Date: Sat, 24 Apr 2004 23:31:04 +0200 From: Leah R. Paltiel-Gedalyovich Subject: Classification of Developmental Language Disorders
EDITOR: Verhoeven, Ludo; van Balkom, Hans TITLE: Classification of Developmental Language Disorders SUBTITLE: Theoretical issues and clinical implications PUBLISHER: Lawrence Erlbaum Associates YEAR: 2004
Leah R. Paltiel-Gedalyovich, Ben-Gurion University of the Negev
This volume is an edited collection of papers evolving from a workshop at the Max Planck Institute for Psycholinguistics. A short introduction precedes the main body of the book which is divided into three parts of roughly equal length: etiology, typology and assessment and intervention. The editors state the purpose of the book as 'the classification of subtypes of developmental language disorders' (DLDs, p. 6). This classification is examined with an eye towards assessment and intervention.
In the introduction, the editors briefly review the various descriptions of DLD and the implications of the difficulties for schooling. They then highlight theoretical and practical difficulties in assessment and differential diagnosis, making practical if brief recommendations for what may be involved in the assessment process. Classification of the type of DLD is seen as a primary goal of assessment, however, limitations in diagnostic tools limit the ability to reach such a classification at a clinical level.
Thus, this volume aims to contribute towards remediating this situation. Multi-faceted approaches of this collection is emphasized by the very different research orientations and methodologies of the participants. It brings a range of theoretical orientations, concentrating on translating theory into clinical practice.
The stated target population is anyone interested in SLI. It is also recommended as an introduction to SLI for undergraduate and graduate students.
PART 1 ETIOLOGY
In the opening chapter, Chapter 1 'Characteristics of children with specific language impairment (SLI)', Nicola Botting and Gina Conti-Ramsden describe the difficulty in reaching a positive criterion for SLI and review the negative criteria (what to rule out). They then describe their research project, a survey of 50% of the children attending special language educational environments in England. They thus have a large sample size (233 children) at one age group (7 years) assessed twice at a one year interval. Note that their group included many children (58) whose standardized language test scores fell within normal limits yet were significantly lower than their non-verbal abilities.
Their research results in a classification of these children into 6 groups which they compare to the classification of Rapin and Allen, although, they prefer to characterize their groups by description of the pattern of performance on a a number of clinical tasks rather than a label. They report the changes which occur in the surveyed population over the year - interestingly, children may change their classification but new classifications were not found. The authors compensate for limited discussion, probably due to length limitations, by references to detailed published accounts for more in depth discussion.
In my opinion this chapter deals with typology and not etiology and would have been more appropriately placed as an opening for the second section of the book.
Chapter 2 ' Neuroplasticity and development: 'The acquisition of morphosyntax in children with early focal lesions and children with specific language impairment' by Judy Reilly, Jill Weckerly and Beverly Wulfeck follows. This chapter compares the patterns of behavior of children with focal brain lesions and SLI with their normally developing peers. Here the emphasis is on the differential patterns shown by different etiologies. The authors conclude that the populations are differentiated in terms of patterns of recovery rather that patterns of disability. As in the previous chapter, the research brought was carried out on an impressively large population. Children with focal brain lesions are found to recover and reach normal levels of language behavior by age 7-8 while children with SLI retain impaired behavior until older ages. The results are related to the question of neuroplasticity. The recovery of the children with focal lesions shows the plasticity of the brain with healthy areas compensating for damaged areas. This is unavailable for children with SLI due to the postulated diffuse nature of the brain impairment.
Chapter 3 is 'Language disorders across modalities: The case of developmental dyslexia' by Pieter H. Been and Frans Zwarts. The authors describe a model of auditory perception, the combined ARTPHONE-SWEEP model. They use this model to simulate behavioral and brain study results of phonological awareness in Finnish and Dutch normal controls, adult dyslexics and in infants at risk for dyslexia. The model shows that dyslexic behavior may results from a deficit in auditory areas which is similar to a visual deficit, based on general perceptual deficits affecting auditory and/or visual areas. This is supported by the modelling of the SWEEP component on models of the visual cortex. However, similar results can be found based on the ARTPHONE model which models an auditory, language specific capacity.
The model has the advantage of making predictions (and recommendations) regarding treatment options, both pharmaceutical and behavioral. It provides a theory of the neural structural deficits underlying dyslexia, specifically, reduced local neural density. Thus deals directly with the question of etiology for this population. The chapter is quite technical as a fair bit of familiarity with neural anatomy and phyiology, as well as modeling is assumed.
Chapter 4 'Neuroimaging measures in the study of specific language impairment' by Paavo H. T. Leppanen, Heikki Lyytinen, Naseem Choudhury and April A. Benasich gives a review of the results of a variety of neuroimaging techniques for children with SLI and related deficits. The accuracy and benefits of such techniques in studying adults is stressed along with the parallel expectation regarding children in general and particularly in clinical populations. Each technique is introduced with a brief description assuming no or little prior knowledge. References to detailed reviews of a variety of techniques are given.
No definitive correlation between anatomical anomaly or neural function and behavioral pattern (subgroup of SLI ) has been found but some trends are evident, particularly in the auditory pathway and in auditory processing. Inconsistencies in techniques, populations and results between studies make drawing conclusions difficult. It is not always clear which behavioral pattern is being examined for a neuro-substrate due to lack of a unified definition of SLI.
The authors conclude that a potential for neuroimaging as a technique for identification and prognosis of children at risk for langauge and language related disorders exists but it is not yet well enough developed to be used for differential diagnosis or definitive prediction.
Chapter 5 'Information processing in children with specific language impairment' by Ronald B. Gillam and LaVae M. Hoffman deals with etiologies of SLI, accompanying symptomatology which may or may not be causal: attention deficits, auditory discrimination and processing deficits, phonological representation limitations.
After considering research into working memory, specifically for auditory speech information, Gillam and Hoffman conclude that the main difficulty is in creating, retaining and using 'verbal codes'. Alternatively, recoding phonological information may cause system overload for these children.
Along with a survey of the relevant research, the authors report two experiments of their own. The first (published) study deals with the question of working memory for verbal codes. The second (in press) study deals with the role of central executive function. they found that children with SLI could not take advantage of the multi-modal clues. Rather than being integrated, the separate visual and auditory clues served to detract from each other. Limitations in verbal storage capacity account for poor performance. They suggest that children with SLI show problems in general memory and response time relative to age-matched peers, concluding that there is a problem with general capacity.
Clinical implications are discussed, specifically, they suggest a "dynamic assessment" process to assess both language skills and "psychological functions that support language development" (p. 150). These psychological functions include: attention, perception, memory, and central executive functions. The testing process involves: pretest, teaching (1-2 sessions), posttest. The posttest assesses the degree and type of change resulting from the teaching phase in order to evaluate the child's information processing abilities. A specific example of the assessment process using the example of narrative skills is given.
Intervention strategies are based on the results of the assessment. Although general treatment may be highly oriented towards ecological validity and pragmatic appropriateness for the social and academic environment, 'mini-lessons' focussing on particular skill areas are sometimes needed. Intervention techniques suggested include: promoting attention, reducing speech rate and improving clarity, promoting phonological coding by drawing attention to various aspects of phonology, choosing treatment topics based on what is familiar to the child, organizing new information and provide memory aids.
Chapter 6, 'Environmental factors in developmental language disorders' by Sienke Goorhuis-Brouwer, Francien Coster, Han Nakken and Henk Lutje Spelberg discusses the role of social communication in the development of language disorders and the converse influence of a language disorder on socio-emotional development.
Three pilot studies are reported: Study 1 examined the behavior of pre-school (1-3 years) children with SLI. Based on a behavior and language questionnaire the only difference between the normally developing preschoolers and those with SLI was in the adaptation (less open in contact with peers). The parent interview however showed behavior problems, again in the area of socialization. Study 2 examined the behavior of kindergarten (3-6 years) children with SLI. Observations assessed by the CBCL (Achenbach, 1991) did not show significant differences between the groups, but again, the parent interview did. Three quarters of the parents reported problems. Study 3 examined the behavior of school children (6-15 years) with SLI. Teachers reported socio-emotional problems which were supported by their observations reported by the teachers version (TRF) of the CBCL.
>From the pilot studies the authors hypothesized that there is a gradual increase in the correlation between reported and observed aberrant behavior with increased age. The combination of behavioral problems and language deficit increasingly interferes with social interactions and communications with adults and peers. This hypothesis was examined in a large study (over 150 participants). The methodology involved the CBCL teacher and parent observation forms. Around one third of the children were found to be within the clinical range for socio-emotional development based on parental ratings, the percentage of children rated by teachers to be within this range was slightly higher. When the ratings were taken together almost half the children were considered to be within the clinical range. Note that there was poor parent-teacher agreement. This is interpreted as indicative of the children's differential behavior depending on the communicative situation and partner. Discrepancy between early perceived behavior and actual behavior was found to affect development of deviant behavior.
Part II Typology
Chapter 7 'Speech output disorders' by Ben Maassen opens this section. Using Levelt's (1989) model of speech output, Maasen aims to arrive at clinical diagnosis. He compares developmental apraxia of speech (DAS) with dysarthria, phonological delay , other (non-specific) articulation disorders and normal speech. As such the symptoms of DAS reflect difficulty 'somewhere between word form retrieval and articulation' (p. 177) in Levelt's model. He follows Bishop (1992) as defining the disorder as a deficit in converting from phonology to motor execution. Five procedures for evaluating intelligibility both quantitatively and qualitatively are described: 1. producing phonetic contrasts, 2. phonological process analysis of a spontaneous sample, 3. word identification, 4. Percentage consonants correct (PCC, Shriberg and Kiawtkowski, 1982), and 5. scales of speech qualities. In addition Maasen demands of a speech evaluation that it should aid in determining etiology and allow differential diagnosis.
An empirical study is reported. Criteria for selecting participants as DAS, dysarthric, SLI or otherwise are described in detail. Experimental tasks include real and nonsense word production as well as rapid syllable string production. Results show that on the word production task, children with DAS produce more errors than children with SLI or dysarthria who produce more errors than their normally developing peers. Dysarthria results in a high percentage of distortions while DAS results in a high percentage of place substitutions. All groups show similar patterns regarding other aspects of sound production. On the maximum repetition task children with dysarthria are slower than children with DAS and SLI who are slower than children with normal development.
Similarities between the SLI and DAS effects leads to a closer analysis which suggests that about half of the children with SLI show dyspraxic symptoms, suggesting a possible shared underlying deficit. This deficit is thought to be a deficit at the level of speech segments and articulatory movement. These are parallel to Levelt's 'segmental spellout' and 'phonetic spellout'. Note that DAS rarely occurs in isolation. Although the chapter is clearly written, I would have benefited from a more detailed (and graphic?) explanation of Levelt's model.
Chapter 8 'Central auditory processing' (CAP) by Jack Katz and Kim Tillery first gives statistics on the prevalence of CAP disorders (CAPD) in the normal population (thought to be 5% -10%); similar or higher percentages may exist in clinical populations. A disorder in CAP is defined as a disorder in what is done with auditory input. CAPD has been associated with learning difficulties, primarily reading but also other language and communicative areas. Its symptoms may be confused with attentional disorders. However, many children with ADHD also suffer from CAPD.
A behavioral CAP assessment battery is described which includes: 1. Staggered spondaic word test (SSW), this yields a quantitative score as well as therapeutically useful qualitative error analysis. , 2. Phonemic synthesis test (PS) which provides quantitative and qualitative information on sound blending skills. , and 3. Speech in noise test.
CAP behavioral studies, together with neural lesion information, have resulted in classification of four subtypes : 1. 'Decoding category' (DEC), 2. 'Tolerance-fading memory' (TFM) category, 3. 'Organization category' , and 4. 'Integration'. The categories are summarized in a clear table. For each category a clear description of the symptoms, the anatomical areas affected and the relation between these is given.
The authors follow with a sample case study which relates assessment results to intervention recommendations. Finally intervention strategies are suggested for each of the categories.
Chapter 9 ''Lexical deficits in specific language impairment' by Laurence B. Leonard and Patricia Deevy looks at children with lexical rather than grammatical disorders (cf. work by van der Lely on a lexical subtype of SLI). They describe the lexicon and its importance, as well as problems deriving directly from lack of word knowledge, secondary effects on grammar and morpho-grammar learning.
Although early symptoms of SLI include late first words, smaller vocabulary, limited comprehension, and decreased lexical diversity, the general pattern of vocabulary composition and use is similar to younger normally developing children. The authors survey some assumptions regarding lexical processing. Each word entry includes: "lexical concepts" (meanings are formed for words and links are made with related meanings), "lemma" (grammatical properties and links to other entries with similar grammatical properties), and "lexeme" (phonological information including links with entries with shared or similar information). Comprehension progresses from lexeme to lemma to lexical concept. For production the process is reversed.
A review of research into lexical skills of children with SLI shows evidence for relatively normal learning of novel words for comprehension but reduced learning for production. An explanation of these results in the context of the given model suggests that this is indicative of the formation of a reduced lexeme. There appear to be two groups of impairment, for one, semantic information is more distorted and for the other, it is phonological information which is primarily affected. Specific difficulty with verbs may also be explained within this model although the authors note that some 'pure' syntactic difficulties remain.
Chapter 10 is 'Morphological disorders' by Dorit Ravid, Ronit Levie and Galit Avivi Ben-Zvi. These authors choose Hebrew derivational morphology as a diagnostic tool in comparison of school children with SLI and NLA. Thus, they aim to contribute to the delay versus deviance debate.
Language in school children is characterized by continued development alongside the interaction of language with growing literacy skills. Children with SLI show problems with: 1. lexicon 2. morphological processing, 3. metalinguistic skills. Children tend to learn the morphology of morphologically rich languages relatively easily, including SLI children. On the other hand, studies of derivational morphology in English have shown that SLI children have difficulty with this area. Previous study by these authors has shown children with SLI have difficulty in production of novel derived nouns and in semantic noun classification.
Adjectives are a unique content word class semantically, pragmatically and syntactically. Adjectives characterize nouns and as such have a morphology which reflects the characteristics of the noun in morphologically rich languages. As background to their study, the authors give a short and clear description of Hebrew verbal, nominal and adjectival morphology.
Counter to predictions children with SLI perform more poorly on comprehension task than age and language matched controls. On production tasks, children with SLI differ from both groups. Resultative adjectives are found to be significantly easier than attributive and denominal. Children with SLI differ from both groups on resultative adjectives but do not differ from language matched controls on attributive and denominal adjective derivation.
The difficulty with comprehension is interpreted as resulting from difficulty in analysis of morphology. However, the experimental task requires production as well, such that it is not clear what causes of the poor performance. The authors note non-language factors which may have confounded results but do not consider expressive language difficulty.
The results are interpreted as supporting a deviance (as opposed to delay) explanation. This is based on a qualitative analysis of errors (quantitative analysis does not shown consistent difference from language matched controls): children with SLI use analytic expressions and semantic responses more often than their LA matched peers as well as one strategy (unconventional adjective). They conclude that this is an appropriate area of study for distinguishing SLI.
In Chapter 11 'Grammatical impairment: An overview and a sketch of Dutch' by Jan de Jong examines theories of the morpho-grammatical deficit which is present in most children with SLI. De Jong reviews theories of grammatical symptoms of SLI. 1. Non-salience of inflectional markings - "surface" hypothesis. Leonard (1989) accounts for cross-linguistic differences where differences in surface saliency of inflections account for whether or not they are affected by SLI. 2. Decreased control of subject-verb agreement. 3. Prolonged optional infinitive stage. 4. Decreased knowledge of abstract features.
Part of the author's doctoral dissertation is reported which investigates the nature of tense and agreement in Dutch SLI and what (if any) of these SLI theories are supported? Results show error patterns in children with Dutch SLI as compared with controls including inconsistent marking of past tense. Errors include omissions, substitution (present tense affixes), infinitive forms. In addition - omissions of inflectional markers (regarding 3rd person), misrepresentation of number (singular instead of plural), form position concordance.
De Jong explains the data as compatible with all but the extended optional infinitive stage hypothesis for omissions, substitutions cause difficulty for the surface theory as well. Adapting the optional infinitive stage to Dutch is difficult since this follows (in the proposed stages) a finite stage. The children use an auxiliary plus infinitive. The optional infinitive account does not predict the agreement (person, number) errors. The agreement hypothesis does not predict the past tense errors. Overgeneralizations argue against a theory that SLI children have no linguistic rules. Research into non-English SLI leads to the need to account for substitutions not just omissions, and to consider phenomena of verb second languages.
In Chapter 12 'Pragmatic disability in children with specific language impairments' by Hans van Balkom and Ludo Verhoeven pragmatic disability is defined as a deficit in communicative use of language. Specifically "an inability to select and match a suitable linguistic form to the most appropriate and effective communicative function. ' (p. 283)
Two approaches modular and functional are described and illustrated in clear figures. I found it interesting that the population studied was very young and that the entire SLI group reached age appropriate behavior as measured by formal tests within two years. Is this representative of children with SLI?
Results of the empirical study show that SLI and NLA are not differentiated by the appropriateness and variety of communicative functions. Rather, the SLI population uses inappropriate linguistic forms for these functions. The authors note the reciprocal relationship between the children's language performance and the parental language behaviors resulting in a poorer linguistic environment. The children have difficulty with discourse processes because of weaker representation of linguistic knowledge, even when this knowledge reaches normal levels. Within the SLI group, higher functioning children perform better than lower functioning children with regard to form-function mapping as measured by discourse coherence. Parent interactions with children with SLI are less involved and less equally distributed than with children with NLA. This linguistic environment provides for the child with SLI only limited opportunities to develop discourse strategies.
PART III Assessment and Intervention
In Chapter 13 'Specific language impairment: Diagnostic dilemmas' by Dorothy V. M. Bishop, the author begins with arguing for the different demands of a classification system for research and clinical purposes. She makes use of the World Health Organization classification to suggest that research deals with impairment, while clinical work deals with disability arising form such an impairment and aims to minimize handicap resulting from such disability.
The aim of clinical assessment is to determine which children will benefit from which specific forms of intervention and educational placement. Traditional criteria like IQ discrepancy scores need to be used discriminantly, again depending on the goal of classification. The type of assessment used is influenced by the model. Qualitative assessment strategies result from a model which views SLI behavior as atypical, eg. Rice's (2000) extended optional infinitive test or non-word repetition. Standardized tests may not be geared to distinguish impaired and non-impaired populations.
Children with a really specific disorder are unusual; usually we find co-morbidity. The search for a very specific diagnosis may result in a too narrow evaluation of the child. She notes that various professionals will diagnose the same symptoms differently. A multidimensional model (instead of labels) allows us to relate to the child's function in a variety of critical areas. However, we need labels to allow appropriate educational placement. Bishop emphasizes, the pure cases are NOT representative of the SLI population. For research purposes co-morbidity raises issues of the relationship between the various impairments. Clinical approaches are more successful in describing underlying processes. Bishop ends by describing some different types of SLI which do have consensus.
In Chapter 14 'A dynamic systems approach to diagnostic measurement of SLI' Paul van Geert defines a dynamic system as "a structure of interacting forces" (p. 327) which undergoes 'self-organization;' becoming increasingly more ordered and structured.
As opposed to the classical Chomskyan nativist view and the "poverty of the stimulus" argument for language innateness, the dynamic systems approach allows rich linguistic competence to develop from poor linguistic input by not assuming the axiom of entropy. According to the Chomskyan view, deficient language development should first of all reflect a deficit in the innate substrate; according to the dynamic systems approach, every step in the self-organization process causes the following step.
This system is iterative; thus the input to the next stage is the preceding stage plus environmental input. Finding the cause of impairment requires reconstructing the developmental process. The system may reach an "attractor state" where the output stage is similar to input stage which will be semi-stable unless external factors are changed (e. g. an individual is given a hearing aid).
The model adopted of development is a linear increase model: Each successive state depends on the current state plus input. If increase and decreasing factors are of statistically equal magnitude the result is a 'random walk' showing random fluctuations. The use of the model to explain fluctuations (loss and recovery) is exemplified by Landau-Kleffner syndrome.
Since more than one skill area competes for the resources (such as attention) the amount of competition influences the outcome. The model predicts a bi-modal pattern of development for two competing factors. The possibility of innate language disorder or innate language knowledge is compatible with the model. An innate disordered state influences the amount of resources which must be allocated.
A dynamic model predicts (as is found) changes in the pattern across time. An impaired genotype may have different manifestations as resources are competed for - resulting in different patterns of strengths and weaknesses.
Psychological properties cannot be accurately measured by tests because they are the interaction of the individual with the context and environment. Therefore, assessment involves describing a range of characteristic performances for a skill area (e.g. morphosyntax) for a child under different contexts (e.g. test performance, normal communication, stress, etc. ) . Thus, one part of the description of the child's ability addresses the variability of his/her performance. In this view, test-retest variability is expected. Rather than giving a child a specific label, van Geert suggests stating that a diagnostic label is "characteristic of a specific person to a degree of approximately. . . . (x). " (p. 345)
I found the detailed explanation of the model and the use of figures and examples extremely helpful in understanding a view of language acquisition and disorder which was totally new to me.
In Chapter 15 'Early detection of developmental language disorders' Hanneke de Ridder and Heleen van der Stege discuss a screening instrument for language impairment introduced as part of a national health center protocol for infants/toddlers. The screening instrument (VTO)is based on early precursors of communication evident in parent-child interactions in children 0-3 years. The instrument schedule follows regular check-up schedule for health centers: 1,2,3,6,9,12,15,18, and 24 mos. Here results of the last 4 visits where measurements occurred are reported. Note that the time allowed is less than 5 minutes for administration of a parent interview with standardized questions.
The areas covered are: production, comprehension, conversation/play. Close to 900 children participated in standardization. For the majority of questions, 90% or more of the children had acquired the surveyed skills, but many questions did not. A look at the detailed chart of the results of this survey shows that for almost all items over 80% of the children performed as expected.
As assessed by this screening instrument, underlying structure emerges at 9 months for comprehension and at 12 months for production - by age 24-30 months production versus comprehension problems can be distinguished. No structure for interaction is shown. The cut off score for normal language behavior is determined by the estimated incidence of language disorders in the country.
Predictive validity is shown in a follow up study, as a correlation was found between failing the screening test and language performance 1 and 3 years later. There were 10% false negatives, information on false negatives is not yet available. The authors note that despite the success of the screening instrument a major problem remains in parents refusing to follow up recommendations for further testing.
Chapter 16 'Early intervention for children with language impairments' by Steve F. Warren and Paul J. Yoder discusses the basis of intervention as the notion that the quality of the input influences development. The authors present three different models of intervention: responsive interaction, milieu, and didactic teaching. They describe how each works and then argue that children at different stages of cognitive and language development can best benefit from different types of therapy. Responsive interaction is found to be the best option for children at the stage of development above MLU 2. 5, milieu for children under MLU 2. 0 and didactic teaching for children with MLUs over 2. 5 and mild cognitive deficits. Also, the therapeutic goal affects method, e.g. milieu teaching has been found to be effective for vocabulary teaching. The message is to choose the therapy method of best-fit to the specific child at the specific stage of the child as opposed to opting for one type of therapy throughout. Recommendations which a clinician can take and apply to his/her clinical work are made.
Chapter 17 'Benefits of speech manipulation for children with language disorders' by Ludo Verhoeven and Eliane Segers discusses intervention based on compensation for assumed limitations in temporal processing at the basis of language disorders. Specifically, they look at research showing the effect of lengthening of formant transitions for individuals with SLI , as well as research on related intervention programs. This research has involved phonemic (between phoneme boundaries) or phonetic (allophonic variation) discrimination. Individuals with language disorders have been found to have more difficulty than controls in discriminating between CVs with a stop consonant. Studies using a speech continuum find differences between 'slow learners' and normal controls but differences overall tend to be smaller. Overall lengthening formant transition was found to improve discrimination for language disordered children (but not adult aphasics).
Intervention programs following this research includes computer assisted training: one example is the fairly intensive Fast ForWord program. In this program computer training is adapted to the child's progress, the program includes syntax and morphology as well as auditory training. Research evaluating this and similar problems is methodologically problematic. Furthermore, the intervention programs do not closely follow the techniques researched.
Although a clear pattern of auditory perceptual difficulty is shown, the authors question the causal relationship between this difficulty and the language disorder.
The closing chapter, Chapter 18 'The close association between classification and intervention of children with primary language impairments. ' by James Law examines classification as related to the existence of differential intervention strategies. He recommends distinguishing transient from persistent language disability by diagnostic therapy designed to child's individual needs, matched to a specific set of behaviors. He reviews research into the effectiveness of intervention, summarizing 22 group studies. Conclusions include: parent-directed treatment is as effective as direct clinician treatment, there is transfer from intervention in syntax to phonological skills and vice versa, positive intervention outcomes are found for articulation/phonology, expressive language, and receptive language, and finally, auditory training did not improve articulation/phonology. Usually intervention moved children into the normal range of language performance.
Law notes that further research is needed to evaluate whether the effects of intervention are only short terms or are real long term changes in the child's psycholinguistic abilities.
Working backwards, behavioral patterns which respond to intervention can be considered classification categories, e. g. if work on articulation improves performance we can identify a category of children with articulatory impairment. Consequently, the less we know about intervention with a certain pattern of behavior (such as pragmatics), the further we are from a useful classification.
I found the volume interesting reading throughout. Although a wide range of theoretical points of view are represented and within the topic, a wide range of sub-topics, continuity is maintained by the progression from the theoretical to the practical. Dorothy Bishop's paper at the beginning of Part III aids this transition in her discussion of the differences between research and clinical needs regarding classification. The fact that many chapters include references to other chapters in the book, either for theoretical basis/practical application or for a contrasting view, adds to the coherence of the volume as a whole. I found some chapters to assume more previous knowledge or familiarity with the topic than others, something which may need to be considered when recommending the book to students. However, all chapters are clearly written and well referenced, pointing the reader in the appropriate direction for filling in any missing background.
Bishop, D. V. M. (1992) 'The underlying nature of specific language impairment', Journal of Child Psychology and Psychiatry and Allied Disciplines, 33, 3-66.
Leonard, L. B. (1989) 'Language learnability and specific language impairment in children', Applied Psycholinguistics, 10, 179-202.
Levelt, W. J. M. (1989) Speaking: From intention to articulation. Cambridge, MA: MIT Press.
Rice, M. L. (2000) 'Grammatical symptoms of specific language impairment', in D. V. M. Bishop & L. B. Leonard, eds., Speech and language impairments in children: Causes, characteristics, intervention and outcome, pp. 17-34. Hove, UK: Psychology Press.
Shriberg, L. D. and Kiawtkowski, J. (1982) 'Phonological disorders III: A procedure for assessing severityof involvement', Journal of Speech and Hearing Disorders, 47, 56-270.
ABOUT THE REVIEWER:
ABOUT THE REVIEWER
Leah Gedalyovich is currently assisting in research into Hebrew G-SLI at the Department of Foreign Literatures and Linguistics, Ben-Gurion University of the Negev. She combines research with part-time work as a speech-language pathologist in pre-school and school settings. Research interests include normative first language acquisition (primarily of Hebrew), language disorders, the interaction of semantics and pragmatics and the clinical application of linguistic theory.