Onset of Talking with Treatment in a Non-Verbal Nine-Year Old Annual Convention of the American Psychological Association Chicago, IL, USA August 15-19, 1997 ABSTRACT An eight year 4 month old non-speaking boy was given a parent-implemented cognitive-behavioural treatment program to reduce his non-compliant disruptive behaviour and to stimulate expressive language. The intervention, lasting 16 months, resulted in the emergence of a functional vocabulary, increased mean length of verbal utterance, and improved scores on the Griffiths test of intelligence. Results are discussed in light of a sensitive period for language acquisition, the role of expressive language for the assessment of intelligence, and the potential clinical utility of the treatment program. Lenneberg (1967) first proposed a critical period for language acquisition, i.e. a period of peak learning sensitivity followed by a substantial decline of language learning abilities, and by underlying maturational mechanisms. He cited indirect evidence from deaf and retarded children to suggest that language learning reached a plateau at puberty when biological maturity occurs. The notion of a sensitive period for language acquisition is widely accepted (Hurford, 1991). The ethical impossibility of experimentally manipulating age of exposure to a first language obviously restricted the empirical study of this hypothesis. Thus, rare naturally occurring cases of children deprived of language have served as the data base for making inferences about the relative role of nature for first language acquisition. The famous case of Genie, deprived of normal linguistic, social, and environmental input from 20 months of age until her discovery at 13 1/2 years, provided detailed documentation over a seven year period of language learning in a child who exceeded the sensitive period for expressive language acquisition (Curtiss, 1977). Curtiss (as cited in Pinker, 1994) also reported on the case of Chelsea, a deaf woman isolated from any linguistic input until 32 years of age. Following intensive therapy by a rehabilitative team, she learned two thousand words of spoken vocabulary but could not manage to learn the English syntax, rendering her speech strikingly abnormal. Studies of maturational change in the ability to learn have examined various domains including bird song learning to imprinting in animals and second language learning and American Sign Language in humans. In general, this research supports the notion of a sensitive period for language acquisition (Newport, 1991). Young children presenting delayed language acquisition provide still another avenue for studying the sensitive period hypothesis. It appears that the longer children experience success using non-verbal forms of communication, the more difficult it may be to achieve verbal expression (Lalinec, 1995; Fischel, Whitehurst, Caulfield & DeBaryshe, 1990; Zelazo, 1989). Moreover, there is a strong association between behavioural problems and delayed language acquisition (Lalinec, 1995; Zelazo, 1989; DeBaryshe & Caulfield, 1987; Zelazo & Kearsley, 1984; Zelazo, Kearsley, Stiles & Randolph, 1985; Whitehurst, 1985; Funk & Ruppert, 1984) rendering cooperation or compliance with the teacher problematic. Whitehurst, Fischel, Arnold & Lonigan (1992) question the efficacy of traditional pre-school speech and language services and report the effectiveness of their own cognitive-behavioural picture book program for children with mild expressive language delays. Zelazo, Kearsley and Ungerer (1984) set forth a parent-implemented treatment program for pervasively developmentally delayed and autistic children, based on the conversion of an oppositional style to a compliant one. The effectiveness and efficiency of this program was demonstrated for expressive language, play and intelligence test scores with pre-school-aged children (Kruzynski, Lalinec, Zelazo, Reid & Kay, 1996; Kruzynski, Lalinec, Zelazo, Reid & Thompson, 1996; Lalinec, Zelazo, Rogers & Reid, 1995; Laplante, Zelazo & Kearsley, 1991; Zelazo, et al., 1996; Zelazo, 1989; 1997a; 1997b; Zelazo & Kearsley, 1984). We examined the effectiveness of this program (Zelazo, Kearsley & Ungerer, 1984) for stimulating expressive language in a mentally retarded child at an age that approached the latter end of the sensitive period for acquisition - late pre-puberty. We examined also whether late language acquisition can reduce delays on conventional intelligence tests. Research with delayed autistic toddlers has shown that expressive language acquisition can increase developmental tests scores (Lalinec et al., 1995; Zelazo, 1989; 1997a; 1997b; Zelazo & Kearsley, 1984). We examined whether there is a comparable effect for the acquisition of expressive language on intelligence test scores at a much later age. Participant Presented here is a case report of a severely mentally retarded eight year, four month old boy, diagnosed with Williams Syndrome, Attention Deficit-Hyperactivity Disorder and Global Dysphasia, who was enrolled in a parent-implemented treatment program devised by Zelazo et al. (1984). Physicians told his parents that he would not speak. He was taught a handful of signs and pointed to pictures to communicate. Despite 4 years of extensive specialised services, he showed minimal effective communication, functioning at the expressive level of an 18 month old (D.Q.=17) on the language subscale of the Griffiths test of intelligence (Griffiths, 1970). The child is from a French speaking family and the therapy with both the family and child were conducted in French. Procedure The treatment program, Learning to Speak: A manual for parents (Zelazo et al., 1984), requires twelve minutes of daily therapy in a structured setting and continuous systematic generalisation to other contexts during naturally occurring daily activities. The therapy sessions encourage developmentally appropriate, behaviourally structured interactions. Non-compliant, resistant responses to adult requests are extinguished and compliance is shaped through contingent positive reinforcement, first with actions, then with words. Compliance to requests is required for verbal imitation, a necessary but insufficient condition for expressive language development (Speidel & Nelson, 1989). Treatment sessions were conducted for 5 to 7 days each week. Behaviours shaped during the structured sessions were generalised systematically to other contexts during the daily interactions. Essentially all direct treatment was administered to the child by the parents at home and all testing was done by the examiner in the laboratory. Parents provided 12 consecutive months of therapy during the first year, but only 8.5 months during the second year due to the need for vacations and relief from the pressures of being therapists and parents. Parents were observed by a therapist weekly for the first three months and bi-weekly subsequently to adjust behavioural and language goals, coach the parents about their teaching technique, solve behavioural management problems and provide support. During all laboratory visits, observations of both parent-implemented and therapist-implemented teaching sessions were recorded on videotape. Behavioural Measures Behavioural observations of the child's expressive language were reduced from the first 12 minutes of the therapist-implemented videotaped bi-weekly sessions and averaged every two months over 16 months of intervention. Three mutually exclusive categories were operationally defined: clear words, word approximations and vocalisations. (1) A word was coded if all appropriate sounds comprising the word were heard distinctly. (2) A word approximation was coded if at least one syllable or a fraction of a syllable of the requested word was distinguishable. The syllable had to occur either at the beginning or at the end of the word. (3) A vocalisation was coded if the child produced a vocal utterance that did not fit into the above categories and was not a reflexive reaction such as a burp, sneeze, cough, or grunt. The coding scheme also distinguished whether the expressive verbal utterances were spontaneous or imitative in nature. The following three examples illustrate the difference: (1) Imitative language: Mother: "What's this?" Child: No response Mother: "Apple" Child: "Apple" (2) Spontaneous language: Mother: "What's this?" Child: "Apple" (3) Spontaneous language: Child: "Mommy!" (grabbing at her) Reliability Reliability was indicated by a percentage score based on the number of agreements divided by the number of agreements plus disagreements. Two independent coders reduced a third of the 25 videotaped sessions conducted by the therapist, reaching 86% inter-rater reliability for words, 91% for word approximations, 78% for vocalisations. Baseline Baseline was established using the more conservative strategy of comparing test sessions from the first two months of treatment with later ones rather than data from observations gathered before the intervention was implemented (Kazdin, 1982). Measures Used for Analysis Total number of words (spontaneous and imitative) produced during the first 12 minutes of the therapist-implemented videotaped session were tallied and used for evaluation. Number of different words produced during the videotaped sessions was tallied and used for comparison with the three conventional vocabulary measures (PPVT-R, EOWPVT-R and MCDI-R). Length of Utterance The mean length of utterance (MLU) was calculated based on the number of spontaneously produced words in each utterance divided by the number of utterances. Two-word utterances were coded if the child spontaneously produced either a two-word sequence or one word and one word approximation in sequence; three-word utterances were coded if the child spontaneously produced a three-word sequence or two words and one word approximation in sequence; and so on. Measures of Developmental Status All standardised measures of developmental status were administered before treatment (T0), after 12 months of the parent-implemented intervention (T1) and after two years (T2). Griffiths Developmental Scales This widely used conventional test of intelligence provides both a quotient for global development and for six subscales that cover the main areas of development. Mental age (MA), obtained from the total number of items passed on each subscale, is divided by the child's chronological age (CA) and multiplied by 100 to produce developmental quotients (DQ). The average of all subscales provides a General Intelligence Quotient. A measure of rate, the Intervention Efficiency Index (IEI), was used as the principal dependent variable and indicates how rapidly developmental change has occurred over the period of intervention, using a value of 1.0 per month of chronological age. The IEI factors out maturation and time of testing from the effectiveness of treatment (Bagnato and Neisworth, 1980). Raw scores were used for the three vocabulary measures because the child was older than ceiling ages and, therefore, could not be assessed in terms of percentile rank. They provide a common unit of comparison among the measures and with number of different words reduced from the videotaped test sessions. Peabody Picture Vocabulary Test-Revised The PPVT-R (Form l; Dunn et Dunn, 1981) measures receptive vocabulary by pointing to the picture labelled by the examiner out of four printed pictures presented. Expressive One Word Picture Vocabulary Test-Revised The EOWPVT-R (Gardner, 1990) measures expressive vocabulary by naming printed pictures presented by the examiner. MacArthur Communicative Development Inventory-Revised The MCDI-R (Fenson, Dale, Reznick, Thal, Bates et al., 1993), a parental report, provides a more representative sample of expressive vocabulary not influenced by performance factors and organised in semantic categories. Results demonstrate the acquisition of a working vocabulary and multi-word sentences at an age that approaches the end of the presumed biologically determined sensitive period for expressive language development (Hurford, 1991). Moreover, results show improved scores on conventional measures of intelligence. No statistical analyses were performed. In single-case experiments, visual inspection is the more commonly practiced method of comparison (Kazdin, 1982). Kazdin outlined four criteria for proper analysis of single case data through visual inspection: change in means, change in the level of performance, change in trend and the latency of change. Behavioural Measures Total Number of Words  Figure 1 illustrates the increase in the total number of words (both spontaneous and imitative) produced during the test session. The child averaged 5 spoken words over the 12-minute session during the first treatment period (baseline) that increased to a mean of 49 words after 14 months of intervention (Period VII).
Number of Different Words A vocabulary of ten different clear words, produced over the baseline test sessions, increased to 73 words by Period VIII. Length of Utterance Mean Length of Utterance (MLU), the dominant form of spontaneous speech, was 1.23 after one year of intervention reflecting essentially single word utterances with occasional two-word utterances. The mean number of spontaneous single-word utterances produced over 12-minute test sessions more than doubled during the first intervention period from a mean of 4 at baseline to a mean of 10.7 at the end of Period I. Single word utterances continued to increase more than six fold by Period VII (M=31.7). Two-word spontaneous utterances appeared two months after the beginning of treatment with an average of two per session and remained stable for six months. After 14 months of treatment, two-word utterances sharply increased to an average of nine per 12-minute session. Three-word spontaneous utterances first appeared after ten months of treatment and reached an average of three per 12-minute session by the 14-month observation. Twelve months after the beginning of treatment, the child spontaneously produced one four-word utterance during the 12-minute sessions (" Je serre la main" - I shake the hand). 
These changes, shown in Figure 2, indicate a clear upward trend in the child's level of performance. Closer inspection of sentence structure in the child's spontaneous speech reveals the acquisition of basic word order
(Table 1) e.g., "bébé bois lait" - baby drink milk; "Maman tient bébé" - mommy hold baby, which both follow the normal subject-verb-object word order.
Measures of Developmental Status Griffiths Developmental Scales After one year of treatment, overall gains were sufficient to move this child from the severely to the moderately retarded range, a category change further enhanced at the second follow-up testing. The continuous increase in General Intelligence Quotient, from 32 on initial testing to 41 following one year of treatment (IEI of 1.1) to 45 on the second year follow-up (IEI of 1.0), shows a clear upward trend for the child's overall mental development. Given an initial IQ of 32 in global development (T0), the expected rate of development would be an IEI of 0.33, i.e. a gain of 4 months in mental age over the course of 12 months. Instead, the child's global development over the first year was that of a normally developing child (IEI rate of 1.0), i.e. he gained 12 months over a 12-month period. By the second year follow-up (T2), his overall IQ increased by 13 points. 
Table 2 contains the improvements on all scales of the Griffiths intelligence test. The child's progress on most of the subscales far exceeded expectations associated with mental retardation. Substantive gains were made on Hearing and Speech and Practical Reasoning, the two language-based scales of particular interest for predicting later intellectual functioning (Largo, Graf, Kundu, Hunziker & Molinari, 1990). The expected Hearing and Speech developmental rate (IEI of 0.17) improved to 1.6 following the first year of treatment. By the 2-year follow-up (T2), the child progressed at a normal rate (IEI of 1.1) and recorded an 18-point increase in IQ. On Practical Reasoning, his estimated IEI rate of 0.25 increased to 0.8 at T1 and 1.1 at T2. His sharp increased rate of development is associated with an IQ increase of 18 points over the two years of treatment. Sizeable gains were measured on non language-based scales such as the Personal-Social and Locomotor scales also. The Personal-Social scale, which measures the child's competence and independence in daily activities, had an expected IEI rate of 0.25, but increased on both T1 and T2, to an IEI rate of 1.5. His IQ on the Personal-Social subscale increased by 24 points by the second follow-up testing. The Locomotor scale, a measure of gross motor ability, had an expected IEI rate of 0.56 that increased to rates of 1.1 and 1.0 at T1 and T2 respectively. His IQ on the Locomotor subscale increased by 13 points over the same period. The remaining scales, Eye-Hand Coordination and Performance scales, measuring fine motor skills, speed of execution and oculo-motor coordination reveal developmental progress rates closer to expected given their initial IQ. Peabody Picture Vocabulary Test-Revised Gains were minimal on the PPVT-R receptive vocabulary scale but nevertheless indicate an upward trend as the child moved from 30 items identified by pointing on T0 to 37 on T1 and 42 on T2. Expressive One Word Picture Vocabulary Test-Revised There was a sharp increase in expressive vocabulary on the EOWPVT-R, from no labelling at all on T0 to 6 pictures on T1 and 27 pictures on T2. MacArthur Communicative Development Inventory-Revised The parents confirmed the gains in expressive vocabulary, indicating a dramatic increase of 237 different spoken words. The child's vocabulary increased from 29 words at initial testing to 179 words by the first year follow-up and 266 words by the end of the second year of treatment. Results extend the efficacy of the cognitive-behavioural treatment program for stimulating expressive language, set forth by Zelazo, Kearsley and Ungerer (1984), from toddlerhood to the tail end of the sensitive period - the onset of puberty. These data reveal clear acquisition of spontaneous spoken vocabulary and simple spontaneously generated communicative sentences in an 8 year 4 month old boy who did not speak previously. He gradually increased the quantity and complexity of his expressive language, from word approximations to single words, to two- and three- or more word utterances. Moreover, his spontaneous sentence structure contained appropriate word order. These results are consistent with the route followed by young first and second language learners (Newport, 1991; 1990). These results also accent the confounding role of expressive language for the assessment of mental ability using conventional measures of intelligence. A nine-point gain in General IQ over the first year of treatment produced a category change from severe to moderate mental retardation. Despite a reduction in the intensity of treatment during the second year including a two-month vacation, this improvement was further secured with an additional four-point gain in General IQ. Treatment specifically targeted expressive language and both language-based scales on the Griffiths (Hearing & Speech and Practical Reasoning) reveal developmental rates indicative of normal intelligence after two years of intervention (IEI rate of 1.1). During the first year, rates on the Hearing & Speech scale progressed 60% faster (IEI rate of 1.6) than predicted by normal intelligence, a necessary condition if delays are to be reduced. The emphasis on converting an oppositional style to a cooperative one serves as the cornerstone of this treatment program and, although the program emphasizes spontaneous language, gains were also measured in domains of development not directly targeted by the treatment sessions such as the Locomotor scale (IEI rate of 1.1) and the Psycho-Social scale (IEI rate of 1.5). These changes appear to occur because compliance is a core behaviour that is central to a number of areas of development (Rogers, 1997; Zelazo, 1997b). As with all single case studies, the generalisability of these results is questionable. It is necessary to apply the treatment program, Learning to Speak: A Manual For Parents (Zelazo et al, 1984), to other children approaching puberty to establish whether the program is effective with children 9 to 11 years of age. Until the sample is increased in number, the study merely raises the possibility that the program is generalisable and that pre-adolescents can learn to speak. Moreover, it appears that the acquisition of spoken language at pre-adolescence can improve intelligence test scores. Bagnato, S.J. & Neisworth, J.T. (1980). The intervention efficiency index: An approach to pre-school program accountability. Exceptional Children, 46(4), 264-269. Curtiss, S. (1977). Genie: A psycholinguistic study of a modern day "wild child". New York: Academic. DeBaryshe, B.D. & Caulfield, M.S. (1987, April). Behavioral correlates of early expressive language delay. Paper presented at the Society for Research in Child Development Biennial Meeting, Baltimore, USA. Dunn, L.M. & Dunn, L.M. (1981). Peabody Picture Vocabulary Test-Revised. Circle Pines, Minnesota: American Guidance Service. Fenson, L., Dale, P.S., Reznick, J.S., Thal, D., Bates, E. Hartung, J.P., Pethick, S., & Reilly, J.S. (1991). MacArthur Communicative Development Inventories: User's guide and technical manual. San Diego: Singular Publishing Group. Fischel, J., Whitehurst, G., Caulfield, M. & DeBaryshe, B. (1990). Language growth in children with expressive language delay. Pediatrics, 83(2), 218-227. Funk, J.B. & Ruppert, E.S. (1984). Language disorders and behavioral problems in preschool children. Developmental and Behavioural Pediatrics, 5, 357-360. Gardner, M.F. (1990). Expressive One-Word Picture Vocabulary Test-Revised. Novato, California: Academic Therapy Publications. Griffiths, R. (1970). The Abilities of Young Children. London: Child Development Research Center. Hurford, J.R. (1991). The evolution of the critical period for language acquisition. Cognition, 40, 159-201. Kazdin, A.E. (1982). Single-Case Research Designs. New York: Oxford University Press. Kruzynski, A.K., Lalinec, C., Zelazo, P.R., Reid, C. & Kay, D. (1996, April). Parent-implemented treatment and improved play in delayed toddler. Poster session presented at the 10th International Conference on Infant Studies, Providence, USA. Kruzynski, A.K., Lalinec, C., Zelazo, P.R., Reid, C. & Thompson, S. (1996, August). Impact of a parent-implemented treatment on free play and directed play in delayed toddlers. Poster session presented at the XIVth meeting of the International Society for the Study of Behavioural Disorders, Quebec, Canada. Lalinec, C. (1995). Impact of a 6-month parent-implemented behavioural intervention for children with developmental delays of unknown etiology. Unpublished doctoral dissertation, McGill University, Montreal, Canada. Lalinec, C., Zelazo, P.R., Rogers, C.-L. & Reid, C. (1995, August). Developmental delay: impact of a parent-implemented behavioural treatment. Poster session presented at the annual meeting of the American Psychological Association, New York, USA. Laplante D.P., Zelazo, P.R. & Kearsley, R.B. (1991). The effect of a short-term parent-implemented treatment program on the production of expressive language in 32 month-old toddlers. Society for Research in Child Development Biennial Meeting Proceeding, 336. Largo, R.H., Graf, S., Kundu, S., Hunziker, U. & Molinari, L. (1990). Predicting outcome at school age from infant tests in normal, at risk and retarded children. Developmental Medicine and Child Neurology, 32, 30-45. Lenneberg, E. (1967). Biological Foundations of Language. USA: Wiley. Newport, E.L. (1990). Maturational Constraints on Language Learning. Cognitive Science, 14, 11-28. Newport, E.L. (1991). Constrasting concepts of critical period for language acquisition. In S. Carey & R. Gelman (Eds.), The Epigenesis of Mind (pp. 111-130), Hillsdale, NJ: Erlbaum. Pinker, S. (1994). The Language Instinct. NY: HarperPerennial. Rogers, C.L. (1997). Covariations of behaviors for developmentally delayed pre-schoolers following treatment. Unpublished doctoral dissertation, McGill University, Montreal, Quebec, Canada. Speidel, G.E. & Nelson, K.E. (Eds.)(1989). The many faces of imitation in language learning. New York: Spinger-Verlog. Whitehurst, G.J. (1985, April). A home-based intervention program for two-year-olds with expressive language delay. Paper presented at the Society for Research in Child Development Biennial Meeting, Toronto, Canada. Whitehurst, G. J., Fischel, J.E., Arnold, D.S. & Lonigan, C.J. (1992). Evaluating outcomes with children with expressive language delay. In S.F. Warren & J. Reichle (Eds.), Causes and Effects in Communication and Language Intervention. Toronto, Ont.: Paul H. Brookes Publishing Co. Zelazo, P.R. (1989). Infant-toddler information processing and the development of expressive ability. In P.R. Zelazo & R.G. Barr (Eds.), Challenges to Developmental Paradigms, Hillsdale, NJ: Lawrence Erlbaum Associates, Inc. Zelazo, P.R. (1997a). Infant-toddler information processing assessment for children with pervasive developmental disorder and autism: Part 1. Infants and Young Children, 10(1), 1-14. Zelazo, P.R. (1997b). Infant-toddler information processing treatment for children with pervasive developmental disorder and autism: Part 2. Infants and Young Children, 10(2) 1-13. Zelazo, P.R. & Kearsley, R. (1984). The identification of intact and delayed information processing: An experimental approach. Symposium: Early information processing and later cognitive functioning in delayed and normal infants. Infant Behavior and Development, 7, 393 (Abstract). Zelazo, P.R., Kearsley, R.B., Kruzynski, A.K., Ciccone, N., Reid, C. & Thompson, S. (1996, August). Improved play with compliance training in delayed toddlers. Poster session presented at the annual meeting of the American Psychological Association, Toronto, Canada. Zelazo, P.R., Kearsley, R.B., Stiles, K. & Randolph, M. (1985, April). Parental facilitation of expressive language. Paper presented at the Society for Research in Child Development Biennial Meeting Proceedings, Toronto, Canada. Zelazo, P.R., Kearsley, R.B. & Ungerer, J.A. (1984). Learning to Speak: a manual for parents. Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.
|
