragmatic Skills.

Autism Spectrum Disorders (ASDs) are a set of complex neurobiological disorders, considered to be one of the most profound disorders of childhood. ASDs affect each child differently, to different degrees of severity. However, all children with ASDs share difficulties in 3 areas: qualitative impairments in social interaction, qualitative impairments in communication and restricted, repetitive and stereotyped patterns of behavior, interest and activities (Diagnostic and Statistical Manual of Mental Disorders, 4th edition {DSMIV}, 1994). The onset is always in childhood and the symptoms persist throughout life. Hence, the term pervasive developmental disorders (PDDs). As per DSM-IV-TR published in 2000, the PDDs includes Autism, Asperger disorder, Retts disorder, Childhood disintegrative disorder, Pervasive developmental disorder (not otherwise specified). Autism is more common in males with the average male to female ratio of four to one (Bailey et al. 1995; Fombonne 1999). ASDs roughly occur in 1 of every 150 individuals (Centers for Disease Control and Prevention, 2007). The etiology of autism spectrum of disorders is unknown. The genetic component of autism was confirmed by the first twin study in 1977 showing significantly higher concordance rates for monozygotic twins (MZ) (36-95%) compared to dizygotic twins (DZ) (0-23%) (Folstein & Rutter 1977; Steffenburg et al. 1989; Bailey et al. 1995).

A current estimate for the recurrence risk of autism in the siblings is ~3%, and the heritability estimate is over 90% (Folstein and Rosen-Sheidley 2001). In British twin Study (Bailey et al., 1995) an examination of 16 MZ pairs concordant for autism or autism spectrum disorders showed clinical heterogeneity even when pairs shared exactly the same segregating genetic alleles. Autism is one of the language disorder primarily characterized by inability to relate to other people and communicate effectively (Bernard-Opitz,1982). Regardless of age, level of intellectual functioning, and developmental level, all individuals with autism demonstrate deficits in social-communicative domain (Wing, 1997; Tager-Flusberg, Joseph, & Folstein, 2001) i.e. pragmatic skills. Pragmatics is the linguistic domain concerned with the appropriate use of language across a variety of social contexts that provides for a listener's accurate interpretation of the speaker's intentions and references (BerkoGleason, 2005). Pragmatic aspect of language acquisition accounts for childrens growing communicative competence, rather than focusing on the structural forms (syntax) or content (semantics) of their language. Review of literature identifies a number of studies on pragmatic deficits in children with PDDs / ASDs. Aarons and Gittens (1987); Wing (1988) have even suggested that pragmatic disability is just another term for autism. (Ball,

1 Junior Research Fellow, Dept. of Speech-Language Pathology, All India Institute of Speech and Hearing, 2 Professor, Dept. of Speech-Language Manasagangothri, Mysore570006, email:shri_shilpa_15@yahoo.co.in, Pathology, All India Institute of Speech and Hearing, Manasagangothri, Mysore570006, email:shyamalakc@yahoo.com.

59

JAIISH, Vol. 27, 2008

Pragmatic skills in Autism Spectrum Disorders

1978; Cantwell, Baker & Rutter, 1978; Paul & Cohen 1985; Loveland & Landry, 1986) have reported that the autistic childrens language and gestures are pragmatically deficient, even when level of language acquisition or IQ is taken into account. These children show pragmatic deficits both in how they communicate (communication means) and how they express intentions (communicative intents) (Rollins, 1999). The present study is an attempt to investigate the performance of non verbal identical twins with autism spectrum disorders on six pragmatic skills namely, giving on request, pointing / visual gestures for requesting, joint attention, gaze exchange, non verbal turn taking and non verbal indication of negation.

Subject details (A1 & A2) As per the information obtained from the mother during clinical interview, the onset of the symptoms was before 12 months for the two subjects. Comprehension skills were reported to be poor with no speech, only vocalization (clinical condition for speech was the same at the time of recording). Motor development was reported to be normal with poor socialization skills, poor imaginative play and stereotyped repetitive behaviors. (Subjects demographic data are given below).
Client Report Age of onset Medical history Motor development Hearing Vision Speech language skills SUBJECTS AI A2 Before one Before one year year - ve - ve Normal Normal Normal No speech (only vocalization) Poor Normal Normal Normal No speech (only vocalization) Poor

Aims and Objectives


The objectives of the study were manifold: 1. To study the type of pragmatic skills used by the two twin subjects during the course of interaction with the mother. 2. To study the frequency / percentage of each pragmatic skill used by mother and the two twin subjects during initiation of communication. 3. To study the frequency / percentage of each pragmatic skill used by mother and the two twin subjects during response course. 4. To compare difference in the performance between the two twin subjects.

Social skills Imaginative Absent Absent play Stereotyped repetitive Present Present behaviors Table-1: Demographic data

Method
Subjects: 5 years old male identical twins (A1 and A2) primarily diagnosed as delayed speech and language with autism spectrum disorders by qualified speech language pathologist served as subjects for the present study. The subjects fulfilled the following criteria 1. They had Kannada as their mother tongue. 2. Subjects had no medical history. 3. Subjects had normal hearing sensitivity and vision. 4. Subjects had no history of regression in motor development.

Interventions for both the subjects were started at 4 years of age. Both the subjects under went speech-language therapy and occupational therapy for two days per week at the duration of 45 minute each. Speech-language therapy was mainly focused on improving prelinguistic skills and communication skills using Picture Exchange Communication System (PECS) focused mainly on functional skills. Materials used: The materials used to elicit responses consisted of several toys, puzzles and building blocks. Procedure: An informed consent was obtained in writing from the mother, prior to the study. The procedure undertaken in the present study consisted of audio - video recording of motherchild interaction using semi-instructed method. Sony (DCR-DVD703E) digital video camera recorder was used for video recording. Prior to video recording, mother was instructed to feel free and to focus on play activity and not to the camera.
60

JAIISH, Vol. 27, 2008

Pragmatic skills in Autism Spectrum Disorders

The mother was demonstrated the method of using toys / materials given to elicit the target behaviors to be studied. Mother was also instructed to interact naturally and to play with the child using toys/materials given by introducing series of questions to elicit different pragmatic functions. Six pragmatic functions were tapped with these questions. These functions are 1. Giving on request (the act of giving objects, toys, eatables, etc to the partner on request). 2. Pointing / Visual gestures for requesting (the act of addressing desire for an object, action, etc). 3. Gaze exchange (the act of maintaining eye contact in long alternating intervals). 4. Joint attention (the act used to direct others attention to an object, event or topic of communicative act). 5. Non verbal turn taking. (The act of interactional behavior, where the partner should wait for his / her turn during play activity). 6. Non verbal indication of negation (the act of confirming the absence of an object, person, etc, nonverbally) One hour audio-video sample of mother-child interaction was collected. Each child was recorded separately in 3 sittings for 20 minute duration each. Recording was done at the home and at therapy room. All the sittings were recorded with in a weeks time. Analysis: The six pragmatic skills studied were analyzed in terms of type of pragmatic skills used and percentage of occurrence / frequency and functional appropriateness of use by mother and the two subjects. The audio-video recorded sample of mother-child interaction was analyzed by three judges, including he 1st author all the three were Master degree holders in speech-language pathology. Before the analysis of the data, latter two judges underwent training for the duration 3 hours. During the training period an audio-video sample of typically developing child interacting with the mother was introduced and the two judges were trained for familiarization of the terminologies used in the present study and identifying the pragmatic skills which were functionally appropriate to the context. Once both the judges were confident in identifying the pragmatic skills, audio-video sample of 40 minute duration of mother-child interaction of each individual subject were shown to the judges separately. Judges were instructed to identify the type and frequency of answering appropriate to the context by the mother and the two subjects separately.
61

Statistical analysis: Reliability analysis was carried out to find inter and intra judge reliability. Frequency of each pragmatic skill used by mother and the two subjects were analyzed in terms of percentage of occurrence using population pyramid graph,

Results and Discussion


Inter and intra judge reliability for frequency of occurrence of pragmatic skills was found to be 0.9. Figure 1 & 2 shows the compiled results of all the 3 judges. Combined communicative strategies (verbal and nonverbal) used by the mother were considered for analysis of percentage of occurrence of each pragmatic language skill.

RQ: Requesting; GE: Gaze exchange; JA: joint attention; NVTT Nonverbal turn taking; NVIN: Nonverbal indication of negation. Figure-1: Frequency of use of pragmatic skills (expressed as percentage) by the mother during the course of initiation of communication with the two subjects A1 and A2.

Percentage of initiation of each pragmatic skill by the Mother w.r.t the subject 1 (A1) As shown in fig 1, 67.66% of the time, mother introduced questions for requesting objects, action, initiation of new task etc. Gaze

JAIISH, Vol. 27, 2008

Pragmatic skills in Autism Spectrum Disorders

exchange was introduced and maintained for 10.53% of time. Joint attention was introduced and maintained for 10.53% of times during play activity (examples of play activity introduced, cricket, playing with soap bubbles and action for rhymes) and conversational task. 7.52% of the time nonverbal turn taking task was introduced during the play activity and 3.76% of the time mother requested the child to identify the item/object which was not present in the surrounding at the time of recording. Percentage of initiation of each pragmatic skill by the Mother w.r.t the subject 2 (A2) Requesting task dominated with 53.57% of occurrence fallowed by gaze exchange (14.29%). Introducing and maintaining Joint attention was found to be 14.29% during play activity (examples of play activity introduced, playing with soap bubbles and action for rhymes) and conversational task. . While 13.09% of the times nonverbal turn taking task was introduced during the play activity. The least of all was introducing question for indication of negation (4.76%). It is clear from figure-1 that, mother maintained uniformity when introducing different questions to elicit responses for the five different pragmatic skills. Requesting task was introduced more frequently and least was the questions on negation. But, examining for the percentage of occurrence of each pragmatic skill, w.r.t the A1 & A2, there were slight variations among them. The reason for introducing requesting questions more frequently is possibly because of the influence of training method used at therapy session and at home using Picture Exchange Communication method, in terms to encourage the subjects to perform giving task on request. The reduced frequency of occurrence of other pragmatic language skills (joint attention, nonverbal indication of negation) could be due to poor responses on the part of the two subjects (see fig: 2) Nonverbal responses obtained from subject-1 (A1) As shown in fig-2, out of 67.66% of mothers requesting, contextually appropriate response was given for 20.00% of the time. Gaze exchange was maintained well with 21.43%. Maintenance of Joint attention for activities was less i.e. 7.14% for 1O.53% from the mother side. Involvement for non verbal turn taking was equally good with
62

30.00%. Non verbal response for negation was nil i.e 0% (No response). Non verbal responses obtained from subject-2 (A2) Out of 53.57% of requesting A2 showed good response (42.22%). Gaze exchange was maintained well with the percentage at 33.33. 0% (No response) was obtained for maintaining Joint attention. Involvement for non verbal turn taking (play activity) was more (45.45%). Non verbal response for negation was nil i.e. 0% (No response).

RQ: Giving on Request; GE: Gaze exchange; JA: joint attention; NVTT Nonverbal turn taking; NVIN: Nonverbal indication of negation. Figure-2: Frequency of use of contextually appropriate pragmatic skills by the two subjects (A1 & A2), (expressed as percentage).

From the results obtained (refer fig 2) it is clear that, there were individual differences seen in percentage of use of pragmatic skills even though there are certain similarities in terms of type of pragmatic skill used by the two subjects. This result is in agreement with British twin study (Bailey et al., 1995) that clinical heterogeneity is commonly seen in monozygotic twin pairs with ASDs However, further research is warranted on the same line for generalization of the results. It was also seen that, the percentage of occurrence of giving on request, gaze exchange and nonverbal turn taking (taking part in play

JAIISH, Vol. 27, 2008

Pragmatic skills in Autism Spectrum Disorders

activity namely, cricket, action for rhymes and playing with soap bubbles) was highest compared to other two pragmatic functions namely, joint attention and non verbal indication of negation. This pattern was similar among the two subjects. This pattern may be mainly because of the affect of speech-language therapy and occupational therapy attended. Frequency of initiation of pragmatic skills by the two subjects (A1 & A2) during the course of interaction with the mother Initiation of pragmatic language skills by the two subjects was restricted for requesting. Requesting was mainly for eatables and toys of their interests. Picture cards were used to indicate their requirements. Other pragmatic skills namely, initiation of gaze exchange, initiation of joint attention, initiations of non verbal turn taking, nonverbal questioning for identification of non existence of items/objects, was not introduced. Response from the mother (i.e giving the requested object) was 100% for the two subjects. The result indicates poor performance by the two subjects for initiation of pragmatic questions/skills during the course of mother-child interaction. The obtained results are in agreement with the studies reporting that pragmatic skills are generally affected in children with autism spectrum disorders (Aarons and Gittens, 1987; Wing, 1988; Rollins, 1999).

References
Aarons, M., & Gittens, T. (1987). Is this Autism? Windsor, Berkshire: NFER-Nelson. American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: American Psychiatric Association. American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed. text revision). Washington, DC: Author. Bailey, A., Le Couteur, A., Gottesman, I., Bolton, P., Simonoff, E., Yuzda, E., & Rutter, M. (1995). Autism as a strongly genetic disorder: evidence from a British twin study. Psychological Medicine, 25, 63-77. Ball, J. (1978). A pragmatic analysis of autistic childrens language with respect to aphasic and normal language development. Unpublished Doctoral Dissertation, Melbourne University. Melbourne. Australia. Berko-Gleason, J. (2005). The development of language (6th Ed.). Boston: Pearson/Allyn & Bacon. Bernard-Opitz, V. (1982). Pragmatic analysis of the communicative behavior of an autistic child. Journal of Speech and Hearing Disorders, 47, 99-109. Cantwell, D., Baker, L., & Rutter, M. (1978). A comparative study of infantile autism and specific developmental receptive language disorder IV. Analysis of syntax and language function. Journal of Child Psychology and Psychiatry, 19, 351 362. Centers for Disease Control and Prevention. (2007). Prevalence of autism spectrum disordersautism and developmental disabilities monitoring network, six sites, United States, 2000. Morbidity and Mortality Weekly Report, 56(SS-1), 111. Folstein, S., & Rutter, M. (1977). Infantile autism: a genetic study of 21 twin pairs. Journal of Child Psychology and Psychiatry, 18, 297321. Folstein, S.E., & Rosen-Sheidley, B. (2001). Genetics of autism: complex etiology for a
63

Conclusions
The results of this study has shown that, during the course of mother-child interaction, initiation of pragmatic questions were mainly by the mother as compared to the two subjects who were limited only for requesting. On the other hand the two subjects responded well for the pragmatic questions introduced by the mother. Even though the two subjects showed similar performance in use of contextually appropriate three of five pragmatic functions namely, giving on request, maintain gaze exchange and non verbal turn taking task during play behavior, they differed in percentage of use of each function. This is an indicative of both pragmatic skill deficits in children with autism spectrum disorders and heterogeneity of behaviors among the group. Hence, early identification of the condition and individualized therapy program assume great importance in such clinical population.

JAIISH, Vol. 27, 2008

Pragmatic skills in Autism Spectrum Disorders

heterogeneous disorder. Nat Rev Genetics, 2, 943-955. Fombonne, E. (1999). The epidemiology of autism: A review. Psychological Medicine, 29(4), 769-86. Loveland, K.A., & Landry, S. (1986). Joint attention and language in autism and developmental language delay. Journal of Autism and Developmental Disorders, 16, 335 349. Paul, R., & Cohen, D.J. (1985). Comprehension of indirect requests in adults with autism disorders and mental retardation. Journal of Speech and Hearing Research, 28, 475 479. Rollins, P.R. (1999). Early pragmatic accomplishment and vocabulary development in Preschool children with autism. American Journal of SpeechLanguage Pathology, 8, 181 190. Steffenburg, S., Gillberg, C., Hellgren, L., Andersson, L., Gillberg, I.C., Jakobsson, G., & Bohman, M. (1989). A twin study of autism in Denmark, Finland, Iceland,

Norway and Sweden. Journal of Child Psychology and Psychiatry, 30, 405-416. Tager-Flusberg, H., Joseph, R.M., & Folstein, S. (2001). Current directions in research on autism. Mental Retardation and Developmental Disabilities Research Reviews, 7, 21-29. Wing, L. (1988). The continuum of autistic characteristics. In F. Shopler & G.B. Mesibov (Eds.), Diagnosis and Assessment in Autism. New York: Plenum Press. Wing, L. (1997). Syndromes of autism and atypical development. In D. J. Cohen & F. R. Volkmar (Eds.), Handbook of autism and pervasive developmental disorders. New York: John Wiley & Sons, Inc.

Acknowledgements
The authors would like to thank Dr. Vijayalakshmi Basavaraj, Director, All India Institute of Speech and Hearing, Mysore, for granting permission to carry out the present study. The authors thank all the participants for their constant cooperation throughout the study.

64

JAIISH, Vol. 27, 2008

Continuum of Developmental Language Disorders

Continuum of Developmental Language Disorders: Where Does PLI Fall?


1

Shivani Tiwari, 2Ashwini Bhat & 3Rajashekhar Bellur

Abstract
Pragmatic Language Impairment (PLI) is a developmental communication disorder, characterized by semantic and pragmatic deficits; relatively adequate phonology and syntax; and mild autistic features. The symptom profile overlaps with Pervasive Developmental Disorder (PDD) and Specific Language Impairment (SLI). The present paper attempts to highlight the differential diagnosis of PLI with common developmental language disorders as SLI and PDD. Two children with speech and language characteristics suggestive of PLI are discussed. Two children (5 yrs each) participated in the study with the complaint of inadequate speech and language skills. Speech and language skills of these two children were evaluated using formal and informal methods. The responses were recorded and were further transcribed for analysis. The obtained profiles of the two children were compared across the profiles of developmental disorders as PLI, SLI and PDD. Case1 presented an early onset, delayed developmental milestones with poor pre-linguistic skills and significant semantic and pragmatic deficits. Case 2, in contrast, showed delayed speech milestones, relatively better pre-linguistic skills and near normal semantic and syntactic skills. Both the cases presented with clear, fluent speech without articulation errors. Pragmatic deficits were prominent in both the cases, with case1 having more severe difficulties. Both case1 and case2 had mildly impaired social skills.Differential diagnosis of PLI with other developmental language disorders is controversial issue. PLI shares features of linguistic processing deficits with SLI on one hand, and deficits in pragmatics, social skills and stereotyped repetitive behaviors with PDD on the other hand. There are no well-defined boundaries amongst these disorders. Additionally, course of PLI is such that it changes the diagnostic category as the clinical picture varies with time. The two cases discussed in the present paper had pragmatic deficits in common, though of varied severity and associated features.
Key words: Specific language impairment, Pervasive developmental disorders, Semantics, Pragmatics, Syntax.

Rapin in 1982, defined Pragmatic Language Impairment (PLI) as a developmental communication disorder, characterized by fluent, well-formed sentences, clear, loose, tangential or inappropriate speech, with difficulty in understanding discourse, having illogical train of thought and relatively better social skills. Children with PLI speak fluently and clearly in long utterances. They may have a severe impairment in pragmatics and semantics, preservation, and significant word finding
1

difficulty. They usually have excessive variation in pitch and loudness. Children with PLI may have milder deficits in phonological and syntactic skills in the early childhood. They may have semantic deficits like delayed semantic development, usage of words only in the limited contexts, difficulty in the comprehension of meaningful verbal messages, questions, idioms, slang expressions, abstract words and the words that relate to feelings and emotions, and tendency to interpret messages quite literally. These children give inappropriate answers to questions, may show semantic

Lecturer, Dept. of Speech and Hearing, Manipal College of Allied Health Sciences, Manipal University, Manipal-576104, email: shivanitiwari8@yahoo.co.in, 2Student, Dept. of Speech and Hearing, Manipal College of Allied Health Sciences, Manipal University, Manipal-576104, 3Professor, Dept. of Speech and Hearing, Manipal College of Allied Health Sciences, Manipal University, Manipal-576104.

65

JAIISH, Vol. 27, 2008

Continuum of Developmental Language Disorders

paraphasias, and use of circumlocutions (Rapin & Allen, 1983; Adams & Bishop, 1989; Sahlen & Nettlebladt, 1993; Bishop, 2000b). Pragmatic deficits seen in these children include delayed pragmatic development, and impaired communication functions. These children have poor turn taking skills, difficulty in establishing discourse topics, inability to repair communication breakdown, and tangential speech. Their speech is characterized by excessive, irrelevant, preservative utterances, use of self-directed speech in the middle of conversation, and show tendency to answer their own questions (Adams & Bishop, 1989; Leinonen & Letts, 1997; Rapin & Allen, 1983). Several etiologies of PLI have been proposed needing experimental verification. Some of these are environmental, genetic (ContiRamsden, Crutchley & Botting, 1997) and neurological factors (Sahlen & Nettlebladt, 1993). Rapin and Allen first described the condition in 1983, and proposed the term as semantic pragmatic deficit syndrome. As these children have problem in the specific areas of language, Bishop and Rosenbloom in 1987 changed the term into semantic pragmatic language disorder. Bishop (2000a) proposed the label Pragmatic Language Impairment; as children diagnosed SPLD do not necessarily have semantic problems. Further, Bishop (2000b) termed PLI-plus for children whose pragmatic problems are disproportionate to their other language limitations, and are not obviously the result of these limitations. In addition, the term PLI-pure, was termed for children who have only pragmatic deficits and normal language skills. Children with PLI initially present with a picture of language delay and receptive language impairment, who then learn to speak fluently, clearly and in complex sentences, with semantic and pragmatic abnormalities becoming increasingly evident as their verbal proficiency increase. Whereas, at first they may be difficult to differentiate from other developmental language disorders, the pattern of verbal deficits looks more distinctive as they grow older (Adams & Bishop, 1989). As PLI exists in close boundaries of developmental language disorders like SLI and PDD, the differential diagnosis becomes an essential part of assessment. PLI is differentiated with SLI on presence of pragmatic deficits, social skills deficits, and stereotyped
66

utterances/behaviors. Further, children with SLI essentially have impairments in phonological and syntactic skills (generally not reported in children with PLI). Within PDD spectrum, Aspergers can be distinguished from PLI having late onset, relatively normal language skills in presence of poor social skills. However, no distinct symptoms anchor difference between PLI and autism. Autism and PLI are sorted only based on the severity of impairment (PLI having milder symptoms). Deficits in social interaction, stereotyped range of interests, theory of mind, echolalia, eye contact, pretend play, semantic and pragmatic deficits are of lesser degree in PLI compared to PDD (Boucher, 1998; Rapin, & Allen, 1983; Bishop, 1989). Perseveration is another feature of PLI, which is otherwise not seen in developmental language disorders. Childrens Communication Checklist (CCC; Bishop 1998; see Appendix) is of the tests of its type developed to distinguish language-impaired children having pragmatic difficulties and those of typical forms of SLI. This test provides the cutoff scores on pragmatic component (>132 on pragmatic composite indicates SLI), differentiating children with PLI and SLI. Scores between 145 and 156 (on pragmatic composite of CCC) indicate normal range. Bishop and Norbury (2002) conducted a study on 21 children (6 to 9 yrs) with language impairment, who were further categorized based on CCC into 13 children with PLI, and 8 children with typical SLI. Children with PLI scored less than 133 on the pragmatic composite of CCC whereas, children with SLI scored above 132. Thus, CCC can serve as a useful diagnostic instrument in diagnosis of PLI. Although, this instrument lacks validity, yet it can be used as a screening tool. Precise diagnostic criterion for PLI is not available. Moreover based on the literature findings, there exists variability in the range of features of PLI. The present study attempt to highlight the differential diagnosis of PLI with developmental language disorders as SLI and PDD. Two children with communication features suggestive of PLI with different symptoms and varied severity are discussed.

Method
Two children (5 years old each), with pragmatic deficits without the diagnosis of autism were taken up for the study. Psychological

JAIISH, Vol. 27, 2008

Continuum of Developmental Language Disorders

evaluation was done to identify the deficits in nonverbal intelligence and social skills. Benite Kamat Test (BKT) (measure of nonverbal intelligence) was administered to find out the intelligence quotient (IQ). The social skills were measured using Vineland Social Maturity Scale (VSMS) to obtain social quotient (SQ). Language tests like REELS (Receptive Expressive Emergent Language Scales), Extended REELS & Pragmatic

Checklists (Shipley & McAlfee, 1998) was administered. In addition, informal language assessment was carried out using black and white line drawings (With a little bit of help, language training manual) incorporating tasks like picture naming, picture description, narration, reciting numbers, alphabets and rhymes. The response obtained was transcribed using IPA, and analyzed for different language parameters.

Results
The analyzed language of the subjects are summarized as follows
Name Age / Gender Language Chief complaints CASE 1 Master N 5 years / Male Kannada attention to speech Repetition of few sentences Delayed responses to questions Symptoms were first noticed at the age of 1.6years as language delay & poor eyeto-eye contact. No significant birth & medical history Family history: -ve Slight delay in motor development Delayed language development: -babbling at 8 months -first words at 1.6 years -phrases after 2 years Social development: delayed recognition of father (at 3 years) CASE 2 Baby A 5 years / Female English Language delay level of activity Symptoms were noticed by the teacher at the age of 3 years (behavioral deviations as poor social skills, echolalia) No significant birth & medical history Family history: -ve Normal motor milestones Delay in language acquisition: -two word utterances at 2.6 yrs -3 word utterances at 3.6yrs Social development: unable to differentiate between family members & strangers (extra friendly with strangers) Normal hearing IQ: average intellectual functioning (BKT) VSMS: average social & adaptive functioning CARS: non autistic Reported to have autistic like features earlier attention span Good eye contact with family members Highly distractible Respiration: normal Phonation: pitch & quality age adequate Articulation: cluster reduction & metathesis Prosody: inappropriate intonation pattern Fluency: fluent speech.

Onset

History

Audiological evaluation Psychological evaluation

Normal hearing IQ deficits could not be ruled out (BKT) VSMS: Borderline deficits in social & adaptive functioning CARS: mild autistic features attention to speech Poor eye contact Respiration: normal Phonation: pitch & quality age adequate Articulation: cluster reduction & distortion of /s/ Prosody: inappropriate intonation pattern & unable to imitate intonation for rhymes Fluency: fluent speech.

Prelinguistic skills Speech skills

Table 1: Case history, pre linguistic and speech skills of the two cases.

The details of the children and their speech and language behaviors are given in Tables 1 & 2. Both children were initially brought with the complaint of language delay. On a detailed psychological and speech-language assessment, case1 was found to have more problem than case2.
67

IQ assessment could not be completed in case1 due to deficits in comprehending the instructions. On the administration of VSMS, he was found to have borderline deficits in social and adaptive behavior. He was diagnosed as having mild autistic like features. Case2 presented with

JAIISH, Vol. 27, 2008

Continuum of Developmental Language Disorders

average intellectual functioning and average social and adaptive functioning. She was diagnosed as non-autistic. Both cases had fluent and intelligible

speech. The prelinguistic skills of case1 were poor. Case 2, in contrast had good prelinguistic skills.

Language Test results

Morphology Semantics

CASE1 RLA= 30-33 months ELA= 24-27 months Scattered findings No comprehension & expression of any morphological markers Vocabulary: restricted & limited use Can name common lexical items in a few categories Cannot comprehend emotions & facial expression of others Irrelevant utterances & perseverations noticed Presence of delayed & immediate echolalia Comprehends 1-step commands on several repetitions Time & place concept absent Cannot comprehend stories Picture description absent Could recite only 1-2 lines of a rhyme with lots of prompts in flat intonation Uses 2-3 word utterances Adjectives: comprehension is present st nd Pronouns: uses 1 & 2 person pronoun; does not rd comprehend 3 person pronoun Does not comprehend gender markers. Uses few prepositions Does not comprehend tenses & plurals st nd Uses 1 & 2 person possessive markers Difficulty with polar questions Inappropriate answers to questions

CASE2 RLA= 4.65 years ELA= 4.6-5 years Age adequate usage Vocabulary: adequate to age Can name common lexical categories Can comprehend emotions & facial expression of others Irrelevant utterances & self talk noticed. Perseveration was absent. Echolalia: absent Comprehends simple & complex commands. Gross time & place concept present Comprehends & expresses story episodes in a sequence with minimal prompts Express fairly good on picture description tasks Recites rhymes without prompts with appropriate intonation

Syntax

Uses 4-6 word utterances Uses adjectives Pronoun usage is present Gender confusion was observed Preposition confusion is present Uses regular tense & plural marker correctly, has confusion with the irregular tense & plural markers. Uses all the possessive markers appropriately Does not have difficulty with polar questions Answers appropriately to questions with occasional confusion

68

JAIISH, Vol. 27, 2008

Continuum of Developmental Language Disorders

CASE1 Pragmatics Communicative functions: -attention seeking absent -rarely requests for objects -occasionally protests & denies undesired items -social greetings only with prompts -giving & seeking information absent -comprehension of feelings & facial expression absent Discourse: -difficulty in topic initiation, maintenance, topic transition & turn taking -giving information absent Imaginative play absent Role-taking not attained Modification in the manner if communication is absent

CASE2 Communicative functions: -attention seeking present -requesting is present No request when wants to go to toilet, indicates by pointing -occasionally protests & denies undesired items -social greetings without prompts -giving & seeking information present -comprehends feelings & facial expression Discourse: -difficulty in topic, maintenance, topic transition & turn taking ; topic initiation present -gives excessive information, inappropriate to the context - word finding deficits Imaginative play absent Role-taking not attained Modification in the manner if communication is absent Attending school since 1 year Able to read & write alphabets, numbers (1-10), small words, her name, & can solve 1-digit addition & subtraction Hyperactive & easily distractible Prefers solo play

Secondary skills

language

Attending school since 2 years Cannot read & write

Behavioral deviations

Stimulation Previous treatment

Hyperactive & irritable in nature Prefers to be alone; plays only games like running & chasing with peers Vacant stares; biting of shirt collar & putting fingers into mouth repeatedly was noticed. Poor stimulation both at home & school No previous treatment.

Good stimulation for language, reading, & writing She was attending special school along with IEP in USA as she was diagnosed to have autistic features. With treatment, there was reduction in echolalia and improvement in all aspects of language including pragmatics. Demonstration therapy was given for a month. Attention enhancement training, activities to improve semantic, syntactic skills, pragmatic, & cognitive skills.

Present treatment

Demonstration therapy was carried out for 2 days. Activities for improving attention, vocabulary, comprehension of questions, reading, & pragmatic skills were carried out.

Table 2: Description about the language skills of the two cases.

Case1 had deficits in syntax, semantics and pragmatics whereas case2 had age adequate speech and language skills. Case1 had pragmatic deficits along with the deficits in other areas of language. The second case had only pragmatic deficits.
69

Pragmatic deficits were again more severe in case1. Case2 was previously diagnosed as having autistic like features, at the age of 3 years. She had received Individualized Education Program (IEP) in California for 1-year duration. Parents reported

JAIISH, Vol. 27, 2008

Continuum of Developmental Language Disorders

significant improvement in the performance with respect to reduction in the echolalia, improvement in pragmatic and social skills along with the improvements in other areas of the language. Communicative functions as request, denial,

seeking and giving information reported to emerge over the course of treatment. Thus, treatment given for case 2 further accounts for the differences in performances between the two cases.

Table 3: Comparison of the two cases with developmental language disorders. (SLI and PDD)

Table 3 provides comparison of the two cases with PDD and SLI. The check mark () indicates presence of a behavior and the cross symbol (x) indicates the absence of that behavior. The table shows absence of communicative functions in children with PDD indicating severely impaired social skills. In case of SLI, the social skills are relatively normal and may have significant deficits in the syntactic and phonological aspects. The difference in the linguistic symptoms between these cases is evident in the table. Case1 shows milder impairment in semantic and syntactic aspects of the language (not seen in case 2). However, the communicative functions of the case2 are better than case1. Both the cases exhibited echolalia, whereas perseveration was observed only in case1. Looking into these characteristics, case 1 was diagnosed as PLI-plus and case2 as PLI-pure.

intelligence could not be ruled out in case 1, owing to deficits in following the instructions. Case 2 had average intellectual functioning. Although autistic like features such as social interaction problems and echolalia were present in these cases, one must notice that the communication functions were relatively better and behavioral problems were not very severe. Hence, diagnosis of Autism was ruled out. Further, the presence of language delay in case 1 and history of language delay in case2, ruled out the possibility of the diagnosis of Aspergers syndrome. Age of onset (18-24 months for case1) with relatively better social skills in two children, rejects the diagnosis of Pervasive Developmental Disorder-Not Otherwise Specified (PDD-NOS). Despite the fact that both the cases had milder deficits in semantic and syntactic skills, a diagnosis of SLI was ruled out, in presence of pragmatic and mild social skills deficits. The social skills deficits in the two children were not as prominent as seen in PDD and more severe in comparison to children with SLI. Hence, these children were eventually placed under the category of PLI, based on presenting symptoms in accordance with literature (Rapin & Allen, 1983; Adams & Bishop, 1989; Sahlen & Nettlebladt, 1993; Bishop, 2000b). Further look into the nature
70

Discussion
The two children presented with language deficits predominantly in pragmatics, though scattered. Authors attempted for a differential diagnosis of the cases with similar profiles of PDD and SLI. Due to the stringent criteria each label subscribes to, there was a confusing picture. On psychological evaluation, the deficits in nonverbal

JAIISH, Vol. 27, 2008

Continuum of Developmental Language Disorders

of semantic syntactic and pragmatic deficits, fluent speech with perseveration and echolalia in case1, a diagnosis of PLI-plus was provided. Case2, alternatively presented with relatively normal semantic, morphological, and syntactic skills (age adequate receptive & expressive language skills) along with echolalia and pragmatic deficits, and hence was given a diagnosis of PLI-pure. There exists unresolved controversy concerning the diagnostic criteria for PLI, the controversy focusing on the differential diagnosis of PLI, Autism and SLI. Some authors argued quite strongly that PLI and autism are not distinct condition; rather, PLI is an identifiable form of Autism, perhaps a subtype, or mild or atypical manifestation of prototypical autism (Brooks & Bowler, 1992; Shields, Varley, Broks & Simpson, 1996). Bowler and Lister-Brook (1998) rejected the use of the term mild autism, considering it misleading in view of the persistent though subtle nature of PLI childrens social impairment. Boucher (1998) argued that PLI might constitute a developmental language disorder in its own right, independent of either Autism or SLI. PLI constitutes a distinct subtype of communication disorder. There are qualitative differences between the pragmatic impairments associated with PLI and those associated with Autism (Rapin & Allen, 1983). Neither of the two internationally recognized set of diagnostic criteria for mental and behavioral disorders, DSM-4 and ICD-10 recognize the existence of PLI. One well recognized difficulty is that criteria for identifying children with PLI is controversial, so the criteria used to select participants vary from study to study with, not surprisingly, confusing results.

social and language disproportionate severity.

impairment

of

References
Adams, C., & Bishop, D. V. M. (1989). Conversational characteristics of children with semantic-pragmatic disorders I: Exchange structure, turn taking, repairs and cohesion. British Journal of Disorders of Communication, 24, 211-239. Bishop, D. V. M. (1989). 'Autism, Asperger Syndrome and Semantic-Pragmatic Disorder: Where Are the Boundaries?' British Journal of Disorders of Communication, 24, 107-121. Bishop, D. V. M. (1998). Development of the Children's Communication Checklist (CCC): A method of assessing qualitative aspects of communication impairment in children. Journal of Child Psychology and Psychiatry, 39, 879-891. Bishop, D. V. M. (2000a). What's so special about Asperger's syndrome? The need for further exploration of the borderlands of autism. In A. Klin, F. R. Volkmar & S. S. Sparrow (Eds.), Asperger Syndrome (pp. 254-277). New York: Guildford. Bishop, D. V. M. (2000b). Pragmatic Language Impairment: A correlate of SLI, a distinct subgroup, or part of the autistic continuum? In Bishop, D.V.M. and Leonard, L., editors, Speech and language impairments in children: causes, characteristics, intervention and outcome. Hove: Psychology Press. Bishop, D. V. M., & Baird, G. (2001). Parent and teacher report of pragmatic aspects of communication: Use of the Children's Communication Checklist in a clinical setting. Developmental Medicine and Child Neurology, 43, 809-818. Bishop, D. V. M., & Norbury, C. F. (2002). Exploring the borderlands of autistic disorder and specific language impairment: A study using standardized diagnostic instruments. Journal of Child Psychology and Psychiatry, 43, 917-929. Boucher, J. (1998). SPD as a distinct diagnostic entity: logical considerations and directions
71

Conclusions
Developmental language disorders comprise a spectrum of disorders with varied severity and symptomatology. PLI is one such disorder sharing features of linguistic processing deficits with SLI on one hand, and deficits in pragmatics, social skills and stereotyped repetitive behaviors with PDD on the other hand. Hence, supporting the views of Bishop, (1989) it is not helpful to adopt a rigid response to diagnostic labels, rather a flexible approach is especially appropriate as we come to recognize the broader spectrum of language disorders and increasingly encounter children with

JAIISH, Vol. 27, 2008

Continuum of Developmental Language Disorders

for future research. International Journal of Language and Communication Disorders, 33, 71-81. Bowler, D. M., & Lister Brook, S. (1998). SPD and autistic spectrum disorder. International Journal of Language and Communication Disorders, 33, 91-94. Brook, S. L., & Bowler, D. M. (1992). 'Autism by Another Name? Semantic and Pragmatic Impirments in Children'. Journal of Autism and Developmental Disorders, 22(1), 61-81. Conti-Ramsden, G., Crutchley, A., & Botting, N. (1997). The extent to which psychometric tests differentiate subgroups of children with SLI. Journal of Speech Hearing and Language Research, 40, 765-777. Leinonen, E., & Letts, C. (1997). Referential communication tasks: Performance by normal and pragmatically impaired children. European Journal of Disorders of Communication, 32, 53-65. Rapin, I., & Allen, D. (1983). Developmental language disorders: Nosologic considerations. In U. Kirk (Ed.),

Neuropsychology of language, reading, and spelling (pp. 155-184). New York: Academic Press. Sahlen, B., & Nettlebladt, U. (1993). 'Context and Comprehension: A Neurolinguistic and Interactional Approach to the Understanding of Semantic-Pragmatic Disorder'. European Journal of Disorders of Communication, 28(2), 117-140. Shields, J., Varley, R., Broks, P., & Simpson, A. (1996). 'Hemispheric Function in Developmental Language Disorders and High-Level Autism'. Developmental Medicine and Child Neurology, 38, 473486. Shipley, K. G., & McAlfee, J. G. (1998). Assessment in Speech-Language Pathology: A resource manual (2nd ed.). San Diego, London: Singular Publishing Group Inc.

Acknowledgments
We acknowledge the subjects and their family members for their cooperation throughout the study.

Appendix: The Childrens Communication Checklist


For each statement, the rater is asked to judge whether the statement DOES NOT APPLY, APPLIES SOMEWHAT, or DEFINITELY APPLIES. The option unable to judge is also given, but raters are discouraged from selecting this unless they have not had the opportunity to observe the behavior in question. For each scale, the base score is 30. For negative items (shown as ), 2 points are deducted from this total for each item coded DEFINITELY APPLIES, and 1 point is deducted for APPLIES SOMEWHAT. For positive items (shown as +), 2 points are added to the total for DEFINITELY APPLIES and one point is added for APPLIES SOMEWHAT. The pragmatic composite is the sum of scales C to G. A: Speech 1. + people can understand virtually everything he/she says 2. people have trouble in understanding much of what he/she says
72

3. + seldom makes any errors in producing speech sounds 4. mispronounces one or two speech sounds but is not difficult to understand; e.g. may say th for s or w for r. 5. production of speech sounds seems immature, like that of a younger child, e.g. says things like, tat for cat, or chimbley for chimney, or bokkle for bottle 6. seems unable to produce several sounds; e.g. might have difficulty in saying k or s, so that cat and sat are both pronounced as tat 7. leaves off beginnings or ends of words, or omits entire syllables (e.g. bella for umbrella) 8. it is much harder to understand when he/she is talking in sentences, rather than just producing single words. 9. + speech is extremely rapid

JAIISH, Vol. 27, 2008

Continuum of Developmental Language Disorders

10. seems to have difficulty in constructing the whole of what he/she wants to say: makes false starts, and repeats whole words and phrases; e.g., might say can I- can I- cancan I have an have an ice cream? 11. +speech is clearly articulated and fluent B: Syntax 12. speech is mostly 2 to 3 word phrases such as me got ball or give dolly 13. + can produce long and complicated sentences such as: When we went to the park I had a go on the swings; I saw this man standing on the corner 14. tends to leave out words and grammatical endings, producing sentences such as: I find two dog; John go there yesterday My grandma cat been ill 15. sometimes makes errors on pronouns, e.g. saying she rather than he or vice versa C: Inappropriate initiation 16. talks to anyone and everyone 17. talks too much 18. keeps telling people things that they know already 19. talks to himself/herself in public 20. talks repetitively about things that no-one is interested in 21. asks questions although he/she knows the answers 22. it is sometimes hard to make sense of what he/she is saying because it seems illogical or disconnected 23. + conversation with him/her can be enjoyable and interesting D: Coherence 24. + can give an easy-to-follow account of a past event such as a birthday party or holiday 25. + can talk clearly about what he/she plans to do in the future (e.g. tomorrow or next week) 26. would have difficulty in explaining to a younger child how to play a simple game such as snap or happy families 27. has difficulty in telling a story, or describing what he/she has done, in an orderly sequence of events 28. uses terms like he or it without making it clear what he/she is talking about 29. doesnt seem to realise the need to explain what he/she is talking about to someone
73

who doesnt share his/her experiences; for instance, might talk about Johnny without explaining who he is E: Stereotyped language 30. pronounces words in an over-precise manner: accent may sounds rather affected or put-on, as if child is mimicking a TV personality rather than talking like those around him/her 31. makes frequent use of expressions such as by the way, actually, you know what?, as a matter of fact, well, you know or of course 32. will suddenly change the topic of conversation 33. often turns the conversation to a favourite theme, rather than following what the other person wants to talk about 34. conversation with him/her tends to go off in unexpected directions 35. includes over-precise information in his/her talk, e.g. will give the exact time or date of an event, e.g. when asked when did you go on holiday may say 13th July 1995 rather than in the summer 36. has favourite phrases, sentences or longer sequences which he/she will use a great deal, sometimes in inappropriate situations 37. sometimes seems to say things that he/she does not fully understand F: Use of context 38. tends to repeat back what others have just said 39. his/her ability to communicate clearly seems to vary a great deal from one situation to another 40. takes in just one or two words in a sentence, and so often misinterprets what has been said 41. +can understand sarcasm (e.g. will be amused rather than confused when someone says isnt it a lovely day! when it is pouring with rain) 42. tends to be over-literal, sometimes with (unintentionally) humorous results. For instance, a child who was asked Do you find it hard to get up in the morning replied No. You just put one leg out of the bed and then the other and stand up Another child who was told watch your hands when using scissors, proceeded to stare at his fingers

JAIISH, Vol. 27, 2008

Continuum of Developmental Language Disorders

43. gets into trouble because he/she doesnt always understand the rules for polite behaviour and is regarded by others as rude or strange 44. may say things that are tactless or socially inappropriate 45. treats everyone the same way, regardless of social status: e.g. might talk to the head teacher the same way as to another child G: Rapport 46. ignores conversational overtures from others (e.g. if asked what are you making? just continues working as if nothing had happened) 47. seldom or never starts up a conversation; does not volunteer information about what has happened 48. doesnt seem to read facial expressions or tone of voice adequately and may not realise when other people are upset or angry 49. poor at using facial expression or gestures to convey his/her feelings; he/she may look blank when angry, or smile when anxious 50. + makes good use of gestures to get his/her meaning across 51. seldom or never looks at the person he/she is talking to: seems to actively avoid eye contact 52. tends to look away from the person he/she is talking to: seems inattentive or preoccupied 53. + smiles appropriately when talking to people H: Social relationships 54. + is popular with other children 55. + has one or two good friends 56. tends to be babied, teased, or bullied by other children

57. is deliberately aggressive to other children 58. may hurt or upset other children unintentionally 59. a loner: neglected by other children, but not disliked 60. perceived as odd by other children and actively avoided 61. has difficulty making relations with others because of anxiety 62. with familiar adults, he/she seems inattentive, distant or preoccupied 63. overly keen to interact with adults, lacking the inhibition that most children show with strangers I: Interests 64. uses sophisticated or unusual words; e.g. if asked for animal names might say aardvark or tapir 65. has a large store of factual information: e.g. may know the names of all the capitals of the world, or the names of many varieties of dinosaurs 66. has one or more over-riding specific interests (e.g. computers, dinosaurs), and will prefer doing activities involving this to anything else 67. + enjoys watching TV programmes intended for children of his/her age 68. seems to have no interests: prefers to do nothing 69. + prefers to do things with other children rather than on his/her own 70. prefers to be with adults rather than other children Adapted from Bishop & Baird (2001). Parent and teacher report of pragmatic aspects of communication: use of the Childrens Communication Checklist in a clinical setting.

74

JAIISH, Vol. 27, 2008

Analysis of Oral and Written Narratives of Language Impaired Children

Analysis of Oral and Written Narratives of Children with Language Impaired Learning Disabilities
1

Siddiqi Tehniat & 2Mukhopadhyay Sourav

Abstract
This study investigated and compared oral and written narratives of children with Language Impaired Learning Disabilities (LILD) and typically growing children as control group using strict reliability measures. The influence of context defined in terms of three narrative elicitation tasks for this study were story retell, story generation and spontaneous narratives.. Three children with LILD subjects aged between 9-12 years were compared to three typically growing children of same age group. The narratives were analysed at a micro-structural level using measures of productivity and form complexity. A specific analysis of writing was also performed. The results of this study indicate that the individuals with LILD did not perform well compared to control group in most of the oral and written task. However, the differences were not found to be statistically significant for oral task. It was also found that oral narrative productions for individuals with LILD were better than their written task
Key words: Oral and written narrative, Discourse productions, Elicitation tasks, Language impaired learning disability.

Narratives are a form of oral or written discourse consisting of extended units of texts (Owens, 1999). The production of narratives is a skill used for communication and children narrate stories to their parents, teachers, and peers every day (Wright & Newhoff, 2001). To produce narratives successfully, children must be able to organize the ideas of their narrative so as to provide an introduction to the story, to maintain the relationship among events or actions of the story, and to present a logical conclusion. Studies investigating the development of narratives indicate that the oral narratives produced by children go from being unstructured sets of utterances to a well-formed narrative. By the age of around six years or by the time children start school they have acquired the basic structure of narratives, which tend to follow a full adult pattern (Applebee, 1976). In the early school years spoken and written narrative are not so highly differentiated (Gillam and Johnston, 1992). With increased mastery of the mechanical aspects of writing, spoken and written narratives start to become differentiated. This differentiation between oral and written
1

modes starts to emerge between the ages of 9-12 years. Initially oral narratives are superior to written narratives. However as the children gain control over their written productions, written narratives become superior to oral narratives (Gillam and Johnston, 1992). Evidence suggests that the grammatical and syntactical organisation of spoken and written forms is distinct to each other. Written texts contain sentences, whereas spoken texts are typically made up of clausal complexes, which may not have a clear syntactic structure (Kress, 1982). The study of narrative discourse is becoming popular. This is because many individuals score within normal limits on standardised language tests; nevertheless deficits in language are apparent within discourse production and processing (Van Leer and Turkstra, 1999). Discourse analysis is also useful for both those population groups for whom standardised language tests are not available, for the assessment of language in a naturalistic setting and for the analysis of language beyond the sentence level. Narrative analysis is an important diagnostic tool for the assessment of language for different

Speech Language Pathologist & Audiologist, Speech and Language Therapy Services, Gaborone Botswana, 2Lecturer, Special Education (Speech Pathology & Audiology), Dept. of Educational Foundations, University of Botswana, email:souravspeech@yahoo.co.in,

75

JAIISH, Vol. 27, 2008

Analysis of Oral and Written Narratives of Language Impaired Children

population groups such as learners with learning disabilities, or who have aphasia. It is observed that narratives produced by different population groups are qualitatively different from each other (Owens, 1999). Narratives allow for the analysis of many different features of language, such as syntax, morphology, pragmatics, phonology as well as word finding difficulties (German and Simon, 1991). These may be assessed from a single elicitation or a compilation of different narratives elicited from an individual. A large number of researchers have tried to identify the characteristic features of narratives produced by learners with learning disabilities (LD)1 as well as determine the influence of the nature of the elicitation task or context on the production of narratives. These studies mostly compared story retelling tasks and story generation tasks (Liles, 1993; Ripich and Griffith, 1988). Liles (1993) cited a study carried out by Merritt and Liles (1989). In that study, the authors compared the story retelling tasks to story generation tasks. It was found that for both LILD and control groups, retold stories are longer and contain a greater amount of information. Spontaneous narratives were less frequently used as a method of data collection. It can be assumed that spontaneous narratives are closer to story generation in genre. However, the difference would lie in the fact that for story generations task one would attempt to produce a literate adult type model, whereas spontaneous narratives task reflect true internalisation of this adult type model. Newcomer and Barenbaum (1991) provided a review of the different aspects of the written narrative abilities of learners with LD. These studies indicate that learners with LD made more mechanical and spelling mistakes than normal subjects. The essays written by these group of learners were poorly planned and used a fewer number of words than typically developing children. Furthermore, learners with LD used

fewer novel words. Despite these deficiencies in the performance of learners with LD, it was also noticed that there was no difference between the complexities of the syntactic constructions used by learners with LD when compared with typically growing children. It was observed that both learners with LD and typically growing children used the same number of T-Units2 (Newcomer and Barenbaum, 1991). Gillam and Johnston (1992) investigated the relationship between both spoken and written narratives in children with LILD between the ages of 9-12 years. Their findings indicated that spoken narratives contained longer sentences with more linguistic connective devices but written sentences were more complex than spoken sentences. In essence these studies indicate that learners with LD across all grade levels tended to produce less coherent oral and written narratives than typically developing children. In terms of writing it was also observed that learners with LD produce more errors of writing than normal individuals. In majorities previous studies failed to find significant differences between the performances of LILD and control groups (Ripich and Griffith, 1998 cited in Henshilwood, 1998). Strong and Shaver (1991) suggested that these conflicting findings on narrative productions might be due to the unreliability of results. Therefore caution must be taken when interpreting data from these studies, as strict reliability measures were not always employed (Henshilwood & Ogilvy, 1999; Strong & Shaver, 1991). Present study aimed to assess the microstructure parameters of narratives and compared the performance of LILD children and typically growing children in oral and written narratives across three narrative tasks: story retell, story generation and spontaneous narratives whilst employing strict reliability measures. Furthermore, this study aimed to assess whether single-task narratives assessment was clinically more useful compared to multi-task narrative assessments.

1 Children with learning disabilities are defined as those with normal intelligence, intact sensory and emotional functioning but who exhibit a disorder in one or more of the basic psychological processes involved in understanding or using language, spoken or written (Donahue, Pearl and Bryan, 1982, p.397). The learning disabled group is diverse and their difficulties are not necessarily obvious on standardised test.

_________________________________
2

T-unit is a main clause and any other subordinate clause that may be attached to it or embedded in it. For written narratives punctuation and capitalisation were ignored when calculating T-units (Paul and Smith, 1993; Houck and Billingsly, 1989). Segmentation into T-units was done according to the procedure outlined by Vorster (1980) for the Test for Oral Language Production (TOLP). This form of segmentation is required for the micro-structural analysis of transcriptions. T-units were judged to be complex if they were grammatically complete and correct and contained a main clause together with one or more additional coordinating, subordinating, complementing or relative clauses (Gillam and Johnston, 1992: p.35).

76

JAIISH, Vol. 27, 2008

Analysis of Oral and Written Narratives of Language Impaired Children

Method
Research Design A multiple case study design was used. This increased the reliability of the data obtained and also controlled for the heterogeneity in the LILD population. Three narratives elicited using three different narrative elicitation tasks were used over two sessions to ensure that the narratives were

representative of the subjects true ability, thereby increasing the internal reliability of the data. Participants Three male children with known LILD and three aged matched typically growing male children without learning disabilities from the Cape Town, South Africa participated in this study. Table 2.1 provides a summary description of the subjects and controls.
NLD2 10.5 LILD3 10.8 NLD3 10.6

Chronological age at time of testing (yrs) IQ

LILD1 11.11

NLD1 11.10

LILD2 10.3

Below Average

Grade Medication Previous Therapy

4 Fixonase (for hayfever) Regular school with Remedial Teaching

With in normal range 6 None Regular school

Below Average 4 None Regular school Remedial Teaching

With in normal range 5 None Regular school

Above Average 4 Ritalin Regular school Remedial Teaching Occupational Therapy (2 yrs) Speech Therapy 4.0

With in normal range 5 None Regular school

Current Therapy and Years in Therapy

Speech Speech Therapy- 2.6 Therapy-2.0 Occupational Occupational Therapy-2.0 Therapy-4.0 Table 2.1: Biological and Educational Information of LILD and NLD subjects.

Procedure Each participant was individually assessed across two sessions. This was to increase reliability of the samples and to rule out performance differences arising from subject variables. Context defined as three different narrative tasks: story telling, story generation, and spontaneous speech were used to analyse microstructural features of narratives produced under different contexts. Each subject was required to produce both oral and written narratives in the same session on the same task. That is if they retold a story orally they were required to retell the same story in writing. Data on oral narratives were obtained first for all types of narrative tasks. Narrative elicitation task Story retell task was elicited by using Frog Story. This story was specifically constructed for story grammar research. It consists of seventeen
77

events and thirteen story grammar events (Ripich and Griffith, 1988). For story generation task, a single picture was shown to the children. This picture depicts a scene of a lion and a lioness fighting viciously. The child was expected to generate a story around this event. The child and one of the researchers were both engaged in a drawing activity for eliciting spontaneous narrative task. Whilst the researcher drew a picture she related one of her own narratives. The child was encouraged to do the same by asking him/her if anything like that has ever happened to him/her. Environment and recording Session one consisted of story retelling and session two included story generation and spontaneous narratives. No time limits were imposed for any task. Subjects were to write the written narratives on lined paper with a pen and

JAIISH, Vol. 27, 2008

Analysis of Oral and Written Narratives of Language Impaired Children

erasing equipment such as tippex was not provided. Oral narratives were video and audio taped. Both types of recording were used to reduce transcription errors due to recording variables. Each subject was required to read the written narratives aloud so that misspelled or illegible words could be resolved. In this study the protocol of Gillam and Johnston, (1992), was strictly used. Transcription and treatment of narrative samples Oral narratives were transcribed verbatim and then treated by excluding pauses and fillers such as uhm and yeah. Unintelligible utterances were transcribed phonetically and included as one word. Written narratives were typed without alteration to their content, spelling punctuation, capitals, paragraph structure and format. Crossed out words were ignored. Illegible words were verified from the video recording of the subject reading the narrative. Treated narratives were reformatted according to the transcription format used in the Computerised Profiling (Long, Fey, Chanelle, 2000) computer program. Analysis of Data Frequency counts and the percentages were made for all the tasks across all the six subjects and displayed in the tables for planned comparisons across the tasks and within the subjects (see table 3.1; 3.2; 3.3) and later on analysed statistically. This involved scrutiny at the word and sentence level. The focus was to scrutinise language at the level of syntax, semantics, phonology or spelling in written narratives. The measures give an indication of the complexity and depth of the narrative produced. Productivity and form complexity were measure for oral narratives and specific analyses of writing were performed. Productivity: The total number of words and number of different types of words per t-unit3 were calculated (Vorster, 1980) Form Complexity: The number of adverbs, prepositions and co-verbs4 per t-unit were calculated as outlined in Test of Oral Language Production (Vorster, 1980) For Written narratives a specific analysis of writing was performed. Mechanical aspects of
3 The following formula was used: The number of complex, grammatically correct T-units X 100 Total number of T-units (Gillam and Johnston, 1992)

writing productivity, syntactic maturity, vocabulary and mechanics were analysed based on the scoring procedures used by Houck and Billingsley (1989). Reliability In order to determine the reliability, inter-rater and intra-rater reliability as well as coder reliability were used. The intra-rater and inter-rater transcription and coder reliability are presented in Table 2.2. Liles (1993) indicated that conflicting findings in many previous studies may be influenced by poor reliability. An arbitrary value of 90% reliability suggested by Strong and Shaver (1991) was decided as being the minimum acceptable reliability for the present study. Both transcription and coder reliability were calculated for the narrative to ensure that the results obtained for the analysis were valid. According to this criterion both intra-rater and inter-rater measures of reliability met the criterion, implying the results accuracy. Transcription Coder Intra-rater Inter-rater 96.2 % 94.9% 94.0% 91.2%

Table 2.2: Inter-rater and intra-rater transcription and coder reliability for narratives.

Results and Discussion


Productivity analysis Productivity analysis was carried out for both children with and without LILD. Table 3.1 displays the productivity analysis. It was found that oral and written narratives of LILD subjects contained fewer t-units than controls narratives. The number of words used by children with LILD and the control group in oral narratives were equally distributed. In the written narratives children with LILD used fewer numbers of words compared to typically growing children. Oral narratives of children with LILD had a higher number of words per t-unit than the oral narrative of controls groups. But written narratives of children LILD have a lower number of words per t-unit than the written narratives of controls groups. The type-token ratios (TTR) for oral and written narratives were lower for children with
Co-verb indicates relationships between noun and the main verb and forms a setting for action of the main verb for this research. [E.g. Lions are fighting outside; [are] was considered as co-verb for this research].
4

78

JAIISH, Vol. 27, 2008

Analysis of Oral and Written Narratives of Language Impaired Children

Participants LILD1

Narrative Types T1 T2

T-units Oral Written

Words Oral Written

Word/ t-unit Oral Written

Type-Token Ratio Oral Written

10 7 77 57 7.7 8.1 3.6 2.8 4 5 39 31 9.8 6.2 3.1 2.2 T3 16 5 167 45 10.4 9 4.7 3.1 T1 10 7 65 46 6.5 6.6 3.4 2.6 LILD 2 T2 30 12 213 74 7.1 6.2 3.5 3.3 T3 11 4 76 27 6.9 6.8 3.8 2.6 T1 6 6 66 43 11 7.2 3.8 2.7 LILD 3 T2 5 4 37 29 7.4 7.3 3.0 2.6 T3 7 3 87 19 12.4 6.3 3.6 2.4 T1 12 10 82 75 6.8 7.5 3.3 3.8 Control 1 T2 8 6 75 66 9.4 11 4.3 3.7 T3 25 12 220 97 8.8 8.1 5.4 4.2 T1 8 8 58 64 7.3 8 3.9 3.8 Control 2 T2 5 6 54 48 10.8 8 3.0 3.6 T3 7 2 43 32 6.1 10.7 3.7 3.5 T1 10 6 73 55 7.3 11 4.1 3.5 Control 3 T2 13 10 86 87 6.6 8.7 4.0 4.2 T3 14 10 128 83 9.1 8.3 4.4 4.2 Key: T1-Task 1 (story retell); T2-Task 2 (story generation); T3-Task 3 (spontaneous narrative); Tunits- number of t-units; Words- number of words; Word/ t-unit- number of words per t-unit (one decimal place). Table 3.1: Productivity analysis of oral and written narratives for three different narrative types for all participants.

LILD than for control groups. Mann-Whitney U test was used to check if these differences were statistically significant. It was observed that the differences between children with and without LILD in oral and written task were not statistically significant. Analysis across oral and written tasks was also carried out. It was observed that the number of t-units and the number of words in the written narratives were lower for the children with LILD, but a similar trend was not observed for the control participants. Both the number of words and the number of t-units dropped for children with LILD but similar drop was not seen in the number of words per t-unit across oral and written narratives. The Type-Token Ratio for the children with LILD was lower in the written narratives than the oral narratives, a difference that was not present for the control groups. Nevertheless the differences were not found to be statistically significant. Analysis between narrative types showed that specific trends for the number of t-units, number of words and the number of words per t-unit were not observed across the different narrative types in the oral and written modes for LILD and control groups. The TTR tended to be higher in oral spontaneous narratives than retold or generated narratives for most LILD and control groups. At
79

the same time, these differences were not found to be statistically significant.

Figure 3.7: Percentage of Complex T-Units in oral and written task for all the subjects.

Form complexity The results of the form complexity analysis of oral and written narratives produced by all participants are displayed in Table 3.2. The percentage of correct complex t-units in the oral and written story retell tasks (T1), the story generation tasks (T2) and the spontaneous narrative tasks (T3) are graphically presented in Figure 3.7. It is observed from the figure 3.7 that

JAIISH, Vol. 27, 2008

Analysis of Oral and Written Narratives of Language Impaired Children

only one subject produced six adverbs, and others did not use adverbs in both oral and written tasks. The prepositions were present approximately equally by both LILD and control groups in both the oral and written modes. Children with LILD used higher number of co-verbs in their oral narratives than control group. However, reverse trends were present in the written narratives of typically developing children. They were completely absent in the written narratives of LILD group. A higher number of correct complex t-units were present in the oral and written t-units
Adverb Prep

narratives of children with LILD than their written narratives. However for the control group, the number of correct complex t-units decreased slightly. A fewer number of connectives per t-unit were used in written narratives compared to oral narratives. However, this difference was not statistically significant between the two groups and between oral and written task. A close observation across tasks revealed that prepositions were most commonly present in the story-retelling task whereby the topic influenced the inclusion of prepositions into narratives; co-verbs were most commonly present in the story generation. They
Coverbs % Complex T-units Connectives Oral Written Oral Written

Narrative Types Participants LILD 1 T1 T2

Oral Written Oral Written Oral Written

0 0 3 3 3 0 20 0 1.3 0.9 0 0 0 0 2 1 0 0 0.5 0.6 T3 0 0 9 3 4 0 12.5 20 1.3 0.8 T1 0 0 3 4 0 0 10 0 0.7 0.6 LILD 2 T2 1 0 1 1 2 0 10 8.3 0.9 0.5 T3 1 0 0 2 0 0 9.1 0 0.5 0.5 T1 0 0 1 2 0 0 33.3 0 1.8 1.5 LILD 3 T2 0 0 0 0 1 0 20 25 0.8 1.5 T3 0 0 2 1 1 0 0 0 1.3 1.0 T1 0 0 0 3 0 1 8.3 30 0.9 0.5 Control 1 T2 1 0 2 1 2 2 37.5 50 0.6 0.3 T3 0 0 4 0 0 1 12 16.7 1.7 0.3 T1 0 6 0 3 0 0 25 25 1.3 0.1 Control 2 T2 0 0 0 0 0 2 20 16.7 1.6 0.2 T3 0 0 0 1 0 0 0 50 1.4 1.0 T1 0 0 2 2 1 0 50 33.3 1.6 1.2 Control 3 T2 0 0 1 1 1 2 0 10 1.3 1.2 T3 0 0 6 4 1 3 14.3 40 1.4 0.8 Key: T1-Task 1 (story retell); T2-Task 2 (story generation); T3-Task 3 (spontaneous narrative); Adverbsnumber of adverbs; Prep- number of prepositions; Coverbs- number of coverbs; %complex t-unitsthe percentage of correct complex t-units; connectives- the number of connectives per t-unit Table 3.2: Form complexity analysis of oral and written narratives for three different narrative types for all subjects.

of control group versus those produced by LILD group which was found to be statistically significant using Mann Whitney U test (p=.001). This increase was particularly noticeable for written narratives. The number of connectives per t-unit present was generally equally across LILD and control groups narratives. It was observed that the uses of prepositions were fairly distributed across the oral and written modes in both the groups. However, LILD group used fewer numbers of co-verbs in the written narratives than the oral narratives. The number of correct complex t-units was higher in the oral
80

were always present in the orally generated narratives for the LILD subjects. For most oral narratives a higher numbers of complex t-units were present in the generated narratives of children with and without LILD. In summary we can say that oral and written narratives of children with LILD were less productive and less complex then the oral and written narratives produced by control children. In addition the oral narratives of LILD children were superior to their written narratives. Lastly oral spontaneous narratives were more productive in terms of the higher Type-Token Ratio and oral

JAIISH, Vol. 27, 2008

Analysis of Oral and Written Narratives of Language Impaired Children

generated narratives were more complex with regards to the number of complex t-units present. However a larger sample would be needed to assess if this fact can be generalised as an established trend. For the present study it can be assumed that the different types of elicitation tasks
Narrative Types Written Productivity

high numbers of words were occasionally present in the sentences produced by children with LILD than the narratives used by control group. This could relate to punctuation errors whereby actual sentences were marked within the written text as they were extremely long. If the intended number

Syntactic Vocab Mechanics Maturity Participants Number of Word/ Morpheme/tWords>7 % correct % correct Sentences sentence unit letters caps spellings T1 1 57 8.9 1 16.7 80.7 LILD 1 T2 1 31 8 1 25.0 67.7 T3 1 45 9.6 0 50.0 73.3 T1 4 11.5 7.1 2 60.0 84.8 LILD 2 T2 4 18.5 6.6 9 61.5 58.1 T3 1 27 7.5 0 33.3 85.0 T1 1 43 7.8 2 33.3 90.7 LILD 3 T2 3 9.7 8.8 4 33.3 65.5 T3 1 19 6.3 2 75.0 89.5 T1 8 9.4 8.6 6 100 98.7 Control 1 T2 6 11 12.8 6 100 97.0 T3 10 9.7 9.3 2 41.5 100 T1 6 10.7 9.5 3 85.7 95.3 Control 2 T2 4 12 8.8 1 100 100 T3 2 16 12.3 3 66.7 93.8 T1 4 13.75 12.2 6 100 96.4 Control 3 T2 5 17.4 10 12 100 100 T3 5 16.6 9.9 6 100 98.8 Key: T1-Task 1 (story retell); T2-Task 2 (story generation); T3-Task 3 (spontaneous narrative); Word/sentence- number of words per sentence; Morphemes/t-unit- number of morphemes per t-unit; Words>7 let- number of words greater then 7 letters; % correct caps- Percentage of correct capitalisation; % correct spellings- Percentage of correct spelling; VocabVocabularyInsert table 3.3 about here

Table 3.3: Written productivity, syntactic maturity, vocabulary and mechanics analysis of written narratives in three different narrative types for all subjects.

do not result in more productive or complex narratives. Regardless of presentation modality, children with LILD process language more slowly, thus reducing the ability to integrate information successfully (McFadden & Gillam, 1996) and retell the essential parts of the story when asked to do so. Including all story grammar parts in the analysis or total number of recalled story grammar parts may add stronger support to the notion that the children with LILD process language more slowly. Specific Analysis of Writing The third specific aim of the research was to perform a specific analysis of written narratives. The results for this analysis for the three different narrative types produced by all six subjects are shown in Table 3.3. The results in Table 3.3 indicate that a greater number of sentences were used by the control group compared to children with LILD and a very
81

of sentences had been counted these values would have been much different. A lower number of morphemes per t-unit, number of words containing seven letters or more, the percentage of correct capitalisations and the percentage of correct spellings were present in the narratives of LILD subjects versus those produced by control subjects. Mann-Whitney U test was run to see the differences between the groups. Except for words per sentences all the other parameters were statistically significant (p = .001). As expected an analysis of the mechanical aspects of writing across the three different tasks did not indicate any particular trend. Results from the specific analysis of writing are in agreement to the findings of Newcomer and Barenbaum (1991). That is children with LILD have significantly compromised writing ability in terms of the mechanics of writing.

JAIISH, Vol. 27, 2008

Analysis of Oral and Written Narratives of Language Impaired Children

Conclusions
The present study aimed to evaluate oral and written narratives produced by children with LILD and compare them to the narratives produced by age matched control group across three narrative elicitation tasks, namely the story retell, story generation and spontaneous narratives. The study also aimed to assess the influence of context on three narrative elicitation tasks. The narratives were analysed in terms of microstructure measuring productivity and form complexity. A specific analysis of writing was also performed. Stringent transcription and coder reliability measures were employed. The transcription and coder reliability measures the arbitrary criteria of 90% reliability set by Strong and Shaver (1991). From an examination of the influence of context on the narrative production, it appeared that the story retell task was frequently the most superior form of narrative production because this is essentially a short-term memory task. The narratives produced were superior due to the fact that an adult type model was available to the subjects. The differences in the story retell and spontaneous narrative task lay in the fact that in story retell task children imitated an adult type model, whereas spontaneous narratives, information had to be retrieved from long-term memory and formulated by the children. The story generation task was self-contained less influenced by context compared to the other two tasks and hence it was the most salient task for the assessment of narratives. The poorest performances by children were for spontaneous narratives and differences between the two groups were highlighted in both the written and the oral mode. This may be the result of the high level of dependency on the surrounding context for spontaneous narratives. It may be concluded that the story generation task was a test of a childrens ability to produce an adult like narrative and the spontaneous narrative task was most useful in highlighting differences between the two groups. Differences between the two groups were highlighted in the written mode as well. There was a larger discrepancy between the oral and written production of the children with LILD whereas this discrepancy was not evident in the control group. Thus due to difficulties with the mechanical aspects of writing and metalinguistic deficits for writing (Newcomer and Barenbaum, 1991) the
82

children with LILD were unable to perform equivalently across the two modes. With regards to the methods of analysis employed it appeared that the Type-Token Ratios were the most useful in differentiating between the two groups of subjects in both the oral and written modes particularly for spontaneous narratives. However large differences were apparent for all measures in the written mode. However the lack of normative data made it difficult to determine the adequacy of the productions. This not only motivates the need for narrative analysis as an assessment toll but also makes narrative analysis difficult to interpret. The shortcomings in the present study were that subjects were not tested across time for the same narrative task, thereby ensuring the results reflecting the subjects true ability for that particular type of narrative were questionable. Secondly the study sample was small. Thus the study should be replicated using a larger subject sample. However the study illustrated that children with LILD have poorer narrative abilities compared to children without LILD. It is also illustrated that written narratives are poorer than oral narratives for children with LILD perhaps due to the fact that the mechanical demands are so high in written narratives that the individual may lose site of the message that they are conveying (Robson, 1988). Lastly the spontaneous narrative task was the most crucial for differentiating between the two groups. The story retell task appeared essentially a short-term memory task, useful for assessing the production of narratives following the presentation of an auditory model.

Implications and Future Research


This study highlighted the need for multiple narrative elicitation tasks during language assessment particularly the spontaneous narrative task. At present little information is available on the narrative production of older children with LILD. Hence further investigation in this population would enhance the language teaching model. Future studies may focus on comparing different types of narrative tasks and linguistic ability, whilst obtaining representative narrative samples for the same task over time. Lastly, Computerised Profiling (Long, Fey & Channell, 2000) appeared to be a timesaving tool for analysis of narratives. It was simple to use and provided quantified data regarding language.

JAIISH, Vol. 27, 2008

Analysis of Oral and Written Narratives of Language Impaired Children

References
Applebee, A. N. (1978). The childs concept of story: Ages two to seventeen. USA: The University of Chicago press. German, D. J. & Simon, E. (1991) Analysis of childrens word-finding skills in discourse. Journal of Speech and Hearing Research, 34, 309-316. Gillam, R. B. & Johnston, J. R. (1992) Spoken and written language relationships in language/learning-impaired and normally achieving school age children. Journal of Speech and Hearing Research, 35, 13031315. Henshilwood, L. & Ogilvy, D. (1999) Narrative discourse productions in older language impaired learning disabled children: employing stricter reliability measures. The South African Journal of Communication Disorders. 46, 45-53. Houck, C.K. & Billingsly, B. S., (1989) Written expression of students with and without learning disabilities: differences across the grades. Journal of Learning Disabilities, 22 (9), 561-568. Kress, G. (1982) Learning to write Great Britain. London: T. J. Press Ltd. Liles, B., Duffy, R., Merrit, D. & Prcell, S (1995). Measurement of narrative discourse ability in children with language disorders. Journal of Speech and Hearing Research, 38, 415425. Liles, B. Z. (1993) Narrative discourse in children with language disorders and children with normal language: a critical review of the literature. Journal of Speech and Hearing Research, 36, 868-882. Liles, B. Z. (1985) Cohesion in the narratives of normal and language disordered children. Journal of Speech and Hearing Research, 28, 123-133. Long, S. H., Fey, M. E. & Channell, R. W. (2000) Computerised profiling (version 3.9.6).

Cleveland OH: Case Western Reserve University. McFadden, T. U. and Gillam, R. B. (1996) An Examination of the quality of narratives produced by children with language disorders. Language, Speech and Hearing Services in Schools, 27, 48- 56. Newcomer, P. L. & Barenbaum, E. M. (1991). the written composing ability of children with learning disabilities: A review of the literature from 1980 to 1990. Journal of Learning Disabilities, 24(10), 578-593. Owens, R. E. (1999) Language Disorders: A functional approach to assessment and intervention. Boston: Allyn and Bacon Paul, R. & Smith, R. L. (1993). Narrative skills in 4-year-olds with normal, impaired, and latedeveloping language. Journal of Speech and Hearing Research, 36, 592-598. Ripich, D. N. & Griffith, P. L. (1988) Narrative abilities of children with learning disabilities and non-disabled children: Story structure, cohesion and propositions. Journal of Learning Disabilities, 21(3), 165-173. Strong, C. J. & Shaver, J. P. (1991) Stability of cohesion in the spoken narratives of language-impaired and normally developing school-aged children. Journal of Speech and Hearing Research, 34, 95-111. Van Leer, E. & Turkstra, L. (1999). The effect of elicitation task on discourse coherence and cohesion in adolescents with brain injury. Journal of Communication Disorders, 32, 327-349. Vorster, J. (1980). Test of Oral Language Production. Pretoria: South African Institute for Psychological and Psychometric Research. Wright, H, H & Newhoff, M. (2001). Narration abilities of children with language-learning disabilities in response to oral and written stimuli. American Journal of SpeechLanguage Pathology, 10(3).

83

JAIISH, Vol. 27, 2008

Phonological Processes in Typically Developing Children

Phonological Processes in Typically Developing Kannada Speaking Children


1

Sreedevi N. & 2Shilpashree H.N.

Abstract
The present study investigated the various phonological processes occurring in 2.6- 3 year old typically developing Kannada speaking children. 8 children, i.e., 4 boys & 4 girls served as subjects. 50 simple Kannada words, which commonly occur in the utterances of normal young children, were selected for the study. These target words were picturized and were used to elicit the target response from the subjects. The responses were audio recorded and the data obtained were transcribed using IPA transcription. Sound by sound analysis was carried out to identify various phonological processes. The results indicated that out of the 12 processes observed, none of the phonological processes qualified as a significant one in the speech of children in the age group of 2.6 to 3 years. However, the most commonly seen processes were final vowel deletion, retroflex fronting, /h/ deletion etc. This study indicates that most of the phonemes in Kannada including fricatives and trills are achieved by 3 years of age and this warrants revision of our existing norms on articulation development. The results also expand our understanding of child phonology in the critical early language learning period.
Key words: Phonological process, Kannada, Retroflex fronting, Typically developing children.

Children are not haphazard in their mispronunciations of words, but they are in fact quite systematic in their production (Ingram, 1976). The concept of phonological process was first introduced by Stampe (1973) and according to him learning of sound system requires suppression of a number of innate simplifying processes and simultaneously increasing number of contrast sounds. Hodson and Paden (1983) defined phonological process as regularly occurring deviation from standard adult speech patterns that may occur across a class of sounds, a syllable shape or syllable sequence. According to Lowe (1996) phonological processes are systematic simplified adult production of children. In short, processes are description of regularly occurring patterns observed in childs speech, which operate to simplify adult targets. Study of phonological processes provide a more comprehensive and adequate descriptive framework for error analysis because they describe the structural as well as the systemic
1

simplifications in the speech patterns and where ever necessary take account of contextual factors influencing production of sounds. It also provides a parsimonious basis for selecting those classes of phonemes which need immediate attention and intervention. The phonological development of children learning English as their first language has been well described. However, as Ingram (1981) points out, despite numerous studies on languages other than English, we know relatively little about phonological development in other languages. This necessitates the need for phonological process analysis in Indian languages as well. Literature reports that there are more than forty such different processes operating during childrens phonological development (Hodson, 1980). Stoel-Gammon and Dunn (1985) reviewed the studies of occurrences of phonological processes and identified the processes which disappeared by three years of age as unstressed syllable deletion, final consonant deletion,

Lecturer in Speech Sciences, Dept. of Speech Language Sciences, All India Institute of Speech and Hearing, Mysore570006, email:srij_01@yahoo.co.in, 2Junior Research Fellow, Dept. of Speech Language Pathology, All India Institute of Speech and Hearing, Mysore-570006,

84

JAIISH, Vol. 27, 2008

Phonological Processes in Typically Developing Children

consonant assimilation, reduplication, velar fronting, diminutization and prevocalic voicing. And the processes, which persist after three years, were identified as cluster reduction, epenthesis, gliding, vocalization, stopping, de-palatalization and final devoicing. Haelsig and Madison (1986) studied 50 children in the age group of 3-5 years in Native American English and reported that gliding of liquids, cluster reduction and weak syllable reduction to be the most frequently occurring processes in this age range. However, the percentage of occurrence of each of these processes declined with age. They also reported that the greatest reduction in the use of phonological processes occurred between 3 and 4 years of age. Roberts, Burchinal and Footo (1990) found that liquid gliding, fronting and deaffrication were dropping between 2.6 and 3 years and cluster reduction between 3.6 and 4 years based on their analysis of the speech sample of 145 children. The age at which a process occurred in less than 10% of the sample was interpreted as the age at which the process had dropped out for the group. There are sparse reports of phonological process analysis in languages other than English also. Becker (1982) studied 10 monolingual Spanish children aged four years and found that de-affrication, /r/ deficiencies, cluster reduction, epenthesis, weak syllable deletion and alveolar assimilation to be the most used processes in these children. Later Martinez (1986) reported Tap/Trill deficiencies, consonant sequence reduction, deaffrication, stopping, affrication, fronting, assimilation and sibilant distortion in 3-year-old Spanish children. Topbas (1997) studied the phonological acquisition in Turkish children and reported that from a cross linguistic perspective, the phonological patterns exhibited coincide broadly with universal tendencies, although some language specific patterns were also evident. In Turkish children /r/ was substituted by /l/, i.e.

liquid realization of another liquid where as in English, the /r/ is usually replaced by /w/ or /j/ a gliding process. The same finding is reported in Italian also (Bonoleni and Leonard, 1991). This is explained on the basis that some phonemes are more common where as some are rare in some languages. For example /w/ and /j/ are rare in Turkish and Italian where as they are more frequent in English. On similar lines, in the Indian context, Rahul (2006) reports that usage of /l/ for /r/ is a frequent finding in children in the age group of 2-3 years with Hindi as their native language. He reasons that both /l/ and /r/ are liquids and /l/ is easier among the two and is more preferred. Paulson (1991) studied 30 normal developing children of Mexican descent in the age range of 25 years. She found that the 2 year olds used phonological processes most frequently and the 4year olds least often. Her subjects used syllable reduction, consonant sequence reduction, prevocalic singleton omission, strident deficiencies, and /r/deficiencies. And miscellaneous error patterns were stopping, gliding, vowel deviation, epenthesis, substitution of /l/ for /r/ and sibilant distortions. Hua and Dodd (2000) studied the phonological acquisition in Putonghua language (Modern Standard Chinese) and suggested that vowels and final consonants were mastered earlier than initial consonants. As noted earlier, the literature on phonological processes is relatively fewer in Indian languages. However, in the recent past a number of such studies have been attempted in several Indian languages focusing on the normal phonological process usage and these have been briefly reviewed in Table 1. The present study intended to analyze the speech sample of typically developing Kannada speaking children (2.6 to 3 years) to identify the type of phonological processes present in their utterances.

85

JAIISH, Vol. 27, 2008

Phonological Processes in Typically Developing Children

Language Age Common processes seen Group 1 Sunil, T J (1998) Kannada 3-4 Fronting, Cluster reduction, Initial consonant deletion and years Affrication 2 Jayashree, U P Kannada 4-5 Fronting, Cluster reduction and Stopping (1999) years 3 Ramadevi et al Kannada 5-6 Stridency deletion, De-aspiration and Retroflex fronting (2002) years 4 Sreedevi et al (2005) Kannada 2-2.6 Retroflex fronting, Trill deletion, Depalatalization, Affrication, years Stopping, Cluster reduction etc 5 Sreedevi, N (2008) Kannada 1.6-2 Retroflex fronting, Initial Consonant deletion, Vowel lowering, Trill years deletion, Cluster reduction etc 6 Sameer, P. (1998) Malayalam 3-4 Cluster reduction, Final consonant deletion, Epenthesis and Deyears affrication 7 Bharathy, R (2001) Tamil 3-4 Epenthesis, Cluster reduction, Gliding, Nasal assimilation, Voicing years De-affrication & Fronting. 8 Santhosh, M (2001) Hindi 3-4 Cluster reduction. Epenthesis, Fronting, Gliding, Metathesis years Nasalization etc 9 Rajeev Ranjan Hindi 4-5 Cluster reduction, partial reduplication and aspiration (1999) years 10 Rahul Banjariya Hindi 2-2.6 Retroflex fronting, Deaspiration, /h/ deletion, Gliding, Initial (2006) years consonant deletion 2.6-3 Affrication, Denasalization, Monothongisation, Devoicing etc years Table 1: Review of Phonological Processes in various Indian Languages.

Author

Method
Subjects: Eight typically developing children, 4 boys and 4 girls in the age range of 2.6 to 3 years were considered for the study. All the children had the Mysore dialect of Kannada as their native language. They were screened for normal speech and language skills, hearing acuity and normal cognitive and motor development. Subjects were pooled from daycare centers and individual homes. Test Material: 50 simple Kannada words, which commonly occur in the utterances of typically developing young children, were selected for the study. Most of the test words used was bisyllabic, except for a few trisyllabic ones. The word list had the vowels of Kannada in the initial position and the consonants were in the initial and medial position of the target words. The target words were picturized on white cards of 4 x 6 size in color. The picture cards were arranged in order as per the sequence of Kannada alphabets. The target words included a few clusters also. The list of target words is given in Appendix 1. Procedure: Each individual subject was brought into a small noise free room and seated opposite to the examiner. Once the rapport was established, the examiner presented the target pictures one after another. The subjects were encouraged to name the
86

item in the picture card. The response obtained were audio recorded using a high fidelity portable digital Sony tape recorder. If any of the subjects failed to identify a target word, additional cues were presented by the examiner. In spite of additional cues, if the child failed to name the target picture, the child was asked to repeat after the examiner. Data Analysis: The data obtained from all the 8 subjects were transcribed by two speech language pathologists using broad and narrow IPA transcription. The inter judge reliability was approximately 86%. Following this a qualitative analysis was carried out for each subject separately. Each word utterance of the subjects was analyzed sound-by-sound. Based on the sound changes occurring, the phonological process operating was identified. Further, frequency of occurrence of the various processes was determined. The qualitative analysis of each childs speech sample was carried out using the following format:
Target word Phonemic structure of target CVC Response of the subject /bt/ Phonemic Structure of the response CVC Phonological process used Stopping

/bs/

JAIISH, Vol. 27, 2008

Phonological Processes in Typically Developing Children

Results and Discussion


Twelve various phonological processes were identified in typically developing children aged 2.6 to 3 years speaking Kannada as their native language. Each subjects utterances were analyzed separately for the type of the process used and the number of times it occurred. The overall percentage of occurrence of each process was obtained by computed by the following formula (Newman & Creaghead, 1988).
Number of times a process occurred ----------------------------------------------------Total number of words spoken X 100

i.e.
Total number of the same type of processes exhibited by all the subjects -----------------------------------------------------------Total number of target words spoken by all the subjects

X 100

colloquial Kannada. The next commonly seen process was retroflex fronting (2.5%). Sreedevi (2008) has reported that retroflex fronting was the most dominant (18%) phonological process operating in children aged 1.6 to 2 years in Kannada. This can be attributed to the complexity in its production, which involves curling the tongue to contact the palate. Therefore most often, a retroflex sound was substituted by an easier dental sound, which is more frontally placed with better visibility and which also requires relatively lesser exertion of the tongue. It is interesting to note that in the present age group of 2.6 -3 years, retroflex fronting had reduced substantially to a scanty 2.5%. As per the earlier literature reports, retroflex sounds are mastered after 3 years (4.6 years - Templin, 1956 (English); 3.6 years Babu, Bettagiri & Rathna, 1972 (Kannada). However, the present study indicates that retroflex sounds are produced correctly by 3 years of age in Kannada. Retroflex fronting was followed by /h/ deletion (1.75 %). This is again a common feature in colloquial Kannada. This was followed by cluster reduction and velar fronting. Cluster reduction was fewer in occurrence as the test sample contained only two clusters. Clusters were not included in more numbers in the test sample as the earlier literature reports suggested that clusters are mastered by 7 years or so and this study targeted a much younger age group. The finding of velar fronting, although negligible, is quite surprising. This can be explained on the basis that, though velar is an early sound, as per the developmental norms, velars are consistently produced by 4 years or so (Fundala & Reynolds, 1986). The remaining seven processes seen namely, affrication, vowel raising, stopping, diphthongization, /r/ deletion, non-geminate to geminate cluster and palatalization were all operating at one or less than 1% level and can be considered as an incidental finding. It is appealing to find that processes which are most commonly reported in Western literature like stopping or substitution of a stop for a fricative and /r/ deletion are almost negligent in this group of 2.6-3 years. This indicates that fricatives and trills are achieved almost consistently by 3 years of age in Kannada. Stopping and /r/ deletions are frequently reported in the Western studies even in children of older age groups (Stoel-Gammon & Dunn, 1985).

Percentage of occurrence of 5% or more was considered high, following Newman & Creaghead, 1988. The processes identified are listed in descending order of occurrence based on the percentage of occurrence in the 8 subjects studied in Table 2.
Phonological process Total observed % 1. Final vowel deletion 3.00 % 2. Retroflex fronting 2.50 % 3. /h/ Deletion 1.75% 4. Cluster Reduction 1.25% 5. Velar fronting 1.25% 6. Affrication 1.00 % 7. Vowel raising 1.00 % 8. Stopping 0.75 % 9. Diphthongization 0.75 % 10. /r/ deletion 0.50 % 11. Non geminate to geminate 0.50% cluster 12. Palatalization 0.50 % Table 2: Phonological processes seen in 2.6 3 years.

The results reveal that out of the 12 processes observed, none of them qualified as a significant phonological process in the speech of children in the age group of 2.6 to 3 years in Kannada. This is because none among the 12 processes crossed the critical value of 5% (Newman & Creaghead, 1988). However, the most commonly seen process among all the 8 children was final vowel deletion (3%). Final vowel deletion was mainly observed for borrowed words from English like /bus/, /ka:r/, /ril/ etc. which end with the vowel /u/ in
87

JAIISH, Vol. 27, 2008

Phonological Processes in Typically Developing Children

The overall results indicate that none of the 12 phonological processes observed qualified as a significantly operating process in the age group of 2.6 to 3 years in Kannada. Most of the processes sparingly occurred and perhaps with a few more repetitions, probably the child would have self corrected the errors. This observation certainly indicates that we need to revise our articulation test norms which were standardized in the sixties and early seventies. Also the processes seen were not similar to the ones reported frequently in Western languages. This is because of structural differences across the languages.

and Hearing Services in Schools, 17,107114. Ingram, D. (1976). Cited in Lass N.J. Speech & Language: Advances in Basic research and practice. Vol.8. NY. Academic Press, 1982 Ingram, D. (1981). Procedures for the phonological analysis of childrens language. Baltimore: University Park Press. Jayashree, U.P. (1999) Development of phonological processes of 4-5 year old children in Kannada speaking population. Unpublished Masters dissertation submitted to the University of Mangalore. Lowe (1996). Cited in Hegde, M. N.(2000). Assessment and treatment of articulation and phonological disorders in children. Austin, TX: Pro-ED. Rahul, B. (2006). Study of Phonological Processes in 2-3 Years old Hindi Speaking Normal Children. Un published Masters Dissertation, University of Mysore. Ramadevi, K J & Prema, K S. (2002). Phonological process in Hearing impaired Children. In Proceedings of the 4th ICOSAL, Chidambaram. Ranjan, R. (1999). Development of phonological processes of 4-5 year old children in Hindi speaking population. Unpublished Masters dissertation submitted to the University of Mangalore. Sameer, P. (1998). Development of phonological processes of 3-4 year old children in Malayalam speaking population. Unpublished Masters dissertation submitted to the University of Mangalore. Santhosh, M. (2001). Development of phonological processes in normal Hindi speaking children in the 3-4 years age group. Unpublished Masters dissertation submitted to the University of Mumbai. Sreedevi, N., Jayaram, M., & Shilpashree, H. N. (2005). Development of phonological Process in 2-2.5 year old children in Kannada. Abstract in Souvenir, Sixth International Conference on South Asian Languages (ICOSAL 6), Hyderabad Sreedevi, N. (2008). Study of Phonological processes in normal Kannada speaking
88

Conclusions
It can be said that understanding the pattern of reduction of phonological processes during the course of phonological development form the basis for dealing with the clinical population. The present study indicates that most of the processes are suppressed considerably by three years of age in typically developing children. Hence based on the results obtained it can be predicted that todays children acquire speech sounds much sooner as compared to their earlier counterparts. However, further standardization of the data on a larger population with more complex words is warranted for generalization. The findings of this investigation also augment in screening the appropriateness of a childs phonological skills especially in the clinical population.

References
Bharathy, R. (2001). Development of Phonological Processes in Tamil: 2-3 years. Masters Dissertation, University of Mysore. Deanine, M.M., and Hodson, B.W. (1994). Spanish Speaking Children's Phonological Assessment and Remediation, Seminars in Speech, language and Hearing Research, 3, 137-148 Fundala, J B., & Reynolds, W M.(1986). Cited in Pena Brooks and M N Hegde (2000), Assessment and treatment of articulation and phonological disorders in children. Austin, Texas: Pro- Ed. Haelsig, P.C., & Madison, C. L. (1986). A study of phonological processes exhibited by 3-, 4and 5- year old children. Language, Speech

JAIISH, Vol. 27, 2008

Phonological Processes in Typically Developing Children

children: 1.6-2 years. Interdisciplinary Journal of Linguistics, 1,103-110. Stoel-Gammon, C., & Dunn, C. (1985). Normal and disordered phonology in children. Clinical linguistics and Phonetics, 4, 145160.

Appendix 1

Acknowledgements
This study is a part of the ARF project titled Phonological processes in Kannada speaking children: A Profile. The authors wish to express their gratitude to Dr. Vijayalakshmi Basavaraj, Director, AIISH for permitting to carry out this study and Dr. M Jayaram, former Director, AIISH, for sanctioning the project. Also the authors wish to acknowledge the subjects who participated in the study.

89

JAIISH, Vol. 27, 2008

Aetiological & Therapeutic Searches by Netizen Parents on Autism Spectrum

A Profile of Aetiological & Therapeutic Searches by Netizen Parents/ Caregivers of Children on the Autism Spectrum
1

Venkatesan S. & 2Purushotham K.

Abstract
Autism is a growing tragedy of contemporary information age. A sizeable population of their parents is computer savvy, knowledgeable and active denizens on the new virtual reality available on world-wide web. They spend several hours on the computer keyboard asking questions, seeking answers, sending messages or exchanging notes between themselves or other professionals in virtual chat rooms, internet groups or through personal emails. The present study seeks to empirically profile the frequency, intensity and extensity of thematic concerns/searches by an organized group of internet netizens dedicated to autism. Content analysis of transcripts derived by data mining 3436 email exchanges of the netizens in the studied sample is classified/presented under three heads: (a) spread of information themes/choices; (b) etiology based issues/ transactions; and, (c) themes related to therapy or treatment of individuals with autism. The highest number of discussions center around the theme of therapies/treatments applicable for persons on the autism spectrum (N: 1697; 49.39 %). This is followed by online concerns of the group members on behavior problems or its management (N: 1488; 43.31 %) in their children. Their etiology based explorations cover four categories: (a) Biological; (b) Environmental; (c) Sensory; and, (d) Diet related causes of autism. Content analysis reveals as many as 238 types of treatment related key-words as discussed by members in the internet group. They are classified as nutrition-based, sensory, and education-behavior based therapies, Alternate Medical Systems, biomedical therapies, and Instrument based therapies respectively. The implications of the study are discussed in the context of parental quest for understanding the continuing challenge and enigma called autism
Key words: Internet groups Content analysis.

Autism is increasingly becoming the tragedy of contemporary information age. The kids with autism inhabit their inner worlds as their parents/caregivers desperately search for an understanding of its ever elusive causes and therapeutic management in the outside world. Many modern day parents are computer savvy active denizens of the new virtual reality available on the world-wide web. They spend several useful hours dabbling on the computer keyboard asking questions, seeking answers, sending messages or exchanging notes between themselves or other professionals in virtual chat rooms, internet groups, or through personal emails. These voyages are made over and above their regular and routine consultations with friends and well wishers,
1

doctors, hospitals, rehabilitation professionals, and others in the real world. Their preoccupations and queries relate to the unfortunate predicament of their children. There are major and recurrent themes of information needs and exchanges that are shared by these netizens. Various e-based diagnostic, counseling and therapeutic services, either professional or otherwise, charged or freebie, are increasingly becoming the order the day. Subscriber based Virtual Expert Clinics close the gap between high demand for special needs services and the limited accessibility and high existing costs of such services. These services provide planning, training and problem solving strategies to assist parents, caregivers, educators and professional service

Professor, Dept. of Clinical Psychology, All India Institute of Speech and Hearing, Mysore:570006, email:psycon_india@yahoo.co.in, 2Research Assistant, Dept. of Clinical Psychology, All India Institute of Speech and Hearing, Mysore: 570 006.

90

JAIISH, Vol. 27, 2008

Aetiological & Therapeutic Searches by Netizen Parents on Autism Spectrum

providers. Some e-therapy sites offer paid and/or freebie services via synchronized chats, others use emails alone. Online communication is fast, easily accessible from home, cost effective, time efficient, etc. Of course, this form of therapy also carries certain disadvantages. It is not suitable for every ones needs. It cannot be intensive and it cannot respond immediately and effectively in crisis situations. Internet groups related to persons with disabilities, including autism has become a reality in India. Several internet users (called denizens) make use of groups and networks like Autism India Network, Autism India Group on Yahoo, India Developmental Disabilities, Autism India, Help_Autism, India_autism_forum, India_autism_forum2, etc. A few examples of such groups based abroad is ASD Friendly, www.aspergersnw.org, AutismSpot, The Autism Connection, The Autism Depot, www.templegrandin.com,Auties.org, autismkey.net, lakeautismgroup@yahoogroups.com, etc. These groups or forums have their own entry/exit services, policies, and philosophies. Several interested or affected individuals, parents, caregivers, professionals and others seek and enter into membership into these groups. They are then regularly posted with information, introductions, discussions, articles, essays, or other details on the subject matter of autism. They are also given an opportunity to post their own questions, experiences, requirements for the peer group to respond through an effective moderator. Aims & Objectives It is the objective of this study to a. Empirically profile the frequency, intensity and extensity of diagnostic, etiological and therapeutic concerns by a group of internet netizens regarding their children with autism; b. Attempt a qualitative and quantitative profile in the patterns or content of their concerns as expressed by the target sample/s on or regarding their children with autism.

passive member of this group. The current membership strength of this internet group is over 1500 denizens. During its first year, the number of registered members and their online transactions was meager. Data collected during that period was not considered for inclusion in this study. An overall of 3793 email transactions between the members was collected and individually recorded in the form of written transcripts from the period between January, 2005 and December, 2007. Within this figure, there were 357 repeat emails that were discarded. The final inclusion comprised 3436 emails as the sample for content analysis in this investigation. Content analysis (or textual analysis), as defined in this study, involves a set of procedures for collecting and organizing the non-structured email information into a standardized format that allows one to make inferences about the target phenomenon (Krippendorff, 2004; Berelson, 1971; Holsti, 1969). It is aptly recognized as study of recorded human communications, such as, books, websites, emails, paintings and laws (Babbie, 2005). Content analysis of the transcripted emails was carried out in terms of the most typical themes, notes, messages, information, queries, answers, or other forms of information exchange happening between/within members of the internet group. The shared messages between members of the internet group were recorded on daily basis using non-participant objective observation methods. The transcript pro-forma recorded the title of message, thematic concern/content, date/s of their postings and the continued frequency or duration of messages on that given topic of discussion. The names of netizens, notes of self introduction, often congratulatory and/or acknowledgement mails were not recorded for ethical reasons. Content analysis of the netizens concerns from the transcripts was carried out in three broad domains: (a) spread of information themes/choices; (b) etiology based issues/transactions; and, (c) themes related to therapy or treatment of individuals with autism. All exchanges and discussions between the virtual group members were classified as such.

Method
The sample for this longitudinal study was derived from an internet group for autism in India. The group is in existence over four years. All online transactions occur in this virtual group only in English. One of the authors is an observing and
91

Results and Discussion


The results of content analysis of transcripts derived by data mining email exchanges of the netizens are classified and presented under the following three heads:

JAIISH, Vol. 27, 2008

Aetiological & Therapeutic Searches by Netizen Parents on Autism Spectrum

(a) Spread of Information Themes/Choices Out of the 3436 emails eventually included for content analysis in this study, the highest number (N: 1697; 49.39 %) of discussions centered on the theme of therapies or treatments for persons with autism. This is followed by emails on self introductions (N: 1684; 49.01 %) by the netizens, and exchange of individual/institutional contact addresses (N: 1679; 48.78 %). The information needs of parents/caregivers querying availability of services/professionals is reported in several studies (Peshwaria, Menon, Ganguly, Roy, Pillay and Gupta, 1995; Garshelis and Mc Connell, 1993; Gowen, Christy and Sparling, 1993; Bailey, Blasco and Simeonsson, 1992; Bailey and Simeonsson, 1992). This is followed by online concerns of the group members about behavior problems (N: 1488; 43.31 %) in their children and its
Sl. No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

management, consulting (N: 1464; 42.61 %), or seeking peer approval on the diagnosis/ therapeutic practices being followed by them (N: 1224; 35.62 %). The internet group discussions offer a splendid opportunity for the denizens to share experiences, ventilate doubts, or deliberate on ongoing day-to-day problems being faced by them in the rearing and caring of their children with autistic disorders. About a third of the discussions by the members pertain to exploring the possible cause/s of the condition (N: 1089; 31.69 %) in their children. The group is also equally preoccupied in seeking mutual clarifications about the diagnosis in their children (N: 1087; 31.64 %). This is particularly true of new entrants into the group who describe the individual signs and symptoms of their wards/children and posting questions whether they signify anything related to disorders on the autism spectrum.

Information Themes

Rank 15 4 7 11 5 3 18 8 10 9 14 16 13 12 6 2 17 1

Advocacy 601 17.49 Behavior Problems 1488 43.31 Causes 1089 31.69 Conferences & Seminars 838 24.46 Consulting 1464 42.61 Contact Addresses: Institutions or Individuals 1679 48.78 Courses 272 7.92 Diagnostic Clarifications 1087 31.64 Enquiries 984 28.64 Greetings & Congratulatory Messages 1030 29.98 Information on Forthcoming Events 623 18.13 Information on Relevant Websites 581 16.90 Introductions/Welcome Notes 628 18.28 Notes, News, Views & Reviews 830 24.16 Peer Approval 1224 35.62 Self Introductions 1684 49.01 Send Articles 328 9.55 Therapies, Treatments & Techniques 1697 49.39 TOTAL 3436 Table 1: Spread of Information Themes Discussed by Netizens.

There are a sizeable number of social greeting and congratulatory messages exchanged between the members (N: 1030; 29.98 %). These greetings relate to local festivities, commencement of year/s, seasonal salutations, etc. These exchanges lubricate the social dynamics and lend a tone of informality to the interactions between members of the virtual group. Despite the fact that the internet group
92

smacks of members who are highly educated and belong to the elitist class of intellectuals, it is seen that their concerns have not yet reached the level of active or belligerent advocacy (N: 601; 17.49 %) for the cause of autism in the country. There are fewer exchanges on information related to websites (N: 581; 16.90 %), books and articles (N: 328; 9.55 %) or

JAIISH, Vol. 27, 2008

Aetiological & Therapeutic Searches by Netizen Parents on Autism Spectrum

about online/offline curses related to the field of autism disorders. (b) Etiology Based Issues/Transactions The etiology based concerns of the internet group members as reflected in their emails pertaining to possible causes of autism was next taken up for analysis. This causal domain
Biological Allergies Amygdala Neurons Autoimmune Disease Brain Abnormalities Chromosomal Familial Food Additives Genetic/Heredity Lead Maternal Antibodies Mercury Mercury Toxicity Metabolic Disturbances Missing Brain ProteinCdk5 MMR Vaccine Oxidation Stress Pesticides Synapse Disconnection Teratogens Toxicity Viral Infection Vitamin Deficiency N Sum Mean SD Sumsq Variance N 28 14 22 18 16 8 29 11 33 7 18 78 19 8 99 10 16 4 13 6 12 24 22 493 22.4 22.9 22139 528.2 Environmental Accidents Bad Parenting Birth Trauma Emotional Trauma Excess Hygiene Folic Acid Perinatal Environment Pesticides Pollution Poor Stimulation Prenatal Environment Refrigerator Mother Social Factors Stress Ultrasound Watching TV N 12 6 7 10 7 6 4 14 12 19 11 7 9 11 7 27

analysis revealed as many as 341 possible causes emerging from the transcripts as being perceived, reported and discussed by the 1089 internet group members (31.69 %). The area wise causal perceptions of the netizens are classified into four categories: (a) Biological; (b) Environmental; (c) Sensory; and, (d) Diet Based respectively (Table 2).
Sensory Auditory Gustation Hypersensitive Hyposensitive Olfactory Tactile Vestibular Visual N 30 7 21 9 5 7 4 3 Diet Based Alpha Protein Deficit Diet disturbances Enzyme Dysfunction Food Allergies GFCF Leaky Gut Syndrome Nutritional Deficiency Poor Nutrition N 13 67 23 32 111 42 26 27

16 169 10.6 5.8 2281 33.1

8 86 10.7 9.6 1570 92.2

8 341 42.6 31.9 21701 485.5

(SS-BG: 6330.8693; df: 3; MS: 2110.29; SS-WG: 19398.6307; df: 50; MS: 387.97; F: 5.44; p: 0.002572) Table 2: Spread of Etiology Themes Discussed by Netizens.

There is obviously greater preoccupation of the internet group members on perceived/reported biological causes for disorders on the autism spectrum (N: 22), followed by environmental cases (N: 16) and equally on sensory and diet based causes (N: 8) respectively. However, within these causal domains, discussions on GFCF (Gluten Free Casein Free) diet stands out as the most discussed theme among all the supposed
93

causes of the netizens (N: 111; 10.19 %), followed by chat about MMR vaccines (N: 99; 9.09 %), mercury toxicity (N: 78; 7.16 %), diet disturbances (N: 67; 6.05 %), etc. On the whole, the group is less attentive to remediable environmental and/or sensory issues related to individuals on the autism spectrum. These differences are also found to be statistically significant (F: 5.44; p: <0.002).

JAIISH, Vol. 27, 2008

Aetiological & Therapeutic Searches by Netizen Parents on Autism Spectrum

(c) Treatment/Therapy Based Issues / Transactions Many members in the studied internet group are apparently more interested on treatment/therapy based issues (N: 1697) than upon etiology based concerns (N: 1089) related to autism. Content analysis of available transcripts revealed as many as 238 types of key therapeutic or treatment related terms being highlighted or discussed by the members in the internet group. They are classified as nutrition based therapies, sensory therapies, education-behavior based therapies, Alternate Medical Systems, bio-medical therapies, and Instrument Based Therapies respectively.
Sl.no A B C D E F G Category Nutrition-Diet Based Therapies Bio-medial Therapies Education-Behavior Based Therapies Sensory Therapies Alternative Medical Systems Medication/Drug Therapy Instrument Based Therapies Total Types 51 24 40 50 40 24 9 238 N 459 337 287 199 189 133 93 1697 Percent 27.05 19.86 16.91 11.73 11.14 7.84 5.48

Therapy, Conductive Education, Floor Time Intervention, Discrete Trial Training (DTT), Early Intensive Behavioral Intervention (EIBI), Face Talk, Facilitated Communication, Family Therapy, Family Focused Early Language Approach Program, Home Schooling/Teaching, Hypnotherapy, Portage Program, Prompt Therapy, Relationship Development Intervention (RDI), Social Skills Training, Verbal Behavior Intervention (VBI), etc. There is copious discussion in the group on the pros and cons as well as possibilities or potentials of about 40 different Alternate Medical Systems in the treatment of autism. A few examples of this category included Acupressure, Acupuncture, Ayurveda, Biosys or Magnet Therapy, Brain Gym, Chinese Medicine or Quigong, Chiropractic Treatments, Dr. Oswals G Therapy, Faith Healing, Hair Analysis and Treatment Protocol, Homeopathy, Love Therapy, Meditation, Miracle Therapy, Nature Cure, Naturopathy, Neuro-Linguistic Program, Pranic Healing, Reiki, Siddha Medicine, Spiritual Medicine, Tibetan Medicine, Unani, Yoga, etc. Among the bio-medical therapies (N: 24) were included in the discussions protocols or program to treat Candida Albicans, Antibiotic Treatment, Chelation, DAN Doctors, DMG, DMPS Therapy, Detoxification, Fertility Treatment, Live Cell Therapy, Maggot Debridement Therapy/ MDT/Larva Therapy, Stem Cell Therapy, etc. Simultaneous discussions were carried out on the experiences of interacting members related to their use of various drugs/medications like Adderal, Benedryl, Botox Injections, Carbamazepine, Cucrumin, Pemoline, Encephebol, Learnol Plus, Olanzapine, Mentat Tonic, Nystatin, Phenol Injections, Piracetam, Prozac, Quetiapine, Resperidol, Ritalin, Secretin, Senetin, Straterra, etc. There was exchange of information and discussions between the denizens on Instrument Based Therapies (N: 9), such as, Advanced Biomechanical Rehabilitation, Electrical Stimulation, Therapeutic Electrical Stimulation, Foot Nerve Therapy, Light Sound Therapy, Neurofeedback, Neuro-developmental Therapy, Vibroacoustic Therapy, Video Therapy, etc. Nutrition-diet based therapies (N: 459; 27.06 %) was the most talked about theme of members in the internet group followed by discussions on biomedical therapies (N: 337; 19.86 %), educationbehavior based therapies (N: 287; 16.91 %),
94

Table 3: Spread of Therapeutic Themes and Choices Discussed by Netizens.

There are 51 subtypes of nutrition-diet based therapies mentioned in the transcripts such as, Almond, Aloe Vera Juice, Brown Barley, Brown Rice Therapy, Calcium, Carbohydrate Diet, Cod Liver Oil, EFA (Essential Fatty Acids) Therapy, Feingolds Diet Therapy, Feldenkraiz Therapy, Fish Oil Therapy, Gluten and Casein Free Diets, Multivitamin Therapy, No Phenol Diet, Omega-3 Fatty Acids, Rice Milk, Vitamin Therapy, Zeolite Therapy, etc. About 50 subtypes sensory therapies were discussed including, Animal Assisted Therapies, Aroma Therapy, Arts Based Therapy, Auditory Integration Therapy, Brushing Therapy, Craniosacral Therapy, Dance Therapy, Deep pressure, Deep Sound Therapy, Drum Therapy, Group Touching Therapy, Holding Therapy, Hydrotherapy, Laughter Therapy, Movement Therapy, Music Therapy, Phototherapy, Rolfing, Swimming Therapy, Therapeutic Massage, etc. Several education-behavior based therapies (N: 40) considered in the group discussions included Applied Behavior Analysis (ABA), Assertiveness Training, Biofeedback, Cognitive Behavior

JAIISH, Vol. 27, 2008

Aetiological & Therapeutic Searches by Netizen Parents on Autism Spectrum

sensory therapies (N: 199; 11.73 %), alternate medical systems (N: 189; 11.14 %), medication/drug therapy (N: 133; 7.84 %), and instrument based therapies (N: 93; 5.48 %) respectively.

Handicapped Infants. The Journal of Special Education. 22.1.117-127. Bailey, D.B., Blasco, P.M., and Simeonsson, R.J. (1992). Needs expressed by mothers and fathers of young children with handicaps. American Journal of Mental Retardation. 97. 1-10. Berelson, B. (1971). Content Analysis in Communication Research. Glencoe: Free Press. Garshelis, J.A.., and Mc Connell, S.R. (1993). Comparison of Family Needs Assessed by mothers, individual professionals and interdisciplinary teams. Journal of Early Intervention. 17. 1. 36-49. Gowen, J.P., Christy, D.S., and Sparling, J. (1993). Information needs of parents of young children with special needs. Journal of Early Intervention. 17. 2. 194-210. Holsti, O.R. (1969). Content Analysis for the Social Sciences and Humanities. New York: Reading Mass. Krippendorff, K. (2004). Content Analysis: An Introduction to its Methodology. Second Edition. Thousand Oaks: Sage Publications. Peshawaria, R., Menon, D.K., Ganguly, R., Roy, S., Pillay, R.P.R.S., and Gupta, A. (1995). Understanding Indian Families Having Persons with Mental Retardation. Secunderabad: National Institute for the Mentally Handicapped.

Conclusions
In sum, the present study opens the internet, web world and virtual reality as an increasingly new or exciting vista/dimension for investigation of social group dynamics in the field of disability rehabilitation. Parents, caregivers and advocates for the cause of children with special needs do not merely exist in the real world alone. There is a growing population of computer savvy service providers as well as receivers with an equally growing redoubtable knowledge base which seeks to answer uncensored anything or everything one wanted to know from the virtual world. However, considering the expanse of themes and discussions on autism by the members in the studied virtual group, it is seen that there is likelihood of one to get lost in the quagmire of information overload unless one is discerning in the quest for understanding the continuing challenge and enigma called autism.

References
Babbie, E. (2003). The Practice of Social Research. Tenth Edition. Wadsworth: Thomson Learning. Bailey, D.B., and Simeonsson, R.J. (1988). Assessing the Needs of Families with

95

JAIISH, Vol. 27, 2008

Auditory Memory and Sequencing in Children

Auditory Memory and Sequencing in Children Aged 6 to 12 Years


1

Devi N., 2Sujita Nair & 3Asha Yathiraj

Abstract
Objectives: The study aimed to investigate auditory memory and sequencing ability in typically developing children. The study also compared the performance of children with suspected auditory memory problems with that of typically developing children. Method: Using the Auditory Sequencing Test developed by Yathiraj and Mascarenhas (2003), auditory memory and sequencing abilities were checked on ninety-six typically developing children in the age range of six to twelve years. Ten children with learning disability with suspected auditory memory problems were also tested. Results: The results indicated that auditory memory scores increases with advance in age up to ten years in the normal children, after which a plateau was obtained. There was no significant difference across gender. Auditory sequencing ability was also found to improve with increase in age up to seven years, after which a plateau was attained. A significant difference was obtained for auditory sequencing abilities across gender in two age groups, with the males out performing the females. Also the scores on the ten children with suspected auditory memory problems, was compared with the obtained data. Conclusions: The results revealed that the majority of children with learning disability, who had indications of memory problems, were identified as having auditory memory deficits. Hence, the obtained data on typically developing children can be used to confirm the presence of auditory memory deficit. It could also be used to determine the efficacy of management procedures on children with an auditory memory deficit.
Key words: Auditory memory, Auditory sequencing, Auditory processing disorder, Learning disability.

A (central) auditory processing disorder [(C)APD] is defined as a deficit in the processing of information that is specific to the auditory modality, that may be exacerbated in unfavorable acoustic environments and that may be associated with difficulties in listening, speech understanding, language development and learning (Jerger & Musiek, 2000). The underlying conceptual and philosophical approach one has regarding auditory processing disorders will determine the testing procedures used for evaluation. The testing procedure can be focused specifically on the auditory processing disorder without the contamination of language, memory, and attention. It can be nonlinguistic stimuli, psychophysical methodology and / or electrophysiological methods used for revaluation. On the other hand, the difficulties experienced in everyday life situations involve various cognitive

processes that are intimately intervened to assess memory, attention and decoding (ASHA Task force on Central Auditory Processing consensus development, 1996; Jerger & Musiek, 2000). (C)APD has been defined as a deficit in the neural processing of auditory stimuli that is not due to higher-order language, cognitive or related factors (ASHA, 2005). The quality of ones memory has traditionally been characterized in terms of the quantity of ideas or the number of aspects of events that are recalled (Rhodes & Kelley, 2005). Chermak and Musiek, 1997 have cited studies providing information regarding memory in children. These studies indicate different aspects of the development of memory in children. They report of a study by Howe and Ceci (1978), which indicated that children gradually acquire knowledge and appreciation of retrieval cues and effective strategies for coding, organizing and retrieving items in memory. In 1979, Howe and

1 Lecturer, Dept. of Audiology, All India Institute of Speech and Hearing, Manasagangothri. Mysore-570006, email:devi_aiish@yahoo.co.in, 2Audiologist, email:sujita_n@hotmail.com, 3Professor, Dept. of Audiology, All India Institute of Speech and Hearing, Manasagangothri, Mysore-570006

96

JAIISH, Vol. 27, 2008

Auditory Memory and Sequencing in Children

Celci reported that by age 6 years, most children demonstrate some awareness of the limitations of memory and the factors affecting memory. By age 8 to 10 years the children were found to demonstrate a planned approach for encoding and retrieval, becoming aware of mnemonics and their benefits. Locke (1968) has suggested that a discrimination impairment seen in those with an auditory processing disorder may be a byproduct of or coexist with an auditory memory deficit. Weisner, Tomblin, Zhang, Buckwalter, Chynoweth and Jones (2000) have noted that auditory memory deficits, seen in children with learning disability, can be attributed to phonological loop impairment. This in turn plays an important role in the acquisition of vocabulary. According to Cusimano (2001), students with auditory memory deficiencies often experience difficulty in developing a good understanding of words, remembering terms and information that have been presented orally. They also experience difficulty in processing and recalling information that they have read to themselves. Widely used measures of auditory memory span involve the use of digits, words, sentences, nonsense syllables, paragraphs and stories which are to be recalled following a single presentation, when the number of stimuli presented is increased. The examiner tests the number of elements the subject is able to retain and retrieve (Underwood, 1964). However, there is a limit to the maximum number of items that can be successfully remembered in this way. An individuals auditory memory span is about 6 or 7 items (Roediger, Knight and Kantouwitz, 1977; Jarold, Baddeley, Heves, Leeke & Philips, 2004). Cusimano (2001) opined that it is important to understand that each aspect of auditory memory is specific unto itself. While one area of the brain involves the intake of a series of unrelated letters, another involves numbers, another word and there are others that involve a contextual series of words, sentences, and whole passages. Hence, students need to be tested to determine if they can recall the number of items in a series proficiently for their age. Howe (1965) reported that if recall is requested as soon as presentation of a list of items is completed, the items that occur at the beginning of the list are generally found to have become
97

more highly consolidated in memory than the items that occurred later. Memory for the early items in a list is more resistant than that for later items to the disrupting effects of various activities. According to Jarold et al. (2004) it also depends on the nature of the to-be-remembered stimuli. Auditory memory spans are smaller for words, which sound alike or are phonologically similar (example cat, bat and hat) than words, that are phonologically dissimilar. In addition, spans are shorter for multi-syllabic words, that are longer in duration (example Helicopter and police man) than for monosyllabic words, that are shorter (example pig and shoe). Owing to the fact that memory plays an important role in spoken language processing and learning, strengthening memory may benefit individuals with a learning disability. In order to detect the presence of auditory memory problems, it is essential to evaluate children with an auditory memory and sequencing test. The test should have age appropriate norms to make accurate diagnosis and suggest the necessary rehabilitation strategy. Absence of data in Indian children instigated the present study. The study aims at obtaining age appropriate data on auditory memory and sequencing in typically developing children in the age range of 6-12 years in both boys and girls. In addition, it also aims at determining whether children with suspected auditory memory problems can be identified based on the data obtained on normal children.

Method
The participants involved in the study comprised of two groups. The study was carriedout initially on 96 typically developing children who were reported to have no academic difficulties as reported by their teachers. Later 10 children with a known history of learning disability were also evaluated to check the utility of the test in determining auditory memory problems. The 96 normal children were in the age range of 6-12 years. The Screening Checklist for Auditory Processing (SCAP) developed by Yathiraj and Mascarenhas (2003) was administered to rule out any auditory processing disorder. These participants were divided into 6 age groups having 16 children in each group. The age groups were 6 years 6;11 years, 7 years 7;11 years, 8 years 8;11 years, 9 years 9;11 years, 10 years 10;11 years, and 11 years 11;11

JAIISH, Vol. 27, 2008

Auditory Memory and Sequencing in Children

years. Of the 16 children in each group, 8 were boys and 8 were girls. These children were taken from primary and middle schools in Mysore city. Children who passed the checklist and met the following criteria were selected: Had English as a medium of instruction for at least one year and were familiar with the language, Had normal IQ based on Kaufmann assessment battery for children, Had no history of hearing and speech problems, No history of otological or neurological problems, Hearing sensitivity within normal limits (i.e. air conduction threshold of less than or equal to 15 dB HL in the frequency range of 250 to 8 kHz in both ears and air bone gap of less than 10 dB HL at any frequency), No report of speech identification problems and, No illness on the day of testing. The second group of participants also met the same participant selection criteria as the first group except that they failed the Early Reading Skills Test (Rae & Pother, 1973) indicating that they had learning disability. These children also obtained less than 50% scores on the Screening Checklist for Auditory Processing (Yathiraj & Mascarenhas, 2003) necessitating further (C)APD evaluation. Further, the children were included in the study only if they failed on at least one item in the SCAP that indicated the possible presence of a memory problem. Procedure The Auditory Sequencing Test developed by Yathiraj and Mascarenhas (2003) was used as the test material. In this Indian-English test, the length of the word sequence increased from a three-word sequence to an eight-word sequence. Each sequence group was referred to as a token. There were 2 tokens in the 3 and 4 word sequences and 4 token each in all the other sequence (i.e. 5, 6, 7 & 8). The interval between words in each sequence was 500 msec, while the interval between tokens (i.e. between groups of words) was 10 seconds. The testing was done in a quiet room that was free from distraction. The signals were presented

at a comfortable level through a CD player (PHILIPS AZ2160). Each child was tested individually. The participants were seated one meter away from the player at a zero degree azimuth. Each child was instructed to listen to the group of words and repeat them in the correct order. The responses were recorded on a scoring sheet. The children from both groups were tested in a similar manner. A score of one was awarded for every correct word that was recalled. An additional score of one was awarded if the words were recalled in the correct sequence. The maximum attainable score was 104 for the auditory memory subtest. Likewise a similar score was attainable for the auditory sequencing subtest. The raw scores obtained for the auditory memory and the sequencing subtests were tabulated across different age groups and gender. Descriptive statistics was done to find out the mean and standard deviation. ANOVA and Duncans post hoc test were carried out to find out the significance of difference between the scores.

Results and Discussion


The mean and standard deviation values of the auditory memory subtest are depicted in Table 1 and that of the sequencing subtest are shown in Table 2. This information is given for males and females, across the different age groups. Tables 1, 2 reveal that with increase in age, generally there was a steady increase in auditory memory and auditory sequencing abilities. This increase was more for the auditory memory than for the auditory sequencing subtest. A similar trend was seen in the males as well as females.
Age Male Female (in years) Mean* SD CI Mean* SD CI 66;11 42.12 5.59 37-47 54.37 7.24 48-61 77;11 57.37 10.64 48-67 60.00 7.83 53-67 88;11 61.62 4.95 57-66 58.12 6.72 52-64 99;11 68.37 7.68 61-75 57.37 10.35 48-67 1010;11 74.00 8.33 67-81 61.87 5.43 57-67 1111;11 73.00 4.37 69-77 70.37 10.37 61-81 Table 1: Mean Scores, Standard Deviation (SD) and Confidence Interval (CI) of the Auditory Memory Subtest across gender and age. * Maximum score = 104

98

JAIISH, Vol. 27, 2008

Auditory Memory and Sequencing in Children

Age (in years)

Mean*

Male SD

CI

Mean*

Female SD

CI

6 6;11 7 7;11 8 8;11 9 9;11 10 10;11 11 11;11

22.12 32.75 33.37 34.25 43.37 39.00

7.98 9.31 10.47 9.96 16.93 5.07

15-29 24-41 24-43 25-43 29-56 34-44

25.12 10.13 16-34 27.25 7.75 20-34 23.62 3.50 20-27 26.87 8.21 20-34 27.37 5.26 22-32 33.75 12.05 23-45

Age in 6Years 6;11 6-6;11 7-7;11 NSD 8-8;11 NSD 9-9;11 NSD 10-10;11 SD 11-11;11 SD

77;11

88;11

99;11

1010;11

1111;11

NSD NSD NSD NSD NSD NSD NSD NSD NSD

NSD

Table 2: Mean scores, Standard Deviation (SD). and Confidence Interval (CI) of the Auditory Sequencing Subtest across age and gender. * Maximum score = 104

Note: SD = significantly different, NSD = not significantly different.

Table 4: Significance of difference between means for the auditory sequencing subtest cross ages.

The results of the one-way ANOVA test indicated that the auditory memory scores were highly significant across the age groups [F (6, 96) = 14.071, p < 0.001] but it was not significantly different across gender [F (6, 96) = 1.078, p > 0.05]. However, auditory sequencing scores were found to be significantly different across ages [F (6, 96) = 3.316, p < 0.01] as well as gender [F (6, 96) = 10.32, p < 0.01]. Since there was a significant difference, the Duncans post hoc test was used. The results of the post hoc test on the auditory memory scores and sequencing scores, across age are given in Tables 3 and 4 respectively.
Age in Years 6-6;11 7-7;11 8-8;11 9-9;11 10-10;11 11-11;11 Note: SD = 66;11 77;11 88;11 99;11 1010;11 1111; 11

In the auditory sequence subtest, the older two age groups (10 years 10; 11 years and 11 years 11; 11 years) differed significantly from the youngest age group (6 years 6; 11 years). There was no significant difference between the other age groups (Table 4). However, there was a steady non significant increase in scores with advance in age, both in the male as well as female participants, as can be seen in Table 2. The ANOVA test revealed that there was no significant difference across gender for the auditory memory subtest while it was present for auditory sequencing subtest. Further analysis of gender difference for the auditory sequencing for different age groups was done using Duncans post hoc test. The results revealed that this significant difference was present only in two age groups (8 years 8; 11 years and 10 years 10; 11 years). No significant difference was observed for the other age groups. The gender difference probably occurred due to individual variability. The males in these two age groups had a higher standard deviation and confidence interval when compared to the females in the same age groups. Such variability was not seen for the other age groups. This could account for the gender difference in these two age groups. Comparison between the scores of the two subtests, auditory memory and sequencing, for different ages showed that there was a significant difference between them. In general it was noted that the auditory memory subtest resulted in the children having higher scores when compared to auditory sequencing subtest. Hence, it is recommended that both the subtests be administered and scored separately while evaluating children. It is highly possible that the processing of auditory sequences takes place in one area of the brain while that of auditory memory taps another area. This could account for difference in scores obtained in the two subtests. This is similar to the
99

SD SD NSD SD NSD NSD SD SD NSD NSD SD SD SD SD NSD significantly different, NSD = not significantly different.

Table 3: Significance of difference between means for the auditory memory subtest across ages.

From Table 3 it is evident that for the auditory memory subtest, the 6 years olds had significantly different scores when compared to all the older age groups. Likewise, the oldest age group (11 years 11; 11 years) differed significantly from the younger age groups. In general, the older groups did not differ significantly from the adjacent age groups but did so from those who were one to two years younger or older than them. It was generally noted that as the children grew older their auditory memory scores improved (Table 1). This improvement was seen till age ten after which there was a plateau in the responses.

JAIISH, Vol. 27, 2008

Auditory Memory and Sequencing in Children

findings of Cusimano (2001) who noted that different parts of the brain are responsible for processing different aspects of memory. The scores obtained by the ten children with learning disability, who were suspected to have an auditory processing problem were compared with the age appropriate scores obtained. This was done for the auditory memory as well as auditory sequencing subtest. The scores obtained by the ten children with learning disability are given in Table 5, Figures 1 and 2.
Case Age /Sex 7/F 8;6/F 9/M 10/F 11/F 12/M 12/M 12/M 12/F 12/M Auditory Memory Scores 35 53 74 27 56 24 36 61 34 65 Auditory Sequencing Scores 13 25 43 18 17 15 18 24 15 26 Interpretation*

The result indicates that eight of the ten children had deviant scores in auditory memory and sequencing. Thus, it can be inferred that the majority children with learning disability who have indications of a memory problems based on *Based on the confidence interval given in Tables 1 and 2. the SCAP, do have an auditory memory and Table 5: Scores of the auditory memory and sequencing problem. However, not all of them sequencing subtests obtained by the have such a problem. Based on this finding, it is children with learning disability. suggested that children with a learning disability should be screened using the SCAP or any other (C)APD screening checklist. Those showing an indication of a memory difficulty should be assessed for the presence of an auditory memory or auditory sequencing problem. Appropriate remedial steps should be provided for those who are found to have deviant scores.

1 2 3 4 5 6 7 8 9 10

Deviant Not Deviant Not Deviant Deviant Deviant Deviant Deviant Deviant Deviant Deviant

Figure 2: Comparison of Auditory Sequencing scores of children with Learning Disability with age appropriate norms.

Conclusions
The present study has provided data on auditory memory and sequencing for typically developing children in the age range of 6-12 years. The findings indicated that with an increase in age, the children showed an increase in auditory memory and sequencing abilities. The increase was more significant for the auditory memory subtest when compared to the auditory sequencing subtest. No significant difference across gender was observed for the auditory memory subtest. However, there was a significant gender difference for the sequencing subtest in two of the age groups. This difference was probably on account of the large variability in scores that the males had in these age groups.

Figure 1: Comparison of Auditory Memory scores of children with Learning Disability with age appropriate norms.

100

JAIISH, Vol. 27, 2008

Auditory Memory and Sequencing in Children

It was also found that the test was useful in determining whether children with suspected auditory memory problems as determined through a screening checklist do have an auditory memory / sequencing problem. The test results can be used to make suggestions for remedial help for children having deviant scores. Thus, the test can be used for diagnosis of auditory memory / sequencing problems in children with suspected auditory memory problems. In addition, it can be used to determine the utility of management techniques in children with auditory processing problems.

of Experimental child Psychology, 26, 230 245. Jarold. C., Baddeley. A. D., Hewes. A. K., Leeke. T. C., & Philips. C. E. (2004). What links verbal short term memory performance and vocabulary level? Evidence of changing relationships among individuals with learning disability. Journal of memory and language, 50, 134 148. Jerger, J., & Musiek, F., (2000). Report of the Consensus Conference on the Diagnosis of Auditory Processing Disorders in SchoolAged Children. Journal of American Academy of Audiology, 11(9), 467-474. Locke, J., (1968). Discrimination learning in childrens acquisition of phonology. Journal of Speech and Hearing Research, 11, 428434. Rae, G., & Pother, T. C., (1973). Informal Reading Diagnosis: A Practical Guide for the Classroom Teacher. New Jersey: Prentice Hall, Inc. Rhodes M. G. & Kelley. C. M., (2005). Executive processes, memory accuracy and memory monitoring: an aging and individual difference analysis. Journal of memory and language, 52, 578 594. Roediger H. L. III, Knight J. L., & Kantowitz B. H., (1977). Inferring decay in short term memory: The issue of capacity. Memory and cognition, 5, 167 176. Underwood B.J., (1964). The representativeness of rote verbal learning. In A. W. Melton (Ed.): Categories of Human Learning. New York: Academic Press. Yathiraj, A., & Mascarenhas, K., (2003). Effect of Auditory stimulation in Central Auditory Processing in Children with central Auditory Processing Disorder. A project funded by the AIISH research fund, AIISH, Mysore.

References
ASHA Task force on Central Auditory Processing consensus development, (1996). Central Auditory processing: current status of research and implications for clinical practice. American Journal of Audiology, 5(2), 41 54. American Speech Language Hearing Association, (2005). Roles of speech-language pathologists in the identification, diagnosis, and treatment of individuals with cognitivecommunication disorders: Position statement. ASHA Supplement, 25. Chermak, G. D., & Musiek, F. E., (1997). Central Auditory Processing Disorder: New Perspective. California: Singular Publishing. Cusimano, A., (2001): Auditory Memory, Learning Disability: There is a Cure. Pennsylvania: Achieve Publications. Howe M. J. A., (1965). Learning and Human Memory. In Howe, M.J.A., (Ed.) The Psychology of Human Learning. New York: Harper & Rou. Howe M. J., Ceci S. J., (1978). Semantic knowledge as a determinant of developmental differences in recall. Journal

101

JAIISH, Vol. 27, 2008

Regenerative Myringoplasty

Regenerative Myringoplasty A Case Report


1

Rajeshwari G. & 2Sundara Raju H.

Abstract
Dry perforation of the tympanic membrane cause hearing impairment and predispose the middle ear to infections. Small dry perforations can be closed by chemical cauterization using silver nitrate in an office setting. In a chemically induced myringoplasty all the five layers of tympanic membrane would be present akin to the normal tympanic membrane. Its an effective means of tympanic membrane closure. In selected patients it restores hearing back to normal and allaying the patient of a irritable symptoms like tinnitus. It reduces the surgical waiting list. It also saves time and money for the patient, surgeon and hospital. We present a case report of 35 year old male diagnosed to have permanent perforation syndrome he was managed successfully in an office setting with complete closure of the perforation restoring his hearing to normal.
Key words: Regenerative myringoplasty, Induction myringoplasty, Silver nitrate induced myringoplasty.

Perforations of tympanic membrane secondary to trauma or otitis media is an indication for surgical repair of the tympanic membrane. Surgical repair is a rule than exception. Dry perforations can be successfully closed by chemical means in an office setting in appropriately selected cases. Such a conservative managements saves the patient the risks of surgery

anesthesia and hospitalization.

Method and Results


A 35 year old male presented with history of otohorrea, decreased perception of hearing and tinnitus of left ear of six month duration. The otohorrea responded to oral antibiotics and aural drops. The decreased perception of hearing and tinnitus persisted. Microscopic examination of his

Perforation of the tympanic membrane

Table: 1

Normal tympanic membrane

1 Reader, Dept. of ENT, All India Institute of Speech and Hearing, Manasagangothri, Mysore-570006, email:drrajg@yahoo.com, 2 Reader, Dept. of Otorhinolaryngology, All India Institute of Speech and Hearing, Manasagangothri, Mysore-570006

102

JAIISH, Vol. 27, 2008

Regenerative Myringoplasty

left ear revealed a large central perforation measuring about six millimeters in diameter thats about 3/5th of the surface area of the pars tensa. No focus of infection was present in nasal cavity, throat or paranasal sinuses. Tuning fork test revealed conductive pathology on the left side. The facial nerve, mastoid and vestibular system was normal. Rest of otorhinolarynogological examination, systemic examination and general examination of the patient was normal. Patient underwent an audiological evaluation, pure tone thresholds showed. Moderate hearing losses of conductive type (Table 1) tympanogram shows B wave on the left side, absent a acoustic reflexes on the left side, the middle ear pressure and static compliance-no peak attainable. However, the speech discrimination scores was 95% on the left side. Routine Haemotological investigations were within normal limits and radiological examination of the mastoids showed well pneumatised mastoids. A diagnosis of chronic otitis media safe variety was made. In the second visit in a office setting 4% topical xylocaine solution about 2 ml was used as local anesthetic for the tympanic membrane. Injection atropine was the only pre anesthetic medication given. Ensuring ear was dry the rim of the perforation was cauterized with a bead of silver nitrate solution (2%). A paper patch soaked in anti biotic solution was used to cover the tympanic membrane perforation. The ear canal was packed with anti biotic soaked merocel. The patient was kept on a follow up at fortnightly intervals to observe the regeneration of the tympanic membrane. By 4 weeks microscopic examination showed closure of tympanic membrane perforation. After six weeks patient was sent for a repeat audiological evaluation. Pure tone thresholds had returned to normal with closure of air bone gap, the tympanogram showed A type of curves, acoustic reflexes had returned to normal, patient confirmed subjective improvement in hearing to almost normal and tinnitus had disappeared.

Table 3

Discussion
Tympanic membrane perforations are of special interest. Since its persistence can lead to variety of problems like permanent perforation syndrome, cholesteatoma tubo tympanic disease, extra cranial complications of otitus media, intra cranial complication of otitis media. The pars tensa has five distinct layers, an outer most epidermal layer, a thin dermis of fibrous tissue, an outer radiate fibrous layer, an inner circular fibrous layer and a mucousal layer. A perforation closure induced by a repeated acid cautery of its rim often results in a perfectly normal appearing, tympanic membrane possessing all the five layers. Whereas, spontaneous closure is only 2 layered devoid of fibrous layer giving rise of thin tympanic membrane. 3 guiding principles promote healing of tympanic membrane perforation by chemical cauterization. i. The edges of tympanic membrane perforation lined by statifed squamous epithelium which prevent spontaneous closure of perforation must be destroyed to permit fibro blastic proliferation of the fibrous layer. ii. The rim of the perforation must be kept moist because drying would caused death of the young fibro blast. iii. The edges of the perforation must be bleeding since hyperemia induces fibro blastic proliferation.

103

JAIISH, Vol. 27, 2008

Regenerative Myringoplasty

Regenerative myringoplasty must be attempted in a perforation not involving more than 65% of the surface area of the pars tensa. Derlacki, who has reported the largest series of chemical cauterization of tympanic membrane perforation about 75% of 131 perforations recovered. Repeated cauterization may be required in few cases. Induction myringoplasty may not be recommended in the following conditions. 1. Large perforation involving more than 65% of the surface area of the pars tensa. 2. Narrow external auditory canal preventing view of the anterior edge of the perforation. 3. Patient who refused series of weekly treatment. 4. When ingrowth of epidermis is suspected of forming an incipient or active cholesteatoma.

membrane and this would result in normal hearing post procedure. Since its an office procedure done under local anesthesia, the patient is saved from the risk of general anesthesia hospital stay, hospital cross infections and cost of the hospitalization. For the surgeon it reduces the surgical waiting list.

References
Derlacki, E.C. (1958). Repair of Central Perforation of Tympanic Membrane, Arch Otolaryngol, 58, 405-420. Goldman, N.C. (2007). Chemical Closure of Chronic Tympanic Membrane Perforations. ANZ Journal Surgery, 77(10), 850 851. Glasscock, M. E., Shambaugh, G. E. (1990). Surgery of the Ear, 4th. Edition. Canada :BC Decker. Wolferman, A. (1970). Reconstructive surgery of the middle ear. New York: Grune & Stratton. Scott- Brown, W.G. (1996). Scott Brown: Otology 6th Edition.Oxford: ButterworthHeinemann.

Conclusions
For the central perforation of the tympanic membrane measuring about 2 - 6 mm in diameter without a middle ear/mastoid disease, inductive myringoplasty can be attempted. The healing which is promoted by such induction would result in formation of all the layers of tympanic

104

JAIISH, Vol. 27, 2008

Universal Hearing Screening

Guidelines to Establish a Hospital Based Neonatal Hearing Screening Program in the Indian Setting
1

Ramesh A., 2Nagapoornima M., 3Srilakshmi V., 4Dominic M. & 5Swarnarekha

Abstract
The challenges to implement universal neonatal hearing screening (UNHS) in India are limited funding, manpower shortages, inadequate support services, low public awareness and uncertainty regarding commitment from health care practitioners. Nevertheless there are isolated groups in India who have been implementing UNHS. St. Johns Medical College Hospital, Bangalore has been implementing UNHS since September 2002. Till date we have screened 5100 neonates. In this article we detail the steps we followed to establish the program and make it a standard of care in our hospital. A qualitative design is used to describe every stage. A team consisting of faculty from Neonatology, Audiology and speech Pathology, Otorhinolaryngology, Child psychology and Medico-Social work constituted the screening team. Due to a high birth rate the strategy we followed was as follows. All infants not at risk were screened by behavioural audiometry using 60 and 70 dB warbled tones. The infants at risk were screened by a two stage otoacoustic emissions (OAE) screening as well as behavioural audiometry. As a safeguard against false negatives of our strategy all the infants irrespective of the results were given a language and hearing milestone chart for parents to report if the age appropriate milestone was absent. Screening in the NICU caused a lot of referrals so it is best avoided . If a sound proof room is available adjacent to the NICU it would be ideal. In our experience a pilot program should be run for at least one year to get an idea of the inputs required as well as aid in formulating a screening strategy. At regular intervals an audit should be conducted and the data published in indexed journals so that others who want to start UNHSP can benefit. Also a national dataset should be collated based on these data to guide institutions to initiate similar programs so that not a single hearing impaired child in this country is neglected.
Key words: Neonatal hearing, Screening program, Behavioral audiometry.

The 2007 statement of Joint committee on Infant Hearing (JCIH, 2007), American Academy of Pediatrics states that every state in the United States of America has a Universal Neonatal Hearing Screening Program (UNHSP) and 95 % of the new borns are screened before discharge from the hospital.(JCIH , 2007).In India we have a long way to go before we can reach this goal. Limited funding, manpower shortages, inadequate support services, low public awareness and uncertainty regarding commitment from health care practitioners are the challenges in a developing nation like India.(Olusanya BO et al, 2004).Nevertheless there are isolated groups in
1

India who have been implementing UNHS. (Nagapoornima et al 2007 ; Malik M et al 2007 , Mathur NN et al , 2007). St. Johns Medical College Hospital, Bangalore has been implementing UNHS since September 2002. Till date we have screened 5100 neonates. We have published our work in a Pub Med/Medline indexed journal and cited an incidence of 5.6 hearing impaired neonates per 1000 screened.(Nagapoornima et al 2007).This article details the steps we followed to establish the program. Also discussed is the logistics of running the program so that it is an established standard of care in our hospital. This information can be used by other hospitals to adapt these guidelines and

Associate Professor, Dept. of Otolaryngology Head and Neck surgery, St Johns Medical College Hospital , Bangalore560034, email:lavirams @ yahoo.com, 2Audiologist and Speech Pathologist, St Johns Medical College Hospital , Bangalore-560034, 3Audiologist and Speech Pathologist, St Johns Medical College Hospital , Bangalore-560034, 4 Professor, Community medicine, St Johns Medical College Hospital , Bangalore-560034, 5Professor, Pediatrics, St Johns Medical College Hospital , Bangalore-560034.

105

JAIISH, Vol. 27, 2008

Universal Hearing Screening

implement UNHS. The UNHSP at our hospital commenced on September 1, 2002. It has been functional for the last 6 years. The description of the work has been examined in four major steps. 1. 2. 3. 4. Establishing a case for UNHSP. Building a UNHS team. Initial pilot run for 2 years. Establishing a full fledged UNHSP.

would be practical to run a pilot program for 2 years using a hired Otoacoustic emission (OAE) screener. Accordingly a proposal was drawn up and placed before the management for approval. On July, 2002 the executive council of the institute granted permission to start UNHS using a hired OAE screener from SRC Institute of Speech and Hearing, Bangalore. Building a team On August, 2002 a meeting of the faculty members who volunteered to participate in UNHSP was convened. The following speciality members formed the team. Neonatology, Audiology and Speech Pathology, Otorhinolaryngology, Child Psychology and Medico-Social work. At the end of the meeting each team members job description was defined. Neonatologist Counsel parents for hearing screening and take consent. Otolaryngologist Examines the external auditory canal for any abnormalities that may cause a false positive result on OAE Screening. Audiologist and Speech Pathologist: Performs OAE screening and Auditory Brainstem Response (ABR) as per the protocol shown in figure 1.

A qualitative design is used to describe every stage. We start every stage with description of the technique followed by us and the limitations of the technique. The results are discussed in the light of the JCIH, AAP, 2007 statement which comprehensively summarises the state of UNHSP around the world. We have concluded by putting forth guidelines which could be adopted by various hospital based institutions. Creating a case for UNHS On January, 2002 a seminar was organized at our hospital to review the state of pediatric audiology in India. Faculty from the specialities of Pediatrics, Audiology and Speech Pathology, Otolaryngology, Child Psychology, Community medicine and Medicosocial work participated in the seminar. At the end of deliberations it was felt that the incidence of hearing impairment from

Figure 1: Protocol used to screen for hearing impairment in neonates at St Johns Medical College, Bangalore

Indian and western literature warrants commencement of UNHS in our hospital. One faculty member from each speciality volunteered to participate in the program if it was established. There was a consensus to write up a proposal for submission to the hospital management. We felt it
106

Follows up the screening passed, not at risk infants at 1 year of age and at risk infant at 6 months and 1 year of age. Receptive expressive emergent language scale (REELS) was the tool used to assess age appropriate language development.

JAIISH, Vol. 27, 2008

Universal Hearing Screening

Arranges for appropriate rehabilitation of hearing impaired infants. Child Psychologist For grief counseling and preparing the parents to be active partners if rehabilitation is required. Medicosocial worker Ensuring adequate follow up of the infants and contact infants who miss the appointment. It is crucial to ensure that each member chose to be part of the team. This made managing the work responsibilities easy. There was no extra incentive for the additional work that was taken up by each member.

At the end of the pilot program for 2 years, the data was presented to the management team consisting of the hospital administrator and heads of the departments involved in the program. The issue of purchasing an OAE screener for the hospital was discussed at the executive council meeting. Partial funding was promised by Christoffel Blinden Mission (CBM), a charitable funding agency. On July 2004, the administration approved purchase of ILO USB I, OAE analyzer. From January 2005, we commenced universal hearing of all infants seeking care at our hospital. Implementation of universal neonatal hearing screening The logistics of the program in our hospital is as follows: The protocol for screening was altered because we could not do OAE testing of all the neonates. All not at risk infants were screened by behavioural audiometry using 60 and 70 dB warbled tones. Behavioural response index was used to standardize the response. The atrisk infants were screened by OAE as well as BOA. All the at risk infants who passed were followed up at 6 months and one year to examine for age appropriate hearing and language milestones. Infants who had disability were rehabilitated. Children with multiple disabilities were asked to follow up on Wednesdays where a multispeciality rehabilitation team provided a single window service for physical, mental, visual and hearing disability. This service called the Unit of Hope provides highly subsidized care for children with multiple disabilities. Hearing aids are procured directly from the company and provided at a much lower cost than maximum retail price (MRP). Every 6 months an internal audit was conducted by the person co-ordinating the program to evaluate the performance and compare it with JCIH 2000, recommended parameters. All the data was collected and published in a Pub Med indexed journal in 2007 for the benefit of other groups in India, wanting to start similar programs. Data recording Initially from 2002 to 2004 all the data was recorded in a proforma (Appendix 1) over a 2 year period. At the end of two years we realized

Establishing a pilot run The logistics of establishing a pilot program at our hospital were the following: Signing of a memorandum of agreement with SRC Institute of Speech and Hearing to hire their OAE screener and staff for 2 hours a week. A charge of Rs. 110 was levied for two OAE screens. St. Johns Hospital collected the amount and reimbursed to SRC Institute on a monthly basis. Neonates in the NICU who were at risk based on the JCIH, 2000 were screened. Screening was done in the NICU for one year. As the number of referrals for second screen was very high, due to unacceptable noise levels in the NICU, after one year we shifted the venue of screening to ENT OPD. The referrals for second OAE screen reduced considerably after this but we missed a lot of babies to follow up. We followed a screening protocol adapted from JCIH, 2000. Our strategy consisted of a two tiered OAE screen. The first screen was completed by 6 weeks of birth and the second screen by 3 weeks if the first screen failed. For a period of 6 months a hand held screener donated by Voita Institute, Germany was used to screen not at risk infants. The data recording details are described as a separate subheading.

107

JAIISH, Vol. 27, 2008

Universal Hearing Screening

that the records were occupying a large amount of space. To reduce this we made a register with all the details printed in columns. By this method we reduced the space requirement by 90 %. The alpha numeric data was coded and entered in another register. This was entered in MS Excel spread sheets in the computer. Statistical package for social sciencers version 15 (SPSS) was used to perform statistical analysis. Establishing a hospital based universal neonatal hearing screening program Lessons learnt from St. Johns program. The absence of any data on sensitivity and specificity of OAE data in the Indian setting was a major handicap in the initial stages of establishing our program. Studies from the European and American programs were used to formulate our initial strategies for screening. Our stumbling block was the large number of neonates to be screened due to a high delivery rate. The western UNHSP did not have this scenario. We altered our strategy of screening in the NICU to screening at first visit or before discharge. Also we screened not at risk neonates by behavioural audiometry to cope up with high delivery rate. To avoid missing out on false negative screens and delayed onset hearing loss, we adopted a strategy to follow up at 6 months and one year. Our learning regarding data management was that custom made registers were more easy to maintain and occupied lesser space compared to proforma sheets.Regarding follow up of screened neonates we discovered that about 60 % of our screened neonates were lost to follow up. This rate, especially in the at-risk failed neonates is alarming. Even in American screening programs loss to follow up is upto 50 % in some centers. We are in the process of improving on our follow up rates by employing a person exclusively to ensure follow up. Another measure we have incorporated is to give a follow up card to the mother which will outline the main milestones of hearing development. This we intend will aid in adequate follow up and detection of late onset hearing loss. Universal neonatal hearing screening initiatives in India Need to publish in indexed journals. UNHSP is in its embryo stage in India. A Pub Med search using the key words infant / neonatal hearing screening in India revealed only 11 articles. In 1990, a major initiative was undertaken

to review the state of pediatric audiology in India during a workshop conducted by All India Institute of Medical Sciences, New Delhi. The workshop highlighted the findings of a large scale multicenter survey conducted with the support of Indian Council of Medical Research (ICMR) as well as work of other researchers involved in pediatric audiology. Following this there were no published reports till 2002. Since 2002 there have been isolated small scale observational studies. In 2002 a national consensus building workshop was organized at All India Institute of Speech and Hearing, Mysore to frame guidelines to establish a national program for early detection and rehabilitation of hearing impairment. Following this there has been a report of normative data for TEOAE in Indian children. The year 2007, saw two reports of large scale screening exercises in indexed literature. The report from New Delhi discussed strategy of screening after 6 weeks to reduce referrals. A report from Bangalore, examined the incidence data in a large neonatal population. Though some groups in Chennai and Trivandrum and Mumbai have been involved in neonatal screening there is no published reports from these centres.All the groups involved in this work should publish their data so that a database can be created for India. This will assist future program to benefit from the experience of these groups.

Conclusions
We have put forth the following guidelines for tertiary care hospitals in the non governmental sector who wish to start UNHSP. 1. A team consisting of faculty from Neonatology, Audiology and speech Pathology, Otorhinolaryngology, Child psychology and Medico-Social work should be constituted. Each member should choose to contribute to the program. The job description of each member should be defined at the outset. 2. A pilot program for at least one year will give an idea of the inputs required from the management as well as help in formulating a screening strategy. 3. A two stage OAE screen with first screen at the first follow up visit to the hospital is more effective for at risk infants. Screening in the

108

JAIISH, Vol. 27, 2008

Universal Hearing Screening

NICU causes a lot of referrals so it is best avoided as a venue for screening. If a sound proof room is available adjacent to the NICU it would be ideal. 4. At the outset of the program a dataset should be formulated. A hard copy as well as soft copy of the data should be created and regularly updated. A register with columns for all the details and rows for each case is more space efficient compared to proformas. 5. At regular intervals an audit should be conducted and the data published in indexed journals so that others who want to start UNHSP can benefit. Also a national dataset can be collated based on these data.

Malik, M., Pradhan, S.K, Prasana, J.G, (2007). Screening for psychosocial development among infants in an urban slum of Delhi. Indian J Pediatr, 74 (9), 841 5. Mathur, N.N, Dhawan.R (2007). An alternative strategy for universal infant hearing screening in tertiary hospitals with a high delivery rate within a developing country using transient evoked otoacoustic emissions and brainstem evoked response audiometry. The Journal of Laryng and Otol, 121, 639 643. Nagapoornima,M., Ramesh,A., Srilakshmi,V., Suman Rao P. N, Patricia P L, Madhuri Gore, Dominic M, Swarnarekha (2007). Universal hearing screening . Indian Journal of Pediatrics, 74(6), 545 549. Olusanya BO, Luxon LM, Wirz SL ( 2004) Benefits and challenges of new born hearing screening for developing countries. Int J Pediatr Otolaryngol. 68(3), 287 305.

References
Joint committee on infant hearing screening (2007) Position statement. Pediatrics Vol 120 (4), 898 921.

109

JAIISH, Vol. 27, 2008

Universal Hearing Screening

Annexure 1 Dataset of St Johns Neonatal Hearing Screening Program


Name : Duration of NICU stay : Number Neonatal : Hospital : Address : Risk factors : 1.Hereditary hearing loss : 2.Ear deformities : 3.Hypoxic ischaemic encephalopathy : Y/N APGAR: 5 mts - 10 mts- Bag Mask Ventilation : Y/N Phone number : Sex : DOB : Birth weight : Gestation :

Endotracheal and ventilation : Y/N Furosemide :

4 .Duration of Medications - Genatmycin : Amikacin : other ototoxic medications : 5. Infection : Sepsis - Y/N Meningitis : Y/N 6.Hyperbilirubinemia: Peak level : 7. Pre term : Y/N Duration:

Phototherapy:Y/N ExchangeTransfusion:Y/N SFD/IUGR:Y/N

LBW : Y/N

Method used to screen OAE BOA ABR Maternal Problems Natal History Delivery : First screen : Second screen

Date

Result

: Home / Hospital : Normal/Breech/Caesarean/Forceps

Family History Type of family Consanguinous marriage Literacy of

: Nuclear/Joint : : Father : Mother

110

JAIISH, Vol. 27, 2008

Reversible Sudden Sensory Neural Hearing Loss

Reversible Sudden Sensory Neural Hearing Loss A Case Report


1

Sundara Raju H. & 2Rajeshwari G.

Abstract
Idiopathic sudden sensorineural hearing loss is a clinical diagnosis characterized by a sudden deafness of cochlear or retro cochlear origin in the absence of clear precipitating factor. Most often it is taught irreversible. If it is identified early and medical intervention is done it shows a good prognosis. Factors affecting prognosis include age, presence of vertigo, unilateral or bilateral pathology, associated systemic diseases and time duration between the onset of symptoms and treatment. The main stay of treatment includes corticosteroids, anti-inflammatory drugs, neurovitamins, antioxidants, carbogen therapy and adequate control of systemic illness. Sudden Sensori Neural Hearing Loss through rare it is one of the medical emergency in otologists practice. Intra tympanic injection of steroids is one of the common accepted methods of management. I am a reporting a case of sudden sensori hearing loss where oral steroid therapy is also equally effective in certain cases if the hearing loss is unilateral and if the patient is of younger age group and if the patient is without any systemic illness.
Key words: Sudden sensory neural hearing loss, Reversible sensory neural hearing loss, Calorie test, Steroids.

Idiopathic sudden sensorineural hearing loss (ISSHL), characterized by new-onset unilateral or bilateral hearing loss that develops rapidly within 24 to 72 hours, remains a diagnostic and therapeutic challenge for the clinician. The cause and pathogenesis of ISSHL remain unknown. Proposed theories of the pathogenesis of ISSHL include viral cochleitis (1), vascular occlusion (2), and membrane breaks. Definitions of sudden hearing loss have been based on severity, time course, audiometric criteria, and frequency spectrum of the loss. A commonly used criterion to qualify for the diagnosis is a sensorineural hearing loss of greater than 30 dB over 3 contiguous pure tone frequencies. The vast majority of cases of sudden hearing loss are unilateral, and the prognosis for some recovery of hearing is good. Usually sudden sensory neural hearing loss presents as unilateral loss of hearing, bilateral involvement is rare and simultaneous bilateral involvement is very rare. This case is reported to stress the importance of oral steroid therapy in certain cases of unilateral sensori neural hearing loss in younger age group without any systemic illness.
1 2

Methods and Results


A 35 year old male presented with a history of sudden hearing loss in the right side 2 days duration. He had associated symptoms of Tinnitus, Vomiting, Headache, Numbness in the right side of the face. Hearing loss sudden onset and static in nature Patient did not have any prior history of ear discharge, ear pain. Patient had tinnitus which was intermittent and low pitched. There is no history of acoustic trauma, hypertension, diabetes, drug intake for the past one week. No history of upper respiratory infection.

On examination patient is anxious about the problem of sudden hearing loss. Pulse 82/minute Blood pressure 130/86 mm of hg Systemic examination NAD ENT examination reveals both the tympanic membranes are normal with all the normal land marks.

Reader, Dept. of Otorhinolaryngology, All India Institute of Speech and Hearing, Manasagangothri, Mysore-570006, Reader, Dept. of ENT, All India Institute of Speech and Hearing, Manasagangothri, Mysore-570006, email: drrajg@yahoo.com.

111

JAIISH, Vol. 27, 2008

Reversible Sudden Sensory Neural Hearing Loss

Pre aural and post aural region normal. Mastoid region normal. Facial nerve intact and normal. A tuning fork test reveals right sided sensory neural hearing loss. Caloric test were normal. Hematological investigations - HB %, Blood total count, Differential count, RBS, Serum cretin, Blood urea, were within normal limits. x-ray of the both mastoids were well pneumatised and appears normal Audiological evaluations

PTA reveals moderate to severe sensory neural hearing loss right side with maximum affections in the mid frequencies SIS less than 60% Impedance was normal with a type tympanograph with absent of acoustic reflex(table 1) ABR normal way morphology with normal interpeak latency. Recruitment

Table 1

In office setting patient was treated on tablet Methyl Prednisolone 1 mg/kg body weight in 3 divided doses for a period of 2 weeks. Peripheral vasodilators like cinnarizine 25 mg was also given for 4 weeks. A non steroidal antiinflammtory was also given. Vitamine B complex and alongwith anti oxidents were prescribed. Patient was advised on low salt diet.

Neurological opinion reveals no neurological deficits. 2 weeks later patient felt 70% improvement in the symptoms of hearing loss and tinnitus. Methyl Prednisolone was given on tapered dose for another 2 weeks. All the other drugs were continued for 2 more weeks. After one month patient felt completely normal in his symptoms. Audiological evaluations were done again after one month and was revealed normal hearing. (Table 3)

112

JAIISH, Vol. 27, 2008

Reversible Sudden Sensory Neural Hearing Loss

Table 3

Discussion
Pathophysiology The postulated pathophysiology for idiopathic sudden sensory hearing loss (ISSHL) has 4 theoretical pathways. These are labyrinthine viral infection, labyrinthine vascular compromise, intracochlar membrane ruptures, and immunemediated inner ear disease. A disease process involving any of these theoretical possibilities could have sudden hearing loss as a symptom. Each theory may explain a fraction an episodes of sudden sensory hearing loss, but none of the existing theories individually could account for all episodes. Viral infection The evidence to implicate viral infection as one of sudden idiopathic sensory hearing loss is circumstantial. Studies of patients with ISSHL show a moderate prevalence of recent viral type illness. Sometimes, evidence of recent viral seroconversion inner ear histopathology consistent with viral infection is present.

Vascular compromise The cochlea is an end organ with respect to its blood supply, with no collateral vasculature. Cochlear function is exquisitely sensitive to changes in blood supply. Vascular compromise of the cochlea due to thrombosis, embolus, reduced blood flow, or vasospasm seems to be a likely etiology for ISSHL. The time course correlates well with a vascular event, a sudden or abrupt loss. A reduction in oxygenation of the cochlea is the likely consequence of alterations in cochlear blood flow. Alterations in perilymph oxygen tension have measured in response to changes in systemic blood pressure or intravascular carbon dioxide partial pressure (pCO2).

Conclusions
Young patients with no systemic illness and unilateral sensory neural hearing loss will have better prognosis. This case is reported because of the treatment protocol was taken up at office base

113

JAIISH, Vol. 27, 2008

Reversible Sudden Sensory Neural Hearing Loss

setup without any admission of the patient and without any administration of the injections. We know that parentral use of steroids is the established protocol for the sudden sensory neural hearing loss. In this study we are reporting a case of oral steroid therapy is also equally effective in certain cases if the hearing loss is unilateral, and if the patient is without any systemic illness and younger age group patients.

(report of the first one hundred patients).Ann Otol Rhinol Laryngol , 66, 963-980. Rasmussen, H. (1949).Sudden deafness. Acta Otolaryngol, 37, 65-70 Schuknecht, H.F., & Donovan, E.D (1986).The pathology of sudden deafness. Arch Otolaryngol, 243, 1-15 Yoon, T. H., Papirella, M.M., Schachern, P.A., & Alleva, M. (1990). Histopathology of sudden hearing loss. Laryngoscope, 100, 707-715. Pitkaranta, A., Vasama, J.P., & Julkunen, I. (1999). Sudden deafness and viral infections. Otorhinolaryngol Nova, 9, 190197. Tucei, D.L., Farmar, J.C., Kitch, R.D., & Witsell, D.L. (2002). Treatment of sudden sensorineural hearing loss with systemic steroids and Valacyclovir. Otol Neurotol, 23, 301-308. Scheinman, R.I., Cogswell, P.C., Lofquist, A.K., & Baldwin, A.S. (1995). Role of transcriptional activation of I kappa B alpha in mediation of immunosuppression by glucocorticoids, Science, 270, 283-286. Scheinman, R.I., Gualberto, A., Jewell, C.M., Cidlowski, J.A., & Baldwin, A.s.(1995). Characterization of Mechanisms involved in transrepression of NF kappa B by activated glucocorticoid receptors. Mol Cell Biol, 15, 943-953.

References
Wilson, W.R., Byl, F.M., & Laird, N.(1980). The efficacy of steroids in the treatment of idiopathic sudden hearing loss. A double blind clinical study. Arch Otolaryngol,12, 772-776. Moskowitz, D., Lee, K.J., & Smith, H.W. (1984). Steroid use in idiopathic sudden sensiorineural hearing loss. Laryngoscope, 94(5 Pt 1), 664-666. Kitajiri, S., Tabuchi, K., & Hiraumi, H. (2002). Is corticosteroid therapy effective for sudden onset sensorineural hearing loss at lower frequencies. Arch Otolaryngol Head NeckSurg , 128, 365-367. Chen, C.Y., Halpin, C., & Rauch, S.D.(2003). Oral steroid treatment of sudden onset sensorineural hearing loss: a ten year retrospective analysis. Otol Neurotol, 24, 728-733.

Van Dishoeck, H., & Bierman, T. (1957). Sudden perceptive deafness and viral infection

114

Journal of All India Institute of Speech & Hearing (JAIISH) "Instructions to Authors"
Submission of Manuscript: Authors should submit three hard copies of the manuscripts (as per the instructions given under "Submission of Manuscript") along with a soft copy on a CD to the Director/Editorial Chief, All India Institute of Speech and Hearing, Manasagangothri, Mysore 570006. Title page: Title page must be submitted as a separate page. The title page should include (a) complete manuscript title; (b) authors' full names, highest academic degree, and present affiliations; (c) name and address for correspondence, including fax number, telephone number, and e-mail address of the corresponding author. Abstract and key words: The abstract containing not more than 300 words must be submitted in a separate page. It should contain the following information: The objectives of the study; the method of the study; the main results of the study and; the conclusions of the study along with clinical implications or applicability. Three to five key words should be provided below the abstract. Authors should follow APA conventions with regard to use of units, symbols and abbreviations. Text: The text should be organized under the following main headings: Introduction, Method/s, Results, Discussion, Conclusions and References. Abbreviations should be expanded at the first mention in text. Manuscript Style: Manuscripts that do not adhere to the following instructions will be returned to the corresponding author for technical revision before undergoing peer review. The manuscript should be prepared based on the following guidelines: The matter should be in English. Microsoft Word is the required file format for the text of the manuscript. Set all margins to 1 inch All text, including abstracts, references, figure legends, etc., must be 1 line spaced. Use Times New Roman font at 12 point. Justify the text. . The material should be sent in A4 size paper. . Use tab key and centering functions to align headings, paragraph indents etc. Do not use space bar. . Headings must be used to designate the major divisions of the paper. . Number the pages consecutively in the upper right hand corner. . All spellings should be checked for UK spelling . Use IPA to represent material in languages other than English . The overall length of the paper should not exceed 20 A4 size pages. References: All the references cited in the text should be included in the reference list. Conversely, each entry in the reference list must be cited in text. The references should be listed alphabetically without numbering. References should be written as per the latest APA format (5th edition). Acknowledgements should be at the end of the text. . . . . Figures: The figures should be numbered in the order in which they are discussed, using Arabic numbers. All figures should have brief legends included at the bottom. Figures should also be provided separately in jpeg format (should not be in word file) in addition to those provided in the text of the article. Tables: The tables should be numbered in the order in which they are discussed using Arabic numbers. All tables should have brief legends included at the bottom.

Похожие интересы