cover next page >

title: Child Language Disability. Vol. 2, Semantic and Pragmatic

Difficulties Multilingual Matters (Series)
author: Mogford-Bevan, Kay.
publisher: Multilingual Matters
isbn10 | asin: 1853591289
print isbn13: 9781853591280
ebook isbn13: 9780585125954
language: English
subject Language disorders in children.
publication date: 1991
lcc: RJ496.L35
ddc: 618.92/855
subject: Language disorders in children.

cover next page >

< previous page page_i next page >
Page i

Child Language Disability

< previous page page_i next page >

< previous page page_ii next page >
Page ii
Multilingual Matters
Attitudes and Language
COLIN BAKER
Breaking the Boundaries
EUAN REID and HANS H. REICH (eds)
Critical Theory and Classroom Talk
ROBERT YOUNG
Deaf-ability Not Disability
WENDY McCRACKEN and HILARY SUTHERLAND
Emerging Partnerships: Current Research in Language and Literacy
DAVID WRAY (ed.)
Fluency and Accuracy
HECTOR HAMMERLY
Gender in Education
EILEEN M. BYRNE (ed.)
Language Acquisition: The Age Factor
D.M. SINGLETON
Language Policy Across the Curriculum
DAVID CORSON
The Management of Change
PAMELA LOMAX (ed.)
Managing Staff Development in Schools
PAMELA LOMAX (ed.)
Oral Language Across the Curriculum
DAVID CORSON
The Path to Language: Bilingual Education for Deaf Children
DANIELLE BOUVET
Parents on Dyslexia
S. van der STOEL (ed.)
Performance Indicators
C.T. FITZ-GIBBON (ed.)
TVEI at the Change of Life
DAVID HOPKINS (ed.)
Please contact us for the latest book information:
Multilingual Matters Ltd,
Bank House, 8a Hill Road
Clevedon, Avon BS21 7HH
England

< previous page page_ii next page >

< previous page page_iii next page >
Page iii

Child Language Disability Volume II: Semantic and Pragmatic Difficulties

Edited by Kay Mogford-Bevan and Jane Sadler

MULTILINGUAL MATTERS LTD

Clevedon Philadelphia Adelaide

< previous page page_iii next page >

< previous page page_iv next page >
Page iv
Library of Congress Cataloging in Publication Data
Child Language Disability
Papers of conference held Apr. 23, 1988 and Apr. 21, 1990 organized by Dept. of
Speech of the University of Newcastle-upon-Tyne in association with Sunderland
Education Authority.
Vol. 2 edited by Kay Mogford-Bevan and Jane Sadler.
Includes bibliographies and index.
Contents: Vol.1: Implications in an educational setting. Vol.2: Semantic and
pragmatic difficulties.
1. Language disorders in children. I. Mogford-Bevan, Kay. II. Sadler, Jane. III.
University of Newcastle-upon-Tyne, Dept. of Speech. IV. Sunderland Education
Authority.
RJ496.L35C47 1989 618.92'855 89-3177
British Library Cataloguing in Publication Data
A CIP catalogue record for this book is available from the British Library
ISBN 1-85359-128-9
ISBN 1-85359-127-0 (pbk)
Multilingual Matters Ltd
UK: Bank House, 8a Hill Road, Clevedon, Avon, England BS21 7HH.
USA: 1900 Frost Road, Suite 101, Bristol, PA 19007, USA.
Australia: P.O. Box 6025, 83 Gilles Street, Adelaide, SA 5000, Australia.
Copyright © 1991 Kay Mogford-Bevan, Jane Sadler and the authors of individual chapters.
All rights reserved. No part of this work may be reproduced in any form or by any means without permission in writing from
the publisher.
Index compiled by Meg Davies (Society of Indexers)
Photoset printed and bound in Great Britain by The Longdunn Press Ltd, Bristol.

< previous page page_iv next page >

< previous page page_v next page >
Page v

Contents

Preface vii
Notes on Contributors ix
1
Semantic and Pragmatic Difficulties or Semantic-Pragmatic
Syndrome? Some Explanations
Kay Mogford-Bevan and Jane Sadler 1
2
Is There Such a Thing as Conversational Disability?
Michael F. McTear 18
3
Teaching Vocabulary: The Case for a Semantic Curriculum
David Crystal 43
4
How Can We Study Pragmatic Disabilities?
Gina Conti-Ramsden 59
5
Analysis of Language-Impaired Children's Conversation
Catherine Adams 68
6
A Teacher's and A Speech Therapist's Approach to Management
Susan Hyde-Wright and Barbara Cray 75
Index 101

< previous page page_v next page >

< previous page page_ii next page >

intentionally left blank

< previous page page_vii next page >
Page vii

Preface
Throughout the 1980s studies of child language emerged which looked beyond the familiar descriptive categories of structural
linguistics, designed for analysis at the level of the sentence, to consider the pragmatic aspects of verbal interaction. These
studies resulted in an awareness, also on the part of many working with children with language difficulties, of the limitations of
decontextualised analyses of child language in terms of phonology, semantics and syntax, and of the importance of the
interactional aspects of language use.
In fact, as researchers turned their attention to examining the conversational skills of language disordered children it became
apparent that in some cases their problems might well be explained in terms of a more specific difficulty with the use, as
opposed to the structure, of language. Naturally this information was seized upon in turn by many practitioners, relieved to have
found a possible explanation for the previously ill-definable language difficulties of some of their children, and gradually the
term semantic and pragmatic disorder was adopted.
However, problems of this nature are still only just beginning to be investigated and scarcity of related literature has meant that
there is a risk of over-application of this new term, of its becoming an umbrella term to label any child whose problems with
communication do not fall clearly into the better-known categories traditionally used to describe language disability.
Articles in journals, letters in professional publications and requests for courses on this subject indicated to us the need of
practitioners for more usable information and led us to convene a conference on Children with Semantic and Pragmatic
Disabilities on 21 April 1990. An attendance of over four hundred participants (teachers, speech therapists and educational
psychologists) confirmed the wide interest in this topic and a desire to learn more about it.
This book, which in the main is based on the oral presentations made by invited speakers at the conference, aims to provide up-
to-date information for professionals working with these children in a variety of educational and

< previous page page_vii next page >

< previous page page_viii next page >
Page viii
therapeutic settings. All contributors are pioneers in this new area of enquiry and practice. They generously agreed to share what
they know, think and do about the problems as they see them at this early stage of research and understanding.
While editing their contributions it became clear to us that there was the need for a preliminary chapter containing some
additional explanations. This, we hope, will make the chapters that follow more accessible to any readers who are relatively new
to the area of child language disability.
Finally, our sincere thanks for support in this initiative go to Sunderland Education Authority, in particular Mike Vening and
Eileen Richardson; to Gillian Cavagan and Lesley Watson of the Department of Speech at Newcastle University for their
secretarial assistance; and to Mike and Marjukka Grover of Multilingual Matters without whom this volume would not have
been published.
KAY MOGFORD-BEVAN
JANE SADLER
UNIVERSITY OF NEWCASTLE UPON TYNE
SEPTEMBER 1990

< previous page page_viii next page >

< previous page page_ix next page >
Page ix

Notes on Contributors
Cathy Adams is a speech therapist who has worked as research assistant on a project to investigate the problems of children with
semantic and pragmatic disorders. She is a lecturer in speech pathology in the School of Education of the University of
Manchester. Her other research interests include language acquisition and development and Asperger's Syndrome.
Gina Conti-Ramsden is a senior lecturer in the Centre for Educational Guidance and Special Needs in the School of Education
of the University of Manchester. She is editor of Child Language Vol.7 and co-author of Children's Language Disabilities. Her
research interests include language development processes, parent-child interaction and language disorders.
Barbara Cray has been a teacher at Dawn House School, Rainworth, Notts, for several years. Her background is in special
education and her specific interest includes the development of reading and computer-assisted learning. She is responsible for a
class of children with semantic and pragmatic difficulties and is developing a practical ongoing assessment procedure, including
the use of video, for this group.
David Crystal is Honorary Professor of Linguistics at the University College of North Wales, Bangor. He is the author of several
books on language problems, including Clinical Linguistics and Profiling Linguistic Disability, and is editor of Child Language
Teaching and Therapy. His latest publication is the Cambridge Encyclopedia, and he is currently working on an English series
for the National Curriculum.
Susan Hyde-Wright is a practising speech therapist and teacher currently working at Dawn House School. She is involved in
developing and applying specific language teaching techniques which can be integrated into an overall curriculum. Her other
interests include the ways that self-cueing strategies can assist children with word-finding difficulties, and the nature of teacher-
pupil talk in a special school.
Michael McTear is a linguist. He is senior lecturer in Information Systems at

< previous page page_ix next page >

< previous page page_x next page >
Page x
the University of Ulster. His books include Children's Conversation, The Articulate Computer and Pragmatic Disability in
Children. His research interests are computational pragmatics, language acquisition and language disability.
Kay Mogford-Bevan is a speech therapist and developmental psychologist. As a lecturer in the Department of Speech at the
University of Newcastle upon Tyne, she teaches undergraduates who are training to be speech therapists and experienced
teachers on the courses in Child Language and Language Disability. Her research interests centre on the development of
communication and related abilities in children with a range of developmental disorders affecting language development.
Jane Sadler is a teacher and a speech therapist. She is a lecturer in the Department of Speech at the University of Newcastle
upon Tyne where she co-ordinates the part-time and full-time Diploma and Masters Courses on Child Language and Language
Disability for experienced teachers. Her research interests include teacher interaction with language disordered children.

< previous page page_x next page >

< previous page page_1 next page >
Page 1

1
Semantic and Pragmatic Difficulties or Semantic-pragmatic Syndrome? Some Explanations
KAY MOGFORD-BEVAN and JANE SADLER
Introduction.
This book is written primarily for professionals working with children with language disabilities in any educational or
therapeutic setting. As the title suggests, all of the chapters concern problems with those aspects of language which linguists
refer to as semantics and pragmatics. Although it is almost impossible to be working in the field of language disability and be
unaware of problems which are labelled semantic and/or pragmatic, the terms themselves are by no means clear-cut. As a result,
there may be some confusion about how semantic and pragmatic difficulties are recognised and described. For example, readers
will find that the contributors to this book use several different terms to refer to broadly similar language difficulties. The fact
that each contributor is concerned either with a different aspect of a similar problem or with a particular group of children helps
to explain some of the differences. But this is only part of the story.
For this reason, and because a number of the readers may be relatively new to the area of childhood language disabilities, the
editors decided to write an introductory chapter with some additional explanations to provide a framework for the chapters
which follow and in order to help to reconcile some of their apparent contradictions.

Terminology
Problems of terminology arise because language disability in childhood

< previous page page_1 next page >

< previous page page_2 next page >
Page 2
has become the concern of several disciplines, each with its own theoretical framework, specialised language and methods. Each
discipline has contributed different elements to the field: for example, medicine is concerned largely with identification and
diagnosis; linguistics with the description and analysis of language; psychology with explaining how the learning capacities and
experience of children contribute to their patterns of behaviour and communication. Speech and language therapists have
attempted to weld together, from each of these disciplines, the methods and insights which allow them to assess and treat the
difficulties of affected children.
In the multidisciplinary area of child language disabilities, the challenge for teachers and therapists is to master the different
areas of knowledge required to understand the problems they must deal with and to steer a path through the minefield created
when different disciplines converge on a problem. The terms semantics and pragmatics, for example, can present a real
difficulty. What abilities in language and language use are implied by definitions of these key terms? A brief and simple
explanation of these and some associated terms is given here. A more extended discussion can be found in Mogford & Bishop
(1988).

Semantics
Semantics is the term used to describe the study of meaning in language. There are broadly two different ways that meaning has
been studied in language development. One branch of study is known as lexical semantics; the second branch is semantic
relations.

Lexical Semantics
The aspect of semantics which first springs to mind, perhaps, is the study of the meaning of words in a language. The words that
comprise a language are known collectively as the lexicon and hence the branch of semantics concerned with word meaning is
called lexical semantics.
In language acquisition, the study of lexical semantics has not been concerned so much with which words a child knows and
uses (i.e. vocabulary) but with the way that children use their vocabulary to refer to things. We cannot assume that when a child
uses a word, the child means something similar to an adult speaker. As in other aspects of language acquisition, errors reveal
how the development of lexical semantics progresses. In the early stages this is seen in the overextension or underextension of
meanings. By

< previous page page_2 next page >

< previous page page_3 next page >
Page 3
overextension we mean that a word is used in an overgenerous or over-inclusive way and applied to a larger class of referents
(the term for the entity referred to) than is normal in adult usage: for example, a child may refer to all four-legged animals as
`cats' or to all men as `daddy'. By underextension we mean that the word is applied in an over-restricted or narrow way: for
example, a child may refer to only a specific animal as `dog'.
The development of the lexicon and semantics is closely related to conceptual development, but they are not identical. Thus a
child may demonstrate that he can manipulate concepts such as quantity or colour, for example matching like with like, before
he has learned the appropriate lexical item to denote that concept.
More sophisticated aspects of lexical semantics are concerned with the use of terms that indicate the relative situation of the
speaker in time or space, such as `here/there', `this/that', `now/then'. These are known as deictic terms. Also included in this
category are some verbs such as `come/go', `bring/take' and the use of pronouns to denote speaker/listener roles (`I/you'). Failure
to develop this aspect of deixis is evident often in autistic and visually handicapped children who use the term `you' rather than
`I' when referring to themselves.
Semantic Relations
The second branch of semantics which has been applied to the study of child language development is called semantic relations.
This concerns the relationship between words at the level of the sentence and how this affects their meaning. The meaning of a
sentence in adult language is derived from its grammar as well as on the meanings of individual words. Take a pair of sentences
in which the vocabulary is identical but the word order is different:
(1) George kicked Sarah.
(2) Sarah kicked George.
We have no difficulty in distinguishing the different semantic roles of the elements which make up these sentences: George is
the agent in (1) and the patient in (2). We can also distinguish their grammatical roles: George is the subject in (1) and the
object in (2).
However, grammatical and semantic roles are not identical.
Take for example:
(3) George is kicked by Sarah.

< previous page page_3 next page >

< previous page page_4 next page >
Page 4
In this sentence George is the grammatical subject but the semantic patient, yet the meaning is similar to that expressed in
sentence (2) above where George was again the semantic patient but this time the grammatical object.
This way of looking at the meanings encoded in utterances was employed by Bloom (1970) in a study of children's early multi-
word utterances when she recorded the same word combinations used in different contexts. By using the context to interpret the
meanings of each word combination, Bloom was able to show that children use word order consistently to express the same
basic set of semantic relations. This set of semantic relations employed the following semantic roles (agent, action, object,
locative, possessor, possession, attribute, entity, demonstrative) in combinations of first two and then three elements. Here are
some examples, with two elements, to illustrate this point:
boy go agent + action
push car action + object
Mummy tea agent + object
Daddy chair agent + locative
Sweetie shelfentity + locative
Timmy hat possessor + possession
ball blue entity + attribute
there teddy demonstrative + entity

Later, children develop the three-word stage by combining phrases expressing two of these semantic relations, for example:
Daddy kick + kick ball  Daddy kick ball
(agent + action) + (action + object) ® (agent + action + object)
Before children are able to employ grammatical devices other than word order, semantic relations are an effective way of
describing the meanings that can be expressed in two-word utterances and the expansion of two-into three-word utterances.
David Crystal's chapter on semantics concentrates on the area of lexical semantics.

Pragmatics
The study of pragmatics covers a huge domain and may be broadly defined as the `rules governing the use of language in
context' (Bates, 1976). Here context refers to `the interrelated conditions in which something exists or occurs' (Prutting, 1982).
For language this includes the social situations, the purposes for which language is used and the form of discourse. Discourse

< previous page page_4 next page >

< previous page page_5 next page >
Page 5
refers to any stretch of language longer than a sentence. It need not be spoken (it could be written or sung) and it need not be a
conversation (it could be a story, or a poem), though conversation is a form of discourse.
In the linguistic context, the term conversation refers not only to exchanges of little significance or `chit-chat', but to any
sequence of utterances that are not random: that is, one speaker relates what is said to what has gone before. At its simplest, this
involves two speakers but can involve more. The relationship between the speakers (e.g. mother and daughter, teacher and child,
husband and wife) influence the way that language is used. For example, a child speaking to a headteacher might choose to use
more polite and indirect forms of speech than a mother speaking to her daughter, who would employ much more direct forms.
The linguistic competence of the listener to whom the language is addressed will also influence the speaker's language. A
number of researchers have shown that adults simplify their language to children in a number of systematic ways (e.g. Snow,
1988, for a review) and older children have been shown to simplify their language to younger children still (Sachs & Devin,
1976; Dunn & Kendrick, 1982). In other words, there are a number of different ways of expressing an idea and context
influences the speaker's choice. Suppose you were with a companion and wished to offer them a sweet, one might say any of the
following:
(4) I was wondering if you would like a sweet?
(5) Want a sweetie?
(6) I was thinking of having a sweet.
(7) Do you fancy a sweet?
The appropriateness of these different forms will depend upon the social relationship between the speaker and listener:
deferential and polite forms are used to adults in authority but language to children is usually more simple and direct. The study
of pragmatics looks at the choices that are made in the use of grammatical forms in systematic ways. There are few features of
language that are not affected by pragmatic factors.
When language is studied in context it emerges that the meaning of an utterance intended by a speaker is not determined simply
by the meanings of the words and the semantic relationships between them, but depends also on the context of that utterance.
Just as there are different ways of expressing a proposition, so there are also different meanings that may be given to an
utterance of a given form, depending on the context. For example, `the kettle is boiling' might be fulfilling one of several speech
acts. It might be a warning to a child, an invitation to a responsible adult to make the tea, a

< previous page page_5 next page >

< previous page page_6 next page >
Page 6
reassurance to an invalid that refreshments are soon to be served. In each of these cases the act the speaker carries out is
different even though its form is the same.
It follows from this that the relationship between a speech act and grammatical form is by no means predictable. This is evident
in the examples above. Sentence (6) functions indirectly as an inquiry but unlike (4), (5) and (7) is not in the grammatical form
of a question. The intention or meaning of the speaker does not come simply from the form of an utterance but from the social
context and paralinguistic cues such as stress, intonation, eye gaze and gesture. The listener has to take this into consideration in
understanding what is said in a conversation. A literal interpretation of an utterance, which ignores contextual information, will
lead to misunderstanding. This is one difficulty which is noted by several authors in this book in children with pragmatic
difficulties.
Although each child eventually has to master the skills necessary for different kinds of discourse, the earliest developing form,
and arguably the most fundamental, is conversation. Here analysis can focus on a number of skills necessary to make
conversations run smoothly and develop productively. These skills include: alternation of speaker and listener roles (turn-
taking), though this becomes more complex than simple alternation when conversations involve more than two people; the
ability to share and maintain a topic of conversation; knowing when and how to change the topic; and recognising when the
topic has been shifted. In addition, the child must learn to recognise when communication has broken down and how to seek and
give clarification as necessary. To do all these things, a child must be able to signal his intention in recognisable form and
appreciate the intentions of the other speaker; this can involve interpreting non-verbal as well as verbal signals. Successful
communication also depends on the speaker assessing what the listener does and does not know and selecting the right forms to
distinguish between new and previously mentioned information. Further aspects of the pragmatics of conversation are
introduced in this book, and wherever possible examples and explanations are provided.
The danger in providing simple definitions is that they may suggest that semantics and pragmatics are discrete aspects of
language and may be unaffected by disabilities of language form, i.e. phonology and syntax. However, it should be evident from
the discussion above that the ability to use pragmatic skills must be closely related to control of syntactic forms. For example,
without the ability to form questions the choices available to perform certain speech acts must be necessarily restricted; some
polite forms depend on the use of modal auxiliaries (e.g. Would you mind? May I go?). By

< previous page page_6 next page >

< previous page page_7 next page >
Page 7
the same token, given that the grammar and meaning of an utterance are related, certain semantic relations will depend on
syntactic development. For example, certain lexical verbs require the ability to control more complex sentence structures (e.g.
compare: `he runs' intransitive, `she throws a ball' transitive, `he puts it in the bin' di-transitive).
In her chapter, Conti-Ramsden approaches the question of whether certain kinds of pragmatic disability are independent of
syntactic difficulties and whether children who show these problems are different from those whose pragmatic disabilities are
secondary to their syntactic disability.

Semantic and Pragmatic Difficulties and Semantic-pragmatic Disorder/syndrome.

Having attempted a brief explanation of the terms `semantics' and `pragmatics', the next discussion concerns the use of two
similar but distinct terms: semantic and pragmatic difficulties as used in the title of this book, and semantic-pragmatic
disorder/syndrome as used by contributors consistently in the later chapters (Adams, Hyde-Wright and Cray).
Semantic and/or Pragmatic Difficulties
These terms can be used either singly or, more controversially, together, and imply no more than a description of those aspects
of language in which some children with language disability experience difficulty. The term implies nothing about the cause of
the difficulty or any conditions that give rise to or may be associated with these language problems.
Semantic-pragmatic Disorder/syndrome
The second term, semantic-pragmatic disorder/syndrome, implies, firstly, that semantic and pragmatic difficulties co-exist and,
secondly, that they are to be found in a group of children who have already been diagnosed as having a specific language
impairment: that is, that none of the factors known to cause language disabilities in childhood have been identified or can
explain the slow and disordered acquisition of language.
When the term `semantic-pragmatic disorder/syndrome' is used it usually implies a complex of linguistic symptoms and
associated behavioural and psychological characteristics, even though it is not yet certain that these can be

< previous page page_7 next page >

< previous page page_8 next page >
Page 8
listed or identified reliably. Adams, in her chapter, appears to argue a third position, that the term `semantic-pragmatic
disorder/syndrome' should refer only to the linguistic characteristics displayed by children within this diagnostic category and
should not carry any further associations with behavioural or psychological characteristics, since there is no agreement or
certainty about these.

Classifying Language Disability

To fully appreciate the distinction in the terms discussed above it is necessary to understand something about the framework
that is used to classify different kinds of language disability in childhood. The traditional medical approach is to classify on the
basis of the medical condition or impairment of function that is thought to give rise to the difficulty with language development.
The linguistic approach, on the other hand, is concerned with describing linguistic difficulties.
The Medical Approach
There are clearly a range of factors that explain why children are impaired in the acquisition of their mother tongue and in using
their language effectively to communicate. Many of these factors are known but not all are fully understood. Additionally, not
all of them lie within the child or are connected with disease or impairment, though many are. For example, it is known that
children who are prevented from gaining adequate exposure to models of the language either through social isolation or
profound deafness in infancy will have language disabilities. Slow-learning children and those with genetic syndromes are
known to have associated language disabilities though the connection is less well understood (e.g. Down's syndrome, Fragile X
syndrome). Still other children are born with, or acquire early in infancy, conditions such as autism that have profound
consequences for the development of language.
In all those cases listed above, the language disability is seen to be secondary to some other condition. These children can be
distinguished from children who fail to develop adequate language in the absence of any conditions or disadvantages known to
interfere with language acquisition. These latter children are said to have a specific language impairment, though this diagnosis
tells us little more than that certain causes of language delay and disorder have been excluded (at least as the sole explanation),
and tells us little about what these children may have in common.

< previous page page_8 next page >

< previous page page_9 next page >
Page 9
The Linguistic Approach
Another approach to childhood language disabilities is to group them on the basis of a description of the aspects of language that
are impaired. This approach is one favoured by linguists whose interest is primarily in describing and explaining the disordered
language simply in terms of failure to acquire, or apply, linguistic skills in the conventional way of the child's language
community. Such a descriptive approach is aimed primarily at defining the problems to be solved in intervention rather than
providing a taxonomy or classificatory system, though this can be done (see Wood, 1982). From a practitioner's point of view,
however, this may not be at all straightforward as several aspects of language may be disordered in any one individual and the
relationship between non-linguistic aspects of the child's development and their disability cannot be predicted from the pattern of
language disability alone. The practitioner cannot intervene without taking the learning capacities, behaviour and associated
impairments of the individual child into consideration.
Reconciling the Medical and the Linguistic Approach.
Bishop & Rosenbloom (1987) argue that these two approaches, the linguistic and the medical, are complementary. They put
forward a two-way classification of language disorders constructed by arranging causes or aetiology of language disabilities
along one axis, forming columns, and the primary areas of linguistic/speech impairment along the other, forming rows. `The
entries in the cells of the table indicate where a particular medical category is associated with a particular type of language
disorder' (Bishop & Rosenbloom, 1987).
The value of this table is that it illustrates clearly a vital point, namely that it is not only children with specific language
impairment (i.e. language disorders of unknown cause) who can show problems with either semantic or pragmatic areas of
language. If we take pragmatic difficulties alone, Bishop and Rosenbloom indicate that children with conditions such as Cocktail
Party syndrome and mild autism may show difficulties with semantic and pragmatic aspects of language. Cocktail Party
syndrome is a form of disorder that has been reported in children with hydrocephalus (Tew, 1979) where children exhibit fluent
and sophisticated language which on examination often appears devoid of meaning. There are other conditions also, not shown
in Bishop and Rosenbloom's table, where semantic and/or pragmatic disabilities have been described at some stage in
development. For example,

< previous page page_9 next page >

< previous page page_10 next page >
Page 10
recent research suggests that some children with Fragile X syndrome 1 show pragmatic disabilites in that their pragmatic
abilities appear delayed with some deviant features associated with a tendency to perseverate (Brunton, 1990).
This is why the descriptive term `semantic and pragmatic difficulties' was chosen for the title of this book: to indicate the
relevance of the contents to a broader group of children than those referred to by `semantic-pragmatic disorder/syndrome'. The
chapters by McTear and Crystal take a more cautious and descriptive approach, though McTear includes an interesting
discussion about possible causes of what he calls conversational disability. He prefers this term to either of those discussed
above, because he is unhappy with the bracketing together of two very different aspects of language without a clear
demonstration of the co-existence of difficulties.
Origins of the Term `semantic-pragmatic Disorder/syndrome'
If it is the case that there is little hard evidence for the association of semantic and pragmatic difficulties in children, this leads
us to ask where the term `semantic-pragmatic disorder/syndrome' came from. The term was first mentioned in the literature by
Rapin & Allen (1983) who were trying to sub-classify children who had been diagnosed as having a specific language
impairment. They wished to distinguish these children from others whose main difficulty centred on acquiring language form
(i.e. phonologic-syntactic syndrome). Other workers had described children with a similar pattern of difficulties but
distinguished them from other children with specific language impairment on different grounds. Bishop (1981), for example, has
stressed their relatively fluent and complex language when contrasted firstly with children whose primary difficulty was in using
language expressively, and secondly with those who lacked expressive language primarily because they had a specific difficulty
in understanding language.
Recognising Children with Semantic-pragmatic Disorder/syndrome
For professionals helping children with language disabilities, especially those not working in special schools or units, the
appearance in the journals and professional publications of articles and letters about a new kind of language difficulty must have
been puzzling. Was this a new name for an old problem or was it an entirely new set of difficulties? Practitioners began to
realise that the new terms applied to children whose difficulties in the past may

< previous page page_10 next page >

< previous page page_11 next page >
Page 11
well have been largely ignored or gone unrecognised. This, we suggest, may have been for any of a number of reasons. Perhaps
their difficulties were often masked by their social isolation. Alternatively, individual practitioners may only have encountered
single cases whose difficulties were so exceptional that they regarded them as idiosyncratic. Unles practitioners encountered
other children with similar difficulties they would not have recognised the problem that these children represented. Other
therapists and teachers may have considered that the difficulty was not primarily linguistic or attributable to a mild form of
autism. Some children, especially those first referred for delayed onset of language, may have been discharged as they began to
develop language because they showed competence in the very language abilities that therapists and teachers were working hard
to remediate in the majority of other language-disabled children. Finally, these children may not have been identified because
there was no specific clinical procedure for systematically assessing, describing and analysing their difficulties.
There are also several reasons why the existence of these children has become more generally acknowledged. One was renewed
efforts to discern sub-groups of children diagnosed rather crudely as having specific language impairment. First attempts to find
patterns of disability and groups sharing similar difficulties were carried out in special schools and units where access to
relatively large groups of children was assured. Once this group of children with similar problems was recognised, their
difficulties were brought to the attention of applied linguists to enlist their help in characterising the linguistic side of their
problems more effectively. At more or less the same time, speech pathologists began to develop research procedures and clinical
tools for assessing pragmatic abilities (e.g. Prutting, 1982; Prutting & Kirchner, 1987). It is to problems of assessment and
identification that we now turn.

Assessment and Identification

Several of the contributors to this book allude to the difficulty of finding appropriate methods of assessment for semantic, but
more especially pragmatic, difficulties. Several authors note that existing language assessments and standardised tests are of
limited value in capturing and measuring the nature and extent of such disabilities. There are three main reasons for this. Firstly,
there are no standardised tests as such for semantic and pragmatic abilities. Secondly, pragmatic abilities and disabilities can
only be revealed within a natural communicative context. And thirdly, and a fundamental issue, there are no standards of
appropriacy which we can confidently apply.

< previous page page_11 next page >

< previous page page_12 next page >
Page 12
Standardised Tests
With the exception of some more recent publications, for example Dewart & Summers (1988) and the Bristol Language Scales
(Gutfreund, Harrison & Wells, 1989), which are not `tests' in the generally accepted sense of this word, there are few, if any,
assessments that attempt to assess pragmatic aspects of communication. Thus we could not expect that most existing tests or
procedures would capture the nature of pragmatic disabilities. Such methods that do exist have largely been described in journal
articles (Prutting & Kirchner, 1987; Adams & Bishop, 1989) or broad analytical frameworks described in books, for example by
McTear (1985). Another strategy is to adapt methods designed primarily for research on specific aspects of pragmatic abilities
(see for example Prinz & Ferrier, 1983; Leonard, 1986; Brinton, Fujiki & Sonnenberg, 1988; Hargrove, Straka & Meddars,
1988).
A similar picture exists for assessment of semantic development in children. The Bristol Language Development Scales
(Gutfreund et al., 1989) provide some limited assessment of semantics. Crystal's Prism (1982) provides a more comprehensive
profile of semantics.
The Context of Assessment
The second point is that pragmatic abilities and disabilities can only be revealed within a natural communicative context. As
Prutting (1982) says, `it has been tempting for researchers and clinicians to want to control context rather than account for it'.
Standardised tests often exemplify what Prutting calls `context-stripped' situations. McTear makes a similar point in his chapter,
pointing out that conversations cannot be easily resolved to single or even paired turns of initiation and response. Nor can a
conversation be `standard' or `controlled' without losing sources of variation which add to rather than detract from its assessment
potential. The reasons for conversational difficulty can rarely be located within a single speaker since, by definition, a
conversation is a text created, over time, by at least two participants. Assessment must take place within the context which may
be giving difficulty, namely conversation, and conversation is something which is the product of joint endeavour and develops
its character through time. Its features also derive from the social context: the social relationship and roles of its participants;
their familiarity with one another; their shared history. These are not things that easily lead to standardisation. Indeed, most
standardised linguistic assessments try to exclude or control these aspects of communication and the conversation that takes
place between a subject and the tester is thus a

< previous page page_12 next page >

< previous page page_13 next page >
Page 13
highly specialised one with special characteristics and constraints. Children often attempt to reinstate the more familiar
conditions of conversation under testing conditions and use their conversational skills to override this artificial task.
Adams & Bishop (1989) tried to find a standard conversation situation for assessment in a research study. They provided means
to control the topic with a single partner, in a clinical situation, to assess certain aspects of children's conversational abilities in a
quantitative form. As they report, this attempt did not provide a great deal of information though their attempt to look at
different kinds of conversational inappropriacy (Bishop & Adams, 1989) was rather more productive. However, the precise basis
for the identification of inappropriacy has not been satisfactorily defined.
On the other hand, it could be argued from a clinical point of view, that in standardising the situation for research purposes,
Adams & Bishop (1989) actually exclude some of the variables that reveal both abilities and disabilities in conversation. The
variables include those associated with the topic, the situation and the conversational partner. Since the character of
conversations varies with the situation and its formality or purpose, it is important to know whether a child can alter his
conversation to be appropriate to different speakers who might employ different kinds of conversational styles, or who might be
more or less familiar, or of different age and status.
An alternative approach to assessment would try to explore these variables, recording naturally occurring conversations and
making observations in a broader range of situations. This would undoubtedly be more time-consuming but perhaps more
productive. The data could then be examined systematically over a range of pragmatic skills used in conversations. Such a
profile has been adapted from McTear (1985), for clinical use, by Milroy (1989). A case study employing these methods
(Wilcox, in progress) demonstrates how one boy, previously diagnosed as having semantic-pragmatic disorder, was shown to be
able to vary the forms he used with adults and children. In addition, while able to initiate and communicate with familiar
children within a language unit, he was less able to do so in integrated settings because unfamiliar children failed to recognise
his initiating moves. This information was invaluable in addressing remedial goals and strategies.
Standards of Appropriacy.
A further difficulty with analysing failures in conversation is that it is difficult to decide when a given instance is of clinical
significance. This is less

< previous page page_13 next page >

< previous page page_14 next page >
Page 14
of a problem with immature or deviant phonology or syntax where we have fairly clear developmental descriptions and models
to use as reference criteria. Failure in conversation is more difficult to identify as being deviant since descriptions of normal
conversational development are in their infancy, though the Adams & Bishop (1989) quantitative study provides some useful
data. More importantly, as McTear points out, all so-called normal conversations, particularly those between adults and
children, are subject to breakdown and repair. These breakdowns occur for a variety of reasons. Some of these, in the case of
children, have to do with the general immaturity of speakers, their immature phonology and syntax and their lack of experience
and skill in communication. Many examples of conversational failure or inappropriate utterances could be explained away as a
one-off or normal misunderstandings between children and adults, especially when the adult is a stranger with apparently rather
special aims and intentions. Wells (1983) has illustrated how a child with manifest communicative competence in her own home
can appear to be communicatively incompetent when faced with the special conversational aims of teachers. It is worth
emphasising the point that McTear makes eloquently, that single examples of conversational breakdown do not indicate
pathology. Pragmatic disability is only revealed through cumulative experience across a variety of different conversations with
different partners. Behavioural observation may indicate inability to initiate verbally where non-verbal behaviour suggests an
intention to communicate. Conversations begun may fail to develop, or the progress and direction of the conversation may be
repeatedly crippled through misunderstanding and the inability of either participant to repair the situation without abandoning
the topic or effort altogether. This insight must be absorbed if clinical assessment and identification of conversational difficulties
are to succeed.

Conclusion
This chapter provides information for non-specialists on those aspects of language that are referred to by the terms semantics
and pragmatics. An explanatory framework for the terminology used by each contributor has been given. Finally, issues that
arise directly from attempts to assess the nature of pragmatic and semantic disabilities have been considered.
This collection deals rather more with pragmatics than semantics. McTear questions the linking together of these two different
aspects of language. Very little has actually been said explicitly about the semantic difficulties of children in association with
evident conversational problems. It remains unclear which aspects are judged to reflect semantic difficulties.

< previous page page_14 next page >

< previous page page_15 next page >
Page 15
Difficulties in accessing words (word-finding difficulties) may have been the primary reason for this association, but there is no
particular justification for thinking that in semantic-pragmatic syndrome, word-finding difficulties represent a distinctive
feature. Children with a history of phonological and syntactic problems are often found to show these problems, and so too are
some children who bear the label developmental dyslexia.
Clearly, aspects of `meaning' are broadly involved in semantic-pragmatic syndrome, both in the reported tendency to interpret
speech acts from their form rather than the speaker's intention (though this is a pragmatic rather than a semantic failure) and in
the formation of so-called rigid concept boundaries. Here again, although this would appear to be a difficulty with lexical
semantics, it is arguable whether this is always a semantic problem, bearing in mind the distinction between concepts, a
psychological term, and the system of words used to signal those concepts, which is the linguistic aspect. Difficulties with
deixis, however, noted in children with specific language impairment would qualify as a semantic difficulty. Until it is possible
to demonstrate more clearly the distinctively semantic dimension of these disabilities it might be better to avoid using the term
in this hyphenated form but also, and more especially, in a descriptive way, if it is not justified.
Arguments about terminology seem set to continue but practitioners, teachers and therapists have to get on with the job of
assisting children whose language impairments are so demonstrably disabling. It would seem that we may learn about the
underlying difficulties and the causes of conversation problems by finding effective ways of treatment and education. This work
is in its early stages but is of no less importance than work that is addressing itself to describing and analysing the symptoms
and developing means of assessment. By their nature, difficulties with pragmatics must be tackled within the context of daily
communication. This point is very effectively illustrated in chapter 6 by the team from Dawn House School; they report on the
methods that are developing there, where a unique opportunity for intervention presented itself. However, there is no reason why
work of similar value could not be undertaken on a single-case basis, in any situation where teacher or therapist is faced with
children with semantic and/or pragmatic disabilities.

Note
1. A genetic syndrome causing mental retardation and associated delay in speech and language development with characteristic
physical and behavioural features.

< previous page page_15 next page >

< previous page page_16 next page >
Page 16
References
ADAMS, C. and BISHOP, D.V.M. 1989, Conversational characteristics of children with semantic and pragmatic disorder. I:
Exchange structure, turntaking, repairs and cohesion. British Journal of Disorders of Communication 24, 211-40.
BATES, E. 1976, Pragmatics and sociolinguistics in child language. In M. MOREHEAD and A. MOREHEAD (eds) Language
Deficiency in Children: Selected Readings. Baltimore: University Park Press.
BISHOP, D.V.M. 1981, Varieties of childhood language disorder. Northern Ireland Speech and Language Forum Journal 7, 20-
37.
BISHOP, D.V.M. and ADAMS, C. 1989, Conversational characteristics of children with semantic and pragmatic disorder. II:
What features lead to a judgment of inappropriacy? British Journal of Disorders of Communication 24, 241-64.
BISHOP, D. and ROSENBLOOM, L. 1987, Classification of childhood language disorders. In W. YULE and M. RUTTER
(eds) Language Development and Disorders. Clinics in Developmental Medicine 101/102. Oxford: MacKeith Press, Blackwell
Scientific Publications.
BLOOM, L. 1970, Language Development: Form and Function in Emerging Grammars. Cambridge, MA: MIT Press.
BRINTON, B., FUJIKI, M. and SONNENBERG, E.A. 1988, Responses to requests for clarification by linguistically normal and
language-impaired children in conversation. Journal of Speech and Hearing Disorders 53, 383-91.
BRUNTON, C.D. 1990, Phonological and pragmatic abilities of children with Fragile X Syndrome. Unpublished B.Sc.
Dissertation, Department of Speech, University of Newcastle-upon-Tyne.
CRYSTAL, D. 1982, PRISM. Profiling Linguistic Disability. London: Edward Arnold.
DEWART, H. and SUMMERS, S. 1988, The Pragmatics Profile of Early Communication Skills. Windsor: NFER-Nelson.
DUNN, J. and KENDRICK, C. 1982, The speech of two- and three-year-olds to infant siblings: `Baby Talk' and the context of
communication. Journal of Child Language 9, 579-595.
GUTFREUND, M., HARRISON, M. and WELLS, G. 1989, Bristol Language Development Scales. Windsor: NFER-Nelson.
HARGROVE P.M., STRAKA, E.M. and MEDDARS, E.G. 1988, Clarification requests of normal and language impaired
children. British Journal of Disorders of Communication 23, 51-62.
LEONARD, L.B. 1986, Conversational replies of children with specific language impairment. Journal of Speech and Hearing
Research 29, 114-19.
McTEAR, M. 1985, Children's Conversation. Oxford: Blackwell.
MILROY, L. 1989, Profile for analysing conversational ability. Unpublished manuscript, Department of Speech, University of
Newcastle-upon-Tyne.
MOGFORD, K. and BISHOP, D. 1988, Language development in unexceptional circumstances. In D. BISHOP and K.
MOGFORD (eds) Language Development in Exceptional Circumstances. Edinburgh: Churchill Livingstone.
PRINZ P.M. and FERRIER, L.J. 1983, `Can you give me that one?' The comprehension, production and judgement of directives
in language-impaired children. Journal of Speech and Hearing Disorders 48, 44-54.
PRUTTING, C.A. 1982, Pragmatics as social competence. Journal of Speech and Hearing Disorders 47, 123-34.

< previous page page_16 next page >

< previous page page_17 next page >
Page 17
PRUTTING, C. and KIRCHNER, D.M. 1987, A clinical appraisal of the pragmatic aspects of language. Journal of Speech and
Hearing Disorders 52, 105-19.
RAPIN, I. and ALLEN, D. 1983, Developmental language disorders: Nosologic considerations. In U. KIRK (ed.)
Neuropsychology of Language, Reading and Spelling. New York: Academic Press.
SACHS, J. and DEVIN, J. 1976, Young children's use of age appropriate speech styles in social interaction and role playing.
Journal of Child Language 3, 81-98.
SNOW, C.E. 1988, Conversations with children. In P. FLETCHER and M. GARMAN (eds) Language Acquisition, 2nd edn.
Cambridge: Cambridge University Press.
TEW, B. 1979, The `Cocktail Party Syndrome' in children with hydrocephalus and spina bifida. British Journal of Disorders of
Communication 14, 89-101.
WELLS, G. 1983, Talking with children: The complementary roles of parents and teachers. In M. DONALDSON, R. GRIEVE
and C. PRATT (eds) Early Childhood Development and Education. Readings in Psychology. Oxford: Blackwell.
WILCOX, A. (in progress) Assessment and intervention in semantic-pragmatic disorder: A case study. Department of Speech,
University of Newcastle-upon-Tyne.
WOOD, M.L. 1982, Language Disorders in School-age Children. Englewood Cliffs, NJ: Prentice-Hall.

< previous page page_17 next page >

< previous page page_18 next page >
Page 18

2
Is There Such a Thing as Conversational Disability?
MICHAEL F. McTEAR
Introduction
Conversation is such a pervasive human activity that we tend to take it for granted. Everywhere we look we can observe people
of all ages engaging in conversational activities. Some of these conversations are directed towards particular goals such as
asking for directions, giving instructions, or completing a business transaction. Other conversations seem to be less goal-
oriented and instead serve social purposes, such as initiating and maintaining social relationships. Whatever the purposes of
conversations, it is clear that they are an important aspect of our everyday activities and that any impairment in the ability to
participate in conversation is likely to have serious consequences for the person concerned.
However, despite the fact that people normally engage in conversation quite successfully and without giving it a moment's
thought, there are many occasions where conversation breaks down. Most of the time these breakdowns are fairly trivial and
there are accepted ways of overcoming them. For example: two people might start speaking at the same time, thus violating one
of the principles of conversational turn-taking. Another example is where one speaker shifts topic without making this obvious
to the other person, so that confusion might result. Or a speaker might wrongly assume that the listener knows some other
person and refers to that person by name without explaining who the person is. These sorts of breakdown occur routinely in
conversation and there are equally routine repair mechanisms available to conversationalists to put the conversation back on a
smooth path, usually without anyone noticing that anything unusual has occurred.
In recent years there has been an upsurge in studies of conversation how it is structured, how it can break down, and how these
breakdowns can be

< previous page page_18 next page >

< previous page page_19 next page >
Page 19
repaired. Linguists, psychologists, sociologists, anthropologists, and computer scientists working in artificial intelligence have
contributed to our understanding of conversation (see, for example, McLaughlin, 1984; Wardhaugh, 1985; Beattie, 1983;
Atkinson & Heritage, 1984; McTear, 1987; Reilly, 1987). There have also been many studies of the development of
conversation in young children (see, for example, Ervin-Tripp & Mitchell-Kernan, 1977; Ochs & Schieffelin, 1979; McTear,
1985), while books on conversational disability in children are just beginning to appear (McTear & Conti-Ramsden, 1991;
Smith & Leinonen, 1991).
This chapter will examine cases of conversational breakdown in which a person's ability to participate in conversation is
seriously impaired. Conversations involving such persons are fraught with misunderstandings, confusion and frustration. The
chapter will focus on conversational disability in children, although it is clear that similar problems are also to be found in
conversations involving adults.
Some Examples.
It might be helpful at this point to give some examples of conversations in which one of the participants appears to be
conversationally impaired. In these examples it will be seen how some children's use of language tends to disrupt the normal
flow of conversation. They may fail to respond appropriately to what their dialogue partner has just said; they may pay undue
attention to the literal meaning of what is said rather than its intended meaning; they may be unaware of what their listener
knows and does not know; they engage in odd associations and reasoning; they are unable to detect and repair
misunderstandings.
Example 1
(1) Adult: did you go camping in the woods?
(2) Child: camping in the woods? [shouts]
(3) Adult: yeah, did you ever do that?
(4) Child: yeah, he ever do that
.
.
.
(5) Adult: when are you going home?
(6) Child: um he IS going home
This example is taken from a case study of an eight-year-old who had presented severe communication difficulties (Greenlee,
1981). One problem

< previous page page_19 next page >

< previous page page_20 next page >
Page 20
which had been noted concerned his responses to questions. Almost half of the questions addressed to the child received no
response, while a quarter received inappropriate or unrelated responses. The child's utterance in line (2) is a case in point. We
cannot be sure what the child's intent behind this utterance might have been. It may have been intended as a request for
confirmation; however, the unusual loudness gives rise to a feeling of inappropriacy. This feeling is further supported as the
conversation proceeds, as lines (4) and (6) introduce further difficulties. Both utterances are complicated by problems with the
use of pronouns. Bearing this in mind, (4) might be an appropriate response in terms of accuracy of content; (6) is clearly
deviant as it fails to respond appropriately to the when question.
Example 2
[The conversation takes place in Nottingham, about three hours travelling time from London]
(1) Child: we go home every day
(2) Adult: do you go home every day?
(3) Child: yes
(4) Adult: where do you live Stephen?
(5) Child: London
(6) Adult: so you don't go home every day
This example, taken from an article by Stubbs (1986), illustrates a problem of odd reasoning in which the child appears to have
difficulty with temporal and causal knowledge. Here the problem is that the child's responses are implausible, as it is unlikely
that he makes a three-hour journey home every day. One possibility is that there is some ambiguity in the use of the word home
here the child might have been referring to his place of residence during the week. Further evidence would be necessary to
support our feeling of inappropriacy, although similar cases of odd reasoning have been widely noted in children with severe
communication difficulties.
Example 3
[The child is discussing with his speech therapist a forthcoming sports day at school]
(1) Adult: which race would you like to be in?
(2) Child: I like to be in Longtree in the sports day [Longtree is a town several miles from the school]
(3) Adult: in Longtree?
(4) Child: yes
(5) Adult: what do you mean?
(6) Child: I mean something?

< previous page page_20 next page >

< previous page page_21 next page >
Page 21
(7) Adult: is there a sports day in Longtree?
(8) Child: there is not there is a sports day in Abbeyvale [Abbeyvale is the school]
(9) Adult: then what's Longtree got to do with it?
(10) Child: nothing
(11) Adult: then why did you mention it?
(12) Child: indeed I did mention it
(13) Adult: why did you mention it?
(14) Child: I don't know
This example, taken from McTear (1985), shows a similar problem with odd reasoning as well as an inability to resolve
misunderstandings. Apart from the inconsistency over where the child intended to be on sports day, this conversation develops
into a tangle of misunderstandings out of which neither the speech therapist nor the child can successfully emerge.
One final example, taken from a report by Maureen Aarons (personal communication), involves a 12-year-old child meeting his
speech therapist for the first time. The child said: can you see how I look different? The therapist was unable to answer and the
child continued I've had my hair cut. Obviously the child had difficulty in assessing accurately what his listener might or might
not know specifically, that she could not know how he was different as this was the first time they had met. For this reason the
exchange appears odd.
These examples could be multiplied ad infinitum and speech therapists and teachers should be able to come up with their own
examples. A few points are worth noting at this stage:
1. There is a striking agreement among speech therapists and teachers that they have experienced children whose conversations
are similar to these.
2. Although it is often possible to explain away any individual example in terms of a simple misunderstanding or the child being
under pressure and thus performing under par, for example, the cumulative effect of many such instances from the same child
gives the impression of some sort of difficulty in the use of language which is in some way abnormal when compared with
children of a similar age.
3. It is often necessary to have a clear record of the context and the background information in order to appreciate what the
problem is, or indeed that there is a problem at all (as in Example 2, where it is necessary to have the real-world knowledge
about where the child lives and where he attends school).
4. Although the examples all involve an impairment affecting conversation, the explanations for these impairments seem to vary;
thus it may be that there are different factors which can singly or in combination give rise to conversational disability.

< previous page page_21 next page >

< previous page page_22 next page >
Page 22
The Nature of Conversational Disability
These examples suggest that some children are impaired in their ability to engage in conversation. We can note that the children
in these examples do not appear to have any obvious difficulties in structural aspects of language for example, their utterances
are syntactically well-formed. For this reason a distinction has often been drawn between those children who have difficulties
with structural aspects of language and those who have problems with the use of language. Rapin & Allen (1987), for example,
distinguished between children with phonologic-syntactic syndrome, who were characterised by difficulties with language form
but who did not have problems with using language for communication, and children with semantic-pragmatic disorder, who
had relatively little difficulty with language form, but experienced problems in using and understanding language in
communicative situations. Bishop & Rosenbloom (1987) have provided a useful list of the features typically reported for this
second group of children:
1. Early history may include echolalia, jargon, poor symbolic skills, poor social relationships, hyperactivity.
2. Delayed speech and language development, but once language develops it appears clear and fluent, with syntax and
phonology developing fairly normally but with problems in semantics and pragmatics, for example:
a. odd associations and reasoning;
b. tangential or inappropriate remarks;
c. undue attention to literal rather than underlying meaning;
d. problems in understanding normal conversation, for example, descriptions of sequences of everyday events which are related
temporally or causally;
e. lack of awareness of what knowledge is shared between speaker and hearer.
3. Poor social skills either withdrawn or eccentric and over-friendly.
We can note in passing that the term `semantic-pragmatic' has been used to characterise impairment in the use of language. This
is, however, an unfortunate term as it encompasses two distinct levels of language. Normally, semantics is used to refer to those
aspects of meaning which are conventional and part of a language system, while pragmatics is used to refer to those aspects of
meaning which are non-conventional: for example, the meanings which an utterance might have in a particular context which
are different from its literal meaning. In this chapter the more neutral term conversational disability will be used with the focus
on the following questions:
1. Is conversational disability a unitary phenomenon or are there different types which may or may not co-occur in a particular
child?

< previous page page_22 next page >

< previous page page_23 next page >
Page 23
2. How does conversational disability relate to a child's level of linguistic ability?
3. How does conversational disability relate to non-linguistic abilities, in particular, to a child's cognitive and socio-cognitive
abilities?
An examination of these questions should lead to a deeper understanding of the nature of conversational disability and should
provide a more solid foundation for decisions regarding assessment and intervention.

Conversational Disability: Is It A Unitary Phenomenon?

It seems reasonable to assume that there might be different types of conversational disability, as it is generally agreed that
conversation involves the integration of a variety of abilities, including linguistic, cognitive and social skills (Shatz, 1983). Thus
a child could be competent in one skill relevant to conversation but deficient in others. Or, alternatively, a child could be
competent in each of the relevant skills but unable to integrate them appropriately to serve the demands of communication.
A useful starting point is the following categorisation of sub-types of conversational disability:
1. Problems due to linguistic impairment.
2. Problems involving conversational rules.
3. Cognitive deficits affecting the use of language.
4. Socio-cognitive deficiencies.
5. Affective and emotional difficulties.
Some of the problems which children have with the use of language seem to be primarily linguistic in origin, such as the
inability to provide an accurate and discriminating description because of a lack of the necessary vocabulary and grammatical
constructions. In these cases the child's difficulties with language give rise to problems in engaging effectively in conversation.
Other problems may have little to do with actual linguistic ability but more with specific aspects of conversational competence,
such as taking turns in conversation, performing speech acts appropriately, or observing the principles of co-operative
conversation. In other cases children appear to have cognitive deficits affecting their common-sense world knowledge, so that
they are unable to make sense of what happens around them or to make inferences about information that is not stated explicitly
in the conversation. A different problem is where children appear to have deficiencies in socio-cognitive aspects of
communication, as in their inability to assess the beliefs and intentions of other people and to predict their behaviour in everyday
social

< previous page page_23 next page >

< previous page page_24 next page >
Page 24
interactions. And finally, there can be affective or emotional problems, which may be due to previous communicative failures
and which can result in socially inept behaviour, such as social withdrawal or inappropriate over-friendliness. These different
aspects of conversational disability will now be described in greater detail in the following sections.
Conversational Problems Due to Linguistic Impairment
It would seem obvious that impairment in language might lead to conversational problems, as some knowledge of linguistic
structure is necessary for conversational tasks such as making requests, producing narratives, or describing objects, persons and
events. Indeed most research on conversational disability has been based on the assumption that deficiencies in language
structure will result in conversational difficulties (see Conti-Ramsden, this volume). For example: in group design studies, in
which language-impaired children are compared with normally developing children matched for age as well as with younger
children matched for language level, it is predicted that the use of language by the language-impaired group will be similar to
that of their language-matched peers. In other words, children whose language is less developed will be similar in their use of
language, irrespective of differences in age or life experiences. In many cases, this prediction is upheld.
There are many conversational tasks which require some degree of linguistic ability. Referring to persons or objects not present
in the immediate environment can be achieved by the use of a relative clause as in the boy who took my pencil, in which who
took my pencil specifies which boy is being referred to. Highlighting the main informational content of a message requires the
ability to distinguish between old and new information in relation to the conversation so far. Knowing the need for this
distinction is one part of the problem; however, being able to make the distinction requires an accurate usage of definite and
indefinite articles as well as the correct placement of the nuclear tone. For the listener, the problem is in reverse recognising the
significance of these linguistic cues for the interpretation of the discourse. Similarly, responding to requests for clarification can
involve more than just repeating the problematic utterance: in many cases, it is necessary to paraphrase in some way. This can
involve re-ordering sentence constituents, changing sentence mood (e.g. from imperative to interrogative), as well as the use of
ellipsis in order to avoid repeating parts of the utterance which are judged to have been understood already. As a final example,
the ability to make polite request forms depends on the prior acquisition of verb forms such

< previous page page_24 next page >

< previous page page_25 next page >
Page 25
as could and would as well as complex grammatical structures such as would you mind + verb + -ing (e.g. `would you mind
shutting the door').
It is interesting that some children who have conversational difficulties also have word-finding problems, for example, in
targeting the required words. In some cases, as shown in a wide range of naming tasks, this seems to be a result of not being
able to access known words sufficiently rapidly (German, 1979). This disability often leads to the adoption of compensatory
strategies, such as circumlocution, or the use of imprecise words, and other indications of word-finding difficulty such as
hesitations and false starts (Wiig & Semel, 1984). The effects on social interaction of such word-finding problems can be
serious. For example: Silliman (1984) reports that the classmates of language-learning disabled children who had such word-
finding problems found great difficulty in attending to the narratives of the learning-disabled children on account of the
numerous dysfluencies. Similar social consequences could be expected in other contexts.
Children may also have difficulties because of unusual or over-restricted use of lexical items, such as spatial prepositions
(Smedley, 1989). The problems may be due to `rigid concept boundaries'. For example: a child saying the clock is by the wall
may believe that on is used only to refer to locations on the horizontal plane. More generally, children may restrict their use of
words to their most obvious, literal meanings so that their expressive language is limited or they are unable to understand less
literal and metaphorical meanings of words.
Deficiencies in language for expressing temporal and causal relations will affect children's ability to order events in a narrative
and to perceive relations between events involving consequence or motivation. Children may have difficulty in understanding
and responding appropriately to when, how and why questions, or they may use because to mean and or so (for example: it is six
o'clock because Mike is getting up). Smedley (1989) found children with these difficulties whose non-verbal cognitive processes
were not impaired, which suggests that their problems were linguistic, with consequences for their ability to produce and
comprehend extended discourse.
Language-impaired children have particular difficulties with certain aspects of syntax and tend to omit determiners, auxiliaries
and sentence subjects in contexts where these are obligatory (Fletcher & Garman, 1988). Omission of determiners means that
the child will not be able to mark the distinction between old and new information which enables listeners to follow the
connections in a piece of discourse. A lack of auxiliaries means that the child is unable to express certain meanings conveyed by
auxiliary verbs, such as degree of possibility, while omission of the sentence subject makes it

< previous page page_25 next page >

< previous page page_26 next page >
Page 26
difficult for the listener to retrieve the referent of the omitted subject. As with other examples, it is more problematic if we wish
to explain these omissions. A failure to use determiners could be a result of the child not yet being aware of a need to
distinguish between old and new information, while omission of sentence subjects may be due to the child's assumption that, if
the referent of the sentence subject is mutually available to speaker and listener, then there is no need to specify it. Thus
syntactic immaturity may give rise to conversational deficiencies, although equally the child's syntactic deficiencies may be
attributable to conversational factors.
Children's linguistic deficiencies may also affect their ability to participate effectively in conversation. In a study of the turn-
taking abilities of language-impaired children, Craig & Evans (1989) found that the children had difficulties in those aspects of
turn-taking which required specific lingustic abilities, in particular, the skills involved in encoding and decoding sufficiently
rapidly for precision-timed turn-taking. A difficulty with conversational initiations could also be a result of linguistic
deficiencies which make it difficult for a child to introduce new topics appropriately or to describe the referents of the discourse
for the listener. As far as speech acts are concerned, language-impaired children often seem to be characterised by a restricted
range of speech acts. For example, it has been frequently reported that language-impaired children fail to use language to
describe and spontaneously inform; instead they use language mainly to direct the listener's attention. In other cases language-
impaired children are non-assertive in conversation and their contributions are limited to back channel responses (i.e. verbal and
non-verbal devices such as umhmm and nodding, which indicate attention without taking the conversational initiative). If the
children's range of speech acts is inferior to that of their age-matched peers but equals that of their language-matched peers, then
it is reasonable to attribute their problems to linguistic deficiencies.
There are, however, other possible explanations. The adoption of a passive role in conversation could be a consequence of poor
comprehension skills, as an inability to understand the messages of their partners would make conversational participation
difficult for these children. To test this hypothesis it would be necessary to test the children's comprehension skills. An
alternative explanation, in terms of interactional factors, is suggested by a study of adult-child interaction which showed that the
mothers of language-impaired children with normal comprehension initiated dialogue significantly more often than mothers of
non-language-impaired children, while the language-impaired children initiated dialogue significantly less often than the non-
language-impaired children (Conti-Ramsden, 1990). The mothers of the language-impaired children used more questions and
directives, thus forcing

< previous page page_26 next page >

< previous page page_27 next page >
Page 27
their children into a more passive role, possibly in an effort to engage the child more effectively in dialogue and to maintain the
interaction. Conversely, the mothers of the non-language-impaired children tended to use semantically related utterances such as
recasts and other contingent replies which are less regulative discourse devices. Thus the passive role adopted by the language-
impaired child can be explained in terms of interactional factors, such as different maternal strategies which are in turn a
response to the children's speech characteristics, such as restricted intelligibility on the part of the language-impaired children.
Finally, there are socio-affective explanations, particularly in the case of autistic children, who are often reported as being non-
assertive in conversation. In this case the explanation lies not in their linguistic ability so much as in a lack of the required
sociability. So the same symptoms can have different origins and the relationship between the observable behaviour and its
explanation can only be determined by obtaining relevant supporting data, whether these are scores of comprehension or
sociability ratings, or by careful analysis of the interactional context and the respective contributions of the conversational
participants.
Linguistic impairment can also affect a child's ability to make appropriate conversational adjustments and revisions. Modifying
syntactic constructions to take account of different types of listener or to express various degrees of indirectness may place
undue demands on the linguistic resources of language-impaired children (Fey & Leonard, 1983). Similarly, the ability to revise
utterances in response to clarification requests by using paraphrases or lexical substitutions is restricted for children with limited
linguistic knowledge (Gallagher & Darnton, 1978; Brinton, Fujiki, Loeb & Winkler, 1986; Brinton, Fujiki & Sonnenberg,
1988). However, as far as adjustments to the listener are concerned, there is the alternative explanation of an inability to assess
the listener's needs. To rule out this possibility it would be necessary to have independent evidence that the child's socio-
cognitive skills were unimpaired.
It is clear from the examples presented here that the ability of some children to use language effectively in conversation is
restricted by their deficient linguistic resources. However, some caution is required when suggesting relationships between
linguistic abilities and the use of language. The following points are worth noting.
1. We need to examine carefully the correspondence between particular syntactic constructions and particular conversational
skills. Some obvious examples have already been discussed, such as: the use of the definite article to mark information which is
mutually known to speaker and listener; the use of interrogative constructions in certain types of request for action and in

< previous page page_27 next page >

< previous page page_28 next page >
Page 28
requests for clarification, or the use of relative clauses in providing more specific information about a referent. However, there
are many cases where no clear correspondences are to be found. What, for example, is the syntactic correlate of a persuasive
appeal or of the justification of a non-compliant response to a request?
2. In relating linguistic knowledge to the use of language, we need to be clear about what we mean by knowledge of linguistic
constructions. In many studies a global estimate of syntactic ability is used in some cases, it is as global as Mean Length of
Utterance (MLU). Much more precise information is required: for example, whether the child uses a particular structure, such as
relative clauses, for other functions but not for the function under investigation; whether knowledge of the construction is
inferred from evidence of the child's linguistic production or comprehension; whether the construction is used spontaneously; or
whether it can be traced to an imitation of some previous adult utterance (whether the imitation is partial, deferred, extended,
etc.). In the absence of such information the claim that the child is syntactically competent is uninformative.
3. Finally, as already suggested, it is necessary to determine where the problem lies. If a child fails to produce polite requests, is
it because of a deficiency in the required syntactic constructions, an inability to use them when required, or a lack of social
judgement (for example, inability to assess listener attributes or to appreciate the implications of differences in these for socially
appropriate speech)? In a study by Donahue (1981), children were required to produce requests to listeners of different ages or
social status. Language-learning disabled boys produced more polite requests to peers than to high-power targets as well as
fewer and a smaller range of persuasive appeals. However, the fact that they made any adjustments at all demonstrated that they
had an adequate linguistic repertoire. The fact that they produced different forms according to listener type also demonstrated
that they had the social ability to perceive distinctive features of listeners. What they lacked was the ability to appreciate the
implications of these differences for socially appropriate speech. In other words, the question of how to apply language
knowledge in conversation requires an understanding of the social and socio-cognitive issues involved a more complex theory
than what is implied by a simple knowledge/use dichotomy.
Problems Involving Conversational Rules
In addition to linguistic problems which give rise to conversational disability, a child may have difficulty with specific
conversational rules which

< previous page page_28 next page >

< previous page page_29 next page >
Page 29
are independent of linguistic knowledge. These include turn allocation, attention-getting and attention-directing, using speech
acts appropriately, and clarifying misunderstandings.
The ability to assign and take turns in conversation evolves out of reciprocal interactions between caregivers and young infants
in the prelinguistic stage. Turn allocation is dependent not on linguistic ability but rather on the recognition that it is appropriate
to take turns in conversation. Language-impaired children do not appear to have difficulties in turn allocation. Kysela,
Holdgrafer, McCarthy & Stewart (1990) found that language-impaired children at the one-word stage compared favourably with
their normally developing peers in their ability to take turns in conversation. Similar findings for older children are reported by
Craig & Evans (1989) as far as sentence-initial overlaps were concerned that is, there were few cases in which adults and
children attempted to start talking at the same time and any cases of simultaneous speech arising from such turn exchange errors
were usually repaired by the child giving up the floor. On the other hand, it has been reported that other groups of children, in
particular, autistic children, have specific problems with appropriate turn-taking (Fay & Schuler, 1980). Baltaxe (1977) found
that autistic children have difficulty in moving between speaker and hearer roles in conversation, with the result that they may
remain in the speaker role for too long (Bernard-Opitz, 1982) or in the hearer role for too long (McCaleb & Prizant, 1985).
Autistic children also fail to use eye contact to signal turn-taking (Mirenda, Donnellan & Yoder, 1983). Thus autistic children
appear to have a specific difficulty with turn-taking which is independent of their language abilities.
Initiating conversational exchanges depends on specific conversational knowledge. Attention-getting requires the ability to
determine whether the listener is already attending or not, as well as the selection of an appropriate attention-getting device.
Autistic children seem to be deficient in the use of appropriate attention-getting devices such as establishing eye contact or using
the listener's name (Fay & Schuler, 1980). As far as attention-directing is concerned, children with learning difficulties appear to
be deficient in providing information which would help their listener to identify the referents of their messages, a deficiency
beyond what would be predicted by their linguistic and cognitive abilities (King, 1989), while autistic children have difficulties
with distinguishing between old and new information (Baltaxe, 1977) and tend to initiate inappropriately by asking the same
question over and over again (Hurtig, Ensrud & Tomblin, 1982).
There are several possible explanations for these difficulties. The children may fail to appreciate the conversational principles
involved in initiating

< previous page page_29 next page >

< previous page page_30 next page >
Page 30
conversation, such as the need to get the listener's attention, to adequately specify the objects referred to in the message, or to
assess the appropriacy of a question. On the other hand, a child who is aware of these principles may still have difficulties in
applying them appropriately because of a failure to assess the listener's needs whether the listener is already attending, can
identify the objects being referred to, or has understood information which has been conveyed. An inability to assess the
listener's needs rather than to be aware of which conversational devices to use is a consequence of socio-cognitive rather than
conversational difficulties.
Some children adopt a mainly non-responsive or a non-assertive role in conversation. An unwillingness to initiate may be due to
various factors other than the child being unaware of the social obligation to be responsive in conversation. Non-responsiveness
may be due to personality factors, as in the case of an extremely shy or withdrawn child. Alternatively, this unwillingness may
be the result of previous communication failures which have convinced the child that defensive strategies in conversation will
decrease the likelihood of further such failures (Bryan, 1986; Donahue, 1987). Another possibility is that the child has linguistic
deficiencies which make responding in conversation and maintaining an ongoing topic or introducing a new one problematic
(Rosinski-McClendon & Newhoff, 1987). Indeed, the cause of the difficulty could involve a combination of these factors.
Using and understanding speech acts in conversation depends on conversational knowledge being aware of the appropriate
conditions for the use of speech acts and using contextual information to distinguish between literal and indirect meanings of
utterances. Using speech acts appropriately requires social knowledge of when a particular form of a speech act should be used.
Autistic children often violate conversational principles of acceptability and politeness (Baltaxe, 1977). Similarly, language-
learning disabled children have difficulty in making adjustments in conversational style to suit the characteristics of their
listeners or the nature of the task in hand (Donahue, 1981). However, it is important to examine the causes of these deficiencies
carefully. To use a speech act appropriately a child must have:
1. The linguistic resources;
2. The ability to discriminate between listener and situational features;
3. An appreciation of how to vary speech according to these features.
As far as the comprehension of speech acts is concerned, language-impaired children, and autistic children in particular, have
difficulty with the comprehension of indirect speech acts and tend to respond to the literal rather

< previous page page_30 next page >

< previous page page_31 next page >
Page 31
than indirect meanings. Thus being able to decide whether an utterance such as can you put the ball into the truck is a question
or a request may depend on the extent to which the child can use prior context or can combine evidence from the current
situation, as well as an assessment of the speaker's likely intentions (Shatz, Bernstein & Shulman, 1980).
Clarification sequences involve a combination of linguistic and conversational knowledge. Clarification requests occur when a
listener realises that an utterance is unclear. This involves the linguistic ability to process utterances and the conversational
ability to interpret them relative to the current context. Responding to clarification requests requires linguistic skills for making
revisions as well as the conversational skills for adjusting the utterance to the needs of the listener and the situation.
Understanding the obligations of speakers and listeners to keep conversation flowing smoothly and to repair misunderstanding is
a basic conversational skill. Language-impaired children appear to be aware of the obligation to respond to and make
clarification requests, although the level of sophistication of their requests and responses is determined by their linguistic
abilities and they seem less aware of the needs of their listeners (Brinton et al., 1986; Brinton, Fujiki & Sonnenberg, 1988).
However, some children appear to be deficient in the production of clarification requests even when they have the necessary
linguistic skills. For example: in experimental studies children have failed to request clarification of ambiguous or unclear
messages, even in situations where, as a result, they would be unable to select the correct object (Donahue, Pearl & Bryan,
1980). Several explanations are possible. Perhaps the children failed to notice that the messages were inadequate: this
explanation was ruled out in one study in which the children demonstrated in a follow-up task that they could recognise
message inadequacy (Donahue, Pearl & Bryan, 1980). Another possibility is that the children avoided clarification requests as
they did not wish to draw attention to their inability to understand possibly as a result of previous experiences of communication
failure. Thus conversational factors may combine with linguistic impairment and lead to a failure to perform appropriately in
everyday conversation.
In summary: there are some specific conversational rules which prove difficult for children and which affect their ability to use
language effectively in communication. However, these factors are often difficult to disentangle from other problems such as
linguistic impairment or an inability to adjust to the listener's needs. Careful examination of the evidence and subtle
experimental techniques are required to enable clinicians and researchers to uncover the source of the difficulties experienced by
a particular child or groups of children.

< previous page page_31 next page >

< previous page page_32 next page >
Page 32
Cognitive Deficits Affecting the Use of Language
It has been argued that the ability to use language depends on appropriate world knowledge. Background (or world) knowledge
is what enables people to make sense of what is said in a conversation by relating it to previous experience and by making
predictions based on knowledge about objects and persons in the everyday world. For example: to be able to make sense of a
story about everyday events, such as a shopping expedition, we need to have some knowledge of what sequences of events are
likely to occur, what the likely participants in the events will be, and so on. In other words, we need to have knowledge of what
Schank & Abelson (1977) have described as scripts, goals and plans, or what Nelson (1986) has described as general event
representations.
Event knowledge enables us to make sense of stories and events in everyday life in several ways:
1. We can make predictions from what we have just heard to what we are likely to hear next and so constrain the range of
possible next utterances, thus reducing the load on our processing capacity.
2. We can make inferences from what we have actually heard to what is likely to lie behind what was said. For example: if we
hear that someone went shopping, we can assume, without being told, that they are likely to have spent some money and
acquired some goods in return.
3. We can make sense of deviations from the normal course of events for example, if someone has gone shopping and has
returned empty-handed, we can assume they were unable to obtain what they were looking for.
4. We can re-organise our cognitive structures in the light of new experiences to provide more general or more detailed
structures.
There is a wide range of evidence from cognitive psychology to support the view that world knowledge is used to assist
language understanding (see, for example, Bransford & McCarrell, 1974). There is also more recent evidence that young
children have well-organised knowledge about familiar events involving temporally and causally related sequences. Olson &
Nickerson (1979) showed how children use world knowledge to aid their comprehension of stories. Another example would be
the types of routine in which children learn how to greet, take leave, give thanks, as well as engage in other socially appropriate
and culturally specific forms of behaviour (Gleason & Weintraub, 1976). As far as conversation is concerned, Nelson &
Gruendel (1979) have suggested that children have scripts for activities that they talk about and that these scripts enable them to
conduct a more coherent dialogue. Where scripts are held in common with other children or adults, interaction is

< previous page page_32 next page >

< previous page page_33 next page >
Page 33
likely to be more predictable as some degree of mutual understanding can be assumed. Indeed, as Nelson & Gruendel found,
there is considerable commonality among children in terms of their knowledge of familiar events involving causally related
sequences. This knowledge is to be found in children as young as three years, and even though older children produced more
elaborate scripts, these were based on the same basic scripts as those of younger children.
It has frequently been observed that some children have difficulty in processing extended discourse such as stories and in
keeping track of topics during the course of a conversation, in spite of normal comprehension abilities when tested for words
and isolated sentences. One study which examines the effect of deficient world knowledge on a child's use of language is by
McTear (1989), in which it is shown how a 10-year-old child appeared unable to discern connections between everyday events
and to go beyond the surface characteristics of an action or an utterance to its deeper, underlying significance. The analysis was
based on a conversation which arose during a description task involving a problem situation depicted on a card and based on an
everyday social scene. In this case the picture showed a child standing on the doorstep of a house, looking rather lost and having
emptied the contents of his pockets on the ground beside him. To understand the situation it is necessary to go beyond these
surface details and make the inference that the child has lost his key and is unable to enter the house. However, the subject in
the case study failed totally to make this inference and persisted in describing superficial details in the picture, such as the boy's
pencil which was lying on the ground. It was only after considerable prompting that he was able to show evidence of
understanding the picture.
What was interesting in this study was that, although the subject failed to make the required inferences to demonstrate an
understanding of the situation, his responses were generally reasonable and appropriate on their own terms. For example, he
inferred that the boy had lost his dinner money, was locked out and had to wait for his mother, and could try to get in by the
back door. He was able to reason about everyday events and to understand relations between events at a local level such as
losing the dinner money and not being able to have dinner but was unable to integrate these isolated pieces of knowledge into a
larger coherent structure.
In order to appreciate this point, it is necessary to examine what knowledge is required to understand the event depicted in this
task. In the present case, the fact that the child is standing outside the house leads to the inference that he has the goal of
entering the house. However, this is not a logical inference and it could be cancelled on the basis of subsequent

< previous page page_33 next page >

< previous page page_34 next page >
Page 34
information. For example: the child may have been looking at the house because he heard a noise from inside, while emptying
the contents of his pockets on the ground may have been due to something unrelated, such as looking for his bus fare. However,
in the absence of this information it is usual to make the default inference that the child has the goal of entering the house. This
being so, and given that he seems unable to achieve his goal, we can assume that there is some problem for example, there is no
one at home or he has lost his key. We can see each of the actions in this scene as having various preconditions and effects. For
example: to be able to enter a house a person either needs a key or there should be someone to open the door. To use a key you
have to be able to find it. To find it you may have to look in your pockets, and so on. McTear (1989) presents a goal tree setting
out some of the knowledge about entering a house on which a child would have to draw to understand the depicted situation. As
only some of the information is represented in the picture, it is necessary to use world knowledge to infer that information which
is implicit. The subject in this case study appeared to be deficient in this ability to reason about implicit information.
Of course, many questions are raised when we begin to consider the relationship between a child's world knowledge and
language use. How has the child come to have deficient world knowledge in the first place? As might be imagined, an
understanding of these issues requires a comprehensive theory of understanding. Most work in cognitive psychology focuses on
components of such a theory without reference to the whole theory. Very little work has been carried out in speech pathology in
relation to these issues, so this is an area where fruitful co-operation is possible between speech pathology, cognitive psychology
and artificial intelligence.
Socio-cognitive Deficiencies.
In order to participate effectively in conversation a person has to be able to make social inferences about the actions, beliefs and
intentions of other persons in order to understand their behaviour and to be able to adapt messages to their needs.
There are different levels of social inference. The most basic level involves the ability to recognise social attributes of listeners
and to adjust messages accordingly. The studies of Donahue and colleagues suggest that children who are language-learning
disabled are able to vary requests according to differences in listener status, although boys seemed insensitive to differences in
age when giving instructions to peers and younger listeners (Bryan, 1986;

< previous page page_34 next page >

< previous page page_35 next page >
Page 35
Donahue, 1987). However, more research is needed on this aspect of communicative competence.
Experimental studies of children's role-taking abilities have shown how younger children often find difficulty in adjusting their
messages to the knowledge level of their listeners. Language-learning disabled children are less likely to convey information
which is appropriate and adequate. As indicated earlier, these problems may be due to linguistic impairments or to an inability to
assess the listener's requirements and provide a description which is informative.
The most sophisticated level of social inferencing has been investigated in studies of the development in young children of a
`theory of mind', which enables children to infer mental states such as beliefs and intentions from the observable behaviour of
other people and to make predictions or adjust their messages on the basis of these inferences. For example: in one study
(Wimmer & Perner, 1983) the child subject, along with another person, observes a set of events (for example, a mother putting
some chocolates in a round tin). Then, in the absence of the other person, the subject observes an unexpected change in the
events (for example, the mother moves the chocolates to a square tin). The subject knows that the chocolates are in the square
tin but should predict that the other person will look for them in the round tin, because the subject should realise that the other
person, not having observed the mother's action of changing the location of the chocolates, still believes that they are in the
round tin. Thus, to succeed in this task, the subject has to distinguish between his or her own beliefs and the false beliefs of the
other person. Wimmer & Perner found that most four- to five-year-olds responded incorrectly by predicting that the person
would go to the actual location of the chocolates, while most six- to nine-year-olds correctly predicted the other person's false
belief (i.e. that the chocolates were still in the round tin).
The Wimmer & Perner experiment was concerned with first-order representations of mental states. First-order states involve the
ability to represent a mental state for example, to believe that another person has a belief or intention. This ability enables the
child to understand the behaviour of others, as demonstrated in their ability to predict behaviour in the experiment. An additional
complexity in this experiment was that the other person's beliefs could be different from those held by the child. The next level
is second-order representations, which are more important for understanding social interaction as they involve the concern
which people have about each others' mental states for example, what a child believes the other person believes about the child's
beliefs or intentions. The following example, based on an experiment by Perner & Wimmer (1985), illustrates second-order
states.

< previous page page_35 next page >

< previous page page_36 next page >
Page 36
Children were told stories in which two characters (John and Mary) were independently told that an object had been
unexpectedly transferred to a new location. For example: an ice-cream van, which both knew was originally in front of the post
office, had been moved to the town hall. Both John and Mary would know that the van was now at the town hall, but would not
know that the other person knew this. So John's second-order belief about Mary's belief would be incorrect i.e. John believes
that Mary believes that the van is at the post office (and similarly for Mary's belief about John's belief). It was not until the ages
of seven to eight years that children could reliably make the correct judgements of such second-order mental states.
The ability to represent mental states is important for social interaction for example, when distinguishing between a speaker's
intentions and what they say. First-order states are necessary to judge when a person is mistaken, that is, that they have a false
belief. However, second-order representations are required to distinguish between a lie and a joke. In the case of a lie the
speaker wants the listener to believe that what the speaker says is true, while the speaker believes it to be false. In other words,
a deceptive speaker wants the listener to think a false statement is true. In the case of a joke, on the other hand, the speaker does
not expect the listener to be misled, as both know that the statement is false and both know that the other knows this is the case.
Several studies have investigated children's deficiencies in the ability to infer the mental states of others (Baron-Cohen, Leslie &
Frith, 1985; Frith, 1989). This deficiency, which is demonstrated in particular by autistic children, has wide-ranging implications
for conversation, which depends on speaker and listener being able to take account of each other's mental states. Indeed, it
appears that, because of this deficiency, autistic children, while being able to produce syntactically correct utterances and even
to vary their language as a function of gross social attributes of their listeners, fail to use language meaningfully and
communicatively (Baron-Cohen, 1988).
Affective and Emotional Difficulties
In addition to being able to infer the beliefs and intentions of others, a socially competent conversationalist also has to be able to
interpret and understand their emotional states. There is no doubt that there exists a relationship between affect and the ability to
interact socially and communicate. For example: a child who is uncooperative, becomes upset easily, or is unable to fit in with
peers, may have experienced severe communication difficulties on earlier occasions and now adopts an avoidance strategy as a
way of coping with the problem.

< previous page page_36 next page >

< previous page page_37 next page >
Page 37
As yet there has been little research on the relationship between conversational disability and emotional and affective problems.
Investigators interested in this relationship have tackled the problem by looking at special populations (for a review see Prizant,
Audet, Burke, Hummel & Maher, 1990). On the one hand, they have looked at children with emotional-behavioural problems
and found out the incidence of communication problems within this population (Baltaxe & Simmons, 1988). On the other hand,
researchers have taken a population of language-impaired individuals and investigated the incidence of emotional-behavioural
difficulties these children might have (Baker & Cantwell, 1982; Cantwell & Baker, 1985). Both approaches have found an
association between communication disorders, emotional disorders and behavioural disorders in children and adolescents.
Baltaxe & Simmons (1988) found a high incidence of language problems in child psychiatric populations. The figures vary in
the literature but it appears that around 75% of children who present to psychiatric institutions also have communication and/or
learning problems (Cantwell & Baker, 1985). Similarly, Cantwell and Baker also found that of 600 children with speech and
language difficulties, 50% had a diagnosable affective and/or behavioural problem. This figure rose to 86% when children with
learning as well as communication difficulties were included in the sample. Interestingly, both approaches have also noted
qualitative differences. Baltaxe (1977), for example, noted that autistic German-speaking children had problems with the more
pragmatic aspects of language use rather than with speech intelligibility and/or grammar. Thus, she found that autistic children
confused the polite and familiar forms of address in German (Sie and Du), a confusion which may arise from a lack of
understanding of social roles. She also documented turn-taking difficulties and problems with taking the listener's point of view
or theory of mind in terms of marking new and old information in conversation. In the same vein, Baker & Cantwell (1982)
found that certain types of communication difficulties were associated more often with emotional-behavioural disorders than
others. In their sample, those children who were having communication problems as well as diagnosable emotional-behavioural
problems were children who presented with language difficulties as opposed to speech difficulties. Their analysis of these
distinctions was rather crude, but nonetheless they were able to conclude that language use, comprehension and processing
appeared to be the more prevalent problems for children with communication and affective disorders. Conversely though,
Camarata, Hughes & Ruhl (1988) found that their group of children with behavioural difficulties had mainly language
difficulties which could be described as syntactic rather than semantic in nature. Unfortunately, the authors used a tool which
looked mainly at syntax and which may have thus biased their results. The authors themselves conclude that a conversational
analysis would

< previous page page_37 next page >

< previous page page_38 next page >
Page 38
have been desirable. There is no doubt that further qualitative research is needed in this area.
From the above findings four major hypotheses have arisen:
1. Emotional-behavioural disorders cause communication difficulties.
2. Communication difficulties cause emotional-behavioural disorders.
3. Communication difficulties and emotional-behavioural disorders are related to each other because they are both related to
some factor that causes both of them.
4. Communication difficulties and emotional-behavioural disorders are interrelated because they are both aspects of the
developing individual. The child is in continuous interaction with the environment and the environment in turn affects the child.
For example, a child may develop a difficult temperament that affects mother-child interaction; this may result in a reduction of
mother-child interaction which in turn may lead to language delay.
Interestingly, there appears to be evidence which supports each of these possibilities. The reason for this would appear to be a
consequence of the type of emotional-behavioural disorder we are looking at, the heterogeneity of special populations, and the
type of communication difficulties we are concentrating on. Most researchers today would agree that a relationship exists but it
is early days in terms of defining the nature of this relationship. Further investigations are needed in this important area,
especially in light of the current interest in conversational disability.

Concluding Remarks
In this chapter some evidence has been presented which supports the view that conversational disability can have serious causes.
Linguistic impairment, difficulties with conversational rules, cognitive deficits, impaired socio-cognitive abilities, and emotional
factors can all give rise to conversational disability. Obviously it is important that anyone assessing a child's conversational
difficulties and devising a programme of remediation should carefully consider which factors are involved in a particular case.
As yet there are no clear procedures for assessing conversational difficulties in this way, although there are a variety of methods
available from which the practitioner can select (for some examples, see McTear & Conti-Ramsden, 1991, Chapter 8.)
It is important that conversational disability should be properly understood, as a child's difficulties with the use of language in
communicative

< previous page page_38 next page >

< previous page page_39 next page >
Page 39
contexts affects social relationships and social development. Studies of children's friendships indicate that there appear to be
differences in conversational skills between those children perceived by their peers as popular and those perceived as unpopular
(Gottman & Parker, 1986). For example: popular children use the strategy of trying to establish a common frame of reference
with other group members in terms of activities and goals, while unpopular children tend to use the strategy of drawing attention
to themselves. Popular children are also better at clarifying misunderstandings and resolving conflict, for example, by giving
justifications or using polite strategies. Similarly, it has been reported that learning-disabled children are less popular in peer
contexts because other children find difficulty in interacting with them (Donahue, 1987). For these reasons, it is important to
examine carefully the nature of children's conversational difficulties and to attempt to understand the underlying cause. This is
not just an academic exercise; it is a means of helping children whose social and educational development would otherwise be at
risk.

References
ATKINSON, J.M. and HERITAGE, J. (eds) 1984, Structures of Social Action. Cambridge: Cambridge University Press.
BAKER, L. and CANTWELL, D.P. 1982, Psychiatric disorder in children with different types of communication disorders.
Journal of Communication Disorders 15, 113-26.
BALTAXE, C.A.M. 1977, Pragmatic deficits in the language of autistic adolescents. Journal of Pediatric Psychology 2, 176-80.
BALTAXE, C.A.M. and SIMMONS, J.Q. 1988, Pragmatic deficits in emotionally disturbed children and adolescents. In R.
SCHIEFELBUSCH and L. LLOYD (eds) Language Perspectives (2nd edn). Austin, TX: Pro-Ed.
BARON-COHEN, S. 1988, Social and pragmatic deficits in autism: Cognitive or affective? Journal of Autism and
Developmental Disorders 18, 379-402.
BARON-COHEN, S., LESLIE, A.M. and FRITH, U. 1985, Does the autistic child have a `theory of mind'? Cognition 21, 37-
46.
BEATTIE, G.W. 1983, Talk: An Analysis of Speech and Nonverbal Behaviour in Conversation. Milton Keynes: Open
University Press.
BERNARD-OPITZ, V. 1982, Pragmatic analysis of the communicative behaviour of an autistic child. Journal of Speech and
Hearing Disorders 47, 99-109.
BISHOP, D. and ROSENBLOOM, L. 1987, Classification of childhood language disorders. In W. YULE and M. RUTTER
(eds) Language Development and Disorders. Oxford: MacKeith Press.
BRANSFORD, J.D. and McCARRELL, N.S. 1974, A sketch of a cognitive approach to comprehension: Some thoughts about
understanding what it means to comprehend. In W.B. WEIMER and D.S. PALERMO (eds) Cognition and the Symbolic
Processes. Hillsdale, NJ: Lawrence Erlbaum.

< previous page page_39 next page >

< previous page page_40 next page >
Page 40
BRINTON, B., FUJIKI, M., LOEB, D.F. and WINKLER, E. 1986, Development of conversational repair strategies in response
to requests for clarification. Journal of Speech and Hearing Research 29, 75-81.
BRINTON, B., FUJIKI, M. and SONNENBERG, E.A. 1988, Responses to requests for clarification by linguistically normal and
language-impaired children in conversation. Journal of Speech and Hearing Disorders 53, 383-91.
BRYAN, T. 1986, A review of studies on learning disabled children's communicative competence. In R.L. SCHIEFELBUSCH
(ed.) Language Competence: Assessment and Intervention. London: Taylor and Francis.
CAMARATA, S., HUGHES, C. and RUHL, K. 1988, Mild/moderate behaviorally disordered students: A population at risk for
language disorders. Language, Speech, and Hearing Services in Schools 19, 191-200.
CANTWELL, D.P. and BAKER, L. 1985, Interrelationship of communication, learning, and psychiatric disorders in children. In
C. SIMON (ed.) Communication Skills and Classroom Success: Vol. I, Assessment. San Diego, CA: College Hill Press.
CONTI-RAMSDEN, G. 1990, Maternal recasts and other contingent replies to language-impaired children. Journal of Speech
and Hearing Disorders 55, 262-74.
CRAIG, H.K. and EVANS, J.L. 1989, Turn exchange characteristics of SLI children's simultaneous and nonsimultaneous
speech. Journal of Speech and Hearing Disorders 54, 334-47.
DONAHUE, M. 1981, Requesting strategies of learning disabled children. Applied Psycholinguistics 2, 213-34.
1987, Interactions between linguistic and pragmatic development in learning-disabled children: Three views of the state of the
union. In S. ROSENBERG (ed.) Advances in Applied Psycholinguistics: Vol. 1, Disorders of First Language Development.
Cambridge: Cambridge University Press.
DONAHUE, M., PEARL, R. and BRYAN, T. 1980, Learning disabled children's conversational competence: Responses to
inadequate messages. Applied Psycholinguistics 1, 387-403.
ERVIN-TRIPP, S. and MITCHELL-KERNAN, C. (eds) 1977, Child Discourse. New York: Academic Press.
FAY, W.H. and SCHULER, A.L. 1980, Emerging Language in Autistic Children. London: Edward Arnold.
FEY, M.E. and LEONARD, L.B. 1983, Pragmatic skills of children with specific language impairment. In T.M. GALLAGHER
and C.A. PRUTTING (eds) Pragmatic Assessment and Intervention Issues in Language. San Diego, CA: College Hill Press.
FLETCHER, P. and GARMAN, M. 1988, Normal language development and language impairment: Syntax and beyond.
Clinical Linguistics 2, 97-113.
FRITH, U. 1989, A new look at language and communication in autism. British Journal of Disorders of Communication 24,
123-50.
GALLAGHER, T.M. and DARNTON, B. 1978, Conversational aspects of the speech of language disordered children: Revision
behaviours. Journal of Speech and Hearing Research 21, 118-35.
GERMAN, D. 1979, Word-finding skills in children with learning disabilities. Journal of Learning Disabilities, 12, 176-81.
GLEASON, J.B. and WEINTRAUB, S. 1976, The acquisition of routines in child language. Language in Society 5, 129-36.
GOTTMAN, J.M. and PARKER, J.G. 1986, Conversations of Friends: Speculations on Affective Development. Cambridge:
Cambridge University Press.

< previous page page_40 next page >

< previous page page_41 next page >
Page 41
GREENLEE, M. 1981, Learning to tell the forest from the trees: Unravelling discourse features of a psychotic child. First
Language 2, 83-102.
HURTIG, R., ENSRUD, S. and TOMBLIN, J.B. 1982, The communicative function of question production in autistic children.
Journal of Autism and Developmental Disorders 12, 57-69.
KING, M.F. 1989, Development of Pragmatic Skills in Children with Moderate Learning Difficulty. Unpublished M.Sc.
Dissertation in General and Applied Linguistics, University of Ulster at Jordanstown.
KYSELA, G.M., HOLDGRAFER, G., McCARTHY, C. and STEWART, T. 1990, Turntaking and pragmatic language skills of
developmentally delayed children: A research note. Journal of Communication Disorders 23, 135-49.
McCALEB, P. and PRIZANT, B.M. 1985, Encoding of new versus old information by autistic children. Journal of Speech and
Hearing Disorders 50, 230-40.
McLAUGHLIN, M.L. 1984, Conversation: How Talk is Organized. Beverly Hills: Sage.
McTEAR, M. 1985, Children's Conversation. Oxford: Blackwell.
1987, The Articulate Computer. Oxford: Blackwell.
1989, Semantic-pragmatic disability: A disorder of thought? In D.M. TOPPING, D.C. CROWELL and V.N. KOBAYASHI
(eds) Thinking across Cultures: The Third International Conference on Thinking. Hillsdale, NJ: Lawrence Erlbaum.
McTEAR, M. and CONTI-RAMSDEN, G. 1991, Pragmatic Disability in Children. London: Whurr Publishers Ltd.
MIRENDA, P.L., DONNELLAN, A. and YODER, D.E. 1983, Gaze behaviour: A new look at an old problem. Journal of
Autism and Developmental Disorders 13, 397-409.
NELSON, K. 1986, Event Knowledge. Hillsdale, NJ: Lawrence Erlbaum.
NELSON, K. and GRUENDEL, J.M. 1979, At morning it's lunchtime: A scriptal view of children's dialogues. Discourse
Processes 2, 73-94.
OCHS, E. and SCHIEFFELIN, B. (eds) 1979, Developmental Pragmatics. New York: Academic Press.
OLSON, D.R. and NICKERSON, N.G. 1979, Language development through the school years. In K.E. NELSON (ed.)
Children's Language, Vol. 1. New York: Gardner Press.
PERNER, J. and WIMMER, H. 1985, `John thinks that Mary thinks that . . .' Attribution of second-order beliefs by 5- to 10-
year-old children. Journal of Experimental Child Psychology 39, 437-71.
PRIZANT, B.M., AUDET, L.R., BURKE, G.M., HUMMEL, L.J. and MAHER, S.R. 1990, Communication disorders and
emotional/behavioral disorders in children and adolescents. Journal of Speech and Hearing Disorders 55, 179-92.
RAPIN, I. and ALLEN, D.A. 1987, Developmental dysphasia and autism in preschool children: Characteristics and subtypes.
Proceedings of the First International Symposium on Specific Speech and Language Disorders in Children, pp. 20-35. London:
Association For All Speech Impaired Children (AFASIC).
REILLY, R.G. (ed.) 1987, Communication Failure in Dialogue and Discourse. Amsterdam: North-Holland.
ROSINSKI-McCLENDON, M.K. and NEWHOFF, M. 1987, Conversational responsiveness and assertiveness in language-
impaired children. Language, Speech and Hearing Services in Schools 18, 53-62.
SCHANK, R.C. and ABELSON, R.F. 1977, Scripts, Plans, Goals, and Understanding. Hillsdale, NJ: Lawrence Erlbaum.
SHATZ, M. 1983, Communication. In J.H. FLAVELL and E.M. MARKMAN (eds) Handbook of Child Psychology: Vol. 3,
Cognitive Development. New York: Wiley.

< previous page page_41 next page >

< previous page page_42 next page >
Page 42
SHATZ, M., BERNSTEIN, D. and SHULMAN, M. 1980, The responses of language disordered children to indirect directives
in varying contexts. Applied Psycholinguistics 1, 295-306.
SILLIMAN, E. 1984, Interactional competencies in the instructional context: The role of teaching discourse in learning. In G.
WALLACH and K. BUTLER (eds) Language Learning Disabilities in School-age Children. Baltimore, MD: Williams and
Wilkins.
SMEDLEY, M. 1989, Semantic-pragmatic language disorder: A description with some practical suggestions for teachers. Child
Language Teaching and Therapy 5, 174-90.
SMITH, B.R. and LEINONEN, E. 1991, Clinical Pragmatics. London: Chapman and Hall.
STUBBS, M. 1986, Educational Linguistics. Oxford: Blackwell.
WARDHAUGH, R. 1985, How Conversation Works. Oxford: Blackwell.
WIIG, E. and SEMEL, E. 1984, Language Assessment and Intervention for the Learning Disabled. Columbus: Merrill.
WIMMER, H. and PERNER, J. 1983, Beliefs about beliefs: Representation and constraining function of wrong beliefs in young
children's understanding of deception. Cognition 13, 103-28.

< previous page page_42 next page >

< previous page page_43 next page >
Page 43

3
Teaching Vocabulary: The Case for a Semantic Curriculum
DAVID CRYSTAL
There have been so many articles and books written on the grammar and phonology of language handicap in recent years that it
seems almost unfashionable to raise the question of vocabulary. But it is essential that we begin to look at vocabulary with the
same objective, systematic intent that we have paid to these other areas. Indeed, it is even more important that we do so, bearing
in mind (a) that vocabulary carries most of the meaning of a sentence, and (b) that there is so much more of it. No one knows
how much vocabulary there is in a language, but it is quite obvious that there are thousands more words than there are sounds
and grammatical constructions. Sorting out the structural organisation underlying 40-odd phonemes is likely to be a simple
matter, compared with the problem of finding order within the chaos of a child's word-hoard, which may number thousands of
words.

The Size of the Problem

It is as well to appreciate that we are dealing in thousands, even in the field of language handicap. It is so easy to obtain the
opposite impression, especially if our only exposure to a child is an occasional lesson or a half-hour clinic session. In the
assessment sessions in language handicap I used to run at the University of Reading, mainly with language `delayed' children of
preschool or early school years, the number of different words produced in a half-hour period was often less than 100, but on
the occasions when parents kept a vocabulary diary for me, it was clear that this was only a fraction of the child's real
vocabulary size. (Parental vocabulary diaries, by the way, are invaluable supplements to professional observations. I find it is
not possible to rely on such diaries for information about pronunciation or grammar, but most parents have little trouble learning
how to keep a list of words used by

< previous page page_43 next page >

< previous page page_44 next page >
Page 44
the child during the day. The main thing they have to remember is to make a note about context, otherwise it is not clear
whether top, for example, refers to the summit or the toy.)
The only way in which we can discover the real vocabulary size of a child (or, for that matter, of an adult) is to record it and
count it and the awesome size of the task involved is the main reason why this has not been done. But these days, we are
beginning to accumulate some data on the point, using radio-microphone techniques. And the general impression from this early
material is that traditional estimates of vocabulary size in children have been hopeless underestimates. Wagner's study, for
example, reported in 1985 (and summarised in Crystal, 1986), set out to determine how many words children said in a day. He
and his students used mini-microphone transmitters to capture everything which was said by a range of children of different ages
and social backgrounds; when this was transcribed, the totals discovered were those displayed in Table 3.1. There are two
columns in this table: the first gives the total number of words (word tokens that is, every instance of the same word was
counted separately); the second gives the number of different words (word types). It seems that German children speak an
average of between 20,000 and 30,000 words a day, from their third year, and have a daily active vocabulary of around 3,000
different words. I should be surprised if the total for English children were very different, but I do not know whether this is so,
for no one has yet carried out a corresponding project in English.
TABLE 3.1 Words used in one day's recording by 10
children
Total number Number of
of different words (i.e.
words (i.e. types)
tokens)
Katrin, 1 year 5 months13,800 1,860
Nicole, 1 year 8 11,700 not available
months
Andreas, 2 years 1 20,200 2,210
month
Carsten, 3 years 6 37,700 4,790
months
Gabi, 5 years 4 months 30,600 2,490
Frederik, 8 years 7 24,700 3,960
months
Roman, 9 years 2 24,400 3,630
months
Markus, 11 years 4 37,200 5,020
months
Christiane, 12 years 2 22,600 3,580
months
Axel, 14 years 10 22,900 3,040
months

These figures do not tell us about the total vocabulary size of any of these children, in terms of word types, as they were based
only on one day's recording. Doubtless, if another recording day were to be arranged near Christmas, on holiday, out shopping,
on a birthday . . . the vocabulary

< previous page page_44 next page >

< previous page page_45 next page >
Page 45
collected would have continued to grow. One would soon reach 5,000 items for a preschool child. But even lacking this further
perspective, it is quite evident that our estimates of vocabulary size for non-handicapped children are in need of revision
upwards. For example, the average productive vocabulary size for an 18-month child is said to be 50 words, and to reach about
200 words by age two (see the review of the literature in Garman, 1979, and Griffiths, 1986). (The estimates for comprehension
are less certain, given the greater difficulty of ascertaining that a young child has really understood a word, but are generally
said to be greater than production, at 18 months, by a factor of 4 or 5. This ratio, incidentally, of 1 word in production to 4-5
words in comprehension, is a useful one to bear in mind when establishing an early lexicon in a handicapped child.) But what do
we then say about Katrin, in Table 3.1, who produced just over 800 word types in just under four hours (the total in Table 3.1 is
an extrapolation to a standardised 12-hour sample)? She would not seem to be particularly exceptional, according to Wagner,
but these figures bear no clear relationship to the previously cited acquisition norms.
All of this leads to two conclusions. First, it confirms the view that vocabulary is `the big one', as far as language assessment
and intervention is concerned. And second, it indicates that we are a long way from having reliable frequency norms about
vocabulary. So where do we go from here?

Frequency is Never Enough.

The first thing is to drop the focus on frequency, and to replace it by a notion of the quality of vocabulary. But dropping the
frequency focus is not easy, as it has been around for a very long time. Most teachers and therapists, I think, have been brought
up to believe that size of vocabulary is the critical factor, and that their job is to increase the number of words. Many tests and
teaching procedures are restricted to this point, and judgements tend to be phrased in quantitative terms. `He has a vocabulary of
about 20 words . . . 100 words . . .' Such observations are commonplace, but their only value is in the context of screening.
When it comes to the question of what words to teach next, numbers alone are not enough. To have established that a child has
50 words does not answer the question `Which word will we teach as number 51?' And to establish that two children both have
vocabularies of 50 words does not mean that they are at the same level, in terms of semantic development. One child may be
using the words in a simple labelling way, each word referring to one object; whereas the other may be using (the same) words
in a creative way, each word referring to a range of objects and ideas. For the first child, cold might mean only a certain kind of
weather; for the second, it might

< previous page page_45 next page >

< previous page page_46 next page >
Page 46
be used to refer to weather, water, ice-cream, and even to a grim face. It is the senses which are important in vocabulary
learning, not the words alone.
The first step in determining the quality of a child's vocabulary is to see how far the words group themselves into sets, or fields,
and begin to define each other. The notion of a `semantic field' is very important. The words apple, banana, orange, etc. belong
to the semantic field of `fruit'. Brother, uncle, father, etc. belong to `family'. Red, green, blue, etc. belong to `colour'. These are
some of the clear-cut cases. Less obviously, bag, basket, cup belong to `containers'. Arrive, come, approach belong to `motion
(in a certain direction)'. It is not always easy to decide which field a word belongs to, and you have to look carefully at the
context. A hospital, for example, is in one sense a building (and thus belongs to the field of `buildings', along with church,
factory and house); on the other hand, it is a place where you get treatment (and thus belongs to the field of `health', along with
ward, consultant and nurse). The context is usually clear enough, when a child uses the word; where it is not clear, of course, it
may not be possible to reach a decision. (A fuller account of semantic fields is given in Crystal, 1981, and a procedure which
can plot children's developing use of fields is outlined in Crystal, 1982.)
The important thing to appreciate about vocabulary learning is that, after the very initial stages, words are not learned `one at a
time', as the frequency model suggests. A new word has to find its place in relation to the other words already acquired. As its
meaning comes to be appreciated, so the meaning of the other words is made to alter. For example, an 18-month-old may have
the word dog in his vocabulary, but he uses it to refer to all animals (the phenomenon of overextension). However, note that as
soon as he picks up the word cat, the meaning of dog changes. Dog and cat now divide up the semantic field of animals. If cow
comes along next, the field is divided yet again. A great deal of debate has taken place as to the way in which this is done which
features of meaning the child pays particular attention to, the order in which these features are acquired, and how long it takes to
complete the process, so that words like dog and cat end up with their adult meaning (see Clark & Clark, 1977, Chapter 13).
With some fields, the process of sorting out all the semantic features which distinguish words can take several years even for a
field such as `family' (where even at age seven or eight children may still be unclear about the reciprocal meaning of some
words if one has a brother, one is a brother or not know the meaning of such words as cousin). In fact, `family' is a good
example of a field which continues to cause problems even in adult life. Evidence? How many readers are totally clear about the
difference between second cousin and cousin once removed?

< previous page page_46 next page >

< previous page page_47 next page >
Page 47
What is going on, when a semantic field is being learned? All kinds of cognitive and social learning, of course, but for the
present my focus is on the purely linguistic issue involved which is that words are being used to help learn other words. Very
little vocabulary learning is carried out by having the word refer directly to the object in the real world. At the very beginning,
this process (of ostensive naming) takes place, of course, as parents point to objects and label them. But this stage does not take
very long. As interests widen, and mobility increases, it becomes increasingly impracticable to define by pointing. A child who
hears the word giraffe and asks `What's a `raffe?' is told something like `It's an animal with a very long neck'. It would be
somewhat abnormal to reply, `I'll show you when we next see one'! And of course for the vast majority of notions, there is
nothing obvious to point at anyway. How would you point at think, or colour, or angry? Or, for that matter, red (pointing to a
red object does not necessarily help, as, from the child's point of view, you might be pointing at its shape, or its size, or even
suggesting that the object itself is called a `red').
Words define words. Giraffe, in the above example, is defined with reference to animal and long neck (one of its distinguishing
semantic features). And so it is, with the whole vocabulary. A dictionary is a book in which all the words of the language are
defined by using the words in the language. Occasionally, it proves easier to see an object than to define it (as with the classic
conundrum of how to describe the word spiral without using your hands), and dictionaries then rely on supplementary pictures.
But in the vast majority of cases, the words are enough. And it is the same when we are asked to define words ourselves, in
everyday life. We give a definition, as best we can, using the same procedures as are illustrated in dictionaries. We rely on the
sense-relations which exist between words the relationships of meaning which we sense intuitively to be present, so that we
`know' that two words `have the same meaning' (are synonyms), `have opposite meanings' (are antonyms), `have one meaning
included in the other' (are hyponyms), or `can't be both things at once' (are incompatible). Here are examples of each category,
illustrated from parent reactions to child questions:
synonymy: `automobile means car'
antonymy: `ugly means not very pretty'
hyponymy: `nectarine is a kind of fruit'
incompatibility: `that's not blue, that's purple'.
These four kinds of relationship between words are the essential ones for developing a qualitative approach to vocabulary
teaching and learning. (For other categories, and for further examples of the above, see Crystal, 1981.) It is these relationships
which help to define the structure of the semantic fields of

< previous page page_47 next page >

< previous page page_48 next page >
Page 48
the language, and they provide the main means whereby teachers and therapists can extend a child's vocabulary in a structured
and systematic way.

Extending Vocabulary
An important preliminary point to appreciate about vocabulary teaching is that there is a difference between this area and that of
phonology and grammar, arising out of our lack of empirical information about acquisition: there is no obvious first place to
start. At least with phonology, it is possible to identify certain basic sound contrasts which children seem to make use of at an
initial developmental stage; and with grammar, there is a small range of basic sentence types which seem to be fundamental. But
with vocabulary, there is as yet little sign of a universal pattern of acquisition, at least as regards the learning of individual words
(and even the order of acquisition of semantic fields shows considerable variation across children). There is no such thing as a
universal `first word'. And even if we make a comparison of a large number of words used by several children in the early stages
of development, it is remarkable how little overlap there is. Table 3.2 gives the first 50 words of three children: there are only
eight in common (adapted from Stoel-Gammon & Cooper, 1984).
In a way, such differences are not surprising. If a child lives by the sea, one would expect sea-vocabulary to appear early on; if
on a farm, farm-vocabulary; and so on. There are undoubtedly certain parallels in the kind of lexical learning that takes place,
which research into semantic acquisition is beginning to uncover (see the relevant chapters of Fletcher & Garman, 1986), but we
are a long way from here to identifying factors which underlie the selection of individual words. So how is a teacher or therapist
to proceed when, faced with a child with little or no productive vocabulary, the question has to be addressed, `What words to
teach first?' which (in the light of the above) means `What semantic field to teach first?' Without normative data to rely upon, is
the only alternative guesswork?
At the very outset, the answer must be yes or, at least, informed guesswork. Obviously, one wants to choose a field which is
motivating to the child, and this means finding out about the child's interests, using whatever techniques one can. A parental
account of the child's home situation and daily routine is essential. Are there animals at home? What are the favourite toys (toys
are a representation of the real world, and can display, in microcosm, all semantic fields)? What members of the family exist?
Much of this information is routinely available, as cognitive/social data, in case history reports, of course; but for linguistic
purposes it may need to be interpreted anew. It is one

< previous page page_48 next page >

< previous page page_49 next page >
Page 49
TABLE 3.2 The first 50 words of three children
Daniel Sarah Will
1light 1baby 1uh-oh
2uh-oh 2mommy 2alldone
3what's that 3doggie 3lightlight
4wow 4juice 4down
5banana 5bye-bye 5shoes
6kitty 6daddy 6baby
7baby 7milk 7don't throw
8moo 8cracker 8moo
9quack (quack) 9done 9bite
10cookie 10ball 10three
11nice 11shoe 11hi
12rock [noun] 12teddy 12cheese
13clock 13book 13up
14sock 14kitty 14quack-quack
15woof-woof 15hi 15oink-oink
16daddy 16Alex 16coat
17bubble 17no (no) 17beep-beep
18hi 18door 18keys
19shoe 19dolly 19cycle
20up 20what's that 20mama
21bye-bye 21cheese 21daddy
22bottle 22oh wow 22[siren sound]
23no 23oh 23grrr
24rocky [verb] 24button 24more
25eye 25eye 25off
26nose 26apple 26tick tock
27fire 27nose 27ball
28hot 28bird 28go
29yogurt 29alldone 29bump
30pee-pee 30orange 30pop-pop [fire]
31juice 31bottle 31out
32ball 32coat 32heehaw
33whack (whack) 33hot 33eat
34frog 34bib 34neigh-neigh
35hello 35hat 35meow
36yuk 36more 36sit
37apple 37ear 37woof-woof
38Big Bird 38night-night 38bah-bah
39walk 39paper 39hoo-hoo [owl]
40Ernie 40toast 40bee
41horse 41O'Toole 41tree
42more 42bath 42mimi [ferry]
43mommy 43down 43s: [snake]
44bunny 44duck 44ooh-ooh [monkey]
45my 45leaf 45yack-yack [people talking]
46nut 46cookie 46hohoho [Santa]
47orange 47lake 47bye-bye
48block 48car 48doll
49night-night 49rock 49kite
50milk 50box 50Muriel

< previous page page_49 next page >

< previous page page_50 next page >
Page 50
thing knowing that there is a younger brother in the house; it is quite another knowing what he is called by the family. A
semantic case history checks on the semantic fields and routine lexical items in use in the child's environment, and draws
attention to any which seem to be attracting his interest (cf. the communicative history form' in the appendix to Crystal, Fletcher
& Garman, 1976).
Sometimes it makes more sense to select a semantic field for teaching on the basis of what demands are being made of the child
in school or clinic especially important when the child stays in a single place for long periods, such as at a residential school.
The routine of the child's day, or the interaction with the peer group, will readily indicate the relevance of certain fields food,
games, clothing, furniture, and so on. But there is one important difference between the professional and the home environment,
as far as vocabulary is concerned: there is usually a much greater gap between the vocabulary of comprehension and that of
production: that is, between the words the child uses and those used by the adults in the environment. Teachers routinely use
many words, as part of the organising `language' of intervention, which the children themselves would not be expected to say
point, show, tell, ask, happen, find, and many more. These are among the most frequent words used in the language-teaching
environment, but they would not normally be considered a priority in vocabulary teaching (indeed, it is usually assumed that
these words are `caught', not `taught', though this is a problem in its own right).
Let us assume, then, on pragmatic grounds, that a semantic field has been identified as a focus for teaching. I will use `body
parts' for illustration, though everything which follows applies equally to other fields. The assessment part of the task is to
establish whether the child knows the words for the various parts of the body, and whether he knows how to use any of them.
The teaching part of the task is to teach these words, in both comprehension and production. Let us assume that the assessment
is clear-cut: the child in question has no knowledge of body parts at all, and has never been heard to name one. How, then, does
the teaching begin?
What are the Relevant Words?
The first thing is to be clear in one's own mind what exactly the semantic field of `body parts' consists of. How many words
does it contain, and how are these organised? This question is in fact more complex than it seems. For how does one draw up a
complete list of relevant body parts? The question of relevance is critical. A really complete list of body parts would be the
equivalent of Gray's Anatomy. Obviously, some notion of `basic' body parts is

< previous page page_50 next page >

< previous page page_51 next page >
Page 51
involved, `appropriate to children of such-and-such an age'. But that kind of information is not readily available. Who is to say
whether wrist, for instance, should be considered appropriate for a three-year-old, or four-year-old, or five-year-old . . . (or for
language-handicapped children at equivalent language ages)? There is a lot of guesswork here, and a decision will often have to
be made on pragmatic grounds that is, you may not know whether it is normal for five-year-olds to have the word wrist, but you
decide nonetheless that it is going to be very useful for your particular child to know this word (perhaps because he owns a
watch).
It makes a very interesting exercise for the members of a team to write down, without consulting each other, which body-part
words they think (say) a five-year-old should know. They then compare notes afterwards, and see how far they agree. One day,
there will be vast statistical surveys of child vocabulary to provide some answers. For the moment, there are only our intuitions
and, as you will discover, if you try this exercise, these often differ.
A convenient way of getting a basic list of potentially relevant words together is to use a thesaurus or a dictionary, or one of the
new kinds of reference book which brings these two notions together. If you look up various body parts in Roget's Thesaurus,
for example, you will quickly accumulate a wide range of relevant vocabulary; but there is a problem here, for the words are
given no definitions. Conversely, if you look up a body part in a dictionary, you will be given a definition, but no sets of words.
However, in the Longman Lexicon a new venture in lexicography (McArthur, 1981) you are given both, and this makes the
book very helpful for people engaged in teaching vocabulary. You look up a word in the index at the back, and this refers you to
a semantic field, in which all the main words belonging to that field are listed in alphabetical order, with definitions and
examples given. An example from the body-parts field is given in Figure 3.1. (Note that the field includes some verbs as well as
nouns.)
I find the Lexicon an essential pedagogical tool, when it comes to selecting and presenting vocabulary. But any good dictionary
will do. The point is that the dictionary should be seen as an essential piece of `clinical' or `remedial' equipment, for people
dealing with language handicap as essential as a stethoscope to a doctor. And to be `good', I mean it needs to be up-to-date, to
ensure that new words and phrases are included and new developments in dictionary methods incorporated. It needs to contain at
least 15,000 entries, to ensure that it gives a basic coverage (but for more advanced work it should approach 50,000). And a
bonus for working with language handicap (available with some of the Longman series) is that there should be a restricted
defining vocabulary that is, the words used in the definitions should all be

< previous page page_51 next page >

< previous page page_52 next page >
Page 52

Figure 3.1
A page from the Longman Lexicon

< previous page page_52 next page >

< previous page page_53 next page >
Page 53
taken from a basic set (of 2,000 words, in the Longman Dictionary of Contemporary English). There is little point in referring to
a dictionary for a word meaning, if the definition is so complex that you need a dictionary to understand it (as in old definitions
of dog which began `carnivorous quadruped . . .').
One thing you quickly discover, when you start putting together a semantic field, is that the words begin to group themselves
together, into `subfields'. This can be seen from Figure 3.1, where we see such headings as `the head', `the face'. Indeed, the
field of `body parts' can be analysed into a dozen or more subfields including such less obvious fields as `outside features' (hair,
beard, skin, nails, etc.) and `inside features' (blood, bones, brain, etc.). And there are the problem words too, as one soon
discovers when working with children such as the scrap who pointed out that tongue could be both! It quickly becomes apparent
that, far from being a very narrow area, the notion of `body parts' is in fact quite large, and needs to be broken down into
subfields before one can answer the question, `Which words should I teach first?'
But even within a subfield, there is a considerable lexical choice 14 words in the Lexicon's listing for `face', for instance. One
cannot teach all 14 words at once. How is a selection to be made? Here, certain guidelines can be suggested. First of all, it is
important to choose words which have some motivation for the child where there is some reasonable chance that the word will
relate to aspects of the child's life. The risk is to teach `body-part' words in an arbitrary way: to assume, for example, that
because eyes, nose and mouth are the `obvious' features, they must be the ones to start with. The question, however, should
always be: `Why should a child want to talk about . . . eyes?' (for example). And indeed, it is not easy to see why one should
find it interesting to go through the parts of a face, labelling them. If one could ask these children which parts of the face are the
most interesting bits, what would they say, I wonder? Having done this once or twice, I can report that the answers are not
always what one would expect beard, glasses, spots . . .
The point to be appreciated is that the words for the different facial parts are not equivalent in their pragmatic force. Nose
becomes particularly important when the child has a cold, and has to keep wiping it. Teeth are important at certain times of the
day, when they have to be brushed. Eyes are important when something gets into them, or when people point out the different
colours. And so on. The value of imaginative story-telling, in this connection, should be obvious: this technique enables us to
`set up' a situation in which the focus is entirely on the vocabulary we wish to teach a land where one-eyed and two-eyed
people live, or a child who cannot open his

< previous page page_53 next page >

< previous page page_54 next page >
Page 54
mouth. Animals provide a useful real-life way of focusing on body parts, because of their anatomical differences finding where
the eyes are on a hamster, a chicken, or a whale. But here, too, the selection of features needs to be motivating. There would
seem to be little point in comparing a hamster and a chicken, for instance, unless they can be shown to inhabit some real or
imaginary world together. It is often easier to set up situations in which all the animals are pets, or farm animals.
Moving on to Features
Once we have selected the semantic subfield, and the specific body part to focus on, and the motivating context to use, are we
now ready to teach the word, in comprehension and production? Not yet. So far we are in the position of being able only to
name an object, in an ostensive way pointing to a person, a hamster, or whatever, and saying eyes, That's an eye, What is it? and
the like. As already mentioned, this does not lead very far. To learn a word is to integrate it into a growing vocabulary, and the
child needs to be taught how to do this, which means deciding on which other words to relate it to. Words should never be
taught in isolation from each other. On the contrary, each word should be presented in a linguistic context which gives part of its
definition.
This is often done, quite automatically, by the teacher or therapist, as the following therapist/child dialogue shows:
T: What are those, do you know? [pointing to own eyes]
P: ____
T: We can see with them, can't we.
P: ____
T: I've got two, and you've got two. [pointing to child's eyes] They're our eyes.
P: eye.
T: That's right, we see with our eyes, don't we . . .
Some of the words needed for the definition of eyes are already being used in the interaction: see and two. (Also, it should be
noted, some of the relevant grammar is coming in as well, such as see with.)
But how does one know to which other words the item should be related? If you were teaching eye, how would you decide how
many relevant words there are, as part of the definition? Here again the dictionary is essential, as it provides most of the data
one needs to answer these questions. The Lexicon definition of eye, for example, is given in Figure 3.1: `the organ of sight, of

< previous page page_54 next page >

< previous page page_55 next page >
Page 55
which there are two at the front of the human head'. Not very elegantly phrased (that is often the way, when you are writing
with a restricted defining vocabulary the words are simpler, but the grammar is awkward), but it gives the essential information:
eye is defined with reference to organ, see, two, front and head. (If you looked the word up in a different dictionary, the
definition would be different in certain respects, but that is not the point: you could still use that definition to pull out the
essential defining features.) Now, armed with this information, it is possible to see the kind of teaching problem you face, if you
decide to teach the linguistic meaning of eye. You will presumably dispense with organ, at this level, but what about the other
words? What is the relevance of two, for instance? Very relevant indeed, it would seem, judging by the way some mothers
`count' their baby's eyes while washing them (one eye, two eyes . . .), or draw a contrast between the `twoness' of the eyes and
ears, and the `oneness' of the nose and mouth. And so one might continue, evaluating the relevance of each of the other words,
before deciding how much to `tell' the child. After all, T in the above dialogue might have continued for a couple more turns,
saying such things as `They're at the front of our heads'. And, in a more explicitly structured teaching exercise, it would be
possible to focus on these features in turn (looking at dolls to see whether they have eyes at the back or at the front, whether
they have one or two, whether they are open or closed (aspects of seeing), and so on).
Analysing a word's definition, to extract its main semantic features, is a prerequisite for systematic, structured work on
vocabulary. Without it, one simply does not know what problems lie ahead. It is very easy to choose a word to teach, and then
get into difficulties as one realises that the word is unexpectedly difficult to explain, as it uses words which are well beyond the
level of the child. Here are some everyday words, with their Lexicon definitions, which illustrate this point:
soil: the material found generally on top of rock, in which plants can grow. (The first part of the definition is not so important,
but the reliance on the notions of `plant' and `grow' is critical.)
factory: a building or group of buildings where goods are made, especially in great quantities, by machines.
Factory is a good example where an unthinking definition can cause trouble. In answer to the question `What's a factory?' one
teacher said `A place where you make things', and this led the child to think that his classroom was a factory. The lack of
reference to `machines' was critical, in this case.
This example also illustrates the way in which it might take a while to lead up to the whole definition. It would be inappropriate
to `flood' the child with the entirety of this definition, at an early stage of learning. It might even be

< previous page page_55 next page >

< previous page page_56 next page >
Page 56
enough to gloss the word by `It's a kind of building', at the outset in other words, to teach an overextension. Then, as the child's
abilities develop, one might add, by degrees, `a building where you make things', `a building where you make things with
machines', and finally to something like `a building where you make lots of things with machines'. The whole definition might
take several years to learn but recall that this is a perfectly normal process. (For developmental stages in the learning of
definitions, see Litowitz, 1977.)

A Semantic Curriculum
Because the learning of a word's meaning is not a once-and-for-all event, it seems essential to plan vocabulary teaching in an
appropriately long-term manner. One needs more than a syllabus of words to be taught week by week. One needs a syllabus of
senses, of definitions, of features. One needs to devise ways of keeping track of a word, as it is introduced into a child's
vocabulary, so that one can look back and see whether there are features of the word's meaning that the child has never been
taught. And one needs to grade the words within their semantic fields, to determine which are likely to be the easiest to teach
(i.e. make the fewest assumptions about what other words have already been learned), and to anticipate the problems which will
arise. Exceptional cases need to be taken into account (specifically, the problem of idioms, which cannot be analysed in a
straightforward semantic way). And strategies need to be devised to ensure that an analysis, once taught, is reinforced by others
involved with the child. It could be confusing for one adult to gloss the word factory as `building', and another as `shed', and
another as `place', during the period when the child is first being introduced to the word; but it is likely that, in the absence of
any standardising guidelines, a great deal of this variation takes place.
Can all this be done? It never has, as far as I know. Most vocabulary teaching proceeds on a serendipity principle, in which
children are assumed to be able to pick up most of a word's meaning from their informal contacts. This assumption works well
enough with normal children (where, I suspect, the roles of television and reading are critical in providing an increased range of
contexts within which to develop their sense of word meaning), but I do not think it is at all satisfactory as a way of proceeding
with language-handicapped children. All too often, one encounters these children, especially from around age six or seven,
whose vocabulary is superficially large but (on investigation) inadequate, because words are only partially understood,
relationships between words are vague, and there is no ability to define (a skill which comes to be increasingly relied upon as a
child moves up through junior

< previous page page_56 next page >

< previous page page_57 next page >
Page 57
school) in short, their vocabulary has no structured foundation. The aim of a semantic curriculum would be to make good this
kind of deficiency.
Can it be done? Already I hear the old argument in the wings: but won't it be very time-consuming? The answer is: yes and no.
The first occasion one works through a whole semantic subfield in the way outlined above, it takes a great deal of time. But the
crucial point to appreciate is that this only needs to be done once. After a set of words has been analysed and organised in a
semantically principled way, the donkey-work is done, and the information can then be used with any child in any setting.
Moreover, it is not necessary to `take on' the whole of a subfield, when commencing to work systematically with vocabulary.
All that is required, in day-to-day practice, is a few minutes to take the small number of words one has decided to teach, to look
them up, and to identify their main features. This exercise alone can be extremely illuminating, and can lead to immediate
suggestions for a systematic or structured approach which can readily save time.
If, then, the people who do this were to take a few more minutes to transfer their analysis on to cards, or on to a computer, for
other people to refer to, it would not take long before a core vocabulary of several hundred words were publicly accessible. And
if the procedure could be standardised in some way, the permanent gain would be invaluable. But this is to think ahead. At
present, what is necessary is for people to experiment with various procedures in different semantic fields to try out techniques
for bringing vocabulary more closely under the microscope, in relation to the daily demands of teaching. The dearth of
published material in this area is such that a report on the analysis and teaching of even a single lexical item would be of general
interest. I very much hope, therefore, that small-scale projects will be initiated by teachers and therapists interested in ways of
improving lexical intervention, and that some centres may proceed with larger projects. Vocabulary is the last large mountain to
be scaled, within the domain of language structure, and it is about time we set up some base camps.
Acknowledgement
This paper first appeared in Child Language Teaching and Therapy 3, 1 (1987), and is reprinted by permission.

References.
CLARK, H.H. and CLARK, E.V. 1977, Psychology and Language. New York: Harcourt, Brace, Jovanovich.

< previous page page_57 next page >

< previous page page_58 next page >
Page 58
CRYSTAL, D. 1981, Clinical Linguistics. Vienna: Springer; London: Edward Arnold (1987).
1982, Profiling Linguistic Disability. London: Edward Arnold.
1986, Listen to Your Child. Harmondsworth: Penguin.
CRYSTAL, D., FLETCHER, P. and GARMAN, M. 1976, The Grammatical Analysis of Language Disability. London: Edward
Arnold.
FLETCHER, P. and GARMAN, M. (eds) 1986, Language Acquisition, 2nd edn. Cambridge: Cambridge University Press.
GARMAN, M. 1979, Early grammatical development. In P. FLETCHER and M. GARMAN (eds) Language Acquisition.
Cambridge: Cambridge University Press.
GRIFFITHS, P. 1986, Early vocabulary. In P. FLETCHER and M. GARMAN (eds) Language Acquisition, 2nd edn.
Cambridge: Cambridge University Press.
LITOWITZ, B. 1977, Learning to make definitions. Journal of Child Language 4, 289-304.
McARTHUR, T. 1981, Longman Lexicon of Contemporary English. London: Longman.
STOEL-GAMMON, C. and COOPER, J.A. 1984, Patterns of early lexical and phonological development. Journal of Child
Language 11, 247-71.
WAGNER, K.R. 1985, How much do children say in a day? Journal of Child Language 12, 475-87.

< previous page page_58 next page >

< previous page page_59 next page >
Page 59

4
How Can We Study Pragmatic Disabilities?
GINA CONTI-RAMSDEN
Introduction
Lisa was an enthusiastic support teacher spending her first term teaching seven-year-old children with special needs. She had
been told that the children she was responsible for were a `hodge podge': some had learning difficulties, some had language
problems, some had attention problems and so on and so forth. Lisa decided to make up her own mind about the children's
problems so she drew up a schedule for observing their behaviour and performance carefully. She decided that she would like
the children to be in a relaxed and comfortable atmosphere when working with her so before deciding on the composition of the
groups she would like to form, she went around asking the children who they were friendly with and who they liked to play with
in the playground. Most children named or pointed to at least one friend. Smiles were exchanged, elbows nudged but one child
called Robbie sat there, looking at her, as if trying to understand what was being asked of him. She decided not to put Robbie
on the spot but instead find out more about him later from his classmates, fellow teachers and parents. `He likes to play on his
own, Miss' said one child; `He doesn't say much', said another; `Sometimes we ask him to play a game but he doesn't always
follow the rules', said yet another child. Fellow class teachers described him as `odd'; `He is a serious little boy. He lacks a sense
of humour. He hardly ever laughs in the right places in a story.' His parents said Robbie played mostly on his own and had no
friends in the neighbourhood. `It isn't because people can't understand him', said his mother; `He speaks very clearly but not
very much. We're usually the ones that ask him questions, he rarely starts a conversation'. Lisa asked them about understanding:
does he understand what is said to him? Can he follow commands? `Oh, yes he understands', they said, and his father then
continued: `But he can get a bit muddled up when he has to think for himself, you know, when we talk there are some things
that are not said but are sort of

< previous page page_59 next page >

< previous page page_60 next page >
Page 60
assumed he doesn't always seem to be able to do that, we sometimes have to sit and go over something step by step with him.'
Lisa must now decide the nature of this child's difficulties.

What are Pragmatic Disabilities?

How we say something, to whom we say it and what we say are all important aspects of communication. A conversation
between two people may take different forms. For example, they may have a discussion, tell a story or describe a memorable
event. Each different type of interaction has its own set of rules which competent speakers will follow. In addition, both speaker
and listener need to know which communicative act is appropriate; for example, we need to know when to disagree, qualify,
change the topic and so on. We also must be able to learn to participate at the right moment in the conversation as well as to
keep the conversation running smoothly.
The term `pragmatics' is often used to encompass those aspects of language use which are subject to systematic variation
according to the social context. Children, like Robbie, with problems in recognising and satisfying the social rules of language
are usually described as having pragmatic disabilities. These children often have difficulty in school, in making friends and
coping with everyday social interaction. Nonetheless, their difficulties are not always obvious. There is an urgent need for more
research into the nature of their problems (Beveridge & Conti-Ramsden, 1987).
If we wish to find out more about pragmatic disability in children, a good place to start is to find some children and study their
conversation. Following are some excerpts from pragmatically disabled children's conversations. What sorts of questions can we
ask from these and other sources of data?
Example 1
[Speech therapist and eight-year-old boy sitting talking]
T: What's wrong with Miss X? [child's teacher] Why isn't she here?
C: She's in bed.
T: She's sick?
C: Yes, she's in bed
T: Oh dear! How do you think she is feeling if she's in bed?
C: Ehm. . . . ehm. . . . Alright
T: You think she's feeling alright?
C: No. . . . Ehm. . . . maybe.
T: Maybe she's feeling alright. But if she's in bed and she is a bit sick, how do you think she feels, sad or happy?

< previous page page_60 next page >

< previous page page_61 next page >
Page 61
C: Sad.
T: I would say so, I would say sad. A bit sad and a bit tired. Maybe we can send her a card.
C: Send her a card?
Example 2
[Teacher and nine-year-old looking at a picture of a baby with a dirty nappy]
T: Why does a baby have to wear a nappy? Do you know?
C: Yeah.
T: Why?
C: He can't go in the toilet.
T: He can't go in the toilet. So what happens to his nappies, do they get clean or dirty?
C: Clean.
T: Are they clean? [pointing to picture]
C: Yes.
T: What does the baby do in the nappy?
C: Toilet.
T: That's right. The baby's toilet is pretty much the nappy, isn't it?
C: Yeah.
T: What would happen if the baby didn't have a nappy?
C: Don't know.
We should ask ourselves if these conversational problems result from underlying language comprehension difficulties or from
real problems with language use.
There are three main ways in which researchers have tried to investigate pragmatic disability: case studies, group design studies
and group comparison studies. Each of these methods have advantages and disadvantages, as we will discuss.

Case Studies
Case studies provide a means of investigating in depth children who have been identified as having a pragmatic disability,
irrespective of whether they are otherwise language-impaired or exhibit difficulties in other areas. Examples of case studies
include the work of Blank, Gessner & Esposito (1979), McTear (1985), and Conti-Ramsden & Gunn (1986). For the purposes
of this chapter we will take the work of Conti-Ramsden & Gunn (1986) to illustrate the case study approach.
Conti-Ramsden & Gunn (1986) observed a three-year-old boy named Tony for a period of three and a half years in an attempt
to document changes

< previous page page_61 next page >

< previous page page_62 next page >
Page 62
in the child's profile across time. In the area of pragmatic conversational development, Tony began to respond in conversation
well before he began to initiate conversations spontaneously. In addition, Tony began using language spontaneously mainly to
label and describe and it was only when he was nearing six years of age that he used language to request information and/or
action. Such late emergence of requests, in the presence of enough expressive language to enable their formulation, represents a
deviant pattern of development which affected the quality and quantity of Tony's social interaction with peers and adults. The
question then arises as to why this was the case. The above researchers also looked at related areas of development in order to
explore potential relationships between conversational problems and other non-linguistic and linguistic difficulties.
In the area of verbal comprehension, Tony's profile was uneven. At any particular point in time, Tony could achieve later
milestones, e.g. choose an object by shape and colour, while unable to succeed in simple interactional acts, e.g. look attentively
or smile appropriately. As Tony grew older he became better at scoring in standardised tests of verbal comprehension and the
tests thus masked Tony's difficulties with verbal reasoning and inferencing. For example, he had problems answering WH-
questions, especially why, and with interpersonal understanding which requires inferring another person's view of the situation.
In terms of non-verbal abilities, Tony consistently scored above his age level in performance tests of intelligence. The authors
argue that such tests are too global and fail to capture specific cognitive difficulties, especially those which may underlie verbal-
pragmatic problems. In the same vein, Conti-Ramsden & Gunn (1986) argue that the typical picture of good expressive
language (syntax and phonology) in children with pragmatic disabilities, although true in relative terms, was not so in more
complex and subtle ways. For example, they found that Tony had difficulties in marking tense appropriately in discourse
(dialogue and narrative).
There is no doubt that there are advantages to the in-depth analyses provided by case studies, particularly where it is possible to
compare results across cases. A similar rationale formed the basis for the large number of single and small group studies found
in the normal language development literature. A further advantage is that such studies afford the possibility of treating the child
as unique for the planning of therapy. However, there is a danger in proliferating case studies in the absence of a satisfactory
descriptive and explanatory framework. Early case studies fulfilled a need in that they identified and described discourse and
conversational problems experienced by some children. Without some basis for comparison across case studies, however, there
is no opportunity to generalise from individual cases to more general principles.

< previous page page_62 next page >

< previous page page_63 next page >
Page 63
Group Design Studies
Group design studies are based on the assumption that children can be assigned to discrete groups which can then be compared
or subjected to some experimental condition. The advantage of group design studies is that they should provide more general
descriptions than case studies, and experimental conditions should provide control over the reliability of the findings. As far as
research in pragmatic disability is concerned, the approach used in group design studies is to examine language-impaired
children in comparison with language-matched and age-matched peers. Part of the motivation for these studies is an attempt to
extend the study of language impairment by employing linguistic analyses that go beyond the more traditional levels of
phonology, syntax and semantics, to focus on context, discourse and pragmatics, and so is an attempt to explore the possibility
that language-impaired children have problems at the pragmatic level.
The general design of these studies is that a group of language-impaired children is compared on selected pragmatic skills with a
group of age-matched and a group of language-matched peers. For example, a pragmatic skill, such as the use of clarification
requests to resolve verbal misunderstandings, is selected and the different groups are examined for their use of this skill.
(Clarification requests are used when a person does not understand something and feels that it is necessary for subsequent
understanding to resolve the difficulty. He may do so by asking the speaker to repeat or confirm his previous utterance, or ask
specific questions to address the elements which he, the listener, has not understood.) The usual prediction is that the language-
impaired children will be inferior in this ability to their age-matched peers, as it is assumed that their use of language is
impaired. It will also be predicted that their performance will match that of their language-matched peers. In this case, it is
assumed that the ability to use language, such as in making clarification requests, is closely related to the level of a child's
linguistic development.
Results from group design studies have tended both to support and to disconfirm these predictions (see reviews in Fey &
Leonard, 1983; Bryan, 1986). There could be two reasons for these discrepancies. Firstly, the assumptions on which group
design studies are based, chiefly, that language impairment predicts pragmatic disability, could be flawed. This would account
for some of the discrepancies and would lead us to question the validity of this method. A second reason could be that there are
different types of pragmatic disability. Some of these types may be related to language impairment, but others may be largely
unrelated. If there are different types then we might expect that only some are related to language impairment,

< previous page page_63 next page >

< previous page page_64 next page >
Page 64
while other types might be due to non-linguistic impairments. This would explain why some of the predictions of group design
studies have not been supported.
Looking at group design studies more closely we can see some problems. The main assumption of these studies is that
language-impaired children will experience difficulties in the use of language. This assumption is in direct conflict with the
division of language-impaired children into groups with different linguistic characteristics such as phonologic-syntactic and
pragmatic. There may be children with grammatical and/or phonological problems who do not have problems using language,
and children whose main problem is in the use of language and who are relatively unimpaired in formal aspects of language.
Related to this logical inconsistency is the problem of explaining how language-impaired children may or may not be proficient
in the use of language. To take an example, if a child is classified as language-impaired on the basis of grammatical immaturity,
what effect might this impairment have on the ability to make clarification requests? If it could be shown that there was a close
correspondence between the grammatical structures in which the child was deficient and those needed to make clarification
requests, then we might be confident in assuming some relationship between language impairment and pragmatic disability.
However, it is possible that a child knows the syntactic structures which are necessary for requesting clarification, but is unable
to work out when to make the requests because they are not particularly good at monitoring their own comprehension. This
means that they may not make clarification requests, even where these are necessary for mutual understanding. A further reason
might be the desire to avoid appearing slow to understand, especially in a situation where others show signs of easy
comprehension. In other words, the failure to use a particular pragmatic skill might be related to factors which have nothing to
do with a child's formal linguistic abilities. These factors need to be considered fully before assumptions are made about
unidirectional relationships between language impairment and pragmatic disability.
An additional problem of group design studies is the lack of precision in the criteria used for the selection of groups. Measures
such as Mean Length of Utterance (MLU) are often used to match groups. Related to this is the assumption that language-
impaired children are a homogeneous group of individuals and can be examined as such. It is important to point out that unless
the linguistic measures on which these groupings are made are precise, it is likely that many different levels and types of
disability will be subsumed under the one label or group.

< previous page page_64 next page >

< previous page page_65 next page >
Page 65
Nonetheless, group design studies have provided information on questions concerning the relationship between linguistic
impairment and pragmatic disability, and as such provide a tool for further research in this area.

Group Comparison Studies.

In the previous sections we have seen how group design studies are often based on invalid assumptions about the relationship
between language impairment and pragmatic disability, while case studies suffer from a lack of generality. A third approach,
large-scale group comparison studies, offers a promising way forward. These studies involve the use of pragmatic protocols to
classify a range of pragmatic behaviours across a sample of children exhibiting different language disabilities. Examples of such
studies include Prutting & Kirchner (1987), Bishop & Adams (1989) and Adams & Bishop (1989). We will illustrate this
approach with the latter work.
The work of Bishop & Adams (1989) and Adams & Bishop (1989) involved investigating the conversational characteristics of
children with semantic-pragmatic disorders (what we refer to as pragmatic disability) from two perspectives. One perspective
involved the examination of natural conversations of different groups of language-impaired children in an attempt to identify
specific behaviours which differentiated clinical groups. The second approach involved identifying episodes of inappropriate
conversational behaviours and attempting to categorise them in order to gain some insight as to the nature of the children's
difficulties. Bishop and Adams used an audiotaped, semi-structured situation which involved looking at photographs to obtain 5-
to 10-minute conversational interactions between children and an adult female stranger. The conversation was allowed to
develop naturally but the adult could refer to a set of prepared questions when necessary. A sample of 57 language-impaired
children, who could be broadly divided into two groups, participated in the study. Fourteen children appeared to fall within the
category of pragmatic disorders, and 43 children within the category of other language disorders. In addition, approximately 60
normal control children ranging in age from 4 to 12 years were included.
Bishop & Adams (1989) included in their conversational analysis an examination of exchange structure, turn-taking, repairs and
cohesion. They found that children with pragmatic disabilities produced more initiations than other language-impaired children
and this appeared to be the one area which differentiated the groups. Exchange structure, repairs and cohesion were not found to
yield differences. They also found that adults exhibited difficulties in

< previous page page_65 next page >

< previous page page_66 next page >
Page 66
conversing with pragmatically disabled children; adults made a greater number of interruptions and requests for clarification
when interacting with these children. Nonetheless, Bishop and Adams found that their coding scheme was not always reliable in
identifying conversational breakdown. Conversational behaviours falling within the categories of turn-taking and repair yielded
low levels of inter-rater reliability, which brings to the forefront the important question of selecting criteria for making decisions
about the appropriacy of children's and adult's conversational behaviours.
There is no doubt that large-scale group comparison studies have a number of advantages. They are sufficiently large-scale to
permit more general findings, compared with case studies. Also, they avoid the problem of overly delimiting the issues to be
investigated that characterises group design studies. In addition, this approach permits clusters of disabilities to emerge, and
correlation of these clusters with groups of children classified in terms of medical or linguistic categories becomes possible.
However, these studies are only as adequate as the theory of pragmatic competence which underlies them.

Concluding Remarks
Each of the approaches discussed above advances our knowledge in different ways, but some approaches have more advantages
than others. Currently, in the area of pragmatic disability we are just beginning to refine our methodology and lay the
foundations for future research. The tools currently available are blunt when it comes to applying them to pragmatic difficulties
and we need to keep re-examining our approaches. This state of affairs is related to the absence of a satisfactory descriptive and
explanatory framework for pragmatic disability. As such a framework evolves, so will the means for investigation improve.
Acknowledgement
The author would like to thank Michael F. McTear for the many profitable discussions we had in this area. Some portions of
this chapter are expanded in a book: McTear, M.F. & Conti-Ramsden, G. 1991, Pragmatic Disability in Children. London:
Whurr Publishers Ltd.

< previous page page_66 next page >

< previous page page_67 next page >
Page 67
References
ADAMS, C. and BISHOP, D.V.M. 1989, Conversational characteristics of children with semantic-pragmatic disorder. I:
Exchange structure, turntaking, repairs and cohesion. British Journal of Disorders of Communication 24, 211-40.
BEVERIDGE M. and CONTI-RAMSDEN, G. 1987, Children with Language Disabilities. Milton Keynes: Open University
Press.
BISHOP, D.V.M. and ADAMS, C. 1989, Conversational characteristics of children with semantic-pragmatic disorder. II: What
features lead to a judgement of inappropriacy? British Journal of Disorders of Communication 24, 241-64.
BLANK, M., GESSNER, M. and ESPOSITO, A. 1979, Language without communication: A case study. Journal of Child
Language 6, 329-52.
BRYAN, T.H. 1986, A review of studies on learning disabled children's communicative competence. In R.L.
SCHIEFELBUSCH (ed.) Language Competence: Assessment and Intervention. London: Taylor and Francis.
CONTI-RAMSDEN, G. and GUNN, M. 1986, The development of conversational disability: A case study. British Journal of
Disorders of Communication 21, 339-51.
FEY, M.E. and LEONARD, L.B. 1983, Pragmatic skills of children with specific language impairment. In T.M. GALLAGHER
and C.A. PRUTTING (eds) Pragmatic Assessment and Intervention Issues in Language. London: Taylor and Francis.
McTEAR, M. 1985, Children's Conversation. Oxford: Blackwell.
PRUTTING, C.A. and KIRCHNER, D.M. 1987, A clinical appraisal of the pragmatic aspects of language. Journal of Speech
and Hearing Disorders 52, 105-19.

< previous page page_67 next page >

< previous page page_68 next page >
Page 68

5
Analysis of Language-impaired Children's Conversation
CATHERINE ADAMS
Background
Issues of classification of childhood language disorder were once again brought to the fore in the 1980s by Rapin & Allen
(1983, 1987). Leaving aside discussion of the merits and disadvantages of this type of `clinical-linguistic' classification, Rapin
& Allen's work triggered debate on an interesting new category of developmental language disorder `semantic pragmatic deficit
syndrome' (Rapin & Allen, 1987). Of course clinicians will recognise this as a new label, rather than a new syndrome. In fact,
such cases had already been described in the literature (Byers Brown, 1979). Nevertheless an increase in descriptive studies and
some research resulted from the new label.
Rapin & Allen (1987) describe children with semantic-pragmatic deficit syndrome as having the following characteristics of
communication:
fluent, well-formed sentences
verbosity
expression superior to comprehension
adequate vocabulary
the tendency to interpret language literally
possible echolalia (especially at preschool age)
semantic errors
turn-taking problems
topic maintenance problems
the use of over-learned scripts.
With such a comprehensive description, the clinician should have no problems in identifying a language-impaired child as
falling within this

< previous page page_68 next page >

< previous page page_69 next page >
Page 69
category. However, clinical diagnoses are one thing, empirical research quite another. In a lengthy research project Bishop &
Adams (1989) attempted to pinpoint more precisely what it is about these children which distinguishes them from other
language-impaired individuals, and also to identify which forms of assessment would be most efficient for the speech therapist.
To this end we compared a group of children with semantic-pragmatic problems (as defined by Rapin & Allen's criteria) with a
heterogeneous group of language-impaired children (both 8-12 years old) and a large control group of language-normal children
(in age bands 4, 6, 8, 10 and 12 years), on a number of measured tasks including:
Test for Reception of Grammar (TROG) (Bishop, 1982)
WISC-R Verbal Comprehension (Weschler, 1976)
Mean Length of Utterance (MLU)
Word-finding abilities (Renfrew scale, n.d.)
Inferences task*
Referential communication task.*
(Those marked* are described elsewhere in detail (see Adams & Bishop, 1989; Bishop & Adams, 1989).)
At the outset it was predicted that the semantic-pragmatic group would present a different profile on the test battery from other
language-impaired children. It rapidly became clear, however, that no single measure or set of measures systematically
discriminated between the two experimental groups. This was disappointing, since qualitatively and informally there appeared to
be a difference. For instance, the content of responses on WISC-R Verbal Comprehension from some of the semantic-pragmatic
group was evidently of a different nature from the typical immature responses of the language-impaired group, but the scoring
did not reflect this. The lessons learnt were that: (a) the more formal and structured the assessment the more `normal' the
semantic-pragmatic group appeared, and (b) a descriptive/analytical approach to assessment rather than a statistically based one
was required.
Conversation analysis was adopted, somewhat reluctantly, as a solution to these observations reluctantly, since such analyses are
time-consuming; there are problems of experimental control over topic; and the linguistic theory of conversation is poorly
worked out as yet. The analysis adopted by Adams & Bishop (1989) was based on McTear (1985) with extensive additions and
modifications for this population, including a lengthy analysis of types of inappropriacy. Details of how the adult-child
conversations were sampled can be found in Adams & Bishop (1989).
Several areas of conversation were defined as being of specific interest for

< previous page page_69 next page >

< previous page page_70 next page >
Page 70
this group of children: exchange structure, turn-taking, repairs, cohesion, and appropriacy. Predictions for the semantic-
pragmatic group as compared to age-matched controls and other language-impaired children included:
1. Exchange structure: a possible surplus of initiations, reflecting true verbosity.
2. Turn-taking: violations and a lack of understanding of pragmatic rules, again reflecting verbosity.
3. Repairs: an increased number of repairs due to frequent conversational breakdown.
4. Cohesion: disturbances in the use of cohesive devices (e.g. the use of pronouns to refer back to nouns already mentioned) due
to referential disability.
5. Appropriacy: a significantly higher rate of inappropriate utterances.

Findings.
Reliability ratings for test-retest and inter-rater analyses were established as being satisfactory. The two most revealing areas of
conversation analysis were exchange structure and appropriacy; other hypotheses were not confirmed.
Exchange structure is the way in which conversation is organised sequentially. Typically rules will be obeyed, so that a response
follows an initiation. For example:
A: did you have a party? [initiation]

C: yes [response]

A: and what did you do there? [initiation]

C: played games [response]

[A = adult, C = child]
One of the distinguishing measures between the `semantic-pragmatic' and `other language-impaired' groups was that the former
display a higher rate of initiations.
Appropriacy, and therefore inappropriacy, is hard to define. For the purposes of this project inappropriacy was defined as
occurring when any child utterance appeared odd or bizarre in that context, or in any context. From the data corpus several
categories of inappropriacy were detected as occurring in more than one individual, some of which are illustrated here. The
examples below are drawn mainly from Bishop & Adams (1989) where a comprehensive description of the categories used can
be found.

< previous page page_70 next page >

< previous page page_71 next page >
Page 71
Expressive semantics:
A: how did you get to Macclesfield?
C: we went on a bus because Ian was sick out of the window.
Reiteration/over-elaboration (too much information):
A: would you say the boy looks ill?
C: well no I don't think he would look ill if he was just checking he was alright, cos he doesn't look ill, and if he doesn't look ill
he's alright and he doesn't look ill.
Inappropriate questioning:
[topic = measles]
C: what colour were they?
A: what colour were the measles? I don't know.
Topic shift and unusual content/style:
A: why has that doctor come?
C: to see if he's sick or something
A: yes/and have you been sick like that?
C: that's [1 syll]/he has been sick/ what about people fall in what/what about lakes?
A: what about lakes?
Categories of inappropriacy which distinguished among the groups were: expressive semantics/syntax, too little information, too
much information and unusual content/style.

Discussion
It appears from these findings that conversation analysis presents us with one method of pinpointing the language difficulties of
the child with semantic-pragmatic disorder. In fact there is no reason why the use of conversation analysis should not be
extended to other diagnostic groups provided that appropriate modifications are made to the techniques used and pertinent
research questions are asked.
The remainder of the discussion is concerned with three issues:
a. individual `styles' of language;
b. relationship of the semantic-pragmatic group to other language-impaired children;
c. relationship to autism.

< previous page page_71 next page >

< previous page page_72 next page >
Page 72
Individual Styles of Language
One striking finding which does not emerge statistically from the Bishop & Adams study, but is nevertheless important, is that
individual children in the semantic-pragmatic group, while possessing characteristics typical of the syndrome, presented with
very individual styles of language. For instance, whereas `turn-taking' as a group measurement could not distinguish the two
experimental groups, certain semantic-pragmatic children demonstrated a high frequency of turn-taking problems clashing,
butting-in and so on.
This was also the case with cohesion, to some extent, and with other pragmatic skills. Some children had very high rates of
initiation, explained perhaps as a strategy to avoid being questioned, or equally likely as a genuine pragmatic deficit: a lack of
appreciation of the partner's conversational rights. Group measures tend to mask such individual characteristics which are
nevertheless of considerable importance from the point of view of therapeutic planning.
Relationship to Other Language-impaired Children
Many language-impaired children have semantic and/or pragmatic deficits. Not all of these, however, fit the `semantic-
pragmatic' label as used by Rapin and Allen. For instance, Prutting & Kirchner (1987), in applying their `Pragmatic Profile' to
41 language-impaired children (heterogeneous group), found some pragmatic deficits in all children. In these cases the
pragmatic impairment could be seen as a secondary consequence of an inadequate language system (Prutting & Kirchner, 1983).
There is, perhaps, in Britain a tendency to over-use the term `semantic-pragmatic disorder' to describe any older child with
reasonably fluent speech and residual language problems. This should be avoided since the term was originally used to describe
a syndrome with specified characteristics: that is, the implication of `semantic-pragmatic' is more than the sum of the meaning
of the two individual words. (See Chapter 1 for further discussion.)
There seems, however, to be some advantage in using the term in a descriptive sense because it can reveal further patterns or
types of difficulty. For instance, in the Bishop & Adams study there appeared to be some polarisation in the semantic-pragmatic
group into those children with more `expressive semantic' inappropriacy and less pragmatic inappropriacy, and those with the
reverse pattern.

< previous page page_72 next page >

< previous page page_73 next page >
Page 73
Relationship to Autism
Controversy still surrounds the relationship of semantic-pragmatic disorder to autism in children. Rapin and Allen did not equate
the two, though they frequently found this language pattern in their autistic group. Of course, it rather depends on the definitions
of the boundaries of autism applied an area of discussion that cannot be contemplated in this chapter. Some would say that
semantic-pragmatic syndrome and autism are the same thing; others that semantic-pragmatic syndrome children are not autistic.
The important argument from the former viewpoint is that if you are autistic then the social, and consequently language deficits
are more or less irremediable. It seems that this point of view requires some substantiation for the mildly autistic child with few
residual social difficulties and some insight.
If one examines the various descriptions of the language features of high-level functioning autistic children there is close
resemblance to those of semantic-pragmatic deficit syndrome. Linguistic features of autism described are: obsessive/stereotyped
language, over-learned scripts, repetitive/over-elaborate speech, topic maintenance problems and presupposition problems. Many
of the Bishop and Adams semantic-pragmatic group had a history of autistic features, but none had ever been diagnosed autistic.
Unpublished data from Adams using conversational analysis from children diagnosed as autistic suggests that semantic/syntactic
errors are extremely rare, and the types of inappropriacy shown are dominated by obsessive/stereotyped language. Thus there
may be a tendency for the semantic-pragmatic diagnosed children to present a different profile from that of the autistic child, but
with some overlapping characteristics.
In a sense the autistic/not autistic debate, as applied to semantic-pragmatic disorder, is vacuous. The terms `autism' and
`semantic-pragmatic disorder' exist in different models (see Figure 5.1). Autism is a categorical/diagnostic medical term, based
on behavioural criteria which encompass far more than language behaviours; semantic-pragmatic disorder is essentially a
descriptive linguistic term. Thus we should be able to apply it to language-impaired children, regardless of any medical
diagnosis.
Medical model  diagnosis  autism
Linguistic model  description  semantic-pragmatic disorder
Figure 5.1
Models from which the terms `autism' and
`semantic-pragmatic disorder' derive
There are grey areas which might be illuminated; for example the borderline cases between children with semantic-pragmatic
linguistic features who have been diagnosed as autistic, and those who have not; the overlap with

< previous page page_73 next page >

< previous page page_74 next page >
Page 74
specific semantic disorders; and children whose individual language profile changes over time and who show residual
difficulties involving semantic and pragmatic abilities.

References
ADAMS, C. and BISHOP, D.V.M. 1989, Conversational characteristics of children with semantic-pragmatic disorder. I:
Exchange structure, turntaking, repairs and cohesion. British Journal of Disorders of Communication 24, 211-40.
BISHOP, D. 1982, Test for reception of grammar (TROG). Department of Psychology, University of Manchester, M13 9PL.
BISHOP, D.V.M. and ADAMS, C. 1989, Conversational characteristics of children with semantic-pragmatic disorder. II: What
features lead to a judgement of inappropriacy? British Journal of Disorders of Communication 24, 241-63.
BYERS BROWN, B. 1979, Case presentation. In B. BYERS BROWN and M. BEVERIDGE (eds) Language Disorders in
Children: A Workshop for Teachers. The College of Speech Therapists, London.
McTEAR, M. 1985, Children's Conversation. Oxford: Blackwell.
PRUTTING, C.A. and KIRCHNER, D.M. 1983, Applied pragmatics. In T.M. GALLAGHER and C.A. PRUTTING (eds)
Pragmatic Assessment and Intervention Issues in Language. San Diego, CA: College-Hill Press.
1987, A clinical appraisal of the pragmatic aspects of language. Journal of Speech and Hearing Disorders 52, 105-19.
RAPIN, I. and ALLEN, D. 1983, Developmental language disorders: Nosologic considerations. In U. KIRK (ed.)
Neuropsychology of Language, Reading and Spelling. New York: Academic Press.
1987, Developmental dysphasia and autism in preschool children: Characteristics and subtypes. In Proceedings of the First
International Symposium on Specific Speech and Language Disorders in Children. London: AFASIC.
RENFREW, C.E. n.d. Word finding vocabulary scale. Available from the author, 2a North Place, Old Heddington, Oxford.
WECHSLER, D. 1976, Wechsler Intelligence Scale for Children (revised edn). Windsor: NFER.

< previous page page_74 next page >

< previous page page_75 next page >
Page 75

6
A Teacher's and a Speech Therapist's Approach to Management.
SUSAN HYDE-WRIGHT and BARBARA CRAY
Introduction
Dawn House School, near Mansfield, Notts, is one of the language schools run by the Invalid Children's Aid Nationwide
(ICAN). It caters for children with a variety of speech and language disorders. In this school there is a class composed of
children with semantic-pragmatic difficulties; it has been found that this provides many opportunities for useful teaching and
therapy intervention.
An approach to management is described which involves close colloboration between speech therapist and teacher.
Part I
Susan Hyde-wright, Speech Therapist
The term `semantic-pragmatic' has sparked off a lively debate over its definition as applied to disorders of communication.
Rapin & Allen (1983) were the first to use the term `semantic-pragmatic syndrome' to describe a developmental language
disorder in which the form of language seems relatively normal but the content and use is abnormal. Recognition of these
features within other syndromes such as infantile autism, `Cocktail Party syndrome' (Tew, 1979), Asperger's syndrome (Wing,
1981) and fluent language disorder (Bishop, 1981) has resulted in terminological confusions and overlapping categories
indicating that these children cannot be clearly distinguished from other children in the language-disordered population, or from
other populations of handicapped children.

< previous page page_75 next page >

< previous page page_76 next page >
Page 76
Yet, irrespective of the label or source of referral, teachers and speech therapists are meeting children who present with a cluster
of features that bring them together, although within that group there is considerable variation. Many of their difficulties
preclude them from education in the mainstream school so they are being sent to language units and schools where therapists
and teachers must devise methods to meet their needs.
Broadly speaking, the characteristics of children with semantic-pragmatic difficulties are:
1. That they fail to respond in an appropriate manner to either the linguistic or the social demands of situations. Here
conversation may break down as the pupil fails to see what is most relevant in the other speaker's utterances and may reply
tangentially.
2. That they show a poor interrelation between the content and use of language.
3. That comprehension is more impaired than expression. (Generally, when language is first examined, these children are able to
produce utterances which are syntactically and phonologically well formed, although Bishop & Adams (1989) state that their
analysis of expressive language revealed a wide range of semantic, syntatic and pragmatic peculiarities.)
4. That concept formation is poor and rigid. (An example of rigidity is where one child listed tennis, cricket and rounders as
games but would not accept the suggestion that football was also a game, as he considered this to be a sport.)
5. That they cannot use language as an anchor to reality; they fail to correlate the language they hear with their experience in the
real world.

Assessment
There is no one formal assessment which effectively diagnoses these difficulties. Many testing procedures are inadequate or do
not pinpoint the exact nature of semantic-pragmatic problems. Therefore, it is initially from close observation that staff at Dawn
House School formulate their provisional diagnoses and develop baselines for work with the children with semantic-pragmatic
difficulties.
We are at the stage of empirically describing and profiling the behaviours of such children, so that we gain some awareness of
related aspects of their difficulties. This may in turn highlight those areas where problems occur and indicate patterns of
development over time. A broad range of semantic and pragmatic abilities are assessed alongside other aspects of language, such
as

< previous page page_76 next page >

< previous page page_77 next page >
Page 77
phonology and syntax as well as non-verbal abilities and personality. This method is modelled on the longitudinal case studies
of Conti-Ramsden & Gunn (1986).
Of 200 children entering Dawn House School between the ages of five and 13 years, 30 have presented with semantic-pragmatic
problems. It is interesting to note that these 30 children make up 15% of accepted referrals. This suggests that a significant
proportion of children with language disorders have these difficulties, which pose considerable educational and therapeutic
problems. However, these children may well be only the tip of the iceberg.
The formal language assessments used have indicated that children with semantic-pragmatic difficulties perform well in the
following: Edinburgh Articulation Test (EAT: Anthony, Bogle, Ingram & McIsaac, 1971), Language Assessment, Remediation
and Screening Procedure (LARSP: Crystal, Fletcher & Garman, 1988), and the grammatic closure subtest of the Illinois Test of
Psycholinguistic Ability (ITPA: Kirk, McCarthy & Kirk, 1968). However, they perform poorly in the auditory sequential
memory subtest of the ITPA and the British Picture Vocabulary Scales (BPVS: Dunn, Dunn, Whetton & Pintilie, 1982).
As there are no standardised procedures to investigate conversational skills at present available for use with these children, it has
been necessary to assess conversational interaction using elicitation procedures. Typical areas investigated are:
1. Conversation about something familiar and not present, e.g. birthday, pet.
2. Picture story sequences to see if the pupil (P) can organise and narrate a sequence of events.
3. Role play to see if P is aware of another person's perspective and can put himself outside the `here and now' (particularly
difficult for children with semantic-pragmatic difficulties).
4. Pictures of disasters or problem situations to see if the child shows awareness of the antecedents or outcome of the depicted
episode.
5. Some aspect of reasoning about similarities and differences, e.g. the child is asked to explain the similarity or difference
between an orange and a biscuit.
In addition, P is observed and assessed in a variety of settings with different partners and at different times in order to build up a
comprehensive description of the child's communication skills, both verbal and non-verbal.
Through using this variety of standardised and non-standardised procedures for reassessment over a period of time,
improvements have been noted in the following areas: syntax, phonology, concepts/semantics,

< previous page page_77 next page >

< previous page page_78 next page >
Page 78
sociability, confidence and non-verbal communication. This is more clearly illustrated with reference to a specific child (Y) by
using assessments and comments from reports from the age of 6:4 to 10:1 (see Table 6.1).
However, improvement is slower and problems persist in the following areas: sequencing events, planning, ways of thinking
(e.g. verbal reasoning, picking up clues, making inferences), receptive vocabulary as measured by BPVS, social awareness,
prosody, stereotyped phrases. These observations can be illustrated in terms of a child (Q) using vocabulary assessments, and
comments from reports from the age of 6:6 to 15:6 (see Table 6.2).
The summary on Q's final leaving report reflected both her improvements and her persisting difficulties:
Over 9 years Q has overcome her grammatical and phonological problems satisfactorily. She has shown herself able to
learn many of the rules of discourse and social involvement and can adequately, if not sensitively, extract and express
language meaning. She is not unpopular but has no close friends. She can still miss the point of remarks or jokes and
remains unaware that she has done so.

Intervention
Throughout all aspects of intervention it is important that therapy is relevant to the children's communication needs and has a
practical focus. Children must be helped to see links and associations between concepts for themselves in order to use and
incorporate this knowledge in the development of an organised and flexible understanding of the world. To achieve maximum
relevance, vocabulary and concepts are taught when needed and not detached from the child's classwork or interests. This in turn
helps to break down rigid concept boundaries and link ideas in a new and flexible way. Language is used and taught in a
practical problem-solving situation even though children with semantic-pragmatic difficulties prefer the cosy security and more
predictable routine of pen and paper, worksheets and familiar board games. The use of real-life situations helps to avoid the
confusion that sometimes occurs for these pupils when more table-top activities are used, and they are expected to transfer this
knowledge gained in an artificial therapeutic setting to events in real life. Frequently the child with semantic-pragmatic
difficulties will avoid practical tasks as they create problems possibly due to inflexibility. Furthermore, a joint practical focus for
therapist and pupil ensures that the referents and their meaning are available to be verified and the

< previous page page_78 next page >

< previous page page_79 next page >
Page 79
TABLE 6.1 Observation period 6:4-10:1 for Child Y
6:4-7:1 Chronological age 8:2-9:1 9:2-10:1
7:2-8:1
LARSP stage III V-VI VI-VII
V
8/25 41/50 N/A
Boehm Test of Basic 33/50
Concepts
Raw scores
51 N/A N/A
Phonology 57
EAT Raw Score*
4:7
Age equivalent 5:5
Confidence Needs a lot of Still needs Likes to
attention in Does not ask security ofdirect
class. Shows for help in small and `boss'
fear of class but will group. in
crowds/shops. settle to work Offers to class.
Rarely speaks quickly. At help Anxious
to adult playtime stays others. to please
spontaneously. with adult. Asks but
Poor eye for still
contact. meaning prefers
of own
new company.
words. Joins in
Will group
volunteer games.
to join
group.
* Standard scores not available above 6 years

< previous page page_79 next page >

< previous page page_80 next page >
Page 80
TABLE 6.2Observation period 6:6-15:6 for Child Q
6.6 8.6 Chronological13.5 15.6
age
9.8
BPVS
Vocabulary
75 86 68
Standardised 82 90
score
18
Percentile 5 2 12 26
rank
Age 50 6:7 6:2 10:3 13:4
equivalent
Ways of Response Follows own train Tends to Often interprets Can pick
thinking tangential. of thought. Takes grasp the things at face important
No clear logic figurative language wrong end of value. Not always information
to literally. the aware of shared from
utterances. stick but can information. spoken
Semantic supply passages.
themes are answers to Problems
hard to high with new
link unless level concepts e.g.
perceptually questions. science.
related.
Social Able to No grasp of When teased Responded to T's Understands
awareness distinguish pupil/teacher by verbal errors by pupil/teacher
happy and sad relationship. others lies on patting T on head relationship.
faces the and calling her
only, but when floor or `silly billy'.
asked screams.
how she would Classroom
look behaviour
if Mummy and responses
smacked to
her Q pointed peers are
to socially
happy face. inappropriate.

T = Therapist

< previous page page_80 next page >

< previous page page_81 next page >
Page 81
conversation can be more tightly controlled; this helps to avoid confusion and the spirals of conversation that result when the
topic involves absent referents. Then the child can be assisted, if required, to give greater precision to his communication. This
participation in practical activities and real-life problems provides motivation and experience that cannot be simulated through
practice activities, which isolate and teach language skills outside the context of meaning and functional use (Bricker, 1986).
Obviously, close liaison and co-operation with the class teacher are imperative to find out how intervention in the classroom can
be most effective, and to enhance and contribute to overall class management. This may result in withdrawal of the pupil from
the class or working in the class with a group or individually. To ensure that therapy has immediate and maximum relevance it
should be centred around the classroom topic or related to it in a way that broadens understanding. Central to each topic is a core
vocabulary which has been predetermined by the therapist and teacher together. Reference is made to the vocabulary in as many
different situations as possible, which aims to facilitate vocabulary acquisition by building up a broad reference system which
the child is encouraged to access (see McCartney, Flower & Meteyard, 1986). Children with semantic-pragmatic difficulties
require therapy in many different semantic fields and at individually appropriate levels.
Within the scope of this chapter only a few aspects of therapy and some different approaches which have been found to be
useful can be highlighted. The ideas may be modified and adapted to suit individual needs and may be relevant to several
therapy areas.
Conceptual Expansion
It has been noticed that children with semantic-pragmatic difficulties tend to categorise experiences rigidly and develop narrow
boundaries in concept development. So therapy must involve a variety of activities that encourage these children to categorise
items in different ways for different purposes; recognising their important features, their similarities and their differences;
understanding and later expressing different ways in which they can be grouped. This helps to extend the boundaries of the
concepts flexibly.
Generalisation of new vocabulary is important right from the start. It is often easier to begin by extending concepts already
established and perhaps rigidly demarcated. Consider the example of two children who had difficulties in understanding the
appropriate uses of containers for food. Child M

< previous page page_81 next page >

< previous page page_82 next page >
Page 82
believed that fruit could only be put in a round, glass, fruit bowl (not a plastic one, nor a wicker basket etc.) whereas child L
thought nothing could be carried in a named carrier bag unless it had been bought from the same named shop. So began the
process of simple categorisation tasks, using a wide variety of containers to proceed from gross distinctions to the more subtle,
building in as many alternative uses for containers and a variety of possible food items to be contained.
As the children's understanding developed other activities were incorporated such as:
Lotto games paired by functional association rather than by picture or object matching.
Odd-one-out tasks, where the child is asked to pick an item which is somehow different from, or suggest one similar to, those
already displayed.
Lists: as many of these children are good readers and spellers they are encouraged to write lists of items which share certain
characteristics. As well as helping them to focus only on the essentials for classification, these lists provide a permanent record
for further reference which can be used in other classroom activities.
Matrices: the children's enjoyment of plans and diagrams is used to create a classification matrix. Later squares can be covered
to see if the children can work out for themselves what could be in the square. See Table 6.3 for an example, made in this
instance for food labels when working on the topic of shopping/food/storage.
TABLE 6.3 Classification using a simple matrix
Freezer Fridge Cupboard
Meat Beefburgers Sausages Tin of spam
Vegetables Frozen peas Celery Baked beans
Dessert Ice-cream Jelly Biscuits

Verbal Reasoning.
Often children with semantic-pragmatic difficulties have a logical but idiosyncratic approach to verbal reasoning. What may
appear to be illogical responses in conversation are the result of pragmatic failures; for example, McTear (1985) reports on a
conversation with a child who appears not to have understood the interviewer's questions. Yet when the child's part of the
conversation is analysed, it is shown to progress logically but not in the way

< previous page page_82 next page >

< previous page page_83 next page >
Page 83
the adult expected. Verbal reasoning is a complex area as the child has to extract the relevant dimensions of a problem to arrive
at a solution. As with so many abstract skills, he needs to start with the concrete example and work out a solution by trial and
error. From this experience the child gradually learns to internalise the process through verbal mediation. So one of the first
tasks when solving a concrete problem is to recognise the irrelevant information and physically remove it. At a later stage the
child is encouraged to exclude irrelevant material mentally. Once again in all these types of activities vocabulary and/or
situations are used that are taken from or linked with the class activity.
For example, the child is asked to select from a range of items which ones are needed when going shopping. (This could be
made more precise by specifying items which are needed for shopping when it is raining). Possible items might be cardigan,
money, pair of gloves, carrier bag, list, pencil, some sweets, roller skates. It should be emphasised to the child that these items
are associated only for this particular purpose. Other items would be associated under different circumstances. This helps to
avoid some of the dangers of rigid association. Once again, therapy should be related to real-world situations.
Other activities could involve:
Letting the child be a detective to solve a crime where at first the clues are tangible and can be referred to as needed. A useful
starting point is to be found in Concept 7-9.
Giving simple oral instructions to draw fantasy or mythical characters with particular features, e.g. mouth but no teeth; three legs
but no arms, and so on. Then, after the drawing has been completed it can be used as a reference source. For example, the child
is asked if the character could chew meat, or climb trees, and is encouraged to explain his answers.
Presenting visual information or scenarios with a statement so that the child has to reason whether the statement is true or not, or
if there is insufficient evidence available for him to decide; for example, P is shown a figure in the kitchen opening the fridge
door with accompanying comment by the therapist (T), `Dad's getting the baked beans'.
Conversational Skills
Often children with semantic-pragmatic difficulties fail to respond in an appropriate manner in conversational situations. Two of
the many ways their difficulties show themselves are: (a) they ask a lot of questions, (b) they have problems with turn-taking.

< previous page page_83 next page >

< previous page page_84 next page >
Page 84
(a) Many children with semantic-pragmatic difficulties find it hard to signal when they cannot cope with the conversational
exchange, being unable to fine-tune, compensate or modify their language to repair the conversation. Possibly they resort to
constantly repeating inappropriate questions, so that they can control the conversation and avoid what is to them the less safe
role of listener-responder. Another possibility is that questioning is partly the result of anxiety, for the questions appear to
diminish in secure environments with a supportive adult. Furthermore, with many of these children, it is possible eventually to
deflect the question back to the child who is often able then to work out the answer for himself.
(b) Turn-taking is a defining characteristic of conversation and precedes other pragmatic behaviours developmentally. It begins
with the child taking turns in conversation although his turn may not be in a recognisable verbal form at this early stage. Turn-
taking skills continue to develop until the child is able to produce more structural and recognisable conversational turns and
more advanced levels of discourse. Yet this area presents particular problems for children with semantic-pragmatic difficulties.
There are several publications which give useful ideas for turn-taking activities (Brandes & Phillips, 1977; Cartledge & Milburn,
1986; Rustin & Kuhr, 1989; Spence, 1980). The choice of game or exercise must naturally be considered in the light of the
child's language skills, his age and maturity, as some activities may well be too hard or need modification for younger children.
Working in a school context is particularly helpful for turn-taking exercises as children can work in groups or pairs. Useful
activities include:
A pair of children have to explain to other children what they have to do in order to duplicate their activity, e.g. to make an ice-
lolly. Thus both children have to take turns to impart specific instructions for the various stages involved.
Telling a story using picture sequences where both children take it in turns to tell the stages of the story, ensuring that each
makes a coherent link to the previous episode.
The child may need to be taught to recognise when his turn is being signalled non-verbally or by intonation. This often requires
rather exaggerated modelling by the therapist to enable the child initially to `cue in'. Here a version of Kimm's game can be
played where the therapist recalls a selection of items that she and the child have briefly seen on a tray but which have since
been covered. When the therapist begins to falter in her recall, the child has to recognise his turn and either prompt or answer
for T.

< previous page page_84 next page >

< previous page page_85 next page >
Page 85
Inference
Children with semantic-pragmatic difficulties understand structured, concrete situations more easily than conversations where
events referred to are beyond the `here and now'. The pupil finds it difficult to draw inferences from material in either a verbal
or a visual form. For example, a child shown a picture of a lady with a shopping bag, purse and shopping list is asked `where is
she going?' but is unable to answer, not because he fails to understand the question but because he could not use the information
in the picture to make the appropriate inference. To make inferences the child has to be able to combine new information with
his previous knowledge and experience through logical thinking. Often the child fails to recognise and evaluate the relevance of
prior information or has not retained that information. He may fail to make the rapid associations and generalisations necessary
from the key words or items presented in the problem. In a practical activity the therapist can notice whether P has recalled
pertinent information and/or knowledge.
Situational inference occurs naturally in practical tasks; events can be seen to be related both temporally and causally. Initially,
where possible the child is allowed to see the effect of what he does, whether it is trying to wash up without using hot water or
serving out jelly with a teaspoon. When he perceives that his actions are not the most appopriate the therapist can guide him to
decide how the difficulty may be resolved and test to see if he is correct. The therapist should provide prompts in the form of
questions, comments or new information that can be used to generate a revision of the solution. Thus the revision is child-
initiated and it is the child who is engaged in the process of rethinking solutions to the problem.
Once more the appeal of plans, diagrams and charts can be exploited to show causes and effects, or make inferences or predict
effects if the cause is known, utilising events from previous practical sessions. For example:

pick up dirty dish put clean dish on the side

To encourage flexibility, situations can be created where there are several options which, in turn, lead to different consequences.
This also helps the child to imagine possible outcomes. For example:

These children will also frequently require persistent and continuing help with making inferences from verbal material. The child
needs to consider implicit meanings as well as information which is explicitly given. In addition,

< previous page page_85 next page >

< previous page page_86 next page >
Page 86
frequently the literal meaning of relatively simple utterances is taken rather than the intended meaning, as occurred during a
practical session when the comment was made by the therapist: `Good, we are all ready to begin. Off you go', and a child with
semantic-pragmatic difficulties left his seat and made to leave the room. This suggests difficulty in understanding the speaker's
perspective and integrating verbal inferences within the situational context. Teaching implied requests or action is very strained
and artificial unless it is included during an activity. It is useful to go armed with two or three comments or messages that are
likely to be appropriate during a teaching session: for instance, during a cookery lesson, `it's awfully steamy in here', meaning
`open the window'; `I can't see the timer on the cooker', implying that the child should move out of the way. When inappropriate
inferences are made the therapist can talk the child through the event by referring to and highlighting the salient information and
cues available for making a more probable and accurate deduction. The child should be encouraged to indicate when he does not
understand and so reduce the number of faulty inferences made.
Activities which can also encourage the child to make verbal inferences are:
Variations of oral guessing games or riddles such as `what am I?', initially restricting choices by providing pictures from which
to select the answer. Later the child should be encouraged to do this without the aid of pictures.
The therapist acts out scenes with toys or puppets, in which the characters give clues as to their intentions and the child answers
questions such as `where are they going?' or `what are they going to do?': for example, a child in a push-chair saying to his
mother, `can we feed the ducks today?', or a man walking along a road saying `I have run out of petrol'.
`Nicknames' or `what do I look like?' games, the child guessing the appearance or nature of the person or animal from the name
they have been given. The nicknames can progress from the simplest obvious level (a cat called `Snowy' is white) to the more
sophisticated (a dog called `Patch' has a patch of colour on his body) to quite hard ones, such as a dog called `Domino' has black
and white spots.
Clearly these approaches can be extended or modified to suit individual children and can be linked with many class activities.
I have described methods currently being pioneered in Dawn House School with children with semantic-pragmatic difficulties,
in order to enable them to interpret and use language more effectively. In the future there needs to be a continuing analysis of
their needs, and evaluation of current practices with a thoughtful and open-minded approach to developing further

< previous page page_86 next page >

< previous page page_87 next page >
Page 87
intervention methods. These three things together should engender a deeper understanding of the nature of semantic-pragmatic
difficulties.
Acknowledgement
I would like to thank Corinne Haynes for her helpful comments and encouragement.

References
ANTHONY, A., BOGLE, D., INGRAM, T.T.S. and McIsaac, M.W. 1971, The Edinburgh Articulation Test. Edinburgh:
Churchill Livingstone.
BISHOP, D.V.M. 1981, Varieties of childhood language disorder. Northern Ireland Speech and Language Forum 7, 20-37.
BISHOP, D.V.M. and ADAMS, C. 1989, Conversational characteristics of children with semantic-pragmatic disorder. II: What
features lead to a judgement of inappropriacy? British Journal of Disorders of Communication 24, 241-63.
BOEHM, A. 1971, The Boehm Test of Basic Concepts. London: Psychological Corporation and Harcourt Brace Journovich.
BRANDES, D. and PHILLIPS, H. 1977, Gamesters' Handbook. 140 Games for Teachers and Group Leaders. Oxford:
Hutchinson.
BRICKER, D.D. 1986, Early Education of At-Risk and Handicapped Infants, Toddlers, and Preschool Children. Glenview, IL:
Scott, Foresman.
CARTLEDGE, G. and MILBURN, J.F. (eds) 1986, Teaching Social Skills to Children. Innovative Approaches, 2nd edn.
London: Pergamon.
CONCEPT 7-9. Nottingham: NES Arnold.
CONTI-RAMSDEN, G. and GUNN, M. 1986, The development of conversational disability: A case study. British Journal of
Disorders of Communication 21, 339-51.
CRYSTAL, D., FLETCHER, P. and GARMAN, M. 1988, The Grammatical Analysis of Language Disability. London: Edward
Arnold; 2nd edn, London: Whurr.
DUNN, L.M., DUNN, L.M., WHETTON, C. and PINTILIE, O. 1982, The British Picture Vocabulary Scale. Windsor: NFER-
Nelson.
KIRK, S.A., McCARTHY, J.J. and KIRK, W.D. 1968, Illinois Test of Psycholinguistic Ability (revised edn). Windsor: NFER-
Nelson.
McCARTNEY, E., FLOWER, P. and METEYARD, J. 1986, A study of word retrieval difficulties in language disordered
children. Paper presented at ICAN Conference, Advances In Working With Language Disordered Children. ICAN, London.
McTEAR, M. 1985, Pragmatic disorders: A case study of conversational disability. British Journal of Disorders of
Communication 20, 129-142.
RAPIN, I. and ALLEN, D. 1983, Developmental language disorders: Nosologic considerations. In U. KIRK (ed.)
Neuropsychology of Language, Reading, and Spelling. New York: Academic Press.
RUSTIN, L. and KUHR, A. 1989, Social Skills and the Speech Impaired. London: Taylor and Francis.
SPENCE, S. 1980, Social Skills Training with Children and Adolescents. Windsor: NFER-Nelson.
TEW, B. 1979, The `Cocktail Party Syndrome' in children with hydrocephalus and spina bifida. British Journal of Disorders of
Communication 14, 89-101.
WING, L. 1981, Asperger's syndrome a clinical account. Journal of Psychological Medicine 11, 115-29.

< previous page page_87 next page >

< previous page page_88 next page >
Page 88
Part II
Barbara Cray, Class Teacher
A number of children whose major difficulties lie in the semantic or pragmatic areas of language (often both) have always
formed part of the population of Dawn House School, being approximately 15% at any one time. However, there has been a
recent increase in the number of children being referred to the school with these problems. Between 1985 and 1988 a total of
seven children with `semantic-pragmatic' difficulties were referred. In the last year alone there have been six.
Often classes contained one or two children whose major difficulties were of a semantic-pragmatic nature. Consequently, their
problems were different from those of the rest of the group. It became clear that the curriculum suitable for the language-
disordered population at Dawn House was not always appropriate for these children. They also required different teaching
approaches from the rest of the group, within which they tended to be isolated and withdrawn. For example, in one particular
class of language-disordered children, the two children with semantic-pragmatic difficulties preferred to sit in separate parts of
the room reading books when they first came into the classroom, while the rest of the group were usually chattering together.
Within the playground these children tended to be loners or to play highly structured games. Any friendships they made were
with other children with similar difficulties.
In September 1989, numbers were large enough to form a viable class group. It was expected that this class would not develop a
group identity as such. It was thought that as all the children were highly egocentric, interaction would be minimal. However,
this is not the case; the children interact far more than they did within a class of other language-disordered children.
However, this interaction is not without problems. The peer interaction now taking place is often highly charged; no one makes
allowances for anyone else. The children cannot put themselves in the place of one another. They have problems making
friends, and friendships that are made are often difficult and short-lived. This is where the children's real and most pressing
problems lie, as is recognised by their parents and themselves: those of building and maintaining relationships with their peers.
Nevertheless it is within the group situation that most opportunities for useful teaching and therapy intervention are to be found.
Having the responsibility as class teacher for these children was a challenging and exciting opportunity and continues to be so.
Providing for

< previous page page_88 next page >

< previous page page_89 next page >
Page 89
their needs is a pioneering and ongoing learning process. Outlined below are some of the approaches and methods that have
been tried and have proved useful so far.

Classroom Organisation.
In the class group there are seven children who present semantic-pragmatic difficulties. They are all boys and their ages range
from nine to nearly 12. They are not a homogeneous group, as the degree and nature of their deficits vary. However, it is
possible to define some general aims and objectives that can be applied to the class as a whole. These are as follows:
Aims
1. To promote self-awareness and to enable the children to reach a greater understanding of themselves and of the world in
which they live.
2. To develop more socially and linguistically appropriate ways of interacting with their environment and the people within it.
Objectives
1. To enable the children to develop some understanding of themselves as a person, their feelings and reactions.
2. To increase the children's awareness of other people, their feelings and needs.
3. To encourage the children to see themselves as active rather than passive participants in the world in which they live.
4. To promote and encourage appropriate social interaction with peers and adults in a variety of situations and contexts (in terms
of both verbal and non-verbal communication).
5. To develop and expand basic concepts through concrete and practical activity and appropriate questioning.
6. To develop a better understanding of oral and written language.
7. To develop and use skills already more highly developed, such as reading and writing, to bridge deficits in concepts and
understanding.
8. To provide opportunities to develop creativity and self-expression.
The children's problems necessitate close liaison with all the adults working with them. Frequent liaison between the speech
therapist and teacher is vital, and regular meetings are necessary where problems can be discussed, ideas exchanged, plans
developed and goals set. Frequent and continual revision is necessary. As a consequence of this team approach a measure of
flexibility is incorporated. Individuals or small groups of children can be withdrawn by the speech therapist or teacher to work
within or outside the classroom. Class

< previous page page_89 next page >

< previous page page_90 next page >
Page 90
work and speech therapy are thus reinforced and developed alongside each other.
The children generally work in pairs or individually. As there are a range of activities going on most of the time, various areas
of the classroom have been roughly designated for specific activities: i.e. finding out, library and computer area, play area,
messy area and work area. Fortunately there is some land around the outside of the classroom where the children can grow
vegetables, plants and flowers, which is an important teaching resource.
Group sessions are incorporated into the timetable and include the speech therapist and teacher in one large group or two smaller
group activities. At these times social interaction in a more formal group situation is encouraged. There are some periods when
the children work in a class group: P.E., Swimming, Drama, Music and Story.

Curriculum
Class activity is based around a central topic; this is used to expand and develop the children's interests. The aim is to work
from the child and move outwards, extending the boundaries of rigidly defined concepts and clarifying confused knowledge and
information. Topics such as `Myself' and `Friends' have been chosen because they are most useful for developing awareness of
self and understanding of others. For all of the children concept formation is at a very basic level; language development has
out-stripped concept development, they have fluent speech and often use sophisticated vocabulary superficially and
inappropriately. For example, one ten-year-old when trying to resolve a dispute with another child said: `This is a very immature
situation'.
Ready access to school transport means that a child, a group or the whole class can experience real-life situations outside the
school. This provides an invaluable opportunity to reinforce and expand concepts and vocabulary as well as developing social
skills in a realistic context. Many of the children cannot easily imagine events beyond the `here and now' so that providing
experiences for them that are realistic and purposeful is very important. A great deal of practical and concrete experience is also
provided in school. It is through these activities that problems appear most clearly; these problems can then be worked upon
immediately in the class or in an individual session. The general aim is that the children plan, organise, carry out and evaluate
their own activities. This may be with the speech therapist, teacher or classroom assistant depending on the child and the
activity. The timetable is especially designed to allow for this kind of flexibility and is always open to revision.

< previous page page_90 next page >

< previous page page_91 next page >
Page 91
Each pupil has his or her own timetable; the composition of this depends on the aims and objectives for that child. The timetable
provides a familiar framework to the day but a balance has to be kept between this structure and giving enough flexibility to
follow up activities spontaneously.
The children work on their own topics. They all have favourite interests, which often amount to obsessions. Trains, cars, maps
and computers are popular. Within these areas they may have great depths of knowledge: for example, one child can list all the
stations between two given points on the London Underground. The aim of this approach is not to break obsessions but to
expand and broaden them, relating them to useful and basic concepts. One child had an obsessive interest in cats: by working
upon a `cat' topic, which culminated in conducting interviews with members of staff about their cats, the obsession now seems to
be more of an interest.
Many of these children have good measured skills of literacy. Spelling especially seems to be an area where many excel; for
example, one boy of 11 has a spelling age of over 15 years. They are often able to decode the written word into speech that is
fluent, expressive and containing the correct pronunciations and appropriate intonation patterns. However, it is in the area of
reading comprehension that most problems occur. There is usually a great discrepancy between reading decoding skill and
comprehension level. The children often have great misconceptions and gulfs in their understanding, although they may remain
unaware of this because they make sense of the text for themselves, but often this is not the meaning that the author intended.
There are problems at all levels of understanding: literal, deductive, inferential and predictive. The monitoring of reading
requires specific and perceptive questioning. Activities that are particularly useful are reading modelling activities, 1 reading
thinking activities2 and using playthings to manipulate in specific ways, according to written workcards.3 Some of the children
write their own comments on book review worksheets when they have finished reading a book; this involves further thought and
discussion at a whole book level.
A great deal of creative writing is done on the computer. Working in pairs on the computer or making a first draft in their
exercise books means that, with discussion, unnecessary details can be omitted and essential details added. There is a great
tendency for these children to go off at a tangent when writing, as with oral language. They can often already produce a number
of fluent and well-structured sentences. However, their work does not always meet the requirements of a specific set task, or
take any account of the reader. Developing appropriateness and understanding are the main aims of creative writing.

< previous page page_91 next page >

< previous page page_92 next page >
Page 92

Figure 6.1
Example of written work by a child with semantic-pragmatic
difficulties
Figure 6.1 is an example of a piece of written work produced by a child with semantic-pragmatic difficulties. This work was
produced as part of a `Holistic' Writing Assessment. (The original writing task and holistic scoring assessment was devised by
Greenhalgh & Townsend, 1981.) At Dawn House 21 children were asked to describe `How to build a swing'. They were limited
by the picture shown in Figure 6.2, which contains the items they could use. Some time was spent talking about the picture,
naming the tools and materials and discussing how they could be used. From the sample of 21 children, between the ages of
nine and 12, seven had previously been assessed as having semantic-pragmatic difficulties.
The scores achieved from an assessment of the written work produced showed no significant difference between the group with
semantic-pragmatic difficulties and the rest of the language-disordered population. Never the less the difference in the type of
work produced and the items attracting scores did highlight some interesting points. Only the semantic-pragmatic group used
unusual or inappropriate vocabulary items, for example `Get a hammer and fit the nails inside the rope'. Only one of the 21
children went completely off task and scored zero by describing how to build a remote control stick, and this was a child with
semantic-pragmatic difficulties. The semantic-pragmatic group had a larger number of inconsistencies, for example, changing
the

< previous page page_92 next page >

< previous page page_93 next page >
Page 93

Figure 6.2
How to build a swing
method of suspending the swing half-way through the description. None of the work produced by this group was sketchy or
skeletal but the extra details used did not add to the description; if extra details were included by the rest of the language-
disordered children these did add to the description and so increased their score. For example, `then you varnish the swing' was
used by two children without significant semantic-pragmatic difficulties, whereas a child in the semantic-pragmatic group added
`end of part one'.
Mathematics consists of a great deal of practical problem-solving activities such as weighing, measuring size and capacity, using
money and telling the time. These children are generally good at the mechanical skills of maths and can even solve written
problems, but difficulties arise when faced with more practical tasks and they have to learn strategies for problem solving. For
example, one child could measure anything in the classroom but when asked to decide whether a table would go through the
door had no strategy for attempting this task.
Science also provides many problem-solving activities, showing up considerable gaps in basic concepts and understanding that
can be bridged in a practical context. The children often work in pairs. The aims are to develop basic concepts and
understanding of cause and effect. Skills of organising and sequencing can also be practised at this time.
Art and craft is planned and initiated by the child as much as possible. The child chooses an activity and is responsible for
collecting the materials

< previous page page_93 next page >

< previous page page_94 next page >
Page 94
needed and completing it. When this type of work was begun, one boy was unable even to choose the colour of the paper he
wanted to use. This is an area where planning and organisational skills can be worked upon in a purposeful context.
Gardening is another subject that provides many opportunities for practising skills of planning, organising and sequencing. It is
organised so that the children work in pairs and decide together on their task and the equipment they need. Through negotiation
with their partner they carry out the activity. A gardening log book is used to provide a format. Planning and completing an
activity by themselves provides the children with an even more challenging situation. Worksheets in the form of flow diagrams
are used to aid planning and encourage self-evaluation when an activity is completed.
Cookery provides further challenging problem-solving opportunities. This is a real-life context for pre-task planning, useful
observation during the session and post-task evaluation. The children work in pairs. A discussion takes place the day before the
session and a simple recipe is chosen, then ingredients and utensils are collected. On the day of the cookery session another
discussion takes place during which cooking procedures are discussed with the children. They are asked to plan, organise and
carry out their own cooking task; they prepare and cook together, with an adult helping only if problems occur. The aims are to
develop basic concepts, expand vocabulary and encourage application of new knowledge. There are opportunities to work on
cause and effect and problem solving. Organising and planning skills can be developed. It is also a purposeful way of building
relationships, and of encouraging co-operation and sharing of ideas.
The children all enjoy music. Tunes can be remembered easily. The children also enjoy making music of their own and taping it.
This is an area worth exploring further in the future, as it creates opportunities for self-expression and creativity.
Most of the children are very fond of using the computer, almost to the point of obsession. Some seem to relate better to the
computer than to people. Many are more than usually distracted by the computer screen, when another child is working. The
computer is useful with this group for providing opportunities for discussion, planning and negotiation: this indicates that most
work should be in groups. The problem-solving, logical thinking and classifying type of programs are most useful. Word
processing and open-ended programs are also invaluable. Note, however, that programs which aim to practice skills individually
only reinforce egocentricity.
All of the children have poor motor skills and co-ordination difficulties, of varying degrees. The first half hour of each day is
taken up with activities

< previous page page_94 next page >

< previous page page_95 next page >
Page 95
aimed at developing these skills. This is based on the motor skills programme operating at Dawn House, devised by Elaine
Parkinson. 4 Pupils are screened on entry using an observational checklist. The Stott test (Stott, Moyes & Henderson, 1984) is
sometimes used for validation purposes. Those who are found to have problems are sent to the visiting physiotherapist for
diagnostic assessment. Sensori-motor skills work is carried out individually or in small groups, two or three times per week,
under the supervision of the classroom aides and the support teacher.
Play is an important part of the timetable. The development of basic concepts and understanding of themselves and others has
not been acquired by these children naturally through play in the past. A number of the children cannot play imaginatively and
this has to be facilitated and encouraged. The main aims of play are to encourage the children to develop more self-awareness,
to take the initiative and participate actively in a situation. Opportunities are provided through play for the children to learn to
be more flexible and to co-operate more fully with their peers. It also provides a medium for self-expression. These children see
the world acting upon them; they have not learnt that they can manipulate their own environment. Great fear or anxiety of new
situations is often shown and they feel a need to make sense of the world by constructing rules. Divergence from these rules is a
threat to their perceived image of the world. This is one area we hope to work upon through play.
Three types of play are available to the children; interactive or imaginative play, directed play and free play. Imaginative play is
encouraged by the teacher or therapist joining in a group play situation. Before these sessions a number of the children would
not attempt this sort of play. At first they used the play format or theme provided by the teacher again and again, for example `a
picnic'; now they are able to devise more ideas of their own and often carry this play over into playtime.
Directed play is made available to the children in a 45-minute session during the week. Play activities such as dressing up,
construction, art and craft and playground games are provided. The children work in pairs and move from one activity to
another.
A `free activity' session is one where any activity in the classroom can be chosen. At first the children chose to carry on with
written work, as they found the freedom frightening. Now they feel more confident and some are able to discover new play
activities for themselves.
Non-directive play therapy is also used for some children. It has been found to be most successful with those children who
exhibit high levels of

< previous page page_95 next page >

< previous page page_96 next page >
Page 96
anxiety. The child is invited to `make a world' using a sand tray and a variety of miniature toys adapted from Lowenfeld (1979).
This enables the child to express a personal world in a representational form. The adult does not intervene, interpret or analyse
but facilitates and encourages. The purpose of this type of play is to create opportunities for self-expression and personal
development.
The class also has a weekly drama session. The aims of this are to increase the children's understanding of social conventions
and rules, to develop non-verbal skills and gesture and increase awareness of different styles of language which are appropriate
to different social situations. Creativity and self-expression are also encouraged. A typical drama session may consist of a simple
circle game followed by some role play, which is modelled initially by the adults. An example of this would be the acting-out of
appropriate and inappropriate non-verbal behaviours. Next the children participate in the role play. The children all have
problems putting themselves in the place of others and they found this very difficult at first. However, by choosing everyday
situations that are relevant to their needs, such as taking messages, and structuring the steps involved, the children are now
developing these skills. There is also some evidence of carry over into situations outside the classroom.

Management
To promote effective behaviour management, the adults working with these children met and developed some basic principles
concerning interaction with the group as a whole. These general principles are listed below.
1. Always try to give the children a choice and make them choose.
2. Listen to them and take time.
3. Stand back. Don't do things for them or solve problems for them.
4. Don't answer for them unless they ask you a question directly.
5. Ask their questions back to them, or ask `What do you think?'
6. Listen to their ideas, accept and perhaps develop them, connecting them to basic concepts.
7. When talking to the children don't overload or confuse by encouraging them to follow obscure or highly specific interests.
Instead relate their thoughts to more obvious general and basic concepts.
8. STAY CALM.
The aims, objectives and general principles already stated have many implications for management in the classroom.
Opportunities for play and experiment must be provided and ritualistic play must be discouraged in favour of more purposeful
activity. The organisation of this needs careful planning. If the children are to be encouraged to take the initiative the adult

< previous page page_96 next page >

< previous page page_97 next page >
Page 97
must stand back, observe, encourage, guide and question. Rushing in and helping the child is not always useful. The children
learn best through solving their own problems with guidance. Often these children have learnt that if they do or say nothing
eventually they will be given the answer.
Rules are necessary in any classroom situation. These children enjoy having rules to follow and will obey them to the letter once
imposed. They appear to need the security of a familiar environment and structured day. They must be encouraged to cope with
incidents out of the usual routine and therefore they must be encouraged to be flexible, which they find very hard. Tantrums are
frequent as emotional and social development lags behind other abilities. The children take a form of `Time Out' if they cannot
control their temper; this consists of walking or running around the outside of the building or standing outside the classroom or
in the next room. Problems usually occur because the rules these children have constructed for themselves seem to have been
broken. They cannot step back and look at a problem in a new way. Their egocentricity means that they may not understand the
difference between an adult talking to the group or talking to them individually.
Discussion and acting out of problems seems to be possible sometimes as a means of defusing conflicts. At the right moment
group or individual counselling sessions have helped to resolve problems. Some of the children are able to use techniques of
relaxation therapy but others are not able to do so. Some of the children may need to be more assertive with their peers, others
less aggressive. This is another area for work. One child has made a contract with the teacher guaranteeing his behaviour and
agreeing to be deprived of privileges if he breaks the contract. Developing an awareness of their own difficulties is the main aim.
In this way it is hoped that the children will begin to see their own behaviour problems and work with a teacher or therapist to
overcome them.
Each child is an individual and it is only by getting to know each one that the most effective methods can be used. Perhaps by
helping them to develop more understanding of themselves, enabling them to see themselves as active participants in the world
around them, progress can be made.

Assessment
It is extremely difficult to assess the problems of children with semantic-pragmatic disorders as standardised tests do not show
these very clearly. Abilities such as self-awareness, creativity and understanding of others are extremely difficult to measure
objectively. Often, apparent facility with language can mask huge gaps in understanding. For example, one child can give many
definitions of a difficult word without any real understanding of the

< previous page page_97 next page >

< previous page page_98 next page >
Page 98
concept behind it. Problem-solving tasks are more useful than table-top tasks for pinpointing difficulties; the practical problems
created by everyday real-life situations are the most useful.
Having a tape recorder running continuously during class is helpful, as is having a notebook to hand, so that anything interesting
can be recorded. Incidental comments often show up whole areas of misunderstanding, as the following example illustrates. One
day the class was planting bluebell bulbs, and a discussion took place about what they would grow into. The next day a dead
mouse was found in the school pond. After burying the creature one child remarked, `but what will it grow into?'. This comment
would not be inappropriate for a much younger child, but seems very odd when produced by a ten-year-old with good fluent
speech.

Future Implications.
A discussion of curriculum at this time cannot be complete without considering the introduction of the National Curriculum. The
problem-solving approach adopted by the Craft, Design and Technology and the Science curriculum documents is particularly
appropriate for these children. The danger is that the pressure of covering all core and foundation subjects may lead to the
adoption of an `acquisition of knowledge' model of learning. These children are generally good at acquiring knowledge and
language at a superficial level but this often masks their lack of fundamental understanding. Unwittingly, knowledge and
language could be taught without ensuring that the foundations of basic concepts have been laid.

Notes
1. Modelling activities are described in Developing Independence in Reading, Open University Reading Development Course
PE231, Block 2, Units 5, 6, 7 and 8 (1977).
2. Reading thinking activities are described in D. Moyle (1982), Children's Words (Grant McIntyre).
3. Manipulating play things to increase understanding of written text is described by Eva Grauberg in a meeting of the National
Association of Professionals working with Language Impaired Children (NAPLIC), June 1989.
4. Parkinson, E. 1990, A critical evaluation of some aspects of sensory motor skills programme. Unpublished MEd thesis,
Sheffield Polytechnic.

References
GREENHALGH, C. and TOWNSEND, D. 1981, Evaluation of students writing holistically an alternative approach. Language
Arts 58, 8 (October).
LOWENFELD, M. 1979, The World Technique. London: Allen and Unwin.
STOTT, D.H., MOYES, F.A. and HENDERSON, S.E. 1984, Test of Motor Impairment (Henderson Revision). Ontario: Brook
Educational Publishing.

< previous page page_98 next page >

< previous page page_99 next page >
Page 99

Index
Note: Page references in italics indicate tables and figures

A
Aarons, Maureen 21
Acquisition, vocabulary 46, 48, 49, 81
Adams, C. 8, 68-74
Adams, C. & Bishop, D.V.M. 13, 65, 69
Appropriacy 11, 13-14, 19-20, 30, 66
and autism 73
and semantic-pragmatic disorder/syndrome 69-71, 72, 76
Art and craft, and semantic-pragmatic disorder/syndrome children 93-4
Asperger's syndrome 75
Assessment 11-14, 38, 69, 76-8, 97-8
context 12-13, 21
Attention-directing 29
Attention-getting 29-30
Autism,
and attention-getting 29
and conversational ability 27
and inference of mental states 36
and language development 8
and literal/indirect meaning 30-1
pragmatic difficulties 9, 37
and semantic difficulties 3, 9
and semantic-pragmatic disorder/syndrome 11, 73-4, 75
and social knowledge 30
and turn-taking 29, 37
Auxiliaries, omission 25

B
Back channel response 26
Baker, L. & Cantwell, D.P. 37
Baltaxe, C.A.M. 29, 37
Baltaxe, C.A.M. & Simmons, J.Q. 37
Bates, E. 4
Bishop, D. 10
Bishop, D. & Rosenbloom, L. 9, 22
Bishop, D.V.M. & Adams, C. 13, 14, 65-6, 69-71, 72, 73, 76
Blank, M., Gessner, M. & Esposito, A. 61
Bloom, L. 4
Bristol Language Scales 12
British Picture Vocabulary Scales 77

C.
 Camarata, S., Hughes, C. & Ruhl, K. 37-8
Cantwell, D.P. & Baker, L. 37
Categorisation tasks 82
Causal relations, problems with 20, 25, 33
Circumlocution 25
Clarification, in conversation 6, 24, 27-8, 31, 63-4, 66
Closure, grammatic 77
Co-ordination problems 94-5
Cocktail Party syndrome 9, 75
Competence,
conversational 23, 28-31, 34-6
syntactic 28
Comprehension,
and conversation 26, 30-1, 33, 61-2, 76
reading 91
Computer, and semantic-pragmatic disorder/syndrome children 94
Concept boundaries,
expansion 3, 81-2, 90
rigid 15, 25, 76, 78
Context, language 4-6, 12-13
Conti-Ramsden, G. 7, 59-66
Conti-Ramsden, G. & Gunn, M. 61-2, 77
Conversation,
adjustments 27, 30
and assessment 12-13
breakdown 14, 18-19, 66, 70, 76
cohesion 65, 70, 72
definition 5
disability 10, 14, 19-21
and affective/emotional difficulties 24, 36-8
and cognitive deficit 23, 32-4
and conversational rules 28-31
and linguistic impairment 24-8

< previous page page_99 next page >

< previous page page_100 next page >
Page 100
nature 22-3
and socio-cognitive difficulties 23-4, 34-6
types 23-38
exchange structure 70
initiation 26, 29-30, 59, 65, 70, 72
repair 14, 18-19, 65-6, 70
skills 6
see also turn-taking
Conversation analysis 69-74
Cookery, and semantic-pragmatic disorder/syndrome children 94
Craig, H.K. & Evans, J.L. 26, 29
Cray, Barbara 88-98
Crystal, David 4, 10, 14, 43-57
Cue,
linguistic 24
paralinguistic 6
recognising 84
Curriculum,
semantic 56-7
for semantic-pragmatic disorder/syndrome children 88, 90-6, 98

D
Dawn House School 15, 75, 76-98
Definitions, learning 55-6
Deixis 3, 15
Determiners, omission 25-6
Dewart, H. & Summers, S. 12
Dictionary, as teaching tool 51-5
Discourse, definition 4-5
Donahue, M. 28, 34-5
Down's syndrome 8
Drama, and semantic-pragmatic disorder/syndrome children 96
Dyslexia, developmental 15

E
Echolalia 22, 68
Edinburgh Articulation Test 77
Eye contact, in conversation 29

F
Field, semantic 46-8, 50-4, 56-7, 81
Fluency 22, 68
Fluent language disorder 75
Fragile X syndrome 8, 10
Frequency, vocabulary 45-6
Friendships, and conversational skills 39
G
Games, therapeutic use 82, 84, 86
Gardening, and semantic-pragmatic disorder/syndrome children 94
Group comparison studies 65-6
Group design studies 24, 63-5

H
Hyde-Wright, Susan 75-87

I.
Identification 11-14
Illinois Test of Psycholinguistic Ability 77
Inference 32, 33-4, 85-7
social 34-6, 62
Interaction,
adult-child 5, 26-7, 38
child-child 88, 89
Invalid Children's Aid Nationwide (ICAN) 75

K
Knowledge, shared 22, 27
Knowledge, social, in conversation 30, 32-4
Kysela, G.M. et al. 29

L
Language, obsessive 73
Language Assessment, Remediation and Screening Procedure 77
Language disability,
classification 8-11
disciplines involved 2
linguistic approach 9-10, 73
medical approach 8, 9-10, 73
Language impairment, specific 7, 8, 9, 10, 15
Lexicon 2-3, see also vocabulary
Lexicon (Longman), as teaching tool 51-5, 52
Lists, therapeutic use 82
Literacy, and semantic-pragmatic disorder/syndrome 91
Lowenfeld, M. 96

M
McTear, Michael 10, 12-14, 18-39, 61, 69, 82-3
Mathematics, and semantic-pragmatic disorder/syndrome children 93
Matrices, therapeutic use 82
Mean Length of Utterance 28, 64
Meaning,
and context 5-6
literal/intended 6, 19, 22, 25, 30-1, 86
and semantic relations 3-4
and semantic-pragmatic disorder/
< previous page page_101 next page >
Page 101
syndrome 15
Memory, auditory 77
Mental states,
first-order 35, 36
second-order 35-6
Milroy, L. 13
Mogford, K. & Bishop, D. 2
Motor skills 94-5
Music, and semantic-pragmatic disorder/syndrome children 94

N
Naming, ostensive 47, 54, 62
National Curriculum 98
Nelson, K. 32
Nelson, K. & Gruendel, J.M. 32-3

O
Obsessions, dealing with 91
Olson, D.R. & Nickerson, N.G. 32
Over-elaboration 71
Overextension 2-3, 46, 56

P.
Parents, keeping vocabulary diaries 43-4
Parkinson, Elaine 95
Peers, interaction with 39
Perner, J. & Wimmer, H. 35-6
Phonological-syntactic syndrome 10, 22
Phonology, disabilities 6
Play, and semantic-pragmatic disorder/syndrome children 95-6
Pragmatics,
definition 4-7, 14, 22
difficulties 1-2, 7, 9-10, 37
assessment 11-13, 14
case studies 61-2, 65
definition 60-1
studying 59-66
and syntactic disabilities 7
Prepositions, spatial 25
Prism 12
Problem-solving approach 78-81, 93-4, 96-8
Pronouns, incorrect use 3, 20, 70
Protocols, pragmatic 65
Prutting, C.A. 4, 12
Prutting, C.A. & Kirchner, D.M. 65, 72

Q
Questions,
inappropriate 71, 83, 84
response to 19-20

R
Rapin, I. & Allen, D. 10, 22, 68-9, 72-3, 75
Reading, semantic-pragmatic disorder/syndrome children 91
Reality, and language 76
Reasoning, unusual 20-1, 22, 82-3
Reiteration 71, 73
Relative clause, use of 24, 28
Relaxation therapy 97
Requests,
emergence 62
polite 24-5, 28
Role play 77, 96

S
Schank, R.C. & Abelson, R.F. 32
Science, and semantic-pragmatic disorder/syndrome children 93
Scripts 32-3, 68, 73
Semantic-pragmatic disorder/syndrome 7-8, 15, 22
assessment 76-97
behaviour management 96-7
characteristics 68-9, 72, 76
classroom organisation 89-90
conversation analysis 69-74
and conversational style 72
curriculum 90-6
intervention 78-87
and language impairment 72-3
origins of term 10-11, 68, 75
recognition of 10-11
Semantics,
definition 2, 14, 22
expressive 71
and language disability 1-2
lexical 2-3, 15
semantic difficulties 7, 9-10, 14-15
semantic relations 3-4, 7
Silliman, E. 25
Simplification, by adults 5
Smedley, M. 25
Speech acts 5-6
restricted range 26
Speech therapist, role 75-87, 89-90
Story-telling,
and turn-taking 84
 and vocabulary teaching 53-4
Stott test 95
Stubbs, M. 20
Style, conversational 13, 71, 72
Subject, omission 25-6

< previous page page_101 next page >

< previous page page_102
Page 102
Syntax,
and conversational skills 27-8
disabilities 6
and pragmatic disability 7

T
Teacher, role 81, 88-98
Temporal relations, problems with 20, 25
Terminology 1-2, 14-15, 75
Tests, standardised 12
Theory of mind 35
Therapy, and communicative needs 78-83
Topic,
changing 6, 18, 71
maintenance 6, 68, 73
Topic teaching 90-1
Treatment 15
Turn-taking 6, 18, 23, 26, 29, 39, 66
and semantic-pragmatic disorder/syndrome 68, 70, 72, 83-4

U
Underextension 2-3

V.
Verbosity 68, 70
Vocabulary,
child use 2-3, 23
core 81
diary 43-4
extending 48-56, 78, 90
quality 45-8
size 43-5
teaching 43-57

W
Wagner, K.R. 44, 45
Wells, G. 14
Wimmer, H. & Perner, J. 35
WISC-R Verbal Comprehension 69
Word order, and meaning 3-4
Word-finding difficulties 15, 25
Words,
definition 55-6
linguistic context 54, 56
sense-relations 47-8
Writing, creative 91-3, 92