LANGUAGE IN INDIA

MEAN LENGTH OF UTTERANCE AND SYNTACTIC COMPLEXITY IN THE SPEECH OF MENTALLY RETARDED

Shyamala Chengappa, Ph.D.
Sapna Bhat, Ph.D. Candidate
Jyothi Hiwarale, M.Sc. Candidate

1. THE IMPORTANCE OF COMMUNICATION FOR MEETING NEEDS

Communication is the most essential base for getting along with others and for satisfying both intra and interpersonal needs. The pattern of language development is sequential universally, unless and until interference is caused due to any sensory or motor deficits. Cognition also plays a very important role in the language acquisition. In cognitively impaired children, not only the overall language development is retarded but also the rate at which the language development occurs is disproportionately slower rate than that of their intellectual development.

2. ASSESSMENT OF THE MENTALLY RETARDED CHILDREN

The assessment in mentally retarded children is specifically a cumbersome task. Discrepancy is always seen between the competence and performance of the child. Thus there is need for systematic and structural approach to study the language development from the single word utterance to adult model. Brown has contributed maximally to this field of language assessment. In 1973, he developed five stages of sentence construction (MLU=Mean Length of Utterance) that seem parallel over all language development. From the time it has been introduced, it has been most popular assessment tool in the west, while it has not been studied in India at all. Hence, the present study was undertaken to see if any predictive trend could be established.

3. THE AIM OF THE PRESENT STUDY

Present study aimed at comparing the MLU values of the normal (4-11 yrs) with that of mentally retarded and arrive at a relation between intelligence quotient (IQ) and sentence complexity in mentally retarded population.

4. BROWN'S STUDY

Brown (1973) first found that at comparable MLU's children used the same grammatical structures upto the MLU of about 4. He observed that chronological age was not a good predictor of language development, this he said after analyzing longitudinally, the speech samples of 3 children-Adam, Eve and Sarah. Brown's stages are designated with Roman numerals and are as follows:

Stage I: Semantic roles and syntactic relations (MLU 1.0 - 2.0 morphemes or 1.75 morphemes). Here child puts noun-verb sequences together.

Stage II: Grammatical morphemes and modulation meaning (MLU = 2.0 - 2.5 with average of 2.25 morphemes). The child starts to change word endings to portray grammar.

Stage III: Modalities of simple sentences (MLU = 2.5 - 3.25 with average of 2.75 morphemes). The child begins to use questions and imperatives.

Stage IV: Embedding (MLU = 3.25 - 3.75 with average of 3.5 morphemes). The child begins to use complex sentences.

Stage V: Co-ordination (MLU = 3.75 - 4.25 with average of 4 morphemes). The child may use connectors and more functions.

Brown did not imply that the stages are discrete, but rather that the linguistic development is continuous and that the stages allow comparison and characterizations at different levels of language proficiency.

5. CONCLUSIONS FROM THE CURRENT LITERATURE

Studies of language and cognitive development in the retarded children suggest that, in the overwhelming majority of cases, levels of language ability are at or below the same child's level of functioning in the other cognitive domains. (Beegly and Cicchetti 1987; Miller, 1988).

One can conclude from the existing literature that the mentally retarded children show inadequacy for language in some or all of following ways:

1. Apparent inability to understand the spoken word
2. Impoverished vocabulary
3. Constant repetition of a few words or phrases (i.e.) perseveration
4. Parrot - like copying of adult speech (echolalia)
5. Poor articulation
6. Primitive grammatical construction in the speech

By and large, children with higher mental endowment posses better language ability.

The traditional belief has been that language of the mentally retarded develops in slow motion. The theory of the quantitative delay of the language of the mentally retarded was clearly articulated by Lenneberg (1967).

Some researchers have also noted qualitative differences in the language used by the mentally retarded. Their use of morphemes differ (Menyuk 1971) and as mental age increases, some differences are also observed in the use of inflectional forms (Schiefelbusch, 1972). Ryan (1977) found that vocabulary improved more quickly then did the grammar in the retarded. A study of semantics (Semmel, Banett and Bennett, 1970), indicated that when retarded and normal subjects of the same mental age are compared on the word-association tasks, the retarded fail to shift from synonyms to antonyms at the same mental ages of the normals. Karlin and Strazulla (1953) observed that many of mentally retarded children show poor attention span, accompanied by easy fatigability and distractibility. Majority of these studies have been carried out on down's syndrome children and very little information is available regarding MLU and syntactic complexity in mentally retarded.

6. METHODOLOGY ADOPTED IN THIS STUDY

Subjects: 20 normal children in the age range of 4-11 years and 10 mentally retarded children with comparable mental age were studied. These 10 mentally retarded children were divided into two groups based on the degree of retardation i.e., 5 subjects each were taken having mild and moderate retardation. All the subjects had a mean mental age of 8.5 years and had Kannada as their mother tongue.

Out of 20 normals, 7 of them were females and 11 males. Among the mentally retarded group 5 were females and 5 males.

Primary mode of communication in all the children was verbal but gestures and pantomime along with verbal communication was noted. All subjects were attending school. All the subjects in the study had minimal exposure to therapy (less than 8 months) or no therapy at the time of the study.

Spontaneous speech, elicited/narrated speech using pictures was used to collect the sample of the language from each child.

All the responses were audio taped. Diary of the session of speech interaction was maintained. Each session lasted 20-30 minutes or longer depending on the child's comfort. Each child was tested to elicit 100 utterances each in spontaneous speech and elicited speech verbal interactions were pursued between investigator-child, teacher-child and teacher-child-investigator. The rules for computing MLU are given in appendix-A.

7. RESULTS AND DISCUSSION

The results are discussed under following headings:

I. MORPHEME LENGTH

Table 1 shows distribution of MLU (words and morphemes) as a function of age as well as the mean MLU (words) and MLU (morphemes) in the normal subjects.

Thus, it is apparent that MLUw (words) was always less than MLUm (morphemes). The group of normal children in age range 4.5 to 10.11 years had mean age of 7-6 years and mean MLU (m) of 5.52 and mean MLU(w) was 4.17.

MLU did not increase with increase in age for both words and morphemes. This finding shows good agreement with Miller and Chapman's (1981) conclusion that the variability in MLU increases after 5 years of age.

B. Results in mentally retarded:

The group of the mentally retarded population ranging in IQ from 43-68.5 with mean IQ of 54.2 had a mean MLU(W) of 1.67. The mean MLU(M) for this group was found to be 2.44. thus MLU did not reduce with reduced IQ for both words as well as morphemes (Table 2).

In both mildly and moderately retarded group, MLU(M) was higher than MLU(W). Mildly retarded group has both MLU(W) and MLU(M) higher than moderate group. The lack of reduction in MLU with reduction in IQ in both the mildly and moderately mentally retarded group may be due to increase in MLU(W) of S5 as compared to that of S4 in the mild group and also MLU(W) value of S8 was higher than preceding subject S7 in the moderate group (see table 2). This variation could not be explained by the variables in the present study. Extraneous variables like age at which intervention was done, home training, etc. could not be controlled and may have contributed. They could have probably contributed to enhancement of MLU(W) scores of the subjects S5 and S8 in the mildly and moderately retarded group respectively.

Such a variation was not seen in MLU(M). That is MLU(M) reduced with reduction in IQ when mildly and moderately retarded group were considered separately.

C. Comparison between normals and the mentally retarded population:

On comparison of the means from table 1 and table 2 for the normals and the mentally retarded population are deficient as compared to the normals. That is the MLU(W) in this population was 1.67 and that in normal subjects MLU(W) had a value of 2.52. MLU(M) in normals was as high as 4.17 and in retarded it was reduced to 2.44. This finding is in agreement with previous finding by Klee et.al (1989), where predicted MLU of language-impaired group was lower than normals across the age range studies.

II. Syntactic complexity:

The results obtained were analyzed under 3 sub-categories. They are

1. The number of single word and multiple word utterances
2. The order and frequency of lexical/grammatical categories in the sample analyzed
3. The arrangement of these lexical/grammatical categories within an utterance

A. Distribution of single and multiple word utterances in normals:

Results obtained in the normals revealed that though normals use a large number of single word and two word utterances, their language sample also contains three, four and five word utterances. On an average, proportion of single word utterances was found to be highest.

B. Results in the mentally retarded population:

In mentally retarded population, single word utterances occurred most frequently. However, longer utterances (with two or more words) were less frequently used in all the retarded subjects. On comparing the mild and moderate groups, occurrence of one word and two word utterances was found to be almost equal in both the sub groups.

Considering the IQ, the mildly retarded group showed reduced frequency of occurrence of two or multiple word utterances, with the decrease in IQ. On the contrary, single word utterances increased with reduced IQ. For moderately retarded group, no such conclusions could be drawn. However, excluding one subject, utterances consisting of 3 or more words decreased with decrease in IQ. Such conclusions were not true for single-word and two-word utterances.

C. Comparison between normals and the mentally retarded population:

The normals and the mentally retarded group were compared in terms of single and multiple word utterances. The 3-4 word utterances were found to increase with age in the normal children with the frequency of one and two word utterances reducing with increasing age. Unlike normals, only single word utterances were most frequent in the retarded population. However, if means are considered, it is true for the normal subjects also, i.e., occurrence of single words is highest. Normal subjects' speech consisted of more longer utterances than the retarded group. The longer utterances in the normals consisted of 11 words which were used only once by one subject. In the mentally retarded population, the longest utterance consisted of 6 words only which were used by two subjects.

D. The order and frequency of lexical/grammatical categories in the normals and mentally retarded:

In normal subjects, occurrence of nouns was the maximum. Other grammatical categories showed decreasing frequency of occurrence in the following order:

Nouns, verbs, pronouns, adjectives, adverbs, kinship terms, conjunction, negatives, quotatives, interrogatives, reduplicative and onomatopoeia.

In mentally retarded children also nouns were most frequent followed by verbs. The following order was noticed in decreasing order of frequency:

Nouns, verbs, pronouns, adjectives, adverbs, kinship terms, conjunction, negatives, quotatives, interrogatives, reduplicative and onomatopoeia.

Comparison of mild and moderate retarded group showed that both sub-groups used nouns more frequently as seen in normals. Mildly retarded group tended to use verbs, adverbs, quotatives, interrogatives, kinship terms and reduplicative more as compared to the moderately retarded group. Nouns, pronouns, adjectives, negatives and conjunctions were used more by the moderately retarded population as compared to the mild ones.

8. SUMMARY CONCLUSIONS

In summary, following results were obtained:

1. There is a quantitative and qualitative difference between normals and the mentally retarded population though qualitative differences were subtler.
2. There were differences noticed within the mentally retarded group-mildly retarded group performed differently from the moderately retarded group on almost all grammatical categories except nouns. Thus we see that speech language delay and deficits were seen in the mentally retarded group.

This study is first of its kinds in an Indian linguistic context. Such studies with larger sample size would help in understanding the clinical picture of linguistic deficits present in mentally retarded and their relation to cognitive abilities.

REFERENCES

Beeghly M. & Cicchetti D. (1987) cited by D. Thal, E. Bates; V. Bellugi (1989) "Language and Cognition in two children with Williams Syndrome". Journal of Speech and Hearing Research, 32, 489-500.

Brown R. (1970). Psycholinguistics - Selected papers. The Free Press, New York.

Brown R. (1973). A First Language. Harvard University Press. Cambridge.

Karlin I. W. & Strazzulla M. (1952). Speech and Language Problems of Mentally Deficient Children. Journal of Speech and Hearing Disorders, 17, 286-294.

Klee, T; Schaffer, M; May, S; Membrino, I; and Mougey, K (1989). A Comparison of the Age-MLU Relation in Normal and Specifically Language-impaired Preschool Children. Journal of Speech and Hearing Disorders, 54, 226-233.

Lenneberg E. H. (1968) Biological foundations of language. In D. M. Moorehead & A. E. Moorehead (Eds) (1976). Normal and Deficient Child Language, pp. 200-207. University Park Press, Baltimore.

Menyuk P (1971) The internal language of children's syntax. Cited in K. Nelson (Ed) (1980). Children's Language. Vol.2, Gardner Press Inc., New York.

Miller J. F. (1988) cited in D.Thal, E. Bates, and U. Bellugi (1989). Language and Cognition in Two Children with Williams Syndrome. Journal of Speech and Hearing Research, 32, 489-500.

Miller, J. F., and Chapman, R. (1979) Significance of syntactic development in children. Cited by J. F. Miller (1981). Assessing Language Production in Children - Experimental Procedures. University Park Press. Baltimore.

Ryan J. (1977) Language in Developmentally Disabled Children. Cited in J. G. deVilliers; P.A.deVilliers (1973). Language Acquisition, pp. 228-272. Harvard University Press, London.

Schiefelbusch, R. (1972) Development and Disabilities in syntax. Cited in R. Shiefelbusch, & D.D.Bricker (Eds) (1982) Early Language: Acquisition and Intervention, pp. 300-307. University Part Press, Baltimore.

Semmel M. I.; Barritt L.S.; Bennett S.W. (1970) cited in J.G.DeVilliers and P.A. deVilliers (1978). Language Acquisition. Harvard University Press, London.

APPENDIX-A

RULES FOR COMPUTATION OF MEAN LENGTH OF UTTERANCE [MLU]
(Adaptation from Brown (1970)

1. The first 100 utterances were transcribed. Utterance during story narration was mandatorily included in the count.
2. Unintelligible or partially intelligible utterances were omitted from the count.
3. Stutterings (Mark by repeated effort) at a single word) and all repetitions were counted as one word. Repetition for emphasis should be counted as two words.
4. Fillers such as mm or oh are not counted, but no, yes, etc. was counted as words.
5. All compound words were counted as two words if the child used the constituent morphemes separately in two different linguistic context - Eg. Birthday.
6. All inflections (possession, plural, tenses) were counted as separate morphemes.
7. Imitations and elliptical answers to questions which gave the impression that the utterance would have been more complete if there had been no eliciting questions (Eg. What is that? `My box') were counted.
8. Rote passages such as nursery rhymes, songs or prose passages which have been memorized and which may not be fully processed linguistically by the child were omitted.
9. All partial utterances which are interrupted by outside events or shift in child's focus were excluded.
10. MLU was calculated using the following formula: MLU (W/M) = Number of words or morphemes÷100

*** *** ***

HOME PAGE | A Contrastive Analysis of Hindi and Malayalam | The Language Education Policy of Gujarat Government | Mean Length of Utterance and Syntactic Complexity in the Speech of Mentally Retarded | Towards an Understanding of Sentence in Science in an Indian Language Context | CONTACT EDITOR