journal of all india institute of speech and hearing
DESCRIPTION
Speech Therapy and AudiologyTRANSCRIPT
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 John’s 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
JAIISHJAIISHJAIISHJAIISH ISSN 0973-662X Vol. 27, 2008
Journal of All India Institute of Speech & Journal of All India Institute of Speech & Journal of All India Institute of Speech & Journal of All India Institute of Speech & HearingHearingHearingHearing
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: [email protected]
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 [email protected].
Dr.Vijayalakshmi Basavaraj
Director & Editorial-Chief
Table of Contents
Speech
1. Effect of Vowels on Consonants in Nasalence Gopishankar R. & Pushpavathi M.
3-7
2. Investigation into Voice Source of Monozygotic Twins using Formant based Inverse Filtering Jayakumar T. & Savithri S.R.
8-14
3. Effect of Spectral Variation on Phoneme Identification Skills in 2 - 3 year old Typically Developing Children Powlin Arockia Catherine & Savithri S.R.
15-18
4. Keratosis of the Larynx Rajasudhakar R. & Sundara Raju H.
19-21
5. Nasalence Value for Rainbow Passage: Normative Data for Non- Native Speakers Sangeetha M. & Pushpavathi M.
22-28
6. Rate of Speech/Reading in Dravidian Languages Savithri S.R. & Jayaram M.
29-39
7. Speech Rhythm in Hearing-Impaired Children Savithri S.R., Ruchi Agarwal & Johnsi Rani R.
40-43
Language
8. The Semantic Association in the Mental Lexicon Gopee Krishnan & Shivani Tiwari
44-48
9. Semantic Pragmatic Attributes and Cognition in Acute and Chronic Schizophrenics: A case comparative study Mithila Poonacha, Shivani Tiwari & Rajashekhar Bellur
49-53
10. Pragmatic Skills in Typically Developing Infants Shilpashri H.N. & Shyamala K. Chengappa
54-57
11. Pragmatic Skills in Nonverbal Identical Twins with Autism Spectrum Disorders Shilpashri H.N. & Shyamala K. Chengappa
58-63
12. Continuum of Developmental Language Disorders: Where Does PLI Fall? Shivani Tiwari, Ashwini Bhat & Rajashekar Bellur
64-73
13. Analysis of Oral and Written Narratives of Children with Language Impaired Learning Disabilities Siddique Tehniat & Mukhopadhyay Sourav
74-82
14. Phonological Processes in Typically Developing Kannada Speaking Children Sreedevi N. & Shilpashree H.N.
83-88
15. A Profile of Aetiological Therapeutic Searches by Netizen Parents/Caregivers of Children on the Autism Spectrum Venkatesan S. & Purushotham K.
89-94
Hearing
16. Auditory Memory and Sequencing in Children aged 6 to 12 years Devi N., Sujitha Nair & Asha Yathiraj
95-100
17. Regenerative Myringoplasty – A Case Report Rajeshwari G. & Sundara Raju H.
101-103
18. Guidelines to Establish a Hospital Based Neonatal Hearing Screening Program in the Indian Setting Ramesh A., Nagapoornima M., Srilakshmi V., Dominic M. & Swarnarekha
104-109
19. Reversible Sudden Sensory Neural Hearing Loss – A Case Report Sundara Raju H. & Rajeshwari G.
110-113
JAIISH, Vol. 27, 2008 Effect of Vowels on Consonants in Nasalence
3
Effect of Vowels on Consonants in Nasalence
1Gopi 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 consonant-
vowel-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 co-
articulatory 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)
1Research officer, Dept. of Clinical Services, All India Institute of Speech and Hearing, Manasagangothri, Mysore-570006, email:[email protected], 2Reader in Speech Pathology, All India Institute of Speech and Hearing, Manasagangothri, Mysore- 570 006, email: [email protected].
JAIISH, Vol. 27, 2008 Effect of Vowels on Consonants in Nasalence
4
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 & Foley-
Homan N (1999) compared nasalance values for
17-syllable passage, 6-syllable sentence, and 2-
syllable 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.
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-plus-
oral 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
5
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.
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.
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.
Figure 1: Mean scores for 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 Across the consonants
F value P
value F value
P value
/pa/-/pi/ /pa/-/ta/ /pa/-/pu/ /pa/-/ka/ /pi/-/pu/
F(2,98)=27.345
<0.001
/ta/-/ka/
F(2,98)=2.481
>0.05
/ta/-/ti/ /pi/-/ti/ /ta/-/tu/ /pi/-/ti/ /ti/-/tu/
F(2,98)=34.808
<0.001
/ti/-/ki/
F(2,98)=24.203
<0.001
/ka/-/ki/ /pu/-/tu/ /ka/-/ku/ /pu/-/ku/ /ki/-/ku/
F(2,98)=72
<0.001
/tu/-/ku/
F(2,98)=3.125
>0.05
Table 3: F values on repeated measures of ANOVA.
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
JAIISH, Vol. 27, 2008 Effect of Vowels on Consonants in Nasalence
6
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 consonant-
vowel 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 /,
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,
519–527.
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
7
quality on velarfunction, phonetica, 36, 187-
193.
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;
1989:246–257.
Karnell M P(1995). Nasometric discrimination of
hypernasality and turbulent nasal airflow.
The Cleft palate-Craniofacial Journal,
32,145–148.
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, 4,381–396.
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 palate-
Craniofacial 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 ,67–73.
Watterson T, Lewis KE & Foley-Homan N (1999).
Effect of stimulus length on nasalance
values. The Cleft palate-Craniofacial Journal, 36, 243–247.
JAIISH, Vol. 27, 2008 Voice Source of Monozygotic Twins
8
Investigation into Voice Source of Monozygotic Twins using
Formant based Inverse Filtering
1Jayakumar 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 Fant’s 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
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.
1Junior Research Fellow, Dept. of Speech Language Sciences, All India Institute of Speech and Hearing, Mysore-570006, email: [email protected], 2 Professor, Dept. of Speech Language Sciences, All India Institute of Speech and Hearing, Mysore-570006, email: [email protected].
JAIISH, Vol. 27, 2008 Voice Source of Monozygotic Twins
9
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.
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
10
and its derivative in the time domain and figure 2
shows the log spectrum of voice source.
0 – glottal Onset, P – peak flow, E – Epoch, C –
Closure, T0-Pitch period, TP-Opening Interval, TN-
Closing 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.
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.
Harmonic ratio (H1-H0) is the ratio of energy at
first harmonic and fundamental frequency. EI/EE-
is 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.
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 r-
values of groups. r values suggest that there was
not much variability among trails, except in few
parameters.
E
JAIISH, Vol. 27, 2008 Voice Source of Monozygotic Twins
11
Trial Parameters
1 & 2 1 & 3 2 & 3 T0 0.892** 0.977** 0.948** OQ 0.924** 0.891** 0.850** SQ 0.704* 0.838** 0.871** LQ 0.768** 0.716** 0.938**
EIEE 0.727** 0.916** 0.741**
AR 0.874** 0.785** 0.798** H0 0.380 0.429 0.892**
H0-H1 0.803** 0.875** 0.751** Roll-off 0.979** 0.865** 0.924**
Table 1: r- values of group I. (** p< 0.01, * p< 0.05).
Trial Parameters 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).
Comparison with in group I & group II
Results of one-way ANOVA revealed no
significant difference within twins in all pairs in
various parameters. Table 3 show mean, standard
deviation of group I and Table 4 show mean, and
standard deviation in group II.
T0 OQ SQ LQ EI/EE AR H0 Ho-H1 Roll-off
Pair 1 3.81(.01)**4.08(.02)
.75(.01)
.74(.02) 1.31(.06) 1.35(.25)
.25(.04)**
.19(.06) .67(.02) .65(.07)
1.70(.29)**1.96(.15)
110.1(1.5) 108.6(4.6)
9.7(3.21) 11.5(6.33)
7.03(1.36)*5.62(.09)
Pair 2 4.02(.06)**4.83(.02)
72(.03)**.75(.01)
1.81(.24) 1.81(.17)
.23(.03)**
.19(.05) .61(.04)**.57(.01)
2.21(.14)**1.70(.38)
108(1.4)** 102 (7)
15.5(6.28) 13.8(4.44)
5.53(.16)* 5.65(.09)
Pair 3 4.53(.03)**4.48(.05)
.89(.07)**
.97(.01) 1.29(.16) 1.26(.06)
.22(.05)
.24(.02) .54(.02)**.61(.05)
1.35(.17)*1.26(.07)
108.8(1.6)**101.3(.6)
11.07(1.40)**13.55(2.55)
14.33(1.3)**12.36(.63)
Pair 4 4.54(.25)**4.17(.06)
.86(.04)**
.94(.06) 1.88(.32) 1.74(.09)
.11(.03)*
.13(.03) .39(.04)**.58(.06)
1.96(.29)**1.71(.21)
101.(1.5)** 107.(1)
5.6(.59)** 9.0(.83)
7.17(.35)** 12.47(2.50)
Pair 5 4.02(.03) 4.0(.02)
.93(.03) *
.81(.01) 1.57(.16)**2.16(.10)
.20(.03)**
.16(.01) .50(.02)**.45(.01)
1.56(.15)**2.33(.10)
100.3(.53)**98.3(.22)
9.47(.48) 3.85(.18)
15.36(.82)**15.92(.48)
Pair 6 4.58(.08)**4.46(.09)
.63(.05)**
.77(.04) 1.41(.17)**2.10(.34)
.03(.01)
.05(.02) .61(.04) .56(.09)
1.86(.11) 1.92(.18)
101.5(.53) 103.1(3.1)
3.6(.51) ** 8.3(1.2)
6.54(.74) 6.30(.58)
Table 3: Mean and SD of 1-3 twin pairs. (** p< 0.01, * p< 0.05)
T0 OQ SQ LQ EI/EE AR H0 Ho-H1 Roll-off
UPair 1 4.21(.04)** 4.31(.06)
.84(.01)*
.89(.07) 1.95(.09)** 1.55(.39)
.14(.01)**
.14(.06) .49(.03)* .55(.16)
1.46(.10)** 1.67(.44)
111.9(.72)** 106.1(2.5)
7.52(.49)** 10.7(1.04)
6.11(1.03)** 10.5(.71)
UPair 2 4.33(.10)** 4.43(.07)
.81(.04)*
.86(.05) 2.93(.72)** 1.82(.14)
.06(.03)**
.10(.03) .49(.16) .50(.08)
2.41(.71)* 1.85(.13)
101.3(4.5)** 105.8(1.5)
4.87(1.57)** 7.66(.89)
6.07(.74)** 9.23(3.22)
UPair 3 5.18(.11)** 4.87(.06)
.63(.08)**
.74(.04) 2.31(.67) 2.24(.16)
.09(.06)**
.02(.01) .37(.03)** .52(.05)
2.55(.64)** 2.09(.13)
99.(2.04)** 103.(1.06)
1.75(1.4)** 6.06(.45)
5.66(.44)** 6.24(.24)
UPair 4 4.57(.04)** 4.23(.11)
.79(.10)**
.92(.08) 1.81(.33) 1.63(.33)
.09(.04)**
.19(.06) .39(.05)** .64(.08)
2.08(.53)** 1.45(.18)
101.4(1.53) 101.2(2.19)
6.55(1.80)** 13.15(2.34)
9.47(2.36)** 9.88(.83)
UPair 5 4.42(.07)* 4.66(.03)
.94(.02)**
.80(.05) 2.13(.22)* 2.39(.26)
.16(.03)**
.10(.01) .35(.04) .35(.02)
1.96(.21)* 2.19(.14)
105.5(.86) 102.3(1.0)
6.71(1.24) 6.0(.50)
7.23(1.23) 8.27(.89)
UPair 6 5.18(.11) 4.29(.04)
.63(.08)**
.84(.01) 2.31(.67) 1.95(.09)
.09(.06)**
.14(.01) .37(.03)** .49(.03)
2.55(.64) 1.46(.10)
99.5(2.0) 111.9(.72
1.75(1.4) 7.52(.49)
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 co-
twin 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.
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
12
Unrelated Twins
T0 0.33 (.28) .28 (.29) OQ 0.12 (.09) 0.08 (.06)** SQ 0.62 (.47) 0.36 (.29)** LQ 0.06 (.05) 0.05 (.04)* EI/EE 0.14 (.10) 0.08 (.06)** AR 0.60 (.52) 0.38 (.28)** H0 5.79 (3.6) 4.88 (3.6)* Ho-H1 4.52 (2.35) 4.05 (3.27) Roll-off 2.18 (1.95) 1.78 (2.14)
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 twin’s 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
quotient and speed quotient are parameter which is
more similar in monozygotic twin’s 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 twin’s voice.
Further investigation on twin pairs based on
perceptual and genetic analysis is warranted.
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, 240-
254.
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, 466-
471.
Decoster, W. A., Van Gysel, A., Vercammen, J., &
Debruyne, F. (2001). Voice similarity in
JAIISH, Vol. 27, 2008 Voice Source of Monozygotic Twins
13
identical twins. Acta Oto Rhino
laryngological belg, 55, 49-55.
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.
JAIISH, Vol. 27, 2008 Voice Source of Monozygotic Twins
14
Appendix 1 – Mean and SD of absolute difference values in both groups.
(U-unrelated pair, Pair – twin pair)
U 1 & Pair 1 U 2 & Pair 2 U 3 & Pair 3 U 4 & Pair 4 U 5 & Pair 5 U 6 & Pair 6 T0 .09(.05)
.27(.03) .13(.10) .81(.07)
.31(.14) .06(.05)
.33(.12) .37(.23)
.23(.06) .03(.03)
.88(.12) .13(.11)
OQ .07(.04) .02(.01)
.05(.03) .04(.02)
.10(.06) .08(.06)
.15(.14) .11(.04)
.14(.05) .12(.04)
.21(.09) .14(.05)
SQ .50(.26) .20(.21)
1.11(.72) .24(.15)
.68(.38) .16(.09)
.44(.24) .25(.19)
.32(.30) .59(.16)
.67(.37) .71(.33)
LQ .05(.03) .07(.04)
.03(.03) .07(.04)
.07(.07) .07(.03)
.10(.09) .03(.02)
.05(.03) .04(.03)
.08(.02) .02(.01)
EI/EE .15(.10) .06(.05)
.16(.14) .05(.03)
.15(.04) .07(.04)
.24(.10) .18(.07)
.04(.04) .05(.02)
.12(.04) .09(.06)
AR .34(.36) .27(.20)
.79(.38) .49(.18)
.75(.38) .16(.12)
.68(.53) .39(.29)
.26(.26) .79(.18)
1.5(.10) .17(.11)
H0 5.85(2.29) 3.49(1.61)
5.75(1.03) 6.87(6.58)
4.32(1.73) 7.54(2.09)
3.18(1.19) 6.39(1.30)
3.19(1.70) 2.09(.70)
12.44(2.21) 2.93(1.86)
Ho-H1 3.18(1) 4.72(6.03)
2.90(1.50) 6.14(4.15)
4.30(1.51) 2.63(1.75)
6.60(2.53) 3.42(1.07)
1.37(.93) 5.61(.51)
5.76(1.46) 4.75(.96)
Roll-off 4.40(.98) 1.41(1.35)
3.58(2.9) .16(.11)
.62(.13) 2.27(1.46)
2.19(1.01) 5.31(2.39)
1.23(.68) .91(.75)
1.03(.77) .62(.52)
JAIISH, Vol. 27, 2008 Voice Source of Monozygotic Twins
15
JAIISH, Vol. 27, 2008 Effect of Spectral Variation on Phoneme Identification Skills
16
Effect of Spectral Variation on Phoneme Identification Skills in 2 - 3
year old Typically Developing Children
1Powlin 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 /tA/) 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
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
1Junior Research Fellow, Dept. of Speech-Language Sciences, All India Institute of Speech and Hearing, Mysore, email:[email protected], 2Professor, Dept. of Speech-Language Sciences, All India Institute of Speech & Hearing, Manasagangothri, Mysore 570 006, email:[email protected],
JAIISH, Vol. 27, 2008 Effect of Spectral Variation on Phoneme Identification Skills
17
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 cross-
sectional 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.
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)
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.
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 audio-
presented 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
subject’s 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
18
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
didn’t 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 (Range)
1820 (590)
1620 (630)
1680 (650)
LLPB (Range)
1630 (610)
1485 (650)
1520 (650)
ULPB (Range)
1995 (470)
1880 (610)
1920 (660)
PBW (Range)
365 (240)
395 (460)
400 (430)
Table 1: Mean and range values of measures of /p- t/
continuum (in Hz).
Figure 2: Mean percent identification scores in boys.
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 Sussman’s 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
Sussman’s 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
JAIISH, Vol. 27, 2008 Effect of Spectral Variation on Phoneme Identification Skills
19
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.
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 late-
talking 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.
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). Cross-
language 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, 1309-
1317.
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.
JAIISH, Vol. 27, 2008 Keratosis of the Larynx
20
Keratosis of the Larynx
1Rajasudhakar 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, Case history, Acoustic parameters, Medication.
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 (García, Aranzábal, 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 pre-
cancerouos 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
1Junior Research Fellow, Dept. of Speech-Language Sciences, All India Institute of Speech & Hearing, Manasagangothri, Mysore-570006, email: [email protected], 2Reader in ENT, All India Institute of Speech & Hearing, Manasagangothri, Mysore-570006.
JAIISH, Vol. 27, 2008 Keratosis of the Larynx
21
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 Condition 1 Condition 2 Fundamental frequency (Fo)
/a/ /i/ /u/
135 Hz 148 Hz 136 Hz
124 Hz 128 Hz 127 Hz
Reading Fo 145 Hz 130 Hz Speaking Fo 140 Hz 126 Hz Frequency range 108-216 Hz 102-228 Hz Intensity range 88-105 dB 70-110 dB Extent of Fluctuation in -Intensity -Frequency
2.52 dB 3.19 Hz
1.25 dB 2.51 Hz
Speed of Fluctuation in -Intensity -Frequency
4.50 % 12.5 %
2.31 % 3.87 %
Maximum phonation duration /a/ /i/ /u/ /s/ /z/
15 sec 12 sec 14 sec 12 sec 10 sec
17 sec 15 sec 13 sec 12 sec 12 sec
EGG measures -Open Quotient (OQ) -Closed Quotient (CQ)
59 % 41 %
54.11 % 45.89 %
MDVP frequency, intensity and perturbation related measures affected
Normal
(Condition 1 = before medication; Condition 2 = after
medication)
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)
seen in the anterior one-third region of the vocal
folds. Conservative medication like antibiotics
(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 (after-
medication).
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.
Figure 1: (a) Results of MDVP parameters in condition
1; and in condition 2; (b) (beyond green
circle indicates abnormal).
JAIISH, Vol. 27, 2008 Keratosis of the Larynx
22
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.
Conclusions
Early diagnosis of laryngeal keratosis, a
precancerous lesion, is important since it has the
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
García, R. L., Aranzábal, 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.
JAIISH, Vol. 27, 2008 Nasalence Value for Rainbow Passage
23
Nasalence Value for Rainbow Passage: Normative Data for Non-
Native Speakers
1Sangeetha 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 O’Regan, 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
1Clinical Lecturer, Dept. of Clinical Services, All India Institute of Speech and Hearing, Manasagangothri, Mysore- 570 006, email: [email protected], 2Reader, Speech-Language Pathology, All India Institute of Speech and Hearing, Manasagangothri, Mysore- 570 006, email: [email protected]
JAIISH, Vol. 27, 2008 Nasalence Value for Rainbow Passage
24
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.
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 Age
range Language (Mother tongue)
Kannada Malayalam Tamil Telugu Hindi Males (N=45)
18-30 yrs
12 8 7 3 15
Females (N=70)
18-30 yrs
30 15 6 2 22
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 subject’s speech are sensed by
microphones on either side of a sound separator
that rests on the patient’s 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 manufacturer’s 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 subject’s 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 subject’s 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
25
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 Gender
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.
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
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 (Year)
Language N Subjects Mean S.D
Fletcher et al. (1989)
American 117 Children 35.6 5.20
Adult male 35.0 6.0 Seaver etal(1991)
American 140 Adult female
36.0 6.0
North west American
30 Geriatric male
23.5 5.1 Hutchinson etal (1978) North west
American 30
Geriatric female
32.0 10.7
Mid Atlantic 15 Adult male 36.0 4.0 Litzaw & Dalston (1992)
Mid Atlantic 15 Adult female
37.0 4.0
45 Adult male 27.9 4.1 Present study (2008)
Indian 70
Adult female
31.3 7.3
Table 3: The normative data for Rainbow passage
across various studies.
JAIISH, Vol. 27, 2008 Nasalence Value for Rainbow Passage
26
Sl No.
Author (Year)
Subjects “t” value
“p” value
Interpretation
1 Fletcher et al. (1989)
Children 9.25 <0.001 Significant difference
Males 11.23 <0.001 Significant difference
2 Seaver etal(1991)
Females 5.23 <0.001 Significant difference
Males 7.04 <0.001 Significant difference 3
Hutchinson etal (1978)
Females 0.69 >0.05 No significant difference
Males 12.82
<0.001 Significant difference
Females 6.36 <0.001 Significant difference
4 Litzaw & Dalston (1992)
Total 9.76 <0.001 Significant difference
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)
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.
JAIISH, Vol. 27, 2008 Nasalence Value for Rainbow Passage
27
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.
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
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
JAIISH, Vol. 27, 2008 Nasalence Value for Rainbow Passage
28
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.
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., & O’Regan, 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 cross-
sectional 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
29
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.
JAIISH, Vol. 27, 2008 Rate of Speech
30
Rate of Speech/Reading in Dravidian Languages
1Savithri 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
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; Stephenson-
Opsai & 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
1Professor, Dept. of Speech-Language Sciences, All India Institute of Speech & Hearing, Manasagangothri, Mysore 570 006, email:[email protected], 2Professor, Dept. of Speech Pathology and Audiology, NIMHANS, Bangalore 560 029,
JAIISH, Vol. 27, 2008 Rate of Speech
31
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.
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
32
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.
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.
Age range/ Language
Kannada Telugu Tamil Malayalam
Age M F M F M F M F 3-3.11 5 5 4-4.11 5 5 5-5.11 5 5 2 2 6-6.11 5 5 3 5 7-10 5 5 5 5
11-15 5 5 5 5 5 5 5 5 16-20 5 5 5 5 5 5 5 5 21-30 5 5 5 5 5 5 5 5 31-40 5 5 5 5 5 5 5 5 41-50 5 5 5 5 5 5 5 5 51-60 5 5 5 5 5 5 5 5 61-70 5 5 3 3 5 5 5 5 71-80 5 5 1 0 5 5 5 5 81-90 5 1 2 0 0 1 5 5
91-100 0 0 0 0 0 0 1 2 Sub-total 70 66 36 33 50 53 46 47
Total 401
Table 1: Details of subjects.
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 audio-
recorded and were digitized at 16,000 Hz sampling
frequency. Pauses, if any, were removed from the
waveform using Cool Edit software. Each syllable
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
33
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.
Age M/SD SS SPM WPM
3-3.11 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.
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 Duncan’s post-
hoc test. Values in the same column are not significantly different.
Age Group
1 2 3 4
3-3.11 4.2358 4-4.11 4.2031 5-5.11 4.2358 6-6.11 7-10 11-15 5.7000 5.7000 16-20 7.0300 21-30 6.4900 31-40 7.1800 41-50 6.7900 51-60 6.8640 61-70 6.9540 71-80 6.4400 6.4400 81-90 5.6200
Table 3: Results of Duncan’s post-hoc test for SS
(Kannada).
Age Group
1 2 3 4 5
3-3.11 290.9821 4-4.11 252.1879 5-5.11 254.1461 6-6.11 7-10 11-15 343.10 343.10 16-20 425.10 21-30 385.50 385.50 385.50 31-40 434.20 41-50 410.41 51-60 415.10 61-70 403.60 71-80 389.80 389.80 81-90 337.3400 337.34
Table 4: Results of Duncan’s post-hoc test for SPM
(Kannada). Age
Group 1 2 3
3-3.11 128.6358 4-4.11 91.3535 5-5.11 119.7360 119.7360 6-6.11 92.3000 7-10 84.5000 11-15 103.6844 103.6844 16-20 127.3501 21-30 116.2539 116.2539 31-40 131.0137 41-50 123.6824 51-60 124.1360 61-70 124.8651 71-80 117.7516 117.7516 81-90 101.7867 101.7867
Table 5: Results of Duncan’s 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.
JAIISH, Vol. 27, 2008 Rate of Speech
34
Age M/SD SS SPM WPM
11-15 Mean 7 431 123 SD 1.25 132 37 16-20 Mean 7 439 125 SD 0.61 36 10 21-30 Mean 8 466 133 SD 0.58 36 10 31-40 Mean 6 384 116 SD 0.95 54 23 41-50 Mean 7 389 117 SD 1.04 75 20 51-60 Mean 6 392 110 SD 1.04 69 18 61-70 Mean 6 336 96 SD 0.99 60 17 71-80 Mean 5 309 89 SD 0.19 11 4 81-90 Mean 5.14 309 88.82 SD 0.19 11.31 3.85
Mean SD
6.4 0.76
384 53.81
111 15.87
Table 6: Mean and SD of rate of speech in Telugu
speakers.
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 Duncan’s post-
hoc test. Values in the same column are not
significantly different.
Age Group
1 2 3 4
11-15 6.2460 6.2460 16-20 6.9700 6.9700 21-30 7.4060 7.4060 31-40 7.7100 41-50 6.3750 6.3750 51-60 6.8970 6.8970 61-70 6.4230 6.4230 71-80 5.5990 5.5990 81-90 5.1400
Table 7: Results of Duncan’s post-hoc test for SS
(Telugu).
Age Group
1 2 3
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 Duncan’s post-hoc test for SPM
(Telugu).
Age Group
1 2 3
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 Duncan’s post-hoc test for WPM
(Telugu).
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 M/SD SS SPM WPM
5-5.11 Mean 4.9950 299.4675 113.9975 SD 1.0360 62.3379 23.7109 6-6.11 Mean 4.6484 278.9059 86.0060 SD 1.0709 64.2550 19.8293 7-10 Mean 6.6920 401.5178 123.8661 SD 1.6674 100.0455 30.8635 11-15 Mean 4.5113 271.7749 94.6891 SD .3627 21.8501 7.7201 16-20 Mean 6.5238 383.5347 133.0958 SD .7114 53.0329 18.4800 21-30 Mean 6.5044 390.8467 135.6623 SD 1.2267 73.5010 25.4697 31-40 Mean 5.4654 326.3887 113.5195 SD 1.1116 67.7142 23.5099 41-50 Mean 5.9058 352.6335 122.4497 SD .8357 49.1482 17.2613 51-60 Mean 5.4330 326.3735 114.8765 SD .8551 51.5573 20.3268 61-70 Mean 5.6564 339.9639 118.0024 SD 1.0284 61.8869 21.4296 71-80 Mean 5.9033 353.8821 123.3533 SD .9218 54.2288 18.5347 Average Mean 5.7169 342.1399 117.0508 SD 1.2058 72.3806 24.6119
Table 10: Mean and SD of rate of speech in Tamil
speakers.
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 Duncan’s post-
hoc test. Values in the same column are not
significantly different.
JAIISH, Vol. 27, 2008 Rate of Speech
35
Age Group
1 2 3 4 5
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 Duncan’s post-hoc test for SS
(Tamil).
Age Group
1 2 3 4 5
5-5.11 299.4675 299.4675 299.4675 6-6.11 278.9059 278.9059 7-10 401.5178 11-15 271.7749 16-20 383.5347 383.5347 21-30 390.8467 390.8467 31-40 326.3887 326.3887 326.3887 326.3887 41-50 352.6335 352.6335 352.6335 51-60 326.3735 326.3735 326.3735 326.3735 61-70 339.9639 339.9639 339.9639 339.9639 71-80 353.8821 353.8821 353.8821
Table 12: Results of Duncan’s post-hoc test for SPM (Tamil).
Age Group
1 2 3
5-5.11 113.9975 113.9975 6-6.11 86.0060 7-10 123.8661 11-15 94.6891 94.6891 16-20 133.0958 21-30 135.6623 31-40 113.5195 113.5195 41-50 122.4497 51-60 114.8765 114.8765 61-70 118.0024 71-80 123.3533
Table 13: Results of Duncan’s 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.
Age M/SD SS SPM WPM
11-15 Mean 7.9082 474.4926 115.2427 SD .8486 50.9189 12.3654 16-20 Mean 8.8040 528.9257 128.2894 SD 1.2238 73.3538 17.7127 21-30 Mean 9.2600 557.9370 135.4231 SD .9902 61.7733 14.9202 31-40 Mean 8.2000 492.1712 119.4511 SD .6055 35.5242 8.5534 41-50 Mean 7.9000 476.6454 115.6346 SD 1.2275 80.9450 19.7229 51-60 Mean 7.9390 482.6940 117.1112 SD 1.0498 66.8788 16.1906 61-70 Mean 7.8170 447.6767 109.9117 SD .5858 66.2682 12.4797 71-80 Mean 7.0350 423.1902 102.3253 SD 1.6478 99.4643 23.9974 81-90 Mean 6.6160 397.8133 96.4343 SD .8643 51.8349 12.7868 91-100 Mean 4.2800 257.4613 62.6108 SD .7302 43.9912 10.7018
Mean SD
7.8240 1.4037
468.6865 87.6557
113.8287 21.0098
Table 14: Mean and SD of rate of speech in Malayalam
speakers.
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 Duncan’s
post-hoc test. Values in the same column are not
significantly different.
Age Group
1 2 3 4 5
11-15 7.9082 7.9082 16-20 8.8040 8.8040 21-30 9.2600 31-40 8.2000 8.2000 8.2000 41-50 7.9000 7.9000 51-60 7.9390 7.9390 61-70 7.8170 7.8170 71-80 7.0350 7.0350 81-90 6.6160 91-100 4.2800
Table 15: Results of Duncan’s post-hoc test for SS
(Malayalam).
JAIISH, Vol. 27, 2008 Rate of Speech
36
Age Group
1 2 3 4 5
11-15 474.4926 474.4926 16-20 528.9257 528.9257 21-30 557.9370 31-40 492.1712 492.1712 492.1712 41-50 476.6454 476.6454 51-60 482.6940 482.6940 61-70 447.6767 447.6767 71-80 423.1902 423.1902 81-90 397.8133 91-100
257.4613
Table 16: Results of Duncan’s post-hoc test for SPM
(Malayalam).
Age Group
1 2 3 4 5
11-15 115.2427 115.2427 16-20 128.2894 128.2894 21-30 135.4231 31-40 119.4511 119.4511 119.4511 41-50 115.6346 115.6346 51-60 117.1112 117.1112 61-70 109.9117 109.9117 71-80 102.3253 102.3253 81-90 96.4343 91-100
62.6108
Table 17: Results of Duncan’s 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 Duncan’s 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 Subset SS
Language 1 2 3 Tamil 103 5.6720 Kannada 136 5.7771 Telugu 82 6.6651 Malayalam 93 7.8240 SPM Tamil 103 339.4596 Kannada 136 346.7957 Telugu 82 406.2927 Malayalam 93 468.6865
Table 18: Results of Duncan’s test on significant difference
between languages.
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
Kannada Telugu Tamil Malayalam
V 6.3 8.2 4.8 7.3 CV 73 67 54 74 CCV 21 22 0.4 3.6 CVC 1.6 38 13.4 CCVC 0.5 0.2 CCCV 0.2 VCC 0.1 0.2 VC 2.7 1.1 CVCC 0.3 0.2 Total 100 100 100 100
Table 19: Percent syllable type in four Dravidian
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
JAIISH, Vol. 27, 2008 Rate of Speech
37
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 can’t 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
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
38
Ingham, R. J., & Cordes, A. K. (1997). Self-
measurement 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, 1-
20.
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, 1102-
1104.
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.
JAIISH, Vol. 27, 2008 Rate of Speech
39
Appendix I– Normative data on rate of speech/reading in 4 Dravidian
languages
M = Mean, A= Average
Kannada Telugu Tamil Malayalam
Age SS SPM WPM SS SPM WPM SS SPM WPM
SS SPM WPM
3-3.11 M 4.9 291 129 SD 1.3 78 34 4-4.11 M 4.2 252 91 SD .79 48 17 5-5.11 M 4.2 254 120 5 299 114 SD 1.3 79 37 1 62 24 6-6.11 M 4.3 261 92 4.65 278 86 SD 0.84 50 18 1 64 20 7-10 M 4.1 250 85 6.7 402 130 SD 0.65 41 14 1.67 100 31 11-15 M 5.7 343 104 7 431 123 4.5 272 95 7.9 474 115 SD .83 48 14 1.25 132 37 .4 22 8 .8 51 12 16-20 M 7.0 425 127 7 439 125 6.5 384 133 8.8 529 128 SD .96 58 17 0.61 36 10 .7 53 18 1.2 73 18 21-30 M 6.5 386 116 8 466 133 6.5 391 136 9.3 558 135 SD .96 70 21 0.58 36 10 1.2 74 25 .9 62 15 31-40 M 7.2 434 131 6 384 116 5.5 326 114 8.2 492 119 SD .31 19 6. 0.95 54 23 1.1 68 24 .6 36 9 41-50 M 6.8 410 124 7 389 117 5.9 353 122 7.9 477 116 SD .87 53 16 1.04 75 20 .8 49 17 1.2 81 20 51-60 M 6.9 415 124 6 392 110 5.4 326 115 7.9 483 117 SD .62 37 11 1.04 69 18 .9 52 20 1 67 16 61-70 M 6.9 404 125 6 336 96 5.7 340 118 7.8 448 110 SD .73 52 12 0.99 60 17 1 62 21 .6 66 12 71-80 M 6.4 390 118 5 309 89 5.9 354 123 7 423 102 SD .85 51 15 0.19 11 4 .9 54 19 1.6 99 24 81-90 M 5.6 337 102 5.14 309 88.8 6.6 398 96 SD .97 58 17 0.19 11.3 3.85 .86 52 13 91-100 M 4.3 257 63 SD .7 44 11 A M 5.3 318 104 6.4 384 111 5.7 34 117 7.8 469 114
SD 2.20 133 40 0.76 53.8 15.9 1.2 72 25 1.4 88 21
JAIISH, Vol. 27, 2008 Rate of Speech
40
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
Telugu
Age SS SPM WPM
LB UB LB UB LB UB
11-15 5.4 7.1 315 567 90 162
16-20 6.1 7.9 367 475 105 136
21-30 7.0 7.8 413 464 118 132
31-40 7.3 8.1 441 492 126 141
41-50 5.7 7.1 346 422 100 132
51-60 6.2 7.6 336 443 102 131
61-70 5.7 7.2 342 441 97 123
71-80 4.9 6.2 297 377 85 108
81-90 3.4 6.9 207 411 54 123
Tamil
Age SS SPM WPM
LB UB LB UB LB UB
5-5.11 3.3 6.6 200 399 76 152
6-6.11 3.7 5.5 225 333 69 103
7-10 5.5 7.9 330 473 101 146
11-15 4.3 4.8 256 287 89 100
16-20 6.0 7.0 346 421 120 146
21-30 5.6 7.4 338 443 117 154
31-40 4.7 6.2 278 374 97 130
41-50 5.3 6.5 317 388 110 135
51-60 4.8 6.0 289 364 100 129
61-70 4.9 6.4 296 384 103 133
71-80 5.2 6.6 315 393 110 137
Malayalam
Age SS SPM WPM
LB UB LB UB LB UB
11-15 7.3 8.5 438 511 106 124
16-20 7.9 9.7 476 581 116 141
21-30 8.6 10 514 602 125 146
31-40 7.7 8.6 467 518 113 126
41-50 7.0 8.8 419 535 101 130
51-60 7.2 8.7 435 531 106 129
61-70 7.4 8.2 400 495 101 119
71-80 5.9 8.2 352 494 85 119
81-90 6.0 7.2 360 435 87 106
91-100 2.5 6.0 148 367 36 89
JAIISH, Vol. 27, 2008 Speech Rhythm in Hearing-Impaired Children
41
Speech Rhythm in Hearing-Impaired Children
1Savithri 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 hearing-
impaired children in comparison with normals. It is concluded that both the hearing-
impaired 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 stress-
timed 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.
Language Class
Languages Intervocalic rPVI
Vocalic nPVI
Stress-timed
English, Germany
High High
Syllable-timed
French, Spanish
High Low
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 hearing-
impaired children.
1Professor, Dept. of Speech-Language Sciences, All India Institute of Speech & Hearing, Manasagangothri, Mysore-570006, email:[email protected]. 2Research Officer, Dept. of Speech-Language Sciences, All India Institute of Speech and Hearing, Manasagangothri, Mysore-6, email:[email protected], 3Research Officer, Dept. of Speech-Language Sciences, All India Institute of Speech and Hearing, Manasagangothri, Mysore-570006, email: [email protected].
JAIISH, Vol. 27, 2008 Speech Rhythm in Hearing-Impaired Children
42
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 audio-
recorded 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:
Where, m is the number of intervals and dk is the
duration of the kth interval. The normalized
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.
JAIISH, Vol. 27, 2008 Speech Rhythm in Hearing-Impaired Children
43
Subject 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.
PVIs
rPVInPVI
95
% C
I fo
r S
co
res
80
70
60
50
40
30
20
10
Groups
Normals
HI
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 (stress-
timed, 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 (stress-
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.
Figure 2: Mean rPVI and nPVI values for normal,
hearing-impaired children and adults.
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.
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
44
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 (PVI’s) 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.
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
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, 352-
355.
Vasantalakshmi (2005). Development of C
language program. AIISH, Mysore.
Yamini, B.K. (1990). Dysfluencies in children: 5-
6 years. M. Jayaram, & S.R. Savithri
(Eds.). Dissertation abstract: Research at
AIISH, Vol.2, 171-173.
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.
JAIISH, Vol. 27, 2008 Semantic Association
45
The Semantic Association in the Mental Lexicon
1Gopee 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, Colomẽ, & Caramazza, 2000).
Spreading Activation Theory in Lexical
Selection
The concept of Spreading Activation Theory
– an idea originally introduced by Collins and
Loftus (1975) – has received widespread
acceptance in the contemporary cognitive-
linguistic 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
1Assistant Professor, Dept. of Speech and Hearing, Manipal College of Allied Health Sciences, Manipal University, Manipal-576 104, email:[email protected] , 2Lecturer, Dept. of Speech and Hearing, Manipal College of Allied Health Sciences, Manipal University, Manipal-576 104, email:[email protected].
JAIISH, Vol. 27, 2008 Semantic Association
46
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.
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
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 DMDX’s clock was set on with the
presentation of the second word. 100 word-pairs
JAIISH, Vol. 27, 2008 Semantic Association
47
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 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.
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
JAIISH, Vol. 27, 2008 Semantic Association
48
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 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.
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.
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. Colomẽ, A., & Caramazza, A. (2000).
Lexical Access in Speech Production: The
JAIISH, Vol. 27, 2008 Semantic Association
49
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, 116–124.
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.
JAIISH, Vol. 27, 2008 Semantic – Pragmatic Attributes and Cognition in Schizophrenics
50
Semantic – Pragmatic Attributes and Cognition in Acute and
Chronic Schizophrenics: A Case Comparative Study
1Mithila 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 “Addenbrooke’s 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
1Post 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:[email protected], 3Professor, Dept. of Speech and Hearing, Manipal College of Allied Health Sciences, Manipal University, Manipal-576 104.
JAIISH, Vol. 27, 2008 Semantic – Pragmatic Attributes and Cognition in Schizophrenics
51
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.
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.
Details Acute Schizophrenic (X)
Chronic Schizophrenic (Y)
Age/Gender 30 yrs, female 24yrs, male Language Kannada Kannada Education B.Sc. Computer
science (incomplete)
S.S.L.C (fail)
Complaint c/o no problem c/o hand pain since 4 days
Onset 2 months 2 years Pre-Morbid History
Maintaining well Well-adjusted personality
Post-Morbid History
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.
Decreased interaction, increased suspiciousness and abusive behavior, withdrawal from family, smiling & muttering to self, irrelevant talk, increased religiosity.
Diagnostic Criteria
Auditory hallucinations, delusion of love (?), socio-occupational dysfunction.
Auditory hallucinations, delusion of persecution, delusion of reference, delusion of grandiose, breaks in the train of thoughts, self-absorbed attitude, socio-occupational dysfunction.
Treatment On antipsychotic treatment for a month
On antipsychotic treatment along with electro-convulsive treatment (ECT) for more than a year
Table 1: Case details and demographics.
Materials and Procedure
The subjects were tested for cognition,
semantics and pragmatics:
Cognition: Cognitive abilities of the subjects were
assessed using “Addenbrooke’s 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-to-
picture 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
JAIISH, Vol. 27, 2008 Semantic – Pragmatic Attributes and Cognition in Schizophrenics
52
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.
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).
Acute Schizophrenic (X) Chronic schizophrenic (Y)
Attention and concentration: [Orientation:
10/10, Registration: 3/3, Attention & Concentration:
5/5] Total: 18/18
Attention and concentration: [Orientation:
5/10, Registration: 3/3, Attention & Concentration:
4/5] Total: 12/18
Memory: [Recall: 2/10, Anterograde: 3/7, Retrograde: 1/4, Recognition: 3/5]
Total: 9/26
Memory: [Recall: 3/10, Anterograde:4/7, Retrograde: 3/4, Recognition: 1/5]
Total: 11/26
Verbal fluency: [Letter: 1/7, Categorical: 3/7]
Total: 4/14
Verbal fluency: [Letter: 5/7, Categorical: 3/7]
Total: 8/14
Language: [Comprehension: 3/8,
Writing: 1/1, Repetition: 4/4, Naming: 7/12,
Reading: 1/1] Total: 16/26
Language: [Comprehension: 6/8,
Writing: 0.5/1, Repetition: 4/4, Naming: 6/12,
Reading: 1/1] Total: 17/26
Visuo-spatial abilities: [Visuo-spatial abilities :4/8,
Perceptual abilities: 8/8] Total: 12/16
Visuo-spatial abilities: [Visuo-spatial abilities :7/8,
Perceptual abilities: 8/8] Total: 14/16
Overall ACE Score: 59/100 Overall ACE Score: 62/100
Table 2: Subjects’ performance on Cognition (ACE – R, Kannada).
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
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
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: 8 Total: 8
Word association: 3 Total: 3
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.
JAIISH, Vol. 27, 2008 Semantic – Pragmatic Attributes and Cognition in Schizophrenics
53
Verbal Aspects Variety of
speech acts Topic selection
Topic initiation Topic maintenance
Pause time Lexical selection
X – – – + – + Y – + + – – –
Paralinguistic Aspects Intelligibility Fluency Prosody Vocal quality X – – – – Y – – – –
Nonverbal Aspects Eye gaze Facial expression Gestures Body posture X – – – – Y – – – –
(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 Gordon’s (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
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 lexico-
semantic memory) as well as in discourse
JAIISH, Vol. 27, 2008 Semantic – Pragmatic Attributes and Cognition in Schizophrenics
54
(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.
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, 684-
696.
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. 444-
466).
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.
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.
JAIISH, Vol. 27, 2008 Pragmatic Skills in Infants
55
Pragmatic Skills in Typically Developing Infants
1Shilpashri 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”.
Children’s 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, children’s 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 children’s 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
1Junior Research Fellow, Dept. of Speech Language Pathology, All India Institute of Speech and Hearing, Mysore-570006, email:[email protected], 2Professor, Dept. of Speech-Language Pathology, All India Institute of Speech and Hearing, Mysore-570006, email:[email protected]
JAIISH, Vol. 27, 2008 Pragmatic Skills in Infants
56
& Tomasello, 1998). Mathew (2004) has reported
in her study that children acquire verbal indication
of negation by 1 ½ - 2 years of age.
Studies have also focused on how caregivers
regulate interaction with the baby by selectively
responding to the baby’s gestures. These early
interactions have been referred as proto-
conversations (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.
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. Smiling.
2. Attention.
3. Eye contact.
4. Vocalization.
5. Play behaviors.
6. Non verbal turn taking.
7. Giving on request.
8. 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. Age in months
sex Sl. No. Age in months
sex
1 06 F 5 07 M 2 09 F 6 08 M 3 09 F 7 08 M 4 11 F 8 09 M
Table 1: Demographic data.
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.
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.
JAIISH, Vol. 27, 2008 Pragmatic Skills in Infants
57
Females (age in months) Pragmatic skills
F 106 F 209 F 309 F 4 11
Smiling 1 1 1 1 Attention 1 1 1 1 Eye contact 1 1 1 1 Vocaliza -tion 1 1 1 1 Play behaviors 1 1 1 1 Non verbal turn taking
1 1 1 1
Giving on request 0 0 0 1 Non verbal indication of negation
0 0 0 1
0-Absent, 1-Present.
Table-2: Pragmatic skills of 4 typically developing
infants (females).
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.
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 ½ – 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.
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.
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
Males (age in months) Pragmatic skills
M 107
M 208
M 308 M 409
Smiling 1 1 1 1 Attention 1 1 1 1 Eye contact 1 1 1 1 Vocaliza -tion 1 1 1 1 Play behaviors 1 1 1 1 Non verbal turn taking
1 1 1 1
Giving on request 0 0 0 0 Non verbal indication of negation
0 0 0 0
JAIISH, Vol. 27, 2008 Pragmatic Skills in Infants
58
interaction. Child Development, 55,
1278-1289.
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 master’s
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 master’s
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
JAIISH, Vol. 27, 2008 Pragmatic skills in Autism Spectrum Disorders
59
Pragmatic Skills in Nonverbal Identical Twins with Autism
Spectrum Disorders
1Shilpashri 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 {DSM-
IV}, 1994). The onset is always in childhood and the symptoms persist throughout life. Hence, the
term pervasive developmental disorders (PDD’s).
As per DSM-IV-TR published in 2000, the PDD’s
includes Autism, Asperger disorder, Rett’s
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 (Berko-
Gleason, 2005). Pragmatic aspect of language
acquisition accounts for children’s 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
PDD’s / ASD’s. Aarons and Gittens (1987); Wing
(1988) have even suggested that pragmatic
disability is just another term for autism. (Ball,
1Junior Research Fellow, Dept. of Speech-Language Pathology, All India Institute of Speech and Hearing, Manasagangothri, Mysore–570006, email:[email protected], 2Professor, Dept. of Speech-Language Pathology, All India Institute of Speech and Hearing, Manasagangothri, Mysore–570006, email:[email protected].
JAIISH, Vol. 27, 2008 Pragmatic skills in Autism Spectrum Disorders
60
1978; Cantwell, Baker & Rutter, 1978; Paul &
Cohen 1985; Loveland & Landry, 1986) have
reported that the autistic children’s 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.
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.
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.
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).
SUBJECTS
Client Report AI A2
Age of onset Before one
year Before one
year Medical history - ve - ve
Motor development
Normal Normal
Hearing Normal Normal Vision Normal Normal
Speech – language skills
No speech (only
vocalization)
No speech (only
vocalization) Social skills Poor Poor Imaginative
play Absent Absent
Stereotyped repetitive behaviors
Present Present
Table-1: Demographic data
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 mother-
child 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.
JAIISH, Vol. 27, 2008 Pragmatic skills in Autism Spectrum Disorders
61
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 other’s 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 week’s 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.
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
62
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
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
63
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).
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.
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
children’s 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
disorders—autism and developmental
disabilities monitoring network, six sites,
United States, 2000. Morbidity and
Mortality Weekly Report, 56(SS-1), 1–11.
Folstein, S., & Rutter, M. (1977). Infantile autism:
a genetic study of 21 twin pairs. Journal of Child Psychology and Psychiatry, 18, 297-
321.
Folstein, S.E., & Rosen-Sheidley, B. (2001).
Genetics of autism: complex etiology for a
JAIISH, Vol. 27, 2008 Pragmatic skills in Autism Spectrum Disorders
64
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 Speech-
Language 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.
JAIISH, Vol. 27, 2008 Continuum of Developmental Language Disorders
65
Continuum of Developmental Language Disorders: Where Does
PLI Fall?
1Shivani 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
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
1Lecturer, Dept. of Speech and Hearing, Manipal College of Allied Health Sciences, Manipal University, Manipal-576104, email: [email protected], 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.
JAIISH, Vol. 27, 2008 Continuum of Developmental Language Disorders
66
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 (Conti-
Ramsden, 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
utterances/behaviors. Further, children with SLI
essentially have impairments in phonological and
syntactic skills (generally not reported in children
with PLI). Within PDD spectrum, Asperger’s 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.
Children’s 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
67
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
CASE 1 CASE 2
Name Master N Baby A Age / Gender 5 years / Male 5 years / Female Language Kannada English Chief complaints • ↓ attention to speech
• Repetition of few sentences • Delayed responses to questions
• Language delay • ↑ level of activity
Onset • Symptoms were first noticed at the age of 1.6years as language delay & poor eye-to-eye contact.
• Symptoms were noticed by the teacher at the age of 3 years (behavioral deviations as poor social skills, echolalia)
History • 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)
• 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)
Audiological evaluation
Normal hearing Normal hearing
Psychological evaluation
• IQ deficits could not be ruled out (BKT) • VSMS: Borderline deficits in social &
adaptive functioning • CARS: mild autistic features
• IQ: average intellectual functioning (BKT) • VSMS: average social & adaptive functioning • CARS: non autistic
Reported to have autistic like features earlier Prelinguistic skills
• ↓ attention to speech • Poor eye contact
• ↓ attention span • Good eye contact with family members • Highly distractible
Speech skills • 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.
• Respiration: normal • Phonation: pitch & quality age adequate • Articulation: cluster reduction & metathesis • Prosody: inappropriate intonation pattern • Fluency: fluent speech.
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.
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
68
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.
CASE1 CASE2
Language Test results
RLA= 30-33 months ELA= 24-27 months Scattered findings
RLA= 4.6–5 years ELA= 4.6-5 years
Morphology
No comprehension & expression of any morphological markers
Age adequate usage
Semantics
• 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
• 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 2-3 word utterances • Adjectives: comprehension is
present • Pronouns: uses 1
st & 2
nd
person pronoun; does not comprehend 3
rd person
pronoun • Does not comprehend gender
markers. • Uses few prepositions • Does not comprehend tenses
& plurals • Uses 1
st & 2
nd person
possessive markers • Difficulty with polar questions • Inappropriate answers to
questions
• 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
JAIISH, Vol. 27, 2008 Continuum of Developmental Language Disorders
69
CASE1 CASE2
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
• 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
Secondary language skills
• Attending school since 2 years • Cannot read & write
• Attending school since 1 year • Able to read & write alphabets, numbers (1-10),
small words, her name, & can solve 1-digit addition & subtraction
Behavioral deviations • 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.
• Hyperactive & easily distractible • Prefers solo play
Stimulation • Poor stimulation both at home & school
• Good stimulation for language, reading, & writing
Previous treatment • No previous treatment. • 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.
Present treatment • Demonstration therapy was carried out for 2 days.
• Activities for improving attention, vocabulary, comprehension of questions, reading, & pragmatic skills were carried out.
• Demonstration therapy was given for a month. • Attention enhancement training, activities to
improve semantic, syntactic skills, pragmatic, & cognitive skills.
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.
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
70
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.
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
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
Asperger’s 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
JAIISH, Vol. 27, 2008 Continuum of Developmental Language Disorders
71
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 children’s 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.
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
social and language impairment of
disproportionate severity.
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
JAIISH, Vol. 27, 2008 Continuum of Developmental Language Disorders
72
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, 473-
486.
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 Children’s 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
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
73
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- can-
can 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. – doesn’t seem to realise the need to explain
what he/she is talking about to someone
who doesn’t 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 ‘isn’t 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
74
43. – gets into trouble because he/she doesn’t
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. – doesn’t 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 Children’s
Communication Checklist in a clinical setting.
JAIISH, Vol. 27, 2008 Analysis of Oral and Written Narratives of Language Impaired Children
75
Analysis of Oral and Written Narratives of Children with Language
Impaired Learning Disabilities
1Siddiqi 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
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
1Speech Language Pathologist & Audiologist, Speech and Language Therapy Services, Gaborone Botswana, 2Lecturer, Special Education (Speech Pathology & Audiology), Dept. of Educational Foundations, University of Botswana, email:[email protected],
JAIISH, Vol. 27, 2008 Analysis of Oral and Written Narratives of Language Impaired Children
76
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.
1Children 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.
_________________________________
2T-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).
JAIISH, Vol. 27, 2008 Analysis of Oral and Written Narratives of Language Impaired Children
77
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 subject’s 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.
LILD1 NLD1 LILD2 NLD2 LILD3 NLD3 Chronological age at time of testing (yrs)
11.11 11.10 10.3 10.5 10.8 10.6
IQ Below Average With in normal range
Below Average
With in normal range
Above Average
With in normal range
Grade 4 6 4 5 4 5 Medication Fixonase (for
hayfever) None None None Ritalin None
Previous Therapy
Regular school with Remedial Teaching
Regular school
Regular school Remedial Teaching
Regular school
Regular school Remedial Teaching Occupational Therapy (2 yrs)
Regular school
Current Therapy and Years in Therapy
Speech Therapy- 2.6 Occupational Therapy-2.0
Speech Therapy-2.0 Occupational Therapy-4.0
Speech 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 micro-
structural 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
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
78
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
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.
Intra-rater Inter-rater Transcription 96.2 % 94.9%
Coder 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
3The following formula was used: The number of complex, grammatically correct T-units X 100 Total number of T-units (Gillam and Johnston, 1992)
4Co-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].
JAIISH, Vol. 27, 2008 Analysis of Oral and Written Narratives of Language Impaired Children
79
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
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
T-units Words Word/ t-unit Type-Token Ratio Participants Narrative Types
Oral Written Oral Written Oral Written Oral Written
T1 10 7 77 57 7.7 8.1 3.6 2.8 T2 4 5 39 31 9.8 6.2 3.1 2.2 LILD1 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 T2 30 12 213 74 7.1 6.2 3.5 3.3 LILD 2 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 T2 5 4 37 29 7.4 7.3 3.0 2.6 LILD 3 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 T2 8 6 75 66 9.4 11 4.3 3.7 Control 1 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 T2 5 6 54 48 10.8 8 3.0 3.6 Control 2 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 T2 13 10 86 87 6.6 8.7 4.0 4.2 Control 3 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); T-
units- 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.
JAIISH, Vol. 27, 2008 Analysis of Oral and Written Narratives of Language Impaired Children
80
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
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
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
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
Adverb Prep Coverbs % Complex T-units Connectives
Participants Narrative Types
Oral Written Oral Written Oral Written Oral Written Oral Written
T1 0 0 3 3 3 0 20 0 1.3 0.9 T2 0 0 0 0 2 1 0 0 0.5 0.6 LILD 1 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 T2 1 0 1 1 2 0 10 8.3 0.9 0.5 LILD 2 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 T2 0 0 0 0 1 0 20 25 0.8 1.5 LILD 3 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 T2 1 0 2 1 2 2 37.5 50 0.6 0.3 Control 1 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 T2 0 0 0 0 0 2 20 16.7 1.6 0.2 Control 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 T2 0 0 1 1 1 2 0 10 1.3 1.2 Control 3 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); Adverbs-
number of adverbs; Prep- number of prepositions; Coverbs- number of coverbs; %complex t-units-
the 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.
JAIISH, Vol. 27, 2008 Analysis of Oral and Written Narratives of Language Impaired Children
81
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
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
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
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.
Written Productivity Syntactic Maturity
Vocab Mechanics Participants
Narrative Types Number of
Sentences Word/
sentence Morpheme/t-
unit Words>7 letters
% correct caps
% correct spellings
T1 1 57 8.9 1 16.7 80.7 T2 1 31 8 1 25.0 67.7 LILD 1 T3 1 45 9.6 0 50.0 73.3 T1 4 11.5 7.1 2 60.0 84.8 T2 4 18.5 6.6 9 61.5 58.1 LILD 2
T3 1 27 7.5 0 33.3 85.0 T1 1 43 7.8 2 33.3 90.7 T2 3 9.7 8.8 4 33.3 65.5 LILD 3
T3 1 19 6.3 2 75.0 89.5 T1 8 9.4 8.6 6 100 98.7 T2 6 11 12.8 6 100 97.0 Control 1 T3 10 9.7 9.3 2 41.5 100 T1 6 10.7 9.5 3 85.7 95.3 T2 4 12 8.8 1 100 100 Control 2
T3 2 16 12.3 3 66.7 93.8 T1 4 13.75 12.2 6 100 96.4 T2 5 17.4 10 12 100 100 Control 3 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; Vocab-
VocabularyInsert 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.
JAIISH, Vol. 27, 2008 Analysis of Oral and Written Narratives of Language Impaired Children
82
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 micro-
structure 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 children’s
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
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 subject’s 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
83
References
Applebee, A. N. (1978). The child’s concept of
story: Ages two to seventeen. USA: The
University of Chicago press.
German, D. J. & Simon, E. (1991) Analysis of
children’s 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, 1303-
1315.
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, 415-
425.
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 late-developing 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 Speech-
Language Pathology, 10(3).
JAIISH, Vol. 27, 2008 Phonological Processes in Typically Developing Children
84
Phonological Processes in Typically Developing Kannada Speaking
Children
1Sreedevi 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 child’s 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
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
children’s 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,
1Lecturer in Speech Sciences, Dept. of Speech Language Sciences, All India Institute of Speech and Hearing, Mysore-570006, email:[email protected], 2Junior Research Fellow, Dept. of Speech Language Pathology, All India Institute of Speech and Hearing, Mysore-570006,
JAIISH, Vol. 27, 2008 Phonological Processes in Typically Developing Children
85
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 de-
affrication 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, de-
affrication, 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 2-
5 years. She found that the 2 year olds used
phonological processes most frequently and the 4-
year 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.
JAIISH, Vol. 27, 2008 Phonological Processes in Typically Developing Children
86
Author Language Age Group
Common processes seen
1 Sunil, T J (1998) Kannada 3-4 years
Fronting, Cluster reduction, Initial consonant deletion and Affrication
2 Jayashree, U P (1999)
Kannada 4-5 years
Fronting, Cluster reduction and Stopping
3 Ramadevi et al (2002)
Kannada 5-6 years
Stridency deletion, De-aspiration and Retroflex fronting
4 Sreedevi et al (2005) Kannada 2-2.6 years
Retroflex fronting, Trill deletion, Depalatalization, Affrication, Stopping, Cluster reduction etc
5 Sreedevi, N (2008) Kannada 1.6-2 years
Retroflex fronting, Initial Consonant deletion, Vowel lowering, Trill deletion, Cluster reduction etc
6 Sameer, P. (1998) Malayalam 3-4 years
Cluster reduction, Final consonant deletion, Epenthesis and De-affrication
7 Bharathy, R (2001) Tamil 3-4 years
Epenthesis, Cluster reduction, Gliding, Nasal assimilation, Voicing De-affrication & Fronting.
8 Santhosh, M (2001) Hindi 3-4 years
Cluster reduction. Epenthesis, Fronting, Gliding, Metathesis Nasalization etc
9 Rajeev Ranjan (1999)
Hindi 4-5 years
Cluster reduction, partial reduplication and aspiration
10 Rahul Banjariya (2006)
Hindi 2-2.6 years 2.6-3 years
Retroflex fronting, Deaspiration, /h/ deletion, Gliding, Initial consonant deletion Affrication, Denasalization, Monothongisation, Devoicing etc
Table 1: Review of Phonological Processes in various Indian Languages.
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
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
child’s speech sample was carried out using the
following format:
Target word
Phonemic structure of target
Response of the subject
Phonemic Structure of the response
Phonological process used
/bΛs/ CVC /bΛt/ CVC Stopping
JAIISH, Vol. 27, 2008 Phonological Processes in Typically Developing Children
87
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 subject’s 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 -----------------------------------------------------
X 100
Total number of words spoken
i.e.
------------------------------------------------------------ X 100
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 observed
Total %
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
cluster 0.50%
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/,
/ræil/ etc. which end with the vowel /u/ in
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).
Total number of the same type of processes exhibited by all the subjects
Total number of target words spoken by
all the subjects
JAIISH, Vol. 27, 2008 Phonological Processes in Typically Developing Children
88
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.
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 today’s
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 child’s phonological skills
especially in the clinical population.
References
Bharathy, R. (2001). Development of Phonological
Processes in Tamil: 2-3 years. Master’s
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-, 4-
and 5- year old children. Language, Speech
and Hearing Services in Schools, 17,107-
114.
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 children’s
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 Master’s
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
JAIISH, Vol. 27, 2008 Phonological Processes in Typically Developing Children
89
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, 145-
160.
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.
Appendix 1
JAIISH, Vol. 27, 2008 Aetiological & Therapeutic Searches by Netizen Parents on Autism Spectrum
90
A Profile of Aetiological & Therapeutic Searches by Netizen
Parents/ Caregivers of Children on the Autism Spectrum
1Venkatesan 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, bio-
medical 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,
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
1Professor, Dept. of Clinical Psychology, All India Institute of Speech and Hearing, Mysore:570006, email:[email protected], 2Research Assistant, Dept. of Clinical Psychology, All India Institute of Speech and Hearing, Mysore: 570 006.
JAIISH, Vol. 27, 2008 Aetiological & Therapeutic Searches by Netizen Parents on Autism Spectrum
91
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’,
‘[email protected]’, 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.
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
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.
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
92
(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
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.
Sl. No.
Information Themes N % Rank
1 Advocacy 601 17.49 15
2 Behavior Problems 1488 43.31 4 3 Causes 1089 31.69 7 4 Conferences & Seminars 838 24.46 11 5 Consulting 1464 42.61 5 6 Contact Addresses: Institutions or Individuals 1679 48.78 3 7 Courses 272 7.92 18 8 Diagnostic Clarifications 1087 31.64 8 9 Enquiries 984 28.64 10
10 Greetings & Congratulatory Messages 1030 29.98 9 11 Information on Forthcoming Events 623 18.13 14 12 Information on Relevant Websites 581 16.90 16 13 Introductions/Welcome Notes 628 18.28 13 14 Notes, News, Views & Reviews 830 24.16 12 15 Peer Approval 1224 35.62 6 16 Self Introductions 1684 49.01 2 17 Send Articles 328 9.55 17 18 Therapies, Treatments & Techniques 1697 49.39 1 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
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
93
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
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).
Biological N Environmental N Sensory N Diet Based N
Allergies 28 Accidents 12 Auditory 30 Alpha Protein Deficit
13
Amygdala Neurons 14 Bad Parenting 6 Gustation 7 Diet disturbances 67
Autoimmune Disease 22 Birth Trauma 7 Hypersensitive 21 Enzyme Dysfunction
23
Brain Abnormalities 18 Emotional Trauma 10 Hyposensitive 9 Food Allergies 32 Chromosomal 16 Excess Hygiene 7 Olfactory 5 GFCF 111
Familial 8 Folic Acid 6 Tactile 7 Leaky Gut Syndrome
42
Food Additives 29 Perinatal Environment
4 Vestibular 4 Nutritional Deficiency
26
Genetic/Heredity 11 Pesticides 14 Visual 3 Poor Nutrition 27 Lead 33 Pollution 12 Maternal Antibodies 7 Poor Stimulation 19
Mercury 18 Prenatal Environment
11
Mercury Toxicity 78 Refrigerator Mother 7 Metabolic Disturbances 19 Social Factors 9 Missing Brain Protein-Cdk5
8 Stress 11
MMR Vaccine 99 Ultrasound 7 Oxidation Stress 10 Watching TV 27 Pesticides 16 Synapse Disconnection 4 Teratogens 13 Toxicity 6 Viral Infection 12 Vitamin Deficiency 24 N 22 16 8 8 Sum 493 169 86 341 Mean 22.4 10.6 10.7 42.6 SD 22.9 5.8 9.6 31.9 Sumsq 22139 2281 1570 21701 Variance 528.2 33.1 92.2 485.5
(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
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
94
(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 Category Types N Percent
A Nutrition-Diet Based Therapies
51 459 27.05
B Bio-medial Therapies 24 337 19.86
C Education-Behavior Based Therapies
40 287 16.91
D Sensory Therapies 50 199 11.73
E Alternative Medical Systems
40 189 11.14
F Medication/Drug Therapy
24 133 7.84
G Instrument Based Therapies
9 93 5.48
Total 238 1697
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,
Feingold’s 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
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. Oswal’s 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 bio-
medical therapies (N: 337; 19.86 %), education-
behavior based therapies (N: 287; 16.91 %),
JAIISH, Vol. 27, 2008 Aetiological & Therapeutic Searches by Netizen Parents on Autism Spectrum
95
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.
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
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.
JAIISH, Vol. 27, 2008 Auditory Memory and Sequencing in Children
96
Auditory Memory and Sequencing in Children Aged 6 to 12 Years
1Devi 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 one’s 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
1Lecturer, Dept. of Audiology, All India Institute of Speech and Hearing, Manasagangothri. Mysore-570006, email:[email protected], 2Audiologist, email:[email protected], 3Professor, Dept. of Audiology, All India Institute of Speech and Hearing, Manasagangothri, Mysore-570006
JAIISH, Vol. 27, 2008 Auditory Memory and Sequencing in Children
97
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 individual’s 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
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
98
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
Duncan’s 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.
Male Female Age (in years) Mean* SD CI Mean* SD CI
6–6;11 42.12 5.59 37-47 54.37 7.24 48-61 7–7;11 57.37 10.64 48-67 60.00 7.83 53-67 8–8;11 61.62 4.95 57-66 58.12 6.72 52-64 9–9;11 68.37 7.68 61-75 57.37 10.35 48-67
10–10;11 74.00 8.33 67-81 61.87 5.43 57-67 11–11;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
JAIISH, Vol. 27, 2008 Auditory Memory and Sequencing in Children
99
Male Female Age (in years) Mean* SD CI Mean* SD CI
6 – 6;11 22.12 7.98 15-29 25.12 10.13 16-34 7 – 7;11 32.75 9.31 24-41 27.25 7.75 20-34 8 – 8;11 33.37 10.47 24-43 23.62 3.50 20-27 9 – 9;11 34.25 9.96 25-43 26.87 8.21 20-34 10 – 10;11 43.37 16.93 29-56 27.37 5.26 22-32 11 – 11;11 39.00 5.07 34-44 33.75 12.05 23-45
Table 2: Mean scores, Standard Deviation (SD). and
Confidence Interval (CI) of the Auditory
Sequencing Subtest across age and gender. *
Maximum score = 104
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 Duncan’s 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
6-6;11 - 7-7;11 SD - 8-8;11 SD NSD - 9-9;11 SD NSD NSD - 10-10;11 SD SD NSD NSD - 11-11;11 SD SD SD SD NSD - Note: SD = 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.
Age in Years
6-6;11
7-7;11
8-8;11
9-9;11
10-10;11
11-11;11
6-6;11 - 7-7;11 NSD - 8-8;11 NSD NSD - 9-9;11 NSD NSD NSD - 10-10;11 SD NSD NSD NSD - 11-11;11 SD NSD NSD NSD NSD - Note: SD = significantly different, NSD = not significantly different.
Table 4: Significance of difference between means for
the auditory sequencing subtest cross ages.
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 Duncan’s 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
JAIISH, Vol. 27, 2008 Auditory Memory and Sequencing in Children
100
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
Auditory Memory Scores
Auditory Sequencing
Scores
Interpretation*
1 7/F 35 13 Deviant 2 8;6/F 53 25 Not Deviant 3 9/M 74 43 Not Deviant 4 10/F 27 18 Deviant 5 11/F 56 17 Deviant 6 12/M 24 15 Deviant 7 12/M 36 18 Deviant 8 12/M 61 24 Deviant 9 12/F 34 15 Deviant 10 12/M 65 26 Deviant
*Based on the confidence interval given in Tables 1 and 2.
Table 5: Scores of the auditory memory and
sequencing subtests obtained by the
children with learning disability.
Figure 1: Comparison of Auditory Memory scores of
children with Learning Disability with age
appropriate norms.
Figure 2: Comparison of Auditory Sequencing scores
of children with Learning Disability with
age appropriate norms.
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
the SCAP, do have an auditory memory and
sequencing problem. However, not all of them
have such a problem. Based on this finding, it is
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.
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.
JAIISH, Vol. 27, 2008 Auditory Memory and Sequencing in Children
101
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.
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 cognitive-
communication 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
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 School-
Aged Children. Journal of American
Academy of Audiology, 11(9), 467-474.
Locke, J., (1968). Discrimination learning in
children’s acquisition of phonology. Journal
of Speech and Hearing Research, 11, 428-434.
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.
JAIISH, Vol. 27, 2008 Regenerative Myringoplasty
102
Regenerative Myringoplasty – A Case Report
1Rajeshwari 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. It’s 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
1Reader, Dept. of ENT, All India Institute of Speech and Hearing, Manasagangothri, Mysore-570006, email:[email protected], 2 Reader, Dept. of Otorhinolaryngology, All India Institute of Speech and Hearing, Manasagangothri, Mysore-570006
Table: 1
Perforation of the tympanic membrane
Normal tympanic membrane
JAIISH, Vol. 27, 2008 Regenerative Myringoplasty
103
left ear revealed a large central perforation
measuring about six millimeters in diameter that’s
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.
JAIISH, Vol. 27, 2008 Regenerative Myringoplasty
104
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.
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
membrane and this would result in normal hearing
post procedure. Since it’s 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: Butterworth-
Heinemann.
JAIISH, Vol. 27, 2008 Universal Hearing Screening
105
Guidelines to Establish a Hospital Based Neonatal Hearing
Screening Program in the Indian Setting
1Ramesh 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.
John’s 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
India who have been implementing UNHS.
(Nagapoornima et al 2007 ; Malik M et al 2007 ,
Mathur NN et al , 2007). St. John’s 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
1Associate Professor, Dept. of Otolaryngology Head and Neck surgery, St John’s Medical College Hospital , Bangalore-560034, email:lavirams @ yahoo.com, 2Audiologist and Speech Pathologist, St John’s Medical College Hospital , Bangalore-560034, 3Audiologist and Speech Pathologist, St John’s Medical College Hospital , Bangalore-560034, 4Professor, Community medicine, St John’s Medical College Hospital , Bangalore-560034, 5Professor, Pediatrics, St John’s Medical College Hospital , Bangalore-560034.
JAIISH, Vol. 27, 2008 Universal Hearing Screening
106
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. Establishing a case for UNHSP.
2. Building a UNHS team.
3. Initial pilot run for 2 years.
4. Establishing a full fledged UNHSP.
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
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
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 member’s 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.
� 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.
Figure 1: Protocol used to screen for hearing impairment in neonates at St John’s Medical
College, Bangalore
JAIISH, Vol. 27, 2008 Universal Hearing Screening
107
� 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.
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. John’s 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.
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 at-
risk 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
JAIISH, Vol. 27, 2008 Universal Hearing Screening
108
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. John’s 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
JAIISH, Vol. 27, 2008 Universal Hearing Screening
109
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.
References
Joint committee on infant hearing screening (2007)
Position statement. Pediatrics Vol 120 (4),
898 – 921.
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.
JAIISH, Vol. 27, 2008 Universal Hearing Screening
110
Annexure 1
Dataset of St John’s Neonatal Hearing Screening Program
Name : Sex : DOB :
Duration of NICU stay :
Number Neonatal : Birth weight :
Gestation :
Hospital :
Address : Phone number :
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 Endotracheal and ventilation : Y/N
4 .Duration of Medications - Genatmycin : Amikacin : Furosemide :
other ototoxic medications :
5. Infection : Sepsis - Y/N Meningitis : Y/N
6.Hyperbilirubinemia: Peak level : Duration: Phototherapy:Y/N ExchangeTransfusion:Y/N
7. Pre term : Y/N LBW : Y/N SFD/IUGR:Y/N
Method used to screen Date Result
OAE : First screen
: Second screen
BOA
ABR
Maternal Problems
Natal History Delivery : Home / Hospital
: Normal/Breech/Caesarean/Forceps
Family History
Type of family : Nuclear/Joint :
Consanguinous marriage :
Literacy of : Father
: Mother
JAIISH, Vol. 27, 2008 Reversible Sudden Sensory Neural Hearing Loss
111
Reversible Sudden Sensory Neural Hearing Loss – A Case Report
1Sundara 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.
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.
1Reader, Dept. of Otorhinolaryngology, All India Institute of Speech and Hearing, Manasagangothri, Mysore-570006, 2Reader, Dept. of ENT, All India Institute of Speech and Hearing, Manasagangothri, Mysore-570006, email: [email protected].
JAIISH, Vol. 27, 2008 Reversible Sudden Sensory Neural Hearing Loss
112
• 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)
JAIISH, Vol. 27, 2008 Reversible Sudden Sensory Neural Hearing Loss
113
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
JAIISH, Vol. 27, 2008 Reversible Sudden Sensory Neural Hearing Loss
114
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.
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
(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, 190-
197.
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.
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 1½ 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.