Understanding Child Development

Understanding Child Development
Psychology Editor Brian Foss

Professor of Psychology, Royal Holloway College, University of London

Psychological perspectives in an
interdisciplinary field of inquiry
Sara Meadows
First published in 1986 by Unwin Hyman Ltd

Third impression 1989
Simultaneously published in the USA and Canada
by Routledge
29 West 35th Street, New York, NY 10001
This edition published in the Taylor & Francis e-Library, 2001.
Sara Meadows, 1986
All rights reserved. No part of this book may be reprinted or reproduced or utilized in any form or by any electronic,
mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any
information storage or retrieval system, without permission in writing from the publishers.
British Library Cataloguing in Publication Data
A catalogue record for this book is available from the British Library.
ISBN 0415084369 (Print Edition)
ISBN 0-203-13834-1 Master e-book ISBN
ISBN 0-203-17792-4 (Glassbook Format)
Contents
Acknowledgements
Preface
Note on reading
11
Note on pictures
11
Introducing the study of child

development
Conceptual issues; The ecology of child
development; Play
13
Perceiving and understanding

Cognitive development; Piagetian
theory; Logic as a model for cognitive
development; Information-processing
approaches; Infants perception and
cognition; Brain organization and
development; Cognition in infancy
33
The development of cognitive skills

Attention; Remembering; Reading;
Skills of learning from text; Writing;
Arithmetical skills; Metacognition;
Social cognition
69
Language development
117
Language; Methodology; Infants
perception of speech sounds; Infants
production of speech sounds; Beginning
to use words; Using words as names;
Early vocabulary; From single words to
sentences; Adults talk to children;

Language differences and social class;
Language at home and at school; The
functions of language; Childrens
metalinguistic behaviour; Word
meanings and concepts; Early concepts;
Literacy, language and thinking
5 Personality
Physiology and personality;
Temperament; Stage models of
personality; Self-concept development;
Beliefs about control of events; Selfefficacy; Aggression; The development
of pro-social behaviour; Moral
development
143
6 Social relationships
173
Infantadult interaction; Describing
relationships; Parentchild relationships;
Early experience and later effects;
Sibling relationships; Peers; Sex roles,
sex differences, sex typing; Learning the
social world of school; Childrens
independent social worlds; Adolescence;
Delinquency; Overview
Further reading
209
Bibliography
212
Author index
238
Subject index
245
Acknowledgements
The author and publishers would like to thank the copyright

holders below for permission to reproduce the following
material:
Figure 2 from Advances in Child Development and
Behaviour, vol. 17, 1982, reproduced by permission of
Academic Press; Figures 3, 4, 5 and 7 from M. M. Haith
and J. J. Campos (eds.), vol. 2 of the Handbook of Child
Development, 1983, reproduced by permission of John
Wiley & Sons, Inc. Publishers; Figure 9 from R. Kail,
The development of memory in children, 1979, Figure 13
from R. Mayer, Mathematical ability, in R. J. Sternberg
(ed.), Human abilities: an information-processing
approach, 1985, both reproduced by permission of W. H.
Freeman; Figure 11, Table 3 and accompanying text,
and Figure 12 from B. Kroll and G. Wells (eds.),
Explorations in the development of writing, 1983,

reproduced by permission of John Wiley and Sons
Limited; Figure 14 1972 by the American
Psychological Association, reprinted by permission of
the authors; Figure 15 and Table 4 by permission of
British Psychological Society and the authors; Figure 17
1980 by the American Psychological Association,
reprinted by permission of the author; Figures 18 and 19
by permission of The Controller of Her Majestys
Stationery Office; Longman Group Ltd for pages 155 6
of Houlbrooke, The English Family 14501700; Plate 1
Bodleian Library, Oxford; Plate 2 Henri CartierBresson; Plates 3, 4, 6, 7 and 15 G. A. Clark; Plates 8
and 11 John Bignell; Plate 9 by courtesy of the Board
of Trustees of the Victoria and Albert Museum; and
Plate 12 O. P. Marzaroli.
Preface
My concern in this book is to discuss the course of

child development. I regard understanding child
development as uniquely important both for its
practical implications for minimizing unhappiness
and maximizing goodness and fulfilment, and for its
intrinsic intellectual interest. All of us have been
children and most of us will be parents:
understanding child development may ameliorate the
human condition. All of us know a lot about child
development; few of us could make exact statements
about why a particular child or children in general
should have developed in this familiar or that
unfamiliar way: understanding child development is
a riveting intellectual problem. I shall not propose
either child-rearing panaceas or stunning new basic
theories. My more modest aim is to look at work
which is either current or fairly recent and some older
very influential work; and to consider how it may
begin to be fitted together into a good theoretical
flamework. There will be a lot of pointing out what
we dont know, but also, I hope, some pointing out
exciting new questions, answers and ways of
working.
I am a psychologist whose recent teaching has been
work with experienced school teachers doing higher
degrees that included courses on developmental
psychology. My research career began with an
esoteric theoretical problem studied using
experiment and factor analysis, and moved to work
which was much more concerned with a tangle of
real life issues and with the methodological
problems involved in undertaking rigorous but
uninterrupted measurement and observation of the
life of the child. These experiences have shaped my
view of what the psychology of child development is
and could be, and hence the approach I have taken to
writing this book. It is about my current
understanding of what child development is and why
it happens as it does: it is not an encyclopaedia of
experiments or an instruction manual of child-rearing

techniques. I have tried to give references to the
research which substantiate my assertions so that
readers can evaluate them for themselves, as
limitations of space, and the need to keep the
argument going, made it impossible to give much
more detail on the material I have used. Readers may
draw recommendations from these pages about what
to do and what to avoid when dealing with children: I
have not intended that they should do so, and I take no
responsibility for the results! Nor do I claim that my
current understanding of child development is
correct, complete, or what I will myself believe
forever. I am merely presenting a frame of reference
which I think is worth trying, and material to fill it out
which seems to make a useful fit. I am not presenting
a completely worked out and documented theory.
It is not a new theoretical frame, nor is there much
dramatic new evidence. Experts in any of the fields I
discuss will find nothing in my discussion that they
have not known about for a long time, and may, I fear,
be exasperated by my simplifications and omissions.
I hope that they will accept my account as an
approximation to what the non-expert might usefully
think about, before and during the reading of the more
detailed and advanced material I have tried to cite. I
have tried to bring together areas that have been
worked on separately because I believe we are far
enough advanced in our understanding of child
development to try out what an overall picture might
be like. I also believe that such a trying-out may, even
if no coherent picture emerges, provide some insights
drawn from areas outside the experts normal range of
reference which will enrich further work in the area of
detailed study. There are signs of this happening, for
example in growing links between areas of
developmental psychology, and in two-way traffic
between researchers working on adults and
researchers working on children.
9
Prefaces are places for stating why one has done

what one has, and also for acknowledging the help
one has had in doing it. I want first to offer my general
thanks to the international community of
developmental psychologists who, although they
produce far more than I can keep up with, even
allowing for the repetitive, the bad and the boring,
have generated enough ideas and put forward
sufficient data to fill my professional life and, more
importantly, to move us further towards
understanding child development. Nearer home, I
owe an enormous amount to talking with friends,
colleagues, teachers and students over the last
eighteen years, and I am most grateful for this. While
writing this book, I was most generously provided
with unpublished material by members of the
Thomas Coram Research Unit, the Child Health and
Education Study (Director Professor Neville Butler),
the MRC National Survey of Health and
Development, the South London Under Fives
Project, and others. I used libraries in London and in
Bristol which may have by their excellence misled
me about the availability of material: the librarians of
the School of Education of the University of Bristol
were particularly helpful and clever despite being
seriously overworked; they were paragons of
patience and efficiency and I have learned a great deal
10
from them. My secretary, Maureen Harvey, shares all

these virtues, and deserves the highest praise. Among
the friends who read and criticized parts of the text
were Elizabeth Robinson, Maggie Mills, Sandy
Acker, Peter Robinson, Brian Foss, Claire LEnfant,
Steve Whittaker, John Conroy, and John Cowley:
Philip Meadows helped with indexing. Mike Smith
of the History of Art Department, University of
Bristol, and the staff of the Victoria and Albert
Museum, the Ashmolean, and the British Museum
helped with the pictures. My thanks to them all.
When I was 9, my best friend in the junior school
and I ganged up on another little girl, and if I
remember correctly I hit her, though I hope not with
the hammer I think I remember. Certainly she then
had to be placated so that we should not be reported to
the teacher, and in the course of doing this I used as a
bribe the promise to dedicate a book to her when I
grew up and was an author. While this is not the sort
of book we then expected, it has benefited from my
experience of children from my own childhood on,
and I do therefore feel it is appropriate to redeem my
promise now. I was wrong to make it, but it is right to
acknowledge the contribution of the young Margaret
and the young Sara and many half-forgotten others to
whatever progress I have made towards
understanding child development.
Note on reading
Note on pictures
Although I have tried to make this text intelligible and

sound enough to do alone for readers who do not
wish to take things further, I am very anxious that
readers should go on from this book to the more
detailed information that lies behind what I have said
here. I have therefore tried to give sufficient reference
for this to be done; as far as possible I have referred to
review papers which are worth reading in themselves
and also provide citations of further sources.
Limitations of space in many cases preclude citing
the original research papers: I hope readers will track
these down through the review sources given.
Among the criteria which guided me in choosing
references were their importance, intelligibility,
accessibility (a serious problem as libraries costs rise
and purchases fall) and how well they represented
their field. A few non-psychology books are
included, mainly because they provide insights and
enjoyment complementary to that gained from the
standard works.
The pictures in this book are there for three reasons.

First, to provide agreeable resting places in the text.
Second, because each can be related to one or more
important issues in developmental psychology (for
example, the Cartier-Bresson photograph of children
in Seville reflects the relation of play to the
ecosystem, in this case showing enjoyable play and
classic play faces in a city ravaged by civil war).
Third, each has aesthetic merit or historic interest,
possibly on a minor scale.
11
Plate 1 An illuminated diagram from a Natural Science textbook compiled by the monk Byrhtferth at Ramsey Abbey
between about 1080 and 1090. It is used as illustration here because in its picture of the place of Man in the universe it links
the four ages (boyhood and infancy, adolescence, young manhood, and old age) to the influence of the signs of the Zodiac,
the four seasons, the four points of the compass, the four elements of earth, air, fire and water, and the four humours (hot,
cold, wet and dry). It might thus be seen as an early model of the developmental psychology of personality. Childhood and
infancy (bottom left dark circle), ages 014, lie between air and fire, west and north, Capricorn and Pisces, and are wet and
hot; adolescence (top left), ages 1428, between fire and earth, north and east, Aries and Gemini, and is hot and dry. (Young
manhood lasts, readers will be glad to learn, until 48, and old age to 70 or 80.) Which of the four humours dominated the
body had implications for medical treatment and for personality (see Chapter 5), and may have influenced concepts of
education and child-rearing.
12
1 Introducing the study of child development
Conceptual issues
Before we embark on either description or
explanation of the course of child development, let
alone before we consider what efforts we might make
to change its course, I must stress that neither child
nor development could be said to be simple
unproblematic concepts; in particular they are
inextricable from beliefs about how to bring up
children. There are many variations between cultures
on what they believe children naturally are and how
they should behave (Laboratory of Comparative
Human Cognition 1983). There are also historical
changes within societies. The study of the history of
western childhood is only just beginning and its
picture is controversial. Philippe Aris pioneering
study (Aris 1962) argued that strong concern and
affection for children, and a belief that childhood was
an intrinsically valuable period, were historically
recent developments associated with the rise of the
affluent household in the sixteenth and seventeenth
centuries. Previously, he claimed, there was no
concept of childhood: children were regarded with
indifference by their parents or as inferior miniature
adults to be strictly reared and severely punished. De
Mause (1976) put forward an even blacker model of
maltreatment and cruelty to children infanticide,
beatings, sexual abuse and a casual acceptance of
high mortality through infection, accident or
childrearing practices such as wet-nursing or using
opiates to quieten a crying child. Aris and De Mause
have both been accused of selecting their data without
much concern for their representativeness, and of
interpreting dubious facts in unjustified ways
* Full references quoted in text appear in the Bibliography on

pp. 21237.
(Pollock 1983; Houlbrooke 1984). As more evidence

is examined, the picture of what was happening to the
crowds and crowds of little children [who] are
strangely absent from the written record (Laslett
1971) becomes clearer and more complex.
Houlbrooke (1984, p. 1556) summarizes the
relationships between parents and young children (in
the English family between 1450 and 1700) in a
picture which resembles what emerges from the
mainly nineteenth century diaries reviewed by
Pollock (1983) and the autobiographies from a later
period collected by Burnett (1982).
The Middle Ages passed on an ideal picture of the
relationship between parents and their offspring. Children
were welcome gifts from God. Parental love was the most
deeply rooted of all human instincts, and showed itself
especially in the mothers tender and loving care of the
helpless baby. But it was futile to grieve in the face of infant
deaths. Rather should the truly Christian mother be glad
that God had taken her child to himself. The child which
died before seven, the age of reason, was normally
incapable of mortal sin and assured of paradise. Yet the
child inherited original sin, which began to manifest itself
even in infantile greed and rage, and the inborn propensity
grew with intelligence. So it was the duty of the solicitous
parent to correct his child with the rod. Successful
parenthood largely depended upon the ability to curb ones
natural inclinations, and those who did not want their
children to become soft, truculent and good-for-nothing
were wise to put them out in late childhood to undergo
further discipline and training at the hands of strangers.
Much of the inherited picture survived throughout this
period, but it changed in important respects. The Protestant
reformers took over an emphasis on original sin and the
parents corrective responsibility. They also laid greater
emphasis on the parents part in early and thorough
religious education. The humanists, on the other hand,
13

believing as they did in the childs capacity for good and the
moral neutrality of its impulses, sought to protect
innocence and prevent deterioration rather than to correct
inborn vices. To show natural parental affection seemed
less dangerous than hitherto. Humanist readiness to give
expression to natural pleasures and sorrows made more
acceptable the display of delight in children in letters and of
a qualified grief at their loss in memorial inscriptions. From
the sixteenth century onwards there was a greater variety of
patterns of upbringing. Yet the fusion of Christian and
humanist ideas encouraged a middle way in upbringing,
between the extremes of severity and indulgence. It
enhanced the complexity of the task of nurture, and
consequently of fine judgement on the part of those who
undertook it.
Actual experience and practice often stood in contrast
with ideals. Children were not always welcome. Some birth
control was practised, and the danger of infanticide was
widely recognized. Yet despite the ease with which
infanticidal practices could be masked within families,
parish registers suggest that infant and child mortality were
relatively low and very largely explicable in terms of
environment and disease. There is much direct evidence of
the reality of loving care in some families and of parental
grief in face of the loss of children. Women of the upper
classes did not generally suckle their own children, though
humanist and Protestant propaganda may have persuaded a
minority to do so. But this did not preclude care and
solicitude in the choice and supervision of wet-nurses.
Differences between the life patterns of socio-economic
groups had many other effects on the childs prospects. The
offspring of the urban poor, rapidly growing in numbers
during this period, always had the poorest chances of
surviving infancy. It was the poorer groups in society who
had to exploit their childrens labour in the struggle for
survival at the earliest point. Only fathers and mothers who
enjoyed the advantages conferred by economic means,
education and a certain amount of leisure were able to
approach the ideal of intensive parenthood set out in the
literature of counsel, carefully polishing their childrens
manners, inculcating the principles of religion and laying
the foundations of literacy. Such instruction for life as most
children had from their parents was probably gained in byre
and field, at spinning-wheel and oven. But the quality of
parenthood was not of course determined by material
14
circumstances alone, very important though these were.

The unquantifiable and still only partially understood
elements of individual character were crucial in this period
as they still are today (Houlbrooke 1984, pp. 1556).
The relationship between what is believed about

children, what is prescribed as appropriate childrearing, and what is actually done, is not clear.
However different basic theories carry different
practical implications. For example, if children are
seen as firmly predestined from birth (because of
innate qualities such as different levels of
intelligence) to be intellectuals or technicians or
general workers, the most efficient way of educating
them may be to provide separate training in the
appropriate roles. Such a system was recommended
by Plato in The Republic and might be identified in
the tripartite secondary school system set up in
Britain under the 1944 Education Act. The research
problems will centre on how to make the initial
diagnosis of the childs nature (various
psychometricians Burt, for example were
involved in this) and on how to run the education. If,
on the other hand, children are seen as changed and
shaped by their experience, with little or nothing in
the way of predestined characteristics, initial
diagnosis is irrelevant, and child-rearing and
education are just a matter of providing the right
experience. Advocates of this view judge the
rightness of experience according to a variety of
criteria: for Skinner in Walden Two and for the
inhabitants of Huxleys Brave New World, the main
criterion is fitting happily into and serving society,
and a major means to this end is pervasive and skilful
conditioning. Huxley, unlike Skinner hostile to
Utopias, incorporates in his techniques of childrearing and social control elements of genetic
selection (and genetic engineering) and chemical
control of development and behaviour both before
and after birth: he also suggests that for the highly
creative and innovative intellectual even this allenveloping system might not work. Skinner has not
seemed to be troubled by such libertarian qualms,
arguing that an effective system which would make
Introducing the study of child development
all development smooth and happy would be

preferable to the ineffective conflictfilled systems
and the unhappy results which he believes exist now.
Similar disagreements exist about the nature of
human nature. Assuming that the infant is nearer to
what is natural than the adult (a dangerous
assumption which I shall seek to question later),
infant nature has been said to be innately good,
better than adults. Rousseau argued for an education
of maximum freedom so that the childs innate
goodness should not be spoiled or his creativity,
spontaneity and ability to love curtailed. I say his
deliberately, since Rousseau saw women as inferior
and fit only to be trained to serve males: the history of
his own children is unclear, but he claimed in his
autobiography to have sent them as babies to
orphanages. Despite this personal bad example, the
progressive educational movement took up
Rousseaus ideas and this model of the child has
dominated early childhood education. Noninterventionist ideas may also be seen in Piagets
accounts of learning, teaching and cognitive
development.
An alternative view of child nature was the older
one of natural badness, Original Sin, unsocialized
egocentricity, etc. This view emphasized childrens
unpleasant characteristics and prescribed strict,
punitive and intrusive childrearing and education:
these took some extraordinarily harsh forms (see de
Mause 1976; but also Pollock 1983). It is hard to find
a contemporary example, though Sir Truby Kings
prescriptions in the 1930s of conditioning babies to a
rigid four-hour feeding schedule perhaps comes near.
Many psychologists have had in mind, however,
views of children as inferior to adults, being, for
example, egocentric, dominated by animal instincts,
irrational and so forth, and needing to grow out of, or
be trained out of, these undesirable faults.
The biologicaland social causes
of development
Underlying all the debate outlined so far are a knot of

difficult issues. The core one is the question of the
relationship between biological and social factors
in human life, especially human development, since

development is often said to be about how the
biological infant turns into the social adult. How
to think of this crucial question is a very complex
problem, bedevilled by the tradition of separating and
opposing biology and society, heredity and
environment, nature and nurture. Once
opposed, one pole is valued highly and the other
denigrated, and, since neither can alone provide a
satisfactory explanation of all human development,
there is a history of see-sawings between polar
extremes. This is all the more unsatisfactory because
biological and social are not at all clearly
separable. Biological facts, such as the physical
consequences of possessing a functioning Y
chromosome, are acted on by society, which
classifies its members as male or female. Social
preferences have always been among the forces
admitted as working for natural selection; social
changes, such as industrialization, have biological
consequences, such as changes in what causes illness
or death. Even at the level of the genes, exactly how
their instructions work may be strongly influenced by
the environment. Hofer (1981) gives an example:
the amount of dark fur on the feet and nose of Siamese cats
depends on the ambient temperature in which they were
reared as kittens, the skin on the extremities being normally
cooler than other skin areas. The expression of this genetic
predisposition depends on the temperature of the skin
during a critical period of postnatal development. Raised in
an incubator, Siamese cats turn out uniformly light, and if
in an icebox, uniformly dark (Hofer 1981, p. 10).
Thus even if there is a genetic difference between

two individuals, exactly what its influence is will
depend on the environment. If every relevant aspect
of their environment has been identical, differences
between them can be attributed to how their genes
expressed themselves in that environment. Reared in
a different environment, again exactly the same for
both individuals, the genetic difference may express
itself differently and the differences between the two
individuals behaviour be completely changed.
15
Imagine, for example, that one of the two has genes

which predispose that individual to highly aggressive
behaviour and the other does not. Both are reared in a
society which encourages aggression, like Margaret
Meads Arapesh or ancient Sparta. The former will
find this a congenial society, will be very aggressive
indeed and will be regarded as a satisfactory or even
admirable citizen. The latter will find it uncongenial,
will be less aggressive and less well-regarded, and
may have a lesser sense of self-esteem and social
acceptability. Alternatively, suppose that both are
reared in a society which discourages aggression, as
Mead said the Mundugumor did. Success and public
approval will come to the genetically less aggressive
person; dissatisfaction, low social acceptability and
neurosis will come to the other, who is being required
by society to suppress genetically natural
behaviour. Heredity and environment will interact to
produce differences not only in aggression, the only
part of behaviour where there is a genetic difference,
but in other areas of behaviour and thinking which are
related to social experience, such as self-esteem,
social role, and acceptance of social values.
This is quite obviously a gross simplification of the
sort of interaction between genetic programming and
experience that really happens in development. In
particular, we would be rash to assume that any given
environment is identical or works identically for
two individuals, or indeed two groups. If the
environment does differ, it may be these differences
as much as any genetic ones which cause differences
in behaviour. To take aggression as an example a
second time, it is possible that there is a genetically
caused difference in aggression between males and
females. In many species, our own among them,
males are more aggressive in more situations than
females are. However it is quite clear that at least in
the human species there are systematic and pervasive
differences in the ways that males and females are
treated, from birth on, which overlay any genetic
difference. By and large, males are treated as
stronger, more independent, braver than females, and
their aggression is accepted or even encouraged:
females are treated as softer and weaker, more
dependent, more in need of protection, and
16
aggression or even self-assertion is discouraged

except occasionally, in private, or when very
indirectly expressed.
I have written here as though environments
impinged on passive individuals, but this, of course,
is not the case. Individuals behaviour, idiosyncratic
as a result of genetic programming and past
experience, will elicit reactions from the outside
social world which correlate with those
idiosyncracies. The child who has, for example, a
genetically caused articulation problem, such as a
cleft palate, may be less rewarding to listen to, less
encouraged to talk, less rewarded for social
interaction, more introverted, more reclusive and
lower in self-esteem than the child whose
conversation has been more accessible to listeners.
There may be deliberate selection of one
environment rather than another; there may be
selective attention to particular aspects of
environments. It has been suggested that much of
development is a process of negotiation or
transaction between the child and the surrounding
adults or other important social facts of life. This
metaphor too has its dangers, but we will use it at
times.
The implication of the interweaving in
development of biological and social is that
theory and research must deal with them together.
Neither can be reduced to the other. Which is more
important in explaining a particular phenomenon will
vary, will be dependent both on the phenomenon and
on the sort of explanation sought. It has often been
seen as a reason for studying children that they are
nearer the natural, biological state than the adult
who has become a member of the social world. This
is a misleading belief in important ways (Gottlieb
1983). The feral child, the isolated monkey of
Harlows experiments, surviving without any social
contact, is not more natural than the child or
monkey growing up in a family or other social group.
Harlows experiments show that social experience is
necessary for much social behaviour such as
successful mating; similarly, physical experience
such as movement is necessary for much of the
development of physical structures, as well as

behaviour (Hofer 1981).
Scarr and Kidd (1983), in a useful review of
developmental behaviour genetics, discuss the
relationship between human biology and culture.
They point out that there have been no important
changes in brain capacity during the last 30,000 to
100,000 years, though changes in the internal
organization of the brain are possible. Culture has
reduced some selective pressures, for example the
effects of climate which we can cope with by using
artefacts such as clothes rather than by having to grow
thicker fur. It has also imposed selective pressures,
for example to be able to learn to use cultural artefacts
and to live in groups. The possibility of complex
learning is seen as one of the most important of
evolutionary changes. Much of our genetic material
is very similar to that of other primates: what makes
us different is not the instructions for making proteins
which our genes carry but the regulation of protein
manufacture, in particular the prolongation of
infantile characteristics or neoteny which I discuss
in the section on play at the end of this chapter. The
evolution of brain development has selected not for a
specific and limited brilliance, but for generalized
adaptability, including adaptability to culture.
It seems likely in fact that much genetic
programming for behaviour in human beings is
relatively unspecific (Scarr and Kidd 1983). Some
animals, such as insects, do have rigid programming
of behaviour, but are only suited to relatively stable
environments. If the animal is to encounter an
unpredictable environment, one with a variety of
possible events and even experiences never before
encountered by any member of the species, it may do
better if its genes have programmed into it a flexible
repertoire of alternative tactics for coping with
events, and an ability to learn from its interactions
with the environment. Learning by proxy, by
observing other individuals or by being taught by
them, may be a particularly useful strategy. This sort
of animal will need experience in order to cope with
all the learning and development of behaviour that it
must do. It will have a relatively long learning period
childhood and many characteristics designed to
elicit and make use of learning opportunities.

Although it may well have genetic programmes for
certain behaviours, or genetically set goals, these are
likely to be modified by experience: that is the same
function may be served first by a genetically
programmed sequence of behaviour which later is
superseded by a learned one. It also may have
alternative routes to the same final state: not all
individuals will learn exactly the same sequence.
Appropriate genetic information is necessary, but it
is not at all clear that it should be regarded as in any
sense more fundamental, the most essential cause
where other contributions are relatively trivial.
Rather the state of nature is more probably for
human beings to be brought up with other human
beings: a long evolutionary history and a cultural
history which is shorter in terms of years, but very
much faster (probably) in producing change, have
combined to make this possible. It should also be
noted that natural and best are not synonyms: for
a clear discussion of the philosophical pitfalls of
equating the two see Radcliffe Richards (1982, pp.
6780).
The belief that it is sensible to concentrate on one
ultimate essential cause of development can be
extremely strong. It is a belief allied to reductionism,
the attempt to explain the subject matter of the social
sciences in terms of the subject matter of the
biological sciences, or the subject matter of
biological sciences in terms of the physical sciences;
and also to historicism, a belief in tracing events back
to the first in a sequence which is then identified as the
ultimate and most important cause of the final event
in the sequence. Attributing complex behaviour, such
as acts of aggression, to genetic programming or
separation from mother during infancy is both
reductionist and historicist. The danger of such
attribution is that it overlooks other possible
contributing causes which may be just as important.
It increases the possibility that nothing will be done to
intervene to break the sequence, as it is argued that
people who had had the very early experience which
causes aggression much later will inevitably be
particularly violent, so that it is too late to prevent it.
Historicism can thus lead to a very determinist view
17
of development and extremes of optimism or

pessimism. It is clear, as we shall see, that in many
areas early experience may have particularly marked
effects so that an important cause of behaviour may
be traced back through a sequence of several years.
Nevertheless it is rarely the case that early events like
separation from mother or even physical deprivation
invariably lead to inexorable late results. Early
disadvantage quite often but not by any means always
leads to late disadvantage. Early privilege quite often
(slightly more often) but not by any means always
leads to late privilege (see, for example, Jencks et al.
1972). If there is continuity it may be because the
early experience all by itself has such bad (or good)
effects that nothing which happens afterwards can
restore normal development, as is clearly the case in
much early physical and perceptual development
where, for example, rearing kittens without normal
visual input at particular periods in the first few
weeks of their lives will leave the cells in their visual
cortex fewer in number, unusually specified and
unresponsive (see Chapter 2). This sort of effect
cannot, so far as we know, be reversed, so the animals
are blind for ever after. In human psychological
development, however, it seems more likely that
there is continuity of the effects of an early experience
because it sets in train other later experiences which
have similar effects. Thus an embryo may have
inherited a high genetic potential for intelligence
from its intelligent parents; but intelligent parents are
more likely to provide their child with good antenatal
conditions, a trouble-free birth, suitable stimulation
of language and cognitive development, a good
school, support for school activities, a good diet etc.,
while the childs interacting genetic potential for high
intelligence and good early experience are likely to
provoke stimulation, support and high expectation
from parents, teachers, siblings and others. Similarly
it is well documented (Rutter and Madge 1976, Rutter
1985a,b,c) that disadvantage all too often does not
come singly, but in battalions. Epidemiological
studies suggest that single isolated events predict
later psychosocial disorders very poorly, but that if
there are multiple difficulties, especially difficulties
which last for a long period of time, the likelihood of
18
later problems such as persistent juvenile

delinquency or psychiatric illness is much increased.
On the other hand, if there are changes in the
environment (such as adoption produces) there are
likely to be changes in development. For example,
although in three large American studies the IQ
scores at adolescence of adopted children are
positively correlated with the IQs of their biological
parents, as theories which ascribe intelligence to
genetic influence would predict, they are also higher
than would be expected from their biological
parents, raised towards the higher IQs of the
adoptive parents (see Scarr and Weinberg 1983). A
smaller British study of late adoption (Tizard 1977)
described children who had had the same emotionally
bleak orphanage rearing for the first few years of their
lives and were, at the age of 4, socially, emotionally
and intellectually retarded. Those who were then
adopted into well-off families who made great efforts
to give them the best possible life improved; those
who were returned to their natural parent, to families
which were much less rich in resources and were
ambivalent about the return of the child, tended to
deteriorate. Environmental changes for the better
give opportunities for changes for the better in
development; changes for the worse do the reverse.
Continuity and discontinuity in development
It is important to emphasize that continuity in

development is not a simple concept or simple to
measure. As I discuss in the context of temperament
(p. 144), it may not be clear what behaviour is
continuous with what precursor. Kagan (1984), for
example, found in his longitudinal study that boys
who were very timid and withdrawn at ages 3 to 8 had
as adults chosen secure jobs with large employers
rather than choosing risky entrepreneurial
occupations. Aggressive boys were likely to be
bullying adolescents (see also Olweus 1979; and p.
153 this volume). However, the detail of behaviour in
a timid child just starting school and a young man in
his twenties will involve many differences as well as
some similarities: the young man will presumably
have given up thumb-sucking, if not a general
unwillingness to take risks. Unless we have a very

good understanding of behaviour at several different
levels of analysis we may not look at the right things
when trying to assess continuity.
More importantly still, continuity should mean
links over time not just a lack of change. In other
words, we need a fine-grained step-by-step analysis
of what led to what, which will mean the inclusion of
many relevant variables and an assessment of how
inevitable or how weak the links between steps were.
We need to examine how the same behaviour
functions for the child at different points in his or her
life. We also need to know how long the continuity
lasts, and whether it becomes unchangeable or can be
broken by a change in circumstances. Preferably we
would also have a theory of the underlying causes of
continuities or discontinuities and a system of social
or educational intervention or support which worked
against bad effects and for good ones.
At this point I will sketch an example which will be
returned to throughout the rest of this book (it is near
the centre of my professional preoccupations). We
know from a great many pieces of research that there
is an association, perhaps a continuity, between
growing up in a disadvantaged home and showing
behaviour problems as a child and into early
adulthood (for reviews of the literature, see, for
example, Rutter and Madge 1976; Rutter 1985a).
Does the family environment cause the childs
problems, or are there alternative explanations such
as a born to fail child or the effects of other parts of
the social system? What characteristics of the
environment are effective, for worse or better? What
characteristics of the child create, elicit, aggravate or
tone down their effect? Are the effects short-term or
long-term, reversible or permanent? What are the
mechanisms of continuities or discontinuities?
Tentative answers to these questions would be
roughly as follows: they are fully treated later in this
book, particularly in the last chapter. At present there
are many gaps in the evidence which prevent a full
understanding. Family environments do affect their
members and may cause them to have problems, and
so too do other environments such as schools. There
are however also genetic influences which make
problems more or less likely, for example

susceptibility to alchohol, an impulsive temperament
or a low IQ, and some individuals would probably
function badly in most environments. Children affect
their families as well as being affected by them; and
both childrens and families problems may be
associated with some extraneous variable or event,
such as social disadvantage or war. Nevertheless,
despite the validity of these alternative possibilities,
family environments can cause problems in their
children.
Among the general characteristics of families
which are strongly associated with problematic
outcomes for the children there are a few which have
turned up repeatedly, particularly in studies of
persistent delinquency and psychiatric disorder.
Discordant and disorderly family relationships,
parent and peers who model maladaptive behaviour,
lack of effective and warm prevention and control of
misbehaviour and disputes, inability to deal with
crises without extreme stress or helpless collapse,
low intelligence, high socioeconomic disadvantage,
appear, when they occur, to make people function
miserably and to make life choices which increase
their problems. I shall try to present some detailed
studies of relevant processes in my discussion of the
development of cognition, personality and social
relationships. An important and incisive paper by
Rutter (1985a) reviews family and school influences
on the development of socially maladjusted
behaviour.
Stages in development
Stage models of development are relevant to the

concept of continuity and discontinuity in
development. Stages differ in type from theory to
theory. Some are purely idealizations, used by the
writer to evoke particular images in the reader, and do
not refer clearly to anything definite or measurable in
behaviour. Eriksons evocative stages of
psychosexual development (1950) are an example of
this sort of usage. Some stages do refer to
measurable behaviour. Stages here are synonymous
with particular behaviours in an age-related process
19
of change; the crawling stage might be separated

out in the gradual change in ways of being mobile that
we all go through. Descriptive stages like this can be
rather arbitrary and they can be used tautologously
hes being difficult because hes in the 2-year-old
obstructive stage, for example. The better
descriptive stage models include analysis of the
structure and interrelation of the behaviours which
make up a stage: if it is to be worth separating out from
the flow of development, a stages behaviours should
presumably cohere throughout the field of relevance
and be relatively separate from the adjacent stages.
Piagets own stages of cognitive development do
have such underlying models of their structure,
though as we shall see in Chapter 2, there are many
problems with his account: all too many of the
attempts to extend Piagetian stages into moral and
social cognition fall down badly on this. Goldman
and Goldman (1982) for example, had interesting
discussions with a large number of children on their
ideas about sex and various associated issues, but do
very little to justify their forcing of these data into
Piagetian stages. We are not therefore able to judge
whether childrens thinking in this area really goes
through an orderly sequence of wide-ranging and
separate stages as the analysis and presentation
propose.
Even if they do contain a decent description of the
internal organization of a stage and of how it differs
from other stages, descriptive stage models often do
not account for movement from one stage to the next.
If they are to explain development, stage models must
do this (as must models which do not draw
discontinuities in the course of development). Again,
Piagets is the main stage model to try. Explanatory
stage models must specify the behaviours subject to
age changes which make up the stages: they must
propose variables which are responsible for the
changes between stages and the organization of each
stage: they must provide ways of measuring these
variables independent of the behavioural changes
they are supposed to produce. Unless this last
condition is satisfied, stages risk being tautologous
shes sucking her thumb because shes in the oral
stage describes behaviour and has an elaborate
20
model of the causal processes behind the behaviour,

but there are no measures of the oral stage
independent of behaviours like thumb-sucking and
the sentence is descriptive not explanatory.
Judgements of the merit of stage theories vary (see,
for example, Piaget 1970; Brainerd 1978). My own
feeling is that they can be useful shorthand but that
precisely this speed and simplicity makes them
potentially dangerous. Unless the conditions I have
outlined in the last paragraph are fulfilled, stages
tend to segment development in ways which may not
be justified, to simplify the complexity of behaviour
by overlooking the ways in which it varies from
situation to situation, and to give an inadequate
account of transition and continuities. I hope the
remainder of this book will illustrate how much has to
be integrated into a developmental model before
stage analyses can be sound.
The ecology of child development

There is little doubt that the environment children
grow up in is a source of influences on their
development. Work on defining the ecology of
child development is a burgeoning field. There is a
strong awareness of how complex environment is.
It is not just a collection of dyadic relations; in
particular it is not a collection of motherchild
relations. While the ways in which people interact
even in large groups and institutional settings may
bear some resemblance to the way a baby interacts
with its mother, the differences will be obvious, and
they need a different explanation. The view taken by
Bronfenbrenner, that social experience is as a set of
nested structures, each inside the next, like a set of
Russian dolls (Garbarino, 1982, p. 21), is to be
preferred. Bronfenbrenners set forms a framework
for discussing child development, which is relevant
in more than just the social sphere.
Bronfenbrenner (1979, pp. 218) emphasizes
that for human beings the environment is much
more than just the immediate, concrete setting
containing the living creature, though this may
possibly have been an adequate way to think of it
when studying animal behaviour. He asserts that

the ecology of human development involves the scientific
study of the progressive, mutual accommodation between
an active, growing human being and the changing
properties of the immediate settings in which the
developing person lives, as this process is affected by
relations between these settings, and by the larger contexts
in which the settings are embedded (p. 21).
The ecological environment of an individual

consists of micro-, meso-, exo-, and macrosystems.
A micro-system is a pattern of activities, roles and
interpersonal relations experienced by the developing
person in a given setting with particular physical and
material characteristics (p. 22).
The babys life at home with its family is one

example; the babys microsystem is the pattern of
feeding, sleeping, bathing, playing, etc., of relations
with parents, siblings, visitors, etc. (and between
these individuals, of playing the role of baby in a
family playing other roles). As people get older, their
range of microsystems increases: settings like
playgroup, school, church, club, work place and so
forth involve a different range of activities, roles and
interpersonal relations.
A meso-system comprises the interrelations among two or
more settings in which the developing person actively
participates (such as, for a child, the relations among home,
school and neighbourhood peer group; for an adult, among
family, work, and social life) (p. 25).
The nature of the links between an individuals

microsystems is what is involved here. Strong and
varied links between microsystems such as other
people who participate actively in both settings,
communication between settings, knowledge and
attitudes being applicable to both are seen as being
an advantage to the people involved. A weak
mesosystem, minimal links between home and
school, for example, may place the child at risk.
The last two systems are more removed from the
child.
An exosystem refers to one or more settings that do not

involve the developing person as an active participant, but
in which events occur that affect, or are affected by, what
happens in the setting containing the developing person (p.
25).
The macro-system refers to consistencies in the form and
content of lower-order systems (micro-, meso-, and exo-)
that exist, or could exist, at the level of the subculture or the
culture as a whole, along with any belief systems or
ideology underlying such consistencies (p. 26).
Thus exosystems are settings which affect the child

but in which he or she takes no part, for example the
parents work or their friendship networks, or local
government decisions about schools, housing,
playgrounds etc. Macrosystems reflect lifestyles and
belief systems which are in contrast in different
societies and which affect the smaller-scale systems
in which individuals move. A belief in social
competition rather than social co-operation, or in
democracy rather than autocracy would be examples.
So would the degree of priority given to the needs of
children.
Bronfenbrenner goes on to emphasize that life
contains many transitions between ecological
settings, which are a function of biological changes
such as increased maturity and of altered
environmental circumstances, and that these
transitions are both consequences of developmental
processes and instigators of further change. He then
defines human development as follows:
Human development is the process through which the
growing person acquires a more extended, differentiated
and valid conception of the ecological environment, and
becomes motivated and able to engage in activities that
reveal the properties of, sustain or restructure that
environment at levels of similar or greater complexity in
form and content (p. 27).
This is a definition which we will consider further in

the remainder of this book. The rest of this chapter is on
21
Plate 2 Seville, 1933

22
Play, a topic which does not fall neatly into the

categories which organize the rest of the book, but
illustrates some of the issues about the nature of
children and of development which I have discussed.
Play
Children, like many other young mammals, spend a
considerable amount of time and energy playing.
This activity has been evaluated by adults in a wide
range of different ways. One view has been that
although childrens play may be enjoyable to the
participants and charming to fond parents observing
it, it is essentially frivolous, a pastime with no
intrinsic consequences, which is grown out of
without it having made much impact on
development. At one extreme this view was stiffened
by a sterner judgement that frivolities should be
discouraged in favour of useful activities, that
children should play less and work or pray more. It
has almost certainly been a fairly recent development
to take a contrary view and see play as making
important contributions to the childs development,
to consider it to be an important way of learning
things that the adult will need to know. This view has
had a period of considerable dominance in
psychological and educational theory: the
paradoxical slogan play is the childs work has been
implicit or even explicit in many early childhood
curricula.
It seems to be common practice in our society to
draw a distinction between work and play. Parents
do this, so do teachers: so do children, who come
home from infant school and tell their inquiring
parent that they just played, though the teacher no
doubt defined their activities in the classroom as
educationally important and as work. The work/play
distinction is conflated with the distinction between
having to work and being free to do what one wishes,
and play is seen as voluntary and not obligatory.
Voluntary activity is seen as free, absorbing,
spontaneous, enjoyable, not serious and done for
oneself not for other people (see p. 85 for discussion
of similar approaches to study). Given a model of
motivation which rests on the assumption that an
organism will not expend energy on activities such as

play which are neither obligatory nor obviously
useful in terms of drive reduction or immediate profit,
it becomes a problem to explain why anyone should
play. I will discuss here some general theories of play
and the validity of the assertions that have been made
about its nature and functions: because play will be
seen to have many different forms and functions,
detailed accounts of their contributions to
development will be found in the appropriate sections
of the following chapters. Thus the contribution of
play to cognitive development is examined in
Chapter 2: social play in school playgrounds in the
section on peer relations in Chapter 6.
Defining play
I have already implied that judgements of what is

play may be subjective and hard to agree on.
Building a wall is likely to look like play if a 2-yearold does it with wooden blocks, and to look like work
if an adult does it with bricks to earn a living or
improve the look of the garden or even to acquire and
practice a new skill; though the effort, involvement
and pleasure at the finished product may be
equivalent in each case. Dressing up in fancy clothes
and pretending to be someone else is certainly play
in the nursery classrooms Wendy house, and in a
more subtle form it might still be so in the stalls at the
National Youth Theatre: but on the other side of the
footlights, or in the stalls at Covent Garden, is it
play? A casual knocking-about of tennis balls with
tennis rackets in a suburban street is play: the
Wimbledon championship matches barely are,
however much we may extol amateurism.
However, there are certain characteristics which
we may see in prototype instances of play, though
more marginal instances may lack most of them.
Garvey (1977) provides one useful list. First, play is
essentially enjoyable and associated with positive
affect. Second, it is an activity done for its own sake,
rewarding in itself, and not dependent on extrinsic
motivation or the achievement of goals outside the
play. Third, it is spontaneous and voluntary, chosen
and generated by the player not by outside authority.
23
Fourth, it requires active involvement on the part of

the player. Fifth, and very important, it is contrasted
with non-play by being in a sense set aside from
real life: it is not intended literally, is not meant to be
taken at face value. Play involves actions done out of
their normal pattern, perhaps fragmented, rearranged
or repeated, carried out as the player chooses not
organized in ways specified by the achievement of an
outside goal. It is framed by special signals that this
is play, not literally what it appears to be, not
leading to the usual consequences, not to be reacted
to by other people as if it were real fighting, real
boasting, real going shopping. Sutton-Smith (1979)
proposes that it is a sort of performance, not a solitary
act but a communicative act directed at real or
imagined others, even if the others are aspects of
oneself. This contrast with non-play allows play to
be buffered from its normal consequences so that
possibilities can be tried out without responsibility
for what happens: Its all right, Im only playing.
While these defining characteristics are certainly
worth thinking about, cumulatively they lead to a
misleading idealization of play. It has been said to be
free, outside ordinary life, not serious, not literal, not
for profit: absorbing, spontaneous, voluntary, due to
intrinsic motivation: refreshing, flexible, egalitarian
and showing positive affect. It is thus heavily
contrasted with the serious aspects of life, such as
earning a living, passing examinations, getting
promotion or an overdraft, fighting, feeding, etc.,
which are seen as lacking these essential qualities.
While the general tendency may be for instances of
play to deserve these pleasant adjectives, to require
them as constituents which must be present if
something is to be called play is to pile up a very
large number of problem instances. Exactly what is
meant by the free/voluntary/spontaneous cluster, for
example? As Sluckin (1981) has observed, much of
what goes on in the school playground is strongly
constrained and coerced by other children (and to a
much lesser extent by adult supervisors also):
freedom here (as elsewhere) is a nebulous concept.
The playground again provides many examples of
play behaviour characterized by negative affect, by
fights and tears and jibes rather than by smiles and
24
affectionate hugs, and of play which is most certainly

not egalitarian. Indeed we were only playing is
often used in an attempt to excuse too much damage
being done.
Play is not necessarily always functional, either.
Fagen (1981) points out that animals take risks and
incur injuries in play, and this is all too evident in
human beings. Sutton-Smith and Kelly-Byrne (1984,
pp. 31416) give some rather shocking examples of
their students deep play: to their examples of
dangerous play arising from adolescent bravado, one
could easily add many other instances including the
statistics of injuries to young children from more
innocent playground activities. Play certainly
becomes less common if the individual is in adverse
circumstances, ill, hungry, frightened or under stress
(Smith 1984, Gould 1977), which might suggest that
it can involve a degree of additional stress to be
avoided when times are hard. Play, like any other
behaviour, can contain negative aspects within a
generally relaxed and positive ambience.
A final characteristic found in many instances of
play is the special quality or state of being which
Csiksentmihalyi (1979) calls flow. Flow involves
a blissful involvement in what one is doing, a loss of
self-consciousness, being carried away by the
activity but simultaneously participating fully in it
and not being out of control. This sort of experience
can be found in many activities which are not play
academic work, when it goes well, is an example, so
might aesthetic experiences and sexual activities be
and many instances of play would be devoid of
flow, but there is obvious overlap between the two.
Considering childrens activities in terms of flow
might also usefully decrease the idealization of play
and the artificial play/work distinction which has
impeded useful analysis and provision for so long.
An alternative to defining play by constructing lists
of its essential characteristics, such as I have just
discussed, is to construct taxonomies of different
types of play behaviour. An early example is Piagets
threefold scheme (Piaget 1962): practice play, which
involves repetition of actions with elaboration of
means being more important than the ends served,
symbolic play involving the manipulation of
symbols, and games with rules. Hutt (1979)

distinguishes exploration and problem solving from
play which is symbolic or rulegoverned. Other
taxonomies have been devised, including some
which specify the materials of play (Garvey 1977) or
its social interactions (Parten 1932). Some of my own
work (Meadows and Cashdan 1983) rated play
behaviour, like other behaviour, on dimensions such
as social participation, degree of childs involvement,
number of operations, themes or skills involved,
extent to which materials were used, and apparent
goals of the child, which did not reduce tidily to
separate types of play. Approaches such as these are
useful in that they stay close to observable behaviour
and stress the variability intrinsic to the concept of
play.
It should be noted, finally, that play needs to be
studied in the context of its ecological environment.
As we will see, it is an important part of the
microsystems of a young child in the family and of
children together in school, street or playground; it is
affected by the exosystems which, for example,
decide to provide or not to provide playing fields,
amusement arcades or cable television; it is allowed,
curtailed or channelled by the macrosystem which
insists that children must contribute to the economics
of their family from a very early age or not until they
are in their late teens, or encourages football in little
boys and discourages it in girls, or fosters
conceptions of childrens innocence which push their
sexual or scatological interests underground. What
play is, and what it does, will not be understood
without consideration of its ecology.
Theories of the causes of play
Play is found in the young of many species. Because

of this possible biological impetus to play, one strand
of interest in theories of why play happens is the
consideration of evolution and biological function.
Classic theories (for a review see Millar 1968 or
Garvey 1977) described play as being for getting rid
of surplus energy, or for practising skills which will
be needed later in life. An early evolutionary theory
of play, famous, picturesque, influential and total
nonsense, was G. Stanley Halls recapitulation

model. There was at the time (the turn of the century)
a belief that a species evolutionary history was
recapitulated in the development of the individual,
that ontogeny recapitulated phylogeny. (Some
apparently supporting evidence came from
embryology; for example, the embryo was seen at
one stage to have gill slits like a fish, and later looked
rather like a monkey. Freuds theory of psychosexual
stages is also recapitulatory.) Hall proposed that the
successive stages of childrens play recapitulated the
behaviour of their evolutionary ancestors. Babies
enjoyed splashing about in the bath because their
distant ancestors had been aquatic creatures
splashing about in the primeval slime. The pre-school
child climbed, jumped, swung as monkey-like
ancestors had done. The hide-and-seek games and
gangs of older children recapitulated the hunter
gatherer activities of early man. Team games
recapitulated tribal conflicts, stamp collecting the
ascendancy of capitalism. This model is historically
crude, selective in the play it accounts for, and
strongly resembles Kiplings Just-So Stories. It is in
any case rooted in a theory now known to be
fallacious, since ontogeny does not recapitulate
phylogeny (Gould 1977). Halls theory seems,
however, to have rooted itself in common-sense
theories of play: adults general comments on
childrens play and professionals evaluations of
natural materials like the sand and water found in
many pre-school curricula often bear a recapitulatory
resonance.
More recently, advances in evolutionary biology
have produced more sophisticated and better founded
accounts of play. Recapitulation theories, which
identified a series of adult ancestral forms in the
immature baby and child who develops beyond them,
have been replaced by models which see human
development as retarded, not as speeded-up and
extended. Instead of being like adult moneys when
we are children, and developing beyond where the
monkey stopped to be super-adult humans, we retain
many of the infant characteristics we (and monkeys)
25

Table 1 Retardation in humans compared with other primates
Rhesus macaque
Gibbon
Gestation
Complete hair covering
24 weeks
During gestation
30
Ossification centres
in wrist at birth
First teeth (months)
Second teeth (years)
Growing period (years)
Life span (years)
All
34
onset during gestation,
completed after birth
23
Gorilla
Human
37
40
Never
completed
None
0.65.9
1.66.8
7
1.2?
?8.5
9
2.712.3
2.910.2
11
3.013.0
3.010.5
11
6.024.0
6.020.0
20
25
33
35
35
70
had as infants, for far longer than the monkeys do.

Information from Gould (1977) illustrates this (Table
1).
This retardation in development reflects an
evolutionary trend which has been pervasive among
large mammals: instead of having large litters of
young which although born helpless dev elop rapidly
to independence, they have reacted to selection
pressures and produced smaller litters of slow
developers, a long gestation and a long period of the
offsprings dependence on adults. The human pattern
is an extreme version of this, and it includes a
prolongation of infantile physical characteristics. To
quote Gould (1977, p. 371):
In practically all human systems, postnatal growth either
continues long past the age of cessation in other primates,
or the onset of characteristic forms and phenomena is
delayed to later times. The brain of a human baby continues
to grow along the fetal curve; the eruption of teeth is
delayed; maturation is postponed; body growth continues
longer than in any other primate; even senility and death
occur much later.
Gould sees retardation as of adaptive significance in

human evolution because it allowed the retention of
juvenile physical characteristics for longer than was
otherwise possible (neoteny); a psychologically
important example is skull shape and hence brain size
26
Chimpanzee
(Gould 1977, pp. 37699). Retardation made both

possible and necessary some distinctively human
characteristics particular body shapes, sizes and
coordinations, more elaborately developed brains,
longer and better learning of a more open potential
range of behaviour and ideas, longer and more
elaborate parenting, more complex social
relationships which have interacted synergistically
so that human development is more plastic and less
preset than that of other animals. We will return to
some of the implications of this at intervals
throughout this book: I will argue here that the causes
and functions of play can be fitted neatly into the
neoteny framework.
Recent evolutionary accounts of play (e.g.
Burghardt 1984; Byers 1984) use more sophisticated
concepts of how evolutionary processes work
together (see, for example, Gould and Lewontin
1979). They make the important point that the
functions which play in the young of a particular
species now serves are not necessarily at all the same
as the functions which similar behaviour served back
in their evolutionary history. It will be difficult to
clarify the evolutionary functions of play avoiding
the Just-So Stories which amuse but have no
scientific foundation. However, the core is a
recognition of the young animal as being born
immature and very able to learn, needing to live with
its conspecifics and to learn many complex skilled
behaviours, and possessing energy and curiosity and

social affiliativeness enough to do this. Play is a result
of these characteristics and a facilitator of
development.
Some possible functions of play
Arousal modulation
One proposed function of play (Berlyne 1960; Shultz

1979) is to modulate arousal. If there is not enough
environmental stimulation to induce a moderate level
of arousal in the child, he or she will play to increase
arousal level. If arousal level is high, if the child is
anxious or over-excited, play that is stimulus-seeking
will cease and the child may turn to play which is
calming and reduces arousal level. Undertaking a
new behaviour will be arousing, and may, if it is too
much so, be mildly aversive, but anticipating or
achieving control over this novelty will reduce
arousal to the optimum and be pleasant.
Obviously this theory of arousal is a very general
one which could apply to any sort of activity. The
model is particularly appropriate for play, however,
in so far as it is a largely voluntary activity. The child
has opportunities in play to modulate his or her own
arousal level autonomously, indeed to play at
arousal modulation by seeking over-stimulation and
coping with it. The getting over-excited in play, to
which children often succumb, is perhaps the
converse of this function of play: arousal must not get
too far off balance if what Csiksentmihalyi (1979)
calls flow is to be preserved.
Practising adult activities
Kittens play at chasing and catching mock prey. Colts

race each other round their pastures. Little girls play at
making tea, washing up and putting their dolls to bed.
Adults, watching, see all this as practising adult skills
which the young will need when they grow up. Play is
said to be for developing behaviours and rehearsing
roles which will be necessary for individual success
and for the survival and propagation of the species: it is
thus an important component of socialization,
biologically programmed.
Is this explanation and function true? The answer is

yes and no. Human beings even more than other
mammals have to learn complex sequences and
patterns of behaviour which are not genetically
programmed. As I discussed, evolutionary theories
note the contribution that play can make to this
learning because it is voluntary, flexible, and sanction
free. The existence of play does serve the
development of immensely complicated adult skills
such as food gathering, keeping with the herd, and
parenting, but play is not for these things. The most
prominent candidate for genetically programmed
preparationfor-adult-skills play is rough-and-tumble
play (Humphreys and Smith 1984), which is clearly
culturally influenced in who does it, and differs in
many ways from the adult fighting for which it is
supposed to be a preparation. Children are very
subject to adult pressures, both implicit and explicit,
to play properly, and it is these that shape the
content of their play, not an evolutionary plan. Thus
play is rarely simply practising adult activities,
though it does allow the practice of behaviour and
roles which are not irrelevant to adult life.
Allowing behavioural recombination
Play activities tend to occur most in the early stages

of the behaviour systems involved. They serve the
mastery of complex behaviour which requires both
practice and the integration of the behaviour into
other systems and knowledge. After the kittens play
pouncing, pouncing behaviour is used by the cat,
more skilfully and better integrated into the whole
stalkpounce catch routine. But because play is
framed, nonliteral, and concentrates on means not
ends, it allows the recombination of pieces of
behaviour, ideas and consequences in innovative or
repetitive ways. The child has good opportunities in
play for varying what are the normal conditions of
reality, breaking set, introducing new considerations
and choices, and so coming to control in action and
understanding the rules of individual or social
behaviour. It is important to stress that play may do
this but does not always do so: some play is rigid and
inflexible. Further, behavioural and conceptual
27
recombinations may be sought deliberately and nonplayfully, as in certain teaching programmes for
cognitive development (see Chapter 3).
Emotional and social functions of play
A variety of emotional functions have been proposed

for play, notably by Freud and his followers. They
include wish fulfilment, anxiety reduction, and
mastering a traumatic event. Clinically these
functions seem very possible. Play which could serve
them seems to be displaced in time from the height of
the emotional crisis, which could be congruent with
the arousal theory discussed above. Whether play
does reduce anxiety or induce catharsis has been hard
to assess, and no conclusions are possible on present
evidence (Rubin, Fein and Vandenberg 1983).
Play is often engaged in with other people. It
therefore offers opportunities of learning about other
people, adopting different social roles, establishing
group structures, etc. As I discuss in the chapter on
social development, even very young children seek
social interaction and play with other children, and
those who lack play and social experience as children
tend to have social difficulties as adults. It must not be
concluded from this, however, that childrens social
play is uniformly beneficial or that the solitary child
is necessarily in developmental danger.
The development of play behaviours
Whether play contributes to the development of the

individual is a question to which we will return later.
The development of the child is certainly involved in
the changing patterns of play which we can observe
from infancy to adulthood. Garvey (1977) describes
how new skills, experience and knowledge become
the resource or materials for play. Infants play with
motion and in interaction with their caretakers, play
with objects develops then play with language, with
social materials and peers and eventually play with
rules. Those resources and materials for play which
were available early do not necessarily become
redundant as new ones are acquired: instead they are
combined in progressively more complex ways to
suit the childs purpose.
28
As Piaget (1962) has described, infants play with

the movements they can make and the results of these
movements. Such play seems to be enjoyable for its
own sake, but no doubt also contributes to the childs
control and co-ordination of movements and to his or
her physical awareness, through kinesthetic
feedback. Movements are also commonly part of
parentinfant games, for example games of clapping
hands or riding the parents knee. I discuss the social
structure of these games in Chapter 6; early on the
adult is responsible for the maintenance of the
movement game, as also of course for the recitation
that often accompanies it, but as the infant learns the
routine of the game he or she comes to take a more
active part. Traditional rhymes provide not only
opportunities for contrasting movement, but also
linguistic and social information, as in this rhyme
dating back several hundred years:
This is the way the ladies ride,
Nimble, nimble, nimble, nimble,
This is the way the gentlemen ride,
A gallop, a trot, a gallop, a trot;
This is the way the farmers ride,
Jiggety jog, jiggety jog;
And when they come to a hedge they jump over!
And when they come to a slippery space
They scramble, scramble, scramble,
Tumble-down Dick!
(Baring-Gould and Baring-Gould 1962)
Infants come to play with objects in a more and

more complex fashion. Initial object play is often said
to be indiscriminate in that infants assimilate
objects to their repertory of behaviours like grasping,
shaking and mouthing without much adjustment to
the objects particular characteristics (Kagan et al.
1978; Rosenblatt 1977). From the latter part of the
first year onwards, play with objects becomes more
likely to involve the use of more than one object, to
show response to the characteristics and the normal
function of the objects, and to begin to involve
pretence, for example that a doll can drink from an
empty cup. Pretend play increases rapidly at about the
same time as language development is taking off:
several theorists have seen them as two facets of the

childs new insight into the possibility of
representational behaviour (see, for example, Rubin,
Fein and Vandenberg 1983).
Given the opportunity, children enjoy play with
materials that allow them to make constructions.
Building with bricks, fitting bits of meccano together,
glueing bits of paper to cardboard boxes, and the
slightly different activities of painting and
completing jigsaws, all allow the child to set a goal
and then achieve it. They also can fit into, indeed be
done in order to serve, social and pretend play. These
are favourite activities from the pre-school years
onwards, often producing especially long bouts of
concentration (Sylva et al. 1980; Meadows and
Cashdan in press).
As they come to spend more time with age-mates,
children come more and more to play with peers (see
Chapter 6). Early play with peers is very much
dependent on the ability of at least one participant,
and preferably all, to sustain a sequence of
interaction. During their pre-school years children
develop more efficient ways of recruiting,
incorporating and directing peers in joint play (e.g.
Asher and Gottman 1981), but some adjustments to
anothers presence and activity are made even in
parallel play, especially if the children playing side
by side are reasonably familiar with each other. Early
social play with peers is facilitated, as early adult
infant play was, by social routines. Garvey (1974,
1977) provides many examples of ritualized
exchanges of language. For example:
Xs turn
Y s turn
1 Hello, my name is
Mr Donkey
Hello, my name is Mr
Elephant
Hello, my name is
Mr Tiger
2 I have to go to
work
Hello, my name is
Mr Lion
Youre already at
work
No Im not
I have to go to
school
No Im not
Youre already at
school
The scripts of Playing house or Superhero

(Paley 1984) also provide a supporting structure for
young childrens social play. Much of such play is
excruciatingly banal and stereotyped to the adult eye
and ear, and many commentators have been worried
by its sexist and aggressive components, but it seems
to be very resistant to adult attempts to induce a new
role play of egalitarian sweetness and light (e.g. Best
1983; Paley 1984). Script play no doubt reflects
childrens social knowledge, which so far as sex role
goes tends to involve relatively rigid stereotypes
during the infant and junior school years (see p. 197).
Scripts can become more elaborate as children get
older, often deriving from the stories they read.
Alison Luries novel Only Children (Lurie 1980)
contains notable examples of such fantasies.
Play with rules, where the play depends on an
externally fixed set of regulations, is usually seen as
a late emergence in the development of childrens
play, though it could be related to the structured social
routines of adultinfant play. The participants in
games with rules have to recognize, accept and
conform to constraints imposed on their activity.
They have to take turns, for example, accept losing,
not cheat, etc. Especially in Piagets account (Piaget
1962), play with rules centres on co-operating in
order to compete. (Competition is of course not a
salient characteristic of adultinfant routines.)
Games with a competitive element and a rulegoverned structure may involve movement, language
(see Opie and Opie 1967, 1969, 1985), objects, social
interaction and even scripts. They are probably the
most conspicuous activity in many junior school
playgrounds, and in very sophisticated forms such as
football persist into many adult lives.
There are marked variations between children in
the ways that they play. Individual differences in
temperament and perhaps intelligence (Rubin, Fein
and Vandenberg 1983) are one source of variation:
the ecology of the play setting is another (Smith and
Connolly 1981): the materials available for play are
also important (e.g. Meadows and Cashdan, in press;
Fein 1981). Probably the most important source of
29
variations in play is, however, culture. Schwartzman

(1978) provides an excellent review.
On the importance of play
I began this section by saying that after a period when

adults saw play as being frivolous and
inconsequential, we have had a period in which play
has been seen as an important contributor to the
childs development. It is important to note that there
is little conclusive evidence for the case that play is of
unique importance. Experimental studies (see Smith
1984, Rubin, Fein and Vandenberg 1983; and pp. 65
6) do not show for certain that more play causes better
problem solving, language, imagination or social
adjustment. Children deprived of play who show
impaired development have almost always been
deprived of other things as well, indeed rather severe
deprivation is needed to stop play. Lack of more
suitable play facilities contributes to the sort of
delinquency that is antisocial play. It is probably the
case that the benefits of play could accrue from other
30
non-play activities, that it is not altogether essential in

development.
Nevertheless, play cannot be written off as useless.
In the first place, it is a source of enjoyment and
pleasure and may hence make a positive contribution
to the childs emotional well-being. It is a potential
source of feelings of competence and achievement
and so a contributor to the childs self-esteem and
feelings of self-efficacy. It is part of the childs social
worlds of peers and of adults. The contribution of
adults to childrens play, through encouragement,
support and resource provision, should not be
underestimated even if children do have a private
world secluded from their parents and teachers. Play
provides opportunities for trying out skills and
investigating the world. It may also be used to assess
the childs development, and through some therapies
to enhance it. The proverbial wisdom of All work
and no play makes Jack a dull boy has considerable
truth in it.
Plate 3
31
Plate 4
32
2 Perceiving and understanding
Cognitive development
Studying cognitive development, we are concerned
with the child as knower; with someone who thinks,
understands, learns, remembers, and so forth. There
is still no clear, complete and valid account of what
adults do when they think, understand, etc., despite
hard work by philosophers over thousands of years,
more recently by psychologists, and very recently by
computer scientists and neurophysiologists.
Accounting for cognitive development additionally
involves describing what develops, that is, noting
what changes between different ages and explaining
how these changes come about. Quite obviously
these are extraordinarily formidable questions. We
cannot yet answer them; but currently psychologists
who have learned from Piagets partly correct and
partly incorrect answer are putting together an
exciting new account.
I will outline and discuss Piagets achievements
first, and then proceed to the post-Piagetian picture of
what and how thinking develops. Good introductions
to Piagetian theory are provided by Brainerd (1978)
and Brown and Desforges (1979); for fuller accounts
of Piagets work see Flavell (1963, 1977), Vuyk
(1981) and Gelman and Baillargeon (1983).
Piagetian theory
An anecdote which Piaget used many times gives a
good picture of the Piagetian child. A boy aged 5 was
playing with his collection of pebbles. He laid them
out in a line and counted them along the line from left
to right: there were ten. Then he counted them from
right to left, and to his great astonishment the total
was, again, ten. He put them in a circle and counted
them first clockwise and then counterclockwise: full
of enthusiasm he discovered that there were always
ten, that the sum of objects was the same whatever

order they were counted in. He grew up to be a
professional mathematician, attributing his choice of
career to this experience at 5. It had been an
excitement and a delight to achieve a new cognitive
control in the world by putting the pebbles in order,
comparing that order with another pre-existing order,
the string of numerals from one to ten, and so create a
property of the collection, their sum of ten, which did
not exist independent of the counters activity.
Cognition as adaptation
Piagets theory of the development of thinking has at

its centre the child actively trying to make sense of the
world, just as any organism must try to adapt to its
environment. According to the theory, making sense
of, at whatever intellectual level, is a special case of
the adaptive processes which pervade all biological
existence and evolution. It proceeds through the twin
functional invariants of assimilation and
accommodation. Assimilation is the relating of new
information
to
pre-existing
structures
of
understanding, and accommodation is the
development of old structures into new ones at the
behest of new external information or problems. The
two occur together, though one may dominate the
other, and their functioning gives rise to a series of
structures of cognition, that is, cognition is organized
into systems of rules, categories, procedures and so
forth which eventually amount to unified
organizations of logical operations. Cognitive
development proceeds through the steady functioning
of assimilation, accommodation and organization,
which together give rise to a succession of increasingly
complex, differentiated, integrated and flexible sets of
ways of understanding the world. There are such
successions of qualitatively different structures in a
number of different areas of content, such as
33
conservation, perspective-taking, causality and

number concepts: there is also, more importantly, the
succession of such stages in the global area of
cognition-in-general, in so far as the structures of
thinking apply across different contents. Piagets
model insists on the general application of cognitive
structures: there are only minor differences in a childs
level of thinking in different areas or different
materials, and differences between children only in
rate of progress through the invariant order of stages.
The emphasis is thus on cognitive structures and
logical operations which are universal and abstract:
inferred from behaviour, not immediately observable.
Thus the Piagetian child is an active thinker who
continually strives to understand the world in a
coherent way. Trying to understand, constructing a
world-model, adapting to the demands of the
environment, are constant in development and do not
change. Insistence that knowing and thinking are
active not passive ways of coping with the world is
Piagets first important contribution. The model of
assimilatory and accommodatory processes for
taking in information, that is relating what one
discovers for the first time to what one knows already,
would also seem to be an important one, indeed many
psychologists believe that a model without
assimilation and accommodation would be quite
implausible, though they need much more precise
specification (see, for example, Gelman and
Baillargeon 1983; Meadows 1983; Siegler 1983).
From birth (or even earlier) to death we are
processing new information and fitting it to our old
knowledge, thus transforming old knowledge into a
new improved version.
As far as what does develop during cognitive
development is concerned, Piaget suggested that
there was a move during childhood from thinking that
was fragmented, partial and closely tied to
experience, to thinking that was at least capable of
being logical, abstract and very flexible, by
adolescence. We will see that there is good reason to
doubt both whether young children are as
unsystematic and illogical as Piaget supposed, and
whether older children and adolescents (and indeed
adults) are as logical. The role of education will turn
34
out to be rather important in understanding cognitive

change.
Causes of cognitive development
Piagets model (see, for example, Piaget and Inhelder

1969, pp. 1547) of how cognitive development
proceeds was complex. He lists four factors, which
are of rather different types and not clearly
interrelated. The first relevant factor was organic
growth, particularly the maturation of the central
nervous system. It is quite obvious that a properly
functioning brain is normally necessary for adequate
cognition and that the immaturity of infants brains
(brain development, most rapid before birth,
proceeds rapidly for the first year or two after birth
and continues to some extent for much longer) makes
them both unlike adults and capable of becoming like
them. Very recent research is beginning to inform us
both about brain development and about the brains
role in cognition (pp. 589). At present theories of
cognition flourish without much neurophysiological
base, but eventually, one might hope, we will know
enough to link the different levels. Kinsbourne
(1980) addresses the problem.
Piagets second factor explaining how cognition
developed was the role of exercise and of acquired
experience in the actions performed upon objects,
including both direct physical experience, e.g. the
weight of pebbles, and indirect reflective experience
of logico-mathematical rules and relations, e.g. the
total number in a collection of pebbles. Again, some
such necessary factor is incontrovertible, even if it is
not entirely clear what and how much experience is
needed. The restrictions of experience suffered by
some very severely deprived children (Skuse 1984)
or by children who are congenitally paralysed, blind
or deaf, do not necessarily lead to deficient cognitive
development though there may be abnormalities and
problems. It has often been inferred from Piagets
account of the role of experience in cognitive
development that young children can only learn by
active involvement, not through observation or being
taught: we will see that a definition of experience
that restricts it to hands-on experience is too

restrictive.
Other sorts of experience are implicated in Piagets
third factor, social interaction and transmission.
This was the least developed part of Piagets model,
but has received a great deal of attention in recent
years. We will discuss it later, and see that the social
ecology is of great importance to cognitive
development and to its diagnosis by experimenters.
Equilibration
Piagets final factor, equilibration, was invoked to coordinate the diverse contributions of maturation and
physical,
social
and
logico-mathematical
experiences. It was a central concept in his theory,
most important in accounting for how development
occurred. Briefly, he postulated that organisms
needed to maintain a stable internal equilibrium
within the changes and uncertainties of the outside
world. Body temperature in warm-blooded animals is
a good example of the sort of process involved: the
feedback systems of thermostatically controlled
central heating are also analogous. In equilibrated
systems, of which cognition is supposed to be one, the
changes and demands of the outside world produce
small perturbations or conflicts in the system
which automatically adjusts itself to cope with them
and return either to the original steady state or in the
case of cognition to a new and better equilibrium.
There is a strong need for equilibrium: durable
disequilibria constitute pathological organic or
mental states (Piaget 1968, p. 102).
The concept of equilibration explains how
cognitive development occurs in terms of a need for
a coherently organized and consistent way of
thinking. This equilibrium is gradually constructed
as partially adequate ways of thinking conflict with
the data provided by the external world, or with their
own inconsistent processes and results, and have to
be improved. Equilibration implies that there
should be a considerable degree of organization and
coherence in cognition. It also implies that conflict
between ideas or models or ways of doing things will
be a major source of progress. Piagets account of it
also involved an insistence that the series of stages of

equilibrated ways of thinking which he described was
universal. Before the formal operational thought that
is the pinnacle of human cognition can be achieved,
every human being must go in the same order through
the sensori-motor substages, the period of preoperational thought and then concrete operations.
Because
equilibration,
assimilation,
accommodation and organization are complex
and abstract concepts, providing behavioural
examples is difficult. I offer a fictional one, designed
to highlight three points I want to make about the
equilibration model, but not, I hope, a caricature of
what Piaget intended. At the outset, the protagonist of
the example is a moderately experienced cook. He or
she has successfully cooked carrots and potatoes,
carrots by boiling them, grating them raw for salads,
and making carrot soup, potatoes by boiling,
roasting, frying, mashing and making chips. The
initial state of knowledge of ways of cooking carrots
and potatoes could be represented as a matrix, thus:
Method
Carrots
Potatoes
Boiling
Chipping
Frying
Grating/salad
Mashing
Roasting
Soup
not tried
not tried
not tried
not tried
not tried
not tried
The cook now meets parsnips for the first time. The
new vegetable is assimilated to the carrot repertory,
perhaps on the basis of similarity of shape and
texture. Parsnips boil very well, make a rather bland
soup, and although they taste quite pleasant raw are
not a visually attractive ingredient in salad. The cook
also assimilates parsnips to the potato repertoire:
parsnips are disastrous chipped or fried, mash well
and are delicious roasted. The knowledge matrix
after assimilation of parsnips would have one further
column, thus: Assimilation is always accompanied by
35
Method
Parsnips
Boiling
Chipping
Frying
Grating/salad
Mashing
Roasting
Soup
No
No
maybe
maybe
accommodation, however. The schema vegetables

you can mash is augmented with parsnips, for
example, and the importance of butter is emphasized.
The cook may accommodate further by extending the
mashing schema to carrots, and hence to the full range
of vegetable pures. Since roast potatoes and roast
parsnips are successful, the schema roasting may be
tried out on carrots, with success. A fully equilibrated
knowledge system for cooking carrots, potatoes and
parsnips might look something like this:
Method
Carrots
Potatoes
Parsnips
Boiling
Chipping
Frying
Grating/salad
Mashing
Roasting
Soup
No
No
No
maybe
No
No
maybe
maybe
This sort of procedure is obviously infinitely

extendable. The same basic principles of
assimilation, accommodation and organization
leading to a thorough equilibrated knowledge
structure will apply whatever new vegetable the cook
encounters turnips, celeriac, Jerusalem artichokes,
and so forth.
The three points I want to raise from this example
are as follows. The first is the question of what criteria
of content guide assimilation. Why are parsnips
assimilated to the carrot and potato schemas, and not
to the schemas for cooking aubergines or
cauliflowers, or indeed to schemas for playing chess
or cleaning cars? I suggested similarities, of shape,
texture and, implicitly, being root vegetables.
36
Nutritionists, aware of the chemical constituents of

the vegetables, or cookery-book writers, aware of the
long history of cuisine, might have other
suggestions to make. There is no Piagetian answer,
and indeed not much of an answer elsewhere, as we
know very little of how peoples conceptual systems
are organized.
The second point to be raised is how thorough and
how wide ranging we should expect peoples
equilibrated knowledge systems to be. Are all gaps in
knowledge really investigated or thought through,
and are all internal contradictions really faced and
resolved? Is there one unified system of knowledge in
which cooking is a small but integral component? Is
cooking knowledge, even, systematic and
integrated? Piagets work implies the answer yes to
these questions: what evidence there is suggests local
organization with a lot of inconsistencies remaining
(Boden 1982).
My third point is about where knowledge comes
from. Cooking illustrates it particularly well. A truly
Piagetian cook would arrive at his or her knowledge
of ways of cooking vegetables from assimilating new
objects to the pre-existing repertoire of techniques
and accommodating the old repertoire to the new
objects, as I have described. Most of the cooking I
know of consists of doing things because other people
have done them successfully, copying parents or
friends, following the instructions given in a recipe
book or on the packet. Even though some cooks may
vary recipes to fit their taste, pocket, time or available
ingredients better than the original, the origin of the
knowledge is heavily social, and as I suggested in
the last paragraph, what is done, and probably what is
known, may be fragmented rather than unified.
Piaget underemphasized, I think, the role of
apprenticeship to more knowledgeable others in the
construction of knowledge, and the possibility that
much of what we know is what anthropologists call
bricolage, or ad hoc works adequately but not
perfectly ragbags. We will return to these points
later.

Problems with the Piagetian model
Predictions can be derived from the equilibration

model and examined. Is there a need for thinking to
be consistent and free of contradictions? Does
conflict between different ideas lead to their
resolution into a better idea? Is there one universal
series of stages of thinking? There has been much
debate over exactly what the philosophical
implications of these predictions are and what would
be relevant behavioural evidence (Rotman 1977;
Boden 1982; Meadows 1983). The issue is very
complicated indeed. Performance on tasks is less
consistent than Piagets competence model would
appear to propose, the competence model has
problems predicting particular performance, conflict
leads to progress only sometimes, and equilibration is
not an adequate explanation of development (Bryant
1982; Gelman and Baillargeon 1983; Flavell 1982;
Sternberg 1984). Here as elsewhere in Piagets theory
there have been problems in translating his abstract
fundamental mechanisms into terms of measurable
behaviour, and all too often the behavioural evidence
has not supported the existence of the abstract
mechanisms.
Probably the best known aspects of Piagets theory
are the sequence of major stages and the accounts of
childrens behaviour when given problems of
conservation, classification, perspectivetaking, etc.
It is, however, precisely these that are now seen as of
relatively less value, and Piaget himself developed
other theoretical areas at the end of his life (Vuyk
1981). Although there quite obviously is a shift from
sensori-motor action-dominated cognition to
reflective abstract quasilogical cognition as
development proceeds, there is little experimental
evidence to suggest that it follows the picture of
discrete, integrated, general stages that Piaget
proposed. I have discussed the difficulties of stage
theories in Chapter 1; in general the evidence on
stages in cognition suggests first, that cognitive
performace is often inconsistent, varying across
tasks, materials and brief periods of time (see, for
example, Brown and Desforges 1979; Miller 1982),
and second, that correlations between tasks which are

said to require and be indicative of the same cognitive
structure are no higher than correlations between
tasks whose basic structures are quite different (e.g.
Meadows 1975; Klausmeier and Sipple 1982).
Flavell (1982) argues that we need to think carefully
about what sort of homogeneity and what sort of
heterogeneity we expect in cognitive development.
Theorists who have clung to Piagets stage model
have suggested that transitions between stages may
be longer and the differences thus less clear-cut than
was originally proposed (e.g. Beilin 1980); it may be
more useful to abandon the notion of general stage
structures for the moment, and pay more attention to
the variety of cognitive processes used in different
areas and at different times. It is this sort of approach
which seems to be proving fertile at the moment.
(Freeman and Cox 1985 suggest it for the study of
childrens drawings also.)
Piagets accounts of childrens behaviour have
turned out to be problematic in related ways. No one
disputes the brilliance and ingenuity of his
questioning, or that young children do indeed give the
bizarre answers he recorded on what have become
standard tests of concrete operations, though when
his theory first made an impact on English language
psychology in the 1950s a great deal of effort was put
into replications of his experiments to see whether
what he had observed really happened. What is now
disputed, and quite often refuted, is his account of
young childrens failure in terms of the inadequacy of
their logic (and indeed of older childrens success in
terms of their more adequate logic). Other
explanations have been proposed for many of the
failures which occur. One particularly illuminating
example is the case of the transitive inference.
Transitive inferences
Piaget took over transitive inference problems from

his early work on Binets intelligence tests. They
involve the combination of two pieces of relational
information to infer a third relation. Binets test items
were of the form Suzanne is taller than Emilie.
37
Suzanne is shorter than Claudine. Who is the tallest?

Piagets easier task (Inhelder and Piaget 1958)
involved three sticks of just noticeably different
lengths and the spoken and/or demonstrated
information that A is longer than B. B is longer than
C. The transitive inference question was Which is
the longest, A or C? Children over the age of about 7
would be quite sure that A had to be longer than C;
children under 7 said they did not know, that they
would have to compare A and C directly. In related
tests they also had difficulty with measuring tasks and
with seriation. Piaget attributed young childrens
failure on the task to their inability to integrate two
pieces of information and make the required
inference, and older childrens success to doing
precisely that.
Peter Bryant and Tom Trabasso suggested that both
success and failure might have different
explanations, and in a series of experiments
demonstrated both what such explanations might be
and a number of other rather important points for the
diagnosis of cognition. Bryant and Trabasso (1971)
argued, among other things, that failure to know the
relation between A and C might be due not to an
inability to infer that given A>B and B>C then
necessarily A>C, but to simply not remembering one
or both of A>B and B>C. They also argued that
success might be due to having attached the label
long to A and the label short to C. In their
experiments they ruled out both these possibilities by
training children on the longer/shorter relation of all
adjacent pairs of the series A>B>C>D>E (and testing
their memory of these pairs), and questioning them
on the crucial relationship between B and D, which
had never been seen together and which had each
been longer and shorter equally often. Children as
young as 4 were successful on the B>D comparison if
they remembered B>C and C>D: and in further
experiments they were also successful on
comparisons they had only been told about, not seen
for themselves. Memory difficulties are quite clearly
implicated in transitive inference failure though
not necessarily in all cases or as the only cause of
38
failure. Memory is a result as well as a cause of what

is known.
Further work by Trabasso and his colleagues
(Trabasso 1975; Riley 1976) suggests that the
making of a transitive inference is not necessarily
what children (or adults) generally do when faced
with a transitive inference problem. It would seem
that it is normal procedure to construct a mental
picture of an ordered array of the objects concerned,
particularly if they bear physical relations of height,
size, weight, etc., but also if they are related in degree
of characteristics such as niceness, and to scan it and
read off the answer to the question. Here, as in other
instances which we will come to later, it is not that
children fail to be logical while adults succeed, but
that both children and adults may sometimes use nonlogical procedures to solve logical problems.
Logic as a model for cognitive

development
This possibility is particularly interesting because
philosophers and psychologists from Aristotle to
Piaget have tended to assume that there was a close
relationship between logic and reasoning, even that
the rules of logic were the laws of reasoning.
Developments over the last hundred years, as new
forms of logic were invented and as psychology grew
away from its philosophical roots, have left this
relationship looking rather different from what was
assumed (Braine and Rumain 1983, Wason and
Johnson-Laird 1972). It is not clear whether
reasoning is itself non-logical but logic is the
normative standard for evaluating the validity of
reasoning, which the transitive inference data might
suggest, or whether reasoning is logical but errorridden, a more Piagetian position.
The development of childrens handling of logical
problems is obviously relevant here. Braine and
Rumain (1983), summarizing a substantial amount of
work on propositional logic and some on predicate
logic, suggest that there is a considerable
resemblance between the reasoning of children
entering school and that of older children and adults.

Some logical problems are handled correctly even by
very young children, and these are easy for older
people too, while some other logical problems are
handled incorrectly or inefficiently by most people
young or old.
A classic logical problem is of the combining of
pieces of information and testing the validity of
inferences from them. For example, we have a
statement of the form If p, then q which we assume
to be true (If it rains, then we get wet). This is
followed by another statement, which tells us
whether one of p, notp, q or notq, is the case at the
moment, and from this we have to draw a conclusion
about what else about p, notp, q or notq we can be
sure is also the case. Any reader who does not know
the answers may like to try the problems set out as
Figure 1, remembering that If p, then q must be
taken as unquestionably true.
Children as young as 6 can draw the correct
inference in the modus ponens problem, If p, then
q:p, therefore q; children and adults alike commonly
draw incorrect conclusions on problems such as if p,
then q: not p, therefore ? or if p, then q:q, therefore
? let alone on more complex inferences involving
chains of reasoning. Analysis of errors in terms of the
logical structures of problems, the language and
various context effects, has suggested that logical
If p, then q
reasoning involves utilizing a previously learned

inference schema to reach a valid conclusion.
Braine and Rumain (1983), summarizing the
evidence, suggest that some inference schemas are
learned very early, possibly through normal language
learning, some only late and through special training.
Using them involves getting the starting information
into the form of a suitable schema and working
through the schema, resisting errors from
misinterpretation of language or context or response
bias, and neither importing extraneous information
nor forgetting what is relevant. All these errors are
common in children and adults. Language may give
rise to particularly severe problems in young children
because the language-processing needed in logic
tasks requires purer analysis than they are used to. In
understanding a normal sentence (see Chapter 4)
listeners process the words but also use other
information available about the speaker, the world,
the subject matter and the conversation, a muchpractised and very skilled comprehending of
language-in-context. In understanding a sentence in
a logical problem listeners must not do this: instead
they must concentrate on the words and segregate the
information they give. What is important is what is
said, not how it is said, what might be intended, what
was said previously or what is generally known.
Logical problems require slow careful analysis at a
high level of verbal skill, a sort of analysis which is rare
If it rains, then we get wet
p, therefore ?
It is raining, therefore ?
not p, therefore ?
It is not raining, therefore ?
no conclusion possible (we

dont get wet from being rained on, but we may not
be wet or we may be wet from some other source
such as swimming)
q, therefore ?
We are getting wet, therefore ?
no conclusion possible (we

may be getting wet from rain or from some other
source such as having buckets of water thrown at
us)
not q, therefore ?
We are not getting wet, therefore
not p
we get wet
it is not raining
Figure 1
39
culturally and associated with scholarly or highly

literate pursuits. Young children, used to employing
context and general knowledge to understand
language, employ them too in logic problems.
There are other aspects of the ecology of logic
tests which have to be considered. One common error
in propositional logic is to accept a conclusion as
valid when the problem is formally undecidable. In
the conditional syllogism if p, then q: not p, therefore
? it is a common error to complete it by saying not p,
therefore not q (If it rains, then we get wet: it is not
raining, therefore we do not get wet). People faced
with undecidable problems like this tend to come to
an invalid conclusion, to be unwilling to say that one
cant tell whether q or notq is the case. There may be
various reasons for this, including expectations that
logic leads to clear conclusions and that testers do
not ask questions which have no answers, and
unwillingness to admit to the tester (an authority
figure) that one cant tell which in other situations is
an admission of ignorance. Similarly, understanding
what has to be the case and what merely may possibly
be the case is important. Logical necessity,
incompatibility and entailment are crucial notions
which are relevant to various issues in the areas of
language and logic. Childrens understanding of
these problems develops gradually (Russell 1978,
1982).
It would appear then that although some aspects of
logical reasoning appear very early in childrens
development, possibly being derived from ordinary
language use, there is a distinction between formal
reasoning, based on analytic processing, and
ordinary everyday reasoning which is often based on
intelligent guesses about plausibility. Guessing,
importing information, using common sense, are
strategies often brought to bear on formal reasoning
tasks. Disembedded formal reasoning appears to be
rare except in highly educated groups but is probably
quite easily taught (see Donaldson 1978). It is not, on
the whole, likely to be perfectly representative of
ordinary everyday cognition: a model which
encompasses a wider range of task variables and
thinker characteristics is necessary. It seems clear that
ordinary cognition is inseparable from
40
knowledge. While current research does pay more

attention to differences between tasks and between
thinkers, as we shall see, there are as yet no
satisfactory high level models which integrate
findings from different areas and approaches.
Information-processing approaches
A major contribution to the study of cognitive
development has been made by informationprocessing approaches. Work in this tradition
emphasizes precise analysis of how information is
recognized, coded, stored and retrieved, and because
computer simulation techniques are often used the
structures and processes involved in handling
information are relatively tightly specified and
testable. Information-processing studies typically
present people with a problem and examine what
information they select, how they store and organize
it, what models or hypotheses are involved, and what
cognitive processes they use to reach a solution.
Capacity, processes and knowledge
Pascual-Leone (1970) and Case (1978, 1984, 1985)

have proposed neo-Piagetian models of cognitive
development couched in information-processing
terms. They propose that what develops is, on the one
hand, a series of distinct executive strategies for
solving problems, and, on the other, the size of
working memory or the amount of mental space
available for information-processing strategies to
work in. It is certainly the case that the number of
things a person can do at one time increases as they
grow older or more expert in a task, and that
overloading working capacity interferes with
performance and may disrupt it (e.g. Baddeley 1976).
As we shall see when we discuss metacognition and
study skills, even quite young learners can recognize
the need to avoid distraction. It is not clear, however,
whether the developmental increase in the
functioning of memory and information-processing
is due to an increase in the capacity of working
memory or M-space. Pascual-Leone proposes that
M-space gets bigger with age and development, that
is it becomes able to handle more strategies or
programmes or schemes at the same time. The

persons repertory of strategies, etc., also increases
with age and development: each strategy may also
develop so that, for example, it becomes slicker and
more automatic. Pascual-Leone claims that changes
in strategies only account for improvement within a
stage of cognitive development: changes between
stages are due to increases in M-space.
It is a serious problem however to distinguish
between changes in strategies and changes in Mspace either in terms of behaviour or in theoretical
terms, because they necessarily interact. Childrens
performance on a cognitive task will be a function of
the strategy used, the demands which that strategy
makes on M-space and the size of M-space itself. We
know that the first two of these develop with age:
attempts to measure the size of M-space have to hold
strategy and strategy demands constant if they are to
distinguish between changes in the size of M-space
and changes in the way a stably-sized space is used.
Let us suppose that the transitive inference task
requires four units of processing space: one each to
store A>B and B>C, one to store the information that
longer than is a transitive relationship as opposed to
one such as loving which is not (Tybalt loves Juliet
and Juliet loves Romeo do not combine transitively
into Tybalt loves Romeo), and one to make the
inference. Children may come to be able to deal with
the transitive inference task by achieving an M-space
of four units. They may also come to be able to do it
by having their inferential strategy use up less space,
for example by learning the relative sizes of A and B
and of B and C so well that they require only half as
much storage space. Or they could be able to deal with
the task by using a non-inferential strategy such as
reading off from a visual image which needs less Mspace. Only very precise analysis of tasks, of learner
activities and of the interdependencies between
knowledge, strategies and processing can explain the
contributions to cognitive development of growth in
processing capacity, growth in the information
archives, greater sophistication of processing,
greater sophistication of the catalogue and crossreferences of the archives, and interaction among all
these. At the moment the best guess seems to be that
the first of these is the least important in cognitive

development, indeed that the limits of information
processing change little though there are enormous
changes in what goes on within them (Brown,
Bransford, Ferrara and Campione 1983). Cases
recent developments of his model take this line and
include experimental evidence which supports it
(Case 1984, 1985), though Sternberg (1983, 1984)
still argues for an increase in total capacity.
Information-processing approaches to cognitive
development assume that people are in essence limited
capacity manipulators of symbols (Siegler 1983, p.
129). Over the course of development the processes
which people use to manipulate symbols become
more complex, more accessible, more exhaustive, more
flexible and faster, in ways which may be analogous to
the changes which happen in the move from being a
novice to being an expert. There are also developmental
changes in symbol storage as more knowledge is
acquired and as it is organized differently. We will look
at some aspects of how children organize their
knowledge later. Recent information-processing work
is giving more attention to knowledge-rich domains
such asmathematics. Early work used artificialdomains
because they did not call on subjects prior knowledge
and thus complicate the picture with uncontrollable
variation between subjects. It is not clear how
knowledge and processes interact, notevenwhen the
knowledge is knowledge about processes (see Chapter
3, section on metacognition). Prior knowledge has been
shown to influence recall, increasing the frequency of
intrusion errors where recall includes information
which is correct and relevant but comes from what was
known before the experimental presentation of the
material to be learned. More prior knowledge may
sometimes produce more efficient processing, perhaps
because it reduces memory load to deal with
overlearned information, or perhaps because the
connections between items are more accessible, or
perhaps because the important characteristics of the
input are more obvious and their irrelevant aspects can
be more easily set aside. Knowledge of the task and of
41
ones own abilities are particularly important aspects of

performance (see Chapter 3, section on metacognition).
Mechanisms of cognitive development
It is not yet clear in what ways childrens cognition

comes to develop from immature cognitive processes
to more mature ones. Answers to questions of the
mechanisms of cognitive development are still very
tentative indeed; all that is clear is that it must be
mechanisms, plural (Sternberg 1984). Kail and
Bisanz (1982) outline the general features which will
characterize information-processing models of
cognitive development as they emerge over the next
few years. First, they will be concerned with how
information is given internal representation in the
organism. While it is agreed that much information is
represented in symbolic form, there is much debate
about what symbolic systems are used, and when
(see, for example, Cohen 1983 for an introductory
account of this debate). The problem of how external
stimulation is turned into internal representation is
also still a serious one. Developmental theories will
have to describe what features of complex stimuli are
processed by children of different ages. Changes

would seem likely to be related to the development of
the ways in which concepts are linked in childrens
knowledge (see p. 139 below). Information
processing theories assume that concepts are linked
in a network of associations. In the course of
development the network comes to include more
elements; superordinate conceptual links between
them become stronger and more important than
perceptual ones, and processing can involve larger
units of information. Kail and Bisanz (1982, p. 60)
illustrate this with the knowledge-of-fruit structures
of two children (Figure 2).
For the younger child, both apples and peaches are
characterized as fruit (an is a relation) but their
perceptual characteristics (is round, has a stone or
seeds) are stronger, and the child would say that
apples and peaches are alike because they are both
round. For the older child, in contrast, apples and
peaches are more strongly linked to the superordinate
category of fruit than they are to their perceptual
characteristics; the is a relation is more important.
Figure 2 Portions of the knowledge base concerning fruits. (a) The knowledge of a 5-year-old, for whom peaches and
apples are alike primarily because they are both round. (b) The knowledge of an 8-year-old, for whom peaches and
apples are similar primarily because they are both fruits.
Source: From Kail and Bisanz (1982), p. 60.
42
The organization of knowledge in real cases is

likely to be far more complex and hard to discover
and describe. A pioneering but methodologically
flawed study by Chi and Koeske 1983 seeks to map a
4-year-olds knowledge about dinosaurs. Repertory
grid techniques hold promise but are difficult to use
with children (Ravenette 1975; Fransella 1976).
The second aspect of information-processing
models of cognitive development which Kail and
Bisanz (1982) discuss is their account of changes in
the cognitive processes which generate, transform or
manipulate representations of knowledge. They
anticipate that research in this area will identify a very
large number of complex procedures specific to
particular tasks, domains or persons but a relatively
small set of elementary procedures underlying them.
I discuss some of the specific cases below. The
general developmental change is, they suggest,
towards an increasing use of rules which are
sufficient, that is they apply correctly to all cases and
do not need supplements or exceptions, rather than
rules which are only partially correct and apply
successfully only in a limited range of tasks or
settings. Sieglers analysis of the balance task
(Siegler 1981, 1983, 1984) is one example; the
history of science provides many others, such as the
replacement of Newtons model of gravity by
Einsteins. There is a similar developmental shift
towards using more efficient processing, to using
procedures which are more powerful and require less
role repetition and unskilled labour. Such
procedures need fewer resources of attention and
effort to run successfully, at least once they are
established (see Chapter 3); using them thus frees
attentional resources for other bits of processing.
Finally, development in cognitive processes may be
an increase in the speed of the processes used rather
than in what the processes themselves consist of.
How might such changes come about? How do we
manage a development, continued over many years,
which gives us an increasingly full and flexible
conceptual structure, increasingly efficient and
sufficient cognitive processes, and an increasing
availability of attentional resources? Even if we

reduce the scale of the change by allowing, on the one
hand, that very young children are cleverer than some
theorists had supposed, and, on the other, that adults
cognition is often enough inefficient, insufficient and
an unsystematic sort of bricolage or botching
together, rather than brilliant craftsmanship, the
phenomenon of cognitive development is worth
admiration. What sort of system would allow such
changes?
The explanatory framework Kail and Bisanz
(1982) propose has two general components. The
first is increases in attentional resources, which I
discussed above. They see increases through growth
in the total resources available (cf. Pascual-Leone) as
being compatible with increases without total
growth, due to better automatization and chunking of
information (cf. Case), though as we saw it may be
hard to distinguish between them. Their second
general component is a set of processes which modify
the knowledge base. During conceptual development
new elements and characteristics are added to the
knowledge base, that is there are new nodes in the
associative network. Some old nodes are deleted, and
there are new or changed links between nodes.
Among the modifying processes are addition/
deletion processes and processes for strengthening or
weakening links. There are also important processes
for determining that change is needed. They suggest
that these involve the detection of inconsistencies,
that is there are processes which compare processes
with other processes or with goals or external events.
An obvious example is what happens inside a child in
a conservation test, when judgement on the basis of
the changed appearance of the array is inconsistent
with judgement based on the knowledge that nothing
has been added or taken away. A need to resolve
inconsistency is, of course, at the core of the Piagetian
notion of equilibration. Inconsistency detection is
not, however, the only monitoring process Kail and
Bisanz include in their account. Another analysis
involves detecting regularities which recur and
demand resources. If there is repeated use of the same
sequence of processes, or if the activation of one node
43
in the concept network always involves the activation

of the same set of representations, it may make fewer
demands on resources to modify the knowledge base
and set the recurrent regularities up as a sort of
package which can be called up en bloc. Detection of
regularities may also be useful in detecting
redundancies and so streamlining processes. It may
also show up occasions when different processes
have the same result, thereby enhancing the
possibility of higher-order organization of processes,
and perhaps also increasing ones confidence in the
correctness of the result the processes have reached
(Bryant 1982).
Monitoring processes are seen as being heavy on
resources, however. Thus there may not be any
developmental change, even though an inconsistency
or a recurrent regularity has been detected, unless
sufficient attentional resources are available. If total
resources available do increase with growth, growth
may make the use of monitoring processes possible
where they could not be implemented before.
Resources also become available through the
automatization of content in the knowledge base.
Automatization is to be associated with repeated use
of processes and with the strengthening of links
between conceptual nodes and hence should, I would
suppose, be positively related to the detection of
regularities and consistencies. As I read their
account, it would not seem to be as closely related to
the detection of inconsistencies and conflict, so it is
possible that in Kail and Bisanzs model,
development through confirmation, practice,
positive feedback and more polished and efficient
performance of established processes may be easier
than development through detecting inconsistencies
and conflicts, and enforced change in processes or
knowledge organization. This reading goes against
the Piagetian emphasis on conflict-led equilibration,
which I argued against at the beginning of this
section; it accounts quite nicely, however, for various
observations, such as peoples difficulty in giving up
a practised but insufficient strategy for a better one
(e.g. Brown et al. 1983), Bryants arguments about
44
conflict and confirmation in conservation (Bryant

1982) and the difficulty of negative instances in
concept formation and scientific problem solving
(e.g. Bruner, Goodnow and Austin 1956; Wason and
Johnson-Laird 1972).
The final point Kail and Bisanz make about their
knowledge modification processes is that they
interact with the resources available, and each
modifies the other. Changes in the knowledge base
alter the ways in which the cognitive system
investigates or interprets its environment, which in
turn alters the internal and external feedback
monitored by the regularity and inconsistency
detectors a formulation very reminiscent of the
Accommodation and Assimilation model, though
perhaps potentially at a more specific level. Thus
changes in the knowledge base may allow the system
to identify inconsistencies and regularities which
were previously undetectable an account of the
phenomenon of sudden insight, perhaps?
I have dealt with the Kail and Bisanz model at some
length because it seems to me to be an unusually
intelligible account of what sort of account the
mechanisms of cognitive development may be
expected from the information-processing approach.
It is not the only model in the field; Sternberg (1984)
contains six more, with Flavells comments on each.
It does have some notable omissions or
underemphases which need some further
elaboration. One of these is the area of the
accessibility and difficulty of cognitive processes,
which is likely to be of particular interest to people
concerned with individual differences but will also
have to be dealt with by models of general cognition.
Another, which I shall discuss at more length
(Chapter 3), is the currently lively area of executive
control, metacognition and strategic thought. The
underemphasis is surprising, since information
processing in computers is one main source of the
executive control metaphor. It is quite clear that in
cognitive development thinking becomes more
strategic and more controlled by the thinker, as we
will see in areas such as attending, remembering, text
processing and problem solving discussed in Chapter

3. Children also know more about cognition as they
get older, and this too is a flourishing area of research.
Models of cognition will have to account for
metacognition, though the interaction of the two is
unlikely to be simple: it is quite clear that one can
know but not do, and do without knowing, as well as
the two consistent positions of know and do, not
know and not do.
I have described three different models of
cognition, the Piagetian, what one might call the
logicians account, and information processing. The
second of these is not a contender as an account of
how we reason in general; rather, logical problems
seem to be a restricted domain where general
processes have to be used in rather specialised ways.
Piagets model remains an immense intellectual
achievement, worthy of great admiration. Some of its
features now seem to be rather seriously dubious (e.g.
automatic equilibration, the groupings model), and
some need to be given a much more specific form
(e.g. Assimilation and Accommodation): some, like
the tendency to play down the possibility of
development through social interaction and being
taught, have distorted our view of childrens thinking
and education and need to be reversed. The
information-processing approach has not yet
achieved a satisfactory model of cognitive
development. It has produced some very interesting
accounts of processes and rules involved in solving a
number of problems, some of which were derived
from Piagets tasks (see, for example, Klahr and
Wallace 1976), and I would expect there to be
considerable progress in the next few years on higherorder modelling of general cognitive systems.
However, there are certain reasons for caution rather
than optimism about this approach. One important
problem is the difficulty of observing thinking: there
is a tendency for some work in this area to base its
model on data which are only very distantly related to
the observable behaviour of real children. This was a
problem for Piaget too, of course, and it still needs to
be guarded against. It is, of course, a perfectly valid
choice to concentrate on providing a model of
idealized competence rather than
actual
performance; competence models may be interesting
in themselves, and may illuminate performance.

What they do omit, perhaps inevitably, is, of course,
much consideration of the context of cognition.
Questions of why cognition proceeds as it does, how
it is shaped and constrained and fostered in
development, how biology and social experience are
involved, seem very important to me, and hard to deal
with in the traditional information-processing terms
computers do not have much in the way of ecology.
Some philosophers argue anyhow that the computer
metaphor is seriously inadequate in that it sets aside
as negligible the roles of consciousness, experience
and intention (Searle 1984; Russell 1984). I can do no
more here than refer back to the second sentence of
this chapter there is still no clear, complete and
valid account of what adults do when they think,
understand, etc. and add that there is none for
children either, let alone an account of the change
from child to adult. It is, however, the case that we are
increasing our understanding in some more limited
ways, as I shall go on to discuss.
Infants perception and cognition

I must preface my brief account of perception and
cognition in infancy with two remarks to guide the
reader throughout the rest of the discussion. The first
is that I am not myself involved in research on infants:
I have thus had to rely on the work of others to an even
greater extent in this section than elsewhere. The
second is that the degree of controversy in infant
perception and cognition seems to be higher than in
most other research areas: there is profound
disagreement not just about ideas or interpretations
but about behavioural facts. Do infants imitate facial
expressions, babble more if given contingent
reinforcement, search for an object that has moved
out of sight? Yes, say researchers on one side: no, we
cannot replicate this finding, say researchers on
another. Non-replications such as these are not
always attributable to methodological changes or age
differences, or even to experimenter bias:
someremain mysterious, and the experts as yet
cannot resolve them. A non-expert, I have not tried to
45
Plate 5
46
do more than point out issues and indicate useful

sources. Most of the material and evaluation in this
section derives from the second volume of the
Handbook of Child Psychology (Haith and Campos
1983). I will begin by discussing what is known of the
infants perceptual capabilities (visual, auditory and
olfactory, and some aspects of brain development)
and then proceed to the cognitive development that
ensues. Throughout I would wish to emphasize that
ultimately we should link development at
anatomical/physiological levels with development at
behavioural levels, and explain developmental
changes with reference to their functions as part of the
evolutionary push and the environmental pull.
Our present understanding being distinctly limited,
however, we will have to be cautious about the nature
of these links, and particularly about their causes and
directions.
Seeing and hearing

The hardware of vision
Banks and Salapatek (1983) review infant visual

perception. Human infants are born with eyes that
are moderately mature in anatomical and
physiological terms, more mature than cats, for
example, and less mature than macaque monkeys
(Figure 3). As in adults, light passes through cornea,
lens and the body of the eye to the retina at the back
of the eye. One layer of the retina is of light-sensitive
cells, the rods and cones: when light reaches
these receptors, signals are sent via the other layers
of retina to the optic nerve and thence to successive
parts of the brain. Compared to adults, the infants
eyes are relatively short from front to back, there is
less possibility of adjusting the focal distance of
their lenses, and there are immaturities in the retina.
In particular, in the adult eye there is a central area,
Figure 3 Cross-sectional drawings of the adult and newborn eye. The adult eye (a) is a horizontal section of the right
eye. Important structures are labelled. The visual axis is represented by the broken line and the optic axis by the solid
line. The newborn eye (b) is also a horizontal section of the right eye. It is drawn to scale to represent its size relative
to the adult eye.
Source: From Banks and Salapatek (1983).
47
the fovea, which differs in cell type and in sensitivity

from the rest of the retina. In this area visual acuity is
better than in the periphery; as we are all aware, we
normally see objects more clearly when they are in
the centre of our visual field than when we use our
peripheral vision. The newborn baby has a much less
differentiated fovea, and therefore probably does not
see shapes and contours as clearly, or discriminate
them as finely, as adults can using their foveal vision.
The retina develops quite rapidly after birth and is
anatomically mature by about the end of the first year.
Babies vision gets more acute quite rapidly as
developments in the retina (and better control of the
lens) are accompanied by developments in various
parts of the brain. In the next major structure in the

ascending visual pathway (Banks and Salapatek
1983) (see Figure 4), the lateral geniculate nucleus,
which is essentially a relay station between eye and
cortex, there are post-natal anatomical changes, as the
neurons grow, and functional changes, as response to
stimulation becomes less sluggish and less easily
fatiguable, and as visual acuity increases in the cells
serving the developing fovea, the nerve pathways
connecting structures slowly grow a sheath of myelin,
which allows better transmission of signals. In the
visual cortex itself, structural development, which
began before birth, continues fora considerable period
Figure 4 Major pathways from the eyes to the central nervous system. Fibres of the optic nerve from the temporal halves
of each retina remain on the same side of the head; that is, they project to the ipsilateral hemisphere of the brain. Fibres
originating from the nasal halves of each retina cross at the optic chiasma and then project to the contralateral hemisphere.
The lateral geniculate nuclei are part of the thalamus. The superior colliculi are part of the mid brain. The striate cortices
are part of the cerebral cortex.
Source: From Banks and Salapatek (1983), p. 446.
48
after it. Evidence from studies of kittens shows that

although capabilities often appear before they are
needed, and some structural organization emerges
without needing experience, much later-appearing
organization and visual capability is guided by visual
experience: neither cats nor monkeys (nor, possibly,
humans) develop the normal population of cortical
cells which deal with visual input in terms of
binocularity unless they have had experience of both
eyes seeing together. Again, animals raised in
environments dominated by lines at a particular
orientation develop a disproportionate number of
brain cells sensitive to that orientation (Banks and
Salapatek 1983, pp. 4589).
The mechanisms of vision
Eyes are often compared to cameras, but there are

significant differences in how they work. For
example, prolonged stimulation leads to fatigue: if an
eye is immobilized so that light falls on the retinal
receptors for a prolonged period, they cease to
respond to it until the input light changes. Further, the
central foveal area of the retina is more sensitive than
the periphery. These characteristics mean that in
order to see something clearly we have to move our
eyes: in one sort of way to move the retinal image
enough to stop it fading, and in another sort of way to
get the image on the fovea rather than the outer parts
of the retina where our acuity is less. Furthermore,
since we have two eyes, we have to co-ordinate them
and their movements: and compensate for
movements of our heads and bodies: and for any
movements of what we are looking at. Adults manage
to do all this with only occasional difficulties: how do
these skills develop?
The saccadic eye movement system of rapid
changes in fixation, which relocates targets first seen
in peripheral vision on to the more sensitive fovea, is
functional at birth, provided the peripheral target is
not too far off the centre. The effective visual field
which will elicit saccades, so that the eye moves and
the target comes to the centre of the retina, grows
postnatally, and so does the accuracy and speed of the
eye movement. Eye movements which compensate
for the observers movements are made via various
reflexes in adults. Some of these reflexes are present

in newborns, including blind infants, but not until
about 3 months of age are they well coordinated with
other movements, or accurate. Eye movements
which follow a moving object are also possible at
birth though they are jerky like saccades. Smooth
visual pursuit appears first when the target object is
moving slowly, but by about 12 weeks infants can
track even fairly rapid movements quite smoothly.
Human eyes, like cameras, can only be focused for
one viewing distance at a time. Objects at this
distance are in sharp focus, objects nearer or further
away are blurred. Adult eyes adjust automatically
and accurately, largely by changing the curvature of
the lens (accommodation). Newborn infants do not
do this with any accuracy; like a pinhole camera they
have a fairly good depth of focus but at longer and
shorter target distances than this range they see less
clearly and do not accommodate well. There is a rapid
improvement in accommodation over the first three
months of postnatal life.
Infants able to support themselves sitting or
standing use visual feedback as information not just
about the movement of other objects but also about
their own movements. If we move forward,
particularly if we fall forward, things initially near the
centre of our visual field move outwards towards its
edge and things initially near the periphery move
outwards out of sight. Conversely, if we see a
centrifugal visual field, we may feel we are falling
forward (a feeling Stanley Kubrick used in his film
2001). Experiments with babies (e.g. Butterworth
and Hicks 1977) showed that they felt like this when
the room wall in front of them moved forward, and
adjusted their posture appropriately, swaying or even
falling in the direction of the rooms motion. We do
not yet know how much experience of ones own
motion is necessary for this.
Looking at objects and patterns
I have outlined the basic anatomy, physiology and

mechanisms of the infants visual system: we have
seen that even very young infants can detect and look
at objects, but that only postnatal development allows
49
babies to see as clearly as adults do. As far as acuity is

concerned, to begin with babies have difficulty
detecting stimulus differences which convey detailed
information about patterns. The rapid development
of the fovea is one area of improvement, but beyond
that there are probably difficulties due to limitations
in the nervous systems ability to process the retinal
image. There is extensive evidence that visual acuity
improves dramatically over the first year of life:
neonates seem only to attend to large objects,
apparently not discriminating their details, while
older babies notice fine detail much more readily.
A most useful paradigm has been developed for
investigating infants ability to discriminate between
patterns: pioneered by Fantz, it is the visual
preference paradigm. Two patterns are presented
simultaneously to the infant and he or she is observed
to determine whether the two objects are looked at for
different lengths of time. Provided controls for
position preferences, etc. have been properly
implemented, if one object is looked at significantly
longer than the other the infant must have
discriminated some difference between them, and
may perhaps be said to prefer the one looked at more
(if we assume that infants are not given to mortifying
themselves by looking more at objects which are
liked less). Indices like smiles may also be used. The
habituation paradigm is also used for investigating
visual discriminations: habituation indicates that the
pattern is seen as being the same, dishabituation that
a difference is discriminated.
There is now an extensive literature on infants
preferences in patterns (though little on older
childrens or adults). Other things being equal, infants
tend to prefer red and blue to green and yellow, some
pattern to no pattern, curved lines over straight ones,
concentric patterns over non-concentric ones,
symmetry to asymmetry, and so on. Some of these
preferences are no doubt due to hardware and
mechanics properties of the visual system: some
have been explained in terms of a general preference in
the infant for complexity. While such an explanation
has intuitive appeal, it has proved hard to define and
measure complexity (Banks and Salapatek 1983, pp.
497506). It is clear that predictions work best,
however, when based on a model of preference which
50
combines the infants perceptual/cognitive state and

the stimulus characteristics: infants of different ages,
with different capabilities and different knowledge
prefer different patterns (Kagan et al. 1978).
In one instance of pattern preference, evolutionary
adaptiveness and ecosystem demands both seem to
be potential explanations and contributors. Infants
look at, and smile at, faces. There is controversy over
how they process the complex information that a face
provides (Harris 1979, 1983) though detailed
observation has shown agreement on exactly what
they attend to. Under about 2 months old, infants
looking at a face look most at areas of high contrast
such as its outside border. Next they attend most to
features, particularly the eyes, and may differentiate
between artificial faces in terms of eyedness
(Kagan et al. 1978). Then a face comes to be attended
to as a whole, and infants develop ideas about the
relative familiarity or novelty of faces (Fagan 1976).
A preference for faces which leads to more looking at
them would both give the infant more opportunities
for social learning and increase his or her
attractiveness to caretakers (see Chapter 6). Parents
commonly say that they begin to regard their baby as
a real person once mutual gaze and smiling has
begun. Infants who do not look at your face, and
anyone whose facial expression does not respond to
yours, are unattractive or even aversive stimuli. On
the whole, infants find blank faces less attractive than
responsive ones, as adults do.
Infants scanning
Eye movements are necessary for seeing. As we

look at a scene, saccadic eye movements continually
move the fovea from fixation to fixation, from one
feature to another. Observation of where these
fixations fall, usually by photographing reflections
on the cornea relative to the pupil, indicate what
features are being looked at in what order. Although
adults can certainly get information from peripheral
vision, and infants may do too, studies of foveal
scanning do illuminate how information is gathered
from the stimulus.Bank and Salapatek (1983)
review their develpoment (pp.50715).
Figure 5 Scanning patterns of newborns. The left portion shows scanning patterns when newborns were presented a
homogeneousfield.Eachdotrepresents eye position at one timesample.Thelines connectingthe dots simply connecteye
positions at adjacent time samples. The vertices of the triangles represent infra-red marker light positions. Thus, the
trianglesshown wereactually not present duringthese trials.Theright portion of this figureshowsscanningpatterns when
newborns were presented a solid black triangle on an otherwise homogeneous field. The outer triangle on each record
represents the triangles contours. The vertices of the inner triangles represent the marker light positions.
Source: From Banks and Salapatek (1983), p. 511.
Given a simple vertical or horizontal line or edge to

look at, newborn infants fixate more in the region of
the contour than in the rest of the visual field,
particularly if it is a vertical contour. This is probably
because it is easier to make horizontal eye
movements across a vertical contour, as infants do,
than vertical eye movements across a horizontal one.
Given a figure to look at, newborns tend to fixate a
point which has a high degree of contrast (Figure 5).
In the left-hand side of the figure, newborns were
looking at what to them was a uniform visual field,
and they made no consistent fixations. In the righthand side, the newborns looked at a solid black
triangle on an otherwise uniform field, and fixated
one corner of the triangle. Older infants began to scan
more widely, for example along the edges of the
triangle, but not randomly. Their scanning came to be
concentrated on areas of maximum information, as
that of adults is. There is an increasing degree of
perceptual organization. Cognition has become
inextricably mixed with perception.
Auditory perception
I will describe the development of hearing more

briefly than I did the development of vision.
Discussion of infants speech perception may be
found in the chapter on language development. Aslin,
Pisoni and Jusczyk (1983) review auditory
development in infancy and were the main source for
this section. They point out that despite much recent
relevant research, many basic questions about the
development of auditory function remain
unanswered, and indeed our understanding of
hearing in adults is much less complete than our
understanding of vision.
So far as the outer ear is concerned, the main
difference between the adults ear and the infants is
that the latter is smaller in size throughout. This may
lead to the infant hearing best sounds which are at a
slightly different frequency range from the sound the
51
Figure 6 Cross-section of the human ear showing the three major divisions (outer, middle and inner ear), their mode of
operation, and their presumed function.
Source: From Aslin, Pisoni and Jusczyk (1983), p. 586.
adult hears best, sensitivity possibly being shifted

towards higher frequencies. Because sound sources
are localized using differences in the times at which
the sound reaches left and right ear, the fact that the
infant has a smaller head between his or her ears than
the adult means that the time difference is smaller and
infants may therefore have more difficulty localizing
sounds. Bower (1974) points out that because their
heads are growing the inter-ear distance of infants is
52
changing and they will need to adjust to this

continuing change when localizing sound.
The functioning of the middle ear and the inner ear
is complex and not perfectly understood. Much of
middle ear structure seems to be adultlike at birth, but
its smaller size may shift sensitivity towards higher
frequencies. Anatomical comparisons between adult
and infant inner ear structures might also lead to
predictions of sensitivity to high frequencies
developing before sensitivity to low frequencies, but

research studies have not yet tested these predictions.
What functional differences there are is very unclear:
certainly infants and children from 5 months old
onwards are better at detecting high frequency
sounds such as bats squeaks than people with ageing
ears but this may be because the latter have lost their
sensitivity. Infant insensitivity to low sounds has not
been demonstrated.
The complexity of brain involvement in hearing is
very great. Research on cats and rats done in the last
decade or so suggests that there are big postnatal
developments of many different sorts in these
animals. There is little evidence on the functional
development of the auditory brain in humans, so we
do not know whether postnatal changes in sensitivity
and responsiveness to sounds are due to an
improvement in perception or in neural connections
and interactions.
Pregnant women have long since reported that
sudden loud noises led to sudden movements of the
foetus in the womb. Recently, convincing evidence
has been presented that the foetus moved because it
heard the noise, not merely as a response to its
mothers reaction to the noise (Aslin et al. 1983, pp.
6023). The infant in utero can hear noises from
outside, particularly perhaps loud noises at low
frequency, and the ambient noise in the womb turns
out to be fairly loud (85 decibels or so). Recodings of
these ambient rhythmic noises have, of course, been
quite a commercial success as lullabies, marketed as
soothing noises to be provided to the baby who has
lost them by being born.
From at least a few weeks before birth, then, infants
can hear, but the threshold of stimulation needed to
produce a noticeable reaction is higher than adults
need. Newborn infants can do some crude
localization of sound sources, orienting reliably to
sounds at 90 degrees left or right of their midline, but
their reactions are slow, and they do have problems
controlling their head movements. Older infants, 1 to
3 months old, may turn their heads less towards a
sound source, but after this period orientation to the
sound picks up again. Babies can fairly easily be
conditioned to turn their heads to one side in response
to a sound, particularly for a reward of a pleasant sight

or food, as in the classic work by Papousek (1967).
Animal evidence suggests that binaural experience is
necessary for sound localization, much as binocular
vision seems to be necessary for normal visual
development.
Summary
So far as the best-studied senses, vision and hearing,

are concerned, it is a simplification but not a
falsification to say that even newborn babies have a
capacity to function which is not very different from
that of an adult. Developments in the brain, and
growth of knowledge, account for most of the
developmental changes in perception: developments
in the peripheral organs or perception seem to be less
important.
Development in other sensory modalities
I have briefly described the development of vision

and hearing in the two previous sections. There has
been much less work on the other senses; indeed they
are much less emphasized by people generally. Not
being able to see or hear is regarded as a significant
handicap, but having no sense of smell or taste is not
felt to be more than a trivial disadvantage. While it is
certainly true that human beings make much less
conspicuous use of smell, taste and touch than of
vision and hearing, it would be wrong to assume that
these other senses are unimportant, or that they play
no part in the development of behaviour.
Very little can be said at present about the
development of taste and of touch. Some
discrimination between sweet and salty taste is found
at birth or even in utero (Crook 1978), and babies are
said to discriminate between breast milk from
different mothers (Macfarlane 1976). It has been
suggested that there are developmental shifts in
liking and disliking the taste of later foodstuffs,
reputedly that young children positively enjoy the
cod liver oil which adults find disgusting. How clear
such changes are, and how they come about, is not
known. Sensitivity to touch is present from birth:
infants needing to be nursed in incubators thrive
much better lying on a soft fleecy surface than on a
53
smooth hard surface (Harlows infant monkeys also

preferred a soft surrogate mother to a hard one which
fed them). Skin-to-skin contact with the mother
directly after birth has been claimed to be an
important contributor to motherinfant bonding
(Klaus and Kennell 1976), but there is reason to doubt
its necessity (Sluckin, Herbert and Sluckin 1983; and
Chapter 6, this volume).
Recent work has begun to illuminate the
developmental importance of the sense of smell.
There is not a great deal of evidence from humans,
although Macfarlane (1976) showed some time ago
that neonates can discriminate between the smell of
their mothers breast milk and that of another mother.
Work on mice however indicates that their
maturation rate is affected by their olfactory
experience. Specifically, female mice exposed to the
odours of natural secretions from male mice reach
puberty earlier than average (and may possibly,
though there is some controversy about this, be larger
in size), while female mice exposed to smells from
adult females mature sexually more slowly than
average (for a review see Johns 1980). Behaviour and
morphology too are shifted, towards the norm for
male mice in females exposed to male secretions, and
in the female direction for mice exposed to female
secretions.
Prenatal exposure to hormones also affects
behaviour. Female mice who were adjacent to males
in utero (mouse foetuses are lined up like peas in
pods) tend to behave more like male mice, and male
mice who were between females in utero behave
more like females. There may have been a similar
result in humans in the cases where female foetuses
were exposed to high levels of androgen prenatally.
In childhood and adolescence these girls were said to
be tomboyish in their behaviour and masculine in
their outlook (Money and Ehrhardt 1972). Their
parents knowledge of the congenital syndrome
which caused excessive productions of androgens
and its likely effects complicates the picture so that
conclusions are very hard to draw (see Huston 1983;
and the section on sex differences in Chapter 5, this
54
volume). The evidence on whether mothers who

were treated with oestrogens or progesterone when
they threatened to miscarry early in pregnancy have
feminized offspring is also not clear (Ehrhardt and
Meyer-Bahlburg 1981). Measures of sex-typed
behaviour and attitudes in humans are problematic,
and sex differences due to social and cultural factors
could be argued to be far more important than any
hormonal base. These difficulties would appear not to
be so great in mice: the mother mouse (or foster
mother) probably does not treat her infants
differently on the basis of the hormones they were
exposed to prenatally. It is hard to interpret the
differences in mices behaviour, growth rate and
anatomy as the result of cultural influences!
So far as the development of olfaction is
concerned, then, there seems to be an interesting
possibility that some naturally occurring substances
may affect development through the olfactory
pathways. Further research is needed.
The last sense I want to mention is the kinaesthetic
sense, which tells us about the bodys movement and
position. Receptors in muscles, tendons, joints and
elsewhere send messages about their activity to the
cortex; we perceive the relationship between these
messages in a pattern over time which depends on the
size, direction and speed of body and limb
movements, as well as on the resistance to movement
that we encounter. We do not know a great deal about
how kinaesthetic perception develops. Messages
from muscles, etc., are emitted automatically, but
what they are understood to mean depends on
experience. Infants, for example, spend quite a long
time watching their hands and learning to correlate
visual and kinaesthetic information. Deliberate
skilled movement requires the use of kinaesthetic
feedback; children who are clumsy and poorly coordinated often seem to have particularly poor
awareness of their posture and movement (Laszlo
and Bairstow 1985). Kinaesthetic awareness, and
skilled precise motor action, take practice as a major
contributor to their development.

Some aspects of brain development in infancy
I emphasize again my assumption that mind and

behaviour depend on brain and body, that without the
latter the former would not exist. Nevertheless, there
are real problems in relating brain and behaviour,
which are particularly sharp in the study of
development. In most of the cases where we find a
change in brain and a change in behaviour which
coincide in time, we may wish to infer a causal link
between them, but there is rarely evidence to test the
inference of causality. The central empirical problem
is how much and for how long do the two causal
chains below occur.
genetically predetermined
environmental
brain development
stimulation
new behaviours
new behaviours
more complex interactions developmental changes in

with the environment
the brain
The answer is probably different at different stages of

maturation: for example, foetal development before
thirty-six post-menstrual weeks, about four weeks
before term birth, is not much affected by sensory
input, because of limitations in foetal sensory
receptors and transmission to the cortex.
Subsequently, development shifts towards a
predominance of environmental factors in
behavioural and cognitive development: genetic
programming does not explain a great deal of the
variance in behaviour such as reading, chess playing
or ability to run marathons. Even if there is
neurological dysfunction or damage, such as cerebral
palsy, cognitive skills can develop in a strikingly
normal way, given a favourable environment.
Recovery from brain damage at birth is very much
positively correlated with how good an environment
is provided for the infant and young child (Shaffer
1985).
Brains consist of millions of nerve cells, intricately
interconnected, and making up various structures
which are involved in functions ranging from
controlling the breathing, sleeping and waking of the

organism to abstruse cognitive, linguistic, affective
and spiritual activities. Much of the development of
cells and structures originates prenatally, and
throughout early life the brain is nearer its mature
state than any other part of the body (Tanner 1978).
There is however substantial postnatal development
of many anatomical, neurophysiological and
functional features of the brain. Parmelee and
Sigman (1983) review recent work: I will sketch the
extent of the work and indicate some interesting
developmental findings.
Brains contain two sorts of cells, the neurons or
nerve cells, which transmit impulses, and the
neuroglia, which form a support system for the
neurons both by forming a guide for their
development and by supplying substances needed for
the manufacture of nerve fibres and their insulation
and of neurotransmitters. A neuron has a nucleus and
surrounding cell body, like any other cell, but part of
its substance is drawn out into long threadlike
processes called dendrites which branch intricately to
make many connections with dendrites from other
cells (Figure 7).
There are about a million million (1012) neurons in
the brain, and a cortical neuron has about 30,000
nerve processes from 3000 or so other neurons
connecting with it (Tanner 1978, pp. 1056). By the
time a foetus is eighteen weeks old, its developing
brain contains almost as many neurons as this, though
not the full mesh of interconnecting dendrites:
connections between neurons continue to develop for
a long time after birth. In rats and kittens
environmental stimulation after birth can increase
dendritic growth and interconnections. Neuroglia
begin to appear a little later than neurons, and new
ones continue to develop until the second year of life,
accounting for some of the postnatal increase in the
brains weight. Postnatal development also involves
the formation of an insulating sheath of myelin along
the neurons processes.
Certain infections in the mother, or deficiences in
her diet, may lead to the foetus developing fewer
neurons, and hence also fewer dendritic branches and
fewer connections with other neurons. Behaviour
may be disturbed and intellectual retardation result.
55
Figure 7 A typical Golgi-stained pyramidal neuron whose cell body is in Layer 5 of the cortex. The dendrites (apical,
oblique, basal) receive synaptic input from other nerve cells. The axon conducts the neurons output to other nerve cells.
The enlargement presents a segment of the apical shaft as it might appear in the electron microscope. At high magnification
(30,000 to 40,000 power) the dark projections from the dendrite appear as protoplasmic extensions called spines. These
spines form synapses with the vesicle-filled axom terminals,or boutons,from other neurons.Synapses on the cell body do
not utilize spines and hence cannot be visualized with Golgi-staining techniques.
Source: From Parmelee and Sigman (1983), p. 99.
56
Other genetic or environmental difficulties may lead

to malfunctioning of the brain. Among the best
known of these is the genetic defect which causes
phenylketonuria (PKU). This means that the
sufferers body cannot metabolize the amino acid
phenylalanine. Normally phenylalanine from protein
is converted into tyrosine by an enzyme called
phenylalanine hydroxylase: this enzyme is defective
in PKU so phenylalanine from the diet builds up in the
body and cannot be got rid of. Developing nerve cells
are vulnerable to it: it strips off the dendritic spines,
breaking down the connections between neurons, and
severe behavioural disturbance and mental
retardation result. If the infant is fed on a diet which
does not contain phenylalanine, no excess is built up,
and the brain can develop more or less normally. The
genetic defect is still there, but it has no effect in the
special environment where there is little or no
phenylalanine. Once brain development is complete,
the individual with PKU is not harmed by a normal
diet. However, if a woman with PKU, whose own
retardation has been prevented by a restricted diet in
infancy and childhood, takes an unrestricted diet
during her pregnancy, the foetus may be born with
severe mental retardation even though it has inherited
the normal gene for phenylalanine hydoxylase from
its father. The build-up of phenylalanine in the
pregnant woman does not affect her because the
mature nervous system is relatively resistant to it, but
it swamps the neurological development of the
foetus. The child is born with PKU mental
retardation, although it does not itself have PKU
(Konner 1982, Scarr and Kidd 1983).
Brain development and environmental pollution:
the case of lead
High dosages of certain substances, for example

mercury and lead, are known to poison the nervous
system. Some of these substances are, at a low level,
common environmental pollutants. It has been
suggested that their pervasiveness may be such that
they cause brain damage, particularly in the
developing brains of children. Lead is perhaps the
best known of these poisons/ pollutants and so there

has been some public concern about its effects.
Petrol has been thought to be an important
contributor to environmental lead levels since lead
use in cosmetics, in canning and in paint has declined
in recent years. Airborne lead levels are high near
roads with heavy traffic, and lead dust also
contaminates the people, clothes and food it falls on.
Children probably ingest a considerable amount of
lead by licking dirty fingers, dropping and eating
sweets and so forth, in addition to the lead inhaled in
petrol fumes. Plants grown in contaminated soil may
take in lead from it, and lead dust is not altogether
easily washed off leaves. The evidence on what
percentage of the lead found in childrens blood
samples comes from petrol is circumstantial, but it is
certainly a major contributor and probably the major
one (Rutter 1983).
High levels of lead in the blood or the teeth are
associated with intellectual impairment, behaviour
problems such as hyperactivity and, at the highest
levels, gross neuropathology. The severity of these
effects is greatest at highest lead levels, and decreases
as they decrease. Until recently it was thought that
lead levels below 40 micrograms (g) per 100 ml in
the blood had no adverse effects. Recent work
reported in Rutter and Jones (1983), however,
suggests that levels lower than this were associated
with impaired psychological function, such as
systematic in-attention in the classroom and lower
verbal IQ. Measurement difficulties, and differences
between different studies, mean that this association
at low levels is not certain, but Rutter (1983)
summarizes the evidence so far as showing a
consistent small effect of low levels of lead on
psychological functioning. Harvey (1984) regards it
as non-proven.
Interesting problems arise over the cause of the
association. There is evidence on the
neurophysiological damage caused by lead
poisoning (it impairs nerve conduction velocities and
EEG patterns also change), but not on what might be
the psychological deficits linked with these
neurophysioiogical ones. Damage from exposure to
lead over a long period of time may be more like
57
damage from chronic malnutrition than damage from

a severe acute injury to the brain. There is usually
considerable recovery from the latter, but persistent
impairment may be more common after lead
poisoning as after malnutrition. The effects of lead
almost certainly interact with the effects of other
socio-economic variables, being more severe for
socially disadvantaged children than for middleclass ones, indeed being negligible for advantaged
children (Shaffer 1985).
We do not yet know very specifically what
cognitive or behavioural problems lead is associated
with: most of the research is on its association with
attention deficits, emotional reactivity and
hyperactivity, where there is a consistent association.
It is not, however, a strong association: lead level
does not all by itself cause inattentiveness or
educational retardation or delinquency, but it does
contribute to some such problems. Preventing lead
poisoning, provided the substitutes have fewer
adverse effects, will lead to small but significant and
not trivial benefits.
Brain organization and development

The brain is organized into different structures
characterized by different clusters of neurons. Most
of this general organization is recognizable by the
seventh month in utero, by which time there has been
a long and complex progress in development, from an
initial pool of undifferentiated neurons to specialized
neurons located in their proper places, a progress
whose mechanisms are not yet understood. Different
parts of the brain develop at different rates, and so are
most sensitive to environmental influence at different
times. The most advanced parts of the cortex are the
primary motor and sensory areas: the areas where
impulses are compared and integrated develop later.
The brain areas which control movements of the
hands, arms and upper body mature earlier than those
which control the legs, a difference mirrored in the
varying control infants and young children have over
different movements. There are general positive
correlations between maturity of brain structure and
maturity of behaviour, but after the earliest stages
58
both the causal sequences sketched at the beginning

of this section are involved.
Recent studies have looked at the metabolism of the
brain, including changes in the metabolic activity of
the brain during specific activities. New techniques are
allowing identification of which parts of the brain are
most active during the processing of particular stimuli.
There are some interesting results from adults
(Parmelee and Sigman 1983, p. 106; Blakemore,
personal communication) but not as yet any
developmental studies. There are now known to be
developmental changes in the speed and efficiency
with which nerve impulses are transmitted and in the
biochemistry of neurotransmitters. Further research is
needed here, but changes in which neurotransmitters
are present, and in their strength, would help to explain
why certain drugs have one effect on adults and the
reverse on infants.
Most of what I have said so far concerned neurons or
even parts of neurons. However, they do, of course, work
together in complex networks. Electroencephalogram
recordings(EEGs)reflecttheneurophysiologicalactivity
of networks of neurons in the cortex. The rhythm of
activity in EEG recordings changes with what the person
recorded is doing (for example sleeping, sitting quietly,
thinking about a difficult problem), and developmentally.
For example, adults have quite different EEG activity
during sleeping and waking, while newborn infants have
much less consistent patterns, and at a behavioural level
go from sleeping to waking much more rapidly. EEG
scans can show up pathological brain activity but do not
help much in distinguishing within the normal range of
behavioural development, temperament or intelligence
(Parmelee and Sigman 1983, pp. 11417).
EEG patterns can be analysed, perhaps by
computer, to assess the time link between
presentation of a stimulus and cortical response to it.
Embedded in the spontaneous EEG activity is a peak
which is due to the presented stimulus, the sensory
evoked potential. In the immature nervous system,
these responses tend to be rather slow, and less
definite in their shape than in adults. They mirror the
long transmission time taken over immature synaptic
junctions and along immature axons which are not yet
fully myelinated. The existence of sensory evoked
potentials even in pre-term infants indicates that the

sensory receptors can respond to the stimulus, that
there are functioning connections between receptors
and cortex, and that the neurons in the cortex can
respond too, even if the whole process is on the slow
side. They do not, however, indicate what processing
is being done at any of these stages unless
comparisons between stimuli are made. Recent
studies (Parmelee and Sigman 1983, pp. 1202) have
begun to do this, in some cases as a measure of
intelligence.
The last aspect of brain development I want to refer
to is the functional lateralization of the brain
(Kinsbourne and Hiscock 1983). Much has been
made of differences between the left and right halves
of the brain: the popular literature on the subject
abounds with suggestions that the left hemisphere is
abstract, analytic and verbal while the right
hemisphere is intuitive, artistic and concrete, or that
educational failure is caused by being right-brained
in a left-brained school system. We will see that
such assertions are not well founded.
One very basic reason for this is that although some
mental functions, such as perceiving the orientation
of a bar of light, can be very precisely located in the
brain, most higher mental functions are represented
in several brain structures not localized in a single
brain centre. Doing an arithmetic problem, for
example, such as 2 + 3 = ?, will involve reading,
memory, reasoning, computation and a variety of
movements of eyes, head and the hand that writes
down 5. Each of these is certainly complex enough
to involve many brain processes. (There is some
discussion of their complexity in the next chapter.) It
will be obvious that we cannot sensibly talk about
how arithmetic is lateralized when many areas of
the brain are involved and when indeed arithmetic
and ways of doing it vary in complexity so much.
Evidence from a large number of studies has now
shown that most adults have their speech control in
the left hemisphere, though the right hemisphere has
a substantial capacity for language comprehension
and some for control of expressive speech if the left
hemisphere is not functioning. The right
hemispheres functions include control of tone
perception and visual memory. A small proportion of

right-handed people, and something under a third of
lefthanded people, have other patterns of
lateralization of speech, some in the right hemisphere
and some in both hemispheres. Recent EEG studies
suggest that the left hemisphere is more active when
tasks which are highly verbal are being done
(Kinsbourne and Hiscock 1983, pp. 1839). There
are anatomical differences between left and right
hemispheres, though how these are related to
behavioural differences is generally not clear.
Although there are some changes between
childhood and adulthood, both in anatomical
differences between the hemispheres and in
behaviour such as recovery from brain damage,
Kinsbourne and Hiscock argue that the development
of functional lateralization of the brain does not begin
with two neutral hemispheres of equal potential and
end with two definitely differentiated and specialized
ones. They read the evidence as suggesting
asymmetry from at least the first few postnatal weeks;
as well as anatomical asymmetries rather similar to
adults, infants seem to have a bias towards
movement to the right, preferences for using their
right hand or right foot, and better perception of
speech sounds in the left hemisphere and of nonspeech sounds in the right (Kinsbourne and Hiscock
1983, pp. 21336). They attribute changes in
apparent laterality not to changes in structural
properties of the brain, but to different approaches to
the task, shifts in the strategies employed.
Although developmental changes in asymmetry of visual
perception have been attributed to an emerging
hemispheric specialization for certain functions, it is
curious that a brain that is lateralized from birth for the
processing of speech sounds should become lateralized for
other functions only after the passing of several years.
We suggested in the previous section of this chapter that
developmental changes in degree of asymmetry probably
reflect developmental changes in the behavioural
organization of the skill rather than a shifting neural base
for a constantly organized set of component skills. As a skill
develops in the maturing brain, additional lateralized
components may be recruited to the performance of the
skill and asymmetries observed for the first time. When
59

performance of the younger child is symmetric, this need
not imply that the lateralized components have not yet
become lateralized; instead the lateralized components
may not yet be functional or integrated into the organization
of the skill. In support of this explanation, we can point to
evidence (previously described) that perceptual tasks are
susceptible to influence from nonstructural variables, such
as strategy, expectancy, and previous experience with the
stimulus material.
Perceptual tasks are not merely measures of some
structural property of the brain. It seems plausible that agerelated changes in degree of asymmetry reflect different
approaches to the task rather than different degrees of
cerebral lateralization (Kinsbourne and Hiscock 1983, pp.
2301).
I cannot assess the merits of their evidence and

reasoning, but they are emphasizing yet again the
problems of linking brain structure and behaviour in
developmental theory. No facile conclusions can be
drawn.
Cognition in infancy
I have outlined what is known about the physical
status and the functioning of infants perceptual
apparatus in the preceding section. It would seem that
they are able to get perceptual experience from birth,
and in some modalities from before birth, indeed that
perceptual experiences of various sorts are needed for
the proper development of the sensory apparatus.
However, there is a difference between the objective
physical stimulus and what is made of it in the course
of perception and cognition. Most stimuli are
interpreted by the organism which perceives them,
whether that organism be human or non-human,
infant or adult. Sometimes the stimuli need little
interpretation to be meaningful and effective,
sometimes a great deal is needed. Sometimes the
interpretation made is objective, derived from the
stimulus more than from the perceiver and likely to be
identically derived by any other perceiver:
sometimes the interpretation is subjective, a result of
the perceivers idiosyncracies not of the stimulus.
Debates about the extent and the sources of
interpretation of perception have been active in
philosophy since well before any psychology or any
60
systematic study of infants began: Hamlyn (1978)

reviews the course of the arguments between the
empiricists and the rationalists.
Piagets theory
The central theory of infant cognition has been

Piagets. He emphasizes the importance of action for
the development of thought. The infant acquires
knowledge of the world by acting on it, actions which
are initially crude reflexes but which develop into
organized acts linked to their consequences. These
are what Piaget calls circular reactions, that is, acts
which tend to produce results which lead to the reelicitation of the initial action. An example is
sucking: this produces (among other sensations)
mouth pressure which tends to elicit further sucking,
so that the initial reflex meshes with feedback from its
results. The circular reaction becomes differentiated
and refined, so that the infant learns, for example, to
suck in different ways suited to different objects.
They also become integrated with other circular
reactions, as in Bruners experiments when infants
co-ordinate the sucking which brings an attractive
picture into focus with their looking at the picture
(Bruner et al. 1966). The increasing differentiation,
refinement, integration and deliberateness of circular
reactions lead to progress in the infants construction
of a sensori-motor action-based representation of the
world. The functional invariants of Assimilation and
Accommodation are central to this progress, as new
stimulation is interpreted in terms of old knowledge,
and old knowledge extended by new information.
Thus Piagets infant sets out into a world of varied
stimulation equipped with a number of reflexes and
assimilatory/accommodatory powers, and makes
good cognitively by hammering out of the world a
construction derived from his or her own actions
upon it. The activity of the infant is what is
emphasized, more than the possible structuring of
information offered by either the physical or the
social world. These possibilities are put nearer the
centre stage by two other theories; Gibson (1979)
emphasizes the orderliness of the physical world and
the contributions that this makes to cognitive
development: Bruner (1973) and others emphasize

the contribution made by the infants caretakers. I
will spend more time on the former here, as the
contributions of adults to infants cognition are most
obvious in the social and linguistic spheres and are
therefore discussed in Chapters 4 and 6.
Gibsons theory
James J. Gibson is primarily concerned with

nondevelopmental work on perception but the
epistemological implications of his work are in some
contrast to Piagets, and are very relevant to
understanding cognitive development in infancy. An
important assertion is that the environment is rich in
organized information. For example, changes in the
apparent texture of objects indicate distance, as when
we can see the separate blades of grass next to us but
as we look further away the grass blades merge into a
smooth uniform green. There is also a great deal of
organization in features like the relative sizes of
objects. If object A alone is getting bigger as we
watch, while the rest of the visual field remains the
same size, we see A moving towards us: if A and field
both get bigger, remaining the same size relative to
each other, we see A and the field getting nearer to us
or ourselves moving nearer to the field. Treating
perception by analysing it into separate bits of
stimulation tends to obscure this organization.
Another important concept is that of affordances
(E. J. Gibson 1982). An affordance is a collection of
properties of part of the environment relative to the
organism being considered. For example, if a surface
is more or less flat and horizontal, rather than sharp,
slanted and vertical, and is large enough in extent
relative to your size and rigid relative to your weight,
then it affords support. It is stand-on-able,
permitting an upright posture . . . it is therefore walkon-able and run-over-able. It is not sink-into-able (J.
J. Gibson 1979, p. 127). This affordance is an
objective physical property of the world, but it is
relative to you. A larger heavier animal might not be
able to stand on the chair which affords support to
you: you cannot be supported by the twig which
affords support to the insect. A noise in the engine of
a car affords information about how well the car is

running: someone knowledgeable about the car will
perceive this affordance, a nondriver will not.
Gibsons theory suggests that many of these
relationships and affordances are perceived quite
directly. The infant has, from birth, a perceptual
system capable of detecting the rich organized
information which the environment provides. Where
Piaget suggests that information from different
modalities or different moments of experience are
separate to begin with, and have to be brought
together by the infants construction, Gibson
suggests that environmental information comes in
already organized and synthesized, in meaningful
and useful packages, and little construction and
reinterpretation is needed. Experience leads not to
new insights but to clearer tuning and better
discrimination of which features are most distinctive:
the child becomes a more skilled and systematic
observer, rather than Piagets active little
experimentalist.
Some cognitive developments in infancy
Harris (1983) reviews much of the literature on infant

cognition in the light of these different theoretical
emphases. I will discuss cross-modal integration,
development of space perception and the
development of the infants understanding of the
existence of objects, as these areas have received a
great deal of attention and raise important theoretical
points.
Cross-modal integration
Piaget, like the empiricist philosophers, argues that

perception of space is gradually constructed from an
integration of visual information into action. As a
result of failure and success in reaching, and later
locomotion, the infant comes to understand depth and
distance. The integration of vision and touch, which
began as separate modalities, leads to cross-modal
perception, and an understanding that objects may
continue to exist although no longer visible arises from
multimodal knowledge and active experience of
moving and finding objects. At first the world is
61
haphazard, varying unstably from moment to moment

and full of unintegrated stimulation: gradually the
infant constructs a world of integrated schemes and
permanent objects.
The Gibsonian infant has the benefit of linked
sensory systems dealing coherently with linked
information in the outside world. It is not necessary to
construct correspondences between sound and sight,
vision and touch: the physical world ensures that they
often co-vary for the objects that surround the infant.
What he or she has to do is learn which bits of covarying information are most reliably associated with
objects or events: this sort of discrimination does not
have to be derived from the infants own actions.
As far as the infants ability to make links across
sensory modalities is concerned, even infants in their
first month of life can use information in one
modality to guide another. For example, infants will
probably look towards a sound, though it is not clear
whether they also expect to see something at the place
where the sound was heard. By the time they are about
4 months old they prefer to look at an object which is
moving at the same tempo as the sound they hear
(Harris 1983, pp.7089, 7401). Coordination
between sight and touch also seems to originate very
early, though there have been many replication
failures here. Bower, Broughton and Moore (1970)
reported that 2week-old infants reached in the
appropriate direction for the object they saw, shaping
their hand appropriately to the object, and showing
surprise and distress when, owing to some optical
trickery, they reached to the place where the object
appeared to be but really was not. Harris (1983, pp.
709, 742) cites only failures to replicate this study.
More recently, Meltzoff and Borton (1979) have
produced data which suggest that neonates can
recognize a visually presented object after a prior
tactual experience; Bryant et al. (1972) showed that
8-month-olds can certainly do this. Thus the evidence
is not easy to interpret.
Nevertheless, Harris concludes that sight probably
does trigger reaching, and sound trigger looking.
What is not clear is whether the infant really expects
to see the source of the sound or touch the seen object,
that is whether there is cross-modal integration at a
62
deeper level. Infants have been shown to have more

inter-modality coordination than Piaget allowed
them, but they have to learn when to expect that
information from different modalities should be
correlated and when it need not be. For example,
although voices often come from faces, non-speaking
faces are quite common, and the infant will
sometimes hear a voice without seeing a face. Lip
movements, on the other hand, should always fit
speech sounds.
Depth perception
Depth perception is thought by Piaget (and by the

empiricist philosophers) to be gradually constructed
from the infants experience of reaching, grasping,
and moving, which are constructively integrated with
visual experience. Gibson suggests that depth is as
directly perceived from environmental information
as colour or shape is, and precedes the infants
experience of movement. As we saw when reviewing
visual perception, the neonate has some capacity to
achieve binocular vision which is one of the visual
cues to depth, and can certainly achieve exact eye
convergence well before accurate reaching is
possible. From about 5 months old infants are
unlikely to reach for objects which are beyond their
reach, and at about this age also react by moving their
arms between themselves and the object which is
apparently on a collision course with them (Harris
1983, pp. 7105). It appears, therefore, that infants
show behaviour adjusted to spatial distance before
they have had much opportunity to practise moving
themselves, behaviour more consistent with
Gibsons theory than with Piagets. They certainly
learn more about distance and depth perception with
experience of locomotion, however.
Infant search and object permanence
Piaget claimed that infants for most of their first year

believed that out of sight meant not just out of mind
but out of existence. Objects, including people,
existed while the infant looked at them and ceased to
exist if not looked at. When the infant looked again,
the object began to exist again, or a new but identical
object began to exist. When the infant did not look,

the object had no existence.
(While this belief seems bizarre and indeed
infantile, it has been a real philosophical problem to
determine whether something would exist while no
one is aware of it. Two of the nineteenth-century
Oxford limericks propose the problem and a solution
to it which Piaget does not attribute to the infant.
There once was a man who said God
Must think it exceedingly odd
If he finds that this tree
Continues to be
When theres no one about in the Quad.
Dear Sir, Your astonishments odd:
I am always about in the Quad.
And thats why the tree
Will continue to be,
Since observed by Yours faithfully, God.
We only know that something continues to exist

even though neither we nor anyone else are
perceiving it because we have a theoretical model of
the world which assumes objects have a stable
existence, and we have learned a great deal about
different sorts of disappearance and reappearance.)
There is a well-replicated sequence of infant search
behaviours. To begin with, infants do not actively
search at all, they simply stare in the direction of
where the object was before it moved or disappeared.
Next, they show some anticipation of the direction of
movement of a moving object, follow it visually and
manually if it is taken from them, and will retrieve an
object from a hiding place which only partly hides it.
Later, they search for an object which has
disappeared from view completely, but mainly in the
place they are used to seeing it at, not in the new
hiding place where they have just watched it being
put. Faced with the disappearance of the object, the
child immediately ceases to reflect and merely
returns to the place where action was successful the
first time (Piaget 1954, p. 61). This is Piagets stage
4, the most striking part of the infant search/ object
permanence data, and a popular research problem. In
stage 5 the infant does not make this perseverative
error but has difficulty with successive not fully
visible displacements to new hiding places, and from

about 18 months old the stage 6 infant can search for
an object whatever displacements have been made.
There is some suggestion that infants at object
search stages 13 may be a bit more sophisticated in
their search than Piaget proposed, in that they do
seem to treat objects which disappeared by being
gradually covered up differently from objects which
disappeared more suddenly or by fading or shrinking
to nothing, and they may seem slightly surprised if the
object when it reappears looks different. This
suggests a Gibsonian position of the infant using rich
information and gradually becoming more
knowledgeable about the possibilities of objects may
be tenable, but Piagets observations of stages 13 are
not in dispute. Replications of stage 4 do, however,
suggest that infants do not invariably make the
perseverative error which is central to his account of
the stage. The Piagetian infant, having previously
seen and found the object at A, then watches the
object being hidden at B but searches for it only at A,
apparently believing that the existence of the object is
contingent on looking for it at A. Replication infants
quite often search at B (about 50 per cent)
(Butterworth 1975, 1977, 1978); they sometimes
search at A even if the object is visible at B; they make
very few errors if they are allowed to search
immediately the object is hidden rather than being
forced to wait (Gratch et al. 1974); they search more
correctly in a conventional container such as an
upright cup than in an unconventional one or under an
inverted container (Freeman et al. 1980). That is, the
infants search is not so unreasonable or egocentric as
Piaget implies, and the AB error may occur because
the infant is confused about the objects whereabouts,
not because he or she believes its existence is linked
to A.
Harris (1983) places the emphasis on the infant
coming to know where to search and what sort of
thing he or she is searching for. A distinction has to be
learned empirically between single objects and
multiple identical objects. The former behave
lawfully as to their position: an individual biscuit
cannot be simultaneously in the tin and on the plate.
The positions of multiple exemplars are not lawful:
63
taking Biscuit One out of the tin does not preclude

taking another biscuit out of the tin unless Biscuit
One is the only biscuit there is. The infant has to learn
how the successive positions of an object are related,
which will involve considering these positions in
relation to the infant, who can move and so change the
relative position of the object, and in relation to a less
mobile external framework. The biscuit is first seen
in front of me, on the table; next in Daddys hand,
closer in front of me; next in my hand, moving to my
mouth; next on the floor underneath my chair. It is
suggested (e.g. Butterworth 1975, 1978; Bremner
1980) that infants initially code position relative to
their own body, a strategy which works less well once
they begin to crawl and hence to alter the relative
position of the object and their body. A spatial
framework specifying position relative to distinctive
landmarks is developed. Some notions of the
physical characteristics of landmarks and objects
may be included; for example, Freeman et al. (1980)
show that infants search more for an object hidden
and moved about under an inverted container.
Certainly there are differences later in understanding
prepositions such as in, on, and under (Clark
1983).
reviews the evidence on how infants deal with

invariance across changes of orientation and with
objects which are different in some respects but
invariant in being members of the same category.
Habituation techniques show that infants can
distinguish different orientations of faces and
geometric shapes, and, if they have seen several
orientations of an object, they can react to a new
object as if it were different from all that they have
previously seen. That is, the old object, whatever its
orientation, is more familiar than the new object. The
age at which this distinction is made depends very
much on the objects used. Infants shown several
different exemplars of a category, for example male
faces in a study by Fagan (Harris 1983, p. 732), show
dishabituation to a member of a different category, for
example a female face, but not to a new exemplar of
the original category. The breadth of familiarization
which they are given is related to the breadth of the
concept they construct. There is some evidence that
they use either the most frequent features or a form of
averaging to build up a prototype example, just as
adults do. This suggests that abstraction and
deduction are involved in the detection of invariance,
not just the detection of features which the Gibsons
propose.
Recognizing the identity of objects
Detecting the invariant qualities of an object among

the qualities which vary in appearance as the object
moves, or is partially transformed, is an important
cognitive activity in both Piagets and Gibsons
theories. For Piaget, the infant attains the object
concept by deduction from his or her active
experimentation on objects, gradually deducing that
objects are permanent, external to the self and retain
their identity whatever their changes in position
(Piaget 1954). Later, this understanding of qualitative
identity is developed into appreciation of what
quantitative aspects of the object are conserved (see
Chapter 3). For Gibson, on the other hand, much
information about identity is given by the
environment, and little deduction is needed.
Invariance is detected not constructed (E. J. Gibson
1969; J. J. Gibson 1979). Harris (1983, pp. 7319)
64
Recognizing the separation of self from objects
The dominant theory has been for a long time that

infants do not distinguish between themselves and
the outer world. They feel certain sensations but do
not know where they come from, they cannot
distinguish events which they are responsible for
from events which are independent of them. They are
utterly egocentric, subjective and solipsistic, with no
way of distinguishing between internally-generated
and externally-given information.
Recently, Gibson has argued that, on the contrary,
infants do have a number of sources of information
which would allow reliable distinctions to be made
between them and the outside world. One source of
information is the kinesthetic feedback supplied by
any movement of the body. Another is the bits of body
which provide boundaries to the visual field, the nose,
the eyebrow ridge and the infants fat cheeks. These
are a fairly constant part of the visual field; they move

whenever the head moves, faster than objects in the
world beyond, and do not move when the head is still.
Thus the infants can feel their own actions and see
part of themselves, including of course the mobile
bits of body such as hands.
While there is as yet little evidence to test Gibsons
model, Harris (1983, pp. 7447) proposes that it
suggests that from the early months of life the infant
is indeed sensitive to the ways in which the visual
field specifies the existence, location and movements
of the observing self. Full recognition of oneself as a
person involves more than a separation of self from
the external world, of course, and is discussed in the
section on the self-concept (Chapter 5).
Play and cognitive development
It has often been suggested that play is the childs way

of learning. This suggestion has been implemented in
the classic early childhood curriculum of free play,
which provides children in settings such as
playgroups and nursery schools with materials and
opportunities for play and encourages them in selfchosen meansdominated activities rather than
involving them in achievement-directed training.
Directive or didactic intervention by adults is seen as
inappropriate for young children. The play way of
learning is backed up by the idealization of play (see
p. 24), by ideas derived from Piagets views on
learning (p. 34) and from Freudian ideas about play as
therapy (p. 28) and more recently by suggestions
from psychology experiments and observation that
play with objects is an efficient way of learning how
to solve problems (Sylva 1977; Sylva, Roy and
Painter 1980). Smith and Simon (1984) review a
number of such experiments which have compared
the effects of a few free play sessions and a few
directed training sessions on convergent or divergent
problemsolving. Reviewing these studies they
summarize them as suggesting that experience in
play and non-play conditions contributes about
equally to convergent problem-solving, where there
is only one solution, and play experience
contributes more to success on divergent problems
where many and novel solutions are required. The
flexibility and lack of constraint that are assumed to

characterize play would be expected to have this
effect. However the experimental studies suffer from
methodological problems. First, it may be doubted
whether what the children do in the play condition
is play, in the sense of flexible, intrinsicallymotivated activity rather than, for example,
exploration which typically precedes play and is
necessary if all the potential of the material is to be
discovered (Hutt 1979). Second, the demonstration
that play is as good an inducer of learning as the
training given does not mean that play is as good a
way of learning as any training would be. In at least
some of the studies reviewed the training is clearly
not optimal for the subjects, and controls do not seem
adequate. Smith and Simon also express concern
about possible experimenter effects such as
differential provision of hints to play and training
groups or simply a more relaxed style of testing in the
problem-solving phase for children from the more
relaxed play group. There is room for doubt also
that one brief play or training session could be
long enough to do any more than help the child feel at
ease with the experimenter. After concluding that no
conclusion on the question of the contribution of play
experience to learning is possible from the
experimental evidence available at present. Smith
and Simon call for more ecologically valid work.
A considerable amount of observational data is
available now from studies of British pre-schools
done during the late 1970s (e.g. Sylva, Roy and
Painter 1980; Hutt et al. 1979; Meadows and
Cashdan 1983; Tizard, Philips and Plewis 1976).
These studies have tried to assess the quality of
childrens play in part because of concern with its
contribution to cognitive development. Sylva et al.
(1980) used a binary categorization of quality, Tizard
et al. (1976) two composite scales, Meadows et al.
(1983) four cognitive scales. These involved
discrimination of symbolic, goal-directed and
unfocused instances of play, four levels of how
involved the child was in the play activity, how much
use was made of play material, and how many
operations were brought together in the activity.
Rating play on these four dimensions showed
65
positive but low correlations between dimensions,

and differences between materials and children in
which sort of complexity characterized them. There
are some problems in making detailed comparisons
between studies, differences in how the quality of
play was assessed being one, but on the whole they
agreed in two ways important for our present
purposes. First, the general level of cognitive
complexity of childrens play in the wide range of
pre-school groups observed was disappointing.
Much of what children did when playing was
pleasurable but simple, repetitive, unstructured,
uninventive, uninvolving, brief and generally
uninspiring. Second, on the rather rare occasions
when teachers were more than casually involved, and
in association with a limited range of materials which
66
made it easy to define a goal and ones progress

towards it (art activities are the clearest example),
children showed play which was at higher levels.
It would be possible for those who are the greatest
enthusiasts about the value of spontaneous play in
cognitive development to dismiss this picture of how
children play in pre-school groups as having managed
to miss the learning that children were really doing,
during what looked uninspired to someone not participating. However the research on different forms of
early childhood education and the work on cognitive
skills which I discuss in the next chapter both suggest,
I think, that although children can and do learn
through their play we need to recognize more explicitly what is learned through social interaction with
adults and their deliberate teaching or model-giving.
Plate 6
67
Plate 7
68
3 The development of cognitive skills
In the previous chapter, I discussed the current

theories of what cognitive development consists of,
and described perception and cognition in infancy. In
this chapter, I am going to outline what we know
about the development of cognitive skills. After the
influence of Piagets very abstract and general theory,
interest has moved towards research which provides
detailed accounts of childrens behaviour in more
limited areas of cognition. I will describe some of
these areas, and draw some conclusions about the
developmental processes involved.
It may be helpful to the reader to have in mind what
I have emphasized already, that neither cognition
nor development is simple, and that learner
characteristics and task characteristics interact.
Indeed, there should perhaps be implicit in our
discussion something like the organizational
framework proposed in a chapter by Brown,
Bransford, Ferrara and Campione (1983) which is the
best current review of cognitive development that I

know. I give here a slightly elaborated version of their
tetrahedral framework (Figure 8).
Attention
Bearing this interaction in mind, then, we can look
at some of the components of cognition and how
they develop. Attention seems a good aspect to
start with since it could be taken as a necessary
condition for further cognitive activity. It is
however a somewhat vague concept, certainly
polymorphous. Taylor (1978) gives examples:
Attention can be directed, switched, captured,
distributed, divided, narrowed, sustained or
withheld. . . . Distractibility may imply that the
child is not motivated to do the tasks he is given, or
that they are too difficult for him to persist at; it may
mean that he explores all stimuli,or all prominent
Figure 8
69

stimuli, or simply that the values he gives to stimuli are not
that of the rater; it may mean that he becomes fatigued very
rapidly and changes task frequently as a result (Taylor
1978, p. 185).
Nor is attention simple to measure: overt

behaviour, task achievement, introspection and
physiological indices of concentration such as
changes in heart rate or EEG have all been used, and
are all problematic. They do however give a
moderately coherent picture of development. As
development proceeds the direction of attention
becomes more independent of what is conspicuous in
the environment, more systematic and more flexible.
In looking at a picture, for example, young subjects
tend to focus on a point and work haphazardly away
from it, while older subjects are more likely to scan
the picture exhaustively or to attend more to
information which is relevant to the task they have
been set. One paradigm which shows this up very
usefully is that of incidental memory (e.g. Hagen
and Hale 1973). Children are shown an array of cards
and asked to recall where members of a specified
subset (e.g. animals) are, while no instructions are
given about the need to recall another subset (e.g.
toys). They are then tested on their recall of both the
subset they were told to learn (central task) and the
subset which they werent (incidental task).
Performance on the central task improves very
much with age: performance on the incidental task
does not change, and may decline in older, better
educated or more effective learners. The older
children have succeeded in directing their attention
and their effort to learn where they are most needed:
the younger children have not. As well as being less
able to direct attention selectively they are less able to
resist distraction (Taylor 1985).
Many school tasks require persistence and
accuracy over a long period of time, and so he has a
short attention span is quite a serious complaint
about a child. Sustaining attention, vigilance, is
known to be affected by many different factors in
adults: difficulties in remaining vigilant are
commonly found in adults and children with
psychiatric symptoms. However, observations of
children doing school work (e.g. Meadows and
70
Cashdan 1983; Galton et al. 1980) show that

alternation of concentration with daydreaming,
social chat and other activities less tightly related to
the task is common. Hard work and rapid progress
during concentration makes up, on most tasks, for
periods of distracted or unfocused attention. (The
accounts given of creative thinking and the role of
subconscious activity are similarly comforting.)
Sustained vigilance in the sense of attention which
never omits to respond to target stimuli and never
makes incorrect responses seems to be learned late, as
a deliberate cold-blooded skill, and, at least in the
forms which military tasks can require, needs special
training. Given high motivation in daily life,
however, children can be sufficiently vigilant: little
pitchers have big ears!
Low attention span and distractibility have
been related to neurological damage and immaturity
and linked to hyperactivity, but there is little
consensus about how strong the links are or to how
physiological and psychological measures interact.
Here as elsewhere (see, for example, discussion of
aggression, Chapter 5) physiological (or
biochemical) changes may cause or be caused by
psychological ones, or both may be caused by some
third sort of change. Psychosocial factors are also
involved, both in choice of what is voluntarily
attended to, and in the strategic aspects of attention.
Family interaction and communication are probably
involved, though a far more precise analysis of the
ecology of homes is required than has been carried
out so far (McGurk 1977).
Attention is a component of two of the bestknown theories of cognitive style. Witkins Field
dependence vs independence model (Witkin et al.
1979; Witkin and Goodenough 1981) involves
selective attention in the ability to process
information analytically rather than holistically. The
impulsivity vs reflectiveness model of Kagan et al.
(1964; Kagan 1984) also implies differences in what
is attended to and how. Neither model seems to be
making much progress at present (Kogan 1983), and
doubts remain about the conceptual and
methodological usefulness of rather simple models
of cognitive style which are supposed to pervade all
functioning. There is some danger that they become

seen as important independent entities and so precise
study of tasks and processes is precluded. Taylor
(1978), concluding his review on the development of
attention, and indicating how many gaps there are in
our understanding, emphasizes that attention is not
unitary and it is not something we simply get more of
as we get older. He suggests that age (presumably he
means the correlates of age, such as maturation and
experience) brings an increase in the use of
systematic, logical strategies of exploring the world;
in the ability to be flexible and selective in ones
approach to information; and in maintaining ones
responsiveness for longer periods (Taylor 1978, p.
195).
As the reader may already have anticipated, this
sort of conclusion lots of gaps in the evidence, but a
picture of initial strategies and skills become more
flexible and polished, in large part as a result of
practice will pervade my account of cognitive
development.
Remembering
It is obvious that remembering of some sort is
necessary for virtually any human cognitive activity.
As we saw earlier it is hard to believe in the possibility
of an intelligent organism which did not use
accommodation and assimilation in its functioning,
that did not make some comparison between the
present stimulus and stimuli encountered earlier. This
would be impossible without memory. Conversely,
memory is rarely an isolated intellectual skill. Early
researchers found it necessary to concentrate on
memory for meaningless materials such as nonsense
syllables in part because this was the only way to
control for differences in subjects knowledge,
understanding and so forth (see, for example,
Baddeley 1976). It is partly because of the interaction
of remembering, understanding and acting that
developmental changes in memory are important.
It is also obvious that remembering is not one
simple activity. There have been many suggestions
about the structure of memory and the processes
which are involved at each stage. There is no clear
consensus over the details of these models and they

are in any case derived mainly from work done with
adults, indeed mainly with undergraduates, rather
than with children. However some distinctions and
some processes drawn from these models appear in
developmental studies, and point up both changes
and lack of change with development.
Recognition memory
One distinction, due in part to differences in

experimental method, is between recognition and
recall. Recognition memory is investigated by
presenting subjects with the material they are to
remember, and then after an interval representing it,
either asking the subjects to judge whether it is
familiar or novel, or requiring them to discriminate it
from material which they have not seen before. In
recall tests, the material is not presented a second
time, and subjects are required to retrieve and
describe it from memory. All other things being
equal, most people find recognition tasks easier than
recall, and adults recognition of meaningful material
(such as pictures of familiar objects) is extremely
good (Shepard 1967; Standing 1973; Bahrick et al.
1975). Recognition memory is virtually as good in
children of 4 and older (e.g. Brown and Scott 1971;
Brown and Campione 1972). In Brown and
Campiones study, 4-year-olds were shown eighty
pictures of familiar objects. After an interval they
were shown 120 pairs of pictures; in sixty pairs one
picture had been in the original set of eighty while the
other picture had not but showed the same object (if,
for example, the original picture showed a dog eating,
it was paired in the recognition test with the same dog
running). The other sixty pairs were entirely
composed of new pictures. Recognition was highly
accurate; after a two-hour interval, after a day and
after a week. This is comparable with results in work
with adults, though adults would often be shown
more pictures.
Children as young as 4, then, have an impressive
ability to say correctly whether or not they have seen
a picture of an object before. Testing recognition
memory in younger children by asking them to say
71
whether pictures were novel or familiar would be

problematic as it might be hard for a young child to
produce the right words or to understand the
instructions. A paradigm which allows the child to
use well-mastered behaviour, particularly non-verbal
responses, is more suitable for investigating
recognition memory in babies and toddlers.
Habituation techniques and measurement of the
attention a baby pays to novel and familiar objects
have produced some useful results.
In habituation experiments, subjects are repeatedly
shown the same stimulus. On its first showing the
stimulus evokes a lot of attention, as measured by
how long it is looked at or by physiological measures
such as deceleration of heart rate or changes in EEG.
If on each successive showing it evokes less
attention, the subject is said to have habituated:
eventually the stimulus which was initially absorbing
receives only a brief glance. This decline in interest or
habituation could only happen if the subject
remembered the stimulus from one occasion to
another. Since habituation to simple patterns can be
demonstrated in babies from the age of 10 weeks,
when babies start to prefer to look at novel patterns
rather than ones they have seen before (Wetherford
and Cohen 1973), there is clearly some recognition
memory capacity in babies under 3 months old. The
possibility of conditioning in neonates (e.g.
Papousek 1967) suggests that there may be
recognition memory earlier still. At these young ages,
however, the time interval between stimuli must be
very short if the baby is to show any remembering. As
babies get older, they can recognize stimuli over a
longer interval: in a study by Fagan (1973) 6-monthold infants looked more at a novel black and white
photograph of a human face than at a familiar one
after an interval as long as two weeks. They can
probably also as they get older remember more
information about the stimulus and therefore notice
and attend to smaller differences between the familiar
and the novel stimuli (Kagan, Kearsley and Zelazo
1978).
This dimension of complexity of stimuli, and the
associated dimension of meaningfulness, may be
among the most important aspects of the
72
development of recognition memory. Although

recognition memory for simple objects is almost as
good in pre-school children as in adults, and there is
not much more difference between them on harder
stimuli such as abstract pictures or puzzle pieces
(Nelson 1971; Nelson and Kosslyn 1976), older
children and adults out-perform younger children on
recognition memory of complex scenes (Newcombe,
Rogoff and Kagan 1977). Young children seemed to
make less use of the meaningful relationships
between items in the scene: they may also have
scanned the picture less exhaustively, even omitting
to look at some parts of it. We will see that meaning
and processing strategies play an important part in
memory, a part which becomes more prominent as
children develop. Discussion of these points will
pervade the rest of our discussion of memory.
Recall Memory
We have seen that meaning and strategy become

involved in recognition memory, but that even in
cases where no particular effort is made to remember
recognition memory may be very good indeed.
Recall memory seems to be, on the whole, a much
more difficult proposition. If adult subjects are shown
a dozen or so unrelated stimuli once, and asked to
recall them in any order immediately afterwards, they
are likely to recall about seven items correctly; the
last few items and perhaps the first few are most likely
to be correct (recency and primacy effects
respectively). Only if the stimuli can be grouped, or if
some deliberate mnemonic work is put in, or if the
sequence has a meaningful structure, will the number
of items recalled rise much above seven. There are
developmental changes in this number, and indeed
recall memory for a series of numbers digit span
is a standard intelligence test item. Children aged 3
years have a digit span of about two, 4-year-olds of
three or four: the items they remember are likely to be
from the end of the list not the beginning, that is they
show a recency effect but not a primacy effect (Myers
and Perlmutter 1978; Dempster 1981). The primacy
effect in adults is thought to be due to subjects quietly
naming items to themselves as they try to learn the
list: that it is absent in young children implies that

they do not do this naming. Naming or rehearsal of
items is one of the strategies which adults use to
improve their memory. If children failed to use such
mnemonic strategies, this might account for their
poorer performances on recall tasks.
Childrens use of mnemonic strategies
There is a great deal of evidence now that children

under the age of 7 or so rarely use memorizing
strategies in memory tasks. Kail (1979) reviews
findings on childrens use of rehearsal. There is little
spontaneous rehearsal in under-7s; while 7-year-olds
sometimes do some rehearsal, it is likely to be
rudimentary and inefficient, for example limited to
repetitions of only one name. Older children and
adults are more likely to rehearse several items, to
rehearse members of the same category together, and
to adapt their rehearsal strategies to the particular
demands of the task. Thus as subjects get older, their
rehearsal techniques become more flexible and
efficient. Rather similar results are found for other
mnemonic strategies. Before school age children
rarely use them at all. When the strategies are first
used spontaneously, they are not used very well, but
as children get older they use them more efficiently
and more flexibly, and performance improves.
Papers in Kail and Hagen (1977) review research on
various mnemonic techniques: another improvement
with age is the combination of mnemonics a belt
and braces approach which also augments
performance.
Why do young children rarely use mnemonics?
One explanation would be that they are not capable of
doing so. Work by Flavell and his colleagues
demonstrates that this is not the case for rehearsal,
since children of 6 or 7 can be trained to rehearse
successfully. Similarly they can be taught to use
categorization during learning, and during the recall
phase to use cues, category labels and the strategy of
exhaustive search of successive categories (Kail
1979, pp. 1825). These young children showed
themselves perfectly able to use the mnemonic
strategy they had been trained in while the
experimenter required them to use it, so inability to

rehearse, use cues, etc., was clearly not their main
problem. However, a substantial proportion did not,
even after training, use the strategy spontaneously.
There is a gap between what they do off their own bat
and what they can be trained to do. While this is
probably true of us all (my memory for references,
say, and I daresay the readers, could certainly be
improved by the practice of certain mnemonic
strategies) one reason for the gap in small children is
particularly important and interesting. It is that they
do not appreciate how useful and necessary
mnemonic strategies are.
Knowledge about memory
In order to know whether it is worth putting special

effort into memorizing something one needs to know
quite a lot about what the task demands and about
ones own capacities. If the task is judged to be easier
than it really is, or ones capacities are judged to be
greater than they are, then one may conclude that no
effort is necessary to succeed. Young children make
mistakes on both these judgements. A graph from
Kail (1979) illustrates this (Figure 9).
Figure 9 Predicted and actual memory spans as a

function of grade level. N = nursery school children;
K = kindergarten children.
Source: From Kail (1979), p. 42.
73
The young children from nursery and kindergarten

believed that they would be able to remember many
more items after a brief exposure than they actually
managed. This inaccurate overconfidence about
performance seemed to be due to ignorance about
memory limitations, not to a general overestimation
of abilities, since estimates of jumping distance were
much more accurate than estimations of memory
span, and the latter were slightly reduced by
information about how well an average peer would
be likely to do (Kail 1977; Yussen and Levy 1975). It
is fairly obvious that over-confidence is likely to lead
to not making the extra effort which a more realistic
assessment would show is needed for success, and
hence to failure on the task. Failure on a memory task
is peculiarly hard to judge, however; one may know
that one has not remembered everything but not quite
how many items one has failed to remember. Thus
over-confident children may do badly on memory
tasks but not notice their deficit: ecological studies of
pre-school children might well show that they are
rarely asked to do recall tasks and even more rarely
have their failure brought home to them.
Improvement in estimating memory span and in the
use of mnemonic strategies comes after entry to
school, where accurate memory performance
increases in importance. A child who goes to school
is also having to live in the two micro-systems of
home and school, which will involve both relatively
spontaneous reporting of school events at home and
vice versa, and the deliberate carriage of information
from one to the other. No longer can one caretaking
adult take all the responsibility for the childs
remembering; it is only the child who inhabits both
micro-systems and knows all their details.
Improvement in memory estimation and in the use
of mnemonics can be induced in the experimental
situation by giving children feedback on their
performance and showing that improvement is
attributable to the use of mnemonics (Flavell and
Wellman 1977; Whittaker 1983). Continued use of
mnemonics is more likely if the demonstrated
improvement is large, as if the mnemonics have to be
seen to be cost-effective, and some generalization to
other tasks may occur.
74
One of the important considerations here is the best

way to remember in a particular memory task. Tasks
vary in how easy the stimuli are to remember and in
how easy different ways of demonstrating
remembering are. These aspects too show
developmental changes (Kail 1979). Pre-school
children know that, all other things being equal,
familiar items are more memorable than unfamiliar,
and long lists are harder than short ones. They do not
appreciate that a homogeneous set of items which
show a consistent relationship may be very easy to
learn irrespective of its length, thus they predict that
a short paired associate list where there are arbitrary
pairings will be easier to learn than a longer paired
associate list where all the pairs are common
antonyms. By the age of 9 or 10 children judge the
antonym list to be easier (as it would be to adults), and
if asked to generate a list which would be easy to
remember invent one with highly related items. Preschool children know about the rapid decay of shortterm memory, so they advise that having looked up a
telephone number one should use it directly, not stop
and do something irrelevant to telephoning; they tend
to agree with adults that recognition is easier than
recall. It is not until the age of about 10 that they say
that to remember word for word is harder than to
tell you in my own words, i.e. that paraphrase is
easier than verbatim memory. These results suggest
that although children have achieved some
understanding of the demands of memorizing by the
time they enter school, considerable further
experience of memorizing is necessary before they
can gauge task demands with accuracy.
In addition to knowledge of oneself and the
particular problems of the task set, performance is
likely to benefit from self-monitoring of progress so
that under-learned items are not neglected and no
item is over-learned from too much attention. Young
children do relatively poorly in their memory
monitoring: they frequently stop their efforts to
memorize prematurely, and on a second trial they
relearn items already learned and neglect items which
they have failed to learn first time. Here too feedback
can improve performance (Kail 1979). Developing
meta-memory skills seems to be a necessary but not

sufficient condition of good performance on memory
tasks. Metacognition is discussed later in this
chapter.
Knowing and remembering
The first point I made when introducing discussion of

the development of memory was that memory and
knowledge are not independent (except in certain
convenient laboratory techniques such as learning
nonsense syllables where knowledge is deliberately
made irrelevant to the situation). Chis elegant
experiments with chess players illustrate this point
particularly neatly (Chi 1978). She compared
children (average age 10) with adults on immediate
recall of a string of digits and on immediate recall of
stimuli placed within an eight by eight array. The
adults performed better than the children on the digit
list, as expected; but the children remembered more
about the stimuli in the eight by eight array. This
unusual finding of better performance by children is
explained by the fact that the children were
experienced chess players and the adults were not,
and the stimuli placed within an eight by eight array
were chess pieces in a mid-game position on a
chessboard. This array was meaningful and thus
memorable to a chess player, meaningless and
difficult to someone ignorant about chess.
Experience highlights the salient points of an event
which are most crucial for understanding and
remembering it (Brown and De Loache 1978) and
also makes complex events intelligible and easy to
remember because reconstruction and inference can
be used where otherwise it would be necessary to
learn every detail. The novice/expert distinction
(Brown and De Loache 1978; Flavell 1978) needs
more specification (Robinson 1983) but may be
useful in discussing cognitive development. Another
example of the fact that knowing more makes
remembering easier is the increased use of categoryrelated mnemonics as children get older. Pre-school
children may be relatively unlikely to use taxonomic
categories either at the learning or at the retrieval
stage (Kail 1979; Kobasigawa 1977), unless the

instances of a category are prototypical members
such as cat and dog rather than snail and
giraffe. This may be associated with lack of
knowledge, for example, that snails and giraffes are
animals. The degree to which childrens categories
are the same as adults is not yet clear (Rosch and
Lloyd 1978; Clark 1983).
It must, however, be pointed out that prior
knowledge may interfere with memory as well as
help it. One classic example of this is Bartletts
constructive recall evidence (Bartlett 1932) where
subjects heard a strange little folk-tale of an
encounter with ghosts by some North American
Indians. When asked to re-tell the story they
gradually anglicized it, losing or changing details
which had significant meaning in the Indian
mythology but had none to English minds. The
distortion produced in this case is an informative but
extreme example; usually in memorizing text the gist
is retained even though the wording and some details
may be lost. The error that increases with age is that
information which was implied, but not explicitly
stated, is inferred by the listener, so that implicit
information is believed to have been made explicit.
Acceptance of true inferences becomes more
frequent as children get older, though false inferences
continue to be rejected. Memory is normally holistic
and inferential and in most cases this is advantageous,
though it must contribute to the inaccuracy of eye
witnesses which so concerns criminological
psychologists (Loftus 1979).
Cultural demands for memory skills
It is speculative but possible that literal and verbatim

memory require special skills rather different from
what is done in ordinary daily remembering. (Tulving
(1972) has made a possibly analogous distinction
between episodic and semantic memory.) What is
commonly called a photographic memory or
eidetic memory, is rare in adults but apparently
commoner in young children who appear to be able to
examine a stable mental image of the picture they
were required to remember and read off from it
75
details, including details which have no meaning to

them. Eidetic memory is, however, a somewhat
elusive phenomenon and a simplistic model of
memory as stored images which are like photographs
is untenable (see, for example, Bransford 1979, pp.
1902). The possibility of verbatim memory for large
amounts of material is one which interacts
interestingly with the availability of external memory
stores. It has been suggested that the possibility of
recording information in writing and retrieving it by
reading transforms cognition (e.g. Bruner 1966;
Goody 1977; Cole and Griffin 1980; Olson and
Torrance 1983). This is a controversial subject which
receives more attention later. There are, however,
certain groups of people in a variety of cultures, both
literate and non-literate, who are required to recall
long passages of words verbatim. One group studied
though by an ethnographic linguist (Lord 1960)
rather than a psychologist was the itinerant ballad
singers of the remoter parts of Yugoslavia, who
moved from village to village entertaining the men
with traditional folk tales just as earlier reciters did
Homers poems, composed long before they could be
recorded in writing. It had been believed that such
storytellers must have achieved phenomenal powers
of verbatim recall. A closer look at what they did
showed that actually they remembered not every
word but the order of events and a limited range of
stock epithets, and with these and the help of strict
rules of rhyme and rhythm they reconstructed a
version of the story which differed slightly from
previous versions, rather than remembering verbatim
an unvarying text. Exact verbatim recall is required in
the memorization of sacred texts, laws and property
lists, which as Goody (1977) points out were among
the earliest products when writing was invented, and
also in the memorization of literary products such as
poems and plays. There has not been much study of
how this is done and whether it develops particular
mnemonic powers. Wagner (1982) provides a rather
brief account of a study of Moroccan boys,
comparing various groups differing in educational
experience and including boys from a traditional
Islamic school where the curriculum centered on
learning the Koran by heart, elementary literacy
76
being virtually a by-product of this. Their unusually

extensive practice of verbatim memorizing, either by
rote or by the organization of accumulated
meaningful material, had no effect on their short-term
recall or recognition of pictures of animals or of
oriental rugs. It is to be regretted that they were given
no opportunity to show how they remembered
material more like the Koran; whether, for example,
they showed better discrimination of explicitly
presented information from information which is
inferred not given. Groups of Moroccan boys who
had had a more western type of schooling did rather
better on the picture recognition and recall tasks.
They also showed signs of the primacy effect in the
recall task, which may indicate that they had used
rehearsal as a mnemonic technique.
It would be very premature indeed to come to any
conclusions about cultural effects on memorization
or possible universals in memory structure or
memory processing. All that can be said at present is
that training with feedback can induce children to use
mnemonic strategies and to persist in using them if
they are efficient; and that a high degree of practice of
a particular mnemonic technique can produce
phenomenal remembering of material suited to that
technique (see Luria 1969). Practice of one sort of
technique on one sort of material may or may not lead
to improved performance of other techniques or on
other materials. If children are to be set to learn reams
of poety, for example, this should be done for the sake
of the advantages of remembering reams of poetry
(which are various and by no means negligible) not
because the memorizing is expected to train their
memory in any general way. That it may do so should
be an incidental bonus.
Although experimental research has told us a great
deal about the structure of memory and about
remembering processes, we do not know as much as
might be desired about memory in ecologically
natural settings. Early work by Istomina (1975)
suggests that young children may do better on
memory tasks which make sense to them than they do
on artificial ones. In her study the comparison was
between recalling items while playing shopping
and recalling the same items as a formally presented
list. There are probably motivational differences in

the two settings, and more knowledge of the demands
of shopping meant more efficient strategies were
recalled as appropriate. Work like this, a more
sensitive approach to cross-cultural comparisons,
and further investigations of the child as
psychologist (Harris 1983a) would be welcome.
Reading
Frith (1980a) calls reading and spelling complex and
astonishing accomplishments, a description which
is obviously correct. We do not yet have a full account
of what people do when they read. Researchers agree
that very many linguistic, perceptual, attentional,
memory and cognitive skills are involved, but they
vary considerably in which they emphasize.
Research in experimental cognitive psychology
often concentrates on bottom-up analyses (see, for
example, Crowder 1982) and emphasizes the
readers use of, for example, eye movements or
pattern recognition processes. Other investigators
may assert that the readers knowledge of what is
likely to be the meaning of a word or piece of text may
be crucial in whether it can be read, and emphasize
top-down models (e.g. Smith 1978). In some cases
there has been a regrettable tendency to make the
top-down or bottom-up emphases too strong, so
that some accounts of reading as a matter of
comprehension have taken the perceptual
components as uninteresting and mechanical, and
some accounts of reading as a matter of decoding
visual information into a verbal form have excluded
anything more cognitive than word recognition.
Top-down and bottom-up have to be co-ordinated
in theories as they are in ordinary reading, where most
of the time processes at all levels are used. Recently
theories which integrate different levels have
appeared. Mortons logogen system (Morton 1969,
1980) and Rumelharts model (Rumelhart 1977; Ellis
1984) are important examples. It is clear that
reading includes many different activities at
different perceptual, linguistic and cognitive levels,
which no doubt interact in changing ways as the
reader becomes more skilled, or when the reader is
faced with different sorts of reading task or text. The

reading involved in recognizing cereals in a
supermarket is probably not quite the same as the
reading involved in understanding the same word in
a newspaper article on the governments farming
policy, and the reading might well be different again
when cereals appears in a scientific article or a
poem. Similarly, children learning to read may use
different processes to make sense of the written
message, varying according to, for example,
familiarity, the availability of context, and the childs
own preferences such as willingness to guess
(Francis 1982).
Development in reading processes
Although we cannot yet specify the developmental

course of interacting reading processes, it does seem
likely that it is the basic perceptual processes which
change least. Visual discriminations between
symbols like b and d, for example, can be shown even
in infants (see Chapter 2; and Banks and Salapatek
1983). There is, on the other hand, obvious and
enormous development in the childs knowledge of
language, of the world, and of reading and literature.
Variations here are predictive of success or failure in
learning to read. I shall have more to say about these
top-down constituents of reading than about the
perceptual basics, but I would not wish to imply either
that the latter are unimportant or that they remain
unchanged throughout the development of reading.
Children as they begin to learn to read probably
have, then, most of the perceptual capacities eye
movements, pattern recognition and discrimination,
attention which they need. They are also very well
used to dealing with the language that they hear. They
have had much practice in extracting meaning from
it: they have also probably analysed it into segments
at the levels of phrases, words and morphemes
(McShane, personal communication). In reading,
they have to do rather similar things to language
which they see rather than hear. There has been much
debate over whether writing is decoded into imagined
speech which is then processed as if it were really
heard, or whether reading goes straight from symbol
77
to meaning. The possible relationships between

speech and reading are complex, and the evidence is
so too (Crowder 1982; ch. 9; Ellis 1984). The debate
may perhaps be resolving into an agreement that the
skilled reader, at least, may use imagined sound or
may go direct from written word to meaning. What is
done depends on the readers skills, the novelty or
familiarity of words, the difficulty of the text and the
purpose of reading, among other variables.
Soundletter correspondences
Children learning to read often have problems over

the relationship between sound and letter. In English
there are many complexities in the correspondence
between phoneme and grapheme consider for
example a in cat, fate and arm, and c in cat,
ceiling and chuckle. There may be more
regularities in combinations of graphemes. As Stubbs
(1980) points out, Bernard Shaw overstated his case
for spelling reform when he suggested a
pronunciation for ghoti which followed precedents
in other words (gh as in cough, o as in women,
ti as in station, hence ghoti is pronounced fish),
as gh is always pronounced as a hard g when it
appears at the beginning of words (ghost, ghastly,
ghetto). There are however many irregularities
even in common words. The letter string ough has a
different pronunciation each time it appears in the
words bough, cough, dough, lough, nought,
rough and through. Further, Liberman et al.
(1977) show that it is unlikely that speech sound is
experienced as ready-segmented phonemes or that
phonemes blend obviously into words. Young
children and illiterate adults find it relatively easy to
segment words into syllables and almost impossible
to segment into phonemes. Symbol-sound
correspondences are somewhat easier to learn in
regularized alphabets such as the Initial Teaching
Alphabet (i.t.a.) (Downing 1979) but such alphabets
do not allow for regional pronunciation differences,
nor do they preserve the lexical and syntactic
information which irregular spelling carries (for
example the semantic relationship between bomb
78
and bombardment). Stubbs (1980) argues that the

English spelling system works extremely well for a
native speaker who knows its phonological and
morphological rules, that is it is better suited to adult
fluent readers than to children (or foreigners)
learning to read English. Awareness of language
would thus appear to be a most important component
of learning to read. Awareness of sounds is a predictor
of speed of becoming a reader: teaching about sounds
(phonics) seems to be a useful part of teaching
reading (Bryant and Bradley 1985). Phonemic
analysis is quite hard to learn but becoming able to do
it is an important breakthrough in the early stages of
reading. Sounding out words, if successful, reduces
dependence on recognizing their visual pattern and
supplements the childs knowledge of likely meaning
and vocabulary. Experience of reading increases both
the childs general knowledge and his or her
knowledge of the underlying rules of written
language so that gradually correspondences between
letter and sound become less crucial to understanding
what is being read. Adult fluent readers may only use
sound deliberately in their reading when they are
having problems
with
understanding
or
remembering what they read.
Language awareness in learning to read
Various studies have now picked up linguistic

awareness and experience as an important predictor
of learning to read, perhaps even a prerequisite for it.
Wells (1981) found significant positive correlations
between parents and childrens interest in literacy,
particularly the frequency with which stories were
read to the child, and the childs progress in learning
to read, for his representative sample of 120 children.
Francis (1982), in a sensitive case study of ten
children, found that understanding the task of reading
and writing was crucial both for doing it successfully
and for appreciating why it was worth doing.
Children who learn to read early and easily (Clark
1976) tend to have acquired such an understanding
before beginning school. How much the child is read
to, how much he or she sees other people reading,
how much and how explicitly written material is used

in daily activities like shopping, knowledge of
concepts like sound, word and sentence, are
related to rate and efficiency of learning to read.
These activities seem to contribute to achieving
insight into the links between written symbol and
word meaning: they also serve to establish that
reading is or can be useful and entertaining.
Children who lack these experiences tend to be
slower in learning to read. For some, lack of readingrelated experiences may be due to a home
background which also does not provide experience
related to other school activities. Children from such
homes may not know, when they first enter school,
what is required of them. Classroom tasks may be
relatively strange and incomprehensible. It is harder
to learn to do something which makes no sense in
itself and which you cannot link to your other
experience than something part familiar whose
purpose you appreciate, and when the something is
as complex and artificial as learning to read, it may be
the best you can do to do it slowly, weakly and by rote,
as some of the children described in Francis 1982 did.
Only when the children achieved an insight into the
general relation between reading and writing on the
one hand and spoken language on the other, only
when reading and writing became meaningful
activities, did they make much progress.
Reading stages and strategies
The problem is certainly not a simple one of deficient

home background, and it is not so much one of
inadequacies of language or of concentration on the
part of the child as of inexperience in reflecting on
language and how it is used: in other words it is often
essentially a metacognitive and metalinguistic
problem, interacting with the social problem of
adjustment to school life. Children solve the problem
of learning to read in different ways. Ellis (1984)
sketches a common developmental course. The 5year-old beginning to learn to read recognizes and
uses a spoken vocabulary of several hundred words,
and speaks grammatically. The first step made in
learning to read is often to glance-and-guess: a few

words are recognized by shape, otherwise an
unfamiliar word is guessed using the context as a
guide, its graphemic or phonemic characteristics
being rather unimportant in the guess. Unfamiliar
words without context cannot be read at all. Errors
tend to preserve meaning but look and sound different
from the correct word. Later, in the second step of
sophisticated guessing, the vocabulary recognized
by sight is larger, and unfamiliar words met in or out
of context are guessed in terms of their visual
similarity with familiar words, with contextual cues
used where possible. Visual cues from the beginnings
of words are probably more easily used than those
from the middle or end of words: words in context are
easier than words isolated in word lists (an instance of
the usefulness of top-down cues). Poor readers
incidentally may tend to rely on pictures for cues, or
even believe that the story is contained in the pictures
rather than in the text (Francis 1982; Yule and Rutter
1985).
Francis (1982) also reports that even her quicker
readers rarely used phonic cues until they had a fair
grasp of visual and contextual cues. They then used
sounding out unfamiliar words largely to supplement
a not-quite-adequate visual and contextual analysis.
Realizing, or being taught, that there is some
consistency between words in how letters or groups
of letters are pronounced, increases the childs
chance of decoding written word into meaning. It
may be particularly useful to do left-to-right
sounding-out on words which are phonemically
regular but visually nondescript, such as bun; words
which are phonemically irregular but visually
distinctive, such as light, may be more easily
recognized by their shape than by sounding-out.
Bryant and Bradley (1980, 1985) report that some
young children have separate reading and spelling
strategies, and hence can read some words (such as
light) which they cannot spell, and spell others
(such as bun), which they cannot read, a
phenomenon found also among Francis sample.
Insight into analysing words into phonemes,
79

Table 2 Stages of reading development: an outline of the major qualitative characteristics and how they are acquired
1
Stage
designation
2
Grade
range (age)
3
Major qualitative characteristics
and masteries by end of age
4
How acquired
5
Relationship of reading to listening
Stage 0:
Prereading,
pseudoreading
Preschool
ages 6
months6
years
Child pretends to read, retells story

when looking at pages of book
previously read to him/her; names
letters of alphabet; recognizes some
signs; prints own name; plays with
books, pencils, and paper.
Being read to by an adult (or older

child) who responds to and warmly
appreciates the childs interest in
books and reading; being provided
with books, paper, pencils, blocks,
and letters.
Most can understand the childrens

picture books and stories read to
them. They understand thousands of
words they hear by age 6 but can read
few if any of them.
Stage 1:
Initial
reading
and
decoding
Grade 1 &
beginning
Grade 2
(ages6
&7)
Child learns relation between letters

and sounds and between printed
and spoken words; child is able to
read simple text containing high
frequency words and phonically
regular words; uses skill and insight
to sound out new one-syllable words.
Direct instruction in letter-sound

relations (phonics) and practice in
their use. Reading of simple stories
using words with phonic elements
taught and words of high frequency.
Being read to on a level above what
child can read independently to
develop more advanced language
patterns, knowledge of new words,
and ideas.
The level of difficulty of language

read by the child is much below the
language understood when heard. At
the end of stage 1, most children can
understand up to 4000 or more words
when heard but can read only about
600.
Stage 2:
Confirmation
and fluency
Grades2&
3(ages7
& 8)
Child reads simple, familiar stories

and selections with increasing fluency.
This is done by consolidating the
basic decoding elements, sight
vocabulary, and meaning context in
the reading of familiar stories
and selections.
Direct instruction in advanced

decoding skills; wide reading (with
instruction and independently) of
familiar, interesting materials
which help promote fluent reading.
Being read to at levels above their
own independent reading level to
develop language, vocabulary, and
concepts.
At the end of stage 2, about 3000

words can be read and understood
and about 9000 are known when
heard. Listening is still more
effective than reading.
Stage 3:
Reading for
learning
the new
Grades48
(ages9
13)
Reading is used to learn new ideas,

to gain new knowledge, to experience
new feelings, to learn new attitudes;
generally from one viewpoint.
Phase A
Intermediate,
46
At beginning of stage 3, listening

comprehension of the same material
is still more effective than reading
comprehension. By the end of stage
3, reading and listening are about
equal; for those who read very well,
reading may be more efficient.
Phase B
Junior high
school,
79
Reading and study of textbooks,

reference works, trade books,
newspapers, and magazines that
contain new ideas and values,
unfamiliar vocabulary and syntax;
systematic study of words and
reacting to the text through
discussion, answering questions,
writing, etc. Reading of
increasingly more complex fiction,
biography, nonfiction, and the like.
Stage 4:
Multiple
viewpoints
High
school,
grades10
12 (ages
1517)
Reading widely from a broad range of Wide reading and study of the
complex materials, both expository and physical, biological, and social
narrative, with a variety of viewpoints. sciences and the humanities; high
quality and popular literature;
newspapers and magazines;
systematic study of words and word
parts.
Stage 5:
Construction
and
reconstruction
College
and
beyond
(age 18+)
Reading is used for ones own needs

and purposes (professional and
personal); reading serves to integrate
ones knowledge with that of others, to
synthesize it and to create new
knowledge. It is rapid and efficient.
80
Wide reading of ever more difficult

materials, reading beyond ones
immediate needs; writing of papers,
tests, essays, and other forms that
call for integration of varied
knowledge and points of view.
Reading comprehension is better than

listening comprehension of material
of difficult content and readability.
For poorer readers, listening
comprehension may be equal to
reading comprehension.
Reading is more efficient than

listening.
phonemegrapheme correspondences and the

integration of phonic strategies, visual memory and
use of context, with of course further vocabulary
growth and increasing knowledge of the worlds
which reading conjures up, combine to form the basic
achievement of reading. Top-down and bottom-up
strategies can now be used interdependently. They
become, no doubt, more fluent with practice just as
any other skills would, and the balance adopted
between different reading strategies probably
becomes more flexible and more efficient. There is
considerable development, in other words, in the uses
to which reading is put. Chall (1983) provides a table
(Table 2) to illustrate this point: reading is an
instrument for knowing, heavily influenced in its
development by the demands of the ecosystem, and
both an object and a result of practice in study skills
(Chall 1983, pp. 857).
Children and stories
There is a history several centuries old of producing

special story books for children, and behind that a
longer tradition of folk tales and rhymes and other
narratives which also took children for their
audience. Most children hear stories and tell them
themselves: many read them or write them. In this
country, at least (see Schieffelin and Cochran-Smith
1984 for notes on two other cultures), there is a
positive correlation between being read stories
during the pre-school years, on the one hand, and both
metalinguistic competence (p. 137) and prospects of
learning to read and write easily, on the other.
Children learn a great deal about written language
and about the structure of stories from being read to
(e.g. Clay 1979; Clark 1976; Wells 1985). They also
take part in a shared activity with the older person
reading to them (and indirectly with the author).
Where this is enjoyable for both parties, it doubtless
provides motivation for further reading. Involving
parents in hearing their older children read to them
has proved a most effective way of improving
childrens reading skills (e.g. Tizard, Schofield and
Hewison 1981).
Being read to can involve children in several
different ways of taking from text (Teale 1984).
Sometimes the child is required to be a passive

audience. Sometimes adults encourage a great deal of
active participation from the child, requesting
identification of objects in pictures, comments on the
action, and predictions about what will happen next.
Sometimes analogies will be drawn between the
events of the story and the childs own experience. If
such elaborations of the reading and listening activity
are sensitively used, children become adept at
relating stories to a wider context and dealing with
their implications more deeply, skills useful in the
classroom (Hayward 1980; Heath 1982, 1983; Mills
and Funnell 1983). Stories are a constant part of
classroom social life, and a delight in many childrens
home lives.
Stories socialize children in other ways besides the
interaction which is required of a reader or listener.
They can convey the culture to the child and socialize
him or her into culturally approved patterns of
attitudes and values. They can do this both overtly
and covertly, and can carry both admirable and
deplorable messages. Hilaire Bellocs Cautionary
Tales for Children, for example, has overt morals
such as dont tell lies, dont slam doors and dont
play with loaded guns, which are, however, less
memorable, and less enjoyed by children, than his
gleeful callousness. Louisa May Alcotts Little
Women proposes that girls should be quiet, ladylike,
self-sacrificing and dominated by men, a mode of
existence which was impossible and destructive for
Alcott herself (Saxton 1978).
This sort of socialization is hard to quantify. Its existence has however been a source of anxiety to many
adults who wish to censor the unsuitable out of their
childrens reading (and indeed out of other adults
too), and often have wished to substitute more suitable reading. There are a number of discussions of
the socializing effects of traditional fairy-tales, for
example (Tucker 1981; Steedman 1982; Zipes 1983).
We do not know whether women were more self-effacing and long-suffering as a result of being told the
story of Patient Griselda, the archetypal victimized
wife whose patience wins her back the affections of
her back the affections of her dreadful husband, or
whether there was more adultery between wives and
81
Plate 8
82
servants after the publication of Lady Chatterleys

Lover. Many of us may remember being (less
strongly!) influenced by a story we have read. In so
far as stories reflect the entire culture, and so are
reinforced by other experiences, they may have a
powerful effect. There is little clear evidence.
It is also argued that stories are important in
childrens emotional development. There are famous
psychoanalytic interpretations of fairy stories in
terms of childrens need to resolve their Oedipus
crisis or their penis envy, notably Bettelheims The
Uses of Enchantment (1978). Since these resolutions
are normally of unconscious feelings and problems,
it is again hard to demonstrate that stories have the
uses claimed. Stories may also model emotions
more directly, as they illustrate their characters
behaviour. Hayward (1980) told a slightly simplified
version of Watership Down to her class of 4-yearolds, and later observed a little boy with aspirations to
toughness and bravado bump his head, and instead of
crying say Miss Hayward, Ive hurt myself bad, but
dont worry, Im Bigwig the strongest rabbit Im
brave like Bigwig.
Stories obviously commonly evoke emotions like
play, they provide a relatively risk-free and
controllable form of being frightened, excited and
exhilarated.
Children begin to understand the form of stories
quite early in their experience of them. They are not
as dependent on temporal order as Piaget (1969)
suggested: given a logically structured story (his
were ill-formed) they make inferences and build up a
coherent shapely sequence of events even at the age
of 4 (Wimmer 1980; Mandler and Johnson 1977).
They develop story grammars just as they develop
scripts of familiar events in their lives. Although
their initial re-telling of stories contains mainly
surface events, probing with why questions elicits
much inference about the characters motives and
intentions. Young childrens limitations may be
attributed to lack of world knowledge and memory
problems rather than to lack of an ability to make
inferences and other logical connections (Trabasso
and Nicholas 1980). As children encounter more
stories they gain both more real world knowledge
and more knowledge of the conventions of stories.

They begin to appreciate the distancing, reassuring
opening Once upon a time; they expect there to be
good characters who triumph over their troubles and
bad characters who get their due come-uppance; they
know that foxes are sly and witches are wicked.
Traditional stories create a world simpler than the
childs own but not entirely unlike it, so that they can
try an alternative reality just as they do in play.
Skills of learning from text

Being literate becomes increasingly important as one
progresses through the educational system, both
because of the possibilities of gathering, rearranging,
comparing and passing on knowledge which it
provides, and because it is one of the commonest
social measures of intellectual adequacy. As we
have seen, reading involves many different levels of
cognitive processing. Reading a text in order to learn
from it requires the use of comprehension and study
skills.
Comprehension
Collins and Smith (1982) describe some of the

important comprehension skills necessary for
dealing with written information. The first group are
concerned with monitoring comprehension, that is
checking whether the text is being understood, being
aware of a breakdown in understanding and taking
appropriate action to remedy it. The reader may fail
to comprehend text at the level of a particular word,
phrase or sentence, or at the level of fitting bits of text
together. Children have limited vocabulary and
general knowledge, compared with adults, and may
make over-simple assumptions about grammar.
When acting out the spoken sentence The cat was
bitten by the dog, for example, they proceed as if the
passive sentence had the subjectverbobject order
characteristic of simple active declarative sentences.
Similar problems may be found in many early
readers: unusual grammar upsets the childs
comprehension. Unusual vocabulary items may
puzzle the child too, but they are probably more easily
clarified, and children often enjoy grand long words.
83
However, children may have more marked

comprehension problems than adults, even at the
simplest levels. Ellen Markman suggests that they
have major problems over comprehending the
integration of separate bits of text into a consistent
whole. They often do not notice, she says, what are to
the adult glaring inconsistencies or omissions in
verbal material, for example incomplete instructions
about how to play a game (Markman 1979; Robinson
and Robinson 1983).
Childrens comprehension of text, then, runs into
difficulties and the problem arises of how to solve
them. Possible strategies include ignoring the
uncomprehended words or passage; waiting to see if
its meaning later becomes clear; guessing, the guess
being confirmed or disconfirmed later; re-reading the
immediate problem passage or the larger part of the
text in which it occurs; seeking outside help, from
peer, teacher, dictionary or another text. Which of
these strategies the child uses will depend on the task
requirements; casual reading uses the simpler
strategies at the beginning of the list, while detailed
mastery of difficult texts requires much more rereading, analysis and comparison of different
sources. Which strategy is used will also depend on
the characteristics of the reader, and in the case of the
very young reader on how he or she is being taught to
read. However new reading is, the child has used
skills very much like many of these for several years
while trying to understand spoken language. The
most obvious skill to be applicable to written
language but not so readily to spoken language is rereading, both because spoken words are much more
ephemeral and because requests for repetition of an
uncomprehended utterance often produce a
rephrasing of what was said, not a verbatim
repetition. Guessing, ignoring and seeking outside
help (sometimes from the same sources) are however
well-practised strategies. Children, like all who are
relatively ignorant or novices, must rather often fail
to comprehend. It is, however, a question of some
interest to identify what and when they know that
they dont know, and what and when they know why
they dont know. This and other questions of
metacognition I will discuss later.
84
Collins and Smith (1982) point out that as well as

guessing at the meaning of particular words or
phrases, the reader will guess about the meaning of
more global aspects of the text. Guesses will be made,
or hypotheses constructed, about matters like what
sort of story this is, what will happen in this situation,
who are the goodies and who the baddies, and so
forth. (In many cases authors consciously use these
hypotheses either to support the plausibility of their
narrative (as in romantic novels) or to mislead the
reader in a sequence of bluff and double-bluff (as in
whodunnits). Jane Austens presentation of Mr
Darcy and Mr Wickham in Pride and Prejudice is a
particularly brilliant example of both uses.) Again,
children beginning to read will have developed
strategies of making sense of stories or events in
terms of broad hypotheses which define what this sort
of story is likely to be about, or what happens in this
sort of event. Schank and Abelson (1977) called the
latter scripts (see section on social cognition,
below). Careful observation of young children at
home shows that they both use and seek to clarify
scripts (e.g. Tizard and Hughes 1982, 1984) and it
has been argued that this use is a major opportunity
for the development of reasoning (Mills and Funnell
1983). Childrens experience of stories gives them
something like scripts for stories, as their filling-in
of the frame Once upon a time . . . and then they all
lived happily ever after shows (e.g. Applebee 1978,
and discussion of stories and of writing elsewhere in
this chapter).
Study skills
Even when the reader understands the text bit by bit

and as a whole, a considerable amount of work will
have to be done on it if the learning task involves the
memorization or the use of information from the text.
Verbatim memorization requires rehearsal;
remembering which is not verbatim requires other
activities, which may also be appropriate to using the
original text in creating ones own account of the area.
These activities have been called study skills. Ann
Brown and her colleagues (Brown et al. 1983) have
been investigating the types of knowledge and
strategies which students bring to learning from texts.

We are some way from the novice informationprocessors described in the earlier parts of this
chapter, but Brown et al. conclude that there is a
sequence of emergence of study skills very like the
sequence of emergence of remembering, attending
and reading skills. There is an early period of sporadic
use of appropriate activities; these activities become
more and more stable, systematic and consolidated,
and more easily applied flexibly to a range of tasks
and situations, as development proceeds. Mature
learning is in large part the result of strategic
application of rules and principles, and the
suppression of serviceable, but less mature, habits
(Brown et al. p. 90).
It is worth going into some detail on the study
skills reviewed because they illuminate
developmental questions through their progress in
schoolchildren, and because using them at a high and
conscious level contributes to intellectual effort well
beyond school level. Among the skills studied are:
identifying the main points, the most important parts,
of a text: identifying the organizing features of the
text: paying more attention to and using more
mnemonic activity on important and organizing
features: accurately estimating ones current state of
mastery and taking the appropriate action. Among
other investigations, Brown et al. looked at how
students worked on the task of summarizing a text.
They compared schoolchildren of 12, 14, and 17,
college students and teachers of rhetoric (experts).
Two simple rules of summarizing which even the
youngest children used quite accurately were 1 delete
trivia and 2 delete redundancy. Further rules were 3
substitute a superordinate term or a superordinate
event for lists of examples or episodes, a process
which can easily be related to the use of category cues
in recall; and 4 use (or invent) a topic sentence. There
was more use, and more effective use, of these
strategies as practice increased, but even college
students did not invent topic sentences on many
occasions when it was appropriate to do so.
One important point to come out of these studies of
summarization and note-taking was that the simple
strategy 12-year-olds mostly used, that of copy-
verbatim-with-deletions, is to some extent effective

in that it does produce a recognizable summary or
outline. However it is very passive and reduces the
chance that understanding of the particular text will
be linked to understanding of other texts. Experts are
much more likely to depart from the text, to combine,
rearrange, and interpolate information, to make
synopses in their own words and to argue with what
they read. This approach is more effective as a way of
understanding text, but requires more cognitive
resources (both process and knowledge). Students
who are managing fairly well by using passive
summaries (and who are not in their education
required to do more) may never give up their partially
adequate strategy in favour of one which is more
efficient (and more fun, if risky). The general public
level of information gathering and reading is not at all
clear: sophisticated evaluation, comparison,
criticism and recreation of texts is probably a
minority pleasure. However it is clear that both these
highly elaborated ways of playing with text and the
humbler study skills that underly them can be taught
(Brown et al. 1983, 1984; Howe 1984; Biggs 1984/ in
press). They are implicit in much school activity, and
many children gradually learn to use them through
schooling, but they can also be deliberately
developed. My own feeling is that they should be,
more often than they are at present: this is an area
where I regret my own state of bricolage!
Metacognition, motivation and study skills
I want to return for a moment to the Brown et al.

tetrahedron model of interactions between the
characteristics of learner, material, criterial tasks and
learning activities, by which at present I mean study
skills. Recent work by John Biggs (1984/in press)
illustrates that these interactions are crucial to
performance, and highly complex. He describes three
basic motives for study: to obtain a qualification with
minimal effort (Surface motive), to actualize ones
interests (Deep motive), and to manifest ones
excellence publicly by obtaining the highest grades
(Achieving motive). (He is mainly concerned with
motives for study in secondary and tertiary
85
education, but the same motives could no doubt apply

in primary education; young children may read to get
through the reading test, or to develop their own skills
or interest, or to show off what good readers they are.)
There are cases of mixed motives, particularly of
Achieving motives mixed with the other two
Figure 10 An elaborated model of student learning.

86
categories. Motives result from the interplay of

student personality and the task situation.
The different motives are paired with different
strategies. Thus the Deep motive tends to be
associated with strategies which emphasize under
standing the meaning of the task and the material, the

Surface motive with reproducing what is seen as the
essential information or performance and ignoring all
that is superfluous to getting by, the Achieving
motive with organization which optimizes the use of
time and effort. They have different outcomes.
Surface learners work is factual and superficial, and
has little intrinsic interest to the student who may
forget it the moment the qualification is obtained:
Deep learners is more complex, more structured and
more affectively satisfying, hence more permanent;
and so forth. The same study skill behaviours are
differently effective for students with different
motives; an Achieving motive student using an
Achieving strategy generally does well, but a Deep
motive student using the same strategy is likely to feel
dissatisfied, aware of the discrepancy between desire
and performance. Different educational institutions
foster different motives and strategies, through their
control of task materials and criterial tasks, and the
different learning activities they propose or require.
Thus there is a complex interaction between
student characteristics such as ability and selfconcept, situational factors such as course structures
and type of evaluation, the degree to which the
student is aware of his or her learning strategies and
activities, and the sort of knowledge and satisfaction
that results, via the approach that the student takes to
the learning task. Biggs (1984) provides a model
(Figure 10).
This picture suggests that a Deep approach to study
is more closely tied to the personality of the student,
who feels I am doing this for me, and not because the
school makes me. A Surface approach tends to be
seen as distant from the self, adopted in order to get a
task done when it has been imposed from outside. The
former is likely to involve more positive involvement
on the part of the learner, a greater awareness of his or
her learning activities, needs and achievements, and
an outcome which shows more understanding and
more links with other knowledge and less role
learning of facts. Surface learning will be less
involving, less enjoyable, often undertaken to cope
with the demands of outside authorities, and will have
much less impact on the rest of the students

understanding: it may be useful for quiz games such
as Mastermind but will never win you an FRS.
Writing
I have described some of the recent work on reading
and using text. I want now to consider the second R,
writing. Like reading, this is an activity which
integrates many different processes physical,
linguistic, cognitive, even social in different ways
according to the writers age, experience and
purposes.
Scardamalia
(1981)
lists
the
interdependent skills involved, among them
questions of handwriting, spelling and punctuation,
considerations of word choice, syntax and textual
connections, and of overall purpose, organization,
clarity and euphony. Each of these is itself of course
highly complex and subject to developmental
change. I shall focus here on the more cognitive
issues, discussing the relation of writing to reading
and speaking and the production problems,
particularly the composition problems this involves.
Writing in relation to speaking
Written language and spoken language differ, though

linguists disagree about how far the former is
dependent on the latter. Stubbs (1980), reviewing the
debate, points out that the relationship between
written and spoken language differs for different
writing systems, different authorial purposes, and
different cultures. One consistent difference,
however, is that the writer must construct the text
without the assistance of signals from the recipient
about whether the meaning is being understood,
whether more information is needed, whether jokes
are being appreciated or persuasive arguments are
having the desired effect. The reader has to get
meaning from the text in the absence of many of the
signals that accompany spoken language: cues from
the speakers body language, pitch, speed of
speaking, facial expression and so forth are not
available in conventional written text, though devices
of punctuation and typography may be used to make
up for some of this loss. Literary theorists debate how
87
far the meaning of a text is in the text itself, in text

plus additional information about the writer and the
context of writing, in the reader (with or without the
context of reading), or in all of these in
complementary or contradictory fashion. These
debates are typically rather distant from psychology,
though they do point to complexities which will be
relevant to childrens response to stories. My present
point is that writers and readers interact much more
distantly than speakers and listeners. Writers cannot,
alas, monitor their readers understanding. If they
wish to be effective, they must therefore compose and
review their writing more carefully than most
speakers need to do, lest problems in any of the skills
which Scardamalia (1981) lists should prevent
satisfactory communication. They also need to
appreciate some of the conventions which differ
between spoken and written language. Normally,
children learning to write have a good basic
command of spoken language, which they have been
using since early childhood and which they can use to
serve a variety of social functions. Their
communication in speech is, however, still
commonly dependent on context and on
paralinguistic cues to carry subtle or precise
meaning. They may also have little experience of
writing themselves or seeing others write, and
particularly of the uses to which writing can be put.
Writing in relation to reading
Before I discuss childrens achievements in learning

to write, I wish to point out that with reading and
writing or spelling the one is not simply the inverse of
the other. There is evidence that at least at the early
stages reading may be done by visual word
recognition and spelling/writing by a phonological
strategy (e.g. Bryant and Bradley 1980, 1983).
Children and backward readers may read correctly,
but spell incorrectly, words which are visually
distinctive but phonologically difficult, such as
light, and read incorrectly but spell correctly words
which are easy to construct phonologically (letterby-letter) but are visually nondescript, such as bun.
Bryant and Bradleys subjects are producing some of
88
their failures because they do not use their

phonological and visual recognition strategies
together: reading and spelling were unfortunately too
independent. As reading progresses beyond the 8year-old level, the different processes intertwine, and
the experienced writer may find it commoner to know
that a word is spelled incorrectly because it looks
wrong rather than because it sounds wrong. Given the
complexity of the relations between phoneme and
grapheme in English, embodying as it does an
immense amount of historical information, and
overlooking as it does considerable regional and
subcultural differences in pronunciation, sound may
be a less reliable index of spelling correctness than
sight in many cases. Readers certainly come to
recognize, understand and spell correctly words they
have never heard. The hero of How Green Was My
Valley reads misled as the past participle of a verb
to misle, pronounces it to rhyme with drizzled,
and is beaten by his schoolmaster despite his
comprehension of the word.
Writing as a physical skill
Writing involves the use of fine muscle movements in

varied but co-ordinated patterns, with visual
monitoring, at a speed (in experienced writers) far
faster than the brain can send messages to the hand
and receive feedback, though more slowly than
thinking or speaking. As Thomassen and Teuling
(1983) remind us, the hand is an extraordinarily
intricate and delicate mechanism, controlled by a
large number of muscles, and during writing
movements in fingers, wrists, arms and shoulders,
plus movements of the eyes (and perhaps head) for
monitoring writing, have to be co-ordinated. These
muscle movements have to be small, quick and
precise, and because the speed of transmission of
neural control messages along the length of the arm is
relatively slow, the brain cannot wait for one
movement to be completed before the next is begun.
Even at this level the brain must plan ahead to get the
hand in the right position at the right place at the right
time.
The Development of cognitive skills
To do this, we write not in single letters but in larger

units. As an example, compare the ease of writing
written backwards with the difficulty of writing
sdrawkcab nettirw: and note where the pauses and
breaks in writing came. Units may occur over
words; we pause more between clauses than within
them, for example. Overunitting may produce
errors; I planned the first sentence of this section from
Writing to patterns as it is now printed, but wrote
final instead of fine, as a conflation of fine with
the sound at the end of muscle rather than as a
semantic error. Motor skills of the highly complex
sort involved in writing fluently are acquired mainly
through practice. Smith (1982) advocates writing
practice free from demands for correctness in
spelling, punctuation or composition which may
distract from establishing automatic motor patterns.
This may or may not be a sensible bit of pedagogy, but
certainly children do take a fairly long time to
develop fine motor co-ordination sufficiently good
for writing neatly: some people never really do it!
Learning about the basic skills of writing
Given a piece of paper and a pencil, a 2-year-old will

make marks on the paper. Initially these scribbles are
probably not unsuccessful attempts to write, rather
they are successful attempts to make marks on the
paper where there were no marks before. Writing
usually follows attempts at drawing representations
Figure11
of objects. Vygotsky (1978a) points out that writing is

a second-order symbol system: the letters stand for
the spoken word, which itself stands for the object.
Drawings, and idiosyncratic pictographs, are in a first
order symbolic relation to the object they represent.
This account obviously implies that writing will
develop after drawing has begun; writing is a special
sort of drawing which represents language. As
children experience written language by reading or
being read to, they learn that conventionally writing
(at least in English) comes in horizontal lines and
consists of particular sorts of patterns. Young
children may write in a scribble or string of letters
running horizontally across the page, or
accompanying a picture as an adult might provide a
caption. A fascinating observation of a child just
under 5 illustrates this.
Unlike the case in other writing episodes, Coline is not
oblivious to the activity around her. She sits with a group of
boys, rather than alone, and watches intently as Raymond
draws rockets and creates the accompanying sound-effects.
The entire writing episode seems to be surrounded by more
language. For the first time Coline begins the composing by
verbalizing her intentions. She lays two blank pieces of
paper on the desk before her and announces: Im going to
do a picture on that one and write on this one. This, in fact,
is what she does. The numbers beside the p.i.s indicate the
order in which they are written by Coline [Figure 11].
89

This product indicates more progress in Colines writing.
Her handwriting is more controlled, the size of her letters
more uniform, her boundaries more defined than before.
Her composing process shows how demanding this new
step forward is. Coline spends 20 minutes just writing the
first three p.i.s and throughout the composing actively
seeks interruptions. She chats with the boys and comments
on their pictures. She writes two letters, then noisily taps her
pencil on the table. She completes a p.i., then rocks her chair
back and forth on its squeaky back legs. There is much
physical turmoil in these pauses from the writing.
The length of time spent composing and the effort
expended in making well-formed letters possibly affects
Colines page arrangement. Instead of composing
horizontally as in the June piece, she returns to a vertical
page arrangement. While composing, however, she does
not proceed in a neat progression from the top of the page to
the bottom. Instead, she begins by writing the first p.i. at the
top of the page, the second at the bottom, the third in the
middle as follows. The transcription to the right of the p.i.
indicates Colines reading:
1. tkoic
2. kioe
3. okki
cat
sat
mat
Coline pauses, then rereads what shes written as follows:

1. tkoic
3. okki
2. kioe
cat
sat
mat
Although she does not compose top to bottom, she

rereads her writing in this way. She begins at the top of the
page and moves down pointing to each word in the order it

appears on the page, an order quite different from her initial
composing. Still the meaning stays the same. She writes 4.
tkkoic says, Im finished and reads again:
1.
4.
3.
2.
tkoic
ttkoic
okki
kioe
cat
sat
on
the
Got to write mat she says to herself, forgetting that a

few moments before she not only called both okki and kioe
mat but also said she was finished. Each word addition
demands a rereading by Coline. Each rereading in turn
changes the correspondence between p.i.s and her
meaning for them. She writes 5. kktols, presumably to be
mat. But the positioning of the word on the page ensures
that she calls it the instead, as happens:
1.
4.
3.
5.
2.
tkoic
ttkoic
okki
kktols
kioe
cat
sat
on
the
mat
Things seem to be going well for Coline. Her demand

that voice and print match is being satisfied. She rereads
again and maintains the same match. Then, composing at
the point of her pencil, she says eating. Realizing, of
course, that there is no corresponding p.i. for eating she
adds one more, 6. tkok. The story complete, there are now
six clearly defined p.i.s on the page before her and six
clearly defined words in the message in her head. The
reading should be straightforward. This, however, is not the
case. Seven oral rereadings follow until Coline finally
makes the match as charted [Table 3].
Table 3
Written text
1. tkoic
4. ttkoic
3. okki
5. kktols
2. kioe
6. tkok
90
Rereadings
13
Rereading
4
Rereadings
56
Rereading
7
cat sat
on
the
mat
eating
cat sat
on
the
mat
eating
oranges
cat
sat
on
the
mat
eating
cat sat
on
the
mat
eating
oranges

For some reason in the first three rereadings she
pairs two spoken words, cat and sat with the one
p.i. tkoic. Consequently she has one p.i. left over at the
end, tkok, and no spoken word to go with it. On the
fourth rereading she solves the problem by adding
another spoken word, oranges. All should go well
now. On the fifth and sixth readings, however, she
returns to her earlier behaviour of matching one
spoken word to one p.i. As a result she now finds there
is no written p.i. left to match the spoken oranges. At
last, on the seventh rereading she remanoeuvres and
makes the match.
Such persistence! Coline will not stop until she
reaches such closure. She adjusts and readjusts her
division of the message until she succeeds. This
selfimposed reading seems to be her way of gaining
control of the writing.
Following her seventh rereading, she makes her
illustration on the second piece of paper she put aside
at the beginning of the composing. She seems pleased
with herself as she draws a picture of a smiling cat and
affixes a label tkko saying, See, I writed cat at the
top.
The message itself sounds like a rehash of a basal
reader story line and is not interesting as a product.
The process of getting there, however, is fascinating.
Getting the message down in primary inventions is
demanding. Coline pays most attention to the
demands of making print and controlling the letters,
less to her information. It is possible that while she
writes she first makes her random clusters of letters
and only later attributes meaning to them. Certainly
children do this with their art. Perhaps there is a stage
in the development of young writers where they write
first, mean later (Kamler and Kilarr 1983, pp. 1879).
Once children have achieved a significant insight into

what writing is (Jarman 1979, Smith 1982; Kamler
and Kilarr 1983), they have now to curb their
inventiveness and adopt the conventional symbols
and construction techniques and spellings required
to communicate with adults. Children do not write
letters badly or incorrectly because of inadequate
visual processing, but because of production
difficulties: letters have to be precisely formed in

direction, size, joins, spaces and so forth, and the
motor control needed for this is very considerable.
Practice gives us a fluent production system, but
children lacking one make the sort of errors and
attempt the sort of solutions that they do in drawing
(see Freeman and Cox 1985).
Some recent observational studies of young
children writing are integrated by Clay (1983) into an
account of how children develop a theory of writing.
As in developing spoken language or reading, the
childs own experience of writing seems to be an
important determinant of rate of development:
children from highly literate families are more
positively motivated and more knowledgeable than
children whose experience of writing is scant. Some
children invent their own letters (Hughes, 1984,
reports children who invented their own numbers and
arithmetical representations, see p. 103). Many
invent their own spellings. One fairly common
system is to represent consonants but not the vowel
sounds which are less prominent in speech, writing
LTL for little, for example. It is worth pointing out,
perhaps, that vowel deletion is common in strategies
for writing fast, and that some Middle Eastern
orthographies do not write in the vowels. As children
become more familiar with the standard spellings
they meet in their reading they move towards the
conventional system: pressure from adults also
contributes to this shift.
The functions of writing
A reader is essentially concerned with extracting the

meaning of a pre-existing text, with or without
increase in enjoyment or knowledge. A writer may
have any of a wider range of purposes, for example to
inform, to entertain, to persuade, to criticize, to
record, to express something personally felt, and so
forth. This list is obviously similar to one listing the
functions of spoken language: recording is the major
exception, as generally it is better done by writing
than by speaking. Childrens spoken language will
have been used for most of these functions by the end
of the pre-school years (Wells 1985; and see Chapter
91
4 this volume). They may not have experienced all

these functions in writing. Although most adults
write sometimes, the main uses of writing at home are
in response to social pressure or practical
requirements: domestic messages, shopping lists and
family letters. Griffiths and Wells (1983)
investigated use of writing in a sample of Bristol
adults and found the group differences shown in
Figure 12.
The Bristol Language Development Study found
adults literacy was related to that of their children;
parents who write more probably also read more,
thus giving their children more experience of how

writing can be used. It must be rare, however, for preschool children to have experience of adults writing
to express themselves, to tell stories or to record their
own experience, since when adults do these things
they do them privately. There is little evidence on
childrens use of writing outside school. Tizard and
Hughes (1984) describe some mothers teaching their
daughters to write, sometimes as a session of letter
forming or copying words written by the mother,
sometimes as part of an activity such as writing a
Figure 12 Relationships between amount of writing in different categories and sex, social class and education.
Source: From Griffiths and Wells (1983).
92
letter to grandparents. The latter is probably a fairly

common real-life experience for quite a lot of
children; most available examples seem to be
conventional rather than expressive. Parents may
encourage (or require) children to write to
grandparents or to people who have given presents or
hospitality: thus there is a real social purpose to
writing. There are clear rules for beginning and
ending the text of letters and since there is often a
specific interpersonal reason for writing the letter, at
least some of what it must contain is also specified. A
thank-you letter from Z to X acknowledging a gift, Y,
must include at least the phrases Dear X, Thank you
for the Y, from Z. Collerson (1983) collected the
letters his daughter Juliet wrote between the ages of 5
and 9 1/2; they show an increase in length via the
inclusion of informative or expressive material
beyond the minimum demanded by the formal letter
scheme, and an increasing tendency to use written
language as a means to a continuing dialogue with
people who are known but too far away to speak to.
The child learns that letters can be a means of
reporting and interpreting experience, a device for
exchanging information, and a method of
maintaining social interaction among friends
(Collerson 1983, p. 92). Some children have
difficulty seeing these purposes in school writing
(Francis 1982; Tamburrini et al. 1984).
This list of functions that are served by Juliets
letters resembles the lists of language functions
provided by Halliday (1975), which is discussed in
Chapter 4, by Smith (1982, p. 14), and others. Writing
and reading differ from speaking and listening not so
much in the language functions involved as in the
possible distance between the participants. As I said
when discussing the relationship between writing
and speaking, writing is a more abstract activity.
Written text has to convey its meaning more
independently of paralinguistic context as writer and
reader do not have the shared immediate context
which speaker and listener enjoy. The text must
create its own context, so the writer must assess what
knowledge can safely be assumed and what must be
incorporated into the text. The writer must also assess

what is the best order for pieces of text and present
each piece unambiguously and explicitly. It may be
easier for the reader to go back and read text again
than it is for the listener to recall speech to re-examine
it, but the writer cannot adjust content, order or
emphasis as the speaker can in response to the
readers cues of understanding or failure to
understand. On the cognitive and social levels,
writing differs from speaking.
Writing as composition
Cognitive and social considerations are mingled in

the composition of text. Text has to be planned,
composed and revised in terms of, among other
things, its probable success in communicating with
the reader and its own explicitness and coherence.
Martlew (1983) reviews some of the errors that poor
or inexperienced writers make: they write as they
speak, leaving their writing dependent on a context
which the reader does not share; they plan poorly, if at
all, and prepare themselves for writing too briefly to
produce clearly organized text; they do not review or
criticize what they have produced. Some of these
errors no doubt arise because the whole task of
considering the adequacy of the text plus spelling
correctly plus writing neatly plus producing the right
amount to satisfy the teacher overwhelms the writer.
Some, however, probably stem from young writers
uncertainty about what to do in composition and how
to do it. Frederiksen and Dominic (1981) suggest that
important cognitive resources
include the writers knowledge, the already established
strategies and procedures for constructing a meaning and
expressing it, and the general characteristics of their
cognitive systems such as processing capacity and both the
automaticity and efficiency of component processes
(Frederiksen and Dominic 1981, p. 4).
A number of researchers have recently

investigated childrens composition techniques (see,
for example, the collections of papers edited by
Frederiksen and Dominic 1981; Martlew 1983; and
Kroll and Wells 1983; and a review by Brown et al.
93
1983; as well as the paper specified as reference

below). An early problem in composing is generating
content. Young children frequently produce the
equivalent of one utterance on the topic and then stop,
claiming that that is all they can think of to say.
Bereiter and Scardamalia (1982) demonstrate that
devices of various sorts increase the amount
produced. Instructions to produce a large amount, the
opportunity to speak or dictate the text instead of
having to write it, the provision of simple prompts
such as go on or more directive ones such as on the
other hand or also, all increased both number of
words and number of ideas expressed. Children
clearly had not reached the limits of what was
available on the subject. They welcomed prompts
and appreciated their effect. Learning to provide
yourself with prompts is part of the development of
writing skills.
Bereiter and Scardamalias subjects did not,
however, seem to fit their additional content neatly to
the prompts or to the sentence openers which later
experiments provided. (I am glad to say that a group
of Bristol 9-year-olds, less formally studied, did.)
Prompts and openers produced more material, but not
all children provided material on the same side of the
argument as an also prompt suggests, or material on
the opposite side after a prompt of on the other hand.
Their productions resembled what has been called a
knowledge-telling strategy, where everything that
is known on a subject or part of a subject is allowed to
flood out without much evidence of organization.
Writers who use this strategy often pay little attention
to the demands of their title and the limits it sets, and
do not adjust what they say to the characteristics of
their reader. Knowledge-telling writing lacks
goalrelated planning and significant revision. The
text is not interconnected but made up of unrelated
sentences produced one after the other without
reference forward or, more particularly, back. It may
thus contain repetitions and contradictions. Children
seem to have difficulties spotting inconsistencies and
inadequacies in text (Brown et al. 1983; Markman
1979) and find it difficult also to estimate accurately
whether they know a little or a lot about a topic
(Bereiter and Scardamalia 1982). In other words,
they have problems of access to their knowledge.
94
Knowledge-telling is a strategy which even

experienced writers may resort to under conditions of
stress (such as writing examination answers!);
Brown et al. (1983) point out that it may be hard to
give up because it does produce text, indeed text
which is high in quantity even if low in relevance and
organization. Knowledgetelling followed by
rigorous revision and ruthless discarding of weak
material is a perfectly respectable composition
technique. Since, however, only relatively
experienced writers produce the evaluating and
revising parts of this procedure, it may be more
effective for young children to be trained to plan their
text before they write it.
Bereiter and Scardamalia (1982) report various
attempts to induce and improve planning. Planning is
one of the ways in which composing written language
differs from spoken language. Spoken language is
much more likely to be influenced by the recipient
who may interrupt, argue, anticipate and so forth,
thus changing the course of the speech. Planning in
speaking has to be flexible and adapted in use to the
behaviour of the listener: planning in writing is much
less disruptible, and it may be more important to get
it right. Before the teens, young writers seemed to
plan, if at all, on a what next? basis, even when
prompts suggested different developments in a text.
Planning was local, of the present sentence in relation
to the previous one, not in terms of how it might fit
into the text as a whole. One strategy the researchers
used to try to induce larger-scale planning was to
provide the children with a final sentence for their
composition. They report their best discussion,
between 12-year-olds given the task of composing a
story ending with the sentence And so, after
considering the reasons for it and the reasons against
it, the duke decided to rent his castle to the vampire
after all, in spite of the rumour he had heard.
Discussion concentrated, unfortunately, on worldknowledge problems such as why a duke might
consider renting his castle to a vampire, and on
separate literary considerations (this whole story is
getting kinda dumb), without managing to resolve
the tension between the two. Most children stuck with
the world-knowledge problems, and did not get on to
the more abstract problems of creating a good story at

all.
Constructing a complex story in a backwards
direction is a difficult task, as indeed is doing so in a
forwards one. A great many highly esteemed novels
and stories consist essentially of a string of incidents
which might well have been generated by a what
next strategy. The Pickwick Papers and Don Quixote
could be seen as examples. The writers have however
had in mind at least a unifying theme and possibly an
overall shape to the novel which have subsumed the
individual elements into what reads as a coherent
whole. A sequential composition strategy such as
children use works well in the context of an overall
high level plan.
Revision
Experienced writers plan their texts: they also review

and revise them. Children mostly fail to do so.
Egocentrism or attachment to ones own text is not
an adequate explanation of this (or of many other
childish foibles, see Cox 1980; Ford 1979), since
children also fail to revise other peoples productions.
Revision requires a highly developed ability to
correct and improve as well as to generate text,
involving treating the output of the nth attempt at
writing as input to be revised into the (n + 1)th
attempt. Bereiter and Scardamalia (1982) argue that
children do not have an internal feedback system to
use in evaluating their text. Just as young speakers
seem to have problems in distinguishing between
what was meant and what was said (see Chapter 4),
young writers find it hard to assess how well what is
written conveys what is meant. As I have said, the
writer does not have the feedback of
incomprehension which the listener gives the
speaker. If he or she also has no way of getting selfgenerated feedback which represents accurately
what the reader would provide, there is very little
prospect of diagnosing a need for revision, or, further,
of acting appropriately to improve the text.
Scardamalia and Bereiter (1983) trained children
aged 1014 to evaluate each sentence of the text they
had produced asking whats the main problem with
this? They had a list of evaluative phrases such as

People wont see why this is important, I think this
could be said more clearly, and Im getting away
from the main point, and a set of directive phrases
such as Id better give an example or Id better
change the wording. Even the youngest children
could choose evaluative phrases which seemed
appropriate judgements to adult experts rating the
text, and the children also said that the phrases were
helpful and enabled them to review their writing in
ways they did not feel they were able to do in their
normal writing. However, the directives that the
children chose did not appear to be helpful either in
the view of the adult judges or in terms of the changes
made in the text. There were more changes for the
better than changes for the worse, but the revised
essays were not on the whole noticeably better than
the original ones.
These children were able to recognize problems in
their texts but not to make effective revisions. In part
this may be because their revisions seemed to be
mainly of minor details, rather than changes which
coped with major problems at levels nearer the whole
text. Scardamalia and Bereiter suggest that
one likely cause is incompletely developed mental
representations of actual and intended texts. These
representations may be developed to the point where the
child can detect that something is amiss, but not far enough
for the child to discover what it is. This is analogous to the
experience one may have in travelling somewhere over an
indistinctly remembered route. One senses that things do
not look right and therefore begins to suspect that one has
taken a wrong turn, but the mental representation of how
things should look is not sufficiently clear to indicate where
the wrong turn might have been made, or even to establish
definitely that one is off course. A common response in the
travel situation is just to keep going and hope things will
become clearer. This is what children seem to do in writing
(Scardamalia and Bereiter 1983, p. 93).
It seems possible however that young writers who

have the rather difficult and not quite clear task of
producing a better approximation to the ideal text
suffer from attention problems as well as a feeling of
95
not knowing exactly what to do. They may have a

repertoire of alternative phrasings but be mesmerized
by the original version: words on the page often seem
to take on a horrid inevitability. The alternative
phrasings have to be thought of the access to
knowledge problem discussed above and each
evaluated in comparison with the others, which will
place very formidable demands on memory and on
the childs powers of switching between composing
and criticizing. Again, Bereiter and Scardamalia
(1982) suggest there may be some improvement
through exercises such as highlighting the main idea
of a sentence, planning the development of a passage,
prcis-making, reading with an eye open for
technique and the various other activities that have
been traditional parts of learning rhetoric.
These analyses of the demands and difficulties of
planning, composing and revising may seem dry and
more likely to stifle childrens interest in writing than
to increase it. It is certainly the case that the intrinsic
interest of particular tasks often leads children to
write with enormous enthusiasm and care, and at
great length (see, for example, Steedman 1982).
Often the spur seems to be an emotional drive from
inside the child, a need to express and so control
feelings about an important experience. Bereiter and
Scardamalias subjects however seemed to enjoy
applying their skills to tasks they had not chosen, to
value the craft skills they had learned for their own
sake. Their feeling of having learned useful new
techniques which they could use to win increased
success on tasks which are highly esteemed is very
reminiscent of what Griffiths and Wells (1983) report
of English adults, that they did not feel competent as
writers but that writing was important and necessary
and they wished they did it better.
Children who are fluent writers
A number of children write for their own pleasure,

producing a large opus of stories, poems, and nonfiction. How many such children there are is
unknown. The Newsons (1976) found that 23 per cent
of their sample of 7-year-olds wrote a lot at home
for pleasure: girls did so more than boys, and middle96
class children more than workingclass children. This,

however, included writing domestic notes and letters,
and copying words without much regard to their
meaning. So far as creative writing is concerned, one
expert I asked said at least 10 per cent and junior
school teachers estimated that there would probably
be a couple in each class. Habitual writing seems to be
a feature of the early lives of literary figures far more
often than not: the Bront family, writing long
chronicles of their imaginary islands, Jane Austen,
writing a dramatization of her favourite novel Sir
Charles Grandison and a burlesque History of
England, and Edith Sitwell, composing poems and
fairy stories, are typical examples. Some write as
children but give it up in adulthood Daisy Ashford
followed The Young Visiters with some other novels,
also parasites on the sort of novels by adults she was
reading at the time, but as an adult published only
writing done in childhood.
Childrens writing reaches the world beyond the
writer largely through adult attention. Secretive
children who have a fair degree of personal space
may not be known to be writers. Much writing which
comes to the attention of parents (or teachers) gets no
further. In terms of its literary merit this is not
unreasonable, since much of what children write
lacks the craft which would make it satisfactory
reading to anyone at all distant from its producer. As
Steedman (1982) cogently argues, however, adults
often have ulterior motives for preserving and
publishing childrens writing. Literary merit is
considered in some cases, particularly I think in texts
concerned with how to teach writing. Parental pride
is another fairly innocent motive. Very commonly,
however, adults use childrens writing to interpret
childrens development, much as they would also
interpret their play, dreams and talk. They also select
writing because it exemplifies what they believe is
typical of childhood in the mid nineteenth century,
for example, childhoods innocence, transience and
spirituality (Coveney 1967). Some is published
because it amuses adults, its artlessness affording
them relaxation, some because it provides a moral
object lesson and edification. Adults may supervise
and correct childrens writing, and children may

tailor their productions to the susceptibilities of
adults. Given such multiform selection and
censorship, the range of content, of style and of genre
that children produce in their writing can probably
not be discovered.
There are, however, examples which indicate what
the range might be and allow some speculation about
children as writers. The research of Donald Graves
and his colleagues (Graves 1983; Calkins 1983)
provides some material. Bissex (1980) published
detailed observations of her sons writing (and
reading) from age 5 to age 11. Paul Bissex wrote
frequently, and used a considerable variety of forms;
captions, stories, directions, catalogues, newspapers,
messages, school-type exercises, rhymes, planned
schedules, diaries and ultimately codes. Initially his
writings functions involved achieving competence
in the act of writing, then sharing it with his parents,
and using it to name the objects of his world, for
example a sign on a cupboard door read:
PAULZ RABR SAF RABRZ KANT GT EN
(Pauls robber safe. Robbers cant get in.)
(The Zs are examples of his good phonetic spelling.)
As he grew older, he used his writing to categorize and
organize the world, for example drawing up lists like this
one (at age 8:7):
Hard words (to spell)

antidisestablishmenterianism
satisfaction
complicated
carbonated
scientific
corrosive
character
exploition
exploraition
developed
interesting
antimated
fictional
individual
annahilliated
irrataitional
antimated
extinct
puncuation
intellectual
He wrote newspapers, poems, stories with inventive

plots, developed styles appropriate to the recipient of
his letters, and showed rather the same preferences in
his writing as in his reading. Throughout, being
literate was important to him and a source of great
enjoyment.
Steedman (1982) presents a rather different
example. She discusses one main text, a story called
The Tidy House, produced by three 8-year-old girls
over a week in 1976. The characters of the story are
two couples and their children: the plot is simple: it
is concerned with the getting and regretting of
children, with how to bring them up and with what
life is like in the tidy house. The story is interesting
in the literary devices used but also for the picture it
provides of these little girls understanding of
domestic life. In her extensive discussion, Steedman
argues that producing the text was a way of
confronting the social system they lived in, so far as
this could be done by children who were captive both
in the classroom and in the sort of lives they were
describing. It is perhaps significant that the last page
of The Tidy House ends in mid sentence, thus:
soon they went home and had tea and went to bed and
At 8, it may not be altogether easy to accept or to

reject a future consisting of loving children who cry
and whine. These working-class girls display in their
writing and their commentary on it a consciousness of
their families lives which must dispel any belief that
childrens social understanding is egocentric or
nave, and also a sophistication in their planning and
revision of text which corrects some theorists
emphasis on childrens disorderly production of
writing.
Arithmetical skills
Having sketched the development of reading and
writing, I want now to look at childrens arithmetical
skills. This is a busy research area at present, with
many fine-grained analyses of mathematical tasks.
Mayer (1985) provides a brief introduction to this
97
work: among the important collections of papers are

books edited by Brainerd (1982), Ginsberg (1983)
and Lesh and Landau (1983). I will discuss some
aspects of the development of number concepts and
their role in childrens handling of addition and
subtraction. I will try throughout to bring together
task analytic approaches, ecological work on
childrens
spontaneous
mathematics
and
considerations of mathematical language.
Learning number words
One rather basic aspect of using numbers is knowing

number words, their names. Observations of preschool children at home and at school show that they
quite frequently encounter number words and
number word sequences in nursery rhymes, songs,
and adult conversation (see, for example, Tizard and
Hughes 1984; Davie et al. 1984). Adults often make
use of opportunities to use numbers; parents may
deliberately teach their children to count in the
senses both of reciting number words and of applying
those words in finding out how many objects there
are. Fuson and Hall (1983) find that a majority of
children can count to twenty or so by the time they are
5, and most can count to one hundred by the age of 6
given moderate practice in school.
In English, number words are regularly structured
from the teens, with a units part, a decades part and
a hundreds or thousands part as appropriate. Thus
1986 is one thousand, nine hundred, eighty, six (or
nineteen hundred, eighty, six); 35 is thirty, five;
and so on. The number names up to the teens have to
be learned more or less by rote, as do the names of the
decades; the rest can be generated by rule. Childrens
learning of number names seems to be initially by
rote. It is usually after they have learned to recite the
sequence beyond twenty that they realize there is a
repetition of something-one, something-two and so
on, and after they have grasped the sequence to one
hundred that they can really generate numbers
iteratively ad infinitum (Siegler and Robinson 1982).
Understanding zero and infinity (and negative
numbers, fractions, etc.) comes of course later still.
98
Counting
Knowing the number words is not, as Piaget (1952)

pointed out, the most important part of number
concepts. Number words have to be matched to
countable objects if the child is to count. Accurate
counting requires that each and every object must be
tagged with one and only one number word. To do
this, objects and words have to be kept in step, and
objects have to be repeatedly and accurately divided
into already counted and not yet counted groups.
Children often use pointing or touching or moving
objects to help themselves do this correctly: rhythms
in the counting may also help. There is quite a lot of
evidence (Fuson and Hall 1983) that childrens
counting does mostly allocate one number word to
one object; it is less clear whether children regard this
as a necessary principle of counting as Gelman and
Gallistel (1978) suggested. They do not always use
the conventional number sequence; young childrens
number lists are often idiosyncratic with an initial
conventional and correct sequence such as one, two,
three followed by a stable but unconventional
portion five, eight, nine, eleven and a final nonstable portion which varies from occasion to
occasion.
Having applied a stable list to the countable objects
in a one to one fashion, the last number used indicates
the numerosity of the set: its cardinal number. Adults
can produce the cardinal number for small sets (up to
four) and very well learned larger sets (such as the
configurations on dominoes and playing cards) very
quickly and without counting. This capacity is known
as subitizing. It is possible that young children can
also subitize small sets of two or three items. It is
however hard to distinguish between rapid covert
counting and pattern recognition or subitizing (see
Gelman and Gallistel 1978, pp. 648, 219 25; and
Fuson and Hall, pp. 5961): until it is clearer exactly
what subitizing consists of, it will be difficult to find
out how counting and subitizing are related in
development.
To find out how many objects there are in a set, one
should count them and then report the last number
used which is the cardinal word for that set. Young
children often fail to report the last number, but when
they have counted the set aloud they may be

expecting the experimenter to realize that the last
number said while counting is intended as the answer;
it is indeed somewhat disingenuous of the
experimenter to require further telling. It appears that
most middle-class children do repeat their last
counting word with added emphasis to indicate that it
is their answer by the age of 4 (Fuson and Hall 1983).
Appreciating that you would get the same cardinal
number for any correct count of the objects in any
order is a rather later achievement, as Piaget pointed
out years ago, and as in the case of his mathematician
friend (p. 33) it may be an exciting thing to realize. It
is of course an achievement which depends on
accurate counting: if your procedure in counting is a
bit shaky, you may get a different numerosity each
time and not realize the source of the variation.
Comparing numerical quantities
Although there are possible comparisons of numbers

implicit in producing strings of number names and in
counting a single set of objects, once there are two (or
more) sets there are inevitably comparisons which
involve judgements of more, less, fewer, same, equal,
greater than, etc. Siegel (1978) reviews studies of
childrens problems with these terms. One source of
confusion is that comparatives are correlated in the
natural world: certainly when comparing sets of the
same objects, the longer line does more often than not
have more in, or the taller heap contain more objects.
Sorting out the comparative dimensions, deciding
that less is not a synonym for more (Donaldson
and Balfour 1981) and sorting out exactly which
entities are being compared (Siegel et al. 1978) seem
to be among the harder preliminaries to making a
comparison.
Up to a point, perceptual comparison strategies
may give quite successful results. If numbers are
small, or if there is a big difference between the sets
to be compared, or if an approximate answer will do,
relative judgements work quite well, and young
children use them quite readily (Bryant 1974). They
work less well with exact comparisons between
larger sets nearer in size, and do not work at all of
course on number symbols: 71 is not obviously

bigger than 48. New strategies therefore have to be
developed, and the old ones given up, which, as we
have seen, is difficult.
One possibility is matching each item in one set
with an item from the other: if a set has one or more
items left over after this procedure it has more items.
The evidence reviewed by Fuson and Hall (1983)
suggests that if children have perceptual cues like
touching, moving and linking with drawn-in lines
(techniques which also helped with counting, as we
saw in the previous section), they can use matching to
establish relative numerosity by the age of about 6. If
matching is perceptually harder fewer children use it
(Brainerd 1979).
Another possible strategy for establishing relative
numerosities is of course counting. If the cardinal
numbers of sets are the same, they are equal in
quantity: if not, the order of the cardinal numbers in
the sequence of number words shows which set has
more and which fewer items in it at least to someone
who knows the number sequence and that if number
a comes before number b then number a is smaller
than number b. (Exactly how children make these
order judgements is still controversial (Fuson and
Hall 1983, pp. 938; Siegler and Robinson 1982, pp.
267 86). They are probably easier to make if the
smaller number is very small and the gap between the
numbers is large: Siegler and Robinson suggest that
there is categorization of the number sequence into
smallest, small, medium and big numbers. Preschoolers judge nine to be a big number in all
contexts: adults judge it to be a big number in the
context of one to ten but not in the context of one to a
million.)
Counting may be of two separate sets existing
simultaneously, or of one set before and after a
transformation. In the classic number conservation
test, both comparisons are involved. It is a wellreplicated finding that children who have counted
both sets before the transformation, and agreed that
they are equivalent, will deny that same equivalence
when one set of counters is spread out, even though
they may count again and reach the same numbers.
99
Nevertheless, counting does help in conservation of

number tasks (Siegler 1981; Klahr and Wallace
1976). It is presumably not used as the best cue to
numerosity in conservation when the child has a
strong perceptual strategy and a low level of
confidence in his or her own ability to count
accurately.
Figure 13 Three counting models for simple addition of m + n.

Source: From Mayer (1985), p. 141.
100
Ordinal numbers
We have discussed using numbers to find the

numerosity of a set or the relative sizes of several sets.
Numbers can also be used to describe relative positions
within a set: I am the fourth of my parents children, the
second tallest woman in my department, an inhabitant
of the seventh largest city in England. Children do
learn ordinal wordssuch asthese, butsome asthese,but
such as these, but some time after they have learned

the conventional counting sequence, and probably
the ordinal number sequence becomes systematic
and iterative through school practice. Using the
ordinal words to refer to the correct items of an
ordered set indicating the third, fifth or whatever
item is also difficult. Putting items in order of size
or some other physical quality is something preschool children enjoy doing, though as Piagets
seriation experiments show they may stick at
comparisons of pairs of objects and not
systematically complete an ordered set (Piaget 1952).
They find it easier to identify the smallest and largest
or first and last of a set than the middle-sized or
next to last one (Siegel 1972, 1978). How ordinal
number development is related to cardinal number
development is not known: ordinal number
development may lag behind because it is
ecologically less common (though ordinal numbers
are used daily in the calendar) or because its
vocabulary is acquired later or both.
Doing addition and subtraction
There has been a great deal of work recently on

childrens strategies for adding and subtracting
numbers. There are a number of useful reviews of the
area, including Carpenter, Moser and Romberg 1982,
Carpenter and Moser 1983, Resnick 1983 and Siegler
and Robinson 1982. Childrens most basic strategy is
to construct sets of physical objects or fingers to
represent the two numbers to be added, move the two
sets together and then count all the joint set from one
to the total. A more sophisticated (and quicker)
strategy is to count on from the cardinal number of the
first set through the members of the second set to the
total number of the combined set. More efficiently
still, the child may count on from the larger number
through the smaller set to the total the counting on
(min) algorithm. These three algorithms are shown in
Figure 13.
Reaction time evidence (Groen and Parkman
1972) suggests that in their first year of school
children use the counting-on (min) algorithm, as it
took longer to reach the correct answer if the
minimum addend was two than if the minimum

addend was one and so on. The mean response times
for 1 + 6 and 6 + 1 were identical and over a second
shorter than the mean response times for 3 + 4 and 4 +
3, for example. Double sums such as 1 + 1 and 2 + 2
were notably easy: this probably reflects a fourth
strategy for coping with addition, the memorization
of number facts. Resnick (1983) suggests that adults
do most of their addition and subtraction by using
mental shortcuts such as using memorized number
facts and repartitioning numbers to fit well-learned
facts (for example turning 18 + 23 into 18 + 2 + 21).
Even in the early school years, as Groen and
Parkmans data show, children use these strategies
(Figure 14). These addition strategies form a
hierarchy which most children pass through in one
way or another before reaching automatic response.
Carpenter and Moser (1983) describe similar
strategies for coping with subtraction problems. A
concrete strategy is separating from; making a set of
the larger numerosity, separating from it the smaller
number and counting the remainder. This is an
enactment of phrasings such as 5 take away 3,
which Conroy (1984) found to be the preferred and
easiest phrasing for Australian 6 to 13-year-olds.
There is a parallel counting strategy called Counting
Down From; the child counts backwards starting with
the larger number for as many number words as the
given smaller number. The last number in the
counting sequence is the correct answer.
Other strategies include starting with the smaller
number and counting up until the larger one is
reached: the number of objects added is the answer.
The most efficient of all strategies is to combine
counting down and counting up to involve the
minimum action: for example, counting down from
eight is the quickest way to solve 82, while counting
up from six is the quickest way to solve 8 6. As I said
when discussing addition, using number facts
supersedes these counting strategies to a
considerable extent.
The early language of arithmetic
We have seen that children as young as 5 know that
adding increases quantity and subtracting decreases it,
101
Figure 14 Response time depends on the number of increments required in the min model. Number pairs represent the two
numbers to be added; e.g. 13 means 1 + 3.
Source: From Groen and Parkman (1972).
are able to recite number names, and quickly learn

ways of quantifying addition and subtraction. They
have a good understanding of some of the language
expressions of addition and subtraction, mainly the
more concrete forms. For example, the 4-year-old
boy recorded by Martin Hughes does very well with
concrete questions but does not translate his
knowledge into an abstract number system at all:
Adult
Child
Adult
Child
Adult
Child
102
How many is two and one more?

Four
Well, how many is two lollypops and one
more?
Three
How many is two elephants and one
more?
Three
Adult
Child
Adult
Child
How many is two giraffes and one more?

Three
So how many is two and one more?
(looks adult straight in the eye) Six.
(Hughes 1983, p. 211)
One of the things this child has to learn is the formal

code of arithmetic, for example one and two used
alone rather than with accompanying nouns as in one
elephant and two elephants; and as a synonym for
plus; is (or makes) as synonymous for equals.
Corran and Walkerdine (1981) point out that children
rarely learn this formal code from adults before
school, and also that the different ways of saying a
string of mathematical symbols such as 5 + 2 = 7 are
not trivially different. Five and two makes seven for

example emphasizes the addition, the production of 7
from 5 and 2: Five plus two equals seven
emphasizes the equivalence of the two sides of the
equation. To take a second example, Coleman 1982
lists various translations of 8 5 = ? take 5 from
8, reduce 8 by 5, by how much greater is 8 than 5,
what is the difference between 8 and 5 and so on.
These seemed to reflect different ideas about what
subtraction meant and how to do it. Conroy 1984
looked at childrens success rate on different orally
presented subtraction sums; 6-year-olds got what is
2 less than 8 right only 21 per cent of the time (8 2
= ? was right 67 per cent of the time) and what is 7
take away 5 right 70 per cent of the time (7 5 = ?
was right 76 per cent of the time). How children
translate from symbols to metaphoric language and
vice versa can be seen to affect the solution
procedures they use, as I shall describe. At some time
the child has to reach a level of fluency and flexibility
in using these terms that allows recognition that they
are interchangeable verbal forms for the one
symbolic statement.
Hughes (1983) gave 5 to 7-year-olds the task of
showing on paper how many bricks there were on the
table. They found this easy, drawing the objects or an
equivalent number of tallies or writing the
appropriate conventional number. They were also
asked to show what happened when a few more
bricks were added or subtracted. Most just drew the
final number of bricks: there were a few attempts to
represent initial number, final number and what had
been added on or subtracted, for example by showing
hands adding bricks or putting bricks away in the box,
or the ingenious if slow strategy of drawing a line of
British soldiers marching from left to right to
represent added bricks, and a line of Japanese soldiers
marching from right to left to represent subtracted
ones. Not one of the seventy-two children tested used
the conventional operator signs of + and , although
they were using them regularly in their arithmetic
lessons, and, if they were comparable with Conroys
children, probably using them correctly most of the
time.
Resnick (1982, 1983), examines subtracting

algorithms and bugs, the errors which arise from
using algorithms in ways which violate some of their
rules. Many of the bugs she describes resulted from
failure to understand the meaning of subtraction, and
particularly of how to manage subtraction problems
such as 61 37 which require carrying between tens
and units columns. As such understandings became
more secure and childrens algorithms became more
flexible, children could combine different operations
to produce an answer and then check it, thus
debugging themselves. How their informal
selfgenerated strategies map on to the formal skills
taught in mathematics lessons needs further research.
So does the question of the extent, nature and causes
of individual differences in arithmetic skills. It seems
probable that a lack of matching between informal
and formal mathematics is one important component
of poor mathematical achievement. So are
metacognitive skills.
Some further developments in mathematics
I have been describing what children do at the

beginning of their work with numbers, and we have
been dealing with small numerical quantities, mostly
below ten. Dealing with larger numbers hundreds,
tens and units also has to be managed. This
involves consideration of the base ten number
system, computations using written versions of it,
and crucially, a recognition that numbers can be
interpreted as made up of other smaller numbers: 35
= 30 + 5 = 20 + 15 and so forth. There has been quite
a lot of research on how children learn to manage
more advanced forms of addition and subtraction
(e.g. Resnick 1982, 1983) partly because difficulties
are frequently found. Lesh and Landau (1983) also
provide material on childrens handling of
proportions, fractions, geometrical concepts and
algebraic problems. As was the case with simple
addition and subtraction, children need to understand
both the reality context of the calculation, its
mathematic structure and the appropriate algorithms,
and to relate all these, if they are to work successfully.
A great deal more research ideally involving
collaboration between mathematicians, psychologists
and educators, is needed.
103
Metacognition
Metacognition refers to ones knowledge and
control of the domain cognition (Brown et al. 1983,
p. 106). It thus involves a range of different
psychological contents. Among them would be
awareness of cognition, of understanding or not
understanding, remembering or forgetting, at a
particular moment; knowledge about ones own
cognitive skills and about a tasks difficulties (I
wont be able to do this because I cant remember how
to do square roots); and regulation of cognition, for
example planning how to do a task, monitoring
progress on it, checking the solution. Some of this
metacognitive knowledge is stable and stateable,
though it may not be objectively true. For example, I
would know all the time that I remember epigrams
rather badly and the general meaning of a play or
novel quite well, and that in order to record the detail
I should make a note of it. Some metacognition is
much harder to describe and much harder to employ
stably and systematically; as the introspectionist
psychologists found long ago, people who reported
on their thinking while they were working on a
problem found their thinking disrupted (Valentine
1982, Russell 1984).
Recently, metacognition has taken a prominent
place in discussion of cognition and cognitive
development. This is partly because of the influence
of the information-processing models of cognition
which I briefly reviewed in the previous chapter (see
also Klahr and Wallace 1976; Siegler 1981; Sternberg
1984; Case 1985). The concept of a central executive
system which controls cognitive processes has been a
powerful metaphor for developmentalists, who see
cognition as becoming increasingly skilled and
controlled as development and education proceed.
Another reason for the rise of metacognition has been
Piagets emphasis that it is the childs awareness of a
problem in his or her thinking which is crucial for
cognitive development: a contradiction between two
of my schemes of thought forces me to develop a
resolution between them, or conflict between my
104
thought and yours makes each of us review thinking

we were previously complacent about.
A number of questions have been asked, then,
about the development of metacognition. I will
outline some of the relevant areas of research, and
then briefly address the implications of this work for
theories of cognition and for educational
intervention. The sections of this chapter which
discuss remembering and comprehension also
contain relevant material.
Sources of metacognitive experiences
When do children know that they dont know? What

effect does such an experience have? Piaget believed
that experiences of knowing that you had
contradictory ideas or that your ideas contradicted
someone elses were common even in young children
and served as one of the important pressures towards
cognitive development. Autonomous regulation of
thought, automatic small adjustments of actions,
were inherent in cognitive activity: conscious
reflection on thought was a late achievement.
Awareness of the intended outcome and
discrepancies from it comes much earlier than
awareness of what exactly is wrong. Thus children
might be expected to know that there was a
contradiction without knowing what to do about it
and without being able to take appropriate action
(Piaget 1976, 1978). It can be very difficult to make
this progress.
Experiments by Karmiloff-Smith and Inhelder
(1974/5) illustrate one way in which the development
is made. They gave 4 to 9-year-olds the task of
balancing blocks of wood on a narrow bar. The
simplest strategy was to use physical trial and error,
nudging the blocks on to the balancing point: slow but
successful. This strategy was, however, supplanted
by simple theories which worked for some blocks but
not for others, for example to balance the block by the
geometric centre, which was successful for uniform
blocks but not for ones with an uneven distribution of
weight. These latter gave rise to an alternative theory,
so the child operated two juxtaposed theories without
co-ordinating them. Constructing a higher level
procedure, which combined the simpler theories and

left no exceptions, was slow and painful in that it gave
rise to more errors as it was developed than its
predecessors had done. Simple theories applied
perfectly to a limited set of blocks: those blocks the
theory couldnt cover were rejected as impossible to
balance, until there were an intolerably large number
of exceptions. Philosophers of science have observed
a similar reluctance in professional scientists to take
on board the accumulating instances which the
present theory cannot incorporate (Kuhn 1962).
Social conflict as a source of metacognitive
development
Another source of metacognitive conflict which

Piaget emphasized is disagreement with peers.
Experiments by Doise and his colleagues deal with
this area (Doise and Mugny 1984). They have used
various Piagetian paradigms. Pairs of children are
required to reach agreement on a task organized so
that they are likely to begin by giving contradictory
answers. For example, in a conservation of length
task, each child might judge the stick that came nearer
to himself or herself to be longer than the one nearer
the other child: thus their two judgements are
contradictory, since it cannot be the case that both
sticks are longer. (In fact, of course, in the standard
conservation test both children are wrong as the sticks
are identical in length.)
Children who have experience of this sort of
conflicting judgement often do better when working
with the other child than they did on the pre-test, and
the improvement may carry over to the post-test when
they are alone again. The child is more likely to learn
from conflict with peers if he or she already has some
grasp of the principles involved in the task, and if
there has been real involvement in the situation rather
than a desultory partial attention. It is not necessary
for one child to give the correct answer: evident
conflict between two wrong answers also leads
children to realize that they need to rethink their
answer. Research by Emler and Valiant (1982)
suggests that intra-individual conflict, confronting
the child with a previous contradictory answer, can

also induce metacognitive progress.
However, it has to be stressed that awareness of
conflict is not quite the same as awareness of
contradiction, and knowing how to resolve the
disagreement is something else again. There is some
evidence that young children treat judgements of size
as being much like judgements of preference: You
think this ones more, I think that ones more is
regarded as the same sort of situation as You like this
one, I like that one (Russell 1981a, 1981b, 1982).
Children who have some understanding of the
physical principles of Piagetian tasks such as
conservation are less likely to make this sort of
mistake.
Beyond awareness to diagnosing the problem
One source of evidence on childrens ability to

diagnose the problem which is giving them a feeling
of not having understood is the work on
comprehension monitoring referred to on p. 83.
Children may not be terribly good at reporting text
comprehension problems but they do show some
behavioural indices of comprehension monitoring,
such as pausing and frowning (Brown et al. 1983, pp.
11416). Another body of information comes from
studies of verbal communication. Robinson (1983)
reviews work in this area, much of it her own.
Typically children are given an ambiguous message,
for example one where the main referent is not
specified sufficiently so that more than one object
might be referred to, although only one particular one
is meant. Thus the child might be told to pick the man
with the red hat when more than one red-hatted man
was available, or told to use a big brick next to
construct a building when there are several big
bricks of different colours and shapes. Young
children tend to act on such messages with great
confidence that they have understood what was
meant: when things go wrong they are likely to say it
was the listeners fault for not listening properly.
Older children, like adults, blame the speaker for
giving an imprecise message, and say that more exact
information should have been given.
105
It seems likely that this can be related to the ecology

of the small childs life. Examination of the tapes
from the Bristol Language Development Study
suggested that when adults met an ambiguous
utterance from a pre-school child they tended to guess
what the child probably meant. Thus if the child said
wanna drink the adults response was do you want
some orange? or orange or Ribena?; thus the adult
took responsibility for constructing a clear message.
A few adults said things like I dont know what you
mean. Tell me what you want to drink: all the
children who were relatively precocious speakerblamers had had this sort of treatment from their
mothers (Robinson and Robinson 1981). Many
children, however, must be told you havent
listened, now listen carefully, and so forth.
One of the important points that arises from
considering how ambiguity and confusion arise in
real life (by which I mean outside tests of cognitive
development) is that they are frequent and that not all
confusions are important enough to need noticing, let
alone resolving. Often someone else will cope, as in
the example where the adult listener takes
responsibility for clarifying the childs ambiguous
message. Or one may understand a concept very
adequately for day-to-day purposes but not be
familiar with all its more technical manifestations. Or
ambiguity and contradiction may be deliberately
employed as a literary or rhetorical device. At least in
this last case, the point may be precisely in the
pointing of levels of contradictory meaning. What
conflicts are diagnosed as, and how they are
resolved, is an important question for research in the
development of childrens thinking (Rotman 1977).
Knowing what to do about a cognitive problem
There is an extensive literature now on childrens

knowledge about how to tackle cognitive problems:
among useful reviews are Brown et al. 1983,
Cavanaugh and Perlmutter 1982, Flavell and
Wellman 1977, and Robinson 1983. There is also
relevant discussion on memory, text construction and
comprehension and awareness of language
elsewhere in this chapter. One basic finding is that
106
older children know more about their cognition than

younger ones do. There is disagreement about the age
at which children begin to make deliberate efforts to
remember, notice text contradictions, plan their
stories and so forth, partly because it is not always
clear what are reliable indices of such behaviour, and
partly because the early stages of strategies or skills
typically involve fragmentary and spasmodic use.
Thus, as in the Piagetian literature, when evidence
that children cannot do X until they are M + 2 years
old is followed by evidence that on the contrary
children can do x when they are only M years old, we
need to examine very carefully the relationship
between X and x. Babies may distinguish between
two beats at a time and three, but is that the same as
the adults recognition of two and three? (Gelman
1982).
A more interesting question than the age one is how
to describe the differences between the young child
who does not use a metacognitive strategy and the
older child who does. One possible description is the
novice-expert dimension (Brown and de Loache
1978; Shatz 1978). Novices lack the necessary skills
for a task and probably also an adequate feel for the
task as a whole. As managing the component parts of
the task takes up all the available processing space in
the early stages, there is little opportunity for
metacognitive activity such as deliberate selfregulation. As the task parts become more familiar
and take their places in the whole task, there will be
more space for metacognition, and the learner can
step back, consider the way the entire problem is
going and make his or her performance more
systematic and better organized. The children I
described doing the balance task set them by
Karmiloff-Smith and Inhelder (1974/5) seem to
follow this pattern. Children will constantly be
novices on tasks where adults have become experts
with many more well-learned routines. Task
familiarity will be a crucial variable (Shatz 1978).
It seems likely however that adults will not just
have learned many routines well enough to find it
easy to perform the routines and to think about them;
they may also have developed more generalized
metacognitive skills which they can transfer flexibly
from task to task. Thus although the specifics of a task

may be equally new to child and adult, the latter may
both have a better sense of what sort of task this is,
more knowledge of his or her own capabilities and
more general routines for controlling his or her
performance. (On the other hand, he or she may have
acquired more emotional blocks to learning you
cant teach an old dog new tricks, Im too old to
handle micro-computers.) Child novice and adult
novice cannot be assumed to be the same.
Making metacognitive progress
One difficulty in giving an account of how and why

childrens metacognition develops is that much of the
research is cross-sectional, establishing that children
of, say, 5 do not have a particular metacognitive skill
and children of, say, 9, do. This approach tells us
nothing directly about the mechanisms of the change.
Training studies, however, provide one way in to
cognitive change, particularly if we compare
successful training strategies with the events that
occur in childrens normal upbringing. Brown et al.
(1983, pp. 129 46) review studies which have
attempted to train learning skills.
Inducing subjects to use a strategy without any
explanation of why the strategy was important or an
explicit noting that it was effective, generally led to
improved performance on the task while the strategy
was being used, but little transfer to other tasks, and
the strategy might be given up even for the target task
once the experimenter stopped reminding the subject
to use it. Experiments by Whittaker (1983) illustrate
this. He set up paradigms for memory tasks which
forced the child to recognize the difference between
performance using the strategy and performance
without it: this greatly increased childrens
persistence in using the strategy.
Providing subjects with information about the
importance and effectiveness of the activity that
theyre being trained to use greatly enhances the
prospect of them continuing to use it and transferring
it to other tasks. This sort of information seems to be
particularly necessary for retarded children who
rarely show spontaneous transfer of learning.
Convincing information about why it is worth using

a particular strategy presumably increases peoples
motivation to use that strategy. Metacognitive
activity requires some effort on the part of the thinker,
and if there are limitations on thinking space (see
Chapter 2) it may seem to be too much bother or not
worth the effort. Demonstration that it does enhance
performance and produce rewards greater than its
initial cost may make it more likely to be used.
Children are, in practice, rarely given explicit
metacognitive training of the sort used in the
intervention studies reviewed by Brown et al. (1983).
However they do have some opportunities to see
older people making special efforts to remember,
plan, review their learning and so forth. Such
activities are probably more common at school than
at home but do occur there too, particularly perhaps
in conversations between the child and older people
(Tizard and Hughes 1984). It seems likely that
parents undertake some of the responsibility for their
childs metacognition. One example is the joint
activity of looking at books, where mothers often say
to their 2-year-old things like I know you know that
one, Well find you something you know very well
(Ninio and Bruner 1978). Mothers are often used as a
memory by their children (e.g. Kail 1979), as well as
modelling external memory devices such as the use of
shopping lists, which children understand well
(Istomina 1975). Nursery rhymes and stories also
provide obvious opportunities for metacognitive
activities. The adult gradually hands over to the child
more and more responsibility for the success of the
activity. There is some reason to believe that children
whose parents go in for quite a lot of explicit
modelling of metacognitive and metalinguistic
behaviour become advanced in their learning of these
skills (e.g. Robinson and Robinson in press; Mills
and Funnell 1983; Brown et al. 1983). We do not at
present know exactly what effects parents
metacognitive behaviour has, or how those effects
come about, and we certainly cannot conclude that
the more metacognitive modelling there is the better.
I do, however, propose that we should take very
seriously the hypothesis that one important aspect of
metacognitive development is a progress from
107
cognition supported by, and regulated by, other

people more skilled than oneself, to cognition which
is relatively independent and self-regulated. To quote
Brown et al. (1983, p. 124)
mature thinkers are those who provide conflict trials for
themselves, practice thought experiments, question their
own basic assumptions, provide counter examples to their
own rules, and so on. Although a great deal of thinking and
learning may remain a social activity, mature reasoners
become capable of providing the supportive-other role for
themselves. . . .
Social cognition
Most of the work on cognitive development deriving
from the Piagetian or information-processing
tradition centres on an individual thinker trying to
understand the objective physical world. Lately, an
interest has grown in an area which poses alternatives
to each half of this individualist model, and centres on
the thinking individual whose thinking comes from
interaction with other individuals as much as from
within himself or herself, or on individuals thinking
about the subjective social and interpersonal world. It
is this newly important area which is called social
cognition. It is a diverse field, including, for
example, the effects of social conflict on performance
of conservation tasks, childrens descriptions of
friends, the role of emotion in understanding other
people, childrens ideas about techniques for
achieving social goals such as joining groups, the use
of social scripts, and childrens theories of how
social institutions work. There is much debate about
conceptual and methodological issues, and not as yet
a unifying theory (indeed only a high level theory
could unify so wide an area): I will not attempt to do
more than discuss selected work in the field. For
reviews see Isbell and McKee 1980, Shantz 1983,
Flavell and Ross 1981, Butterworth and Light 1982,
and Forgas 1981.
Cognitive development as a social process
There is no room for doubt that childrens cognitive

development takes place within a social world and is
108
influenced by other people. Children use other

peoples knowledge in their own development: the
ways that other people treat the child shape his or her
cognition not just about people but about many
aspects of the non-human world. Knowledge and
influence come not just from the micro-system, as I
shall describe, but from the wider social world.
Religious teaching, for example, was a significant
factor in the historical development of thinking about
the relationship between Man and Nature (Thomas
1984). The questions that need to be answered are not
about whether social interaction influences cognitive
development but when and how it does.
Piaget placed his main emphasis on the dialectic
between the child and the physical world, but
included social interaction as a motivator of
development, particularly through conflict of ideas
between peers. Discussion and criticism involving
peers forced the child to decentre, to resolve the
contradictions between different viewpoints or
opinions. Recently Doise and his colleagues in
Geneva have carried out a series of studies on the
effects of peer interaction on childrens performance
of Piagetian tasks (Doise and Mugny 1984). They
paired children on conservation and perspective
taking tasks set up in such a way that the children
would disagree on the answer. This conflict led to
more advanced judgements on subsequent tasks,
provided that the children had the beginnings of a
grasp of the principles underlying the correct answer.
Children with no understanding at all of conservation
or whatever were likely to treat the disagreement with
the peer as a matter of disagreement over some
subjective preference (Russell 1981, 1982; Light
1983): that is, there was an interaction between
metacognition and ability to learn from social
interaction.
There has been some debate about whether interindividual conflict is a better motivator of cognitive
development than either intra-individual conflict
(children seeing that their present judgement is
incompatible with their previous one) or interindividual co-operation. Self-contradiction has
seemed to be as effective in inducing cognitive
development as other-contradiction (Emler and
Valiant 1982), though Doise and Mackie 1981 argue

that such situations are in fact socially produced (by
the experimenter or another adult setting up the
child) or that the child may have decentred enough
to treat his or her previous and present views as
objective and thus contrastable. In Chapter 2 I
touched on the difficulties of knowing how to resolve
a conflict and pointed out that recognizing its
existence does not in fact solve the problem. Cooperation with another person, on the other hand, may
both provide new information and confirmation of
the participants ideas where they agree (Bryant
1982). This sort of behaviour by adults does seem to
advance cognitive development (Vygotsky 1978;
Bruner 1968; Mills and Funnell 1983; Wood 1980;
and see Chapters 2, 5 and 6). While children are
developing cognitive skills and models within an
area, commentary by other more expert people which
helps to integrate the developing thinking into a
coherent whole may serve to support the novices
understanding. Such behaviour seems to lead to the
optimum cognitive development if it is contingent on
the childs behaviour and interest (Mills and Funnell
1983), and if it is a reflection of the experts expertise
rather than his or her misunderstandings!
Childrens understanding of the properties of
persons and objects
Social cognition implies a distinction between

social and non-social, between persons and other
objects. Exactly what is the basis for such a
distinction is a matter of some debate, particularly in
contentious areas such as whether non-human
animals or complex computers are persons
(Midgley 1979; Searle 1984). The central criteria
seem to be:
persons are agents, that is they are capable of initiating
actions, while objects can only move if something or
someone else initiates the action
persons know, think, learn and have emotions, while
objects do not
persons are alive, that is they develop and reproduce
themselves
persons typically act and react in a wider variety of ways

than objects do, thus they may be less predictable.
Some of Piagets early work (Piaget 1929, 1930)

showed children attributing animate characteristics,
such as independent movement, to inanimate objects,
such as bicycles or the moon. This he called
animistic thinking, and it was supposed to be a
pervasive feature of pre-operational childrens
thought. More recently, a number of studies have
failed to find much animistic thinking even in 3-yearolds. Shields and Duveen 1982, for example, asked
nursery-school children which of a farmer, a cow, a
tractor and a tree could eat, sleep, move by itself, talk,
feel angry, and so forth. The childrens answers drew
a clear distinction between the tractor and the tree,
which could not do any of these things, and the farmer
and the cow which could. (There was some
disagreement over whether cows could talk or have
emotions, but children who claimed they could,
maintained that cows talked with other cows or the
farmer, talk being extended slightly to mean
communicate.) Similarly, studies reviewed by
Gelman and Spelke (1981, pp. 48 51) suggest that
children as young as 2 may make some adult-type
distinctions between animate and inanimate objects
if the objects are familiar to them and if the questions
asked are fairly straightforward. Some of Piagets
animistic answers may have stemmed, it is argued,
from asking problematic questions like Does the sun
know where it is moving? Explaining difficult
phenomena, however, may draw out animistic
reasons: some of Shields and Duveens subjects, like
some of Piagets, said that the wind was caused by the
voluntary movement of the trees.
A distinction between animate and inanimate
objects, and between people and other animate
objects, seems to be discernible in quite young
children. It is not entirely clear what the distinction is
based on and we have only speculation at present as
to how it is built up in the first three years of life. One
probable source is no doubt differences in the ways
persons act on the child and react to his or her actions
compared with the actions and reactions of objects.
Careful analysis of the lives of babies and toddlers
would perhaps inform us about this.
109

Childrens understanding of people
Research on the development of childrens

understanding of people is one of the largest bodies of
work in the social cognition area. It has been heavily
influenced by Piagets theory that children are
egocentric, that is that they are unable to appreciate
that other people have knowledge, feelings or views
different from their own. The notion of
egocentricity no longer seems as credible as it once
did in the light of evidence that even very young
children can and often do show a practical
understanding of other peoples feelings and
knowledge (see p. 157). There is also a strong
probability that children operate an intricate social
world with, on the whole, about as much success as
adults, which again suggests they have a fair
understanding of their familiar peers and adults (see
p. 192). Nevertheless, it is possible that the
understanding which manifests itself in successful
interpersonal action does develop. What evidence is
there?
Shantz (1983) and Rogers (1978) review studies
where children were asked to describe or judge
peoples behaviour. One of the best known of these is
work by Livesley and Bromley (1973). They elicited
free descriptions from each of their 7 to 15-year-old
subjects of eight people personally known to the
subject. They found that the younger childrens
descriptions were predominantly in peripheral and
external terms, such as appearance, age, and
surroundings. As children got older, their
descriptions become more abstract and involved
more inferences about stable psychological traits,
initially rather global ones such as nice or good,
but gradually with more qualifications and balances.
This sort of picture, initial descriptions which are
mostly concrete and based on behaviour and later
descriptions which are more abstract and in terms of
psychological dispositions, emerges from most
studies which have used this sort of approach (Shantz
1983; Rogers 1978).
They have been taken as indicating that young
children have a poor understanding of other people
and older children a better one. Abstract decriptions
based on generalized dispositions may allow more
110
general predictions than specific concrete

descriptions do: if this is the case, and it is still a
debatable issue in personality theory (see, for
example, Cook 1984), the development of person
perception may be from poor to better. However, it is
not really very clear how well childrens free
descriptions represent their ideas about people.
Describing people is a different activity from
engaging in social interaction with them, and may not
use the same information. There are also likely to be
problems of interpreting what is meant, particularly
with children whose vocabulary is limited and who
may not mean by a particular term what an adult
would mean by it. Research which uses more probing
and more rating scales may be needed, even if such
methods do restrict subjects. A study by Flapan,
which Rogers (1978) describes, apparently found
responses to questions were more sophisticated than
free descriptions.
Childrens attributional judgements
Young children do seem to be rather less likely than

older ones to infer inner psychological dispositions
from outward behaviour: there also seem to be some
differences in what evidence they use when making
attributional judgements about the causes of an event.
Attribution theory (Heider 1958; Kelley 1979; Ross
1981; see also Chapters 5 and 6) distinguishes
between personal causal factors, such as a persons
timidity, and impersonal causal factors, such as the
size and muscularity of the threatening opponent,
which come into consideration when decisions have
to be made about questions such as whether it should
be fight or flight. It also uses the co-variation
principle. If an act is done in that situation by most
people most times, but not in other situations, then the
situation, an impersonal causal factor, is more
important than the person, whereas if that person acts
that way in most situations most times then the cause
lies in the person. Attributions like these are
potentially useful in social interaction. They do need,
however, to be based on representative evidence,
preferably a great deal of it; to be available for use and
revision; and to be unbiased (most adults under-
estimate situational factors relative to personal ones

when they are spectators (Ross 1981), though not so
much when they are themselves actors).
It is probably premature to draw conclusions about
what evidence children use and what judgements
they come to in considering causal attributions.
However, the work summarized by Rogers 1978,
Ross 1981, Shantz 1983 and others, suggests that they
do seem to make inferential use of both personal and
impersonal information using the co-variation
principle, though young children emphasize the
situation at the expense of the person. They do
integrate information, though young children may
not manage to handle so many separate pieces of
information and do much better when the person they
are talking about is familiar to them or like
themselves. They also live in social worlds which
may be significantly unlike adults social worlds,
though what the likenesses and unlikenesses are has
not been adequately conceptualized. This too might
account for some of the apparent inconsistencies and
inadequacies of childrens judgements about people.
Understanding social events and institutions
The range of social events and institutions that

children encounter is of course enormous,
particularly if vicarious encounters through reading,
television, etc., are included. I propose only to discuss
one potential general model, the script, and one
social area, aspects of the economic world, here.
Research on childrens knowledge of the world of
school is discussed on pp. 198200.
Scripts as representations of social knowledge
The notion of scripts derives mainly from the work

of Schank and Abelson 1977, which was concerned
with describing a model for computers
understanding of inferences in a story setting. A
script is basically an ordered sequence of actions,
appropriate to their context and organized round a
goal, a generalized set of expectations about who is
likely to do what during this sort of event. There are
obligatory actions and optional ones, points in the
script where one particular thing has to be done and
other points where there are alternative possible

actions, essential rules and props whose omission
would be very surprising indeed, and others which
are less predictable. The best known prototype script
is the restaurant script, which would look something
like this:
Roles Customer, waiter or waitress (chef offstage, possible
cashier)
Customers goal to obtain food to eat (maybe other

goals such as being sociable)
1 go to restaurant, enter it, move to empty table,
sit down (customer or waiter/waitress may choose
table)
2 receive menu, study menu, decide what to eat,
give order to waiter/waitress
3 receive food, eat it (possible repeat of 2 and 3
for later courses)
4 ask for bill, receive it, leave tip for waiter/
waitress, pay bill (pay waiter/waitress or cashier:
order of paying bill and leaving tip may be
reversed), leave restaurant.
This sort of organization seems to be convenient
for adults: elicitation of scripts from children
suggests that they too use an organization based on a
sequence of actions within a particular context
(Nelson 1981). What is particularly interesting about
these scripts is that they are general, not accounts of
one particular episode. For example, a distinct
language style is usually used, the impersonal
pronoun and timeless present tense (such as you go
to the restaurant, you go in and sit down, etc.). The
events included (in temporal order) are more likely to
be the routine components than the exceptional ones,
though some components may be so routine and
inevitable that they are not mentioned. Scripts look
like generalized knowledge about a social routine,
organized by experience and elaborated by further
exposure: they seem likely to scaffold understanding
of past and present events and prediction of future
ones.
111
We do not know a great deal about how scripts are

acquired. It seems likely that they are picked up from
routines and contexts which adults structure for the
child (see Chapter 6). Children take part in many
activities which adults direct and organize. Children
have to play their part more or less as determined by
the adult partner, who may even supply the lines
Say thank you to Granny for the nice present, In
this school we say Good morning Miss Church,
Please Miss Church may I . . . not just Miss, Miss;
its politer. Scripts provide a context of general
expectations which reduce the uncertainty in
particular problems. It seems quite a strong
possibility that children show better understanding
and more social sensitivity if they have shared script
routines with familiar adults and if there have been
explanations for particular events in generalized
terms (Light 1978; Tizard et al. 1982; Mills and
Funnell 1983). It is possible that the scripts
themselves might be better articulated, though as yet
there is no evidence for this.
Having a script for an event or context means that
taking part in it can be automatic to a considerable
extent. Attention and cognitive processing space is
thus freed for other things: the idiosyncrasies of this
particular instance of a general type of event, for
example, or even remembering or planning events
quite separate from the present very routine one. Not
having a script means that the activity is far more
problematic: we are less able to predict what will
happen or to interpret what already has happened.
Children, being inexperienced and ignorant
compared with adults, are more frequently in this
position. They may therefore be preoccupied with
building scripts, and may focus on details which are
incidental rather than typical; their scripts may not
show a good match either to the situation or to other
peoples scripts for it. Nelson and Gruendel (1979 pp.
801) provide an example of two 4-year-olds
comparing scripts and showing great interest in
achieving agreement on their essential features.
G1 And also, at night time, its supper time.
G2 Yeah, at night time its supper time. It is.
112
G1 Its morning.
G2 At morning, its lunch time!
G1 At morning, we already had breakfast. Because
at morning, its lunch time!
G2 RIGHT!
G1 Yeah, at morning, its lunch time.
G2 At morning its lunch time.
G1 But, first comes snack, then comes lunch.
G2 Right . . . Just in school, right?
G1 Yeah, right, just in school.
G2 Not at home.
G1 Well, sometimes we have snacks at home.
G2 Sometimes.
G1 Sometimes I have a snack at home.
G2 Sometimes I have a snack at my home, too.
G1 Uh-hum. Because when special children come to
visit us, we sometimes have snack. Like, like,
hotdogs, or crackers, or cookies or, something
like that.
G2 Yeah, something. Maybe cake. (Laughs)
G1 Cake.
G2 Cake. Yeah, maybe cake.
G1 Or maybe, uh, maybe, hotdog.
G2 Maybe hotdog.
G1 But, but, but, Jill and Michael dont like hotdog.
Dont you know, but, do you know Michael or
Jill?
G2 I know another Michael.
G1 I know, I know another Michael.
G2 No, I know just one Michael. I just know one
Michael.
G1 Do you know Flora?
G2 No! But you know what? Its a, its its one, its
somebodys bro . . . its somebodys brother.
G1 Are you eating your dinner? (Laughs) But not for
real.
G2 Not for real.
G1 Because at morning, its lunch time.
G2 Right, at morning it is lunch time.
G1 Right, at morning it is lunch time.
G2 Yeah.
G1 I think . . . Ill have . . . lunch. [Nelson and
Gruendel, 1979, pp. 8081]
Scripts could apply to a wide variety of events

and understandings, and the notion may prove
fruitful in analysing many phenomena in childrens
development. As well as the social settings discussed
in this section, it seems to be applicable to childrens
construction and reading of stories (p. 83), to their
language development (p. 118) and to their beliefs
about their self-efficacy (p. 150). It might even be
script violation that is the root of many pre-school
emotional upsets: if the getting dressed script
specifies that the left foot should be shod before the
right foot, or that teeth are brushed before hair, the
adult who does otherwise risks protests and tears. Not
having events go according to the script is a
disorientating and unpleasant experience.
Childrens understanding of socio-economic
systems
Childrens development, including their social

cognition, has mainly been studied in terms of
microsystems such as home or school, or
mesosystems such as the relationships of agreement
or discrepancy between home and school
(Bronfenbrenner 1979; see Chapter 1 this volume).
Chil dren also live however within wider worlds:
their part in these worlds is less prominent (people
concerned with childrens rights (e.g. Leach 1979)
argue that it needs more recognition), but it is none the
less of potential interest to developmental
psychologists. There is some work on childrens
understanding of social institutions such as shops,
banks, and governments, mainly conceived of as
exosystems in which other people play roles that the
child can observe, rather than as macrosystems or
ideologies, which are more abstract and inferred
rather than observed.
Studies of childrens understanding of social
institutions tend to involve questioning children,
perhaps necessarily, since they do not normally vote,
have bank accounts, or play many of the roles of
economic or political persons directly rather than
through or on behalf of adults. Their answers to
questions such as What things, jobs, people, are
important in a town? (Furth 1980) are often full of
charm but hard to analyse. Furth, like several other
investigators, has used neo-Piagetian stage analyses

although data about uniformity of stage across
contexts or the separation of stages are not usually
impressive. More precise questions, script
elicitations, and perhaps role play, may provide
complementary information.
Understanding shopping
The central transaction in a shop might appear to be a

simple exchange between customer and shopkeeper
of goods and their value in cash. The shopkeeper has,
and the customer wants, a bar of chocolate costing
15p: the customer gives the shopkeeper 15p and
receives the chocolate in return. This is certainly the
most visible part of shopping but it is far from the
whole. The retail shopkeeper is a customer to the
wholesalers who supply the shop with goods: the
shop has assistants who exchange their labour for
wages, and also various running costs such as rent,
rates, lighting, etc. These components of the shop are
less salient to the shopper, but they have significant
effects on the cost of the items in the shop. Children
seem to have some difficulty, as we shall see, in
grasping the existence of these components and their
effects, let alone their scale. A further source of
confusion is that the passage of money between
customer and shopkeeper may be two-way: the
customer gives 20p and receives the chocolate bar
and 5p change. To a small child with little grasp of the
monetary values involved, this may look like a ritual,
or a profit to the customer who ends up with goods
and money, rather than the exchange it is.
Jahoda (1979, 1984) studied Scottish and Dutch
children. The youngest, aged about 6, seemed to see
the transactions as rituals; goods were given to the
shop, not bought by it, and shop assistants were not
doing a job and so were not paid. Even when these last
two misconceptions were beginning to weaken, at
around 7 to 8, the idea that the price the shop charges
the customer is the same as the price the wholesaler
charged the shop remains strong. The shopkeeper
merely passed on goods at the same price as their
original cost: the overheads of the shop, if recognized
at all, were met from some external source such as the
Mint or the government. Children at about 11, with a
better understanding of money, jobs, and costs, had
113
begun to understand the concept of profit, though

the size of profit made, the extent of the mark-up
between wholesale and retail prices, probably wasnt
grasped it is often enough a surprise to adults!
Understanding banks
Jahoda (1984) also summarizes the results of a study

of childrens understanding of the functioning of
banks as economic institutions. He questioned
Scottish children aged between 11 and 16. The
youngest, who might have shown a little
understanding of profit in the context of shops,
showed no idea of it in banks. They viewed banks as
being much like money boxes and borrowing from
friends: you got out exactly what you put in, to get
more or to pay for a loan would not be fair. Even if
they knew about interest they had little idea where it
came from, and still thought of the bank in
interpersonal terms.
As children moved through early adolescence,
they showed more appreciation of interest and the
banks use of the money deposited in it. Only a
minority, however, mastered the reciprocity of
interest in the banks lending and borrowing, even
after questioning designed to induce new insights.
Again, appreciation of what scale banks profits are
on, and what is done with them, is probably unusual
in adults.
Representing economic inequalities
Emler and Dickinson 1985 have studied childrens

representation of economic inequalities. They
asked children aged between 7 and 12 to estimate
the weekly incomes of doctors, teachers, bus
drivers and road sweepers, to explain why the
incomes differed and to say whether equality of
income would be better. The average estimates of
middle-class children for each job were higher than
those of working-class children: doctors were seen
as having the highest incomes followed by teachers,
bus drivers and finally road sweepers. As Figure 15
and Table 4 show, middle-class children gave much
more differential in their estimates of income than
114
working-class children dic, as well as judging

incomes to be higher overall.
The children were asked about the fairness of the
different amounts they had estimated for each job and
of a hypothetical equality of income. There were no
significant differences in judgement of fairness by
age or by social class, though it was less common for
the oldest children or for middle-class ones to say that
equality of income would be better. Most of the
children justified inequality of income by reference to
some form of equity consideration, such as
differences in the work or the responsibility involved
in the job or the amount of training required for it.
Middle-class children were more sophisticated in
their justification of inequality, producing more (and
more varied) reasons.
Thus, although most children believed that
income inequalities were justified, middle-class
children estimated that the inequalities were larger
(though even they probably underestimated their
scale). The absence of age differ ences argues

Table 4 Ratios of income estimates for doctor and road sweeper by age and social class
Age level
78
WC
MC
89
910
1011
1112
SD
SD
SD
SD
SD
1.31
3.32
0.66
3.25
1.91
2.94
1.34
2.66
2.63
4.79
0.67
5.35
1.59
2.88
0.72
1.22
1.61
2.64
0.52
1.42
Source: From Emler and Dickinson (1985), p. 194.
against this class difference being caused by a

developmental increase in the complexity of
understanding of social systems, faster in the middleclass children than in the lower-class. The different
estimate levels seem more likely to be due to different
information about incomes. Emler and Dickinson
argue that social representations of economic
inequalities are more detailed, extensive and salient
in the middle class, and hence middle-class children
have had more opportunity to assimilate their
communitys shared knowledge and belief about
income inequality. Exactly where in the community
economic knowledge comes from remains to be
investigated.
Sources of socio-econornic understanding
The three studies I have discussed in this section

illustrate some of the recent work on the development
of childrens understanding of the socio-economic

system. They use information derived from their
participation in their community to construct
representations and justifications of the workings of
the economy. Limitations on this information which
are caused by their own status as observers, not
participants, or by the communitys particular beliefs
or values, restrict their models of economic life to
simple versions. Although internal cognitive
limitations may also restrict their conceptualization,
sheer ignorance seems a likely cause of inadequate
models. Children can learn to be sophisticated and
critical observers of at least such parts of the
economic system as advertising (Ward et al. 1977).
Further work on the social origins of childrens
economics could produce findings which will be
relevant also to the wider issue of social
representation.
Plate 9 Pas Bank from The Book of Shops, 1899, by E. V. Lucas, illustrated by Francis D. Bedford
115
Plate 10
116
4 Language development
Language
The development of language provides a good
example of epigenesis, and its later stages also
involve consideration of eco-systems. Human infants
clearly start with a great many capacities and pieces
of behaviour suited to language, but they are also born
into communities which use language and expect the
infant to use it too. All except the severely impaired
develop language in very similar ways, though at
varying rates: but the details of the language and how
it is used are heavily influenced by the childs
experience. Through development, language
functions as a means of communication, as a means
of reflecting on and re-organizing experience, and as
a way to receive and transform the accumulated
knowledge and values of the community. Using
language is thus a central part of human existence.
Methodology
Studies of young childrens language have relied

heavily on naturalistic observation; that is, on
recordings of childrens spontaneous language in
their ordinary activities at home. One reason for this
was an admirable concern for ecological validity, a
belief that what children could be induced to do in
experimental settings was not representative of what
they did outside them. The tendency of children aged
between 18 months and 3 years old to prove
extremely difficult subjects for standardized
experimental methods no doubt reinforced
researchers concern for ecological validity. Reliance
on naturalistic observation, however, meant, here as
in other areas of research, that very large amounts of
data had to be collected in order to find sufficient
instances of interesting phenomena which naturally
occur only infrequently. This meant researchers had
to have access to the child for long periods of time:

thus samples tended to be small in number and to have
parents who were unusually interested in
developmental
psychology,
many
being
psycholinguists themselves. The use of small
samples biased towards the intellectual middleclasses raises problems of the representativeness of
the findings of many studies.
There is need also to look very carefully at the
ecological validity of naturalistic observations. In the
first place, they involve the Observer Paradox as
Labov has put it, to obtain the data most important
for linguistic theory, we have to observe how people
speak when they are not being observed (Labov
1972, p. 113). Further, many studies do not specify
exactly what settings are observed. Language
produced at home playing with mother may differ
from language produced at home when playing with
a baby sibling (Dunn and Kendrick 1982), from
language produced when out shopping with mother,
from language produced at playschool with other
children, from language produced at playschool with
teacher. Systematic studies are needed of how
situational variables (including the presence of an
observer), affect childrens language production and
comprehension. Generalizing from the one situation
observed to all others, particularly if allied to a belief
that if children dont do something it is because they
cant, has led to some false conclusions and
misguided remediation programmes. One notable
instance of this is the debate over social class
differences in language which is discussed later in
this chapter.
At the root of the problem of what ecological
validity really is, there may be the difficulty of
defining what behaviour a particular situation
requires as well as describing what it actually evokes.
117
We do not have more than the very beginnings of

anything relevant to such definitions. Grice (1975;
see also Searle 1975) discusses the co-operative
principle which is necessary for effective
communication. Speakers normally try to ensure that
their contributions to discourse give information
which is neither insufficient nor superfluous in
quantity, which is known to be true rather than false
or merely supposed, which is relevant to the aims of
the conversation and which is not obscure,
ambiguous, wordy or disorderly. Contributions
which flout these requirements usually lead to
breakdowns of communication or to indirect
implications where what is actually said may be only
part of what is meant. Participants in conversations
tend to have an understanding of the implicit meaning
of utterances and events which is affected by who is
speaking, when, how, why and where. They may also
have scripts (e.g. Schank and Abelson 1977; Nelson
1981) which define what ought to happen in a
particular setting or encounter. We will look at what
is known about childrens use of indirect speech acts
and of scripts later. For the moment, my intention is
merely to point out that the significance of what is
observed in a natural setting cannot be fully
understood unless the script or the task demands
of the setting are known. The richness of naturalistic
data forces selection and categorization on the
observer, and unless this is done carefully, explicitly
and consistently error and ambiguity may result.
Corrigan (1982) suggests that naturalistic
observation has to be supplemented by careful
controlled experiment: a full picture of language
acquisition requires information about what children
choose to produce given less structured situations, as
well as the limits of what they are capable of
producing when required to do so (Corrigan 1982, p.
182). Gordon Wells, director of the Bristol-based
Language Development at Home and at School
project, the largest study to use naturalistic
observation, discusses the methodological problems
involved and discloses that with hindsight he would
have collected more experimentally-elicited
utterances (Well 1982). The Bristol study does
nevertheless have more and in some ways better data
118
than most others, and some of its findings are

prominent in the discussion of language development
that follows.
Language can be analysed at many different levels,
including the sounds made, the words used, the
sentences constructed, the meaning conveyed and the
functions served. Each level is, of course, involved in
the others most of the time, and children and adults
alike do not learn or use them in isolation.
Nevertheless the different levels have often been
investigated separately, and there are differences as
well as similarities in their developmental courses.
We have now good descriptions of much of the
sequence of childrens language developments, as I
shall outline, but there remains much to discover
about precisely how and why this development
happens.
Infants perception of speech sounds
At the level of language sounds, newborn infants

have been shown to have a preference for speech-like
sounds over musical or non-speech noises (see
Chapter 2; Eimas et al. 1971; Wolff 1966; for a review
see Aslin et al. 1983), and an ability to discriminate
between the voice of their mother and that of another
woman (DeCasper and Fifer 1980). From a few days
old, if they hear human speech their limb movements
will become synchronized with the rhythm of the
speech (Condon and Sander 1974). These are likely
to be useful characteristics to child and parent alike:
the babys selective attention and close response to
language will both increase its own opportunities for
learning from language and help to make the parent
believe that the baby is particularly interested in and
responsive to the language-producing person.
Interactional meshing, as we will see in discussing
the development of relationships with other people,
has to be got more or less right somehow, if
relationships are to go well. The capabilities of
newborn infants suggest that there may be some
genetic preprogramming, perhaps of brain structures,
which may possibly be very specific about speech
sounds. There certainly has been shown to be, in
infants of a few months, categorical perception of
some consonant sounds (e.g. Eimas 1971; Trehub

and Robinovitch 1972; Streeter 1976), at the same
points as adults place the discrimination. Adults
discriminate between /b/ and /p/, for example, in
terms of whether they are pronounced with vocalcord vibration (voicing); so do infants, including
those growing up in a community using a language
(such as the Kikuyu) where /b/ and /p/ are not
regarded as different in meaning. This suggests that
the auditory system is wired up very early on in
ways that mean there are points of particularly great
sensitivity on the physical characteristics of speech
sounds, and that languages place category limits
between different phonemes at these sensitive points.
However, if a language does not use a particular
discrimination its speakers lose it, and have great
difficulty in hearing (or producing) the appropriate
sound. The programming of babies here too is not
more specific than it needs to be. Ideas of a speciesspecific plan for language perception in humans
may have to be abandoned in the light of recent
evidence that chinchillas (and possibly monkeys)
categorize some speech contrasts as humans do.
Aslin et al. (1983) review the evidence.
Infants production of speech sounds
Just as it could be said that babies begin their

language career by hearing more sounds than adults,
it appears that when they start babbling they produce
sounds which the adults around them do not produce.
Babbling initially involves vowels; consonants start
to appear in a fairly orderly way (see Clark and Clark
1977, ch. 10) in the second half of the first year.
Babies have quite often been observed to do what
looks like deliberate experimenting with sounds,
contrasting the noises produced with different tongue
and lip positions and at times moving mouth and
tongue systematically but without making sounds.
As the child gets older, strings of babble with varying
intonation are produced; gradually the range of
sounds made comes to resemble the range present in
adults language; and particular sounds are regularly
produced in association with particular events. There
may be good reason (Halliday 1975) to regard these
sound-event pairs as the first words. Adults are

certainly likely to interpret them as meaningful, and
babble noises which appear often may become
important parts of adults baby-talk. The expected
words for young children to call their parents, for
example, tend in many languages to resemble babble
noises mama, papa, baba, and so forth. A baby
randomly producing such a noise may get a great deal
of reward from parents who believe or pretend to
believe that they are being addressed by name. Such
reward has been shown to increase not specific noises
in babble but on the one hand the amount of babble in
general and on the other the probability that what was
in its early stages something the baby did when alone
will be done as part of a social activity. Parents clearly
shape their babys early talk, requiring more and
more precise articulation of words and later
meanings, and also requiring the child to take part in
conversation in an increasingly self-directed way, as
we shall see.
The stable sequence of babbling and its appearance
in virtually all children in much the same way at about
the same time suggest that there is a high degree of
genetic programming and maturation involved.
Nevertheless experience is also necessary. Children
who are congenitally deaf start babbling like normal
children but give it up at about 8 or 9 months, almost
certainly because they have not been able to hear
themselves. If they are made able to hear at this point,
babbling is restored and the child learns a spoken
language (Fry 1966).
Maturation (in this case of articulation) is also
clearly involved in the timing of childrens first
conventional words, which commonly appear
between 10 and 15 months. Children have usually
showed understanding of words, and may have
produced their own idiosyncratic words, before then,
and deaf babies learning sign language produce their
first conventional signs for objects at about 8 months
(de Villiers 1979) because the muscular control
needed for manual signs is less fine than that needed
for speech. Spoken words need more or less mature
articulation, and the planning of the sequence of
speech sounds, not just the ability to associate words
with objects or events. The latter is also important,
119
however; children who are growing up in a bilingual

environment where the number of labels attached to
objects is greater, where objects have more variable
names, are commonly slower in their early
vocabulary development.
Early
words
often
involve
phonemic
simplification and the use of context to carry some of
the meaning: /da/, for example, may serve as a
simplified version of dog, duck and daddy and
be acceptably unambiguous most of the time because
only one of the possible referents is present or being
played with or pointed to. Later in phonemic
development, the child may begin to use systematic
rules and strategies to produce intelligible words.
Children commonly, for example, simplify
consonant clusters which are hard to articulate,
saying mack instead of smack. They replace two
different consonants in a word by a repetition of the
same one, as goggy for doggy. They may avoid
the use of words which contain a difficult sound,
something which gets easier to do as ones
vocabulary of synonyms increases. Certain sounds
are difficult for many children and may not be
produced correctly for many years (or ever). Phonetic
mistakes are not always due to incapacity, however,
as in the case of one linguists son (Smith 1973) who
at 25 months said puggle when he wanted to talk
about puddles, but pronounced puzzle as puddle.
A rule of some sort seemed to be involved here. The
same child provided evidence of an ability to perceive
and store discriminations which could not be
produced: he said/maus/for both mouse and
mouth, but would not accept adult usage of /maus/
to mean mouth. In early language comprehension is
quite commonly in advance of production. This may
be analogous to song development in some birds who
learn their song from other birds one season but do not
sing it themselves until the next year (Nottebohm
1970).
Beginning to use words
Children tend to begin by talking about what they

already know, familiar commonplace objects,
people, or events, particularly objects or events that
120
the child is concerned with as actor. Early words tend

to be nouns or regulatory words such as more, no,
up and so forth. They are not necessarily the same as
adult words in how they are used or in their exact
range of reference.
Using words as names
It may seem obvious that when a child says a word he

or she is referring to an object or event, that the word
names or represents the referent. McShane (1980)
argues otherwise. He sees early words as
inseparable from the functions that, in the context,
they perform for the child. Thus, if the child says
mummy only when making requests to his or her
mother and in no other context, mummy is certainly
functioning as part of the childs request. It is not
certain that, as far as the child is concerned, it
represents mummy or is her name; adults are likely
to believe it has these functions but it could just be for
the child a noise one makes when trying to get
something, as please or help or a certain sort of cry
might be. If mummy was used not just as a vocative,
as in requests, but to perform a variety of functions,
for example to answer questions, direct attention,
make statements and so forth, in a variety of contexts,
then it is safer to conclude that the child understands
it is his or her name for mother, and that the word
represents the person.
McShanes suggestion highlights the importance
of considering the functions of language and the
construction of meaning (see pp. 13540). It avoids
some of the difficulties which earlier theories of the
meaning of early utterances have run into (McShane
1980, ch. 2) but rather more evidence is required to
support it. The hypothesized sequence of events is as
follows. The childs caretaker systematically
responds to the childs utterances as if the child
intended to convey a particular meaning (see below
Chapter 6) and the child thus comes to learn that there
are contingent relationships between his or her
utterances and other peoples behaviour. Adults
difficulties in interpreting the child presumably
shape utterances towards a form which is more
intelligible and usually more conventional, for
example if the contingent relationship between

saying wowl and getting what you want is less
regular than the contingency between saying help
and getting what you want, help will tend to be
preferred. The child learns to communicate and to
intend to communicate, but what is communicated is
pragmatics needs, requests, directions, and so forth.
At about the same time, parents are introducing
children to the naming game (Ninio and Bruner
1978). In joint activities such as looking at picture
books the adult helps the child participate in the ritual
of whats that? Thats a doggie. To begin with the
adult plays both parts, but gradually pauses, prompts
and other devices lead the child to make an
increasingly large contribution to the routine. The
child utters names at appropriate points of the ritual,
but does not at first understand that this is naming, a
particular sort of activity relating language and the
world. It is worth pointing out that the concept of
naming is not a simple one: feeling that names are
an inseparable and unalterable part of their referent
(so that milk could not be called ink, for example),
or that ones name is private or even secret (Sinclair
et al. 1978) seems to be rather common. McShane
suggests as a result of taking part in the highlystructured ritual of naming the child comes to the
insight that the words originally embedded in the
ritual are names. That quite a lot of children show a
rapid increase in vocabulary between 18 months and
2 years, and behaviour which looks like asking the
name of every object they set eyes on, supports this
notion, though as Wells (1985) points out, many
parents deliberately teach names at about this time.
One documented instance of insight comes from
Helen Keller, who after being blind and deaf from the
age of 8 months had words finger-spelled to her by the
governess, Anne Sullivan, who was put in charge of
her education when she was 6. Anne Sullivan tried to
spell out on Helens fingers everything we do all day
long, although she has no idea yet what the spelling
means. Helen quickly imitated the hand signs but
made no connection between them and the objects
they symbolized. The insight came in an incident
which has become famous.
They had been wrestling with the words MUG and W

ATER, recorded Helen, and she persisted in confusing
the two. Later they went for a walk by the well-house.
Someone was pumping water. Annie placed Helens hand
under the spout and as the cool stream gushed over one
hand, she (Annie) spelled into the other the word water, first
slowly then rapidly. I stood still, my whole attention fixed
upon the motions of her fingers. Suddenly I felt a misty
consciousness as of something forgotten a thrill of
returning thought; and somehow the mystery of language
was revealed to me. I knew then that WATER meant
the wonderful cool something that was flowing over my
hand. . . . I left the well-house eager to learn. Everything had
a name, and each name gave birth to a new thought. As we
returned to the house every object which I touched seemed
to quiver with life.
Annie Sullivan wrote She has learned that everything
has a name and that the manual alphabet is the key to
everything she wants to know. . . . She has flitted from
object to object, asking the name of everything (Lash
1981, pp. 578).
Not every development of the concept of naming

would be dramatic, and like other concepts further
development is to be expected. One development that
McShane suggests follows the initial insight that
objects have names is that attributes, events and
actions have names too. Linguistic accounts grow in
length and develop grammatical structure to convey
these different sorts of naming and reference. Here
too adult scaffolding and extension of childrens
utterances provides a model and a frame.
Early vocabulary
The limits of early vocabulary items have been seen

as of interest as possibly revealing how young
children associate label and referent. Quite
commonly children use an early word to refer to
many more objects than an adult would, for example
doggy is applied not just to dogs but to cats and
pieces of fur, in what is called over-extension. (The
converse, under-extension, is also found, as when
doggy is only used to refer to the family dog.) Eve
Clark (Clark and Clark 1977, pp. 4927) suggested
that in over-extension children first associate the
word with one or two particularly salient
121
characteristics of the object, for example furriness,

and so use it over-extensively to refer to anything
characterized by furriness. Later they add in other
salient features as requirements for dogginess, such
as four-leggedness, size, ability to bark, narrowing
their definition towards the adult level. Word, and
concept, are seen as proceeding from particular
instance to generalized abstraction by the increasing
specification of features. (See Chapters 2 and 3 for
discussion of cognitive models and semantic
networks.)
It must be noted that childrens early words are not
necessarily simple neutral efforts to label. Some
involve more complex activities such as commenting
and comparing (doggy might be appropriately
glossed in some instances not as thats a doggy but
as thats rather like a doggy or my doggy has fur
like that too). As children get older, simile and
metaphor become more likely. So do other indirect
language uses. One child is recorded as having said
heavy in many situations which involved her in
notable physical effort, including not only pushing
open doors or lifting bricks but also climbing a long
steep flight of stairs. If she had been of the right
generation and subculture, she might have
appropriately used her favourite word to refer to
situations involving social and interpersonal
difficulties: Man, being arrested is a real heavy
scene. Over-generalizations do not unequivocally
indicate inability to distinguish between objects or
events: they may involve confusion, or they may
involve comment, comparison, metaphor or joke.
Social demands may also enter into the situation; if
one feels socially obliged to say something but lacks
the correct vocabulary item, one may produce a
slightly inappropriate word or phrase as an
approximation in verbal terms but an adequate
response in social ones. This sort of experience
comes fairly frequently to people who are not fully
competent users of the language in question.
It should also be pointed out that the vocabulary
children use and their degree of over-extension or
under-extension may be related to adult usage of
language to children. Here adults seem to have a
sense of level of appropriateness related to the
122
distinctions they want the child to make. Plant is

thus seen as a bit too general, rose as a bit too
specific, flower as about right; and contrasts are
drawn through the means of vocabulary between a
flower which can be admired, smelled, not walked
on, etc., and other plants such as grass, tree,
vegetables, which can variously be walked on,
climbed, eaten and so forth. The adults own interest,
and the adults encouragement of the childs interest,
will make notable contributions to the childs later
vocabulary. So will the distinctions the local culture
makes and uses. A country child may quickly learn
distinctions of which a town adult remains unaware.
There is a salutary story, no doubt apocryphal, of a
school inspector testing the general knowledge of
children in a small country school. He showed them a
picture of a sheep, but his Whats this? was received
with puzzled faces. Eventually a child said I aint
never seen one of those its got a face like a Cheviot
but its backs like a Jacob but neither of thems got
horns like that un. Labovs work makes a similar
point (see p. 131).
Generally it would appear that children begin with
only one or a few appropriate vocabulary items but
develop differential ranges. Halliday (1975) provides
examples of initial single words being used in a
variety of situations but being replaced by a range of
words used selectively according to the
circumstances. As we shall see in examining the
development of language as a functional
communication system, this is an important part of
later development and crucial to successful social
life.
From single words to sentences
As we have seen, developing words which refer

stably to objects and events is a tremendous
achievement for the child, but in order to get very far
with communication words have to be combined into
sentences, sentences into longer passages and so
forth. Grammatical rules specify how words may be
combined to express increasingly precise meanings.
Children have to learn to combine their words in ways
which obey the rules. They can do this in rudimentary
ways very early: managing some of the most subtle
language structures requires considerable practice if

not explicit training (e.g. whether the word none
takes a singular or plural verb).
As my comments above on the interpretation of
one-word phrases imply, the beginnings of grammar
are obscure. Only when two morpheme utterances
begin can we look for evidence of word order, which
is, of course, an important grammatical device in
English. Childrens earliest word orders tend to take
the adult form and to express salient grammatical
cases such as action, possession, location and so
forth. Both the meanings and the structures which 2year-olds express are of very similar sorts in most of
the languages studied: they are usually uninflected
words, mainly in the order which an adult would use
to express the same meaning, though the functor
words like the, by, but are omitted in favour of
the words that carry most of the meaning. Some
phrases are taken on wholesale from adult language
and used as unanalysed chunks or single words
wossat, gimme are obvious examples. Peters
(1983) argues with some cogency that this reflects
some fundamental problems in our concepts of
word and syntax. Even adults use some phrases as
units which although potentially analysable are not
normally broken into their components polite
formulae such as how do you do are of this type.
Peters suggests that language learners acquire units
which consist of one or more words or morphemes,
and which then become candidates for segmentation
into smaller units. If a unit can be segmented into
smaller units these are added to the lexicon, and the
original unit may be retained. For example, the
original unit how do you do can be segmented into
its separate words but may also be kept as a social
formula, howdjado, which expects a different
response from its components as well as being used as
a unit. Segmentation also contributes to knowledge
of the languages structures.
Children do of course produce utterances which are
unlike adults (all gone sticky) or which over-extend
adults rules (mouses, goed). Thus their development
of grammar is not learned in any simple automatic
way from adults. Noam Chomsky, impressed by the
speed, regularity and specificity of childrens
language development, by apparently universal
features of language and by evidence of specialized
language centres in the brain, proposed a model of

language development in which an innate
understanding of fundamental linguistic rules
(centring on using syntax to express meaning) was
activated by the language the baby heard and
accounted for the speed and regularity of
development (see, for example, Chomsky 1976;
Clark and Clark 1977; Dale 1978). A great deal of
research has been carried out aimed at constructing
grammars of child language in terms of Chomskys
model of transformational generative grammar (e.g.
Bloom 1970), or in terms of case grammars focusing
on concepts like agent, action, locative (e.g.
Bowerman 1973). (Dale 1976 and Maratsos 1983
provide useful accounts of both grammars and the
associated research.) A Chomskian model of
grammatical competence, which centres on a
syntactic component represented without respect to
meaning and involving transformations, is not now
seen as a good psychological model of young
childrens language development. Beyond that, it is
not as yet clear what grammatical models are most
appropriate for the description of early language, or
indeed for adults language (Maratsos 1983); one
particularly important aspect of the debate is how far
it is sensible to treat syntax separately from semantics
and pragmatics, that is to separate the formal
grammar of what is said from what is meant and what
effect the utterance has or was intended to have. We
will look at these aspects of childrens language
development presently.
The question of whether all children develop
language in the same way is of some interest. It has
been impossible to answer accurately because
research on language development in young children
is very time-consuming at both the data-collecting
and analysis stages. Most samples have been very
small and many have been drawn from middle-class
intellectual families, often the researchers own. The
nearest approximation to a large and socially
representative sample that I know is that studied since
1972 by Gordon Wells and his colleagues in the
Bristol
Language
Development
Research
Programme. After a description of the structure and
methods of this programme I will outline the answers
it gives to the question of whether all children
develop language in the same way.
123
The children studied were 128 Bristol residents.

Each was observed at three-monthly intervals for
twenty-seven months; that is, ten observations. Each
child wore a small radio microphone which recorded
the childs speech and other peoples speech to the
child for a number of brief periods spread through the
day so that neither child nor family knew with
certainty that the microphone was on. The researcher
was not present during this day but checked the tape
with the childs mother in the evening to complement
the sound tape with contextual information. Half the
children were observed from 15 to 42 months, the
other half from 39 to 66 months. Further studies of
some of the sample in school were made (and indeed
some of the sample are currently being studied as
second-language learners). The families of the
children were representative of the entire social class
range, except that the proportions of very high and
very low social class were increased to give an
adequate number for analysis. Thus the programme
has produced a very large quantity of data which may
be analysed in terms of sex, age and social class
variables, which are representative of the
spontaneous speech of British children between 15
months and 51/2 years, and which are probably
relatively unspoiled by the participants
consciousness of being observed.
Wells (1985) presents the resultant picture of the
sequence of language development, and although the
use of spontaneous language gathered at quite long
intervals may mean that language items which occur
infrequently appear later in the data sequence than
they would if attempts had been made to elicit such
items, the picture is almost certainly pretty accurate.
There seems to be a universal sequence of
development, at least in general outline (Wells 1985,
p. 224).
The data do not tell us exactly why we find this
order of emergence. Theorists have suggested that
uses and structures appear early or late because they
are frequent or infrequent respectively in the
language which the child hears, and the Bristol data
lend some support to this hypothesis (Wells 1985, ch.
9). Another suggestion has been that emergence is
correlated with linguistic and cognitive complexity,
124
and although it is not unproblematic to define such

complexity, the emergence of auxiliary verbs,
pronouns, meaning relations and functions did seem
to be correlated with complexity. It is, of course,
rather likely that frequency and complexity of
language influence each other: if we need to say
something often we may gradually simplify it
(acronyms such as BBC, USA and BPS are examples
of this). Mothers adjustment of their language to
what they believe will fit their childrens current
competence will increase the frequency of less
complex items in the childs language, providing
another sort of interaction between frequency and
complexity.
Adults talk to children
It has been one of the most consistent findings in

studies of language development that mothers and
other adults (and indeed older children, for example,
Shatz and Gelman 1973) adjust their speech to the
child developing language. Among the many
adjustments are attention getters and holders, such as
a frequent use of the childs name, a high pitch or
exaggerated intonation, and many gestures and
touches; restriction of semantic content, for example
by talking more than usual about the here and now
and by careful selection of vocabulary by rules like
the level of appropriateness mentioned above;
syntactic restriction to brief and simple sentences
without, for example, passives or subordinate
clauses, but with lots of repetitions; a specialized
strategy of discourse, high on expansions and
extensions of the childs own utterances, high on
contemporaneous comment on ongoing activities,
and high on questions, directions, prompts and
modelling of discourse, and produced more slowly
than language to adults. This sort of language has
been called baby-talk or motherese: it is a
relatively consistent, organized, simplified and
redundant set of utterances. It is thus very unlike the
disorderly and degenerate language which was all the
child in Chomskys account had to learn from:
motherese has been seen as quite the reverse, as a
particularly good source for the child to learn
language from.
Maratsos (1983) raises certain queries about this.

He first points out, correctly, that parents do not use
motherese in order to teach their children language;
they use it because they are trying to keep the child
interested and understanding. Wells (1985, p. 380)
makes the point neatly:
for most of the time the relatively finely tuned modelling of
meanings and forms that the frequency data reveal occurs
incidentally, as adults carry on conversations with their
children for quite other purposes to control the childs
behaviour in the interests of his safety and their joint wellbeing, to share in and extend his interests, to maintain and
enrich their interpersonal relationship and so on. Success in
achieving these aims requires that the majority of the
adults contributions be pitched at a level of complexity that
is not too far beyond the childs linguistic ability. However,
this is achieved quite spontaneously by most adults under
the control of feedback from the childs comprehension and
production and does not require deliberate attention. The
tuning that occurs is thus as much a response to, as a
determinant of, the sequence in the childs learning.
The Bristol data show a great deal of this fine

tuned linguistic input, frequently with an increase in
the complexity of the adults language just before an
increase in the complexity of the childs. Wells
comments on this (p. 381):
It appears, therefore, that the influence of the input on the
childs learning is enabling rather than determining. Once
the child has the prerequisite cognitive understanding of the
distinction which is encoded by a particular linguistic
category, frequent appropriately contextualized
occurrences of the category in the speech that is addressed
to the child provide opportunities for him to make the
connection between linguistic category and non-linguistic
experience.
Fine-tuned input should not be seen as either

necessary or sufficient for children to develop
language, since there are several accounts of children
who do not seem to have been talked to in this way but
did seem to develop normal language (Lieven 1978;
Ochs and Schieffelin 1983; Brice Heath 1983;
Romaine 1984). It does seem possible that fast

developers encounter particularly finely-tuned
language (e.g. Cross 1977, 1978) but otherwise
differences seem to be not in what is possible but in
what is usually done, that is, in the distributions of
tokens, not in the range of types. However the
facilitating functions of adult speech are probably
broader than the categories used in studies of
motherese which concentrate on features like Mean
Length of Utterance, sentence types or syntactic
expansions. Wells lists five relevant types of
intention (pp. 3989): 1 to secure and maintain intersubjectivity of attention, 2 to express ones own
meaning intentions in a form that ones partner finds
easy to understand, 3 to ensure that one has correctly
understood the meaning intentions of ones partner, 4
to provide positive responses in order to sustain the
partners desire to continue the present interaction
and to engage in further interactions in the future, and
5 to instruct ones partner so that he or she may
become a more skilled performer. This last intention
applies only to some sorts of interactions, teaching
vocabulary in conversation about picture books
being a well-documented example. These intentions
underlie the behaviour which has been seen in
accepting or responsive mothers (Nelson 1973;
Lieven 1978). They have the short-term effect of
sustaining the current conversation, and the longterm effect of maintaining the childs general
motivation to interact and especially to converse with
his or her adults. They also provide an opportunity for
modelling to the child ways of doing things with
language which he or she needs and cannot quite
produce alone, utterances which are within the
Vygotskian zone of proximal development. (They
also contribute to the opportunity to discuss cognitive
issues: Wood (1980) and Tizard and Hughes (1984)
stress the usefulness of conversations between adult
and child based on the childs current interest as a
painless way of learning about the world.)
Children who take part in a lot of child-contingent
conversations of this sort seem to develop
linguistically (and cognitively) rather well. However
we cannot yet conclude that the conversations
accelerated their language development. Bates et al.
125
(1982) outline some of the reasons why such a

conclusion would be premature. They point out that
emphasis on motherese as a teaching device
neglects the ability and the willingness of the child to
learn. The childs knowledge of language and the
social world must be considered: we will discuss its
development later (p. 137). So must childrens
ecosystems (see p. 132). Second, even if a feature of
parental language is correlated with a feature of the
childs language development, we cannot assume
that the former caused the latter. It might be the
reverse, that parents are responding to the childs
idiosyncracies, or a two-way relationship where child
and parent each react to the other; or there may be a
more indirect relation between the two measures.
They give an example (p. 51) from a study (reported
at a 1975 conference) where Tulkin and Covitz found
a significant correlation between the prohibition
ratio in parents speech to the child at 10 months and
childs performance on a vocabulary test at 6 years.
Bates et al. decline to conclude that saying no to
children decreases their vocabulary in any direct way,
and I think even people who advocate extremely
permissive childrearing would hesitate to draw this
conclusion. It is more likely that relationships which
include a lot of prohibition by parents have other
qualities which adversely affect the childs
development. Perhaps they are short on joint
attention, child-contingent conversation, parental
warmth, and opportunities for the child to explore the
world at will but with an appreciative adult to put his
or her discoveries into an accessible model. Only
very detailed analyses of the pattern of childrens
experience, in conjunction with experimental
evidence, can sort out the pathways of causation.
Wells (1985, ch. 8) points out that there are marked
differences in the language used in different social
contexts in the Bristol data. For example,
representational speech was high during sessions of
reading or watching TV, controlling speech was
high during caretaking, eating and imaginary play
without adults (which involves a lot of defining or
allocating roles, e.g. you be the dog, you gotta
crawl and bark). Davie et al. (1984), observing 3 to
5-year-olds at home, similarly found context
126
differences in language use. Looking at books, for

example, was associated with highly informational
talk from adults a high rate of naming and labelling
objects in pictures, and a low rate of general
conversation. I will return to this point about context
and language in its social class aspects presently; at
the moment it serves as a warning that we must have
a representative quantity of data before we draw
conclusions about what language input a child is
experiencing and how it is associated with
developmental outcomes.
A quotation from Wells (1985, p. 394) summarizes
the complexity of the enterprise of understanding
language development, making points which we will
see apply just as strongly to our discussion of social
development (Chapter 6).
In various ways, therefore, the differences observed in adult
behaviour may owe as much to differences between the
children with whom they converse as to inherent
differences in the adults themselves. But the reverse is also
true, and so, in seeking to explain the differences in adult or
child behaviour we must recognize that, ultimately they are
as likely to emerge from the interaction between a
particular pair of participants, as they are to be attributable
to either participant considered separately. If we are to
untangle the relationship between features of the input and
progress in language learning, therefore, it will be
necessary to develop models of multiple and reciprocal
causation operating within a matrix of interaction, which,
on any particular occasion, is also affected by the particular
context in which it occurs. Since we are very far from
having such a model, it seems for the moment safer to
conclude more modestly that although the evidence
supports a belief in the potentialy facilitating effect of the
adult input, this facilitating input itself is the product of
interaction to which both child and adult contribute to
varying degrees.
Language differences and social class
As it was once for Jane Austens matchmaking

matrons a truth universally acknowledged, that a
single man in possession of a good fortune must be in
want of a wife, it was for some time a truth
universally acknowledged that a child from a
working-class background must be in want of a
language different from that of his or her home if

failure in school was to be avoided. Both truths have
been intensively attacked: we will consider the latter
here.
Native English speakers perceive differences of
sound, syntax and vocabulary in peoples speech
which they relate to social class. (The concept of
social class is a problematic one (see Giddens and
Held 1982) which I do not propose to analyse here. In
this section what is referred to is social status as
indicated by occupation: in the case of children,
fathers occupation.) Romaine (1984), summarizing
the results of surveys of peoples speech in various
countries, says that, simplistically, the finding is
that the middle-class adhere more closely to the
norms of the high prestige standard language, while
the working-class speak in less standard ways, closer
to the local vernacular. Sometimes research picks up
enormous class differences: Romaine (pp. 856)
cites a study by Trudgill of the use of verb forms such
as he go, verbs without /s/, in Norwich. Speakers
from the middle classes virtually never used this
non-standard suffixless present tense, while
working-class speakers used it from 70 per cent
(upper working class) to 97 per cent (lower working
class) of the time. Often class differences are much
smaller, however, and there is more variation within
classes: indeed many speakers adjust their speech to
their situation and their listener. Nevertheless,
members of the community of native English
speakers hear the differences in peoples speech and
may use them to judge the speakers social class and
other characteristics (Giles and Powesland 1975).
There are also differences in educational
achievement between children from different social
classes. Again simplistically, children from
working-class families are more likely to leave
school at the earliest opportunity and with minimal
formal qualifications, having been more likely to be
poor readers or non-readers from primary school on.
These differences may be larger on verbal measures
(including reading tests and intelligence tests) than
on less verbal ones. It has been suggested, and often
believed, that these two phenomena, language
differences and differences in educational

achievement, are closely related; particularly that the
former cause the latter. Working-class children are
said to fail in school because they 1 do not, and 2
cannot, use language in middle-class ways.
This hypothesis reflects in a popularized form the
ideas put forward by Basil Bernstein. Bernsteins
ideas have changed considerably over the last
twenty-five years: the three volumes Class, Codes
and Control contain papers from the period 1958 to
the early 1970s. The more recent formulations are
much more complex and subtle than earlier ones, and
their psychological and educational implications are
different. Since the early formulations are still current
in some accounts of language and education,
however, I will present a brief account of the
successive models.
The original influential version of the theory was
that there were two distinct types of language, usually
called restricted and elaborated codes.
Elaborated code was said to be grammatically
complex, with frequent subordinate clauses,
passives, impersonal pronouns (One sees), and a
wide and unusual vocabulary of adjectives and
adverbs: these features were used in the service of
precision and explicitness. Restricted code was said
to use short, simple sentences, often incomplete or
elliptical, and was far more repetitive, rigid,
imprecise and implicit: it contained many more
appeals to a shared context of understanding, for
example in the use of phrases like you know, sort
of, innit?. Middle-class speakers were said to use
both language codes, but some (and some came to
imply all in the popular version) working-class
speakers were supposed only to have access to
restricted code. This deficit affected the way they
could express themselves, and, it was postulated, the
way they were able to think (verbally) to themselves.
Further, since the language of school work is said to
require an elaborated code, the inability to use
anything except a restricted code explained a good
part, if not all, of the working-class childs poor
school achievement. He or she had a language
deficiency which would directly prevent good
127
performance in school unless it was remedied by

teaching the use of elaborated code.
This model was quickly given up by researchers in
the area, including Bernstein himself and his
colleagues, although it continued to influence
teachers and classroom remedial programmes (e.g.
Tough 1973, 1977). Among the problems was, first,
lack of evidence that the two codes had a real
existence as distinct entities rather than as different
modes on a continuum of variation. The literature
virtually never contained transcribed examples, and
linguists found the defining criteria unsatisfactory
(Gordon 1981). It was generally the case that the
evidence said to support the existence of separate
codes was merely that one group of speakers tended
to use more of a particular language form (e.g. passive
verbs, implicit pronouns) in a particular situation
than another group of speakers. For example in a
much-quoted experiment by Hawkins (1973, 1977),
5-year-olds were required to tell the story depicted in
a series of pictures given by the tester. Middle-class
children used more nouns to convey who was doing
what in the pictures: lower working-class children
used more pronouns, relying on the pictures to make
their referent unambiguous. Hawkins argued that
these implicit restricted code children were
conveying less information about the pictures: their
version
makes enormous demands on the listener. It means that the
context (i.e. the pictures) must be present if the listener is to
understand who and what is being referred to. It assumes
the listener can see the pictures (Hawkins 1973, p. 87).
Since, however, both the child and the listener/

tester can see the pictures in the test situation, the
demands and the ambiguity are more in Hawkins
judgement than in the social ecology of the setting.
Subsequent work (e.g. Hughes et al. 1979; Labov
1969; Heath 1983; Wells 1979) has amply
demonstrated the fallacy of arguing that tend not to
in situation X implies cannot in any situation.
Subsequent formulations of links between language,
class and education have taken into account the effect
of different social contexts and the question of how
128
use of language relates to language capabilities. It has

also been realized that the relationship between
linguistic form and cognitive complexity is not as
uncomplicated as might have been supposed. Simple
syntax can express logical sequence, hypothetical
possibilities and interdependent propositions:
complex syntax can often enough be used to disguise
banal ideas.
Linguists have also found Bernsteins notion of
implicitness or explicitness in language to be
problematic. In implicit use of language the
meaning of a statement is not completely spelled out,
as the speaker assumes that the listener has enough
basic information to understand the unspoken part of
the message. Strictly, however, all use of language
involves assuming some shared understanding. We
have to assume for example some shared vocabulary,
syntactic knowledge and understanding of the social
rules of discourse: if we could not make this
assumption we could not communicate at all.
Suppose we say the cat wants his dinner. This
simple sentence assumes the listener knows what a
cat is; knows that we are talking about a particular cat,
since weve said the cat, and indeed were very
probably talking about our own familiar domestic
cat; similarly understands wants his and dinner;
and understands that in normal discourse the sentence
would convey (at least to English speakers and
listeners) an obligation to feed the cat. What looks
like an explicit sentence of minimal complexity
actually depends on a large amount of implicit
information and works within rules of discourse that
are also implicit, at least until they are broken (see
Grice 1975; and p. 135 of this volume).
Implicitness is thus a normal part of language use,
and indeed it would be impossible (or at least tedious)
to be perfectly explicit on all occasions. Certainly
there are likely to be problems in communication if
the speakers assumptions about shared aspects of
meaning are incorrect and too much is left implicit;
the listener may not understand what is said. It may
even be a more serious breakdown in communication
than over-explicitness, which would presumably
leave the listener in a state of adequate understanding,
though probably some social resentment against the
speaker (pedantic old bore). What degree of

implicitness is appropriate is a social property of the
participants whole discourse, and one of the aspects
of language we learn from our experience of
communicating with others about our world.
Bernsteins analysis is unsatisfactory in that it deals
with implicitness as a property of the text. As we saw
in the case of Hawkins experiment, paralinguistic
means to explicitness such as pointing are not
included.
There are various grammatical features which are
used to make reference unambiguous (see Romaine
1984, especially pp. 1436); the indices Bernstein
uses, types of pronouns and ratio of pronouns to full
noun phrases, are inadequate as they ignore the
situational context of the speech and, in practice,
underestimate the effect of links between utterances
(Gordon 1981).
Empirical investigations of whether there really
are class differences in explicitness or implicitness of
speech suggest that on the whole there are not. In
Tizard and Hughes study (1984), for example,
mothers of all classes were sometimes implicit in
talking to their daughters, but took good care to be
perfectly explicit when it was felt to be important that
the child should understand. Mothers varied in how
much they demanded the child should be explicit:
Robinson and Robinson (in press) suggest that
parental demand that children should think about
how to make themselves understood is a cause of a
faster understanding of communication (see p. 137).
Wells (1985) also found more variation within
classes than between them.
Bernsteins later work (e.g. Bernstein 1973) moves
codes into a level of abstraction some distance from
observable speech. Codes are now seen as a sort of
underlying competence which give rise to speech
variants which are, roughly, the general range of
syntactic forms, vocabulary items, and sort of
meaning expressed. (See 2 and 3 for problems raised
by competence/performance distinctions.) Syntax
and vocabulary in speech variants are very like those
proposed for the early version of codes: in
restricted speech variants, meaning is said to
concentrate on particularistic implicit meaning

shared between speaker and listener, while meaning
in elaborated speech variants is more concerned to
be universalistic and explicit, appealing to high
level general principles. For example, a restricted
code explanation for a parental command would be
something like Because I say so; an elaborated
explanation would be Because people need to eat
their spinach all up to grow into nice big strong
people.
Much about language is undoubtedly learned
through socialization (see Chapter 6). Bernstein
(1971) differentiates between two sorts of families
which have different status structures and different
communication systems. Positional families are
said to have clear-cut definitions of the role and status
of different family members: children shall be seen
and not heard; daughters-in-law must defer to their
mothers-in-law; the oldest son inherits the land, the
second goes into the regiment, the third becomes a
clergyman and takes up the family living: rigid status
positions and roles are filled according to
characteristics such as age, wealth and sex, not
according to the particular strengths, weaknesses or
desires of individual family members. In personcentred families these status distinctions and
ascriptions of roles are modified and varied in terms
of the idiosyncracies of individuals: people can
achieve their positions on the basis of their merits,
rather than having greatness (or exclusion from
greatness) thrust upon them. Person-centred families
are said to have open communication systems;
having a more fluid status system they need to
communicate, negotiate and explain, and hence use
elaborated code. Positional families have closed
communication systems and thus use restricted code,
with strict social control based on commands and
prohibitions explained only in terms of family
members status Do as your father tells you;
Little girls cant do that. Bernstein suggests that
working-class
families
are
predominantly
positional, and middle-class families personcentred.
129
Figure 16
Diagrammatically, his model is shown in Figure 16,

where heavy arrows indicate a strong link and light
ones a less certain link.
Both types of family structures may be found in
each social class, and family types may vary in their
communication system and codes: the version of the
model drawn by Stubbs (1983) has, rather unkindly,
all possible diagonal arrows drawn in between levels!
However Bernstein (1973) is claiming tendencies to
different class distributions, so there should be more
emphasis on the vertical chains as drawn above, even
if, exceptionally, diagonals may occur. It is expected
that, on the whole, working-class families will be
positional in structure; that, on the whole, positional
families will generate closed communication
systems; that, on the whole, closed communication
fosters restricted code; that restricted code leads to
restricted speech variants; and that restricted speech
variants lead to educational underachievement.
This more elaborate formulation stands up to
investigation little better than its predecessor. Just as
130
there is no evidence for two distinct language codes

there is no evidence for two distinct family types, let
alone two types clearly separable by class, and
differences between families look like variations on a
continuum, or, rather, on several continua which do
not map on to the positionalperson-centred
distinction without loss of information (see Chapter
6). There is very little evidence that different family
types talk in distinctly different ways. A few families
are characterized by continual escalating conflict and
non-communication (Patterson 1975; and see
Chapter 6) but these aside, most families sometimes
use explanations and sometimes not, sometimes ask
or encourage questions and sometimes not,
sometimes get involved in discussion of cognitively
complex ideas, sometimes negotiate, sometimes
command, and so forth. Wells (1985), reviewing the
largest body of data available, emphasizes that by the
time of their entry to school virtually all children had
heard and used at home the same large range of syntax
and language function. There are class differences in
amounts of questioning, discussion, complex use of

language, vocabulary and use of books and
imaginative play between mother and pre-school
child, but the differences are not large (Tizard and
Hughes 1984; Wells 1979, 1985; Davie et al. 1984).
There are more striking variations within classes than
between them.
There is no good evidence at all on whether
different sorts of language structure lead to different
sorts of cognition. The difficulty of assessing
cognitive structures is enormous (see Chapter 2 and
3): so is the difficulty of assessing the cognitive
complexity of a language. Although there is a fairly
plausible case that it is easier to talk about and think
about things we have good vocabulary entries for (so
that the Eskimo is more fluent in discussion of snow
than the Spaniard, or the skier than the surfer), there
is virtually no reason to believe that language
structure severely restricts thought, and, especially,
no good reason to believe that different languages
vary in their possibilities of being explicit or logical
(Labov 1969).
Differences in the use of language, however,
including the social functions of language, may
contribute to at least differences in knowledge about
language and possibly also differences in cognitive
processes. Children who have been encouraged or
required to think about language itself, for example,
seem to be faster in developing some aspects of
knowledge about language. Elizabeth and Peter
Robinson have carried out a careful and thorough
series of experiments on developmental changes in
childrens understanding of communication failures
due to the message having an ambiguous referent.
These showed that children who had had their
attention drawn (rather specifically) to the fact that
the listener had not understood what the child meant,
were better at playing a referential communication
game and showed more understanding of why
messages were not understood and of how to make
them unambiguous (Robinson and Robinson 1982,
1983; and see p. 107 this volume). Children who have
been taught about the sounds of letters do more
phonic processing in their reading and read better
(Chall 1967; Bryant and Bradley 1985; and see p. 78

this volume). Children whose parents have talked to
them in ways similar to those of their teachers
probably understand rather better what their teachers
require of them and how to provide it (Willes 1983;
Wells 1981, 1982).
The emphasis shifts here to add an indirect link
between the childs language and his or her
performance in school, with the teachers judgement
of the child and its consequences (and perhaps the
childs judgement of the teachers and school) being
the crucial intervening variable. Teachers who think
a child is stupid or inarticulate are likely to expect less
and require less than they do of a pupil judged to be
clever. A child who understands school as foreign,
and teachers as people who impose meaningless tasks
and ask bizarre questions, is presumably less likely to
work hard and well than a pupil who finds school
work and talk familiar, and understands why it is
worth doing. This sort of fit or misfit is not likely to be
simple, and the shifted emphasis is far from giving us
an adequately detailed explanation of the educational
failure which was the instigating problem. However
the attention currently being given to the participants
view of education (see p. 198) to the sort of discourse
that characterizes classrooms (see p. 132), to the
patterns of stimulation, support, encouragement and
motivation between parent and child, and teacher and
child (see p. 184), and to self-concept development
(see p. 150), seem likely to lead us to a better
understanding of how to encourage everyone to be as
clever as possible in as many ways as possible.
Language at home and at school
A popularized and radically cruder version of

Bernsteins work argued, first, that children bring to
school the language they used at home; second, that
the middle class had access to what was called
elaborated code while the working class did not;
third, that working-class children failed in school
because they lacked the elaborated code which
school required; and fourth, that special language
programmes should be run to remedy this deficit.
Following a great deal of polemic, some careful
131
collection and analysis of data has shown that there is

a great deal to be said against each step of the
argument just outlined.
It is necessarily true that children enter school
having already developed many aspects of language.
They have after all been using language for four or
five years. If what I have said about the need for
adultchild conversation and scaffolding is at all
true, this is just as well: no teacher could provide it for
thirty children at once. Most children of 5 have shown
themselves to be capable of constructing most of the
sentence types and most of the semantic relationships
of their language, including many complex ones (see,
for example, Wells 1985, discussed earlier in this
chapter). They still have things to learn about
language, as we shall see, but they are basically
linguistically competent. Do they bring this
competence to school, as the first step of the argument
asserts? The true answer is both yes and no, and the
differences and discrepancies look important for the
problems of social class and educational failure.
It is the case that some children are tongue-tied and
monosyllabic when required to talk to teachers in
school, and that such children often come from a
background unlike the teachers. This apparent
inability to respond to the linguistic demands of the
classroom is often interpreted as showing that the
child cannot use language in the required ways, not
merely that he or she does not. Closer observation of
children shows, however, that in many cases they can
and do use language perfectly competently when they
are with their peers or members of their families (e.g.
Wells 1982; Tizard and Hughes 1984; Wood 1981).
This suggests that an explanation in terms of a
deficiency in the ways the child can use language is
untenable. An alternative explanation is that there is
something unusual about the language usage
required by schools, some factor in the demands that
school makes on children which is unlike the
demands of the home. Recent analyses (e.g. Sinclair
and Coulthard 1975; Romaine 1984) of classroom
language and behaviour suggest some focal
differences. Most of the discourse which children
have taken part in at home has arisen from the
participants activity. It has a background of shared
132
knowledge, so that each participant can assume the

other knows roughly what is being talked about and
what the aim of the conversation is. Much of the
discourse is initiated by the child. School discourse is
much more likely to be teacher-initiated, adultstructured and devoid of practical context. For
example, a high proportion of teacher talk consists of
display questions such as what colour is the
house?, where the teacher requires the child to give
an answer which the teacher knows, and which the
child knows the teacher knows. Questions at home
are much more likely to occur because the questioner
(child as frequently as adult) lacks information which
he or she believes the respondent can supply, such as
what would you like for dinner?
Romaine (1984) points out that school
conversations frequently have goals which are
known to the teacher but not to the child, who is rarely
told explicitly what the criteria for correctness are.
She quotes (pp. 712) an example recorded by
Dannequin of a French teacher who has strict but
implicit criteria for what is an acceptable answer but
expresses them so ambiguously that her pupils have
to go through a long problem-solving process to get
the right answer.
T:
Avec quoi prend-on la temprature?

What does one take a temperature with?
P:
Un thermomtre.
A thermometer.
T:
Une petite phrase. Avec quoi prend-on la temprature? Vronique, une phrase.
A short sentence. What does one take a temperature
with? Veronica, a sentence.
V:
Un thermomtre.
A thermometer.
T:
Tu me rsponds par un mot. Je veux une phrase.

Youve given me a word. I want a sentence.
V:
Avec un thermomtre.
With a thermometer.
T:
Ce nest toujours pas une phrase. Tu me rsponds

par un autre mot. Je voudrais une phrase Myriam.
Thats still not a sentence. Youve just given me
another word. I want a sentence Myriam.
M: Maman/prend/la temprature/avec un thermomtre
(utterance is syllable-timed with each chunk forming a separate tone group).
Mother takes a temperature with a thermometer.
T:
Voil. Rpte Vronique.

There. Repeat, Veronica.
These pupils are being required to give up the normal

discourse rule of not giving superfluous information
(see the beginning of this chapter, p. 118) in favour of
the showing-off of grammatical competence, and
what is more, a grammatical competence somewhat
arbitrarily defined by the teacher (une phrase can
mean a phrase in the English sense, so with a
thermometer or indeed a thermometer would be
technically adequate as answer). It is hardly farfetched to feel that Vronique comes across as
justifiably irritated and Myriam as getting rid of a
pointless importunity. This is a particularly neat
example because the teacher is requiring
metalinguistic awareness, that is an ability to talk
about language as having an objective existence, as
being opaque rather than transparent.
Dannequins teacher is, however, representative here
of a very high proportion of teacherchild discourse:
Wells (1982) and Wood (1981) provide other
examples. Heath (1983), in a brilliant study of
children from different neighbourhoods in the southeast of the United States, reports awareness on the
part of some children that different people used
language differently, asking different sorts of
questions, for example. One little boy, being given a
lift in Heaths car a few weeks after he started nursery
school, asked her a series of display questions about
a fire-truck that passed, e.g. What colour dat truck?
What colour dat coat? What colour dat car? Such
questions were rare at home, and Heath expressed her
surprise at him asking questions to which they both
knew the answer. The boy began laughing: he had
been imitating the questions the teacher asked at
school. If, as is probably the case, children playing at
being teacher show even more stereotyped behaviour
than children playing at being mothers, the reason
may be somewhere in the oddness of much teacher
child dialogue.
Children become aware quite early that they are

expected to talk in different ways with different
people. Politeness rules, which are relatively
systematically taught by many parents, are one
example. Edinburgh schoolchildren reported to
Romaine (1984) on the differences between polite
and rough speech, the former being more like
standard English and the latter more like Scots in
features of pronunciation (cannae instead of cant,
not down but doon) and vocabulary (ken instead
of know). Dialect differences such as these are
associated by adults with differences in prestige,
personality characteristics, education, occupation
and so forth (see Giles and Powesland 1975), rural
accented people being seen as nicer than urban
accented people, and those using RP (Received
Pronunciation or BBC/Oxford English) as being
better educated and more powerful, though possibly
not so trustworthy. Romaines informants seemed to
be making these sorts of distinctions by the later part
of their primary school careers; by then, too, girls
language was usually nearer the prestige norm than
boys language, and they used fewer non-standard
forms than boys. By this sort of age children used
more non-standard language to their peers, too,
sometimes seeing this as affirming (or required by)
their social identity. Peer groups and other social
groups often press their members into particular
vocabularies, dialects or uses of language (see
Romaine 1984, ch. 6.3; and p. 191 below).
School and home, then, show somewhat different
ranges of language use. Some homes do include more
school-type language than others, and homes which
encourage relatively abstract and contextindependent uses of language are quite likely to
produce children whose language is advanced and
who do cope well with school language (Wells 1982,
1985). A family which involves the child in literacy
seems to be particularly advantageous (see also
Osborn et al. 1984); no doubt this is partly because
literacy is one of the major preoccupations of schools,
but it is also probable that literacy encourages
abstract and context-independent thought (see p.
140). There are class differences in literacy behaviour
133
Plate 11
134
such as reading, writing and discussing what is read

or written, the middle classes indulging in them rather
more, but otherwise there is as much variation in
language use within social classes as there is between
them. Talk of two homogeneous social classes,
middle and working or lower-working, with two
separate language codes, is seriously misleading (and
to be fair, Bernstein would not have approved of such
a model); so is an assumption that school language is
exactly like any home language. The development of
childrens understanding of the demands of school is
one of the growth areas for research at present (see
Chapter 6). So is the search for explanations of
childrens school failure which examines the
microsystem of the school for causes as well as the
possible deficiencies of the child. So is the study of
what language is used for, the subject which we now
take up.
The functions of language
I mentioned the functions of language in the first

paragraph of this chapter, and they have been implicit
in all that has been said since. The list of language
functions which Michael Halliday (1975) observed
in his young son provides an example of the range of
uses which childrens language can serve. A very
early use is language in an instrumental function, a
way in which the child satisfies his or her needs or
wants; early language is also often regulatory,
controlling others behaviours, or interactional, used
for establishing or maintaining interpersonal contact.
Children also use language in a personal function, to
talk about themselves, in the heuristic function, to
find out about the world, and in the imaginative
function of lets pretend. Using language in the
informational function is a relatively late
development, but achieved well before pre-school.
Initially language functions are simple, but they
become more diverse and more are involved in any
given utterance as people become more sophisticated
speakers. Describing the function of the language
being observed becomes problematic to a
considerable degree.
The work of philosophers such as Austin, Searle
and Grice on speech acts (see, for example, Searle
1969) has provided important insights into how to do

things with words. One important distinction is
between different levels of significance in an
utterance such as the teachers Somebodys making
a lot of noise. The first level is the locution, the
linguistic form itself, in this case a possible neutral
statement. The second is the illocutionary force,
which is the type of speech act the speaker intends,
here that the class should quieten down. The third is
the perlocutionary force, the significance the hearer
finds in the utterance, here perhaps that Miss wants
everybody to be quiet or Sir is making a fuss again.
Here we are mainly concerned with types of
illocutionary forces, the functions which the speaker
intends the utterance to serve.
Speech act theory provides definitions of speech
acts such as requests, promises, threats and so on. For
example, both promises and threats refer to future
acts by the speaker, but promises refer to acts which
will be of benefit to the hearer and threats to acts
which will not. Children have to learn how to produce
and comprehend speech acts in ways which are
appropriate to their social and linguistic community,
to the microsystems in which they take part.
Problems may arise if the conventions for speech act
usage differ between microsystems, such as the
home/school differences I described above. A child
who is used only to very direct speech acts, such as
Shut up, may not see the illocutionary force of
indirect ones, such as Someones making a lot of
noise. Differences will apply at the level of single
utterances and of whole long discourses: Heath
(1983), for example, describes several different
understandings of what stories are and how to tell
them.
Some examples of childrens use of speech acts and
of their awareness of the conditions that underpin
them can be found in books and papers such as
Garvey 1975, Ochs and Schieffelin 1979, Bates and
MacWhinney 1979, Shatz 1983. Olson (1980a)
provides a particularly nice example of the variety of
social forms language can serve in the course of an
interaction, in this case distributing dominoes equally
between the two nursery-school children involved.
135

L: Lets make a domino house out of these.
J:
Okay.
First by grabs.
J:
Lookit how many I got . . . . You took a couple of mine!
L: Now you took a couple.

Then by commands.
L: Now you got to give me three back!
....
L: Now give me just one more and then we got the same.
And then by requestful assertives.
J:
Now, you got more than mee.
And denials.
L: No! We got the same.
By fact collecting, assertions, and inferences.
L: [Begins to count her dominoes] One, two, three, four . . .
twenty-eight, twenty-nine. [Then counts Jamies dominoes]
One, two, three, four . . . eighteen, nineteen . . . [short pause]
twenty-nine.
J:
I got nineteen and you got twenty-nine. . . . You got more

than me.
L: Noo [shouting] I COUNTED. . . . You have the same as me . .

. We got the same.
J:
NOOO!
And when negotiations b reak down again, by grasping.

[There is a shuffle of dominoes across the floor and now Jamie has
more than Lisa.]
And finally, by appeal to authority.
L: You got much more than me now.
J:
Now we got the same [Paul, a volunteer teacher, enters the

room.]
L: Does he have much more than me?

P: Not too many more!
(Olson 1980a, pp. 956).
136
Some examples of these formal links are the

replacement of nouns by pronouns on second and
subsequent occurrences, or the use of items such as
therefore, meanwhile and in addition. These
devices refer back from the present word or phrase to
earlier ones which it is not necessary to repeat
verbatim because they can be understood from the
whole linguistic context that has gone before. My use
of these and they in the last three sentences are
examples.
Very small children may merely string sentences
together without using syntactical linking devices,
but some instances of grammatical devices such as
pronominal reference are produced and understood
by pre-school children (Ervin-Tripp 1978;
Karmiloff-Smith 1979). McTear (1985) describes the
use of discourse connectors such as well, anyway
and so forth by his daughter Siobhan and her friend
Heather. Over the period of recording the two little
girls more such devices emerged and their discourse
became both more flexible in structure and more
continuous in theme. Devices for initiating, restarting
and repairing conversation also became more varied
and efficient. The children were able to monitor their
own speech and the speech they heard and to take
appropriate action to repair conversation and adjust
its course so as to achieve the social interaction they
desired. By the time they reached school-age Siobhan
and Heather were effective as a conversational duo
and in their talking with familiar adults. Differences
of conversational style between these children and
adults were mainly of degree for example the adults,
who had more world knowledge and more power,
would use more indirect speech, innuendo and
oblique references though adults, of course, have
more practice in specialized types of discourse such
as counselling, chairing meetings or seduction
routines. As they get older, children too learn more
about how to use language to get the result they want
(e.g. Clark and Delia 1976; Romaine 1984).
The picture of childrens conversational
competence which is emerging from research at
present shows less of the incompetent egocentric
language use which Piaget (1959) described and
more of children using a variety of conversational
devices in orderly ways similar to adults usage. By

school-age they have shown many complex skills:
later conversational development is predominantly
through increased world knowledge, better social
cognition and mastery of rare linguistic forms. The
integration of literacy and thinking and the world of
school are perhaps the crucial problems and
possibilities.
will often switch to the correct form. The following

dialogue is a good example of how the two forms flit in and
out of consciousness in the course of natural conversation:
Dan:
Hey, what happened last night after we left?

Did Barbara [the baby sitter] read you that
whole story? Remember you were reading
Babar?
Heida:
Yeah . . . and, um, he . . . she also . . . you know

. . . mama, mama, uh, this morning after breakfast, read the whole, um, book of the three little
pigs and that, you know that book, that . . .
Heida:
I dont know when she readed . . .
Dan:
You dont know when she what?
Heida:
. . . she readed the book. But you know that

book that green book that has the gold
goose, and the three little pigs, and the three little bears, and that story about the king?
Dan:
M-hm.
Heida:
Thats the book she read. She read the whole,

the whole book.
Dan:
Thats the book she readed huh?
Heida:
Yeah . . . read! [annoyed].
Childrens metalinguistic behaviour
In Chapter 3 I discussed work on metacognition,

childrens awareness of and control of their thinking.
There is a similar concept in the work on language
development, childrens awareness of and control of
their language. As language has many different
levels, metalinguistic competence has many
components too and develops over a long period of
time. Intuitive judgements about the sameness or
difference of consonants such as p or b can be
traced back to infants discriminations (see Aslin et
al. 1983, and Chapter 2 this volume), but picking the
odd word out in a set containing contrasting items
such as pat, bat, bee and boy is hard for
children beginning to learn to read (eg Bryant and
Bradley 1985).
Observations of pre-school children show use and
awareness of different sorts of language, for example
the child aged 2:10 who was recorded as saying
When I was a little girl I could go geek-geek like
that. But now I can go this is a chair, or the children
who refuse to accept from other people the incorrect
pronunciation or construction which they themselves
use (Clark and Clark 1977; De Villiers 1979). Slobin
1978 provides an example of how flickering
childrens awareness of grammar can be at this age:
Overgeneralizations planted in adult speech elicited protest
from Heida only if the standard form happened to be
momentarily present in her consciousness:
25 (4;7). If she has just used the correct past tense of an
irregular verb, she is annoyed with me if I respond to her
with the overregularization; but if she has used the
overregularization, she does not object to my following
suit. If I follow her incorrect form with the correct form, she
[digression of about one minute]
Dan:
Oh.
Heida:
Dum-dum!
Dan:
Barbara readed you Babar?
Heida:
Babar, yeah. You know, cause you readed some

of it too.
[brief interlude about dressing]
Dan:
Well I just started it.
Heida:
Yeah. She readed all the rest.
Dan:
She read the whole thing to you, huh?
Heida:
Yeah . . . nuuh you read some.
Dan:
Oh, thats right; yeah, I readed the beginning of

it.
Heida:
Readed?! [annoyed surprise] Read! [insisting

on the obvious].
Dan:
Oh yeah read.
Heida:
Will you stop that, papa?
Dan:
Sure.
(Slobin 1978, pp. 523).
137
Heida can monitor her language (and her fathers)

and correct it, but does not do so consistently. The
pressure to communicate may have overcome the
pressure to be grammatical; adults commonly accept
ungrammatical utterances which are intelligible (see,
for example, Wells 1982).
Segmentation is an important part of
metalanguage. Segmenting words into sounds seems
to be difficult for most children under about 7 and is a
particular problem for poor readers (Bryant and
Bradley 1985).
Training children to focus on sounds increased
their metalinguistic awareness and their
performance: in the case of the children Bradley
trained it improved their reading. There are similar
difficulties in segmenting sentences or utterances
into words; again, childrens ability to do this
improves at around 6 or 7, and it seems likely that
language awareness and reading practice influence
each other. Ideas about words often confound the
word itself and its referent: thus book is a long
word because it has lots of letters in it, or
primrose a short one because primroses are small
(Berthoud-Papandropoulou 1978). Words such as
the are not proper words: alternatively a word
given as response to the question tell me a word
might be a complete sentence. Again, reading is
associated with understanding segmentation into
words. Experience with written texts probably also
helps children to deal with unusual grammatical
constructions such as over and over rolled the ball.
Awareness that what is said and what is meant are
not necessarily the same is another fairly late
achievement (e.g. Olson and Torrance 1983). For
example, in one of Olsons experiments children
were read a story about two children sharing some
popcorn and then arguing about the distribution. One
says to the other You have more than me! 5-yearolds asked what was said are likely to reply that the
child in the story said Give me more. 7-year-olds
report verbatim and can show they knew what was
meant as well. In another study children of this age
did not manage to deal with literal and expected
meaning quite so well. Here a Sesame Street
character said Im going to divide this banana up so
138
both of us can have some and then ate the whole

banana and gave his partner the skin. It was not until
at least 8 that children realized that this division did
fulfil the literal meaning of the sentence, though not
the promise it implied. No doubt sad social
experience contributes to understandings of this sort.
Childrens handling of ambiguity is another source
of evidence on their metalinguistic awareness. Again
there are various sorts of ambiguity. It can be due to
lexical double meanings (such as pipe) or syntactic
ambiguities (she hit the man with the glasses, he
told her baby stories). Childrens jokes often exploit
ambiguities of sound, vocabulary, syntax or
expectation. Gleitman and Gleitman (1978, p. 118)
provide some examples:
Sample jokes, classified in terms of the source of ambiguity
A: Phonological
1. If you put three ducks in a box

what do you have? A box of
quackers.
2. Bob coughed until his face
turned blue. Was he choking?
No, he was serious.
B: Lexical 1. How
can hunters in the woods

best find their lost dogs? By
putting their ears to a tree and
listening to the bark.
2. How do we know there was fruit
on Noahs ark? Because the
animals came in pairs.
C: Surface Structure 1. How would you run over a dinosaur? Id start at his tail, run up
his back, then over his neck
and Id jump off.
2. Where would you go to see a
man-eating fish? A seafood
restaurant.
D: Deep
Structure
1. Were going to have my grandmother for Thanksgiving. You

are? Well, were going to have
a turkey.
2. Will you join me in a bowl of
soup? Do you think theres
room for both of us?
E: Morpheme
BoundaryNo
Phonological
Distortion
1. Why can one never starve in the

desert? Because of the sand
which is there.
2. How do trains hear? Through
their engine ears.
F: Morpheme
Boundary
with
Distortion
1. Do you think that if I wash, my

face will be clean? Letssoap
for the best.
2. Did you read in the newspaper
about the man who ate six
dozen pancakes at one sitting?
No how waffle.
Understanding ambiguity, appreciating verbal

jokes which use it, and developing the use of
metaphor (Reynolds and Ortony 1980; Kogan 1983)
are all aspects of language awareness which develop
during the school years. All of them both require and
make possible greater world knowledge and social
understanding.
Word meanings and concepts
Words refer to concepts: concepts are labelled by

words. There is an intimate interaction and mutual
influence between cognition and language. I want
here to look briefly at two aspects of this interaction:
first the childs early use of concepts and of
conceptual language, and then the question of how
literacy affects language and thinking.
Early concepts
Peoples conceptual knowledge may be much more

varied than their words for expressing it; for example,
there are many more discriminably different colours
than there are colour terms, and the cultures
vocabulary of colour words is not a major
determinant of accuracy of colour discrimination,
though it does affect ease of labelling colours (Heider
1971, 1972; Rosch and Lloyd 1978). We do not have
good descriptions of complete concept systems,
though there are some interesting examples of
attempts to map out some limited areas such as a 4year-olds knowledge of dinosaurs (Chi and Koeske
1983), and patients views on the significant people
in their lives drawn up using Repertory Grids (e.g.
Fransella 1976). Young children doubtless have

simpler concept systems, but it requires considerable
ingenuity to elicit information from them.
Gelman and Baillargeon (1983) and Clark (1983)
review some of the recent work on the beginnings of
categorization and concept labelling. Both the
habituation evidence from babies which I discussed
in Chapter 2, and some evidence on very young
childrens manipulation of categorizable objects,
suggest that the roots of categorization lie in the first
year of life, before the child is producing language.
Childrens sorting behaviour can involve a
significant amount of choosing objects consecutively
from the same category before 18 months old, even an
exhaustive search for a category, but at that age, and
for a year or two to follow, sorting is in competition
with pattern making, and the child may move freely
from picking out items of a particular colour to
arranging items in a pleasing pattern. This oscillation
between interests leads to the apparently haphazard
and conceptually confused collections and chain
concepts which Vygotsky (1962) and others
described. Markman, Cox and Machida (1981) report
an experiment where children sorted items into
plastic bags instead of on a table top: this procedure
made arrangements impossible and increased the
frequency of logically based sortings. Thus there is
some categorization of objects round about the time
that children are producing their first words and
sentences. Clark (1983) points out that early words,
which typically first apply to a prototype object, are
quickly used for all objects of that kind, not just the
original prototype. Words such as dirty are applied
to objects which are dirty in different physical ways
but share the conceptual characteristic of being
something you shouldnt touch, something taboo or
literally likely to make you dirty. At this point in
language use we find the over-extensions I
mentioned earlier, and also probably various other
less conspicuous mismatches between word and
referents. Children are already using sets of
properties to define which objects are members of a
category, and over-extended uses do tend to involve
some similarities in appearance or other properties
with the prototype instance. They try out different
139
assignments of objects to categories, picking up the

communitys conventional labels and judgements of
same or different. Success in communication is an
important criterion for the continued use of a
conceptinstance match. Children assume that there
will be a word for a concept which is consistent from
one instance of that concept to the next and which will
also contrast with the words that belong to contrasting
concepts. From very early on they ask for adult forms,
try them and repair their own word choices (Clark
1983, p. 805). The goal is to fill lexical gaps by
finding words for concepts the child wants to talk
about. Gaps may be filled by finding the right word or
by temporary measures such as over-extensions or
the use of general purpose words such as that,
thing, do or go, though these tend to be contextdependent. Children also coin new words, frequently
using the standard devices of the language. In
English, for example, nouns can be converted into
verbs or vice versa, e.g. trumpet, attempt; affixes are
used, e.g. garden gardenerunder-gardener; and
there are compound words, e.g. darkroom, taxpayer.
We need more data on childrens word coinage, but
they certainly use conversion (Can you needle my
shirt; Im gonna gun you), and suffixes (where they
prefer the more frequent -er to -ist), and make
compound words (a smile-person, a knock-thing).
Appreciating the relationship between concepts
and the overlapping of different words referents
requires a considerable amount of world-knowledge.
In the absence of crucial knowledge, using a
superordinate category may be impossible. Gelman
and Baillargeon (1983) give as example some
unpublished work by Susan Carey on childrens
concepts of animals. Characteristics such as breathe,
think, have bones, were recognized as animal-like
properties, and children recognized that mammals
had most of these defining characteristics; then, in
order, birds, insects, fish and worms, were seen as
being the least animal like. However, young
children did not know accurately which animals had
which characteristics and did not reliably judge one
property to be more crucial than another. This
ignorance would lead inevitably to classifications
which are biologically ill-founded. Ignorant adults
140
make the same sort of mistakes; both adults and

children may be deceived by words which are similar
into believing that the objects they refer to are similar
too. Adults use of category hierarchies appears to
influence childrens learning of superordinate
categories (Callanan 1985).
Literacy, language and thinking
I have emphasized throughout my discussion of

childrens cognition and language that they develop
within a social context. Particularly once children
attend school, literacy is an important part of this
context. There are strong arguments in the literature
that language and cognition skills are necessary for
the development of literacy (notably awareness of
language concepts, and awareness of language
sounds, see p. 78) and also that literacy changes
peoples use of language and ways of thinking. It is
this latter possibility that I want to concentrate on
here.
There have been two main bodies of evidence for
the assertion that becoming literate changes the ways
in which people think and use language. One body of
evidence centres on the marked change in childrens
thinking and talking which Piaget and many others
have observed to happen during the early school
years. An important part of this change is a growing
ability to reassign descriptions or categorizations to
the same object or event. This is seen as being related
to the childs learning to read, an achievement which
requires systematic thought about the alternatives
and categorizations possible in language. Reading
also opens up systematic access to the stored
knowledge of the culture, and is thus what Bruner
(1966) called a cultural amplifier. Bruner sees the
cultures supply of amplifiers and the demands that
life in the culture makes on an individual as being
crucial determinants of the powers of mind which
will develop. For example, our society demands
(among many other things) that people go through a
formal educational process. Becoming literate is
essential for coping with this demand; literacy is also
one of the cultural amplifiers which the educational
process offers, not just in school but in the use of
books and other writing and reading outside school.
A similar suggestion comes from the second body

of evidence which relates literacy, thinking and
language. This consists of historical and
ethnographic studies of the consequences of literacy
or other social changes (e.g. Goody 1977; Luria
1976). In societies where literacy is newly
introduced, those who become literate become able
to do new thinking and language tasks rather as
children becoming literate do in a culture where
literacy is already established. Among the areas in
which change has been claimed (for both children and
literate members of developing societies) are
deliberate remembering, logical and scientific
reasoning, and various aspects of language use more
characteristic of writing than of speech (see Goody
and Watt 1968; Olson 1977; and Chapter 3 this
volume).
There is debate about whether the consequences of
literacy are general cognitive changes which extend
far beyond reading and writing to other areas of
language and thought, highly generalizable cognitive
operations that are responsible for intelligent
behaviour; or whether they are specific changes
centering on literacy tasks and extending not terribly
far from the business of reading and writing.
Although the issue is very complex, the present
emphasis seems to be on the specific possibilities
which literacy allows rather than on literacy as a
general cultural amplifier which has strong effects
even when literate skills are not being directly used.
For example, Cole and Griffin (1980) describe
experiments which show that most normal adults
have the ability to correlate pieces of information and
make correct inferences, whether they live in literate
societies or not. What makes literate adults rather
better at such tasks is that they use writing and devices
such as tables of instances to help them remember the
information; if they are not allowed to do this, they

forget information and so reason no better than
illiterates. Similarly, provision of record-making
materials lifted the performance of primary school
children to the formal operations level of systematic
hypothesis testing on Piagets science tasks (e.g.
Brainerd 1978).
Further evidence that literacys effects may be
centred on the skills of literacy comes from Scribner
and Cole (1981). They studied a society which has
several different systems of literacy. The Vai people
in Liberia use a syllable-based script, mainly for
letter-writing and recordkeeping: this script is
learned at home. It is useful but not necessary for the
Vais traditional employments as rice farmers, smallscale entrepreneurs and craftsmen. About 20 per cent
of Vai men use Vai script. Some people have attended
American-type schools where they learn to read and
write in and through English. There is also an Islamic
influence; some Vai read Arabic, mostly just to
decode the Koran and read it in religious services, but
some also use it to write letters and records and to read
commentaries on the Koran. Literates of these four
different sorts were tested on a variety of skills based
on analysis of literate practices, such as coding and
decoding symbols, recalling information, playing a
communication game and solving logical syllogisms.
Test performance was very closely related to the
specific functions of the particular literacy used by
the subject. They argue that the apparent general
effects of literacy are related to what it is used for and
how it is taught. Cultural amplifiers are embedded
in the social context as language and cognition
themselves are. Practice at meeting the demands of
the microsystem makes perfect on the skills the
microsystem requires: not necessarily on the skills it
does not.
141
Plate 12 The Castlemilk lads

142
5 Personality
Just as there has been a tremendous variety of ways of

understanding and investigating personality itself,
there have been many different approaches to its
development.
Developmental
studies
have
encountered all the conceptual and methodological
problems of general personality theories, plus
difficulties associated with the interpretation of
continuity of character or behaviour over time.
Theories of personality development differ in their
basic ideas about the origin of personality, about its
consistency, and about the importance of life events
and possible series of developmental stages. This
chapter looks first at theories which place more
emphasis on permanent dispositions, next at some
stage or life-event approaches, and finally at the
developmental evidence for certain aspects of
personality.
Physiology and personality

The belief that people differ in general character, that
their approach to different situations is, say,
consistently active or lethargic, melancholic or
optimistic, is a very ancient one indeed. So is the
attribution of such differences to an underlying
physiological bias. Such ideas can be traced through
millennia, surfacing strongly in the nineteenth
century in phrenology and physiognomy (Gould
1984, in a vivid description of this work and of early
intelligence testing provides a salutary warning to
psychologists about the social implications of their
research). They appeared again in this century in
Kretschmers and Sheldons work on the link
between body-build and character; and might be seen
to be related also to Eysencks accounting for
introversion extraversion and neuroticism in terms
of the way the central nervous system works (e.g.
Eysenck 1967). So far as the early theories were
concerned, it was a case of anatomy is destiny; for

Sheldon, if you had a round, soft, plump sort of body
you were destined to be a comfort-loving, relaxed,
sociable sort of character, just as for Lombroso if you
had ears which stuck out or unevenly set eyes you
were almost bound to go in for one or the other sort of
criminal activity. Personality was seen as genetically
fixed, and as running in families. Jukes and Kallikaks
bred true, and social reform depended heavily on
eugenic breeding strategies.
These were seductive theories, corresponding
plausibly with common-sense everyday experience,
and they had a long run. However once careful and
large-scale investigations were carried out, the
correlations between body-build and character, let
alone between skull shape and character, turned out to
be too small to be much predictive use. The direction
of causation of any such relationship is, in any case,
likely to be unclear. The ectomorph, long, thin and
weedy in build, with (less measurably) a highly
sensitive nervous system and a low threshold for
sensation, is unlikely to be easily competent or skilled
at vigorous competitive athletic activities, or at
settling disputes by physical means, and so may
indeed concentrate on intellectual and symbolic
activities. Conversely, however, avoiding strenuous
exercise will affect how muscular the body becomes
and indeed how efficiently heart, lungs and so forth
work. Only if there is a self-induced experiment in
the form of a change of lifestyle can we know if bodytype determines personality or vice versa. We know
very little about how much changes in each are
possible, or affect the other. People who change the
amount of exercise they take quite often claim that
they have experienced changes in psychological
areas too: systematic study would be useful, here. So
would studies of whether round, cuddly babies are
treated very differently from hard muscular babies;
143
given the apparent difference in both description and

handling that boy and girl babies receive, it would
seem quite possible that subtle differences in
experience amplify initial small differences in
physiology.
This sort of interaction, and these sorts of
problems, no doubt also apply to Eysencks model
based on differences in the central nervous system.
However there is little evidence about the continuity
of extraversionintroversion over childhood (it
seems to be fairly consistent in the adult years) and
neuroticism, at least in terms of the prognosis from
childhood depression or anxiety to similar states in
adulthood, is not particularly consistent (Barker
1979). The theory is not entirely clear about the
degree to which the balance of the central nervous
system is affected by or determined by
experience. Again, answers to these questions would
be very useful.
Thus there is not at present an adequatelygrounded model of the relationship between physical
and psychological characteristics in general
personality. There is work on relationships between
variation in physical characteristics such as hormone
level and differences in aggressive, nurturant and
other behaviour which we will look at later. One
important and very interesting body of research has,
however, continued to look at continuities in
personality over childhood. Among the workers
involved in studying temperament have been
Thomas and Chess and their colleagues (e.g. Thomas
and Chess 1977; Buss and Plomin 1975; Halverson
and Waldrop 1976; Kohn 1977). Dunn (1980) and
Berger (1985) discuss the methodological issues
involved, and a CIBA foundation symposium (CIBA
1982) reviews recent work.
Temperament
Temperament is seen as continuities of
characteristic style of behaviour which can be
observed and rated in infancy, and pervade all that a
person does. Temperament interacts with experience
to produce personality. Various temperament
dimensions have been suggested, among them
144
activity level, general mood quality, how rhythmic or

predictable behaviour is, intensity of reaction,
approachwithdrawal,
distractability
and
fastidiousness. There is a tendency for clusters on
these variables to be found which Thomas et al. label
as easy, difficult and slow to warm up general
types of temperament. Easy babies are positive in
mood, regular and predictable in behaviour,
moderate in their activity and reaction, and highly
adaptable to changes. Difficult babies are intense,
negative, irregular and slow to adapt. Slow to warm
up babies are inactive, withdrawn and slow to adapt.
These characteristics are said to be highly stable over
short periods of time, though continuities decrease as
the time interval increases from babyhood to
childhood. It is also claimed that early temperament
differences are predictive of later psychopathology.
Children whose early temperament is irregular,
unadaptable, negative, unfastidious and highly
intense are particularly likely to be referred for
psychiatric treatment in childhood or early
adulthood. Children who are slow to warm up may
also have later problems, but these are likely to be of
withdrawal and apathy, where the difficult children
are more likely to have more vigorous and disruptive
behaviour problems (Rutter 1985a; Kohn 1977).
Temperament is generally assessed by means of
detailed reports of the childs behaviour elicited from
the mother or some other adult who knows the child
well. Such reports are generally creditably accurate
about present behaviour, though their accuracy in
looking back over a period of months may be more
doubtful. There are however more serious
methodological problems. There may well be a quite
marked relationship between baby differences,
caretaker differences and continuity. Attentive
mothers, for example, tend to have more alert babies;
there are continuities in the mothers contribution to
interactive style which may help to produce
continuities in the childs (Ainsworth 1979; Dunn
1980). While temperament is often continuous in a
stable environment, a change in caretaker, as in
adoption, is likely to lead to discontinuities in the
childs behaviour (see Chapter 6, and also Tizard
(1977) on late adoption, which is discussed
Personality
elsewhere). Similarly apparent temperament varies

in interaction with different members of the family:
that is it has a substantial interactive or relational
component. Finally, difficulties arise over the
meaning of behaviour and the assessment of
continuity over time. To take an obvious example,
crying on encountering a difficulty might be seen as
adaptive for a baby, for whom it summons help, less
so for a 5-year-old and actually maladaptive for a 10year-old. Similarly, longitudinal studies show up
subtleties in the patterning of continuities over time.
Halverson and Waldrop (1976) found considerable
stability in their general activity dimension
between 21/2 and 71/2 years. At 21/2, ratings of very
high activity play were associated with individual
differences in social participation and impulse
high activity. However these two latter variables
were associated with quite different groups of
behaviour measures at 71/2. High social participation
21/
at
2 was positively correlated with social
participation at 71/2, with high verbal IQ and with
high field independence. High impulsivity of action
had negative implications for intellectual
development. Possibly this difference might be
related to the different demands made on children at
the two ages: a child of 71/2 is expected to be able to
control its impulses and fit them to outside
constraints such as classroom requirements of
sustained concentration and planful activity. A
cognitive style characterized by high impulsivity
has often been seen as militating against academic
achievement, and impulsivity might in excess (or in
extreme lack) also have social implications.
These subtle relationships between characteristics
over time are further complicated by the existence of
differential socialization. Parents try to encourage
ways of behaving in their child which will make the
child better able to cope with the world. One
interesting example though ethnographic rather
than the sort of work on which we are concentrating
here is found in Heaths (1983) account of language
use in different Appalachian communities (see
Chapter 4 this volume). Another, older, example in
one of the pioneering studies of temperament and
socialization, is Kagan and Mosss (1962) findings

on sex differences in dependence and aggression. In
their longitudinal study of Midwestern children
through the 1950s they found that the degree of
dependence in infancy predicted to dependence in
childhood and young adulthood in girls but not in
boys: that is a highly dependent female baby was
likely to be highly dependent as an adolescent girl,
while a highly dependent male baby was not likely to
be particularly dependent as a boy or youth. The
reverse was true for aggression, which was fairly
continuous in relative level in boys, but
discontinuous for girls: aggressive little girls, and
dependent little boys, were taught to change their
behaviour towards the social stereotype of the
passive nurturant female and the go-getting
competitive male.
Stage models of personality

Whatever there may be in terms of personality
differences innate in the person as temperament
was first supposed to be they are not easily
measured or indeed conceptualized, and they will in
any case interact with life stresses and the day-to-day
experience of less extreme events in very complex
ways. We do not know exactly how this happens, but
general models of personality which emphasize
development through a series of life-events have
been put forward. Many have Freudian ancestry,
many see life as a series of crises that have to be
overcome but leave their scars on the developing
personality. Eriksons eight stages (Erikson 1963)
provide an example of this approach. Each stage is
dominated by a conflict which has to be resolved.
Thus, in the first stage the infant has to develop a
sense of basic trust that the world is basically
benevolent and trustworthy: the mothers reliability
and sensitivity (especially to the babys oral needs)
are crucial to this achievement, and if they fail the
infant will have a sense of basic mistrust. The second
stages conflict is between autonomy and shame or
doubt; developing control of anal reflexes and
learning when to give and when to withhold faeces
are the crucial events in this second year. Later
145
conflicts include establishing a secure sense of

identity rather than a sense of role diffusion (the fifth
stage, adolescence), and resolving the contradictions
between intimacy and isolation (the sixth stage,
young adulthood).
As with other stage theories, Eriksons eight are
presented as a series and appear to be discrete and
relatively self-contained, though a poor outcome to a
particular stage-conflict will lead to problems at later
stages. What evidence there is on the existence of
these stages comes from clinical material or from
highly interpretive accounts of data which may not be
much more reliable (Block 1971; Levinson 1978;
Vaillant 1977). The conflicts themselves may have
particularly acute periods at particular stages, but the
difficulties they involve could equally well be
thought to be constituents of human social life, at
least in its modern western form. There is room for
doubt over whether some of the stages appear at all in
other social traditions, though Erikson regarded them
as universal, and over whether they are common even
in western experience. The identity crisis of
adolescence, for example, appears to be Eriksonian in
only a small minority (Coleman 1980). Nevertheless,
Eriksons emphasis on the different demands that are
made on the developing child at different ages is a
useful one. It raises two sets of questions: first, what
are these demands and what stresses do they set up,
and second, how far, and why, do the present stresses
have links to past and future? There has been some
interesting work done on the links between the
demands parents make on their children, including
how these are made, and the outcome in terms of
childrens personality and behaviour. This detailed
work is discussed in the next chapter.
Self-concept development
Parental behaviour less explicitly specified has been
seen for some considerable time as one of the most
important sources of the self-concept. G. H. Mead in
his classic book Mind, Self and Society (Mead 1934)
says that via social interaction the young child begins
to appreciate that other people (notably, parents) have
views of him or her as good, bad, clever, a real
146
boy, big for her age and so forth. These views are
inferred from others behaviour towards him or her as
well as from their talk, and are accepted as evaluative
and categorical labels much like names. It is from
these labels applied by others that the child builds up
his or her self-concept a looking-glass self.
More recent theorists paint a slightly different
picture. Lewis and Brooks-Gunn (1979) make
Meads account part of a wider development. They
describe two rather different aspects of the self. The
subjective aspect, the existential self, is centrally
the distinction of oneself from others, the awareness
of the me who is acting, experiencing, remembering
and so forth. It probably involves the sensory self
which may have a neurological base (Konner 1982)
but it requires a sense of self-permanence which
Lewis and Brooks-Gunn see as analogous to object
permanence, and also an accumulation of learning
about the patterning of actions and outcomes. As we
have seen in discussing infancy, the rudiments of a
distinction between self and not-self seem to become
apparent increasingly clearly in the babys first six
months; by the last quarter of the first year
distinctions between self and other are becoming
independent of specific actions and contexts so that
the child could be said to have a permanent
existential self.
From here on, the second aspect of the self
develops too. This is the objective aspect, the
categorical self. This aspect refers to the
characterization of oneself in terms of categories like
gender, age, competence, attractiveness etc. The
categories used may change between cultures or
historical periods or may be universal, and may
change over an individuals lifetime or remain
constant. To give an example: my own categorical
self would include the following categories, all
relative and not in any order of importance: tall,
which appeared early, will remain constant and could
be universal; female, also early, universal and
constant, although the defining attributes and
connotations of female have undergone historical
and cultural changes; distrustful of technology
which appeared late, is clearly not historically and
culturally universal and possibly ought to be
Personality
changed, in view of what I shall have to say about the

effects of self-fulfilling prophecies! Such selfcategorizations develop from babies understanding
of their own actions and from their use of others
categorizations of them, and language and other
symbolic systems play an increasingly important
part. During the second year, the self-concept
develops rapidly: Kagan (1981) calls this the period
of the emergence of self-awareness. Verbal selfreference and use of pronouns you and I begin
here. Research (e.g. Bertenthal and Fischer 1978) has
identified landmarks in the process: from 10 months
on the child who sees in a mirror a moving obiect
whose movements are contingent on his or her own
begins to be able to use the information in his or her
reflection to locate the object; from about 18 months
old, seeing in his or her reflection a trace of rouge
which had earlier been surreptitiously put on his or
her face, the child reaches out to touch not the
reflection but the rouged spot on his or her own face.
This achievement requires an image of what ones
own face normally looks like and a recognition and
location of the discrepancy. Gallup (1977, 1979) has
demonstrated that given some hours of
familiarization with mirrors, primates such as
chimpanzees could similarly recognize themselves,
but (so far) animals such as macaques, baboons and
gibbons could not. Interestingly chimpanzees reared
in social isolation did not show self-recognition. This
is what Meads account of self-concept developing
through social interaction would predict, and it might
be seen as analogous to the self-reports of humans
who grew up with little social experience (Hartup
1978), but we really need a great deal more
investigation of the competences and pathologies of
social isolates before we can claim to understand the
role of social interaction in the development of selfconcept.
The mirror recognition experiments demonstrate
young childrens interest in observing their own
actions and in discrepancies such as an unexpected
rouge-coloured spot. Kagan (1981, 1984) points out
that discrepancy becomes a preoccupation in the
second year; things which are broken, or not as they
should be, are enthralling and often distressing.
Distress at something that is broken or dirty might,

of course, be seen as a conditioned result of the
scolding, slaps or cold adult faces which accompany
the childs breaking or soiling of objects (and toilettraining is likely to be part of the childs life at this
age), but it could also be related to childrens distress
when their self-categorizations are teasingly denied.
Once children have learned their gender, for example,
the discrepancy of saying that a boy is a girl may
evoke vigorous denial and upset: or may, in more
favourable circumstances, be a big joke. Dunn and
Kendrick (1982), presenting rich and fascinating
data, provide a nice example which illustrates a
childs play with categorization of herself, her teddy
and her baby brother (it also shows her use of an adult
label to describe herself, and a possible confusion of
the referent of you in her response to her fathers
first question).
Sally C:
Child (playing with her teddy) to father, F: Teddys a man.
F: What are you?
C: Youre a boy.
F: Yeah. What are you?
C: A menace.
F: Yeah, a menace. Apart from that are you a boy or a girl?
C: Boy (laughs).
F: Are you? Whats Trevor?
C: A girl (laughs).
F: Youre silly.
(Dunn and Kendrick 1982, pp. 11011 )
Several authors have pointed out that gender seems

to be an early part of self-categorization, and as we
shall see in the next chapter it becomes a salient
aspect of childrens social lives. Relative age seems
to be another fairly early attribution, again salient in
social relations (Lewis and Brooks-Gunn 1979) and,
in Dunn and Kendricks data on the first-born childs
reaction to a new younger sibling, part of what appear
to have been extensive discussions of people. In two
147
examples, a child seems to have a triumphant sense

of always being ahead:
Laura W:
1) C to Mother (after M comments to Baby about cutting
teeth): I was cutting teeth. I was walking before he
was. I walked before him.
2) C to Observer: Hes a walloper. Hell smack me when
hes bigger. Im going to be huge when hes a bit
bigger. Up to the ceiling. Like you.
O to C: Im not up to the ceiling.
C to O: Well, Ill be up there. Ill grow so much. Up to
the ceiling. So high.
(Dunn and Kendrick 1982, pp. 1089)
Comparisons of self with peers and siblings seem to

be increasingly recognized in the literature as
important sources of the self-concept. Again, the
contribution of interactions with other children will
be discussed in the next chapter; but this new
recognition has implications for the concept of
identification which I will discuss here.
Identification
Although it is a crucial concept in several accounts of

the development of the self, the conscience and sex
role, identification has proved somewhat hard to
define and very hard indeed to measure. On the
whole, if A is to be said to be identified with B, A
must act, feel, think like B over a long period of time,
in many different situations, and not by superficial or
deliberate imitation but more unwittingly; and A
must strongly want to be like B. Identification is both
process and outcome; it is taking on a role in such a
way that the role becomes oneself. It is differentiated
from imitation, which is more likely to be
temporary, fragmentary and deliberate, to have less
emotional tone, although imitation can contribute to
the process of identification. (Certain sorts of
behaviour therapy and of actors preparation for new
roles illustrate this.) Simple overt similarity of
behaviour is not a reliable index of identification,
since similarity can result from similar experiences
148
(for example people who are slightly short-sighted

are often said to be aloof and not to greet their
acquaintances, whom they have not seen well enough
to recognize) or from similar genes as in the reports
of similar postures and gestures among twins
separated very young (Watson 1981).
Explanations of why identification happens have
come from two main groups of theorists; Freud and
his successors on the one hand and the social learning
theorists on the other. Both centre their model on
children identifying with their parents and
particularly with the parent of the same sex. One
motivation is to gain social approval and to avoid
punishment; another is to lessen the risk or pain of
losing the parent by becoming ones own parentsubstitute. Maccoby (1980) gives an example
supplied by Anna Freud:
A little girl, just two years old, had always been put to bed
by her mother, and there was a familiar bed-time routine.
For the first time, the mother was away overnight, and the
child was being put to bed by a baby-sitter. The child had
great difficulty going to sleep, and even though she was
very tired, kept her eyes open after she was tucked in and the
sitter had tiptoed out of the room. Through the open door,
the sitter hears the child say, imitating her mothers voice:
Goodnight my dearest (Maccoby 1980, pp. 1415).
Identification here provides comfort: it is also said to

reduce anxiety about loss of love by making it less
likely that the child will want to offend against
parental prohibitions, to provide definition and
models of the skills one ought to acquire, and to allow
for vicarious experience and understanding. It is seen
as especially important in the development of
conscience (see moral development) and of gender
identity and sex-roles. For both, identification with
the parent of the same sex is said to be crucial: if this
identification fails, the conscience will be weak and
gender identity confused. This has been an influential
theory to a wider audience than just professional
psychologists: but this is how Eleanor Maccoby
summarizes the research that has investigated it.
Personality
The research in child rearing that stemmed from this
tradition was imposing both in conception and productivity
. . . the importance of parental nurturance and of the way
parents exercise authority have been amply demonstrated.
Yet the yield of the work with respect to the theory of
identification was disappointing . . . no consistent
relationships were found among characteristics that ought
to have been linked by their common origins in the process
of identification (Maccoby 1980, pp. 1718).
If, then, the definition, measurement and effects of

identification are confused, it may be better to step
back from it as an explanation and investigate at a
more detailed behavioural level who identifies with
what in whom. Recent work on social relationships
looks at some questions bearing on this, and results
are discussed in the next chapter.
Beliefs about control of events

One aspect of personality which has received much
attention from developmental psychologists is the
extent to which people feel that they themselves are
in control of their lives and responsible for their
actions. It is suggested that feelings of being
competent or helpless will affect what people choose
to do, the emotions they have about what they are
doing, and how they tackle problems. Such feelings
have been conceptualized as locus of control (e.g.
Rotter et al. 1972; Nowicki and Walker 1974)
attribution of causation (e.g. Weiner and Kukla 1970)
and self-efficacy (Bandura 1977, 1981). Beliefs
about ones control of events are one likely source of
achievement on difficult tasks, for example academic
ones.
Locus of control
Locus of control is defined as a generalized

expectation of internal or external control of
reinforcement. People with an internal locus of
control believe that they are responsible for what
happens: their own effort brought about success, their
negligence led to failure. Someone with an external
locus of control would interpret success as being due
to good luck or the favouritism of powerful people,
and failure as due to bad luck or persecution by the

powerful. The external would be less likely, it is
argued, to work hard and effectively and therefore
achieve success than the internal would.
Most research studies on locus of control use
questionnaire scores as a measure of how internal or
external a subject is. There is some debate about
questionnaires adequacy as measures (Stipek and
Weisz 1981, and see also the general literature on
personality assessment). Nor is it simply the case that
locus of control leads to achievement: experience of
success or failure might also affect locus of control,
for example. A related point to this is that it is not
clearly a truer understanding of reality to have a more
internal locus of control: whether you get good
examination results or a university place, for
example, depends not only on your own efforts but on
public policy about what percentage of candidates are
to be put in each marking grade or how many
university places are to be available. However it does
seem likely that children with a more internal locus of
control do tend to show higher academic
achievement (Osborn, personal communication on
the findings of the Child Health and Education
Study), possibly because they manage their own
learning and their use of resources supplied by
teachers better than externals do.
Attributional models
An alternative model emphasizes situational

specificity in identifying the causes of events.
Attribution theory (e.g. Weiner 1979; Weiner and
Kukla 1970) examines characteristics of the person
and of the task or situation which may differ in how
internal, how controllable (that is, contingent on your
own actions) and how stable they are. For example,
how much effort you make is seen as being internal,
controllable and unstable; ability is internal,
uncontrollable and stable; task difficulty is external,
uncontrollable and stable. Attributing success to your
own ability and effort is said to result in greater
motivation to achieve, particularly if the task you
have succeeded on is a difficult one, and thus to
greater confidence in taking on further tasks.
Attributing failure to stable characteristics of
149
yourself, such as lack of ability, will be more

demoralizing than attributing it to unstable
characteristics such as lack of effort, or to external
uncontrollable factors such as bad luck or an
excessively difficult task. While most work on the
attribution theory model has been done with adults,
there is some evidence from children (e.g. Nicholls
1975, 1978). The model specifies more relevant
variables than the locus of control model and is
therefore less ambiguous in its predictions about
achievement behaviour and reactions to success or
failure.
Self-efficacy
Banduras work on self-efficacy (Bandura 1977,
1981) links the cognitive and motivational
components of attribution theory to the development
of social understanding. He is concerned with how
children come to think of themselves as efficacious,
and how they act on such a judgement. He sees four
principal sources of informaion on self-efficacy. The
first is childrens own accomplishment or
performance, their history of success or failure on
tasks and their attributions of causes for achievement.
The second is vicarious experience, seeing other
peoples success or failure similarly attributed. The
third is other peoples judgement of the childs own
efficacy: credible suggestions by others that he or she
is bound to fail may persuade the child not to try and
therefore not to achieve success. This is a form of the
self-fulfilling prophecy which may operate in
schools (see, for example, Rogers 1982). Banduras
fourth source of evidence is emotional arousal; if it is
very high, arousal probably debilitates performance,
and judgements of lack of self-efficacy may increase
arousal and thereby lead to the poor performance that
was expected. What children know about how their
emotional state affects their performance is also an
important influence.
How people combine these different sorts of
information, none of which is simple, is not as yet
understood. Bandura (1981, p. 210) suggests that
judgements of self-efficacy will tend to be more
egocentric and emotionally toned than objective and
150
dispassionate. Accurate appraisal is, however, much

more advantageous for effective performance.
Coming to such an appraisal is part of metacognitive
development.
The development of self-efficacy
Even infants seem to take a particular interest in

environmental events which are contingent on their
own actions, and to be better able to learn about
contingent events than non-contingent ones (see
Finkelstein and Ramey 1977; and Chapter 2 this
volume). Children seem to learn language best when
they experience child-contingent conversations with
adults (Wells 1985; and Chapter 4 this volume).
Comparisons with siblings and peers also give
information to the child about his or her competence,
though much more research is needed about how this
happens and what its effects are (Dunn 1984;
Bandura 1981).
School is a very powerful source of information
about academic achievements, including, it is
currently thought, influences which leave girls in coeducational groups less confident about their own
self-efficacy than boys are (Dweck and Elliott 1983,
pp. 6589). School practices may be associated with
not only childrens ideas about their own selfefficacy but also their sources of pleasure in their
achievement and their reactions to their failure. It has
been suggested (Harter 1981) that the intrinsic
interest which infants show in being effective actors
in their environments becomes overtaken by
extrinsic motivation such as teacher approval, gold
stars and the need to get good marks to get a job.
Children who are made to seem ineffective at
academic tasks may react by alienating themselves
from school values and setting up a counter-culture
(see, for example, Hargreaves 1967). Again, much
work needs to be done on exactly how self-efficacy
develops.
Aggression
In considering aggression it is necessary, as in other
areas, to point out the definitional problems. Olweus
(1979) offers this definition:
Personality
any act or behaviour that involves, might involve, and/ or to
some extent can be considered as aiming at, the infliction of
injury or discomfort; also manifestations of inner reactions
such as feelings or thoughts that can be considered to have
such aim.
Aggression thus defined would include some

possibly doubtful cases, such as the professional
activities of a dentist filling teeth and a judge
pronouncing a sentence of imprisonment, and
accidental injury, since can be considered as aiming
at follows an and/or conjunction. It also includes
feelings or thoughts which are not put into action, so
that, for example, hoping that the Prime Minister will
be ousted from office is aggressive in the ordinary
voter as well as in the plotters in Opposition. What is
more it does not specify who is to do the considering
that the behaviour aims at injury and discomfort: all
too often aggressor and victim have had different
ideas about this. Beating children, for example, has
been seen in various groups at various times as
necessary evil, moral duty, parental right and barbaric
brutality: so has not beating children.
Narrower definitions run into similar problems. So
typologies of aggression have been suggested
(Cook 1984), excluding predation and assertion and
distinguishing pro-social and anti-social, intropunitive and extra-punitive, instrumental and
hostile acts, and also the degree of aggression.
This would need to be considered, presumably, in
terms of intent in the aggressor, extremity of
behaviour, and degree of injury or discomfort
inflicted. Bearing these distinctions in mind helpfully
emphasizes the complexity of aggression, and they
do reflect distinctions which we commonly use in
judging the culpability of an aggressive action, but
how accurately any particular action can be classified
is very uncertain. As in the analysis of language, other
events not only before but after will affect the
classification, and different classifiers, with different
knowledge and approaches, may judge differently.
What was done to Steve Biko in the last few days of
his life illustrates the navety of believing that
different types of aggression can be simply and easily
distinguished. The policemen, prison officers and
doctors involved hardly saw themselves as being

aggressive (perhaps they saw Biko as not a member
of the human species and so their behaviour as more
like predation between species); in so far as they
admitted aggression, it was mild, prosocial and
instrumental, in that it contained and eventually
removed someone hostile to the social system. To the
liberals and opponents of apartheid both in and
outside South Africa, the same behaviour was
inexcusable and, in their terms, severe, anti-social,
hostile and extreme. The failure of those who caused
Bikos death to see anything extraordinary or
culpable in what they had done seems to the liberals
to be one further sign of their moral inadequacy.
Since children rarely if ever act like those South
African prison officers (or like the similar examples
which could easily be found in almost any days
newspaper) this example may seem to be an
extravagant digression from the subject of child
development. However the question of how far
childrens aggression is like adults is
unanswerable until we know a great deal more about
the causes, forms and functions of aggression than we
do at present. Physiologists and ethologists have
expended a great deal of effort on studies of
aggression in man and other animals, with results
which as yet just indicate the complexity of the
picture and elucidate a few details (Konner 1982).
Aggressive behaviour involves high activity in
particular
brain
structures,
notably
the
hypothalamus, the amygdala and the septal area, but
also circuits in the midbrain, and at least in mice there
are
psychopharmacological
changes
in
neurotransmitters. Slower-acting chemicals such as
the stress hormones and testosterone seem also to be
involved in aggression; at least in males and in some
species
more
testosterone
can
increase
aggressiveness, but also successful aggression
increases testosterone levels and unsuccessful
aggression decreases them. Differences in hormone
levels in the distant past as well as the present may
also affect the level of aggression shown: crossbreeding experiments show genetically based
differences in aggression in mice, dogs and bulls.
However, even in non-human animals learning,
151
experience and social structure have a great deal to do

with aggressive behaviour. Imitation, place in the
dominance hierarchy, and experience of social
isolation all affect aggression. To quote Konners
summary:
early rearing in social isolation [the animals discussed are
rhesus monkeys] . . . will produce a lifelong tendency to
social hyperreactivity, unaffected by the usual sorts of later
social experience. In males, such hyperreactivity
frequently results in a high level of threat, attack, and
fighting behaviour, often inappropriate and unsuccessful.
To a lesser extent, the same behavioral abnormalities occur
in rhesus monkeys that have been raised normally with their
mothers in the first year of life, but without contact with
peers. [In free-ranging rhesus monkeys] . . . high-ranking
females have female infants and possibly also males that
grow up to be high-ranking themselves and not just for
genetic reasons. Infants of such mothers are frequently
observed to imitate their mothers threat-and-chase
behavior, even in relation to adult animals. Obviously they
are not capable of defeating the adults in individual combat,
but the infants make their moves in the mothers shadow
even if she is not in the immediate vicinity a phenomenon
called by ethologists protected threat. In this context, the
infant has innumerable conventional learning experiences
and opportunities for imitation and social facilitation that
lead eventually to effective dominance behavior and high
rank. (Konner 1982, p. 199)
Thus aggression in animals involves genetic

proclivities, subtle chemical and electrical influences
in the central nervous system, and a history of
learning, particularly learning in social contexts. The
same could be said of human beings, with learning, as
usual, of enormous importance, and with the
existence of a genetically-based trait of aggression
less apparent than in animals which have been
selectively bred to increase, for example, their
tendency to attack a matadors red cape. In so far as
there is evidence of heritability of aggressive
offences it can account for only a small percentage of
the range of aggression found in the population, and
alternative explanations such as heritability of
152
alcoholism (which is associated with aggression,

social disadvantage etc.) cannot be ruled out (Wells
1980; Rutter 1980a). It is necessary to remember,
also, that aggression can take many forms:
precisely this point is made clear in observations of
developmental changes in childrens aggression.
If, as seems wise, intention to hurt, frighten or
distress is seen as a necessary condition for defining
an action as aggressive, children in their first year
are rarely aggressive, though they may be angry
(Lewis and Michalson 1983), and they may do things
that in fact hurt, such as biting the nipple or pulling
hair that comes within reach. Lacking the intention to
hurt, and also knowledge of how their actions affect
other people, they are not properly aggressive.
Bronson (1975), describing behaviour changes in the
second year, found that children showed increasingly
intense frustration and anger at having a favourite toy
taken from them or withheld. Tussles over property
become commoner, but 2-year-olds who hit their
opponent during such a disagreement often seem
surprised that the other child is hurt. Realizing that
ones actions can cause distress in others, and how
best to do this, requires considerable cognitive
sophistication, Maccoby (1980) argues. Thus
although there are many incidents of conflict in preschool groups, not all of these incontrovertibly
involve aggression. They do however serve as a
training ground for learning effective strategies for
initiating and terminating conflicting aggressive
interactions (Parke and Slaby 1983). Early strategies
and conflicts tend to involve physical aggression
(hitting, grabbing), and to be instrumental, that is they
are relatively brief acts directed towards attaining a
toy or other desired resource. These strategies
decrease as verbally mediated interaction becomes
more sophisticated, so that by around 7 aggression
has become more verbal, more personal and more
hostile Dont play with A, hes a rotten smelly
meanie. Retaliatory aggression also becomes more
common as children move from the pre-school
through the primary school. Successful retaliation
does inhibit the original aggressors, while a weak
reaction encourages them to act aggressively again.
On the other hand, the experience of having retaliated
Personality
successfully encourages aggressive behaviour in the

original victim, including preemptive attack as well
as simple retaliation after the event (Patterson et al.
1967). Children develop techniques for keeping
aggression within bounds: these include non-verbal
appeasement gestures (Camaras 1977, 1980), verbal
rituals (Opie and Opie 1967; Heath 1983) and social
rules about the conduct of fighters (Davies 1979;
Sluckin 1981). Children who contravene these peer
group regulations and act in highly aggressive ways
tend to be unpopular with their peers and lacking in
positive social skills (Parke and Slaby 1983). It is not,
of course, a simple causal sequence: aggression may
cause you to be rejected by your peers, or rejection by
your peers may cause you to be aggressive, or both.
The importance of social skills is worth noting,
however. Training in social behaviour which is cooperative and aggression-avoiding has proved to be
an effective way of treating aggressive patients.
There are developmental changes in what and who
makes children angry (Feshbach et al. 1984). As
Parke and Slaby (1983, pp. 5701) describe,
development also brings changes in childrens
understanding of stories about peoples intentions
and thus in their judgement of aggressive acts. There
is an increasing tendency with age to distinguish
between accidental and deliberate provocation,
reacting less strongly to the former. Similarly, more
discrimination is made between aggression
committed with good intentions and aggression
committed with selfish motives. These distinctions
will probably be harder to make in real life, of course.
Parke and Slaby argue that some children who
consistently respond with aggression to behaviour
which is non-intentionally negative may be having
particular difficulty in making such distinctions.
They misinterpret other peoples behaviour and react
to it as if it was more hostile than it was actually
intended to be. Some work by Dodge (e.g. Dodge
1980; Parke and Slaby 1983, pp. 5713) provides
support for this argument. Dodge suggests that there
may be an intricate self-sustaining process involved:
Given a negative outcome in the context of unclear
intentions, an aggressive child may be likely to attribute a
hostile intention to a peer who is responsible for this

negative event. This attribution may confirm his general
image of peers as hostile and may increase the likelihood of
his interpreting future behavior by the peer as hostile.
Consequently, he may retaliate against the peer with what
he feels is justified aggression. Subsequently, the peer, who
has become the recipient of a negative outcome, may
attribute a hostile intention to the aggressive child. This
attribution confirms the peers view of the child as being
inappropriately aggressive in general and increases the
peers likelihood of interpreting future behavior by the
aggressive child as being hostile. Consequently, the peer
may aggress against the aggressive child, which could start
the cycle over again. Given a series of negative outcomes,
which is inevitable, the cycle could turn into a selfperpetuating spiral of increased hostile attributions,
aggressive behavior, and social rejection (Dodge 1980, p.
169).
Dodge is implying here a model of aggression as

part of the network of social action, social cognition
and social affect. Developing cognitive skills,
especially understanding other people and moral
issues, are seen as central to the ability to control
aggression. It is also necessary to act and feel
appropriately; thus social skills are important. While
this sort of socialcognitive model is likely to lead to
a more useful understanding of aggression than did
regarding it as a biologically given instinct, there is
not as yet much evidence that changes in cognitive
skills lead to changes in childrens aggression. Nor
does it account for why people begin to behave
aggressively. Personality differences, notably on
temperament dimensions such as activity,
impulsivity and low tolerance for frustration, seem
likely to be involved.
It is probably a combination of continuity in
temperament and stability of social skills which lies
behind the rather considerable degree of continuity of
aggressiveness found in a number of studies. Olweus
(1979, 1984) reviews the evidence and finds that
boys aggression, as revealed in observations, adult
ratings or peer nominations, is highly stable over
periods of less than a year, and still pretty stable
(correlations of around 0.4 or 0.5) over periods as
153
long as ten years. It is thus comparable with

intelligence in its consistency, and more stable than
its opposite of inhibited, withdrawn behaviour.
Evidence from other studies (Kelso and Stewart
1986; Kohn 1977) contributes further to this picture.
It is well known that boys act more aggressively
than girls. They show more physical and more verbal
aggression from pre-school onwards (Maccoby and
Jacklin 1980; Meadows and Cashdan 1983; Smith
and Green 1974). They almost monopolize violent
crime, though the rate of female delinquency is
increasing (Rutter and Giller 1983). They are
somewhat more likely to initiate aggression and
much more likely to retaliate aggressively. Girls are
said to be more likely to be indirectly aggressive, for
example to set up the situation so that they are not
blamed and someone else gets into trouble (e.g.
Pollard 1985). They are certainly more vigorously
socialized into feeling guilt and shame about
aggression. J. and E. Newson (1968, 1976) document
this particularly well in their Nottingham sample.
Mothers reported that fighting is much more likely to
occur if the child is a boy or if the antagonist is a
sibling. Girls rarely fight outside the family, and
have a lower rate of fighting with siblings often, but

they fight siblings sometimes as frequently as boys.
Girls are more often advised to withdraw from fights
or conflict. The picture is one where girls are
consistently discouraged from fighting, particularly
in public. Biological differences lead to different
social learning experiences.
Family socialization techniques and personality
characteristics contribute to the level of aggression in
children. In a study of boys in their early teens,
Olweus (1980) showed the effect of four variables.
The first was negativism in the basic emotional
attitude of the principal caretaker (the mother). If she
was hostile, cold, indifferent or rejecting this had a
powerful effect on her discipline techniques and also
a direct effect on the boys aggression. A permissive
attitude towards aggression was the second variable,
and this also contributed directly to the boys
aggression: mothers were more likely to be
permissive about aggression if they had a negative
attitude to their sons. The third variable was both
parents use of discipline techniques which relied on
Figure 17 Path model relating child-rearing variables to boys aggression.

Source: From Olweus (1980), p. 650.
154
Personality
the assertion of their power rather than on negotiation

or reasoning; this was more common in negativistic
parents and made some contribution to the boys
aggression. Finally, the boys temperamental level of
activity and intensity interacted with the mothers
negativism, contributed to how much she permitted
aggression and had a direct influence on the boys
aggression. The variables fell into the pattern shown
in Figure 17. Further discussion of child-rearing
practices and social development may be found in
Chapter 6.
Finally, the social ecology of aggression has to be
considered if aggression is to be identified
correctly. For example, Manning et al. (1978)
distinguished different types of hostility in preschool children. Some children tended to get rather
rough during vigorous physical games or to bully
their peers in imaginative games, but to be timid
otherwise. Some children went in for unprovoked
teasing and showed a high level of hostility and
violence. Some children only showed hostility when
specifically frustrated or provoked. These different
types of behaviour had different prognoses for the
childs state four years later, and were linked to
different family backgrounds. Moderate aggression
at the right moment can be socially desirable.
Definitions of moderate and right are of course
offered by the culture (e.g. Walzer 1984). The childs
task is to learn and act on those definitions.
The development of pro-social behaviour

It would appear that psychoanalytic theories of iddominated impulses towards selfish pleasure,
Piagets theory of cognitive egocentrism and
sociobiological theories of the selfish gene
(Dawkins 1976) all imply that the natural state of
unsocialized man and hence of young children is
one where altruistic acts and empathy with other
people will be rare (and capable of explanation in
terms of other, less selfless motives). It has recently
been argued, however, that such a picture is
substantially false: young children are not incapable
of distinguishing other peoples feelings from their
own or of acting on this distinction in ways which
benefit the other. I have already discussed some of the
areas of egocentricity which were claimed to

dominate language and cognition: here I shall discuss
the evidence on emotional egocentricity and the
conditions that facilitate empathy and altruism.
It is not disputed that very young children, indeed
babies, may react to other peoples distress with
distress of their own. However it has often been
claimed that young children, unable to differentiate
between themselves and other people, are unclear
about who is feeling the distress.
Consequently the child probablyreacts to anothers distress
as though his dimly perceived self-and-other were
somehow simultaneously, or alternatively, in distress. As
an example consider a child I know whose typical response
to his own distress, beginning late in the first year, was to
suck his thumb with one hand and pull his ear with the other.
At 12 months, on seeing a sad look on his fathers face, he
proceeded to look sad and suck his thumb, while pulling his
fathers ear (Hoffman 1975, p. 614).
It is quite commonly observed that a child seeing

someone else in distress will try to comfort himself or
herself, or, more sophisticatedly, will offer to the
distressed other his or her own comfort object.
Theorists
who
are
conservative
about
acknowledging empathy in children interpret this as
showing the childs failure to distinguish between
self and other, attributing to the child the (incorrect)
belief that what ends his or her own distress will end
other peoples. Theorists with slightly more liberal
views about egocentricity and empathy consider the
same behaviour differently. Yarrow and Waxler
(1976), for example, suggest that the egocentric
comforting behaviour might represent an attempt to
act out and thereby understand other peoples
feelings. Self-comforting in the face of a parents
sadness, as in Hoffmans example, might also be an
appropriate reaction if the child felt distress at what
he saw in his father: what is more, self-comforting is
particularly appropriate if the usual source of
comfort, the father, is unavailable because of his
distress. Merely offering an inappropriate
155
Plate 13
156
Personality
comfort to someone in distress (and an

inappropriate comfort might through its provision
of distraction or its demonstration of goodwill prove
partially or even completely comforting) is not a
sufficient sign of an inability to empathize or of a
belief that ones own sources of comfort will serve for
everyone else. It is after all a rare adult who can
always provide appropriate comfort.
One of the things children do have to learn in order
to empathize effectively is what other peoples lives
are like; if another persons life is very unlike our own
we may have more difficulty in understanding their
feelings than if we have a lot in common. Thus we
might expect children to show empathy towards
people familiar and like themselves earlier and more
effectively than they do towards people less familiar
and less similar. Dunn and Kendrick (1982) show
unequivocally that children as young as 14 months
demonstrate awareness of the emotional experience
of parents and older siblings, and elder siblings still
under 3 years old respond to the babys state in
appropriate non-projective ways. Awareness and
response can be quite sophisticated, as in the two
examples which follow:
Bruce S (Baby playing with a balloon): He going to pop it
in a minute. And hell cry. And hell be frightened of me too.
I like the pop (Dunn and Kendrick, p. 106).
Laura W Callum repeatedly reaches for and manipulates

the magnetic letters Laura is playing with. Laura repeatedly
says NO gently. Callum continues trying to reach the
letters. Finally, Laura picks up the tray containing the letters
and carries it to a high table that Callum cannot reach.
Callum is furious and starts to cry. He turns and goes
straight to the sofa where Lauras comfort objects, a rag doll
and a pacifier, are lying. He takes the doll and holds tight,
looking at Laura. Laura for the first time is very upset, starts
crying and runs to take the doll (Dunn and Kendrick, p.
116).
As these examples show, awareness of other

peoples potential feelings does not of itself produce
altruistic behaviour. Callum uses his understanding
of Laura to get his own back on her; Iago uses his
understanding of Othello to destroy him. Lack of
understanding of other people will probably make it
harder to please or succour them, that is some degree
of understanding may be necessary for altruism but it
certainly is not sufficient. This is one reason why
attempts to find a causal link between altruistic
behaviour and the cognitive capacities of
perspective-taking, role-taking and moral
reasoning have had inconclusive and unconvincing
results (Radke-Yarrow, Zahn-Waxler and Chapman
1983; Kurdek 1978). There are no strong consistent
positive correlations between frequency of altruistic
behaviour and perspectivetaking ability, stage of
moral reasoning, IQ or even score on a test which
measures empathy (Hoffman 1977). In particular
we do not have good reason to believe that more
advanced understanding causes more altruistic
activity, a point which has to be made also in respect
of Kohlbergs theory of moral development (see next
section). Feeling, and wishing to act, are essential
components of human altruistic behaviour. Acting,
and the consequences of acting, may contribute to
understanding as well as the reverse.
Another reason for the inconsistency of the results
of studies which correlated sociomoral reasoning
and altruistic behaviour is that these have been
assumed to be two homogeneous entities. Few
psychologists have investigated how different sorts
of altruistic or prosocial behaviour are related.
Radke-Yarrow et al. (1983) point out that although
the terms altruistic and pro-social are understood
to refer to the whole spectrum of kind, helpful and
self-sacrificing behaviour, what has actually been
investigated has been a much smaller range of
actions. The dependent variable in experiments, for
example, has often been rather trivial behaviour such
as sharing out sweets or donating small sums to
157
hypothetical poor children. Observational studies

include a wider range of behaviour, but do not usually
have enough data to show how different sorts of
altruism co-occur, alternate or substitute for each
other. We badly need a natural history of pro-social
behaviour, since we really do not know much about
what children do or about what they think in this area.
Anthropological evidence suggests that western
children act helpfully and altruistically much less
often than children in non-western societies (e.g.
Whiting and Whiting 1975). Psychologists observing
western children in school classrooms and
playgrounds also report a low frequency of prosocial acts, though as Yarrow and Waxler (1976)
show the frequency found depends on how
inclusive the definition of prosocial acts is.
Undoubtedly one reason why western children
apparently lack altruism compared with children
from other cultures is the ecology of their lives, or at
least the part of their lives which is most accessible to
researching psychologists. Most observational
studies of western childrens pro-social behaviour are
done at school, that is they tend to be studies of
children in groups whose members are all much the
same age and much the same level of skill, and who
are more often than not within the supervision of
adults. Helping behaviour is much less common in
any such homogeneous group than it is in
heterogeneous ones, where helping, protecting, caregiving and patronizing behaviours are a much more
frequent part of the social interaction (Radke-Yarrow
et al. 1983; see also Chapter 6 this volume). It might
be an overstatement to say that western school
settings emphasize individual competition at the
expense of altruism and co-operation, but they
certainly give fewer opportunities for altruistic
behaviour than settings where children are important
care-givers for other children. In particular, they give
few opportunities for serious altruism, behaviour
whose cost to the actor is serious or which is seriously
needed by the recipient. The success of charitable
appeals made by childrens television programmes
such as Blue Peter suggests that we should not
158
altogether believe in the apparent selfishness of

western children, though there is clearly no room for
complacency about the altruism of the western world.
If, as I have implied, some settings, social systems
and experience make prosocial behaviour more
likely than others do, which are they? There are some
fairly firm islands of evidence in a quagmire of
ignorance, here. Social sensitivity is associated with
parental discipline, in that children whose parents
reasoned with them, negotiated issues and explained
their commands and prohibitions show more
advanced understanding of other people than do
children brought up under rigidly authoritarian or
very laissez-faire rgimes (Light 1979; Maccoby and
Martin 1983; Rollins and Thomas 1979; see also
Chapter 6 this volume). Children who readily
participate in other peoples feelings seem to have
parents who show empathy towards their childrens
distress (Zahn-Waxler et al. 1979). Seeing a model
behave altruistically has increased childrens
altruism in some experiments but by no means all
(Radke-Yarrow et al. 1983), while outstandingly
altruistic children seem often to have modelled
themselves on their own outstandingly altruistic
parents. Attempts at conditioning children to behave
altruistically by giving rewards for helping, sharing
and so forth have had inconsistent results (RadkeYarrow et al. 1983), sometimes leading the child to
act altruistically only when he or she is rewarded
which obviously makes it plausible that the
altruistic act is really a selfish one! Attributional
remarks like I know youre the sort of person who
really cares for other people may have some effect on
the childs behaviour but the effect depends on
whether the remarks appear sincere and believable,
and probably on other aspects of the social situation.
Prosocial behaviour cannot be divorced from social
behaviour in general. Children who are confident and
experienced in their relationships with other children
may be helpful and protective as part of their normal
social role. Children who are anxiously on the fringes
of a social group may seek to share their toys or
sweets, and may give up their own immediate interest
Personality
in favour of a group members, in the hope of buying

their way into a group. Children who are happy about
themselves may be able to be more generous towards
other people than children who are chronically
miserable, or their general complacency may prevent
them from noticing another persons need. We have
so little in the way of a natural history of altruism, and
so little of a convincing philosophical analysis of
altruism, that we can say very little about what
conditions promote it. Since it is an important part of
moral development and social interaction, we must
hope that better data and better theory will be
forthcoming.
Moral development
Moral development in children is a complex subject:
many different emphases are relevant to its analysis,
and because of the social implications of morality,
and the apparent importance of education and childrearing practices to its development, discussion has
often been very value-laden. As we shall see, there
are real differences as to what are and are not moral
problems, what are good or universal moral
principles, and how they develop. As in other areas of
developmental
psychology,
models
which
emphasize biology, social conditioning or
cognition are in competition: so are different views
about the natural goodness or badness of human
beings and the functions of society.
Morality and sociobiology
The most heavily biological account of morality is

relatively new and far from being fully worked out. It
is part of the ambitious claim made by Wilson (1975)
that sociobiology, which asserts that human
behaviour is determined by natural selection in ways
similar to the influence of natural selection on human
morphology, should take over psychology and the
other social sciences. At the centre of the account is
altruism, that is self-sacrificing behaviour which
has good results for people other than the person
actually doing it. This sort of behaviour is often
regarded as morally admirable in humans (Greater
love hath no man than this, that he lay down his life
for his friend), and in other animals, particularly

perhaps if the benefited person is an infant (or a
human being, as may be seen in the popularity of
lachrymose stories in which a dog saves his masters
child from predators or other danger, or keeps vigil at
his grave like Greyfriars Bobby). Wilsons work
brought to more general attention the fact that
altruistic behaviour was to be found in a great many
species where it could not plausibly be explained by
social learning or conditioning. Since altruism
involves sacrificing oneself for others, it is likely to
handicap the individuals who display it and reduce
their chances of living and reproducing themselves. It
is thus hard to account for in terms of an evolutionary
theory which says that genes for characteristics or
behaviour which militate against reproductive
success will tend not to be reproduced, and to die out.
The proposed explanation is that although altruism is
by definition of immediate disadvantage to the
altruistic person, there are indirect advantages which
may add up to a better total. Thus although the mother
lapwing who tempts the predator hawk away from her
nest by pretending to be injured, or the human father
who rushes into a blazing house to bring out his
children, may lose their own lives, if they have saved
the lives of enough of their offspring the net result
may be that as many or more of their genes survive for
later reproduction. The closer the genetic relation, the
greater the possibility of sacrifice will be. Indirect
advantages may also come about through reciprocity
of altruism: if the altruistic action of A towards B
makes it more likely that B (or indeed anyone else)
will act to the advantage of A (or As kin) in future,
then the present risk which A takes may be less than
the future benefits which B may be obliged to shower
on A. Seen in this way, what was apparently morally
admirable self-sacrifice is reduced to a somewhat
sordid calculation of self-interest. To quote the most
enthusiastic proponent of sociobiology in the days
before be began to show much caution about its
claims:
The theory of group selection has taken most of the good
will out of altruism. When altruism is conceived as the
mechanism by which DNA multiplies itself through a
159

network of relatives, spirituality becomes just one more
Darwinian enabling device (Wilson 1975, p. 120).
necessary to look at the whole pattern of motives and

behaviours.
This view is, however, an example of excessive

reductionism; it deflates the smug myth of natural
human goodness but at the cost of setting up an
alternative myth of natural selfishness. The kin
selection model has had some considerable
predictive successes though it works less well where
animals cannot estimate their genetic relatedness to
each other with any accuracy. The view runs however
into crucial difficulties, two of which I want to pick up
as they are central to the field of moral development.
The first, as I stressed in Chapter 1, is that the link
between genes and behaviour is not a simple direct
one.
They [people with crude notions of Darwinism] confuse the

mere fact of competing, that is, of needing to share out a
resource, with the motive of competitiveness or readiness to
quarrel. Where creatures are competing (as a fact), their
success will be decided by whatever tendencies they have
that best help their predicament. These need not be
quarrelsome tendencies at all. A species may prevail
because it is better at finding food or turns to a food that is
more plentiful, or because it grows protective colouration,
or indeed because it becomes less quarrelsome and more
co-operative (p. 132).
The nature of the relation between [the] genome and the

physical realization of the actual animal, or its phenome, is
an extremely complex, and as yet quite unsolved
conceptual problem. That is to say, it is not yet possible to
state in just what way any physical and behavioral feature
of an animal can be said to be determined by its genes. . . .
The conceptual obstacle to providing such an account lies
mainly in the role played by the enormously complex
context in which the genes find themselves in the course of
embryonic and post-embryonic development. (Stent 1978,
p. 18).
Developmental studies of social behaviour in various

primates (see, Trivers 1985, Gottlieb 1983, Hinde
1983) have demonstrated that the most genes can
contribute is a predisposition to learn the complex
behaviours that make up altruism or parenting.
Social experience and the opportunity for imitation
and learning are essential; if they are absent, as in
Harlows deprivation experiments, or adverse, as
seems to have been the case in the Ik (Turnbull 1972),
there is no morality. Even when they are present, it
may be doubted whether we are essentially the
calculating, prudent, consistent persons that
sociobiology at its most Hobbesian claims. Our
altruistic actions may in fact benefit ourselves, but,
and this is the second point, our motives, our ideas
about our actions, will also be worth considering. As
Midgley (1979) points out with admirable clarity, it is
160
The necessary complexity of such capacities points up the

wrongness of an atomizing approach to impulses. It seems
unrealistic to talk as though the tendency to rescue people
were something that could be carried by a single gene . . . in
any fairly complex creature, the undertaking of dangerous
actions must involve other traits in the character besides the
impulse in question; the whole character has to be such as
to permit them. Such behaviour cannot stand alone (pp.
1345).
All the creatures that it makes sense to suppose could
develop positive altruism are already caring for their young
. . . the development of sociability proceeds in any case
largely by this extension to other adults of behavior first
developed between parents and young grooming, mouth
contact, embracing, protective and submissive gestures,
giving food. In fact, wider sociality in its original essence
simply is the power of adults to treat one another, mutually,
as honorary parents and children. It is enriched later with
other patterns largely drawn from the interactions between
infants. . . . But quasi-parental interactions come first. They
work well because they are adapted to soothe, to conciliate,
to forge a bond . . . those who, from whatever cause, are
especially protective and good at rearing young, are likely
to leave a disproportionate number of descendants in
relation to those actually born (p. 136).
That is, moral behaviour is part of the results of

evolution, but simplistic accounts which treat it
atomistically, which ignore motives and reasoning,
and which isolate it from other parts of behaviour,
will be unsatisfactory. That Midgley places adult
Personality
child relationships/behaviour at the centre of the

development of social being is interesting, with
obvious self-justificatory attractions for the
developmental psychologist: it reflects back on my
early assertions about the inseparable development
of the biological and the social. We will return to the
notion, when we have considered some other
accounts of moral development which pay more
attention to Midgleys points about motives,
reasoning and the total complex pattern of behaviour.
Morality and Freudian theory
There is both biology and emphasis on the

importance of parenting behayiour, though more
pessimistically viewed, in another and better known
theory of morality and moral development, the more
elaborate and extraordinarily seductive account
given by Freud. At the centre is a dichotomous view
of the individual and the social system. There is
conceptual (and indeed emotional) opposition
between the self and others, the individual and the
group, personal fulfilment and social obligation.
It is impossible to overlook the extent to which civilisation
is built up upon renunciation of instinct, how much it
presupposes precisely the non-satisfaction (by
suppression, repression or some other means?) of powerful
instincts (Freud 1930/61, p. 44).
The powerful instincts dominate the unsocialized

individual, including the unsocialized child: they are
the instincts of sexuality (Eros), which Freud
discovered, if that is the word, early in his work
(Sulloway 1980), and aggression (Thanatos), which
grew in importance in his thinking after the horrors of
the First World War and the early death of his
daughter Sophie (Clark 1980). The social system,
culture or civilization, exists to protect
individuals and both to restrict the satisfaction of their
instincts and to allow it some limited safe expression.
The individual has powerful instincts which demand
release and reduction of their tension; they are kept
within bounds by external coercion which becomes
internalized as development proceeds.
Civilisation, therefore, obtains mastery over the

individuals desire for aggression by weakening and
disarming it and by setting up an agency within him to
watch over it, like a garrison in a conquered city (Freud
1930/61, pp. 701).
Moral development thus centres on the mastery of

the instinctual drives that seethe in the id and demand
instantaneous satisfaction. The baby quickly
experiences frustration of instinctual demands
milk, warmth, mother, cannot come fast enough or
reliably enough and the second part or aspect of the
personality, the ego, develops or differentiates itself
from the id. The functioning of the ego involves
coping with delay in the discharge of instinctual
energy, adapting the desires of the id to the reality of
the outside world.
The egos relation to the id might be compared with that of
a rider to his horse. The horse supplies the locomotive
energy, while the rider has the privilege of deciding on
the goal and of guiding the powerful animals movement.
But only too often there arises between the ego and the id the
not precisely ideal situation of the rider being obliged to
guide the horse along the path by which it itself wants
togo(Freud 1933/64, p. 77).
The horse is a fierce and powerful one, and the

external worlds rules about how and where to ride
are multiform and stringent. The ego tackles the
problem of balancing the two in increasingly subtle
ways but the problem-solving process involves
anxiety which sometimes threatens to overcome the
ego. Defence mechanisms control and alleviate
anxiety but do so by distorting reality: in reaction
formation for example the ego may deny an
unacceptable desire by focusing on its opposite, so
that it may appear to be, for example, demonstratively
loving rather than jealously hurtful, altruistic rather
than selfish. The sceptical may often see defence
mechanisms at work in the activity of moralists,
particularly those who see themselves as the
guardians of virtues unappreciated by the more lax.
While both id and ego contribute to moral
behaviour, the most important moral aspect of the
personality is the late developer, the super-ego. This
161
has two important parts, the ego ideal, which is what

one thinks one should be, and the conscience, which
tells one how far short of ideal one is, and
consequently how deplorable and worthy of
punishment. The superego is as self-denying as the id
is self-gratifying, and can be as extreme and violent.
People with very strong superegos punish themselves
very harshly, and tend to have impossibly high moral
standards which they may apply to other people as
well as themselves Social justice means that we
deny ourselves many things so that others may have
to do without them as well (Freud 1955).
The superego, and hence moral self-scrutiny, is
said to develop in early childhood, largely as a result
of the tensions of the phallic stage. Earlier, young
infants get their sensual gratification first from oral
activity and may have made rudimentary moral
distinctions between good objects, such as breasts,
which satisfy their needs, and bad ones which do not,
and subsequently from anal activity, which because it
is the object of parental training and anxiety is a
source of ideas about being good or bad oneself.
Both these stages leave traces in personality and
morality. It is however in the third stage, round about
the age range of 3 to 6, when the childs sensuality is
said to be centrally phallic, that Freud places the
social and emotional dilemmas which result in the
development of moral structures such as the
superego, and which have become famous (or
infamous) as the Oedipus complex.
At this stage the developmental courses of boys
and of girls diverge as to their detail and their
consequences, although the central dilemma of being
in love with the parent of the opposite sex is common
to both sexes and Freud believed very strongly that
human nature is inherently bisexual, writing to a
friend in 1899 that he was accustoming himself to
the idea of regarding every sexual act as a process in
which four persons are involved (Wollheim 1971, p.
120). In outline, what happens is as follows. The little
boy is in love with his mother and wants her sexually
and exclusively for himself. This brings him into
conflict with his father, whom he feels threatened by
but also loves; indeed in later versions Freud
proposed that this love for the father was erotic too. In
this crisis of ambivalence and conflict, the child
162
represses his sexuality, metaphorically castrating

himself, and identifies with his father. In this
identification he adopts or introjects all the fathers
beliefs and values as the child sees them; that is he
internalizes a threatening, punitive, hypercritical,
powerful person who is an insuperable obstacle to the
little boys sexual gratification. This internalized
version of parental authority is the superego, the heir
of the Oedipus complex.
Freud says that the little girl also starts by being in
love with her mother, as mothers have been for all
infants the first love-object because of being the first
important supplier of oral gratification. She also
loves and desires her father and comes to realize that
he has a prized object which she does not have a
penis. She blames her mother for this lack, and feels
that she herself has been castrated and is inadequate.
While the boy fears the loss of his penis as a
punishment for his incestuous desires, the girl having
lost hers already has less to fear now. Her penis-envy
leads to a depreciatory attitude towards other people
who lack penises, a possessive attitude towards her
menfolks penises, and a wish for a child as a penissubstitute. Although she does identify with her
mother and introject her values, her identification is
slower and weaker and her superego is consequently
inferior (Freud 1925, v. 19, 1931, v. 21).
This Freudian scenario of the interaction of
psychosexual and moral development derives of
course from the retrospective accounts of their early
life by Freuds patients. There is little in the way of
evidence which unequivocally supports or refutes it:
indeed it has been said that there could be no refuting
evidence since a demonstration that a person did not
experience an Oedipus complex, feel penis-envy etc.
might be interpreted as evidence for the perfect
repression of the persons Oedipus complex, penisenvy, etc. The account has been modified, notably by
feminists reaching to what they took to be its
misogynistic character (Sayers 1982; Archer and
Lloyd 1982). Since the theory is centrally clinical and
therapeutic, whether it is true or not may be less
important than whether it is helpful to patients. For
our purposes in considering moral development, I
emphasize only the fact that Freuds account of the
development of the superego is of an interaction of
Personality
biology (the instincts), society (parental authority)

and cognition (the childs evaluation of parents and
self) which together make up moral processes and
moral standards. That it suggests major changes in
morality at the age of 6 or 7 as a result of the resolution
of the Oedipus conflict, and differences in morality
between males and females, are empirically testable
questions which we will give more attention to later.
Morality and social learning theory
Social learning theory (see, for example, Bandura

1977) accounts for moral development in terms of the
childs reinforcement history. Moral development is
the acquisition of cultural values, the conditioning of
moral anxiety is conscience, and moral character is
learned habits. That which is good is that which is
reinforced. Moral values are inculcated by parents
and teachers, or by the secondary reinforcement of
association with them, and by imitation of models
whom the child can identify with. There are
developmental changes because the childs
accumulated experience is growing, because
children become able to take more and more varied
reinforcement contingencies into account as they get
older, because they become more able to anticipate
and to infer potential reinforcement, and because of
changes in adults expectations and discipline
techniques of the child. Social learning theory, unlike
classical learning theories, does concern itself with
peoples ideas as well as their overt behaviour, and
would allow that moral behaviour is influenced by
internalized moral reasoning, though this is itself
influenced by reinforcement histories. Nevertheless,
the theory predicts situation-specific behaviour as
moral judgement involves a complex process of
considering and weighing various criteria in a given
social situation (Miller 1983, p. 222).
One major strength of learning theories has been
their amenability to experimental testing.
Unsurprisingly, there has been a great deal of
experimental research on social learning theory
predictions about moral development, and the results
have generally supported its account of the learning
processes which could be involved. The results of
observational and correlational studies of more
complex situations such as family interactions,
parental discipline techniques, and the degree of

similarity between parents and adolescents moral
beliefs (e.g. Rutter 1980), which will be discussed
more thoroughly when we consider adultchild
relations, are also in accord with social learning
theory. The suggestion that moral judgement and
moral behaviour may not necessarily be the same,
and that the outcome will vary from situation to
situation although the underlying processes are the
same, seems a useful recognition of the complexity of
morality, though it obviously reduces the degree to
which the theory in its present state can predict
behaviour with any exactness. The increasing
learning theory emphasis on reasoning which
mediates between stimulus and response also of
course makes the theory less testable, but it also
makes it more similar to the influential theories
derived from Piaget which have sought to describe
the structure as well as the process of moral
development. I will discuss these, giving a rather
longer account of them, next.
Concepts of morality
The work outlined so far has concentrated on the

dynamics, the motives, of moral behaviour and
development, rather than on its organization. Piagets
work, and the further cognitive developmental
work that it has inspired, was more concerned with
moral judgement and the principles that people use in
evaluating some action as good or bad, deserving
praise or blame. Questions of how to define what is
good have been a recurring preoccupation of moral
philosophers. There is some agreement about a core
meaning of morality, though less about its content.
Moral rules may compete among themselves, but
they all have the following characteristics (Gewirth
1978). First, they are obligatory, that is they do not
depend on what anyone happens to feel like doing.
Second, they are generalizable, that is what is right or
wrong for any particular person is also right or wrong
for any other individual (assuming that there are no
relevant differences of characteristics or situation).
Third, they are important, that is, the moral rule
which is the best moral rule in a situation should take
precedence over other non-moral considerations,
such as conserving ones energy by not bothering, or
163
justifying contravening the moral rule by appeals to

etiquette or the desire for a quiet life or ones
momentary frivolities. (Although some cultures
have, of course, elevated principles of etiquette to a
pre-eminent moral status, as a great many nineteenthcentury novels would tell us, this is usually seen as
morally inferior to attention to purely moral
principles, except in so far as etiquette itself can be
justified in terms of moral principles. The most
admired characters in novels are often those who
flout etiquette in the service of abstract moral
principles, the most tragic ones those whose moral
strivings are stifled by convention. Mary Garth,
Rosamund Vincy and Dr Lydgate in Middle-march
provide good examples: George Eliot was of course a
moral philosopher worthy of serious consideration as
well as a great novelist.)
The content or justification of moral rules would
seem to be more controversial. Various thinkers have
said that something is good if doing it tends on
balance to increase peoples happiness, or to
contribute to their welfare, or to avoid their harm, or
to be congruent with principles like justice or truth
or beauty or rationality, or to contribute to the
bringing about of some desirable state of affairs
which is the goal of existence or history or whatever
(such as Evolutionary Adaptedness or The
Kingdom of God on Earth or The Revolution or
even the conquest of inflation). It is quite evident
that no single one of these will do as a necessary and
sufficient and unchallengeable condition for every
instance of the good: hence G. E. Moore suggested
that goodness was the same sort of quality as
yellowness, not to be defined in terms of anything
but itself. Alternatively one might sidestep this
debate by suggesting that many of the characteristics
listed above will be involved in the evaluation of an
action. Activities which satisfy all the requirements
will be incontrovertibly good, while those which
only satisfy some will be good but, will be
qualified goods until they satisfy so few
requirements that they have died the death of a
thousand qualifications. This would make it clearly
a psychologically relevant matter as to how the
various principles were evoked and ranked, and how
people decided between them when they were in
164
conflict. It is this sort of matter that the cognitivedevelopmental theorists of moral reasoning address.
They are concerned with how children judge good
and bad, with what sort of principles they invoke
and use in decision and justification. Piaget (1932)
made the first of the attempts to describe the
development of moral reasoning discussed here;
Kohlberg (e.g. Kohlberg 1981a,b,c) is the second
major figure, while some subsequent studies reacting
to Kohlbergs work (e.g. Mussen and Eisenberg-Berg
1977, Turiel 1983), are also touched on.
Piagets theory of moral development
Piaget was an influence on studies of moral reasoning

both through his relatively early work on childrens
moral judgements (Piaget 1932) and through his later
elaboration of the structure of cognitive
development. The Moral Judgment of the Child
pioneered the method which has been much used
since, of telling children a story involving moral
conflict and asking them what the protagonists
should do and why, or how their actions should be
judged. Here are two examples (Piaget 1932, pp.
11718, Penguin edition).
A. A little boy who is called John is in his room. He is
called to dinner. He goes into the dining room. But behind
the door there was a chair, and on the chair there was a tray
with fifteen cups on it. John couldnt have known that there
was all this behind the door. He goes in, the door knocks
against the tray, bang go the fifteen cups and they all get
broken!
B. Once there was a little boy whose name was Henry. One
day when his mother was out he tried to get some jam out
of the cupboard. He climbed up on to a chair and stretched
out his arm. But the jam was too high up and he couldnt
reach it and have any. But while he was trying to get it he
knocked over a cup. The cup fell down and broke.
[. . .] About each of these pairs of stories we ask two
questions: (1) Are these children equally guilty? (2) Which
of the two is the naughtiest, and why? It goes without
saying that each of these questions is the occasion for a
conversation more or less elaborate according to the childs
reaction.
Personality
Similar pairs of stories were used to probe

childrens ideas about lying (incredible but without
any evil intention versus quite probable but told with
intent to deceive), about the appropriateness of
different sorts of punishment (expiatory or
reciprocal), fairness and so forth, and ideas of justice
and equality, and the nature and origin of the rules of
games such as marbles were also investigated. Piaget
discusses all these in terms of two different
moralities. The one which predominated in the
answers of the younger children he called a morality
of constraint or heteronomous morality. The child
believes that there are strict rules imposed from
outside, which must be obeyed and cannot be
questioned, altered or avoided, and are applied to
wrongdoing without taking into account intentions,
mitigating circumstances or the possibility of
avoiding doing wrong. Piaget attributes the source of
this harsh and rigid morality to the childs response to
parental authority, which inevitably means the
issuing of injunctions which make no sense to the
child. He gives examples drawn from observations of
his own children (pp. 1701).
Jacqueline has never been punished in the strict sense of the
term. At the worst, when she makes a scene, we leave her
alone for a little while and tell her we shall come back when
she can talk quietly again. She has never been given duties
as such, nor have we ever demanded from her that sort of
passive obedience without discussion which in the eyes of
so many parents constitutes the highest virtue. We have
always tried to make her understand the why of orders
instead of laying down categorical rules. Above all, we
have always put things to her in the light of cooperation: to
help mummy, to please her parents, to show her sister,
etc. are for her reasons for carrying out orders that cannot
be understood in themselves. As to rules that are
unintelligible to very little children, such as the rule of
truthfulness, she has never even heard mention of them.
But in ordinary life it is impossible to avoid certain
injunctions of which the purport does not immediately
seem to have any sense from the childs point of view. Such
are going to bed and having meals at given hours, not
spoiling things, not touching the things on daddys table,
etc. Now, these commandments, received and applied

before being really understood, naturally give rise to a
whole ethic of heteronomy with a feeling of pure
obligation, with remorse in case of violation of the law, etc.
For example, one evening I find Jacqueline, aged 2;
6(15),* in bed, spoiling a towel by pulling out the threads
one by one. Her mother has already often told her that it is a
pity to do that, that it makes holes, that you cant mend the
holes, etc. So I say to J.: Oh, but mummy will be sad. J.
answers calmly and even with an ill-concealed smile: Yes.
It makes holes. You cant mend . . . etc. I continue my
lecture, but she obviously is not going to take me seriously.
Still hiding her amusement with difficulty, she suddenly
says to me Laugh! in so comic a tone that in order to keep
a straight face I quickly change the subject. J., very
conscious of her powers of seduction, then says to me My
little darling Daddy, and the incident ends. The next
morning, however, J. wakes up full of it. Her first words
refer to what had happened the night before. She thinks
about the towel and asks her mother whether she isnt sad.
So in spite of the first reaction showing such charming
disrespect, my words had told and the command had
brought about the usual consequences.
The evening of the same day, J. begins to pull the threads
out of the towel again. Her mother repeats that it is a pity. J.
listens attentively but says nothing. A moment later she is
calling out and cries till someone comes to her: she simply
wanted to see her parents again and make sure that they bore
her no grudge.
The other major source of this moral realism is

the childs egocentricity. Immediately after the
passage quoted above, Piaget points out that one
begins to differentiate ones own intentional actions
from ones involuntary ones at quite an early age (3
4), and quickly to use this as an excuse (I didnt mean
to, I didnt do it on purpose). But it is harder to
apply this insight to other people, particularly when
one is oneself the victim of the others wrongdoing
but even more when the problem is in a story rather
than part of an immediate practical situation (Piaget
1932,
* 2; 6(15) = 2 years, 6 months and 15 days.
165
p. 177). This last is an important point which has

repercussions for research methodology and theories
of moral development that have not been sufficiently
considered. Piaget here suggests a gap between moral
reasoning in the abstract and moral behaviour in real
life a problem for all the cognitivedevelopmental
theorists but also that real moral behaviour may
be, at least at early ages, more advanced than
discussion of moral stories can be. We will see that a
difference in the opposite direction has been
suspected at later ages.
Although heteronomous morality is strong in the
young child, and can be seen to flourish in society
(Piaget 1932, p. 383), a second (and, as far as Piaget
is concerned, better) morality gradually develops. By
way of interaction with peers and consequently more
egalitarian experience, the child comes to experience
the benefits of co-operation and reciprocity and to
develop a morality of equality and autonomy. Adult
behaviour to the child plays a part in this (Piaget 1932,
p. 307) but equal interaction and particularly cooperation and mutual respect among peers is the main
reason for development. Blind obedience to authority
is no longer esteemed; punishment is no longer to be
expiation but should as far as possible be reciprocal or
make reparation for the wrong done; rules are the
products of social interaction and may be changed by
democratic agreement; justice should take into
account extenuating circumstances, the protagonists
motivation, and so forth.
While social interaction is seen here as a major
source of developmental change, Piaget is concerned
to link levels of moral reasoning with levels of
cognitive functioning. The decline of intellectual
egocentricity is necessary for moral development;
consciousness of oneself as a particular individual
with particular limitations and resources is necessary
for co-operation with others (Piaget 1932, p. 381).
This linking of cognition with moral, social and
emotional development did not progress into detailed
specifications of links in Piagets own work but has
been developed since, notably by the Kohlberg
school (e.g. Kohlberg 1964, 1981; Lickona 1976)
which also modifies the Piagetian theory and extends
its stage sequence at both ends. In these modifications
and extensions more emphasis is placed on the
166
developmental role of changes in cognition than on

changes in social experience where research
evidence is said only to give mixed or weak
support (Lickona 1976). Some more recent evidence
bearing on the question of links between social
experience and moral behaviour will be considered
after Kohlbergs work has been described.
Kohlbergs theory of moral development
Kohlberg suggests that there is a sequence of six

stages of moral judgement. Like Piagets stages of
cognitive development, Kohlbergs moral stages
involve structured reasoning, are invariant in
sequence, hierarchical and universal throughout the
human species. They are concerned with what ought
to be, rather than with actual behaviour, and the
essential characteristic of a stage is the sort of
underlying reasoning which leads a subject to a
particular judgement, rather than the content of the
judgement itself. While the sequence of stages and
the underlying structures of each stage are invariant,
progress through the stages will vary in speed and
some individuals will not reach the later stages.
Progress is a result of an interaction between the
maturation of individuals and their experience,
particularly their experience of social role-taking,
with a neo-Piagetian equilibration playing an
important part in the interaction as well as within the
cognitive moral structures. As Turiel (1973, p. 737)
says
Stages in a developmental sequence are, then, successive
levels of equilibrium in two respects. First, each stage is a
more equilibrated form than the previous one (e.g. there is
more internal consistency). Second, each stage represents a
more equilibrated means of interacting with the
environment. That is, each new stage is a more adequate
way of understanding moral problems and resolving
conflicts encountered.
As in Piagets work, subjects are told stories

involving moral dilemmas and invited to comment on
what the protagonist should do, and why. This is
perhaps the best-known story.
Personality
In Europe, a woman was near death from a kind of cancer.
There was one drug that the doctors thought might save her.
It was a form of radium that a druggist in the same town had
recently discovered. The drug was expensive to make, but
the druggist was charging ten times what the drug cost him
to make. He paid $200 for the radium and charged $2000 for
a small dose of the drug. The sick womans husband, Heinz,
went to everyone he knew to borrow the money, but he
could only get together about $1000 which is half of what it
cost. He told the druggist that his wife was dying and asked
him to sell it cheaper or let him pay later. But the druggist
said, No, I discovered the drug and Im going to make
money from it. So Heinz gets desperate and considers
breaking into the mans store to steal the drug for his wife.
Responses to stories of this sort are analysed, using

a detailed handbook, in terms of the quality of the
judgements, the ways in which the situation is
perceived and what moral principles are considered
and in what way, rather than in terms of the surface
content of the solution recommended to the storys
conflict. The inherent dilemma of the stories is
between an act which would comply with social or
legal rules but would neglect the welfare or human
rights of an individual (for example, Heinz deciding
he must not steal, even though this deprives his wife
of the drug) and an act which violates the rules but
serves human needs (for example, stealing the drug
and giving it to the sick woman). The important
criterion of the stages is however not Steal or Dont
steal but the reasons behind this decision.
Discussion of ideas about laws, rules, authority,
responsibility, equality and justice is at the centre of
the stage sequence and of Kohlbergs conception of
morality.
The six stages are paired into three levels; Preconventional, Conventional, and Post-conventional.
At the earliest level, which Kohlberg says is
characteristic of most children under 9, the person
does not understand societys rules and expectations,
and considerations of personal interest and advantage
are paramount. The first stage in this level is of
heteronomous morality and the emphasis is on
avoiding punishment by obedience to authority.
Piagets heteronomous morality involved respect
for adults: Kohlbergs does not, as he found not
respect but self-interest and fear. In the second

stage, the child is more concerned than before with
the positive aspects of relations with other people:
reciprocity becomes extremely important but on the
pragmatic ground that if you scratch my back, Ill
scratch yours, what Kohlberg at an earlier point in
his work called instrumental relativism (Kohlberg
1971). The beginnings of Piagets autonomous
morality or morality of co-operation may be seen
in this stage: it is also reminiscent of the
sociobiologists reciprocal altruism, discussed
above.
The Conventional level is preoccupied with
maintaining the expectations of the social group and
obeying the law precisely because they are the law
and the social consensus. Conformity has become a
matter of active support and identification not just of
fearful or self-interested compliance. Increasingly as
Kohlbergs work progressed, it appeared that the
majority of adolescents and adults were to be found at
this level. In the earlier of its two stages, stage 3, the
person is concerned to win approval, particularly
from the immediate social group. Behaving well is
still seen as important, but so is meaning well.
Kohlberg called this the good boynice girl stage (it
is not clear whether the different moral implications
of these two phrases are intended at this point: we will
discuss the question of sex bias in the theory later). In
the later stage, stage 4, there is an emphasis on laws
and duties which are seen as necessary for the
maintenance of society, which is now understood as
being wider than ones own social group. Justice is
based on the authority of government, punishment is
to expiate ones debt to society. Equality before the
law is important, as is equality of opportunity, but
efforts to eradicate more specific economic or social
inequalities by positive discrimination or more
generous treatment of the disadvantaged are likely to
be seen as wrong; conserving the social status quo is
preferred to bringing about social change.
It is in the minority of people who move into the
third or Post-conventional level that this dependence
on consensus and normative sources of morality
begins to be replaced by truly autonomous moral
reasoning based on universal moral principles. The
stage 4 thinker believes it is important to obey the law
167
simply because it is the law: the stage 5 or 6 thinker is

prepared to say like Mr Bumble in Oliver Twist, and
with reasons in terms of impersonal principle, that If
the law supposes that . . . the law is a ass a idiot and
should be changed. In stage 5 democratic agreement
and recourse to agreed procedures for changing the
rules are supported: laws are respected in terms of the
purposes such as the providing of equal rights to
life, liberty and the pursuit of happiness which the
laws are intended to fulfil. At this stage there is an
explicit awareness that there may be conflict between
moral principles. In the story of Heinzs dilemma,
Kohlberg says that a stage 5 thinker might believe that
Heinz would be morally right to steal the drug to save
his wifes life but that he would at the same time be
legally wrong so that it would be a judges duty to
punish him. Moral reasoning at this stage also tends
to take up a utilitarian position, to uphold principles
of the maximization of human happiness, which
Kohlberg regards as ethically inferior to the absolute
principles of justice and individual conscience that
characterize stage 6.
At stage 6 the process of making moral judgements
has become autonomous. Neither the egoistic desires
which dominated the Pre-conventional level nor the
social pressures which dominated the Conventional
level and were also considered to some extent in stage
5 are involved any more. The crucial principle is the
application to all individuals of a belief in the dignity
of human life; all individuals must be given
fundamentally equal consideration. This may
involve very careful consideration of what other
peoples needs, wishes and judgements would be in a
situation of moral conflict in terms of the universal
moral principles that should guide actions. Heinz, for
example, would consider whether the druggist would
continue to place property rights over a human
beings right to life if it were the druggist who was
mortally ill and someone else who was withholding
the cure; and also whether the wife would continue to
demand the drug if she were the property-owner and
someone else the invalid. Kohlberg asserts that the
inevitable outcome of this balancing of role-taking
would be to demonstrate quite clearly that the right of
a human being to life takes precedence over another
human beings right to property; if the druggist does
168
not concede this, he is wrong, and Heinzs theft is

morally justified.
Whether or not the principles involved in this stage
are unarguably better than any other, the reasoning
processes involved are clearly very sophisticated. It
is perhaps not surprising that they are described as
isomorphic with formal operations: or that the age
at which we can expect stage 6 to be attained has been
raised in successive accounts from early adolescence
to early middle age. There have also been suggestions
for stages 5 and 6 as well as for formal operations that
not all thinkers progress so far, let alone always
manage to reason at their highest possible level. In a
recent paper (Kohlberg 1984) Kohlberg has linked
his difficulty in finding Post-conventional moral
reasoning with post-Watergate disillusion with
political principle and with the privatism that has
accompanied the economic recession of the last few
years. He has therefore advocated moral education
which tries to bring about stage 4 or social duty
reasoning as this may be the best that can be achieved
in the face of the characteristically stage 2 or stage 3
reasoning of the new conservatism.
Kohlberg, then, is describing a series of stages of
moral reasoning which make a first transition from
self-interest to social interest, and in some cases may
make a second transition from considering the
immediate society as it is to considering wider
universal principles. He puts the moral principle of
justice at the centre of his model, and sees cognitive
development and one of its consequences, improved
role-taking ability, as the main motive force for
development. In many ways it is an impressive
theory, but there are a number of problems both in
how it is conceived and in the quality of the evidence.
In outline, these involve questions about stages,
about moral reasoning based on universal principle,
about justice as that principle, about relations
between reasoning, feeling and acting, about
individual differences in morality, and about research
methodology.
Criticisms of Kohlbergs theory
I have already pointed out some of the dangers of

stage theories. In the case of Kohlbergs model, as
with all the other stage sequences that have
Personality
proliferated so much since Piaget developed a stage

model, there are dangers of creating artificial
separations and artificial uniformities, overlooking
variation within stages and similarities between
stages, simplifying complex behaviours into linear
models, and reifying into a unity what might be more
usefully thought of as multiform. It cannot be said too
often that although we may have as an ideal a belief
that if something exists, we can measure it, we need
not believe that if we can measure something, it
exists. Moral reasoning is a case in point. There is
really very little evidence that moral reasoning is
consistent within stages, hierarchical between
them, accurately diagnosed over a wide range of
situations and so forth. There is also no really clear
account of exactly what they are stages of, moral
development, judgement, reasoning surely
involving a number of different components
understanding, using, feeling, for example which
are conglomerated in Kohlbergs stages. If, as Peters
(1974) argues, the order of the stages is logical rather
than psychological, we are also in danger of doing
pseudo-empirical research uselessly (see Smedslund
1980).
Kohlbergs model places universal principle at the
centre of moral judgement, and is thus an ethical rule
theory (Munsey 1980; Rosen 1980). Ethical rule
theorists assume that one must have a moral rule to
make a justified moral judgement. The general
consideration, the moral rule, is especially important,
as it is necessary both for the moral judgement and for
identifying the relevant particular facts of a case
before the judgement is made. There is an alternative
to ethical rule theory in ethical act theory, which
suggests that possession of a general moral rule may
not be necessary for making a moral judgement, since
relevant facts in a particular case can be identified
without moral rules and may override moral
generalizations. Act theorists see moral rules as
summaries which may admit of exceptions, and
morality as therefore situation-specific; rule theorists
see them as a priori universals. Rosen (1980)
supposes some developments of the Heinz story to
make this point. One of these, and Rosens
subsequent comment, follow and illustrate some of
the difficulties of Kohlbergs work.
Suppose the wife has contemplated suicide for years and

wishes to die with dignity now. Heinz wants her to stay alive
because he and she live off the interest of a trust fund set up
in her name. The druggist is the brother of the wife, knows
his sisters intentions, and while he legally would have to
sell the drug to Heinz if the latter had the money, he uses the
legal excuse not to sell the drug in order to allow his sister
to carry out her own ends.
We would have to know much more about Heinz, the
druggist, the wife, the society they live in and many more
relevant items, in order to construct our conditionals. . . .
Moral dilemmas are real problems, faced by real people in
a real setting. It is no test of an ethical theory, or of the moral
reasoning of people for that matter, to pose artificial
problems. The problems, note, are not artificial primarily
because they are fictional. They are artificial because they
do not represent realistic situations with all their
complexity. Perhaps we should let the poets and the
novelists describe the moral problems for the tests and not
the philosophers and psychologists. The problems would
be more difficult to solve, but at least they would be relevant
to the real problems that human beings have (Rosen 1980,
p. 259).
In addition to the inadequacy (pp. 1634) and

superfluity (above) of universal moral principles, it
has been argued (notably by Peters (1974) and
Gilligan (1977)), that elevating justice to the top of
the list of universal moral principles or even making
it the source of all the rest of the list, is unsatisfactory.
The principle of caring for others is seen as equally
important. This is a particularly interesting point for
developmental psychology because there is a great
deal of evidence that even very young children are
aware of, and frequently sympathetic towards, other
peoples feelings. A model of moral development
centering on sensitivity to others would be a great
deal more positive about young children than
Kohlbergs justice model, and would have obvious
educational implications. Such a model is sketched
by Mussen and Eisenberg-Berg (1977). It would also
reverse what has been seen as a sexist bias. Gilligan
(1977) refutes the tendency in work on moral
development to view women as morally deficient
(Freud provides a slightly earlier example than
Kohlberg, but one could produce many ancient
instances). Kohlberg, finding that womens moral
169
reasoning was frequently at stage 3 (behaving well,

meaning well, pleasing people the good boynice
girl orientation), interpreted this as lower than
mens, and both functional and adequate for them!
With admirable control, Gilligan comments
And yet herein lies the paradox, for the very traits that have
traditionally defined the goodness of women, their care
for and sensitivity to the needs of others, are those that mark
them as deficient in moral development. The infusion of
feeling into their judgements keeps them from developing
(one might prefer expressing) a more independent and
abstract ethical conception in which concern for others
derives from principles of justice rather than from
compassion and care (Gilligan 1977, p. 484).
Gilligans account of the feminine voice in

morality is of compassion and love, with associated
non-violence, at the centre of moral judgement.
While I find this more appealing and more admirable
than Kohlbergs enthroned justice, I want for the
moment only to use it to tie in with a point made by
Peters (1974) and others: it is not enough to know
what is right and wrong; one also has to care. It is
this neglect of feeling and of action that is one of the
most serious reasons why Kohlbergs theory must not
be taken as a complete account of moral
development. The theory does not address itself
adequately to the relationship between judgement
and behaviour, and what little evidence there is
suggests that level of judgement does not predict
behaviour at all well (Kurtines and Grief 1974).
Kohlberg (1981b) says that a stage of judgements of
justice is a necessary but not a sufficient condition for
moral action: given the diagnostic problems I have
touched on above, even this may be a somewhat
generous statement.
Recent work on moral development
Recently, work on moral development has taken

some interesting new directions. One of these is the
suggestion by Turiel and his colleagues that issues of
morality and of social convention are differentiated
by children as young as 4 to 6. Moral rules (such as not
hurting other people) are seen as obligatory,
important and generalizable, and conventional rules
170
(such as wearing clothes to school) as alterable, less

important and more context-specific, except in so far
as they were related to moral principles (such as not
embarrassing people at school by going naked). A
number of studies (e.g. Turiel 1978; Pool, Shweder
and Much 1980; Nucci and Turiel 1978) show
children giving different rationales to account for
rules that have a moral, a conventional or a practical
base (practical rules are such things as clean your
teeth twice a day). This research does not provide a
really convincing demonstration that morality and
convention are separable areas, first because moral
rules were used as the ultimate reason for
conventional ones and second because of the
confounding effect of the severity of the
transgression. However, this approach has provided
some interesting data about childrens social
cognitions, and a corrective (if one was needed) to the
idea that morality is one simple thing.
Another promising initiative (Weinreich-Haste, in
press) examines the role of feeling in the moral
experiences of real life. Helen Weinreich-Haste
regards the use of hypothetical moral dilemmas as
concealing the role of feeling: subjects do not react to
Hans dilemma as they would to a dilemma of their
own, and thus their reasoning about Hans is not
representative of all their moral reasoning. Kohlberg
and his colleagues see moral cognition as leading to
action, not the reverse. A high level of moral
reasoning means that the reasoner perceives his or her
responsibility to act in accordance with the moral
rules involved in the issue: seeing this responsibility
leads to a closer integration of moral judgement and
moral action (Kohlberg and Candee 1984).
Responsibility is an obligation to act appropriately; it
organizes and energizes action. Feelings about the
situation are not considered, so there is no account in
Kohlbergs work of how they arise, how they affect
reasoning, or how they are involved in moral action.
Weinreich-Haste argues that this unidirectional
relationship between cognition and action, which
also gives little place to affect, is inadequate. In reallife moral problems much more is involved. The first
key concept is responsibility, which involves
perception that this problem involves moral issues, a
belief that you can be effective in taking action on
Personality
them, and a belief that it is right and necessary that

you, personally, should be involved, that it is not
enough to leave the responsibility to someone else.
Thus responsibility involves vision, efficacy, and
commitment: all of them blending affect and
cognition. If young children seem to be inadequate as
moral agents, I would suggest that this may be to
some extent because many of the responsibilities
which could come their way are taken for them by
adults, and because they know they are not skilled or
powerful enough to be as effective as adults. A
judgement that somebody else will cope is a barrier
against commitment. We know that children are less
helpful to each other in an adult-led peer group than
in more heterogeneous groups (Radke-Yarrow et al.
1983, and see p. 158) and that participating in
discussions within the family seems to facilitate the
development of social responsibility (Maccoby and
Martin 1983; Radke-Yarrow and Zahn-Waxler 1983;
and Chapter 6 this volume).
Real moral problems commonly last for more than
the few minutes spent considering Heinzs dilemma,
and dealing with them involves a complex sequence
of judgements and feelings. Weinreich-Haste
presents material on five people who have
experienced moral crises which have led to a change
in their lives. For example, Sandra decided to become
a vegetarian. The triggering event was a stay in
France where she witnessed butchers slaughtering

animals for meat and was also revolted by meat
cooked so that it was burnt on the outside and bloody
inside. She felt disgusted, and reflecting on her
feeling decided that eating meat was a moral issue she
could do something about. Her feeling of disgust
turned to a moral feeling, I just cant face eating
something like that which has been killed in that way
and hasnt had its own life, a moral judgement, I
think thats wrong . . . because the animals themselves
havent really had a life. Were just breeding them to
kill them, and a moral action, the decision not to eat
meat.
The model of moral crises derived from this and
other case-studies involves cycles of events, reactive
affect, cognitive reflection, moral affect, feeling of
responsibility and commitment to action. Level of
moral reasoning is important in perceiving the issue
to be morally important and in perceiving personal
responsibility to do something about it, but the
subjects sense of personal responsibility and selfefficacy, and social legitimation and support from
other people, may be crucially important too. The
interplay of affect, cognition and action lasts a long
time, and is complex. There is little data as yet on preadolescents moral crises, but no reason to believe
that they are always simple.
171
Plate 14
172
6 Social relationships
I have stressed throughout this book that childrens

development takes place in large measure through
social relationships. Other peoples behaviour
towards the child, and the childs behaviour towards
other people, influence the development of
cognition, language, personality, emotion and, of
course, social behaviour. If it were possible for a child
to grow up without any social relationships at all, and
it probably is not, that child would not be
recognizably human: would not have spoken
language, would not have the intellectual skills we
revere, would not, probably, have self-awareness or
empathy. Social interaction is necessary for all this:
feral children, who may have completely lacked
interaction with other human beings resemble
beasts, savages or idiots (Zingg 1940; for a
recent review see Skuse 1984). Because such cases
are, fortunately, rare, and usually, unfortunately,
badly documented, we do not know how complete
and how permanent the distortion of their
development is. There are two famous cases where
initial state, training given and degree of recovery are
well documented. Victor, captured in the forests of
Aveyron in 1800 after he had lived there for at least
three years, and probably since early childhood, was
put into the hands of Jean-Marc Itard and his
housekeeper Madame Gurin. He was about 13 years
old. Itard used patient and careful training procedures
which remarkably anticipate twentieth-century
behaviour modification techniques (Lane 1976/7).
Victor progressed from purely sensori-motor
intelligence, little emotional display and no social
skills whatever, to some conceptual thought and
moral and empathic feelings, but his language
development was disappointing and his emerging
sexual feelings were disturbing to the adults
responsible for him (and to the boy himself). A rather
similar picture emerges from accounts of Genie

(Curtiss 1977) who was imprisoned by her psychotic
father from infancy to early adolescence in a state of
almost total sensory and social deprivation. She too
was barely human when rescued, has developed
considerably with sympathetic teaching but shows
remaining deficits in language use and social skills.
Children less totally deprived of social lives may
show more recovery than these two (Koluchova
1976; Clarke and Clarke 1976; Skuse 1984) if given
carefully set up social and educational experience.
Since similarly severe deficits arise in other species,
such as rhesus monkeys reared in isolation (Harlow
1969), we would have good grounds for regarding the
social world as a biological necessity for individual
development.
If something is a biological necessity it is likely
or at least possible that structures and processes
which facilitate it will have evolved through natural
selection and will be programmed by genes or
memes (Dawkins 1976; Trivers 1985). Are there
such structures and processes contributing to the
social development of children? If there are, how do
they work?
Infantadult interaction
It is fairly clear that babies are, and do, and can do,
things which are likely to be useful in their social
world. They have the big shiny eyes, plump cheeks,
high foreheads, fine skin, smell of milk and uncoordinated movements which evoke automatic
reactions of tenderness and nurturance in adults in our
own species as in many others (Tinbergen 1951).
They selectively attend to human faces and voices,
and astonishingly early appear to discriminate the
familiar voice of their mother from other nonfamiliar voices (Mehler et al. 1978), a preference and
173
a discrimination likely to convince mother that she is

special to her baby. Their vocalizations and their
movements indicate what they feel, and receive
social interpretations from their caretakers. Adults
treat them as individuals who are trying to
communicate, as social beings with needs and wishes
and intentions not unlike their own. Recent research
has shown that babies behaviour is also subtly
patterned in ways which resemble the patterns of
social interactions between adults, such as
conversations (e.g. Trevarthen 1978), and the
impressiveness of fit between baby and caretaker in
their interactions has led some theorists to suggest
that the baby is a social being from birth. Kaye (1984)
argues that such a view is mistaken, that the fit is due
entirely to the adult, who seizes on any bit of the
babys behaviour which could possibly be
understood as having a social or communicative
significance and treats it as if it really did, although
the baby was not in fact capable of intending anything
of the sort. There are indeed patterns innate in the
babys behaviour cycles of arousal and the
characteristic rhythm of sucking in a burst pause
alternation are central ones but their function is to
entice the mother into turn-taking interaction with the
baby. When the baby pauses in sucking, for example,
the mother tends to interpret this as loss of interest or
falling asleep on the job and takes steps to recall the
babys attention to breast or bottle by talking, jiggling
the nipple in the babys mouth, or changing the babys
position. Thus while the baby is actively sucking, the
mother is relatively quiet: when the baby stops
sucking, the mother becomes active. This could be
seen as turn-taking or protoconversation. Kaye,
however, points out that the babys part in this is
entirely automatic. The pattern of turns is due
entirely to the mother who learns quite quickly that
jiggling makes it less likely that the baby will suck,
that what provokes sucking is the cessation of a brief
jiggle, and who therefore changes her own behaviour
to affect the childs. Similar turn-taking and change
in turn-taking can be seen in games and
conversations throughout the years of infancy (and
beyond) as we have seen in our discussion of
language development (Chapter 4). Adults fit into
174
childrens behaviour to produce a shapely sequence,

or to achieve a goal, or to prolong interaction. The
baby is genetically programmed to produce
regularities and hence predictabilities of behaviour.
The adult is genetically programmed to pick up and
use (or at least to be capable of picking up and using)
these predictabilities in ways which apprentice the
baby to the social (and intellectual) ways of the adult
expert. By first using rhythms and regularities for
shared activity, later imputing intention and a desire
to communicate that intention (Dunn and Kendrick
(1982) show us children as young as 4 doing this for
their baby siblings), later still using an assumption of
shared memory and shared language, adults
treat the child as more mature and more of a partner than he
really is. Admittedly, there are real cues from the child that
show he understands more than he did last week or last
month. But the higher forms of interaction into which the
adults slip are inevitably more advanced than what the child
is actually capable of at the time. Thus parents are
constantly drawing the child forward into a more
challenging apprenticeship, eventually into a full
partnership (Kaye 1984, p. 68).
Parenting
Kaye uses the idea of adults providing functional

frames for their children, organizing for them the
world of objects, people and events in ways which
reduce potential chaos to intelligible order. Adults
nurture children, meeting their needs for
nourishment and comfort (in both physical and
emotional senses) and in so doing allow and enhance
communication
and
mutual
understanding
(intersubjectivity). They protect children from harm,
ideally while still allowing them to do things which
are not yet quite within their competence. They act as
helpers or instruments, either doing for children
things they cannot do for themselves or modifying the
wished-for activities or objects so the child can
achieve them. They provide feedback on the childs
actions so that consequences can be more consistent
or more salient or less dangerous than in nature. They
provide models and demonstrations of skills and
Social relationships
attitudes. They support and encourage discourse,

which is a means of sharing and enhancing
understanding. They act as a memory for the child and
this helps in the organization of knowledge and the
fulfilment of plans (Kaye 1984, pp. 7783). I would
add that they modulate the childs arousal and invite
participation in culture. To do all this requires a great
deal of adult sensitivity to what the child is feeling
and doing, and a great deal of patience, as many, many
repetitions of various frames will be needed for all
the child has to learn. What good parents do is make
possible the guided reinvention of language (Lock
1978) and of cognition, social convention and so on.
Bonding and maternal instinct
The first three of these frames, and perhaps

modelling and feedback, can be seen in parents of
other species: the full repertoire is distinctively
human. Most parents, many adults and many older
children fit their behaviour to babies and young
children in framing or scaffolding (Ninio and
Bruner 1978) ways. However to produce this fit
consistently and effectively throughout the years of
childhood requires an enormous investment of
goodwill and energy and hence a major emotional (or
professional) commitment to the child. The concept
of maternal instinct has been suggested as an
explanation of why this commitment is usually
forthcoming from the mother. A cultural expectation
has grown up that mothers naturally know whats
best for their babies. More specifically there is a
belief that the hormonal changes of late pregnancy
make the mother ready to fall in love with her
newborn baby so creating a motherinfant bond.
Winnicott (1958) called this primary maternal
preoccupation, describing it as a state of heightened
sensitivity, almost an illness in which the mother has
to identify herself with the baby in order to be able to
identify and serve its needs. This bonding is supposed
to happen rapidly in the first few days (or hours or
minutes) after the babys birth: it is implied that if
bonding goes well all will be well thereafter, and that
if it fails there will be serious and perhaps insuperable
problems. These two implications about the
consequences of bonding or not bonding, and the

existence of maternal instinct and primary maternal
preoccupation themselves, need to be looked at very
carefully.
At least some mothers (Oakley 1980), and given an
opportunity to express it, some fathers (Jackson
1984), report feelings about their newborn child
which resemble primary maternal preoccupation.
Some do not; reporting no emotion, negative
emotions, ambivalence or sheer exhausted
confusion. Many newborn children elicit from their
parents the same stereotyped patterns of looking,
touching and talking, and the same feelings of
tenderness. Obstetric practice during the 1950s and
1960s reduced the amount of time new mothers had
their infants with them, and in the interests of hygiene
and routine separated mother and baby except for
feeding. This separation came to be seen as
interfering with bonding, and a series of researchers
(Klaus and Kennell 1976; Field 1977; Richards 1979;
Trowell 1982) have shown that if mother and baby
have very little contact in the period immediately
after the babys birth the outcome may be worse than
normal, all other things being equal, and if they have
more contact than usual outcomes tend to be better
than normal, again all else being equal. This has had
a beneficial humanizing effect on obstetric practice.
The differences are not however large (or consistent)
and they certainly do not mean either that good early
bonding guarantees good development despite later
problems, or that failed early bonding leads to
inevitable disaster. The reason why mother and baby
were separated also needs examination. If it is
because the baby was ill, prematurely born, or the
product of a difficult labour (Trowell 1982 looked at
caesarian births) it may be the illness, rather than the
separation, which causes problems through the
anxiety and loss of self-confidence which it
engenders in the mother. It is a consistent finding that
babies who are ill at birth and need a few days
treatment in an intensive care unit because of low
birthweight, failure to breathe or to establish a normal
heartbeat and so forth, are likely to show some
developmental deficits in the first year of their lives
but to catch up with the normal healthy baby by the
175
time they reach school, provided they do not add

major social disadvantage in their family to their poor
state at birth. Economic or educational disadvantage
in the parents are much stronger predictors of
developmental problems in the child than whether the
child was medically high-risk at birth or whether
there were difficulties in bonding. Unfortunately in
this area as elsewhere different sorts of disadvantage
co-occur. A socially disadvantaged mother is likely
to have a large number of problems (see, for example,
Rutter and Madge 1976) such as: being less well-fed,
less well-grown, and less well-cared-for before
becoming pregnant, and becoming pregnant when
younger and more immature; having less good food,
health, living conditions, medical care and freedom
from physically demanding work during pregnancy;
having her baby in a worse-staffed hospital or
unexpectedly without good medical care; and to live
with the baby in an overcrowded, unfit, lead-polluted
environment with continuing poor diet, poor healthcare, economic stress and so forth. Every single one
of the disadvantages listed above is among those
thought to have an adverse effect on the development
of children: accumulating, they can seriously
disadvantage parent and child even if the mothers
maternal instinct is strong and mother and child
bonded well at birth. There is controversy over
whether maternal instinct and bonding exist
(Sluckin et al. 1983); whether they do or not they are
clearly neither necessary nor sufficient conditions of
good parenting. A society where most first-time
mothers have little personal experience of babies,
where many expect that babies will resemble dolls or
the photogenic cherubs of advertisements, and where
many such mothers have no training and little support
in dealing with the realities of post-natal life must
expect there to be stress and problems (Boulton
1983). We will examine the extent and the
implications of these later.
Intersubjectivity
I have said that babies behaviour shows preferences

and regularities which make it possible to treat them
as social beings, and also that babies who never
176
experience social interaction seem not to develop the

specifically human characteristics of self-awareness
and representational language unless given a long
period of special care and training. Two issues here
need some further discussion: the first, to be touched
on only very briefly, is how and when the baby
functions as a truly social person, and the second, to
be examined more fully, is what sort of social
interaction is necessary for the good or adequate
development of the baby and young child.
Some theorists have suggested that the baby is born
social, is capable of intersubjectivity (the sharing
of meaning with another person) very early in life
(Bullowa 1979; Trevarthen 1980). What is meant by
this hangs on what is meant by meaning and by
intention which are not at all simple concepts
(Trevarthen 1982; Kaye 1984). The very early ability
of the baby to express emotion in ways which affect
the adult is seen by Trevarthen 1978 as primary
intersubjectivity. Other observers prefer terms like
protointersubjectivity (Hinde 1979), wishing to be
cautious about attributing communicative intention
to a baby of 2 or 3 months old, and also preferring to
explain behaviour in the simplest possible terms. The
babys emotion-laden signals may be intended to
communicate with other people, or they may be an
automatic expression of the babys feelings which
other people respond to as if communication was
intended, though at the time there was no intention. It
will be hard to tell which of these is the case, since as
well as the perennial difficulties of knowing anyones
intentions we have to cope with the difficulties of
working with very young organisms, the baffling
mixture they exhibit of psychological immaturity and
readiness for a mental life in the company of other
persons (Trevarthen 1982, p. 77).
There seems to be disagreement about the best way
to conceive of the earliest stages of social
understanding and interaction in babies; there is,
however, substantial agreement that parents
acceptance of babies as social partners and their
consequent framing or scaffolding of joint
activities and conversation form an extremely
effective context for infants to learn about
themselves, other people, and the outside world.
Current theory emphasizes the contribution which

other people make to every aspect of a childs
development, redressing the underemphasis of this
contribution which characterized Piagetian and other
theories (see Butterworth and Light 1982; Lamb and
Sherrod 1981, etc.).
Attachment
Clinical models and theories of the development of

emotion and personality have consistently seen the
early social relationships of the child as of major
importance. Early good experience and in particular
the establishment of strong emotional ties
bonding to a mother figure who provided good
care were seen as crucial for healthy later life. The
baby, who was at first not notably choosy about
people, comes to show preferences for one person
over another; by the age of 7 months or so these
preferences amount to a strong positive feeling for
one or a few familiar caretakers, comforters and
playmates, and a negative feeling of fear or caution or
lack of interest for unfamiliar others. This differential
positive feeling has been called attachment.
Attachment is, centrally, the affectional bond or
tie that an infant forms between himself and his
mother figure (Ainsworth et al. 1978, p. 302). This
bond is seen as providing the baby with affection and,
most of all, security. It is inferred from the babys
attachment behaviour, that is, from items of
behaviour that share the usual or predictable
outcome of maintaining a desired degree of
proximity to the mother figure behaviour through
which the attachment bond is first formed and then
later mediated, maintained and further developed
(Ainsworth et al. 1978, p. 302). Among the relevant
items of behaviour are enthusiasm in greeting the
mother figure, distress on being left by her (or him),
following him, seeking her attention, showing less
anxiety and more confidence in his presence and fear
in her absence. Babies usually become attached to the
person who gives them attention, opportunities for
joint play, and pleasantly intense social interaction,
particularly if that person is responsive, co-operative
and sensitive. Provision for the babys physical needs
seems to be less important, and bottle-fed babies

become just as attached as breastfed ones. Bowlby
(1977) believes that attachments are normally
formed to people who are seen as stronger and wiser,
and that babies attachments function to protect them
from danger (including being left unprotected from
predators, since he has taken an evolutionary
perspective on the development of attachment,
though he was also influenced by psychoanalytic and
systems theories). Being near to a mother figure, and
becoming anxious if separated from her, would be
adaptive, particularly if there were complementary
feelings and actions in the mother figure. The
biological basis of the model and especially its
synthesis of ethology, systems theory and
psychoanalysis contributed to its considerable
impact on theory and on practical recommendations
for the care of young children.
There can be no doubt, first, that people (including
children) do form attachments to other people
(including their mother figures), and, second, that
these attachments are important both for the present
happiness of the individual and for his or her future
prospects. As in the case of maternal instinct the
theory has brought about some welcome changes in
how young children and their parents are treated by
professionals. A great deal of interesting and
important research has been done on attachment (see,
for example, Murray Parkes and Stevenson-Hinde
1982). There are, however, a number of problems
which need to be worked out in defining attachment
and adequate mothering.
As can be seen from the short list of items of
behaviour which was given in the last paragraph but
one, attachment is seen in a rather varied range of
actions. An even wider range could have been
presented: Main and Weston (1982) seem to be
willing to include angry behaviour like tantrums
because within an established attachment
relationship they may serve the function of increasing
proximity between partners an infants tantrum
may persuade the mother either to approach him or to
permit his approach; in turn, a mothers angry
behaviour often brings her infant towards her (Main
and Weston 1982, p. 33). It is clear that attachment
177
behaviour is heterogeneous and thus attachment is

not a simple unitary concept. It might be possible to
measure each type of attachment behaviour
(minutes of crying, number of frowns or smiles, how
much physical effort is put into removing the barrier
between oneself and ones attachment figure); and it
might be possible to assess the strength of each (more
indicates stronger); and these strengths might
correlate highly positively, so that one could
amalgamate them into an overall measure of strength
of attachment. However it has turned out in a large
number of studies with a variety of species that the
different behaviours may not be correlated and that
they vary from time to time and situation to situation
(Ainsworth 1982; Hinde 1982; Rutter 1981). This
complexity and variation has been seen as
undermining the usefulness of the concept of
attachment: reacting to such criticisms, attachment
theorists like Ainsworth (1982) and Sroufe and
Waters (1977) emphasize the need to look not at
specific attachment behaviours but at patterns of
behaviour, which are much more stable. Ainsworth
has developed a typology of patterns of attachment
displayed by infants towards their mothers. They
were assessed using the strange situation paradigm,
where the children successively experience playing
in a strange room in their mothers presence, being
left alone there, her return and repeated departure, the
entry of a stranger and so forth. The largest and
normative group of babies explored actively when
their mother was present, were upset at her departure
and stopped exploring, and showed a strong interest
in interacting with her, mostly seeking close bodily
contact, when she returned. These babies, who were
distressed by their mothers absence, confident in her
presence and whole-hearted in their greeting of her on
her return, were said to be securely attached. A
small group of babies were anxious before
separation, very upset during it, and ambivalent
during the reunion when they both sought and
resisted contact. A third group of babies showed little
distress at the separation and avoided contact or
closeness with the mother on reunion; some ignored
and some avoided her. Observed at home, their
behaviour was rather similar. The first group,
178
securely attached, cried least, were least anxious

and unco-operative, and had mothers who were
positive, sensitive and encouraged close physical
contact. The other two groups were anxiously
attached, cried more, showed more general distress,
and were negative about close physical contact. Their
mothers were less sensitive to all the babys signals;
the mothers of the second group, the ambivalent
babies, were warm but highly insensitive so that their
warmth came at inappropriate times, the mothers of
the third, avoidant, group were (relatively) cold,
angry and rejecting (Ainsworth 1982, pp. 1617).
Differences like these are said to persist, and to be
associated with differences in the childs social,
emotional and intellectual behaviour, over the first
five years of life (Sroufe et al. 1977; Sroufe 1983).
Describing relationships
We will return to the question of long-term effects of
parentchild relationships later: there are important
and delicate questions to be asked about how any
effects are caused and what cognitive or affective
systems are involved. It must also be recognized that
the mothers side of the relationship needs more
consideration than it has received and that the
ecological context is important. Variation in the
babys behaviour in different settings and towards
different people (Lamb 1978), and the approach of
defining attachment behaviour in terms of whether
they usually or predictably bring about proximity
between baby and mother figure, imply that we must
consider not just the actions of the baby but also the
actions of the person he or she is seeking proximity
with: in other words that attachment is a property of
relationships not just of individuals, and particularly
not just of the baby who is by far the more intensively
studied partner (Hinde 1979, 1982). Most children
develop multiple attachments (Rutter 1981) and
these differ somewhat in intensity, function and
content. We need to consider how we describe and
evaluate relationships.
Robert Hinde (1978, 1979) has suggested a number
of dimensions and principles which may prove useful
in constructing a science of inter-personal
relationships. His model is complex, as the subject

requires, exploratory, as our present state of
knowledge requires, and presented with
characteristic care and wisdom. My discussion of it is
necessarily brief, and is centred on the development
of relationships in childhood. A relationship
involves interactions which happen over an extended
period of time, have some degree of continuity, and
involve each participant taking account of the
behaviour of the other. Thus I would probably not
have a relationship with the shop assistant who sold
me a dress as it would be (unless the dress had to be
returned or it was a favourite shop) a one-off
interaction with no personal past or future for either
of us. I do have a relationship with my (admirable)
secretary because our interactions have happened
over a number of years and what we do during each is
affected by our accumulated history of working
together, so that we each take account of what the
other does now, has done in the past and will probably
do in the future: and of what we ourselves do, have
done and will do.
A relationship has properties which apply to the
relationship, not simply to either partner: it involves
behaviour which (usually) has meaning to the
participants. In describing this behaviour we must
make reference to content (e.g. kissing), to quality
(e.g. kissing tenderly or passionately) and to the
patterning of behaviour. To use Hindes example
(1979, p. 20) Clearly the relationship between a
couple who always kiss after they quarrel will be very
different from that between a couple who always
quarrel after they kiss, even though the total amounts
of kissing and quarrelling are the same in both cases.
Relationships always take place in a social context,
and are affected by the participants social past
outside the relationship and by their other
contemporaneous relationships; and thus by many
different levels of social structure, as Bronfenbrenner
also points out. It is important to distinguish between
social behaviour (which is usually studied
quantitatively) and social relationships (where
quality may be more important); behaviour and
relationships may be related but they are not
identical. Finally, relationships involve not just overt

behaviour but expectations, goals, values, feelings,
assessments,
interpretations,
memories,
categorizations and norms: the participants are active
agents not just passive subjects, and their
understanding of the relationship may be as
important as what really happened.
In describing a relationship, then, we must attend to
the content and the diversity of the interactions within
it; to aspects of quality such as intensity, the sorts of
communication by linguistic and other means
involved, and the meshing of the participants
behaviour; and to the relative frequency and
patterning of interactions. It may be worth looking to
see whether interactions involve reciprocity, so that
the participants show similar behaviour either
simultaneously or alternately as in childrens games
of rough-and-tumble, or whether the interactions are
complementary, as in mealtime interactions between
mother and baby. Relationships also vary in intimacy
or the degree to which participants are willing to
disclose their personal secrets to each other, in
commitment to the relationships duration or content,
and in the relations between the views held by the
participants about the relationship, themselves, each
other and the outside world. These characteristics too
need examination.
Parentchild relationships
It will be clear by now that a science of inter-personal

relationships will be a very complex thing; it will
become clear that there are as yet no studies that
describe, explain or predict relationships or their
effects in the detail that Hinde is proposing on
anything but a very small sample. (Outside
conventional psychology, some biographers
approximate this sort of description: see, for
example, Bate (1975) on Samuel Johnson, Thwaite
(1984) on Edmund Gosse, Rose (1985) on five
Victorian literary marriages.) There is, however,
interesting research which bears on some of the issues
raised, and an account of some of this follows. I
would like first, however, to point out some of the
179
implications of Hindes sketch of a taxonomy of

relationships for the idea that the social experience of
infants and very young children determines or is a
model for their later social relationships. The parent
infant relationship has a number of characteristics
which differentiate it from all other interpersonal
relationships. In the first place, there are marked
cognitive inequalities between parent and infant.
Certainly even very young babies show some
remarkable cognitive and social achievements (see
Chapter 2 and the earlier part of this chapter) and they
develop even more impressive ones very fast;
nevertheless on the whole parents are much more
experienced, knowledgeable, skilled, deliberate and
generally capable than their children for at least the
first few months or years! The parentinfant
relationship is thus more unequal than most child
child or adultadult relationships, especially since
status and power differences are added to the
cognitive ones. Consequently, most parentinfant
relationships involve complementary rather than
reciprocal interactions, as we have seen. The content
and diversity of interactive behaviour will be greater
for the parent than for the infant, as will the range of
communication techniques that may be applied; the
meshing of parents and childs behaviour will be
largely under the control of the parent. Intimacy, in
the sense of deliberately disclosing ones inner self to
another person, and deliberate commitment are
irrelevant to the very young, and the babys and
parents capacity for having views of each other as
persons and of the relationship as a relationship are
markedly different. These differences between
participants are vastly greater than in other
relationships, and although the baby learns fast, the
differences remain considerable for some years.
Moreover, the fact that the baby is changing as time
goes by inevitably means that the relationship
changes as to content, quality, patterning and all the
dimensions I have outlined. Precisely the recognition
of this change and the provision of appropriately
changed behaviour is an important part of being a
180
good parent; indeed, I have suggested that if parents

wish to produce precociously competent children,
what they need to do is slightly overestimate
childrens maturity, providing them with
opportunities for achievements a little beyond their
present ones, and scaffolding them as they extend
their accomplishments. This is part of the teaching
models of Vygotsky (1978) and Bruner (1967) and of
the effective behaviour documented in various areas
by, among others, Kaye (1984) and Wells (1985).
It may be seen that relationships are multi-faceted
complex things; that the parentchild relationship
differs from most other relationships so profoundly
that we might have reservations about taking too
literally any suggestion that it determines later
relationships; and that it is potentially (and in most
cases actually) an important source of learning for the
child (and, incidentally, for the parent), which
implies that it may have important effects on how and
what children learn and so on their lives beyond the
parentchild milieu. I want next to look at what we
know about the effects of different sorts of parent
child relationship, starting with the relationships of
parents and infants.
Early experience and later effects

There has been a long tradition of theory which
proposes that childrens experiences with their
parents determine much of what they are like for the
rest of their lives. Early experience was seen as
particularly important. Weaning and toilet-training in
classic Freudian theory, the neonates experience of
the breast in Melanie Kleins model, the
establishment of a strong and uninterrupted bond
with the mother in Bowlbys 1951 monograph, more
widely the view traditionally attributed to the Jesuits
that the course of the first seven years of childrens
lives established them forever, are examples. All
these accounts of development placed its most easily
influenced time right at the beginning and implied
that change was difficult (or impossible) thereafter.
This is a strong hypothesis which has turned out to be
false, or to be correct through more subtle sequences

of causation than were assumed.
Methodological and conceptual problems
The importance of early experience in terms of later

outcomes is hard to assess because of the problems of
first defining and second explaining continuity (as I
discussed in Chapter 1). It is not always obvious what
outcome or range of outcomes a theory implies that a
particular early event should be expected to bring
about; the Freudian childs reaction to severe toilettraining, for example, might range, via differential
success with various defence mechanisms, from the
extremes of anal-expulsive messiness and
selfishness to the extremes of anal-retentive tidiness
and obsessive orderliness. Similarly the forms taken
by anxiety at the age of 2 may differ from the forms
taken by anxiety at the age of 32. Further, and perhaps
more importantly, the long-term effect attributed to
an early experience may be due not to that experience
itself but to the continuing effects of its short-term
consequences, or to a pervasive chronic problem
which caused both the acute early event and the
longer term effects which are attributed to the acute
event itself. For example, it has long been known that
children whose parents divorce are more likely to
have problems such as delinquency, poor educational
achievement and disturbed social relationships.
Some explanations of this centred on the traumatic
effect on the child of being separated from one parent
(and parents who were contemplating divorce were
urged to stay together for the sake of the children).
It now seems to be a much more viable alternative
explanation to attribute the childs behaviour
problems to the discord in the family which preceded
and accompanied the divorce and separation, and to
the loneliness, impaired finances and changes in
disciplining and affection which the divorce forced
on the single parent left in charge of the child (Rutter
1981; Hetherington 1979). Distinguishing between
different models of causation is important because
their implications for prevention and treatment are
different, as the example of divorces effects on
children makes clear.
Maternal deprivation
Much of the debate of the last two decades on

questions of the effects of early experience centred on
Bowlbys account (1951, 1969, 1973, 1980) of what
has come to be known as maternal deprivation.
Rutter (1981) provides an authoritative review of this
debate. Evidence has accumulated to show that
Bowlby was right to argue that deprivation and
disadvantage have important influences on
childrens psychological development: we will
examine some of the ways in which this happens later.
Deprivation, however, takes many different forms
and these have different effects via quite different
psychological mechanisms; some of the bad effects
are rare, some common; some are irreversible and
some quite easily made up for; some remain severe,
some gradually wear off. Individual differences in
reaction to stresses and disadvantage are a new and
illuminating area of study (Rutter 1978, 1981; Rutter
and Garmezy 1983; Garmezy and Rutter 1985).
Emphasis on the early months and years as a period of
particular vulnerability is also being re-examined,
and the extent to which continuity of effect persists
despite discontinuity of experience is providing some
very interesting findings (e.g. Tizard 1977 on late
adoption; Clarke and Clarke 1976; Doyle et al. 1984).
Child-rearing techniques
A very considerable amount of research has been

done on the effects of parents child-rearing
techniques on childrens personality and
achievement. This work has to be interpreted with
great caution and consideration of its theoretical
bases, methodological soundness and freedom from
cultural, historical or ideological bias. Caution is all
the more necessary because parents are often judged
or advised about how to bring up their children. Some
research base to this advice is probably better than
none, but only if the research is fairly sound.
(Hardyment (1983) reviews advice books from
Locke to Spock: a salutary story.)
One common problem has been theoretical or,
more forgiveably, methodological, focusing on
parent effects on children as if one could expect a one-
181
way influence. On the contrary, it is quite clear that

differences in actual or imagined characteristics of
the child call forth different behaviour in the parent.
Sex differences represent one rather intensively
studied example (see p. 196). A more specific
example comes from a study of motherchild
prelinguistic communication in normal children and
children suffering from Downs syndrome (Jones
1979). Mother child pairs were carefully observed
while playing at home with a supplied set of toys. The
Downs children were as much involved in
interaction and vocalization as the normal children
but they phased their activities in ways which made it
difficult for their mothers to take turns. Their mothers
were correspondingly more directive and restricted.
The result of this was that the Downs children
seemed to be providing themselves with an
environment which was less stimulating in both
social and cognitive terms which might be expected
to contribute to the retardation of language and
cognition usually seen in Downs syndrome children.
Jones makes the crucial point in her discussion. It is
noteworthy that these subtle communication
difficulties were only brought to light when the
motherchild interactive context was taken into
consideration (Jones 1979, p. 194).
Not only do different children call forth different
behaviour from their parents because they
themselves behave differently or are classified
differently, but each side may interpret the same
behaviour differently according to who is doing it.
Again sex differences provide a wealth of examples:
one of the most telling is the Newsons data on
parents treatment of aggression in boys and girls at 7
years old (Newson and Newson 1976, see also my
discussion of the development of aggression
(Chapter 5) and of sex differences (pp. 1958)). This
interaction of behaviour and interpretation may be
very complex indeed: Schachter (1982), for example,
speculates that in families with more than one child
siblings may try to be unlike each other (sibling
deidentification) and one child will identify with
each parent (split-parent identification). While this
is a useful counterweight to the commoner
hypothesis that it is normal and right for children
182
to identify primarily with the parent of the same sex,

all else being risky if not pathological, there is so far
very little evidence on the validity of such hypotheses
and almost none which deals with behaviour rather
than interview or questionnaire data. Identification
as a concept is examined in Chapter 5 and is seen to
have its difficulties.
We know little about how far behaviour is
consistent over time or different situations. There
may be inconsistency in parents behaviour and in
childrens, and in their interpretations: and this
inconsistency may itself be consistent or
inconsistent. For example, it may be that what parents
expect children to do, and how they deal with
transgressions, is dependent on the setting. The
Newsons Nottingham parents said that they
permitted more aggressive behaviour in the privacy
of the family than they did in public places,
particularly for girls (Newson and Newson 1976).
There may be inconsistency over short periods of
time; 5-year-olds are often more tired and more
whiney at the end of the school week; a mother who
suffered from premenstrual tension (Dalton 1983)
might be more irritable and more restrictive in the few
days before her period than she was at other times,
and her children might adjust to Mummy being in
one of her moods again. Consistent inconsistencies
like these pose problems for researchers seeking to
classify parents child-rearing techniques (or indeed
any other sort of behaviour) on the basis of limited
amounts of data, but are very much part of what we
need to learn to be adequate social beings. They seem,
as we shall see, to pose much less of a problem for
children than inconsistent inconsistencies where
what is done is so unpredictable that the child cannot
learn from it.
Inconsistency on a longer time-scale is to be
expected on various levels. It is obviously necessary
that different discipline techniques should be used on
children of, say, 12 months, 4 years and 14 years.
Inconsistency in effects of parenting may also be
found over a long time-scale, as children experience
other influences or change themselves. There have
been few studies with a time-scale longer than early
childhood to pre-adolescence, so there is little data
beyond biography. The exception is work on children

who were parented particularly badly: we will look
at this later.
Most of the data we have on parenting techniques
and child behaviour comes from studies done on
samples limited as to nationality, class, race and
historical period. It is not at all clear how far what is
found with one group may be true of other groups. It
is clear that we risk being egocentric in our
interpretation of what we see and in what we
prescribe as good parenting. John Ravens account
of a Home Visiting project in Edinburgh (Raven
1980) illustrates this. Lower working-class mothers
were encouraged to interact with their pre-school
children in middle-class ways. This experience
changed what the mothers believed it was possible to
do with children to encourage their cognitive
development, but not what they felt they could
actually do themselves: their confidence in their
mothering skills was diminished. We need to realize
that some part of what is good parenting depends on
what society outside the family allows to and
demands of the child and family.
Societys ideas about parenting will introduce
methodological problems in at least two ways. One is
by the biasing of the researchers judgement and
observation: as well as misinterpreting behaviour we
may simply not see what we think is unimportant. The
other is by the biasing of parents behaviour towards
social acceptability. This must be recognized as a
potential problem in both observational and
interview/questionnaire studies. What we actually
do, what we think we do, and what we say we do may
be different things.
It must finally be said that although there are
statistically significant correlations between various
parental practices and various childrens
characteristics, the correlations are not, typically,
large. As in the case of maternal deprivation, some
children do well despite horrific parental behaviour,
and vice versa. The more we know about other
characteristics of child and parent and about the
childs experiences outside the family, the nearer we
may get to understanding this variation in outcome.
As far as research goes, we will need to study the co-
variation of many variables in large samples; as far as

recommendations on parenting go, we must be
cautious and not doctrinaire.
With all these caveats about models, variation,
interpretation, interaction, inconsistency, limited
time spans and limited samples in mind, we can
proceed to look at what variations in parenting are
associated with what child outcomes. The main
dimensions investigated have been parental
warmth or responsivity and parental discipline
techniques.
Warmth and responsivity
Warmth is hard to define, and is certainly not a

simple single trait. In most definitions it involves
parents being deeply committed to the childs
welfare; responsive to his or her needs and actions;
willing to become involved in joint activity with the
child, especially activity that stems from the childs
interests; enthusiastic about the childs achievements
and virtues, and sympathetic and helpful about his or
her difficulties and failures; and sensitive to the
childs emotional needs. No-one could possibly be all
these things all the time, so warmth is relative.
There is a tendency for children whose parents are
high on warmth to be relatively affectionate and
sensitive themselves, to be willing to comply with
reasonable commands, to be securely attached, to be
altruistic, and to have good opinions of themselves, in
other words to show a high degree of positive social
behaviour (Maccoby 1980; Rollins and Thomas
1979; Shaffer and Brody 1981). If warmth is
notably lacking in the parentchild relationship there
is a tendency for children to show the opposite
characteristics and an increased probability of a range
of difficulties (Rutter 1981). We will consider the
pathological extreme of lack of warmth later (p. 185).
A paper by Wadsworth and Wingfield (1986/
forthcoming) reports findings from the unique
second generation part of the MRC National Survey
of Health and Development of a cohort of 5362
people born in 1946. Studied at intervals through
their lives so far (Atkins et al. 1980), they have been
studied as parents as they have produced their first-
183
born children. 1684 children who are the oldest child

of a member of the 1946 cohort form the second
generation. This work has allowed studies of
continuity and change between generations and of
very long-term effects of parenting.
Interview data from the mothers of the second
generation children includes their description of their
emotional relationship with their 4-year-old children.
The 30 per cent of mothers who said they had
reserved or one-sided relationships with their
children were much more likely to say that they
disciplined their child by threatening withdrawal of
love, physical punishment or separation from mother.
They took part in pretend play less, they read or told
stories to the child less, they tolerated the childs tall
stories less, they less often told the 4-year-old
truthfully where babies come from. They less often
sent the child to pre-school playgroup or nursery, and
they more often described the child as highly strung
or backward. When the children were tested at the
age of 8, their scores on tests of reading, sentence
completion and vocabulary were significantly lower
than those of children who had had more
demonstrative and more stimulating relationships
with their mothers at the age of 4. Some of the
difference could be accounted for in terms of the
lower educational level of the less demonstrative
mothers, but even after this had been statistically
controlled through multiple regression analyses,
affection and verbal stimulation were still of
significance for the verbal test scores. Some
information was gathered on the childrens
friendships at the age of 8: children from reserved
relationships were much more likely to be unpopular
at school and not to have any friends. Reserved
mothers were more likely to be rather young (1719
at the birth of their first child), to have had some
hospital treatment for emotional disturbance and to
say that their own childhood had been unhappy. Lack
of warmth seems in this large sample to be associated
with under-stimulation, lower achievement and
difficulties for the child in getting on with other
children. The differences are not large and many
children did not seem to be much affected, however.
184
Discipline and control
The other much-studied dimension of parenting is

discipline or control. Baumrind (1971, 1980)
distinguished three major patterns: authoritarian
behaviour which tended to be coercive, rigid,
intrusive and punitive; permissive, which placed
minimal constraint on the child, avoided controlling
him or her preferring to allow freedom and selfactualization; and authoritative, which balanced
conformity and independence, encouraging
discussion and negotiation in the context of firm
standards. The findings may be summarized as
follows:
Daughters of Authoritative parents tended to be socially
responsible, as well as independent, whereas sons, though
socially responsible, were no more independent than
average. Authoritarian styles, meanwhile, were associated
with less achievement orientation and independence in
girls and more hostility in boys. Somewhat curiously,
similar patterns emerged among children of Permissive
parents. Daughters of Nonconformist parents resembled
daughters of Authoritarian and Permissive parents,
whereas the sons of Nonconformists were significantly
more independent and achievement-oriented. Baumrind
has speculated that the similarity between the children of
Permissive and Authoritarian parents is due to the fact that
both types tend to shield their children from stress and thus
inhibit the development of assertiveness and frustration
tolerance. By contrast, Authoritative parents value selfassertion, willfulness, and independence and attempt to
facilitate childrens attainment of these goals by assuming
active and rational parental roles. Their children, on the
whole, are socially responsible because their parents
impose demands that are intellectually stimulating (that is,
their expectations are demanding and clearly
communicated but not unrealistic), as well as moderately
tension producing (inasmuch as firm discipline necessarily
results in occasional clashes of will).
The findings discussed here dramatize the fact that it is
not particularly valuable to consider isolated parental
attributes like punitiveness, warmth, or control. The effects
of these attributes are only evident when we consider
complex patterns of attributes. In other words, the effects of
firm control, for example, can be understood only when we

know more about parents warmth and punitiveness (Lamb
and Baumrind 1978, pp. 57, 59).
Results like these may be looked at in terms of

underlying characteristics of behaviour instead of in
terms of types of parenting. Rollins and Thomas
(1979) review a large number of studies. One
important dimension is, they say, parents
supportiveness, because supportive parents provide a
responsive and facilitating environment for the
childs activity. Inductive control, which gives
information about causes and reasons and
encourages children to understand the world and the
consequences of their actions is also important.
Coercive control and punishment has bad effects
because it decreases responsivity, the childs
effectiveness and communication. Sociologists
studying families as systems (e.g. Garbarino 1982)
produce similar lists of characteristics of wellfunctioning families: mutual affiliation and
affection, open communication, flexible structure
balancing individual and family needs, spending time
and energy on family matters and joint activity.
It might be a carping criticism to say that this
picture, drawn mainly from work with middle-class
Americans, resembles the cosy ad-mans dream.
More fairly, it must be seen as limited in both
historical and cultural terms, a model probably not
much seen outside twentieth-century western
society. Nevertheless, it is probable that families
which have the reverse characteristics may fairly be
called badly-functioning. Both adults and children
living in families with poor communication, hostility,
low commitment, and generally negative relations
are likely to have social, emotional, educational and
employment difficulties, at least in the context of a
society which subscribes to the authoritative family
as ideal. The British evidence (Rutter 1980a; West
1982, 1985; Wadsworth 1979) suggests that boys
from hostile, disrupted, disordered, disadvantaged
families, which provide models of anti-social
behaviour in other family members and where
discipline and surveillance of the child is lax, are
particularly likely to become not merely delinquent

but recidivist or involved in serious violent crime, or
to avoid delinquency only by having such severe
difficulties in getting on with other people that they
are almost completely withdrawn from any social
activity.
Very poor parentchild relationships
A majority of parents who batter their children had an

unhappy, rejecting and cruel upbringing themselves
(Kempe and Kempe 1978; Rutter and Madge 1976),
though only a minority of children from battering
families grow up to be battering parents and the
evidence is that family members vary in their
parenting style almost as much as the general
population (e.g. McGlaughlin et al. 1980, 1983). It is
unlikely, too, that unhappy families have their ill
effects simply because they provide maladaptive
models, or because their stresses tend to induce
psychiatric disorders in their members, or because
they are unstable, or because they lack resources
(material, economic and political as well as social)
and connections outside the family; no single cause is
sufficient. It is likely that there is a chain of causation
whose details will differ from case to case (Madge
1983): indeed, to quote the first sentence of Anna
Karenina
All happy families are alike but an unhappy family is
unhappy after its own fashion (Tolstoy, translated R.
Edmonds 1954).
We are only just beginning to tease out how it is that

disadvantage, deprivation and dysfunction arise,
recur and are avoided. It is a complex issue on both
methodological and conceptual levels, but of
immense social importance. Among the important
British sources of information are Rutter and Madge
(1976), Wilson and Herbert (1978), Coffield et al.
(1980), Essen and Wedge (1982), Madge (1983),
Brown and Madge (1982).
Very poor parentchild relationships seem to be
likely to set in train a complex sequence of
185
disadvantaging events which may, if not broken by

good fortune, lead to serious disturbance when the
child grows up, including difficulties in secondgeneration parenting. Removing the child from the
disturbed family might be seen as one way of
interrupting the sequence. With the possible
exception of adoption, which seems to be relatively
successful even in the case of quite old children
(Tizard 1977; Hersov 1985), this cannot be seen as a
panacea. Rutter, Quinton and Liddle (1983) found
that poor mothering and generally poor psychosocial
functioning in early adult life were strongly
associated with being reared in an institution; a
harmonious relationship with the father greatly
ameliorated mothering behaviour, but girls brought
up in care were particularly likely to become
pregnant early and to marry men who themselves had
psychosocial problems. Their samples were small
and all the girls came from one Inner London
borough, but the research was very carefully done. It
is not possible to say whether problems are caused by
the disorder and difficulties which originally led to
children being removed from their families, or to
subsequent deficiencies in the institutions or foster
parents ways of rearing the child, or to purely
administrative factors such as the abrupt termination
of support when the child reaches adulthood
(Jackson, personal communication). It is all too
common for children to be brought up in a Home
where they never took decisions about any aspect of
their own lives or even boiled the proverbial egg
and to be expected to function independently the
moment they leave it. Part-time removal from the
family, into a daycare group or the care of a childminder, is likely to help but to leave the child still at a
disadvantage (Clarke-Stewart 1982; Tizard 1974). In
part this may be because the supply of daycare for
children is much less than the demand (Bone 1977;
Hughes et al. 1980), and many providers are very
poorly paid; not surprisingly the quality of provision
is in many cases worryingly low (Jackson 1979;
Bryant et al. 1980), and even in the most privileged
sector, nursery schools, not as good as might have
been hoped (Sylva et al. 1980; Meadows and
Cashdan 1983). Parents who are already having
186
difficulties may find it impossible to get good daycare

for their child or to make the best of what is found.
The Child Health and Education Study reports on the
pre-school experience of a national sample (Osborn
1984). Their results show that it was precisely the
most disadvantaged families who were most likely to
get no pre-school provision at all for their child (16.6
per cent to 2 per cent of the most advantaged group);
similarly children living with both natural parents or
two adoptive parents, and children from small
families, were much more likely to have pre-school
experience than children from other sorts of families.
Direct support to mothers may be a more effective
method of intervention (Bronfenbrenner 1976).
Daycare
The effects of daycare on children have been

investigated but there are limitations in the research
which make drawing conclusions rather dangerous.
The daycare institutions studied have often been of
above average quality, and comparisons of home and
pre-school are made difficult because little is known
of home experience (but see Davie et al. 1984, and p.
191 below) and because children cannot normally be
assigned randomly to home care and daycare:
children are in a pre-school group because of the
choices of their parents or other selective factors
which make them different in unknown ways from
children at home. Nor is it obvious what measures of
effect are appropriate: IQ? behaviour problems?
present happiness? later school performance?
Clarke-Stewart (1982) and Clarke-Stewart and Fein
(1983) review the research done; children who have
attended good nursery schools and similar groups
show advanced cognitive skills on entry to school,
though the effect may wash out, and they are more
socially skilled with their peers than home-reared
children. They are still attached to their mothers, but
somewhat more independent and boisterous. It is
probably the case, then, that it is a good thing for
everyone concerned if the nuclear family rearing
which dominates 1980s Britain is supplemented or
partially replaced by group experiences (see also
Weikart 1978). Experience with adults and children
outside ones own immediate family has after all been
the usual pattern at most historical times and in most

places.
Child effects on parents
Developmental studies of the socialization of children

in the family must be concerned with the sequence of
normative age changes, with individual differences
within each norm, and with how both general and
idiosyncratic sequences come about. As I have
pointed out, we must accept that while parents do
socialize children, children also socialize parents.
Also, there are changes in the child (and quite probably
in parents too) as the child gets older which are
contributors to the socialization process rather than
results of it. As they grow up children get bigger,
stronger and better co-ordinated, for example. As a
result, they pose different control problems at
different ages. Almost all 2-year-olds can be outrun
and overpowered by their parents; most 8-year-olds
can be overpowered but many cannot be outrun, so it
is a matter of first catch your child; most 16-yearolds cannot be easily overpowered or outrun. Parents
haveto changetheir disciplinetechniques appropriately,
and adjusting to changes in the child by changing
their own behaviour is one of the important general
principles that successful parents follow. Physical
growth is relatively free from parents influence after
birth, though I cannot resist a quotation from Dickens
illustrating how parents have influenced it to their
own advantage. The subject is Miss Ninetta
Crummles, from the theatrical family which Nicholas
Nickleby joins:
. . . the infant phenomenon, though of short stature, had a
comparatively aged countenance, and had moreover been
precisely the same age not perhaps to the full extent of the
memory of the oldest inhabitant, but certainly for five good
years. But she had been kept up late every night, and put
upon an unlimited allowance of gin-and-water from
infancy, to prevent her growing tall, and perhaps this
system of training had produced in the infant phenomenon
these additional phenomena (Nicholas Nickleby, ch. 23).
She had also, of course, had her motor development
influenced by her parents training in ballet; sex

differences in physical skills before puberty, and to a
lesser extent after it, can be related to amount of
practice, and variation in strength between
individuals also derives in some measure from
variation in experience. Parents often react to signs of
desirable talents in the child by providing increased
opportunities for them.
Other developmental changes seem to be more or
less universal, among them language development
and improvements in communication (see Chapter
4), metacognitive processes and a decline in
impulsivity (see Chapter 3), a decrease in
dependence and an increased demand for autonomy.
It is not clear how far these are products of
socialization pressure (or of biology, Trivers 1985);
Maccoby (1984, p. 325) suggests that within large
limits these changes are surprisingly independent
[sic] of the way parents treat their children. Whether
or not a change is a result of socialization it must
influence and should change parents treatment of
their children. A decline in impulsivity and an
improvement in communication skills, whether or
not they result from earlier parental practices, will
themselves make direct physical control of the child
less necessary because they allow the use of verbal
discipline techniques such as prohibition and
reasoning. Parental socialization techniques coexist,
that is to say, with developmental changes which
have maturational, self-stabilizing, components; and
also, of course, with the socialization techniques of
other people. Childrens relationships with other
children form a lively new area of study. I will discuss
sibling relationships first, and then peer relationships
outside the family.
Sibling relationships
Most children grow up in a family which contains
other children, brothers and sisters. Often these
siblings are an important part of the childs life, both
because children may spend at least as much time
with their siblings as with their mothers, and more
time with siblings than with their fathers (Dunn
1983), and because relationships between child and
187
parent are likely to be affected by the relationships

that each has with other family members such as
siblings. We have only a small amount of data on what
siblings do together, even less on what developmental
effects sibling interaction may bring about, and very
little in the way of theory about sibling relationships.
This unfortunate state of affairs may change as
researchers get to grips with the complexity of mutual
influences within the family. The necessity of doing
so is highlighted by the findings of behaviour
geneticists such as Sandra Scarr. Although siblings
share an average of 50 per cent of their genetic
material and many aspects of the family
environment, they differ in personality, intelligence
and most sorts of psychopathology almost as much as
unrelated people do (Scarr 1983; Scarr and Grajek
1982). This suggests that we need to examine both
how a family environment differs for the different
members of a family and how siblings affect each
other. Various theorists (see Lamb and Sutton-Smith
1982) have suggested that siblings may often try to be
as different as possible. Schachter (1982) for example
calls this sibling de-identification; she also
proposes split-parent identification, an ugly term
referring to the possibility that if the first child
identifies (or is identified with) parent I the second
child will identify with parent II. What we need here
is both good data on what actually happens, and a
more precise understanding of both identification
(see Chapter 5) and family dynamics. It would seem
to be possible that within families, as within other
social groups, a range of roles are available, and
individuals may have some choice in their ecological
niche-picking.
One useful distinction made in Hindes taxonomy
of relationships (Hinde 1979; see p. 179 above) was
between complementary and reciprocal activities.
Earlier I argued that adultchild relationships, being
in so many ways unequal, will be weighted towards
complementary interaction; the activities of the adult
differ from those of the child and the two partners
complement each other. Relationships between
peers, more alike in competence and status, are likely
to involve more reciprocal interaction where the
partners do the same thing together or in some sort of
188
turn-taking. Siblings interaction can be examined in

terms of this distinction (Dunn 1983), and to do so
illuminates some distinctive qualities of sibling
relationships and their developmental effects.
Complementary interaction
Particularly in non-western societies, and in sibling

groups with a large age-gap between the members,
older children may be the caretakers for their younger
brothers or sisters (Whiting and Whiting 1975). Child
caretakers are responsible for much of the nurturing
and socialization that babies and young children
receive. Even in our society, it is common for older
brothers or sisters to be informally responsible for
keeping an eye on younger siblings, protecting
them in the school playground, fostering their dbut
in the adolescent social world, and so forth. Recent
observational studies have shown that most young
children are concerned and helpful about their
younger siblings (Dunn and Kendrick 1982) and the
younger siblings are attached to the older ones
much as they are to their parents (Schaffer and
Emerson 1964). Children as young as 3 adjust their
speech when talking to their baby sibling in ways
which would be likely to bring about better
communication (Dunn and Kendrick 1982; Shatz and
Gelman 1973, 1977); they exaggerate their
intonation, use simpler sentences and give more
repetitions and explanations. These are some of the
features of motherese (see Chapter 4 on language);
what siblings rarely used from the motherese register
were the various language elicitation behaviours
such as questions and scaffolded dialogues. In part
this was because at least Dunns Cambridge children
talked most to their baby siblings when engaged in
play: they were concerned either to direct the babys
activity or to prohibit the baby from interfering with
the older childs own activity. Teaching baby to talk
would not be a sensible part of such interaction. The
first-born childs advantage in language
development is to have had unshared language
teaching from parents; they themselves do not
provide such bad models of language as has been
suggested (e.g. by Zajonc and Marcus 1975) but they
do dilute the tutorial motherese environment for
their younger siblings. Older children frequently

teach their younger brothers and sisters about skills
and games and their own areas of knowledge, and
may do so more successfully than unrelated teachers
(Cicirelli 1972, 1976). It has been suggested that
teaching ones own competence to someone less
skilled is a particularly good way of achieving more
oneself (e.g. Light 1983), since being faced with the
cognitive difficulties of another person may force one
to reflect on the problem in new and productive ways.
This happens even for parents informally teaching
their children (Tizard and Hughes 1984); it
presumably would have even more potential for an
older child teaching a younger one.
The amount of complementary interaction between
siblings varies. The most frequently examined sources
of differences are sex and age. The evidence on sexbased patterns is inconsistent (Dunn 1983). It has
more often been found that same sex sibling pairs are
more positive in their relationship than different sex
pairs; but sometimes there has been no difference and
sometimes the pairs with one girl and one boy have
been the more co-operative, comforting and friendly.
Possibly older girls are more strongly expected to be
helpful and nurturant, while older boys are seen as
more challenging, but the differences in observed
behaviour are, so far, small or non-existent. The
behaviour which siblings show to each other is
intimately related to the interaction between the
children and their mother (and no doubt any other
significant adult). The patterns are complex (Dunn
1983; Dunn and Kendrick 1982) and change over
time, and we have no firm grasp of what causes what.
Possibly maternal treatment of children which is
responsive and consistent towards each one, avoiding
differential treatment and drawing children into
discussion of their responsibilities towards each
others wants, needs and feelings, is most likely to
lead to relatively good relationships between
siblings.
Reciprocal interactions
I have, so far, discussed complementary interaction

between siblings. Their relationships do, however,
also have strong reciprocal characteristics: siblings

are relatively similar in age, competence and status
within the family. Observational studies of young
siblings show them to be close and familiar, and this
intimate knowledge of each other allows both
warmth and highly effective aggression and
exploitation. For better and worse, the interaction of
brothers and sisters frequently has a strong emotional
tone (e.g. Furman and Buhrmester 1985).
One aspect of young childrens interest in their
siblings which Dunn and Kendrick (1982) emphasize
is their frequent imitation of each other. Initially it is
the older sibling who imitates the baby but as they
both grow older the younger sibling more frequently
imitates the older, and throughout childhood mutual
imitation may be an especially pleasing and exciting
activity. Imitation is more frequent in pairs who are
also friendly and helpful towards each other, and also
in same sex pairs. Warmth and perceived similarity
seem to be important for this sort of reciprocal
behaviour. An affectionate sibling relationship is also
associated with higher levels of social sensitivity and
role-taking skills (Light 1979), and siblings interest
in joint role play enables more experience of
negotiating and enacting roles than the less frequent
and less successful participation of an adult in role
play can do. There is some evidence that children
with siblings get on better with peers (Vandell and
Mueller 1980; Hartup 1978) presumably because
getting on with their siblings has provided them with
practice in getting on with peers. Brothers and sisters
will have experienced the similarity of their own with
each others needs, wishes, skills and interests which
makes it easier to understand another person and
ones self, and hence people generally. They will also
have learned about mutual help and joint activity, and
about mutual antagonism, jealousy and aggression.
These social understandings and social skills can be
applied outside the family, and some later friendships
may strongly resemble sibling relationships in their
interaction of closeness, support and teasing. It is
worth noting that having poor relationships with
siblings is quite strongly associated with later
189
Plate 15
190
pathology and antisocial behaviour. Sibling behaviour

both elicits and maintains aggression, and aggressive
behaviour is often used to settle sibling disputes. The
vast majority of children are at least sometimes
physically aggressive to their siblings (Parke and
Slaby 1983) though there is only rarely any serious
injury. Parents often allow their children to express
aggression within the family while discouraging the
same behaviour in public (Newson and Newson
1976), especially for girls. It may be easier to make
up squabbles within the immediate family than in a
wider social group which is less obliged to go on
living together. In some families, however, coercion
and counter-coercion escalate, or aggressive activity
continues even if the victim reacts in a pleasant or
conciliating way. Boys from these families are
particularly likely to be hyperaggressive out of
control children (Patterson and Cobb 1971, 1982)
with a very low level of sensitivity to other peoples
needs, wishes and actions. Habits learned with
siblings, like habits learned with parents, may affect
a childs approach to the social world beyond the
family.
Peers
An enormous amount of work has been done recently
on childrens relationships with other children. It has
varied so much in research method (ethnographies,
questionnaires, sociometrics, formal experiments) in
underlying theories (neo-Piagetian, behaviourist,
sociological, etc.) and in degree of insight (ranging
from stunning banality and pomposity to an eloquent
communication of shared experience) that I shall not
attempt to summarize and integrate it. There are
major reviews or collections by Hartup (1983), Lewis
and Rosenblum (1975), Foot, Chapman and Smith
(1980), Asher and Gottman (1981), among others,
and an accessible introduction by Rubin (1980). Here
I will focus on peer relations as they shed light on a
number of developmental issues and as part of the
ecology of the childs life.
I have placed a great deal of emphasis so far on the
importance of adults for childrens development. I
must now correct that emphasis by pointing out that
children spend a great deal of time (and energy, in the

widest sense) with other children. In densely
populated and age-graded societies like ours, those
other children will probably be of about the same age,
which implies relative similarity of skills,
experience, interests and status: that is, they can
reasonably be called peers. In so far as this is the
case, childchild relationships are likely to involve
reciprocal interactions rather more than adultchild
relations complementary interaction (Hinde 1979;
see p. 179 above). We do not have much detailed
evidence on what is actually typical of childchild
interactions in enough contexts (or on enough cases)
to know how far this is true, but it does appear that
from the pre-school years onwards children expect to
play with peers not with adults and to get help from
adults not from peers (Edwards and Lewis 1979;
Barker and Wright 1955). Even babies under a year
old seem to take a different attitude to other babies
from that they take to adults, being more friendly and
less wary (Lewis and Rosenblum 1975), though they
cannot sustain social interaction except with a partner
who can scaffold them as an adult would (Vandell
and Mueller 1980). If the participants in a childchild
interaction are unequal in age or competence there is
more complementary interaction, with older or more
skilled children helping younger ones, and younger
children showing dependence on the older (e.g.
Whiting and Whiting 1975). Interactions between
equals involve more give-andtake, both aggressive
and conciliatory. Clearly people need to develop the
social skills used between peers or towards the
weaker as well as those used towards the more
powerful; it may be easier for the child to do this with
other children than with parents. A minority of
children have serious and lasting difficulties in peer
interactions because they are very withdrawn or
impossibly bossy (e.g. Newson and Newson 1976);
poor peer relationships do seem to imply a bad
prognosis for later psychological health (Rutter
1985).
Social skills in peer interactions
What then are the social skills involved in peer

interaction? How do they develop? They clearly
191
begin with mutual interest shown in looking and

vocalizing, and by the time children reach 2 years of
age there are sequences of interaction as well as
isolated contacts (Vandell and Mueller 1980); if the
children are familiar with each other the interaction is
likely to be physically closer and more mature than if
they are strangers (Young and Lewis 1979; compare
also Dunn and Kendrick 1982). Pre-school children
take part in more positive and co-operative
interaction as they get older, especially becoming
better at taking part in joint activities such as rough
and tumble play, cooperative building and
construction and dramatic role-taking play (e.g.
Meadows and Cashdan 1983). They do not, as Parten
(1932) suggested, give up solitary activity or
parallel play (alongside other children but not
noticeably interacting with them); these activities
continue to take up a considerable amount of time for
pre-school children but no longer happen so often
because the child cannot manage to play with other
children. They begin to develop techniques for
gaining access to other childrens games, and also
techniques for excluding would-be participants in
their own games (Putallaz and Gottman 1981; Rubin
1980). There begin to be marked sex differences in
choice of play activity and, largely consequently, sex
differences in social interaction and other experience.
In most nursery studies, girls have chosen the quieter
activities of painting, sewing etc. which are done near
or with the teacher, and the domestic role-play of the
Wendy house: boys choose play with construction
toys such as meccano, vigorous physical activities
and rough and tumble. Through this choice girls
spend more of their time with adults or with small
groups of other children, mostly girls: boys spend
more time in larger groups of children, mostly boys,
and in rough, boisterous and overtly aggressive play.
There is some evidence that same-sex interactions are
commoner and more amicable than cross-sex
interaction, (e.g. Serbin et al. 1977; Maccoby and
Jacklin 1978; compare also Dunn 1983), and this
pattern of self-segregation and mild hostility between
the two sexes has proved very difficult for adults to
change (e.g. Best 1983; Huston 1983). Indeed as
children get older the ideology of separation gets
192
stronger. Boys with feminine interests may be

forced by the ridicule of their male peers to give them
up and adopt an attitude of exaggerated machismo,
while girls who wish to take part in masculine
activities can only do so if they are as talented in them
as the best boys. There is little in the way of a
rapprochement until well into adolescence, though
the public sexist rhetoric does not preclude all private
cross-sex friendships.
Clearly as children move through the primary
school years their social skills develop, becoming
more sophisticated in themselves and being used
differentially according to the setting, the people
involved, the task in hand and the recent history of all
these. They have more contacts with other children
and, as the course of sex-stereotyping in play
illustrates, have to accommodate themselves to the
social norms of the group. They become more aware
of the relative skills and statuses of their peers, and
explicit and stable hierarchies and roles develop. A
number of studies influenced by ethnography have
described examples of these child social groups;
among these are Sluckin (1981) on life in Oxford
playgrounds, Best (1983) on a Maryland class,
Davies (1979) on a group of Australian children and
Fine (1981) on boys playing in Little League baseball
teams. These detailed participant observation studies
provide reminders that childrens peer interactions
are not all sweetness and light, and although much of
what they show about childrens behaviour is
deplorable, they are a vivid counter-balance to some
of the over-sanitized secondhand reports of
childrens social worlds that can be found in the
literature.
Childrens understanding of other children
Theoretical accounts of why childrens interaction

with peers changes as they grow older often place a
great deal of emphasis on cognitive changes (e.g.
Hartup 1983). Young children, it is suggested, are too
egocentric, too limited in language, and too crude in
their understanding of intentionality to take part in
complex social relationships in an effective way. This
hypothesis is weakened by recent evidence that
young children are less deficient in these cognitive

skills than was supposed (see Chapters 2, 3, 4), by
recent observations of highly sophisticated social
behaviour by young children with familiar partners
(see the earlier part of this chapter, especially p. 189)
and by the omission of an account of how cognitive
and social skills interact. The cognitive
developmental model of social development
suggests that immaturities of cognition retard social
interaction and better cognition advances it, but the
connections are not well documented and it could
equally be the case that social interactions advance or
retard cognition (see Light 1983; and Chapter 3
above). However a greater emphasis on the ideas and
understanding of the participants in social interaction
has enriched theory, and the relatively new field of
social cognition seems to be flourishing.
Although the evidence so far is limited, there do
seem to be developmental shifts in childrens
understanding of events and other people (Shantz
1983). Young children are more likely than older ones
to describe the more obvious aspects of what they see
without explaining them or making inferences about
internal state, intention, or causes, although when the
situation is extremely familiar and they have no recall
problems even pre-school children can go beyond
description. Pre-school children appear to assume
that most acts are intended, and so may not recognize
(or, more specifically, label) accidental acts as being
accidental, though again familiarity and recall seem
to be important. Attributions of intention, and indeed
the concept of intention itself, are, of course, by no
means simple things, and it is hardly to be wondered
at that social experience plays an important part in
their development. What children say when asked to
describe another person changes as they get older
from descriptions of observable concrete
characteristics and global evaluations such as nice
or bad, through more abstract inferred psychological
qualities, to explanatory descriptions with more
qualifications and accounts of specific person by
situation interactions (e.g. Livesley and Bromley
1973). Thus children under 7 might talk like
behaviourists or demographers and say Shes a lady,
shes got yellow hair and a loud voice and she wags
her finger a lot; a child of 10 or so might talk like a

trait personality theorist and say Shes bossy and she
always likes to get her own way; an adolescent might
talk in interactionist terms and say well, she seems to
want to appear to be tough and totally in control but
maybe she had to act like that to get where she has,
and remember she did show some feelings when her
son got lost in the Sahara.
Childrens views on friendship
Rather similar changes appear to occur in childrens

free accounts of friendship. There is a shift from
rather concrete behavioural definitions centering on
giving things to and playing with, possibly with an
emphasis on the satisfaction of the child talking, to a
more abstract dispositional description involving
caring for, sharing feelings with, comforting and so
forth, with more emphasis on mutual satisfaction
(Shantz 1983). Reciprocity underlies friendship at all
ages; friends like each other in part because they do
things together. Friends tend incidentally to be rather
similar to each other (Hartup 1983); they also tend to
be more responsive and co-operative than nonfriends (Foot, Chapman and Smith 1980). These
changes in free descriptions have been taken as
indicating identical changes in the underlying ways
of thinking about people. There are problems in this,
however, as in all tasks involving interpreting
verbalization. It is conceivable, though not entirely
likely, that the conceptual structure has changed very
little and all that is being assessed is vocabulary
growth. Further, what is said has to be interpreted:
typically probing is needed, and it is also dangerous
to assume that a particular term means the same to all
subjects, let alone the same to a child as it does to the
researcher. A more structured research technique,
with appropriate statistical analyses, is really needed.
Among the possibilities are developments in
Personal Construct elicitation, such as Beveridge and
Brierley (1982).
Childrens attributions
Childrens answers to questions like why did he do

that? have also been studied and linked to attribution
193
theory (see Chapter 5). Behaviour which a person

shows consistently in most situations most of the time
is attributed to characteristics of the person;
behaviour which most people show in that situation
most of the time is attributed to the situation. Children
by 5 or 6 seem to use these attributional principles as
adults do, though the evidence so far is limited
(Shantz 1983). One area in which there has been
important work is childrens attributions of the causes
of achievement (Weiner 1974; Dweck and Elliott
1983). Success or failure on a task can be attributed to
internal causes such as ones effort, which is
occasion-specific i.e. unstable, or ones ability, which
is (relatively) stable, or to external ones such as luck
(very unstable), teacher favouritism (which could be
stable or unstable) or the difficulty of the task
(relatively stable) or to some interaction of such
causes. Thus if I, for example, won a fortune by
backing the horse that won the Derby, I could not
attribute my success to my own effort, since I made no
effort, nor to my ability, since I have never applied it
to studying form, conditions, and so forth, nor to
favouritism shown me by horse, jockey, bookmaker
or race officials, nor to the difficulty of the task, but
only to luck an external and unstable factor; and
hence I would have no grounds to expect further
success in my betting. If however I were the trainer or
the jockey (or the horse), the success might well be
attributed to aspects of my effort or ability or both,
unless an explanation in terms of cheating by officials
on my behalf or the peculiarly poor quality of the
other horses provided a sufficient explanation. If
there was no such alternative explanation I could
reasonably attribute my success to internal factors
and might expect future successes.
It appears that people differ in their attributions of
their own achievement, possibly because of their
developmental histories (see Chapter 5), and also that
there are overall developmental changes, which I will
outline here. Before the beginning of schooling,
although children try to achieve goals set by
themselves or by others and are pleased by their
successes and sad about their failures, it is not clear
how far they discriminate between different causes of
their achievement. They will often enough say I
194
cant do it and show fear of failure, but it is rarely

clear whether they discriminate between task
difficulty and ability or compare themselves with
peers. Indeed, their judgement of ability is often
heavily dependent on the achievement of a difficult
task, and it is fairly unusual for teachers in preschools to comment with much exactness on
childrens achievement, ability or effort (Meadows
and Cashdan 1983; Wood et al. 1980). Once children
start school, however, they are confronted with
learning tasks where success and failure are more
clear-cut, with teachers who demand effort and
achievement and with peers to compare themselves
with. During the early school years children develop
ideas about their own ability relative to other
childrens and become well aware of the need to put
some effort into the tasks they are faced with. They
start to differentiate between the contributions of
ability, effort and external factors to particular
successes or failures, making attributions on the basis
of other peoples judgements as well as their own.
Somewhere around the beginning of adolescence,
sex differences in attributions of achievement start to
appear. Girls are more likely to attribute their failures
to lack of ability, boys to lack of effort. This
difference rather closely resembles the many reports
of differences in teachers comments to boys and
girls. Comments to girls, particularly on difficult
masculine tasks like mathematics, seem more likely
to be of the well, thats rather hard for you, dont
bother your pretty little head about it type; while
comments to boys are more likely to be on the lines of
you havent taken much trouble with this, have you:
you go back and try properly this time (e.g. Dweck et
al. 1976, 1978). Attributing ones failure to lack of
effort implies that success is still possible and does
not reflect on ones ability; attributing ones failure to
lack of ability implies that one is inadequate, inferior
and unlikely to succeed. The implications of ideas
about achievement for personality development are
discussed in Chapter 5.
Having outlined some of the developmental
changes there seem to be in childrens ideas about
other people, we can return to the question of whether
more advanced social cognition brings about more
advanced social interaction. The evidence is patchy

and inconclusive, and the most that can be said is that
while social cognitive ability may be necessary for
understanding other people it is not sufficient for prosocial behaviour (as I discuss on p. 157). Boys who
are aggressive are more likely to interpret someone
elses ambiguous action as being intentionally hostile
(Dodge 1980); children who dont attend to the
psychological characteristics of those with whom
they interact may have problems interpreting other
peoples behaviour. Those who believe that they are
helpless may fail to achieve anything; those who
believe that they can achieve anything they want, if
they only try, may do so (Dweck and Elliott 1983). All
we can say at the moment is may: it all depends,
and we know very little about what it depends on.
Sex roles, sex differences, sex typing
For reasons which remain obscure, one of the most

common ways the human race categorizes itself is
into male and female. This division seems to have
been made in most cultures (e.g. Rosaldo and
Lamphere 1974) and in most historical periods (e.g.
OFaolain and Martines 1974); it is also one of the
discriminations very young children seem to make.
Along with it goes an assumption that males and
females are significantly different: this assumption
underlies common-sense or lay discussion of what
males and females are like, and has significantly
affected the research literature, which reports
differences far more often than similarities (Maccoby
1980; but see also Archer and Lloyd 1980).
Furthermore, the assumed differences of commonsense and of much research almost all involve taking
males as being more representative of the human
species (homo sapiens), or even simply better (e.g.
field independence vs field dependence), and females
as inferior or deficient in so far as they are different
from males (e.g. Kohlbergs moral judgement work,
Chapter 5). This sort of value judgement has been
under a heavy attack for at least the last hundred years
and has to some extent gone underground, at least so
far as most western psychological literature is
concerned. Elsewhere it remains strong, and it
continues to support discrimination against females

which worsens every aspect of many lives from the
earliest moment that sex can be diagnosed until death
(which, since selective infanticide and abortion
almost always select against females, may not be
separated by a long time gap). I would recommend
that anyone who does not regard this as deplorable
read Janet Radcliffe Richardss The Sceptical
Feminist (1982); it is quite clear that it is a significant
injustice to treat males as the norm and females as
defective males. Such a bias needs to be eradicated:
so do interpretations of sex differences which
attribute them to universals (such as being more
muscular) which just happen to be convenient for the
discriminatory status quo (such as males domination
of politics). Throughout it must be remembered that
with the exception of a few physical characteristics,
mostly directly related to reproduction, all sex
differences are differences between the averages for
male and female populations, and population
distributions overlap. To give a rather obvious
example, although there are more highly aggressive
males than there are highly aggressive females, and
the average level of aggression is higher for males
than for females, our present Prime Minister, though
genetically female, is conspicuously more aggressive
than many of her male colleagues and predecessors.
She disproves the stereotyped belief that only men are
forceful and assertive, if not the stereotyped belief
that women are unsuited to politics.
In the chronology of the development of sex
differences (see, for example, Huston 1983; Tanner
1978), the first cause is genetic: males have an XY
chromosome combination and females have an XX
pair. The effect of having the XY pair is to produce
differentiation of primitive gonadal tissue in the 6week-old foetus into what will be testes; if this
differentiation does not happen, the tissue develops
into ovaries at about the twelfth week. The male
foetuss testes produce more androgen than females
gonads do, and high levels of androgens lead to the
development of penis and scrotum. Female foetuses
exposed to high levels of androgen, and male foetuses
which are insensitive to androgen, develop external
195
genitalia more like those of the other sex. At birth,

gender is normally assigned on the basis of what the
external genitalia look like: and the pink or blue
infant blanket is followed by the childrearing patterns
that I will describe presently.
It is still a possibility that the genetic difference
directly influences sex-typed behaviour other than
that involved in reproduction, that, for example,
males are more aggressive or females more nurturant
because of biologically based factors such as the
current level of hormones circulating in their
bloodstream, or prenatal hormone exposure which
has led to different brain development. Huston
(1983), reviewing the research, points out the many
defects in studies which make conclusions
impossible at present. An interdisciplinary
collaboration between psychologists and biologists
is needed. The existence of a biological basis to any
sex difference would not, of course, mean that social
influences were unimportant, or that nothing Can or
should be done to change either sexs behaviour. That
something is natural does not mean that it is right or
unchangeable, as I argued in Chapter 1.
Whatever genetic influences there may be on sex
differences, there are certainly very pervasive social
ones. I would argue that it is these that turn the
biological sex into the social gender. Children
observe adults: predominantly they see women
fulfilling female roles domestic responsibilities,
service or nurturing jobs, aesthetic or social
recreations and men fulfilling male roles heavy
physical work, technical and scientific jobs, athletic
and aggressive recreations. People as observed in the
mass media, and in childrens literature, are even
more stereotyped than in real life (Lobban 1978;
Delamont 1980). The models that children are
exposed to, then, tend towards two contrasted
stereotypes. Children develop concepts of what are
appropriate behaviours for each sex as part of their
learning about the social world.
Boys and girls are also treated differently by adults.
This begins at birth, when boy babies are seen by their
parents as stronger and more vigorous, and girls as
softer, more fragile and less alert (Rubin et al. 1974;
196
Campos et al. 1983). Fathers show more

differentiation than mothers do: they also later play
vigorous physical games much more with their sons
than with their daughters. Boys are encouraged in
their gross motor activity and in goal-directed tasks:
girls are given more quiet interpersonal stimulation
and more encouragement to play with dolls. There is
consistent evidence from studies of American and
European children that suggests this pattern of
differences in parents behaviour persists through
childhood (Huston 1983; Block 1984). Whats more,
where parents behaviour is less sex-stereotyped,
childrens tends to be less stereotyped too. Girls in
particular are probably more likely to have
androgynous interests, activities and aspirations if
their mothers are employed or career-oriented
(Etaugh 1974, Gold and Andres 1978, Block 1984)
though the evidence is not strong.
Adults playing other social roles towards children
may also treat boys and girls differently. What
evidence there is suggests that, for example, boys get
more vigorous disapproval and scolding from their
teachers, maybe more attention altogether in the
secondary school (Huston 1983). Girls are probably
under less pressure to persist independently with a
task if they run into difficulties (Serbin et al. 1973,
1978); Serbin and others argue that this reduces their
sense of being in control of their own achievement.
A major source of pressure towards separate
gender roles comes from other children. Even in the
pre-school years, interaction with a child of the same
sex seems to be more likely to be mutually responsive
and positive (Lamb and Roopnarine 1979), though
the picture is more complex if the other child is a
sibling (Dunn 1984), when it is heavily influenced by
each childs relationship with their mother. As
children move through the primary school, peer
pressure amounts to a curriculum second only to the
official curriculum of the 3 Rs
it taught the children the traditional role behaviour for their
sex. It taught little girls to be helpful and nurturant. It taught
little boys to distance themselves from girls, to look down
on them, and to accept as their due the help that girls offered.
Through its insistence that boys learn to be boys and girls

learn to be girls, this second curriculum resulted in separate
worlds for the two sexes within the classroom and on the
playground (Best 1983, pp. 45).
This picture of feminine girls and very

masculine boys forming separate and somewhat
hostile groups comes out of most studies of what
children think gender roles and the social world
should be like. There is much peer pressure on
individuals to conform to these stereotypes. A girl
may achieve honorary status as a boy if she has
exceptional talents at some masculine activity, such
as football, and girls social worlds can be somewhat
androgynous in the primary school years. For a boy to
be girlish, at least where his peers can see, seems to
be social death. The children Best studied insisted
that boys must be strong, be good fighters, be able to
take it and dish it out, be the leader, be best at
everything. They must not show any weakness or
sentiment, must not cry, must not associate with girls
or with sissies (boys who associate with girls), must
not do feminine activities such as housework or
helping teacher, must not care more for another
persons welfare or success than for their own. Only
after adolescence do these stereotypes become less
rigid, and androgynous and non-sexist behaviour in
public begin to be a viable possibility.
Childrens sexism, leading them as it does into
separate social groups and different activities, gives
them different experiences during their school years.
On the whole, girls gravitate towards adults more, do
more domestic activities, show more affection and
nurturance towards their friends and to younger
children, and because of their vulnerability to sexual
assault and unplanned pregnancy are more carefully
chaperoned than boys (e.g. Newson and Newson
1976). On the whole, boys keep away from adults
more, play more competitive and vigorous games,
fight more, have gangs rather than intimate pairings,
and are freer to range over longer distances than girls.
Differences in exposure, practice and subjective
importance might well account for sex differences
such as males tendency to be more aggressive and
better at certain visualspatial skills, and females

tendency to be more nurturant and neater at fine
motor activities.
The development of childrens gender roles
through childhood which we find in the literature is
complex and contentious. I would summarize it as
follows. It begins with the chromosomal difference
which leads to differences in the external genitalia,
and thus to social assignment of a particular gender,
and possibly also to other physical differences
relevant to psychological development, though these
have been hard to identify. Thereafter there is a
tendency for parents to treat their sons as more active
and independent and interested in masculine toys,
and their daughters as more fragile and dependent and
interested in feminine toys. Children themselves
come to discriminate between males and females by
the age of 2 or 3 and tend to choose to play the games
of their own sex. They learn through the pre-school
years what are appropriate activities for each sex,
and increasingly through the primary school years
adopt for themselves and prescribe for others sex
typed behaviour which is stereotyped. There is
explicit influence by adults on sex-typed play and
occupational choice boys dont play with dolls,
girls cant be train drivers and more implicit
influence on social behaviours such as independence
and persistence with hard tasks. Peer groups set up
and to a considerable extent enforce separate
stereotypes. The mass media confront children with
virtually polarized sex roles.
This is a depressing picture, all the more so since
children are often so adamant about their sexism.
There are grounds for hope, though, in that sex
stereotypes are changing and becoming more multidimensional. It is more acceptable than it has been in
the recent past for men to nurture their children, show
their feelings, and negotiate rather than fearing loss of
face, all feminine behaviours: and for women to
assert themselves, seek public achievement and
responsibility, and be recognized as having equal
rights to mens. There has not been enormous, or
steady, or uniform progress, but there has been some:
even, as Best (1983) documents, among school
children!
197
Learning the social world of school
Schooling is compulsory in Britain from the age of 5,

and most children are in some sort of schooling by
their fifth birthday. Good behaviour in school is also
more or less compulsory, and success at school tasks,
while not compulsory, is desirable if life in school is
to be satisfying, as well as for the certification which
comes in its later stages, and for the job prospects
which follow schooling and are supposed to be served
and enhanced by it. Children are therefore obliged to
learn what to do in school. Since schools are very
complex social institutions, what is learned is also
complex and in particular operates on many different
levels. I shall do no more here than mention some
recent attempts to investigate childrens social
worlds in school and to indicate some of the issues
that are involved.
Even nursery and kindergarten children quickly
develop a social script which answers questions
about what do you do in school. Fivush (1984)
found narratives about school which used the general
present tense, relatively abstract language and correct
temporal order in 5-year-olds on their second day in
school. A sample protocol went like this:
Play. Say hello to the teacher and you do reading or
something. You can do anything you want to. Clean up, then
you play some more and then clean up. And then you go to
the gym or playground. And then you go home. You have
your lunch and then you go home. And you go out the
school, and you ride on the bus or the train and go home
(Fivush 1984, p. 1708).
British 4-year-olds, interviewed about what they did

in nursery school, produced rather similar narratives
(Meadows et al. 1977; Weaver and Meadows, in
preparation); when asked what their teachers did their
answers echoed the protocols emphasis on clearing
up, and teachers were predominantly organizers and
controllers. Willes (1983) induced her sample to play
with a toy school and complete a story about a school
day, and they too represented the teacher as being in
authority.
It is of course the case that teachers are (and need to
198
be) in authority over their class, though some of the

ideologies of education disguise this, particularly for
young children (King 1978). Part of childrens
learning about school is concerned with learning how
to behave towards teacher, how to negotiate teacher
child interaction. As I discuss in Chapter 4, the
discourse strategies and the etiquette of classrooms
differ in significant ways from those of home. There
is less shared knowledge, less child-contingent
dialogue, a requirement that the child puts up his or
her hand and waits to be called on rather than shouting
out an answer, a much higher rate of questions from
the adult where the adult already knows the answer
and the child knows the adult knows it, and so forth.
Children have to abide by such rules if the class is to
function at the busy hum which the ideology of
early childhood education advocates. Teachers teach
some of these rules explicitly through personal
directives, Put your hand up if you know, otherwise
I cant hear you or statements of general rules and
principles, Kind hands dont grab and snatch, they
share (King 1978). Some emerge more implicitly
from their discourse strategies and the childrens
experience of classroom interaction. Typically, the
classroom rules which young children encounter are
phrased as either inviolable general principles which
everybody always follows, or suggestions which are
designed to benefit the children, and which they will
therefore go along with if they know their best
interests, but which allow for some negotiation of
what is done and when. Complying with such rules is
seen as good and sensible, since they are supposed to
work to the benefit of all involved. Non-compliance
in the very young pupil is silliness or immaturity:
in the worst young offenders, and in older children, it
smacks of wilful disobedience or rejection of school.
All these categorizations of failure to comply with
school rules lead to negative evaluation of the child.
Children who are used to other sorts of rules for
behaviour and ways of achieving compliance and do
not adopt the school conventions quickly may
therefore be at particular risk of being frowned on by
the teacher (Rogers 1982).
I said, and meant, that teachers need to be in
authority over the children in their classes. Pollard
1985 provides an interesting analysis of the

interests-at-hand of teachers and pupils in primary
schools. Teachers are supposed to ensure that the
maximum amount of learning goes on in their
classroom in an orderly way; achieving a reasonable
approximation of this is essential if they are to feel
they are competent teachers, if they are to avoid the
stress of coping with classroom conflicts and the
external threat of criticism from the head teacher and
other authorities, and if their work is to be felt to be
enjoyable and fulfilling. Pollard sees this
maintenance of selfimage as a person who enjoys and
controls their work, autonomously and without
undue stress, as the primary interest-at-hand of the
teacher.
Pupils are also concerned to maintain an
advantageous self-image. They are in the difficult
position of having two reference groups within the
classroom. In so far as they relate to the teacher (or
other adults), it will serve this interest-at-hand to be
successful learners and dutiful pupils, thus avoiding
the stress, indignity and unhappiness of failing or
being reprimanded. The uneven power relationship
between teacher and pupil may make it a sensible
coping strategy to seek to please teacher. However,
the child has his or her peers as another reference
group. Classrooms are crowded with children
(another difference between home and school),
playgrounds are dominated by them, and
relationships with other children are an important
part of school life. They contribute to childrens selfimages, their enjoyment of school, the stress they
feel, and their dignity or indignities. Childrens
informal social systems are complex and demanding
(Davies 1979, Sluckin 1981), and not always
compatible with the demands that teachers make.
Analyses of children in schools have often found
children who belonged to a sort of counterculture
opposed to the teacher-centred ethos. Typically, they
would be doing less well on the tasks of the school
curriculum than their peers, would get little intrinsic
satisfaction from lessons (which are boring) and
find more enjoyment in messing about, having a
laugh, and a culture of toughness of which staging
confrontations with authority was an important part.
They may well be right in believing that the school

system does little for their present happiness and their
future prospects: their rejection of it, however, make
this even more likely.
It is likely that any classroom, and certainly any
school, will contain children who are coping with the
demands of the schools social structures in different
ways: teacher-centred goodies and swots, peercentred gangs and roughies, some children who
are virtually isolates, and some who, clever or
socially skilled, are popular with both teachers and
peers. Current detailed work on interaction in the
classroom and the playground, and on pupils and
adults concepts of school and of the functions of
education, is just beginning to sort out how and why
children transact the social networks of schools, and
what the effects of their coping strategies are. Among
useful sources are Hargreaves 1967, Rogers 1982,
Best 1983, Davies 1982, Delamont 1983, Galton,
Simon and Croll 1980, Hammersley and Woods
1984, King 1978, Newson and Newson 1977, Pollard
1985, Sluckin 1981 and Willes 1983.
As well as coping with the social systems of school,
children have to cope with the formal curriculum of
educational activities. They begin to categorize
school tasks as early as the nursery class (Beveridge
and Brierley 1982; Weaver and Meadows, in
preparation). Links between homes and schools
definitions and methods of education facilitate
learning and the childs satisfaction (Newson and
Newson 1977; J. Tizard, Schofield and Hewison
1981; B. Tizard, Mortimore and Burchell 1981).
Anxiety over hard work is the first source of
unwillingness to go to school discussed by the
Newsons in their study of 7-year-olds: some children
complained of being distracted from their work by
their peers or by non-academic activities such as PE
or prayers. Reading has a crucial place in the
curriculum because it is so often necessary for other
subjects. Taking control of their own learning
processes, that is developing and using their
metacognitive abilities, is another component of
being a successful school learner, the more so
because pupils must share limited access to the
teachers help and limits on the funding of schools
199
have led to sparse provision of books and other

materials for reference. Some children classify
certain areas of the curriculum as outside their
interests or their competence, not always to their
advantage (as, for example, girls deciding not to do
science subjects). Coping with the curriculum, with
peers and with teachers are intricately mingled
throughout school life.
Childrens independent social worlds
Studies of life inside the school and inside the family

form major components of the research literature on
childrens social development. Despite their possible
developmental importance, studies of childrens
unsupervised social worlds are comparatively rare.
Such settings are by definition (and often by the
childrens deliberate choice) away from adults
attention, so that it is difficult for researchers to gain
access to them save through the children themselves
acting as informants. Studies by Best 1983, Davies
1979, Fine 1980, Patrick 1973 and Sluckin 1981
illustrate the possibilities of such a method and
provide fascinating pictures of the groups studied,
but it is hard to know how representative these groups
are. The Opies work on childrens lore and
language (Opie and Opie 1967, 1969, 1985) makes a
notable contribution to our understanding of the
social order of their worlds, but again needs to be
placed in a wider picture of their discourse,
relationships and activities.
There is a limited amount of evidence from large
samples on what childrens out of school activities
are. The Newsons asked their Nottingham mothers
some open-ended questions about the spare time
activities of their 7-year-olds (Newson and Newson
1976) and also have similar unpublished data on these
same 400 children at later ages (E. Newson, personal
communication). In a study whose central concern
was how children were supervised out of school
hours and what recreation they enjoyed, Petrie and
Logan (1984) collected information on the pursuits of
423 London 7-year-olds and 11-year-olds. The Child
Health and Education Study, in its survey of its
national cohort sample (14,000 children) at the age of
200
10, asked the childrens mothers to rate how often

their child took part in each of a list of activities.
Medrich et al. (1982) surveyed the out of school
activities of 764 children of 11 to 12 years old in
Oakland, California.
Of these studies, those by Petrie and Logan and by
CHES are not yet fully analysed, and Medrich et al.
tabulate their data without conveying as vivid a
picture of the childrens lives as the Newsons do. The
Nottingham 7-year-olds were beginning to achieve
some independence of their families, playing outside
and making their own way home from school, though
still expected to get back promptly, say exactly where
they were going before leaving the vicinity of home,
and stay within specified geographical limits. Their
outdoor pursuits including bicycling, playing on
swings, ball-games (football for boys and bouncing
games for girls), skipping (mainly girls), rollerskating, and making dens. Playground games such as
hopscotch and marbles seemed rarely to be played
outside school. As in the Opies studies, special
verbal rituals and magical beliefs flourished within
groups of age-mates.
The most commonly mentioned indoor pursuit for
these 7-year-olds was drawing, painting and
colouring: reading, that is reading plus looking at
books, magazines and comics was another major
interest, particularly for girls and middleclass
children. Writing was also common (see Chapter 3).
Other indoor activities included model making
(mostly boys) and sewing or knitting (girls), making
scrapbooks and collecting things. The Newsons
provide a list of some of the things collected.
The following are some of the things collected by children
in our sample: silver paper, acorns, matchboxes, string,
buttons, tins, nuts and bolts, stones, conkers, tickets, boxes,
religious texts, cigarette cartons, toffee papers, make-up,
matchsticks, free gifts, rubbish, bottle-tops, nails, cheeseboxes, handbags, handkerchiefs, pens and pencils, plastic
gardens, jigsaws, gollywogs, caterpillars, car numbers,
marbles, model planes, soft toy animals, leaves, coins,
chemistry set equipment, records, Meccano, costume dolls,
dolls clothes, books, little cars, feathers, money, Action
Man sets, comics, jewellery, magazine pictures, scraps,

beermats, dolls-house furniture, labels, postcards,
ornaments, railway accessories, drawings, badges,
fircones, tea cards, Lego parts, stamps, soldiers, marbles,
bubble-gum cards, bricks, cactuses, footballers (sic),
sweets, insects, Premium bonds, Scalextrix accessories,
shells, anything thats weird or ghastly, and pictures of
Cliff Richard (Newson and Newson 1976, p. 134, Penguin
edition).
The more recent data provided by Petrie and

Logan, by CHES and by Medrich et al. pick up
watching television as a major indoor pursuit:
activity would probably be an inappropriate word,
since the detailed comments made in the American
study suggest that television viewing was passive,
non-critical, non-selective and non-social, even
when several family members watched together.
Bicycling, sport, swimming and reading were
common activities among these other samples too,
however. In each study, some children went to
museums and cinemas or to organizations such as
Scouts and Brownies. Substantial numbers of the
Oakland sample and of some of the ethnic groups in
the London sample went frequently to church or to
church based activities: the CHES survey did not ask
about this. Music or dance lessons were common in
the CHES and Oakland samples but are not
spotlighted in the London study. In Oakland and in
London there were substantial differences in what
children did associated with cultural and economic
differences: the CHES data have not yet been
analysed in these terms. Some of the activities of
children included in these studies are heavily
dependent on sponsorship and practical support by
adults: the collaboration or at least tolerance of
parents is necessary even for collections of pictures
of Cliff Richard. The less child-centred influence of
the commercial world also, of course, shapes
childrens activities.
Adolescence
I have very little to say about adolescence: my main

point is that I see a discrepancy between theories
about it and the apparent facts. The theories are, on

the whole, highly dramatic and subjective storm
and stress, identity crisis, etc. and the facts, such
as they are, seem to be much drier and more mundane.
In so far as theory and fact can be resolved, I think the
resolution lies in the social psychology of
adolescence, hence its appearance in a chapter on
social development. I am particularly conscious that
I am writing at a time when the economic structure of
society makes certain aspects of adolescence
peculiarly difficult in ways that are possibly
historically new. Being adolescent may not in itself
be a problem, but being adolescent now plausibly is
harder than other statuses now or being adolescent at
other periods.
Theories of adolescent crisis
Coleman (1980), reviewing theories of adolescence,

divides them into two groups. The psychoanalytically
oriented theories, Freuds own, Eriksons, the work
of Blos, Laing and others, stress an adolescent
vulnerability of personality and maladaptiveness of
behaviour, due to the pressure of sexual instincts
surging up after the latency period, and to the need to
disengage from parents and establish an adult
identity. Sociological theories stress that adolescence
is a period of role change: emerging from childhood,
where roles are largely ascribed to the child by other
people, he or she has far more choice of what role to
undertake and of how to play it. This discontinuity,
pressures from other people (and the culture via the
mass media) which may give contradictory messages
about what you should be, and a sense that adult
responsibilities loom while adult rewards may be
further off, lead to a problematic transition period.
Both schools of thought thus emphasize difficulties
and crisis. At the same time, there are the physical
changes of puberty clumsiness, spots, the first
possibility of reproduction, settling with a body
different in shape and size from what youve been
used to. Gloom, anxiety and sturm und drang seem all
too likely. They may seem so inevitable as to be
necessary parts of development, essential
contributors to the entire process of growing up.
201
Indeed, Erikson (1965) suggests that those who do

not suffer an identity crisis as adolescents are less
mature and healthy as adults than those who have a
crisis and resolve it successfully.
Self-concept in adolescence
A theory of adolescence as a period of confused

identity, role change and crisis surely suggests that it
will be a period when peoples views of themselves
change a great deal, and that this change will amount
to a significant disturbance. It is by no means easy to
assess peoples self-concepts (see p. 146) and there is
little really sound evidence on the point at issue.
However, during the early part of adolescence young
people probably are more self-conscious and more
self-critical than they were before puberty, with a
fluctuating and rather unfavourable self-image and a
loss of belief that other people can be relied on for
favour and indulgence. A highly negative self-image
and behavioural maladjustment tend to go together,
those who depreciate themselves most being likely to
be depressed, anxious, failing in school and feeling
incompetent in social relations. The adolescents
studied by Coleman (1980) expressed particular fear
and anxiety in their concept of themselves in the
future; the worsening of job prospects for adolescents
since 1980 suggests that these fears were justified and
the present generation might be worse off (Donovan
et al. 1986).
It is clear, however, that although adolescents do
experience doubts about themselves and do have to
make difficult decisions about their future lives, to
call this a disturbance is appropriate only to a
minority. Most people do not experience an acute or
dramatic crisis, a crucial moment when
development must move one way or another
(Erikson 1968, p. 16): rather they manage a gradual
adaptation. The twig bends, rather than either
snapping or remaining rigidly straight. Nor is
identity ever completely resolved, except perhaps
for the very fortunate extravert: most of us have to
question what we are at points of moral or practical
difficulty well after adolescence. Some theorists,
recognizing this, produce descriptions of the post-
202
adolescence stages of crisis in the self-concept:

Harter (1983) describes the model put forward by
Levinson, which appears to have a new crisis for each
five-year period. These models are fortunately
outside our current concerns, though there could be
work to be done on how the parents (or teachers)
mid-life transition crisis affects the childs
adolescent crisis.
Social relationships in adolescence
Relations between parents and children are popularly

supposed to be at their worst in adolescence, with the
generation gap at its widest. Before adolescence,
children have been dependent on their parents to a
considerable extent; after it, western societies expect
them to be independent. Being independent and
being dependent each have advantages and
disadvantages: children are likely to be ambivalent
about the change. So are parents, for the same
reasons. Inconsistent behaviour will accompany the
inconsistent attitudes. Nevertheless, most of the
adolescents interviewed say that they like, respect
and feel close to their parents, use them as models and
disagree with them mainly about minor issues of
dress, taste in music, and so forth (Coleman 1980;
Maccoby and Martin 1983). The patterns of family
interaction I discussed earlier (p. 181) are relevant
here; some families do show pronounced alienation
and conflict between parents and children in
adolescence, but on the whole they were functioning
in pathological ways before the childrens
adolescence was reached (Rutter 1983).
The next component of adolescent conflict I want
to glance at is the influence of the peer group. Again,
in the popular imagination this is an unhappy
influence; the child, once happy and innocent in the
bosom of his or her family, is seduced away into the
dissident, dissonant, discordant peer group of
adolescence and becomes there interested in wicked
things like sex, drugs, pop music and having a good
time. This stereotype too does not stand up to what
evidence there is (Coleman 1980; Douvan and
Adelson 1966; Hartup 1983). Peer group and parents
tend to agree about fundamental moral principles,
and conflicts seem to centre on socially trivial issues

(such as, for example, taste in music). Hartup (1983)
sees parent and peer influences as being synergistic in
adolescence, just as they are before and after. Thus, if
neither parents nor peer group approve of something,
the adolescent probably wont either; if either one
approves, the adolescents approval will rise but not
so far as it would reach if both parents and peers
approved. In early adolescence, however, there is
much pressure to conform with peers and this may
include conformity with antisocial norms (Berndt
1979). Again, the ecology of the childs life needs to
be considered (Bronfenbrenner 1979); adolescence
is perhaps a time when the individuals various
microsystems are more varied and less linked into a
tightly structured mesosystem than before or after.
Adolescence as marginal social status
A very interesting book by Bernice Martin (1981) is

relevant to this point. She uses the models of
anthropologists and sociologists to argue that
societies have an intrinsic tension between formal
responsible rational behaviour and anarchic free
behaviour. The latter is disruptive and so has a
marginal status, being controlled by restriction to
certain social groups or certain times. Thus the
bohemian very rich and the extremely poor are not
really expected to work, though the rest of us earn our
bread and simultaneously deplore and ogle their
freedom. Or the whole town works hard and is
compulsively respectable for fifty weeks of the year,
but has a few major festivities such as Christmas and
Wakes Week which are marked by large-scale
spending, drinking, new clothes and a greater
appearance of sexuality, what Hoggart (1957) called
excursions into the Baroque. Or, in the few years of
independent income between leaving school and
starting a family, the young are allowed to be a bit
outrageous; sowing wild oats, youre only young
once, gather ye rosebuds while ye may. The
culture of youth is marked by spontaneity, hedonism,
immediacy and a kind of self-centred emotional
intensity which, from some angles, can resemble
individualism, non-conformity or even rebellion
(Martin 1981, p. 139).
How far youth participate in youth culture is an

open question. The likely answer is that most
adolescents live in the interacting cultures of family,
school and peer group as well as youth culture,
seeing the latter as subjectively important but
recognizing its transitory nature. Autobiographical,
ethnographic and interview accounts (e.g. Kitwood
1980; Jenkins 1983; McRobbie and Nava 1984; and
Griffin 1985) show the complexity of the social
picture. At present, many adolescents have difficulty
in
achieving
employment
and
financial
independence and there may be less individualism
and less freedom than there was in Martins focal
period, the 1960s.
It is part of the sturm und drang model to say that
adolescence is a period of high rates of psychological
stress and malfunctioning. The rates of manifest
psychiatric disturbance do not support this view
(Rutter 1980, Rutter et al. 1976). Rutters Isle of
Wight population showed very little increase in
disorder between 10 and 14 overall, though there was
an increase for girls. Psychiatric disorder became less
likely to be associated with severe family discord
than it had been in childhood. Depression seemed to
be the central problem, mild depression being very
common indeed, as it is among some adults (Brown
et al 1978).
Thus although theories of adolescence emphasize
its stresses and problems, most adolescents seem to
manage a fairly smooth transition from being a child
to being an adult. Coleman (1980) suggests they are
able to do this by coping with one problem at a time:
now examinations, now getting permission to stay
out late, now achieving the current peer group
desirable. A number of issues have to be got through,
but they come into focus at different ages and are not
so interdependent that solution of one requires prior
solutions of others. So far as there is evidence, this
focal theory seems to fit it better than the crisis
theories which have dominated the literature. It is
when there are multiple problems which have to be
dealt with simultaneously, or which have become
chronic, that real stress, storm and breakdown occur.
One all too conspicuous example of this is juvenile
delinquency.
203
Delinquency
Young children very rarely commit offences which

bring them before the courts: only a small minority of
adults over 21 come into conflict with the law. A
substantial proportion of male adolescents, however,
are found guilty of, or cautioned for, indictable
offences; the proportion of females is much lower,

but girls too are at their most criminal between 14 and
17. There has been a considerable rise in adolescent
indictments since the 1950s:Figures 18 and 19 show
the Home Office figures on criminal statistics for
England and Wales.
Figure 18 Males* found guilty of, or cautioned for, indictable/triable-either-way offences per 100,000 population in the age group by
age
Notes: *Other offender, i.e. companies, public bodies, etc., are included with males aged 21 and over because separate figures are not
available before 1976; Adjusted for changes in legislation.
Source: From Rutter and Giller (1983), pp. 69; using Home Office figures.
204
Figure 19 Females found guilty of, or cautioned for, indictable/triable-either-way offences* per 100,000 population in the age group
by age.
Note: *Adjusted for changes in legislation.
Source: From Rutter and Giller (1983), p.70; using Home Office figures.
Some of this rise is due to changes in legislation,

police recording practices and a trend towards
dealing with antisocial behaviour formally rather
than informally, but such factors cannot account for
all the increase. Many offences do not get to court,
and we know from self-report studies that many are
committed but not detected; doing delinquent things
must be seen as very common indeed among
adolescents.
Delinquent acts cover, of course, an enormously
wide range from the rather trivial to the very serious
indeed. They include going to an X film when under
age, which I and a majority of adolescents in Wests
sample (West 1982, p. 22) have done (I was taken by
my father), to major offences against property or
persons. Most delinquents offences are from the less
serious end of the range: the commonest concern theft
and handling stolen goods (Maliphant 1979; West
1982) which account for more than 80 per cent of
girls crimes. About half of delinquents, mostly those
who have committed relatively minor offences, are
one-time offenders who never appear before the
courts again. Most boys who are delinquent as
adolescents, even those with several convictions,
lead a normally law-abiding life as adults.
Delinquents who only commit minor offences, once
only or only a few times, and give it up in their late
teens, are very similar in personality and background

to boys who have never been convicted. The picture
is rather different for the persistent offenders and
those who commit serious, particularly violent,
crimes.
Rutter and Giller (1983) provide a very thorough
review of the evidence on why people become
delinquent. One of the most striking features of their
survey is how thin and patchy the evidence is. One
problem is the heterogeneity of delinquents, as I
described in the last paragraph. Another is that
measures are often weak, samples small and studies
correlational. However they do tentatively identify a
list of individual and psychosocial factors which
characterize the prototypical persistent delinquent,
the delinquent who repeatedly commits non-trivial
antisocial acts. He probably comes from a large
family, in worse economic straits than its neighbours
and in worse housing, with poor relations between
family members, ineffective supervision and
discipline by parents and discord between them, and
other members of the family also have criminal
records. He is probably a school failure, has a lowish
IQ and attention problems, has been unpopular with
his peers and troublesome to his teachers since he was
8 or 10, and spends a lot of time alone watching TV,
preferring violent programmes. He lives in an area
205
where there is little constructive amusement and lots

of unsupervised public space. His social competence,
and his attractiveness to potential employers, are both
low. (We should not, of course, draw premature
conclusions about which of these cause delinquency
either directly or indirectly.)
The same sort of picture comes from two important
studies which followed their sample from a very early
age through to adulthood. Wadsworth (1980) reports
on delinquency in the 1946 cohort, West (1982) on a
sample of Inner London working-class boys born in
19524 and studied from the age of 8 to the age of 21
(for some subgroups 24). These studies have the great
advantage that they know something about the lives
of their subjects before they were first found to be
delinquent, and they have similar data on boys who
do not become delinquent. West (1982, pp. 2930)
lists five major factors, each of which had a
significant association with likelihood of
delinquency, that could not be explained in terms of
other, more basic items. The factors were
1 coming from a low-income family (n = 93,
percentage delinquent = 33.3 per cent)
2 having four or more siblings by your tenth birthday (n = 99, percentage delinquent = 32.3 per cent)
3 parents considered by social workers to be
unsatisfactory as child-rearers (n = 96, percentage
delinquent = 32.3 per cent)
4 IQ below 90 (Ravens Matrices Test) (n = 103,
percentage delinquent = 31.1 per cent)
5 parent with criminal record (n = 103, percentage
delinquent = 37.9 per cent)
These adverse factors overlapped considerably in the
sample; of the sixty-three boys who had a combination
of at least three of the five, thirtyone became juvenile
delinquents.
These are unsatisfactorily high rates: each factor
approximately doubled the rate of delinquency in
boys who had it compared with the other boys in the
sample who did not. Each factor was quite common,
too, applying to about one in four of the sample. Boys
206
who were in the worst 10 per cent on each factor, and

among the most troublesome according to teacher
and peer ratings, were very likely to be among the
persisting recidivists, the young men convicted at
least twice before the age of 19 and at least once for
further offences between their nineteenth and
twenty-fourth birthdays. West calls their social
histories little short of disastrous. About half of
these most disadvantaged boys were persisting
delinquents, and of the remainder many more showed
some degree of social disturbance. West and his
colleagues did a special study of the eighteen men
who, although they had a history of serious
deprivation, had no criminal conviction by the time
they were 22, in order to discover how they had
managed to avoid becoming delinquents.
Depressingly, it turned out that some of them actually
had committed offences, and the most prominent
characteristic of the group was their social failure.
Many were chronically unemployed, or employed in
ill-paid low-status jobs; their housing was often very
poor; their social contacts were often limited and their
emotional relationships unsatisfactory. West (1982,
pp. 936) gives some case histories: this is the last
one, which is said to be fairly representative.
Case 011
This mans childhood was marred by impoverished
conditions and extreme conflict between his parents. His
father, who was chronically unemployed, was described as
something of a hermit, hardly communicating even with
his wife. He was 11 when his father died. His mother was an
aggressive, quarrelsome woman with a long history of
psychiatric disorder, diagnosed as paranoid psychosis
with depressive features in a woman of low intelligence.
She made a number of suicide attempts.
At school he was no disciplinary problem, but was a poor
attender and was taken before a juvenile court on that
account. He was thought to be under his mothers
domination. She would excuse his absences by
complaining, falsely, that his classmates were picking on
him. His mother and two of his siblings sustained criminal
convictions.
He declined to be interviewed at 18, his mother writing

on his behalf to order the interviewer to stay away,
commenting that there were plenty of delinquents among
the ignorant lot living in the same building who needed
surveying. She was at the time in great conflict with the
housing authority who would not re-house her because she
was such a troublesome tenant. When he was aged 20 the
social services became involved with the household
because his mother had attacked him with a knife. He was
noted then to be an unemployed labourer who rarely went
out in the evening.
At age 23 he did agree to an interview. He was still living
with his now aged mother, but his siblings had all left and
he no longer saw them. The home was in a very neglected
state with the living-room floor partly eaten away by rats,
the banisters and several doors fallen offand the sink almost
permanently blocked. He had been continuously
unemployed for eighteen months. He never went out in the
evenings and his only two regular excursions each week
were to collect social security money and to window gaze.
He had no outside human contact apart from repair men and
council officials. Asked about offences he replied
pathetically: I cant get into trouble, I never go out (West
1982, pp. 956).
Material on the 1946 cohort (Wadsworth 1979),

while basically in agreement, provides further
illustration of the range of problems associated with
coming from a severely disadvantaged and disrupted
family. Wadsworth shows that children from
disrupted families, as well as being more likely to be
delinquent, were more likely to be admitted to
psychiatric hospitals or to be divorced or separated
before the age of 26, and were more likely to have
suffered from disorders like stomach ulcers or colitis.
Wadsworth puts forward two possible ways in which
there might be causal links between early disruption
and later disorder. One is that the children have
learned to handle stress differently; the other is that
generally held social views of the effect of a disrupted
family life have acted as a self-fulfilling prophecy. If
teachers, social workers and other significant adults,
and perhaps the children themselves too, believe that
coming from a family which lacks unity and
happiness inevitably leads to doom and gloom, they

may behave in ways which make this progress more
likely. The child who is continually told that he will
come to a bad end, and who is given no opportunity to
make good, may be trapped into a delinquent career
which a positive intervention might have avoided.
It is very important to recognize that despite the
tragedies shown in Wests sample and corroborated
in Wadsworths, a substantial number of even the
most disadvantaged boys do not become delinquent.
Wadsworth (1979) shows that life history does not
predict mild delinquency any better than chance
would (p. 124), and although 87 per cent of the worst
delinquents did have broken homes, large families
and so forth in their backgrounds, 13 per cent did not
and would not have been predicted to be delinquent
from their biographies. Family problems should
perhaps be eliminated for their own sake, not because
they cause delinquency.
We also need to investigate what factors have
protected children against their poor life chances.
Rutter, Wadsworth and West all emphasize how
much more research is needed. However there are
someindicationsastowhatmay beworth investigating.
We need to know much more about strategies for
coping with stress and disadvantage. We need to
know about the subtleties of adults behaviour which
may affect both the opportunities that children are
given and their self-concepts. We need to know more
about the peer group, about the ecology of the social
world of the adolescent. Most juvenile delinquency
involves acts undertaken with peers: Wests
delinquents particularly seem often to have begun by
acting as lookout or fence to other boys and
progressed, if thats the right word, to more personal
involvement in the offence. Boys who persist in
spending their time with a group of four or more other
boys are more likely to remain delinquent than those
who move to mixed-sex groups and give up the
gang. Parental supervision of friendships and
moving away from the inner citys street corner
society also decrease delinquency. Marriage,
provided it was to a non-delinquent girl, also seems to
reduce delinquent behaviour, in so far as it produces
a change for the better in the social group and in
207
personal relationships. Being employed possibly

decreases aggression and delinquency, though
decisive evidence is lacking. Living in a pleasant
environment with neighbours feeling responsible for
their surroundings is probably also important (Rutter
and Giller 1983).
The core problem in persistent delinquency does
seem, though, to be the quality of the relationships
and the social training that the child has received in
his family and other social interactions. We do not at
present know enough to identify exactly which
children are going through experiences which will
leave them at risk for delinquency, nor do we know
what to do about identified problem cases. Whipping
a child away from an unsatisfactory family into the
care of the local authority, for example, is not the ideal
solution. Potentially, a great deal could be done to
reduce the disadvantages experienced by many
families and to improve the stability and the harmony
of child-rearing inside and outside the home. More
work needs to be done to establish what prevention of
individual predisposition to delinquency is possible.
Changes in wider social spheres such as schools, the
mass media, local ecology and the economy and
political system we live in could also reduce
delinquency. There are a multitude of influences
which bring about antisocial behaviour; once their
interactions are understood we will be better placed to
208
prevent it. Prevention, here as elsewhere in the social

system, is demonstrably cheaper than cure (e.g.
Weikart 1978).
Overview
In this book I have tried to put together our current
evidence on various aspects of child development to
make a coherent picture of the patterns and sequences
that occur. I have tried to indicate what general
theories of child development might try to describe
and explain, with some detail in certain well-studied
areas. I have argued that we must try to achieve sound
causal explanations of both continuities and
discontinuities in development, not just getting age
differences but explaining in detail what has led to
what within the ecosystems that the child inhabits,
and where the sequence might have been different.
Detailed studies of restricted areas of development
are appropriate, but we must be cautious about
isolating one aspect of childrens development from
others: cognition, emotion, social relationships and
contexts, language, physical state, all co-occur and
mutually influence each other. We cannot assume that
what we see in one setting or one sample or at one time
will be representative in detail of different settings,
samples or times unless we understand very fully why
things are the way they are.
Further reading
Chapter 1 Introducing the study of child

development
Introductory reading
McGurk, H., Growing and changing, Methuen

Essential Psychology (1975)
Wood, D., Models of childhood, in Models of Man,
ed. A. J. Chapman and D. M. Jones, British
Psychological Society (1980)
History
Aris, P., Centuries of Childhood (Penguin 1973)

de Mause, L. (ed.), The History of Childhood
(Souvenir Press 1974)
Hardyment, C., Dream Babies (Jonathan Cape 1983)
Pollock, L. A., Forgotten Children (Cambridge
University Press 1983)
Further reading
Bruner, J. S., Sylva, K. and Jolly, A., Play (Penguin
1976)
Davie, C. et al. , The young child at home (NFER/
Nelson 1984)
Rubin, K. H., Fein, G. G. and Vandenberg, B., Play,
in E. M. Hetherington (ed.) vol 4 of the Handbook
of Child Psychology (New York: Wiley 1983)
Schwartzman, H. B., Transformations the
anthropology of play (NY: Plenum Press 1978)
Smith, P. K. (ed.), Play in animals and humans
(Blackwell 1984)
Sutton-Smith, B. (ed.), Play and learning (New York:
Gardner Press 1979)
Chapter 2 Perceiving and understanding

Scientific issues
Bronfenbrenner, U., The Ecology of Human

Development
(Cambridge, Mass.: Harvard
Kessen, W. (ed.), Handbook of Child Psychology,
vol. 1, series ed. P. H. Mussen (Wiley 1983)
Haith, M. M. and Campos, J. J. (eds.), vol 2 of the
Handbook of Child Psychology
Rutter, M., Family and school influences on
behavioural development, Journal of Child
Psychology and Psychiatry, 26 (3) (1985a), pp.
34968.
Rutter, M., Family and school influences on
cognitive development, Journal of Child
Psychology and Psychiatry, 26 (5) (1985b)
Play
Garvey, C., Play (Fontana/Open Books 1977)
Donaldson, M., Childrens minds (Fontana 1978)

Flavell, J. H., Cognitive development (Englewood
Cliffs, NJ: Prentice Hall 1977)
Howe, M. J. A., A teachers guide to the psychology
of learning (Blackwell 1984)
Further reading
Brainerd, C. J., Piagets theory of intelligence

(Englewood Cliffs, NJ: Prentice Hall 1978)
Bruner, J. S., On cognitive growth, in J. S. Bruner,
R. Olver and P. M. Greenfield (eds.), Studies in
cognitive growth (New York: Wiley 1966)
Butterworth, G. and Light, P. (eds.), Social cog nition:
studies of the development of understanding
(Harvester Press 1982)
Case, R., Intellectual development (New York:
Academic Press 1985)
Flavell, J. H. and Markman, E. M. (eds.), vol. 3 of the
Handbook of Child Psychology (New York: Wiley
1983)
209
Haith, M. and Campos, J. (eds.), vol. 2 of the

1983)
Harris, P., Infant cognition, in M. Haith and J.
Campos (eds.), vol. 2 of the Handbook of Child
Psychology (New York: Wiley 1983)
Kail, R. V. and Bisanz, J., Information processing
and cognitive development, in Advances in child
development and behaviour, vol. 17 (New York:
Meadows, S. (ed.), Developing thinking (Methuen
1983)
Perlmutter, M. (ed.), Intellectual development,
Minnesota symposium on child psychology, vol.
19 (Hillsdale, NJ: Lawrence Erlbaum Associates
1985)
Chapter 3 The development of cognitive

skills
Howe, M. J. A., A teachers guide to the psychology

of learning (Blackwell 1984)
Kail, R. V., The development of memory in children
(San Francisco: Freeman 1979, 1985)
Sternberg, R. J., Human abilities (New York: W. H.
Freeman 1985)
Further reading
Brown, A. L. et al., Learning, remembering and

understanding, in J. H. Flavell and E. H. Markman
(eds.), Handbook of Child Psychology, vol. 3
(New York: Wiley 1983)
Bryant, P. and Trabasso, T., Nature 232 (1971), pp.
4568
Ellis, A., Reading, writing and dyslexia (Erlbaum
1984)
Kail, R. V. and Hagen, J. W., Perspectives on the
development of memory and cognition (Hillsdale,
NJ: Lawrence Erlbaum Associates)
Meadows, S. (ed.), Developing thinking (Methuen
1983)
Mills, M. and Funnell, E., in S. Meadows (ed.),
Developing thinking (Methuen 1983)
Piaget, J. and Inhelder, B., Memory and Intelligence
(Routledge & Kegan Paul 1973)
210
Robinson, E. J., Metacognitive development, in S.

Meadows (ed.), Developing thinking (Methuen
1983)
Chapter 4 Language development

de Villiers, P. A. and de Villiers, J. G., Early language

(Fontana/Open Books 1979)
Wells, C. G., Language, learning and education
(Centre for the Study of Language and
Communication, University of Bristol 1982)
Further reading
Elliott, A. J. Child language (Cambridge University

Press 1981)
Flavell, J. and Markman, E. (eds.), vol. 3 of the
1983)
Heath, S. B., Ways with words (Cambridge
Lock, A. (ed.), Action, gesture and symbol
(Academic Press 1978)
Maratsos, M., Some current issues in the study of the
acquisition of grammar, in J. H. Flavell and E. H.
Markman (eds.), vol. 3 of the Handbook of Child
Romaine, S., The language of children and
adolescents: the acquisition of communicative
competence (Blackwell 1984)
Chapter 5 Personality
Cook, M., Levels of personality (Holt, Rinehart &

Winston 1984)
Maccoby, M., Social development: psychological
growth and the parentchild relationship (New
York: Harcourt Brace Jovanovich 1980)
Rutter, M., Psychopathology and development:
links between childhood and adult life, in M.
Rutter and L. Hersov (eds.), Child and adolescent
psychiatry: modern approaches
(Blackwell
Scientific Publications 1985c)
Further reading
Further reading
CIBA foundation symposium 1982, Ruth Porter and

Geralyn M. Collins (eds.), Temperamental
differences in infants and young children (Pitman
1982)
Bandura, A., Self-efficacy: toward a unifying theory
of behavioural change, Psychological Review, 84
(1977a), pp. 191215
Erikson, E., Childhood and society (Penguin 1963)
Parke, R. D., and Slaby, R. G., The development of
aggression, in E. M. Hetherington (ed.), vol. 4 of
the Handbook of Child Psychology (New York:
Wiley 1983)
Radke-Yarrow, M., Zahn-Waxler, C. and Chapman,
M.,
Childrens prosocial dispositions and
behaviour, in E. M. Hetherington (ed.),
Socialisation,
personality
and
social
development, vol. 4 of the Handbook of Child
Chapter 6 Social relationships

Coleman, J. C., The nature of adolescence (Methuen

1980)
Rogers, C., A social psychology of schooling
(Routledge & Kegan Paul 1982)
Rubin, Z., Childrens friendships (Fontana 1980)

Schaffer, H. R., Mothering (Fontana 1977)
Further reading
Coleman, J. C. (ed.), The school years (Methuen

1979)
Dunn, J. and Kendrick, C., Siblings (Grant McIntyre
1981)
Hetherington, E. M. (ed.), vol. 4 of the Handbook of
Child Psychology (New York: Wiley 1983)
Hinde, R. A., Towards understanding relationships
(Academic Press 1979)
Kaye, K., The mental and social life of babies
(Methuen 1984)
Pollard, A. W., The social world of the primary school
(Holt, Rinehart & Winston 1985)
Richards, M. P. M., The integration of a child into a
social world (Cambridge University Press 1974)
Richards, M. P. M. and Light, P., Children of social
worlds: development in a social context (Polity
Press 1986)
Rutter, M., Changing youth in a changing society
(Harvard: Harvard University Press 1980)
Rutter, M. and Giller, H., Juvenile delinquency
(Penguin 1983)
211
Bibliography
Acker, S., Megarry, J., Nisbet S. and Hoyle, E. (eds.),

Women and Education: World Yearbook of Education
1984 (London: Kogan Page 1984)
Ainsworth, M. D. S., Attachment as related to mother
infant interaction, Advances in the Study of Behaviour,
9 (1979), pp. 252
Ainsworth, M. D. S., Blehar, M. C., Waters, E. and Wall, S.,
Patterns of attachment: a psychological study of the
strange situation (Hillsdale, NJ: Erlbaum 1978)
Applebee, A., The childs conception of story (Chicago:
University of Chicago Press 1978)
Archer, J. and Lloyd, B., Sex and gender (Harmondsworth:
Penguin 1982)
Aries, P., Centuries of childhood (London: Jonathan Cape
1962)
Atkins, E. et al., The 1946 British birth cohort survey: an
account of the origins, progress and results of the
National Survey of Health and Development, in S. A.
Mednick and A. E. Baert (eds.), An empirical basis for
primary prevention: Prospective longitudinal research
in Europe (Oxford: OUP 1981)
Asher, S. R. and Gottman, J. M., The development of
childrens friendships (Cambridge: Cambridge UP
1981)
Baddeley, A., The psychology of memory (New York:
Harper & Row 1976)
Bahrick, H. P., Bahrick, P. O. and Wittlinger, R. P., Fifty
years of memory for names and faces: a cross sectional
approach, Journal of Experimental Psychology:
General, 104 (1975), pp. 5475
Bandura, A., Self-efficacy: toward a unifying theory of
behavioural change, Psychological Review, 84 (1977a),
pp. 191215
Bandura, A., Social learning theory (Englewood Cliffs, NJ:
Prentice Hall 1977b)
Bandura, A., Self referent thought: a developmental
analysis of self-efficacy, in J. H. Flavell and L. Ross
(eds.), Social cognitive development (Cambridge:
Cambridge UP 1981)
212
Banks, M. S. and Salaparek, P., Infant visual perception,

in M. M. Haith and J. J. Campos (eds.), vol. 2 of the
Handbook of Child Psychology (New York: Wiley 1983)
Barker, P., Basic Child Psychiatry (London: Granada 1979)
Barker, R. G. and Wright, H. F., Midwest and its children
(New York: Harper & Row 1955)
Baring-Gould, W. S. and Baring-Gould, C., The Annotated
Mother Goose (New York: Bramhall House 1962)
Barrett, J. H. W., Prenatal influences on adaptation in the
newborn, in P. Stratton (ed.), The Psychobiology of the
human newborn (London: Wiley 1982)
Bartlett, F. C., Remembering (Cambridge: Cambridge UP
1932)
Bate, W. J., Samuel Johnson (New York: Harcourt Brace
Jovanovich 1975)
Bates, E., Bretherton, I., Beeghly-Smith, M. and McNew,
S.,
Social bases of language development: a
reassessment, in H. W. Reese and L. P. Lipsitt (eds.),
Advances in Child Development and Behaviour, vol. 16
(New York: Academic Press 1982)
Bates, E. and MacWhinney, B., A functionalist approach
to the acquisition of grammar, in E. Ochs and B.
Schieffelin (eds.), Developmental pragmatics (New
York: Academic Press 1979)
Baumrind, D., Current patterns of parental authority,
Developmental Monographs, 41, 1 (1971), pt 2
Baumrind, D., New directions in socialisation research,
American Psychologist, 35 (1980), pp. 63952
Beilin, H., Piagets theory: refinement, revisionism or
rejection?, in R. H. Kluwe and H. Spada (eds.),
Developmental models of thinking (New York:
Bell, R. W., A reinterpretation of the direction of effects in
studies of socialisation, Psychological Review, 75
(1968), pp. 8195
Belloc, H., Cautionary tales for children (London:
Duckworth 1918)
Bereiter, C. and Scardamalia, M., From conversation to
composition: the role of instruction in a developmental
Bibliography
process, in R. Glaser (ed.), Advances in Instructional
Psychology, vol. 2 (Hillsdale, NJ: Erlbaum 1982)
Berger, M., Temperament and individual differences, in
M. Rutter and L. Hersov (eds.), Child and adolescent
psychiatry (Oxford: Blackwell 1985)
Berlyne, D. E., Curiosity and exploration, Science, 153
(1966), pp. 2533
Berndt, T. J., Developmental changes in conformity to
peers and parents, Developmental Psychology, 15
(1979), pp. 60816
Bernstein, B., Class, codes and control, vol. 1 (1971); vol.
2 (1973); vol. 3 (1975) (London: Routledge & Kegan
Paul)
Bertenthal, B. I. and Fischer, K. W., Development of selfrecognition in the infant, Developmental Psychology,
14 (1978), pp. 4450
Berthoud-Papandropoulou, I., An experimental study of
childrens ideas about language, in A. Sinclair, R. J.
Jarvella and W. J. M. Levett (eds.), The childs
conception of langauge (Berlin: Springer-Verlag 1978)
Best, R., Weve all got scars (Bloomington: Indiana UP
1983)
Bettelheim, B., The uses of enchantment (Harmondsworth:
Penguin 1978)
Beveridge, M. (ed.), Children thinking through language
(London: Edward Arnold 1982)
Beveridge, M. and Brierley, C., Classroom constructs: an
interpretive approach to young childrens language, in
M. Beveridge (ed.), Children thinking through language
(London: Edward Arnold 1982)
Biggs, J. B., The role of metalearning in study processes,
paper given at BPS Education Section Conference
Cambridge, September 1984 (1984/in press)
Bissex, G. L., Gnys at wrk: a child learns to write and read
(Cambridge, Mass.: Harvard University Press)
Bjorklund, D. F. and Zeman, B. R., Childrens
organisation and metamemory awareness in their recall
of familiar information, Child Development, 53 (1982),
pp. 799810
Blinkhorn, S. F. and Hendrickson, D. E., Averaged evoked
responses and psychometric intelligence, Nature, 295
(1982), pp. 5967
Block, J., Sex-role identity and ego development (San
Francisco: Jossey-Bass 1984)
Block, J. H., Lives through time (Berkeley, California:
Bancroft Books 1971)
Block, J. H. and Block, J., The role of ego-control and ego

resiliency in the organisation of behavior, in W. A.
Collins (ed.), Minnesota Symposium on Child
Psychology, vol. 13 (Hillsdale, NJ: Lawrence Erlbaum
1980)
Bloom, K., Evaluation of infant vocal conditioning,
Journal of Experimental Child Psychology, 27 (1979),
pp. 6070
Bloom, L., Language development: form and function in
emerging grammars (Cambridge Mass.: MIT Press
1970)
Blurton Jones, N. (ed.), Ethological studies in child
behaviour (Cambridge: Cambridge UP 1972)
Blurton-Jones, N., Woodson, R. H. and Chisholm, J. S.,
Cross-cultural perspectives on the significance of social
relationships in infancy, in D. Shaffer and J. Dunn
(eds.), The first year of Life (Chichester: Wiley 1979)
Boden, M. A., Piaget (London: Fontana 1979)
Bone, M. R., Preschool children and the need for daycare:
a survey carried out on behalf of the DHSS (London:
HMSO 1977)
Boulton, M. G., On being a mother (London: Tavistock
1983)
Bower, T. G. R., Development in infancy (San Francisco:
Freeman 1974)
Bower, T. G. R., Broughton, J. M. and Moore, M. K.,
Demonstration of intention in the reaching behaviour of
neonate humans, Nature, 228 (1970), pp. 67981
Bowerman, M., Early syntactic development: a
crosslinguistic study with special reference to Finnish
(Cambridge, Mass.: Cambridge UP 1973)
Bowlby, J., Maternal care and mental health (London:
HMSO 1951)
Bowlby, J., Attachment and loss, vol. 1, Attachment (1969);
vol. 2, Separation (1973); vol. 3, Loss (1980) (London:
Hogarth Press)
Bowlby, J., The making and breaking of affectional
bonds, British Journal of Psychiatry, 130 (1977), pp.
20110
Bradley L. and Bryant, P. E., Categorising sounds and
learning to read: a causal connection, Nature, 301
(1983), pp. 41921
Braine, M. D. S. and Rumain, B., Logical reasoning, in J.
H. Flavell and E. M. Markman (eds.), Hand-book of
Child Psychology, vol. 3, series ed. P. H. Mussen (New
York: Wiley 1983)
213

Brainerd, C. J., Piagets theory of intelligence (Engle-wood
Cliffs, NJ: Prentice Hall 1978)
Brainerd, C. J., The origins of the number concept (New
York: Praeger 1979)
Brainerd, C. J. (ed.), Childrens logical and mathematical
cognition (New York: Springer-Verlag 1982)
Bransford, J. D., Human Cognition (Belmont, Cal.:
Wadsworth 1979)
Bremner, J. G., The infants understanding of space, in M.
V. Cox (ed.), Are young children egocentric? (London:
Concord 1980)
Breslow, L., Reevaluation of the literature on the
development of transitive inferences, Psychological
Bulletin, 89 (2) (1981), pp. 32551
Bretherton, I., Young children in stressful situations, in G.
V. Coelho and P. I. Ahmed (eds.), Uprooting and
development (New York: Plenum Press 1980)
Bretherton, I., McNew, S. and Beeghly-Smith, M., Early
person knowledge as expressed in gestural and verbal
communication: when do infants acquire a theory of
mind?, in M. E. Lamb and L. R. Sherrod (eds.), Infant
Social Cognition (Hillsdale, NJ: Erlbaum Associates
1981)
Bromley, D. B., Natural language and the development of
the self, Nebraska Symposium on Motivation 1977
(Lincoln, Nebraska: University of Nebraska Press 1978)
Bronfenbrenner, U., Is early intervention effective? Facts
and principles of early intervention: a summary, in A. D.
B. and A. M. Clarke (eds.), Early experience (London:
Open Books 1976)
Bronfenbrenner, U., The ecology of human development
(Cambridge, Mass.: Harvard Univ. Press 1979)
Bronson, W. C., Developments in behaviour with age
mates during the second year of life, in M. Lewis and L.
A. Rosenblum (eds.), The origins of behaviour:
Friendship and peer relations (New York: John Wiley &
Sons 1975)
Broughton, J., Development of concepts of self, mind,
reality and knowledge, in W. Damon (ed.), New
directions for child development, 1 (1978), pp. 75100
Broughton, J. M., Genetic logic and the developmental
psychology and philosophical concepts, in J. M.
Broughton and D. J. Freeman-Moir (eds.), The
cognitive-developmental psychology of J. M. Baldwin
(Norwood, NJ: Ablex 1982)
214
Broughton, J. M. and Freeman-Moir, D. J. (eds.), The

cognitive developmental psychology of J. M. Baldwin
(Norwood, NJ: Ablex 1982)
Brown, A. L., The development of memory: knowing,
knowing about knowing, and knowing how to know, in
H. W. Reese (ed.), Advances in child development and
behaviour, vol. 10 (New York: Academic Press 1975)
Brown, A. L., Bransford, J. D., Ferrara, R. A. and
Campione, J. C.,
Learning, remembering and
understanding, in J. H. Flavell and E. Markman (eds.),
vol. 3 of the Handbook of Child Psychology, series ed. P.
Mussen (New York: Wiley 1983)
Brown, A. L. and Campione, J. C., Recognition memory
for perceptually similar pictures in preschool children,
Journal of Experimental Psychology, 95 (1972), pp. 55
62
Brown, A. L. and De Loache, J. S., Skills, plans and selfregulation in R. S. Siegler (ed.), Childrens thinking.
What develops? (Hillsdale, NJ: LEA 1978)
Brown, A. L., Palincsar, A. S. and Armbruster, B. B.,
Instructing comprehension-fostering activities in
interactive learning situations, in H. Mandle, N. Stein
and T. Trabasso (eds.), Learning and comprehension of
texts (Hillsdale, NJ: Erlsbaum 1984)
Brown, A. L. and Scott, M. S., Recognition memory for
pictures in preschool children, Journal of Experimental
Child Psychology, 11 (1971), pp. 40112
Brown, G. and Desforges, C., Piagets theory: a
psychological critique (London: Routledge & Kegan
Paul 1979)
Brown, G. W. and Harris, T., Social origins of depression: a
study of psychiatric disorder in women (London:
Tavistock 1978)
Brown, M. and Madge, N., Despite the Welfare State
(London: Heinemann 1982)
Bruner, J. S., On cognitive growth, in J. S. Bruner, R. R.
Olver and P. M. Greenfield (eds.), Studies in cognitive
growth (New York: Wiley 1966)
Bruner, J. S., Toward a theory of instruction (New York:
Norton 1968)
Bruner, J. S., Organisation of early skilled action, Child
Development, 44 (1973), pp. 111
Bruner, J. S., From communication to language a
psychological perspective, Cognition, 3 (1976), pp.
25587
Bruner, J. S., Goodnow, J. J. and Austin, G. A., A study of
thinking (New York: Wiley 1956)
Bibliography
Bruner, J., Jolly, A., and Sylva, K. (eds.), Play: its role in
development and evolution (Harmondsworth: Penguin
1976)
Bruner, J. S., Olver, R. and Greenfield, P. M. (eds.) Studies
in cognitive gro wth (New York: Wiley 1966)
Bryant, B., Harris, M. and Newton, D., Children and
minders (London: Grant McIntyre 1980)
Bryant, P. E., Perception and understanding in young
children (London: Methuen 1974)
Bryant, P. E., The role of conflict and of agreement
between intellectual strategies in childrens ideas about
measurement, British Journal of Psychology, 73
(1982), pp. 24352
Bryant, P. E. and Bradley, L., Why children sometimes
write words which they do not read, in U. Frith (ed.),
Cognitive processes in spelling (London: Academic
Press 1980)
Bryant, P. E. and Bradley, L., Psychological strategies and
the development of reading and writing, in M. Martlew
(ed.), The psychology of written language (Chichester:
Wiley 1983)
Bryant, P. E. and Bradley, L., Childrens reading problems
(Oxford: Blackwell 1985)
Bryant, P. E., Jones, P., Claxton, V. and Perkins, J.,
Recognition of shapes across modalities by infants,
Nature, 240 (1972), pp. 3034
Bryant, P. E. and Trabasso, T., Transitive inferences and
memory in young children, Nature, 232 (1971), pp.
4568
Bullowa, M., Before speech: the beginnings of
interpersonal communication (Cambridge: Cambridge
UP 1979)
Burghardt, G. M., On the origins of play, in P. K. Smith
(ed.), Play in animals and humans (Oxford: Basil
Blackwell 1984)
Burnett, J. (ed.), Destiny obscure: autobiographies of
childhood, education and family from the 1820s to the
1920s (London: Allen Lane 1982)
Buss, A. H. and Plomin, R., A temperament theory of
personality development (New York: Wiley 1975)
Butterworth, G., Object identity in infancy: the interaction
of spatial location codes in determining search errors,
Child Development, 46 (1975), pp. 86670
Butterworth, G. Object disappearance and error in Piagets
Stage IV task, Journal of Experimental Child
Psychology, 22 (1977), pp. 391401
Butterworth, G., Thought and things: Piagets theory, in

A. Burton and J. Radford (eds.), Perspectives on thinking
(London: Methuen 1978)
Butterworth, G. and Hicks, L., Visual proprioception and
postural stability in infancy, Perception, 6 (1977), pp.
25562
Butterworth, G. and Light, P. (eds.) Social cognition:
Studies of the development of understanding (Brighton:
Harvester Press 1982)
Byers, J. A., Play in ungulates, in P. K. Smith (ed.), Play
in animals & humans (Oxford: Blackwell 1984)
Bynum, T. W., Thomas, J. A. and Weitz, L. J., Truthfunctional logic in formal operational thinking,
Developmental Psychology, 7 (1972), pp. 12932
Calkins, L., Lessons from a child: on the teaching and
learning of writing
(Exeter, New Hampshire:
Heinemann 1983)
Callanan, M. A., How parents label objects for young
children: the role of input in the acquisition of category
hierarchies, Child Development, 56 (1985), pp. 50823
Camaras, L. A., Facial expressions used by children in a
conflict situation, Child Development, 48 (1977), pp.
14315
Camaras, L. A., Childrens understanding of facial
expressions used during conflict encounters, Child
Development, 51 (1980), pp. 87985
Campos, J. J., Barrett, K. C., Lamb, M. E., Goldsmith, H. H.
and Sternberg, C., Socio-emotional development, in
M. M. Haith and J. J. Campos (eds.), vol. 2 of the
Handbook of Child Development (New York: Wiley
1983)
Carpenter, T. P. and Moser, J. M., The acquisition of
addition and subtraction concepts, in R. Lesh and M.
Landau (eds.), Acquisition of mathematics concepts and
processes (New York: Academic Press 1983)
Carpenter, T. P., Moser, J. M. and Romberg, T. A. (eds.),
Addition and Subtraction: a cognitive perspective
(Hillsdale, NJ: Erlbaum 1982)
Case, R., The process of stage-transition: a neoPiagetian
view, in R. J. Sternberg (ed.), Mechanisms of cognitive
development (New York: Freeman 1984)
Case, R., Intellectual development: birth to adulthood
Cavanaugh, J. C. and Perlmutter, M., Metamemory: a
critical examination, Child Development, 53 (1982), pp.
1128
215

Chall, J. S., Learning to read: the great debate (New York:
McGraw Hill 1967)
Chall, J. S., Stages of reading development (New York:
McGraw Hill 1983)
Chi, M. T. H., Knowledge structures and memory
development, in R. S. Siegler (ed.), Childrens thinking:
what develops? (Hillsdale, NJ: LEA 1978)
Chi, M. T. H. and Koeske, R. D., Network representation
of a childs dinosaur knowledge, Developmental
Psychology, 19 (1) (1983), pp. 2939
Chomsky, N., Reflections on language (London: Temple
Smith/Fontana 1976)
CIBA foundation symposium: Ruth Porter and Geralyn M.
Collins (eds.), Temperamental differences in infants and
young children (London: Pitman 1982)
Cicirelli, V. G., Concept learning of young children as a
function of sibling relationships to the teacher, Child
Cicirelli, V. G. Sibling structure and intellectual ability,
Cicirelli, V. G., Sibling influence throughout the life span,
in M. Lamb and B. Sutton-Smith (eds.), Sibling
relationships (Hillsdale, NJ: Erlbaum 1982)
Clark, E. V., Meanings and concepts, in J. H. Flavell and
E. M. Markman (eds.), vol. 3 of the Handbook of Child
Psychology, series ed. P. H. Mussen (New York: Wiley
1983)
Clark, H. H. and Clark, E. V., Psychology and language
(New York: Harcourt Brace Jovanovich 1977)
Clark, M. M., Young fluent readers (London: Heinemann
1976)
Clark, R. and Delia, J., The development of functional
persuasive skills in childhood and early adolescence,
Child Development, 47 (1976), pp. 1008 14
Clark, R. W., Freud: the man and the cause (London:
Jonathan Cape 1980)
Clarke, A. M. and Clarke, A. D. B., Early experience: myth
and evidence (London: Open Books 1976)
Clarke, R. V. G., Delinquency, environment and
intervention, Journal of Child Psychology &
Psychiatry, 26 (4) (1985), pp. 50524
Clarke-Stewart, A., Day Care (London: Fontana 1982)
Clarke-Stewart, A. and Fein, G., Early childhood
programs, in M. M. Haith and J. J. Campos, vol. 2 of the
Handbook of ChildPsychology, series ed. P. H. Mussen
216
Clay, M., Reading: the patterning of complex behaviour

(Heinemann 1979)
Clay, M., Getting a theory of writing, in B. Kroll and G.
Wells (eds.), Explorations in the development of writing
(Chichester: Wiley 1983)
Coffield, F., Robinson, P. and Sarsby, J., A cycle of
deprivation? A case study of four families (London:
Heinemann 1980)
Cohen, G., The psychology of cognition (London:
Cole, M. and Griffin, P., Cultural amplifiers reconsidered,
in D. R. Olsen (ed.), The social foundations of language
and thought (New York: W. W. Norton 1980)
Cole, M. and Means, B., Comparative studies of how
people think: an introduction (Cambridge, Mass.: HVP
1981)
Coleman, J., The nature of adolescence (London: Methuen
1980)
Coleman, M., Take one, leave one: young childrens
concepts of subtraction (Unpublished M.Ed. thesis,
University of Bristol 1982)
Collerson, J., One child and one genre: developments in
letter writing, in B. Kroll and G. Wells (eds.),
Explorations in the development of writing (Chichester:
Wiley 1983)
Collins, A. and Smith, E. E., Teaching the process of
reading comprehension, in D. K. Detterman and R. J.
Sternberg (eds.), How and how much can intelligence be
increased? (Norwood, NJ: Ablex 1982)
Condon, W. S. and Sander, L. W., Neonate movement is
synchronised with adult speech: interactional
participation and language acquisition, Science, 183
(1974), pp. 99101
Conroy, J. S., Towards a framework for language
functioning in learning mathematics, Paper given at the
Fifth International Congress on Mathematics Education
Adelaide, Australia, August 1984
Cook, M., Levels of personality (London: Holt, Rinehart &
Winston 1984)
Corran, G. and Walkerdine, V., The practice of reason, vol.
1 (London: Thomas Coran Research Unit, 1981)
Corrigan, R., Methodological issues in language
acquisition research with very young children,
Developmental Review, 2 (1982), pp. 16288
Corsaro, W. A., Friendship in the nursery school: social
organisation in a peer environment, in S. R. Asher and J.
Bibliography
M. Gottman (eds.), The development of childrens
friendships (Cambridge: Cambridge UP 1981)
Coveney, P., The image of childhood (Harmondsworth:
Penguin 1967)
Cowie, H. (ed.), The development of childrens imaginative
writing (London: Croom Helm 1984)
Cox, M. V. (ed.), Are young children egocentric? (London:
Batsford 1980)
Crook, C. K., Taste perception in the newborn infant,
Infant behaviour and development, 1, pp. 5269
Cross, T., Mothers speech adjustments: the contribution
of selected child-listener variables, in C. Snow and C.
Ferguson (eds.), Talking to children: language input and
acquisition (Cambridge: Cambridge UP 1977)
Cross, T., Mothers speech and its association with rate of
linguistic development in young children, in N.
Waterson and C. Snow (eds.), The development of
communication (Chichester: Wiley 1978)
Crowder, R. G., The Psychology of reading: an introduction
(Oxford: OUP 1982)
Csiksentmihalyi, M., The concept of flow, in B. SuttonSmith (ed.), Play and Learning (New York: Gardner
Press 1979)
Curtiss, S., Genie. A psycholinguistic study of a modernday wild child (New York: Academic Press 1977)
Dale, P. S., Language Development (New York: Holt,
Rinehart & Winston 1976)
Dalton, K., Once a month: the menstrual syndrome
(London: Fontana 1983)
Damon, W. and Hart, D., The development of selfunderstanding from infancy through adolescence,
Davie, C. E., Hutt, C. J., Vincent, E. and Mason, M., The
Young Child at Home (Windsor: NFER Nelson 1984)
Davies, B., Life in the classroom & playground (London:
Routledge & Kegan Paul 1979)
Dawkins, R., The selfish gene (Oxford: OUP 1976)
Decarie, T. G., A study of the mental and emotional
development of the thalidomide child, in B. M. Foss
(ed.), Determinants of infant behaviour, vol. 4 (London:
Methuen 1969)
Decasper, A. J. and Fifer, W. P., Of human bonding:
newborns prefer their mothers voices, Science, 208
(1980), pp. 11746
Delamont, S., The sociology of women (London: George
Allen & Unwin 1980)
Delamont, S., Interaction in the classroom (London:

Methuen 1983)
De Mause, L. (ed.), The history of childhood (London:
Souvenir Press 1976)
Dempster, F. N., Memory span: sources of individual and
developmental differences, Psychological Bulletin, 89
(1981), pp. 63100
de Villiers, P. A. and de Villiers, J. G., Early language
Dodge, K. A., Social cognition and childrens aggressive
behaviour, Child Development, 51 (1980), pp. 16270
Doise, W. and Mackie, D., On the social nature of
cognition, in J. P. Forgas (ed.), Social cognition (New
York: Academic Press 1981)
Doise, W. and Mugny, G., The social development of the
intellect (Oxford: Pergamon 1984)
Donaldson, M., Childrens minds (London: Fontana 1978)
Donaldson, M., Conservation: what is the question?,
British Journal of Psychology, 73 no.2 (1982), pp. 199
208
Donaldson, M. and Balfour, G., Less is more: a study of
language comprehension, British Journal of
Psychology, 59 (1981), pp. 46171
Donovan, A., Oddy, M., Pardoe, R. and Ades, A.,
Employment status and psychological wellbeing: a
longitudinal study of 16-year-old school leavers,
Journal of Child Psychology and Psychiatry, 27, (1)
(1986), pp. 6576
Dorval, B. and Eckerman, C. O., Developmental trends in
the quality of conversation achieved by small groups of
acquainted peers, Monographs of the Society for
Research in Child Development, 49 (2) (1984), serial no.
206
Douvan, E. and Adelson, J., The adolescent experience
Dowdney, L., Skuse, D., Rutter, M., Quinton, D. and
Mrazek, D., The nature and qualities of parenting
provided by women raised in institutions, Journal of
Child Psychology & Psychiatry, 26 (4) (1985) pp. 599
626
Downing, J., Reading and reasoning (Edinburgh:
Chambers 1979)
Doyle, A. B., Gold, B. and Moskowitz, D. S. (eds.),
Children in families under stress (San Francisco: JosseyBass 1984)
Dunn, J. F., The first year of life: continuities in individual
differences, in D. Shaffer and J. F. Dunn (eds.) The first
217

year of life: psychological and medical implications of
early experience (New York: Wiley 1979)
Dunn, J. F., Individual differences in temperament, in M.
Rutter (ed.), The scientific foundations of developmental
psychiatry (London: Heinemann 1980)
Dunn, J., Sibling relationships in early childhood, Child
Development, 54 (1983), pp. 787 811
Dunn, J., Sisters and brothers (London: Fontana 1984)
Dunn, J. and Kendrick, C., Siblings: love, envy and
understanding (London: Grant McIntyre 1982)
Dweck, S. and Bush, E. S., Sex differences in learned
helplessness I, Developmental Psychology, 12 (1976),
pp. 14756
Dweck, C. S., Davidson, W., Nelson, S. and Enna, B., Sex
differences in learned helplessness II & III,
Dweck, C. S. and Elliott, E. S., Achievement motivation,
in E. M. Hetherington (ed.), Socialisation, personality
and social development, vol. 4 of The Handbook of
Child Psychology, P. H. Mussen (ed.) (New York: Wiley
1983)
Dweck, C. S. and Goetz, T. E., Attributions and learned
helplessness, in J. H. Harvey, W. Ickles and R. F. Kidd
(eds.), New directions in attribution research, vol. 2
Edwards, C. P. and Lewis, M., Young childrens concepts
of social relations: social functions and social objects, in
M. Lewis and L. A. Rosenblum (eds.), The child and its
family: genesis of behaviour, vol. 2 (New York: Plenum
Press 1979)
Ehrhardt, A. A., Gender differences: a biosocial
perspective, Nebraska Symposium on Motivation 1984
(Lincoln, Nebraska: University of Nebraska Press 1985)
Ehrhardt, A. and Meyer-Bahlburg, H., Effects of prenatal
sex hormones on gender-related behaviour, Science,
211 (1981), pp. 131218
Eimas, P. D., Siqueland, E. R., Jusczyk, P. and Vigorito, J.,
Speech Perception in Infants, Science, 171 (1971), pp.
30226
Elder, G. H., Children of the great Depression: Social
change in life experience (Chicago: University of
Chicago Press 1974)
Elder, G. H., Nguyen, T. V. and Caspi, A., Linking family
hardship to childrens lives, Child Development, 56
(1985), pp. 36175
Ellis, A. W., Reading, writing and dyslexia: a cognitive
analysis (London: Erlbaum 1984)
218
Emler, N. and Valiant, G., Social interaction and cognitive

conflict in the development of spatial co-ordination
skills, British Journal of Psychology, 73 (1982), pp.
295304
Emler, N. and Dickinson, J., Childrens representation of
economic inequalities: the effects of social class, British
Journal of Developmental Psychology, 3 (2) (1985), pp.
1918
Ennis, R. H., Conceptualization of childrens logical
competence: Piagets propositional logic and an
alternative proposal, in L. S. Siegel and C. J. Brainerd
(eds.), Alternatives to Piaget (New York: Academic
Press 1978)
Erikson, E., Childhood and Society (Harmondsworth:
Penguin 1963)
Erikson, E., Identity: youth and crisis (London: Faber
1968)
Eron, L. D. Parentchild interaction, television violence
and aggression of children, American Psychologist, 37
(1982), pp. 197211
Ervin-Tripp, S., Some features of early child adult
dialogues, Language in Society, 7 (1978), pp. 35773
Essen, J. and Wedge, P., Continuities in childhood
disadvantage (London: Heinemann 1982)
Etaugh, C., Effects of maternal employment on children: a
review of recent research, Merrill-Palmer Quarterly,
20 (1974), pp. 7198
Eysenck, H. J., The biological basis of personality
(Springfield, Illinois: Charles C. Thomas 1967)
Fagan, J. F., Infants delayed recognition memory and
forgetting, Journal of Experimental Child Psychology,
16 (1973), pp. 42450
Fagan, J. F., Infants recognition of invariant features of
faces, Child Development, 45 (1976), pp. 3516
Fagen, R., Animal Play Behaviour (New York: OUP 1981)
Falbo, T., Only children in America, in M. Lamb and B.
Sutton-Smith (eds.), Sibling relationships (Hillsdale,
NJ: Erlbaum 1982)
Fehr, L. A., Methodological inconsistencies in the
measurement of spatial perspective taking ability: a
cause for concern, Human Development, 21 (1978), pp.
30215
Fein, G. G., Pretend play: an integrative review, Child
Development, 52 (1981), pp. 1095118
Feldman, C. F. and Toulmin, S., Logic and the theory of
mind, in Nebraska Symposium on Motivation (Lincoln:
Univ. of Nebraska Press 1976)
Bibliography
Feshbach, S., Feshbach, N. D., Cohen, R. S. and Hoffman,
M., The antecedents of anger: a developmental
approach, in R. M. Kaplan, V. J. Konecni and R. W.
Novaco (eds.), Aggression in children and youth (The
Hague: Martinus Nijhoff 1984)
Field, T. M., Effects of early separation, interactive
deficits, and experimental manipulations on infant
mother face-to-face interaction, Child Development, 48
(1977), pp. 76371
Field, T. M., Sostek, A. M., Vietze, P. and Leiderman, P. H.,
Culture and early interactions (Hillsdale, NJ: Erlbaum
1981)
Fine, G. A., The natural history of preadolescent male
friendship groups, in H. C. Foot, A. J. Chapman and J.
R. Smith (eds.), Friendship and social relations in
children (London: Wiley 1980)
Bibliography
Fine, G. A., Friends, impression management, and
preadolescent behaviour, in S. R. Asher and J. M.
Gottman (eds.), The development of childrens
friendships (Cambridge: Cambrige UP 1981)
Finkelstein, N. W. and Ramey, C. T., Learning to control
the environment in infancy, Child Development, 48
(1977), pp. 80619
Fivush, R., Negotiating classroom interaction, The
Quarterly Newsletter of the Laboratory of Comparative
Human Cognition, 5 (4) (1983), pp. 837
Fivush, R., Learning about school: the development of
kindergartners school scripts, Child Development, 55
(1984), pp. 1697709
Flavell, J. H., The developmental psychology of Jean
Piaget (Princeton, NJ: Van Nostrand 1963)
Flavell, J. H., Cognitive development (Engle-wood Cliffs,
NJ: Prentice Hall 1977)
Flavell, J. H., The development of knowledge about visual
perception, Nebraska Symposium on Motivation 1977
(Lincoln, Nebraska: University of Nebraska Press
1978a)
Flavell, J. H., Comments on Brown & DeLoachs paper in
R. S. Siegler (ed.), Childrens thinking: what develops?
(Hillsdale, NJ: Erlbaum 1978b)
Flavell, J. H., On cognitive development, Child
Flavell, J. H., Botkin, P. T., Fry, C. L., Wright, J. W. and
Jarvis, P. E., The development of role-taking and
communication skills in children (New York: Wiley
1968)
Flavell, J. H. and Ross, L. (eds.), Social cognitive

development (Cambridge: Cambridge UP 1981)
Flavell, J. H. and Markman, E. M. (eds.), Cognitive
development, vol. 3 of the Handbook of Child
1983)
Flavell, J. H. and Wellman, H. M., Metamemory, in R. V.
Kail and J. W. Hagen (eds.), Perspectives on the
development of memory and cognition (Hillsdale, NJ:
Erlbaum 1977)
Foot, H. C., Chapman, A. J. and Smith, J. R., Friendship
and social relations in children (Chichester: Wiley
1980)
Ford, M. E., The construct validity of egocentrism,
Psychological Bulletin, 86 (1979), pp. 116988
Forgas, J. (ed.), Social cognition (New York: Academic
Press 1981)
Francis, H., Learning to read: literate behaviour and
orthographic knowledge (London: George Allen &
Unwin 1982)
Fransella, F., The theory and measurement of personal
constructs, in K. Franville-Grossman (ed.), Recent
advances in clinical psychiatry (London: Churchill
Livingstone 1976)
Frederiksen, C. H. and Dominic, J. F., Writing: the nature,
development and teaching of written communication,
vol. 2, Writing: process, development and
communication (Hillsdale, NJ: Erlbaum 1981)
Freeman, N. H. and Cox, M. V. (eds.), Visual order: the
nature and development of pictorial representation
(Cambridge: Cambridge UP 1985)
Freeman, N., Lloyd, S. and Sinha, C. G., Infant search tasks
reveal early concepts of containment and canonical
usage of objects, Cognition, 8 (1980), pp. 24362
Freud, S., Three essays on the theory of sexuality, vol. 7
in J. Strachey (ed.), The standard edition of the complete
Psychological works of Sigmund Freud (London:
Hogarth Press 1905/53)
Freud, S., The ego and the id, vol. 19 (1923/61)
Freud, S., The dissolution of the Oedipus complex, vol.
19 (1924)
Freud, S., Some psychical consequences of the anatomical
distinction between the sexes, vol. 19 (1924)
Freud, S., Civilisation and its discontents, vol. 21 (1930/
61)
Freud, S., New introductory lectures on psychoanalysis,
vol. 22 (1933/64)
219

Freud, S., Analysis of a phobia in a five year old boy, vol.
10 (1955)
Frith, U., Reading and spelling skills, in M. Rutter (ed.),
Scientific foundations of developmental psychiatry
(London: Heinemann 1980a)
Frith, U. (ed.), Cognitive processes in spelling (London:
Academic Press 1980b)
Fry, D. B., The development of the phonological system in
the normal and the deaf child, in F. Smith and G. A.
Miller (eds.), The genesis of language: a
psycholinguistic approach (Camb., Mass.: MIT Press
1966)
Furman, W. and Buhrmester, D., Childrens perceptions of
the qualities of sibling relationships, Child
Development, 56 (1985), pp. 44861
Furth, H., The world of grown ups (New York: Elsevier
North Holland 1980)
Fuson, K. C. and Hall, J. W., The acquisition of early
number word meanings: a conceptual analysis and
review, in H. P. Ginsburg (ed.), The development of
mathematical thinking (New York: Academic Press
1983)
Gallup, G. G., Self-recognition in primates: a comparative
approach to the bidirectional properties of
consciousness, American Psychologist, 32 (1977), pp.
32938
Gallup, G., Self-recognition in chimpanzees and man, in
M. Lewis and L. Rosenblum (eds.), The child and its
family: the genesis of behaviour, vol. 2 (New York:
Plenum 1979)
Galton, M., Simon, B. and Croll, P., Inside the primary
classroom (London: Routledge & Kegan Paul 1980)
Garbarino, J., Children and families in the social
environment (New York: Aldine 1982)
Garmezy, N. and Rutter, M., Acute reactions to stress, in
M. Rutter and L. Hersov (eds.), Child & adolescent
psychiatry: modern approaches (Oxford: Blackwell
1985)
Garvey, C., Requests and responses in childrens speech,
Journal of Child Language, 2 (1975), pp. 4163
Garvey, C., Play (London: Fontana/Open Books 1977)
Garvey, C., Childrens talk (London: Fontana 1984)
Gelman, R., Basic numerical abilities, in Sternberg, R. J.
(ed.), Advances in the psychology of human intelligence,
vol. 1 (Hillsdale, NJ: Erlbaum 1982)
Gelman, R. and Baillargeon, R., A review of some
Piagetian concepts, in J. H. Flavell and E. M. Markman
220
(eds.), Handbook of Child Psychology, vol. 3, series ed.

P. Mussen (New York: Wiley 1983)
Gelman, R. and Gallistel, C. R., The childs understanding
of number (Cambridge, Mass.: Harvard UP 1978)
Gelman, R. and Spelke, E., The development of thoughts
about animate and inanimate objects: implications for
research on social cognition, in J. H. Flavell and L. Ross
Cambridge UP 1981)
Gesell, A., Wolf child and human child (London: The
Scientific Book Club 1942)
Gewirth, A., Reason and morality (Chicago and London:
Univ. of Chicago Press 1978)
Gholson, B. and Beilin, H., A developmental model of
human learning, in H. W. Reese and L. P. Lipsitt (eds.),
Advances in Child Development and Behaviour, vol. 13
(1979)
Gibson, E. J., Principles of perceptual learning and
development (New York: Appleton-Century-Crofts
1969)
Gibson, J. J., The ecological approach to visual perception
(Boston: Houghton-Mifflin 1979 )
Giddens, A. and Held, D., Classes, power and conflict
(Basingstoke: Macmillan 1982)
Giles, H. and Powesland, P. F., Speech style and social
evaluation (London: Academic Press 1975)
Gilligan, C., In a different voice: womens conceptions of
self and morality, Harvard Educational Review, 47
(1977), pp. 481517
Gilligan, C., In a different voice: Psychological theory and
womens development (Camb., Mass.: Harvard UP
1982)
Ginsburg, H. P. (ed.), The development of mathematical
thinking (New York: Academic Press 1983)
Gold, D. and Andres, D. Comparisons of adolescent
children with employed and unemployed mothers,
Merrill-Palmer Quarterly, 24 (1978), pp. 24354
Goldman, R. and Goldman, J., Childrens sexual thinking
(London: Routledge & Kegan Paul 1982)
Goldsmith, H. H., Genetic influences on personality from
infancy to adulthood, Child Development, 54 (1983),
pp. 33155
Goodnow, J. J., Compensation arguments on conservation
tasks, Developmental Psychology, 8 (1) (1973), p. 140
Goodnow, J., and co-authors, Adult social cognition, in
M. Perlmutter (ed.), Cognitive perspectives on
childrens social and behavioural development,
Bibliography
Minnesota Symposium on Child Psychology (vol. 18)
(Hillsdale, NJ: Lawrence Erlbaum Associates, in press)
Goody, J., The domestication of the savage mind
Goody, J. and Watt, I., The consequences of literacy, in J.
Goody (ed.), Literacy in traditional societies
Gordon, J. C. B., Verbal deficit: a critique (London: Croom
Helm 1981)
Gottlieb, G., The psychobiological approach to
developmental issues, in M. M. Haith and J. J. Campos
(eds.) vol. 2 of the Handbook of Child Psychology (New
York: Wiley 1983)
Gould, S. J., Ontogeny and phylogeny (Cambridge, Mass.:
Harvard UP 1977)
Gould, S. J., The mismeasure of man (Harmondsworth:
Penguin 1984)
Gould, S. J. and Lewontin, R. C., The spandrels of San
Marco and the Panglossian paradigm: a critique of the
adaptationist programme, Proceedings of the Royal
Society of London, B205 (1979), pp. 58198
Gratch, G., Appel, K. J., Evans, W. F., LeCompte, G. K. and
Wright, N. A., Piagets Stage IV object concept error:
evidence of forgetting of object conception?, Child
Graves, D., Writing: teachers and children at work
(Heinemann 1983)
Grice, H. P., Logic and conversation, in P. Cole and J. L.
Morgan (eds.), Syntax and semantics, vol. 3, Speech acts
Griffin, C., Typical girls? Young women from school to the
job market (London: Routledge & Kegan Paul 1985)
Griffiths, M. and Wells, G., Who writes what and why?, in
B. Kroll and G. Wells (eds.), Explorations in the
development of writing (Chichester: Wiley 1983)
Groen, G. J. and Parkman, J. M., A chronometric analysis
of simple addition, Psychological Review, 97 (1972),
pp. 32943
Grotevant, H. D. and Cooper, C. R., Patterns of interaction
in family relationships and the development of identity
exploration in adolescence, Child Development, 56
(1985), pp. 41528
Guttentag, M. and Longfellow, C., Childrens social
attributions development and change, in C. B. Keasey
(ed.), Nebraska Symposium on Motivation vol. 25
(Lincoln, Nebraska: Univ. of Nebraska Press 1977)
Hagen, J. W. and Hale, G. H., The development of attention

in children, in Minnesota Symposium on Child
Psychology, vol. 7 (1973), pp. 11739
Haith, M. M. and Campos, J. J. (eds.) Infancy and
developmental Psychobiology, vol. II, Handbook of
Child Psychology, ed. P. H. Mussen (New York: Wiley
1983)
Halliday, M. A. K. Learning how to mean (London: Edward
Arnold 1975)
Halverson, C. F. and Waldrop, M. F., Relations between
preschool activity and aspects of intellectual and social
behaviour at age 71/2, Developmental Psychology, 12
(1976), pp. 10712
Hamlyn, D. W., Experience and the growth of
understanding (London: Routledge & Kegan Paul
1978)
Hammersley, M. and Woods, P., Life in school: the
sociology of pupil culture (Milton Keynes: Open
Hampson, S. E., The construction of personality (London:
Hardyment, C., Dream babies: from Locke to Spock
(London: Jonathan Cape 1983)
Hargreaves, D., Social relations in a secondary school
Harlow, H. F. and Harlow, M. K., Effects of various
motherinfant relationships on rhesus monkey
behaviours, in B. M. Foss (ed.), Determinants of Infant
Behaviour, vol. 4 (London: Methuen 1969)
Harris, P. Perception and cognition in infancy, in K.
Connely (ed.), Psychology Survey No. 2 (London:
George Allen & Unwin 1979)
Harris, P. L.,The child as psychologist, in M. Donaldson,
R. Grieve and C. Pratt (eds.), Early childhood
development and education (Oxford: Blackwell 1983a)
Harris, P. L., Infant cognition, in M. M. Haith and J. J.
1983b)
Harter, S., A model of intrinsic mastery motivation in
children: individual differences and developmental
change, Minnesota Symposium on Child Psychology,
vol. 14 (Hillsdale, NJ: Erlbaum 1981)
Hartup, W. W., Children and their friends, in H. McGurk
(ed.), Issues in childhood social development (London:
Methuen 1978)
221

Hartup, W. W., Peer relations, in E. M. Hetherington (ed.),
vol. 4 of the Handbook of Child Psychology, series ed. P.
H. Mussen (New York: Wiley 1983)
Harvey, P. G. Lead and childrens health recent research
and future questions, Journal of Child Psychology and
Psychiatry, 25 (1984), pp. 51722
Hawkins, P. R., Social class, the nominal group and
reference, in B. Bernstein (ed.), Class, codes and
control, vol. 2 (London: Routledge & Kegan Paul 1973)
Hawkins, P. R., Social class, the nominal group and verbal
strategies (London: Routledge & Kegan Paul 1977)
Hayward, C., Literary theme development in the nursery
classroom (Unpublished M.Ed. thesis, University of
Bristol 1982)
Heath, S. B., What no bedtime story means: narrative skills
at home and at school, Language in Society, 11 (1982),
pp. 4976
Heath, S. B., Ways with words (Cambridge: Cambridge UP
1983)
Heider, E. R., Focal color areas and the development of
color names, Developmental Psychology, 4 (1971), pp.
44755
Heider, E. R., Universals in color naming, Journal of
Experimental Psychology, 93 (1972), pp. 1020
Heider, F., The psychology of interpersonal relations (New
York: Wiley 1958)
Hersov, L., Adoption and fostering, in M. Rutter and L.
Hersov (eds.); Child and adolescent psychiatry: modern
approaches (Oxford: Blackwell 1985)
Hetherington, E. M., Divorce: a childs perspective,
American Psychologist, 34 (1979), pp. 8518
Hinde, R. A., Interpersonal relationships: in quest of a
science, Psychological Medicine, 8 (1978), pp. 37386
Hinde, R. A. Towards understanding relationships
(London: Academic Press 1979)
Hinde, R. A., Attachment: some conceptual and biological
issues, in C. M. Parkes and J. StevensonHinde (eds.),
The place of attachment in human behaviour (London:
Tavistock 1982a)
Hinde, R. A., Ethology (London: Fontana 1982b)
Hinde, R. A., Ethology and child development, in M. M.
Haith and J. J. Campos (eds.), vol. 2 of the Handbook of
Child Psychology (New York: Wiley 1983)
Hinde, R. A., Easton, D. F., Meller, R. E. and Tamplin A.,
Nature and determinants of preschoolers differential
behaviour to adults and peers, British Journal of
222
Hinde, R. A. and Tamplin, A., Relations between mother

child interaction and behaviour in preschool, British
Journal of Developmental Psychology, 1 (1983), pp.
23157
Hines, M., Prenatal gonadal hormones and sex differences
in human behaviour, Psychological Bulletin, 92 (1982),
pp. 5680
Hirsh-Pasek, K., Gleitman, L. R. and Gleitman, H., What
did the brain say to the mind? A study of the detection and
report of ambiguity by young children, in A. Sinclair, R.
J. Jarvella and W. J. M. Levelt (eds.), The childs
conception of language (Berlin: Springer-Verlag 1978)
Hofer, M. A., The roots of human behavior: an introduction
to the psychology of early development (San Francisco:
W. H. Freeman 1981)
Hoffman, M. L., Developmental synthesis of affect and
cognition and its implications for altruistic motivation,
Hoffman, M. L., Empathy, its development and prosocial
implications, in C. B. Keasey (ed.), Nebraska
Symposium on Motivation, vol. 25 (Lincoln: Univ. of
Nebraska Press 1977)
Hoffman, M. L., Is altruism part of human nature?,
Journal of Personality and Social Psychology, 40
(1981), pp. 12137
Hoffman, M. L. and Levine, L. E., Early sex differences in
empathy, Developmental Psychology, 12 (1976), pp.
5578
Hoggart, R., The uses of literacy (Harmondsworth:
Penguin 1957)
Horn, J. M., The Texas Adoption Project: adopted children
and their intellectual resemblance to biological and
adoptive parents, Child Development, 54 (1983), pp.
26877
Hotopf, W. H. N., An examination of Piagets theory of
perception, in B. A. Geber (ed.), Piaget and knowing:
studies in genetic epistemology (London: Routledge &
Kegan Paul 1977)
Houlbrooke, R., The English family 14501700 (London:
Longmans 1984)
Howe, M. J. A., A teachers guide to the psychology of
learning (Oxford: Blackwell 1984)
Hughes, M., What is difficult about learning arithmetic?,
in M. Donaldson, R. Grieve and C. Pratt (eds.), Early
childhood development and education (Oxford:
Blackwell 1983)
Bibliography
Hughes, M., Carmichael, H., Pinkerton, G. and Tizard, B.,
Recording childrens conversations at home and at
nursery school, Journal of Child Psychology and
Psychiatry, 20 (1979), pp. 22532
Hughes, M. and Donaldson, M., The use of hiding games
for studying the coordination of viewpoints,
Educational Review, 31 (1979), pp. 13340
Hughes, M., Mayall, B., Moss, P., Perry, J., Petrie, P. and
Pinkerton, G., Nurseries Now (Harmondsworth: Penguin
1980)
Humphreys, A. P. and Smith, P. K., Rough-andtumble in
preschool and playground, in P. K. Smith (ed.), Play in
animals and humans (Oxford: Blackwell 1984)
Huston, A., Sex-typing in E. M. Hetherington (ed.), vol. 3
of the Handbook of Child Psychology (New York: Wiley
1983)
Hutt, C., Exploration and play, in B. Sutton-Smith (ed.),
Play and learning (New York: Gardner Press 1979)
Isbell, B. J. and McKee, L., Societys Cradle: an
anthropological perspective on the socialisation of
cognition, in J. Sants (ed.), Developmental Psychology
and Society (London: Macmillan 1980)
Istomina, Z. M., The development of voluntary memory in
preschool-age children, Soviet Psychology, 13 (1975),
pp. 564
Jackson, B., Fatherhood (London: George Allen & Unwin
1984)
Jackson, B. and Jackson, S., Childminder (London:
Jahoda, G., The construction of economic reality by some
Glaswegian children, European Journal of Social
Psychology, 9 (1979), pp. 11527
Jahoda, G., The development of thinking about
socioeconomic systems, in H. Tajfel (ed.), The social
dimension, vol. 1 (Cambridge: Cambridge UP 1984)
Jarman, C., The development of handwriting skills
(London: Basil Blackwell 1979)
Jencks, C., Inequality: a reassessment of the effect of family
and schooling in America (Harmondsworth: Penguin
1975)
Jenkins, R., Lads, citizens and ordinary kids: workingclass youth life-styles in Belfast (London: Routledge &
Kegan Paul 1983)
Jennings, K. D., Harmon, R. J., Morgan, G. A., Gaiter, J. L.
and Yarrow, L. J. Exploratory play as an index of
mastery motivation: relationships to persistence,
cognitive functioning and environmental measures,

Johns, M. A., The role of the vomeronasal system in
mammalian reproductive physiology, in D. MullerSchwarze and R. M. Silverstein (eds.), Chemical Signals
(New York: Plenum 1980)
Johnson, C. N. and Wellman, H. M., Childrens
developing conceptions of the mind and brain, Child
Development, 53 (1982), pp. 22234
Jones, O. H. M., A comparative study of mother child
communication with Downs syndrome and normal
infants, in D. Shaffer and J. Dunn (eds.), The first year
of life (Chichester: Wiley 1979)
Jorm, A. F., Share, D. L., Maclean, R. and Matthews, R.,
Cognitive factory at school entry predictive of specific
reading retardation and general reading backwardness,
Journal of Child Psychology and Psychiatry, 27 (1986),
pp. 4554
Kagan, J., Information processing in the child: the
significance of analytic and reflective attitudes,
Psychological Monographs, 78 (1964), p. 578
Kagan, J., The nature of the Child (New York: Basic Books
Inc. 1984)
Kagan, J., Kearsley, R. B. and Zelazo, P. R., Infancy and its
place in human development (Cambridge, Mass.:
Harvard UP 1978)
Kagan, J. and Moss, H. A., Birth to maturity (New York:
Wiley 1962)
Kail, R., The development of memory in children (San
Francisco: W. H. Freeman 1979)
Kail, R. and Bisanz, J., Information processing and
cognitive development, in Advances in Child
Development and Behaviour, vol. 17 (New York:
Kail, R. and Hagen, J. W. (eds.) Perspectives on the
development on memory and cognition (Hillsdale, NJ:
LEA 1977)
Kamler, B. and Kilarr, G., Looking at what children can
do, in B. Kroll and G. Wells (eds.), Explorations in the
development of writing (Chichester: Wiley 1983)
Karmiloff-Smith, A., A functional approach to child
language: a study of determiners and reference
Karmiloff-Smith, A., Getting developmental differences
or studying child development?, Cognition, 10 (1981),
pp. 1518
223

Karmiloff-Smith, A., and Inhelder B., If you want to get
ahead, get a theory, Cognition, 3 (1974/5), pp. 195212
Kaye, K., The mental and social life of babies (London:
Methuen 1984)
Kaye, K. and Furstenburg, F. F. (eds.), Family
development and the child, special section of Child
Development, 56 (2) (1985), pp. 279501
Keasey, C. B. (ed.), Social cognitive development
Nebraska Symposium on Motivation, vol. 25 (Lincoln,
Nebraska: Univ. of Nebraska Press 1977)
Keil, F., Childrens thinking: what never develops?,
Cognition, 10 (1981), pp. 15966
Keil, F. C., Mechanisms in cognitive development and the
structure of knowledge, in R. J. Sternberg (ed.),
Mechanisms of cognitive development (New York:
Freeman 1984)
Kelley, H. H., Personal relationships: their structures and
processes (Hillsdale, NJ: LEA 1979)
Kelso, J. and Stewart, M. A., Factors which predict the
persistence of aggressive conduct disorder, Journal of
Child Psychology and Psychiatry, 27 (1986), pp. 7786
Kempe, R. S. and Kempe C. H., Child Abuse (London:
Fontana 1978)
King, R., All things bright and beautiful? A sociological
study of infants classrooms (London: Wiley 1978)
Kinsbourne, M., Brain-based limitations on mind, in R.
W. Rieber (ed.), Mind and body (New York: Academic
Press 1980)
Kinsbourne, M. and Hiscock, M., The normal and deviant
development of functional lateralisation of the brain, in
M. M. Haith and J. J. Campos (eds.), Infancy and
Developmental Psychobiology vol. 2 of Handbook of
Child Psychology, series ed. P. H. Mussen (New York:
Wiley 1983)
Kitwood, T., Disclosures to a stranger: adolescent values
in advanced industrial society (London: Routledge &
Kegan Paul 1980)
Klahr, D. and Wallace, J. G., Cognitive development: an
information-processing view (New York: Erlbaum
1976)
Klaus, M. and Kennell, J., Maternalinfant bonding (St
Louis: C. V. Mosby 1976)
Klausmeier, H. J. and Sipple, T. S., Factor structure of the
Piagetian stage of concrete operations, Contemporary
Educational Psychology, 7 (1982), pp. 16180
Kobasigawa, A., Retrieval factors in the development of
memory, in R. V. Kail and J. W. Hagen (eds.),
224
Perspectives on the development of memory and

cognition (Hillsdale, NJ: Erlbaum 1977)
Kogan, N., Stylistic variation in childhood and
adolescence: creativity, metaphor and cognitive styles,
in J. H. Flavell and E. M. Markman (eds.), vol. 3 of
Handbook of Child Psychology, series ed. P. H. Mussen
Kohlberg, L., Development of moral character and moral
ideology, in L. W. Hoffman (ed.), Review of Child
Development Research, 1 (New York: Russell Sage
Foundation 1964), pp. 383431
Kohlberg, L., Stages of moral development as a basis for
moral education, in B. Munsey (ed.), Moral
development, moral education and Kohlberg
(Birmingham, Alabama: Religious Education Press
1971)
Kohlberg, L., From is to ought: how to commit the
naturalistic fallacy and get away with it in the study of
moral development, in T. Mischel (ed.), Cognitive
development and epistemology (New York: Academic
Press 1971)
Kohlberg, L., Essays on moral development, vol. 2, The
Psychology of moral development. The nature and
validity of moral stages (San Francisco: Harper & Row
1984)
Kohlberg, L., and Candee, D., The relationship of moral
judgement to moral action, in L. Kohlberg (ed.), Essays
on moral development, vol. 2, The Psychology of moral
development. The nature and validity of moral stages
(San Francisco: Harper & Row 1984)
Kohlberg, L., Levine, C., Hewer, A., Moral stages: a
current formulation and a response to critics (Basel:
Karger 1983)
Kohn, M., Social competence, symptoms and
underachievement in childhood: a longitudinal
perspective (New York: Wiley 1977)
Koluchova, J., Severe deprivation in twins: a case study,
in Clarke, A. M. and Clarke, A. D. B. (eds.), Early
experience: myth and evidence (London: Open Books
1976)
Konner, M., The tangled wing: biological constraints on
the human spirit
(London: Heinemann 1982;
Harmondsworth: Penguin 1984)
Kopp, C. B., Risk factors in development, in M. M. Haith
and J. J. Campos (eds.), vol. 2 of Handbook of Child
1983)
Bibliography
Kroll, B. and Wells, G., Explorations in the development of
writing (Chichester: Wiley 1983)
Kuhn, T. S., The structure of scientific revolutions
(Chicago: Chicago University Press 1962)
Kurdek, L. A., Perspective taking as the cognitive basis of
childrens moral development: a review of the
literature, Merrill-Palmer Quarterly, 24 (1978), pp. 3
28
Kurtines, W. and Grief, E. B., The development of moral
thought: review and evaluation of Kohlbergs work,
Psychological Bulletin, 81 (1974), pp. 45370
Laboratory of Comparative Human Cognition, Culture
and Intelligence, in R. J. Sternberg (ed.), Handbook of
Human Intelligence (Cambridge: Cambridge UP 1982)
Labov, W., 1969, The logic of non-standard English, in
Keddie, N. (ed.), Tinker, Tailor . . . the myth of cultural
deprivation (Harmondsworth: Penguin 1973)
Labov, W., Sociolinguistic Patterns (Philadelphia: Univ. of
Pennsylvania Press 1972)
Lamb, M. E. and Baumrind, D., Socialisation and
personality development in the preschool years, in M. E.
Lamb (ed.), Social and personality development (New
York: Holt, Rinehart & Winston 1978)
Lamb, M. E. and Roopnarine, J. L., Peer influences on sexrole development in preschoolers, Child Development,
50 (1979), pp. 121922
Lamb, M. E. and Sherrod, L. R. (eds.), Infant social
cognition: empirical and theoretical considerations
Lamb, M. E. and Sutton-Smith, B., Sibling relationships:
their nature and significance across the lifespan
Lane, H., The wild boy of Aveyron (London: George Allen
& Unwin 1976)
Lash, J. P., Helen and Teacher (London: Allen Lane 1980)
Laslett, P., The world we have lost (London: Souvenir Press
1971)
Laszlo; J. I. and Bairstow, P. J., Perceptualmotor
behaviour: developmental assessment and therapy
(London: Holt, Rinehart & Winston 1985)
Lawton, J. T. and Coleman, M., Parents perceptions of
parenting, Infant Mental Health Journal, 4 (1983), pp.
35261
Leach, P., Who cares? A new deal for mothers and their
small children (Harmondsworth: Penguin 1979)
Lesh, R. and Landau, M. (eds.), Acquisition of mathematics

concepts and processes (New York: Academic Press
1983)
Levinson, D. J., The seasons of a mans life (New York:
Knopf 1978)
Lewis, M. and Brooks-Gunn, J., Social cognition and the
acquisition of self (New York: Plenum 1979)
Lewis, M. and Michalson, L., Childrens emotions and
moods: developmental theory and measurement (New
York: Plenum 1983)
Lewis, M. and Rosenblum, L. A. (eds.), The child and its
family (New York: Plenum 1979)
Liberman, I. Y., Shankweiler, D., Liberman, A. and Fowler,
C., Phonetic segmentation and reading in the beginning
reader, in A. S. Reber and D. L. Scarborough (eds.),
Toward a psychology of reading (New York: Wiley
1977)
Lickona, T., Research on Piagets theory of moral
development, in T. Lickona (ed.), Moral development
and behaviour: theory research and social issues (New
York: Holt, Rinehart & Winston 1976)
Lieven, E. V. M., Conversations between mothers and
young children: individual differences and their possible
implication for the study of language learning, in N.
Waterson and C. Snow (eds.), The development of
communications (Chichester: Wiley 1978)
Lieven, E. V. M., Context, process and progress in young
childrens speech, in M. Beveridge (ed.), Children
thinking through language (London: Edward Arnold
1982)
Light, P. H., The development of social sensitivity
Light, P. H., Social interaction and cognitive development:
a review of post-Piagetian research, in S. Meadows
(ed.), Developing Thinking (London: Methuen 1983)
Liss, M. B. (ed.), Social and cognitive skills: sex roles and
childrens play (New York: Academic Press 1983)
Livesley, W. J. and Bromley, D. B., Person perception in
childhood and adolescence (London: Wiley 1973)
Lobban, G., The influence of the school on sex-role
stereotyping, in J. Chetwynd and O. Hartnett (eds.), The
sex role system: psychological & sociological
perspectives (London: Routledge & Kegan Paul 1978)
Lock, A. (ed.), Action, gesture and symbol: the emergence
of language (London: Academic Press 1978)
Loftus, E. F., Eyewitness testimony (Cambridge, Mass.:
Harvard UP 1979)
225

London, P., The rescuers: motivational hypotheses about
Christians who saved Jews from the Nazis, in J.
Macauley and L. Berkowitz (eds.), Altruism and helping
behaviour (New York: Academic Press 1970)
Lord, A. B., The singer of tales (Cambridge, Mass.:
Harvard UP 1960)
Luria, A. R., The mind of a mnemonist (London: Cape
1969)
Luria, A. R., Cognitive development: its cultural and social
foundations (Cambridge, Mass.: Harvard UP 1976)
Lurie, A., Only children (Harmondsworth: Penguin 1980)
Maccoby, E. E., Social development: psychological growth
and the parentchild relationghip (New York: Harcourt
Brace Jovanovich 1980)
Maccoby, E. E., Socialisation and developmental change,
Maccoby, E. E. and Jacklin, C., The psychology of sex
differences (Stanford, California: Stanford UP 1978)
Maccoby, E. E. and Jacklin, C., Sex differences in
aggression: a rejoinder and reprise, Child Development,
41 (1980), pp. 96480
Maccoby, E. E. and Martin, J. A., Socialisation in the
context of the family: parentchild interaction, in E. M.
Hetherington (ed.), vol. 4 of the Handbook of Child
1983)
Macfarlane, J. A., Olfaction in the development of social
preferences in the human neonate, CIBA Foundation
Symposium on Parent infant interaction, 33 (1976), pp.
10313 (Amsterdam: Elsevier)
McGlaughlin, A., Empson, J., Morrissey, M. and Sever, J.,
Early child development and the home environment:
consistencies at and between four pre-school stages,
International Journal of Behavioural Development, 3
(1980), pp. 299309
McGlaughlin, A. and Empson, J. M., Sisters and their
children: implications for a cycle of deprivation, in N.
Madge (ed.), Families at risk (London: Heinemann
1983)
McGurk, H. (ed.), Ecological factors in human
development (Amsterdam: North-Holland Publishing
Co. 1977)
McGurk, H., Issues in childhood social development
McLaughlin, B., Early bilingualism: methodological and
theoretical issues, in M. Paradis and Y. Lebrun (eds.),
226
Early bilingualism and child development (Lisse: Swets

& Zeitlinger BV 1984)
McRobbie, A. and Nava, M., Gender and generation
(London: Macmillan 1984)
McShane, J., The development of naming, Linguistics, 17
(1979), pp. 879905
McShane, J., Learning to talk (Cambridge: Cambridge UP
1980)
McTear, M. Childrens conversation (Oxford: Blackwell
1985)
Madge, N. (ed.), Families at risk (London: Heinemann
1983)
Main, M. and Weston, D. R., Avoidance of the attachment
figure in infancy: descriptions and interpretations, in C.
M. Parkes and J. Stevenson-Hinde (eds.), The place of
attachment in human behaviour (London: Tavistock
1982)
Maliphant, R., Juvenile delinquency, in J. C. Coleman
(ed.), The School Years (London: Methuen 1979)
Mandler, J. M. and Johnson, N. S., Remembrance of things
parsed: story structure and recall, Cognitive Psychology,
9 (1977), pp. 11151
Manning, M., Heron, J. and Marshall, T., Styles of hostility
and social interactions at nursery, at school and at home.
An extended study of children, in L. A. Hersov, M.
Berger, and D. Schaffer (eds.), Aggression and antisocial behaviour in childhood and adolescence
(Oxford: Pergamon 1978)
Maratsos, M., Some current issues in the study of the
acquisition of grammar, in J. H. Flavell and E. M.
Psychology series ed. P. H. Mussen (New York: Wiley
1983)
Markman, E., The facilitation of part-whole comparisons
by use of the collective noun family, Child
development, 44 (1973), pp. 83740
Markman, E. M., Realising that you dont understand:
elementary
school
childrens
awareness
of
inconsistencies, Child Development, 50 (1979) pp.
64355
Markman, E. M., Comprehension monitoring, in W. P.
Dickson (ed.), Childrens oral communication skills
Markman, E., Cox, B. and Machida, S., The standard
sorting task as a measure of conceptual organisation,
Developmental Psychology 17, (1981), pp. 11517
Bibliography
Martin, B., A sociology of contemporary cultural change
Martlew, M., The psychology of written language:
developmental
and
educational
perspectives
Mayer, R., Mathematical ability, in R. J. Sternberg (ed.),
Human abilities: an information-processing approach
(New York: W. H. Freeman 1985)
Mead, G. H., Mind, self and society (Chicago: Univ. of
Chicago Press 1934)
Meadows, S. A. C., The development of concrete
operations: a short-term longitudinal study, (PhD
thesis, University of London 1975)
Meadows, S., An experimental investigation of Piagets
analysis of class inclusion, British Journal of
Psychology, 68 (2) (1977), pp. 22937
Meadows, S. (ed.), Developing thinking: approaches to
childrens cognitive development (London: Methuen
1983)
Meadows, S. and Cashdan, A., Teaching styles in nursery
education: final report to SSRC (1983)
Meadows, S., Philps, J., Weaver, J. and Mably, S., Adults
and childrens views on education, and their behaviour at
home and in nursery school: a report on two pilot
studies, Paper given at the British Psychological
Society Conference, London (December 1977)
Medrich, E. A., Roizen, J. A., Rubin, V. and Buckley, S.,
The serious business of growing up: a study of childrens
lives outside school (Berkeley, California: Univ. of
California Press 1982)
Mehler, J., Bertoncini, J., Barrier, J. and JassikGerschenfeld, D., Infant recognition of mothers
voice, Perception, 7 (1978) pp. 4917
Meltzoff, A. and Boston, R. W., Intermodal matching by
human neonates, Nature, 282 (1979), pp. 4034
Meltzoff, A. and Moore, M. K., Imitation of facial and
manual gestures by human neonates, Science, 198
(1977), pp. 758
Midgley, M. Beast and man: the roots of human nature
(Brighton: Harvester Press 1979)
Millar, S., The psychology of play (Harmondsworth:
Penguin 1968)
Miller, P., Theories of developmental psychology (San
Francisco: Freeman 1983)
Miller, S. A., On the generalizability of conservation: A
comparison of different kinds of transformation, British
Journal of Psychology, 73 (2) (1982), pp. 22130
Mills, M. and Funnell, E. Experience and cognitive

processing, in S. Meadows (ed.), Developing Thinking
Mischel, T. (ed.), Cognitive development and epistemology
(New York & London: Academic Press 1971)
Money, J. and Ehrhardt, A., Man and woman, boy and girl
(Baltimore, Maryland: Johns Hopkins UP 1972)
Morton, J., Interaction of information in word
recognition, Psychological Review, 76 (1969), pp. 165
78
Morton, J., The logogen model and orthographic
structure, in U. Frith (ed.), Cognitive processes in
spelling (London: Academic Press 1980)
Munsey, B. (ed.), Moral development, moral education and
Kohlberg (Birmingham, Alabama: Religious Education
Press 1980)
Mussen, P. and Eisenberg-Berg, N., Roots of caring,
sharing and helping: the development of prosocial
behaviour in children (San Francisco: W. H. Freeman
1977)
Myers, N. A. and Perlmutter, M., Memory in the years
from two to five, in P. A. Ornstein (ed.), Memory
development in Children (Hillsdale, NJ: LEA 1978)
Nelson, K., Structure and strategy in learning to talk,
Monographs of the Society for Research in Child
Development, 38 (1973), serial 149
Nelson, K., Social cognition in a script framework, in J.
H. Flavell and L. Ross (eds.), Social cognitive
development: frontiers and possible futures
Nelson, K. and Gruendel, J., At morning its lunchtime: a
scriptal view of childrens dialogues, Discourse
processes, 2 (1979) pp. 7394
Nelson, K. E., Memory development in children: evidence
from non-verbal tasks, Psychonomic Science, 25
(1971), pp. 3468
Nelson, K. E. and Kosslyn, S. M., Recognition of
previously labelled or unlabelled pictures by 5 year olds
and adults, Journal of Experimental Child Psychology
21 (1976), pp. 405
Nesselroade, J. R. and Baltes, P. B., Longitudinal research
in the study of behaviour and development (London:
Newcombe, N. E., Rogoff, B. and Kagan, J.,
Developmental changes in recognition memory for
pictures of objects and scenes, Developmental
Psychology, 13 (1977), pp. 33741
227

Newson, J., Intentional behaviour in the young infant, in
D. Schaffer and J. Dunn (eds.), The first year of life
Newson, J. and Newson, E., Four years old in an urban
community (London: George Allen & Unwin 1968)
Newson, J. and Newson, E., Seven years old in the home
environment (London: George Allen & Unwin 1976)
Newson, J. and Newson, E., with Barnes, P. Perspectives on
school at seven years old (London: George Allen &
Unwin 1977)
Nicholls, J. G., Causal attributes and other achievementrelated cognition, Journal of Personality & Social
Psychology, 31 (1975), pp. 37989
Nicholls, J. G., The development of the concepts of effort
and ability, perception of academic attainment, and the
understanding that difficult tasks require more ability,
Child Development 49 (1978), pp. 80014
Ninio, A. and Bruner, J. S., The achievement and
antecedents of labelling, Journal of Child language, 5
(1978), pp. 115
Nottebohm, F., Ontogeny of bird song, Science 167
(1970), pp. 9506
Nowicki, S. and Walker, C., The role of generalised and
specific expectancies in determining academic
achievement, Journal of Social Psychology, 94 (1974),
pp. 27580
Nucci, L. P. and Turiel, E., Social interactions and the
development of social concepts in pre-school children,
Oakley, A., Women confined: towards a sociology of
childbirth (Oxford: Martin Robertson 1980)
Ochs, E. and Schieffelin, B. (eds.), Developmental
Pragmatics (New York: Academic Press 1979)
Ochs, E. and Schieffelin, B., Acquiring conversational
competence (London: Routledge & Kegan Paul 1983)
OFaolain, J. and Martines, L., Not in Gods Image
Olson, D. R., From utterance to text: the bias of language
in speech and writing, Harvard Educational Review, 47
(1977), pp. 25781
Olson, D. R. (ed.), The social foundations of language and
thought: essays in honor of Jerome S. Bruner (New York
& London: W. W. Norton & Co. 1980)
Olson, D. R., Some social aspects of meaning in oral and
written language, in D. R. Olson (ed.), The social
foundations of language and thought: essays in honor of
228
Jerome S. Bruner (New York: W. W. Norton & Co.

1980a)
Olson, D. R. and Torrance, N. G., Literacy and cognitive
development, in S. Meadows (ed.), Developing
Thinking (London: Methuen 1983)
Olweus, D., Stability and aggressive reaction patterns in
males: a review, Psychological Bulletin, 86 (1979), pp.
85275
Olweus, D., Stability in aggression and withdrawn,
inhibited behaviour patterns, in R. M. Kaplan, V. J.
Konecri and R. W. Novaco (eds.), Aggression in children
and youth (The Hague: Martinus Nijhoff 1984)
Opie, I. and Opie, P., The lore and language of
schoolchildren (ondon: Oxford UP 1967)
Opie, I. and Opie, P., Childrens games in street and
playground (Oxford: Clarendon 1969)
Opie, I. and Opie, P., The Singing Game (Oxford: Oxford
UP 1985)
Osborn, A. F., Butler, N. R. and Morris, A. C., The social life
of Britains five year olds. A report of the Child Health
and Education Study (London: Routledge & Kegan Paul
1984)
Paley, G., Boys and girls: superheroes in the doll corner
(Chicago: Univ. of Chicago Press 1984)
Papousek, H., Experimental studies of appetitional
behaviour in human newborns and infants, in H. W.
Stevenson, E. H. Hess and H. L. Rheingold (eds.), Early
behaviour (New York: Wiley 1967)
Paradis, M. and Lebrun, Y. (eds.), Early bilingualism and
child development (Lisse: Swets & Zeitlinger BV 1984)
Parke, R. D. and Slaby, R. G., The development of
aggression, in E. M. Hetherington (ed.), vol. 4 of the
Parkes, C. M. and Stevenson-Hinde, J. (eds.), The place of
attachment in human behaviour (London: Tavistock
1982)
Parmelee, A. H. and Sigman, M. D., Perinatal brain
development and behaviour, in M. M. Haith and J. J.
Development, series ed. P. H. Mussen (New York: Wiley
1983)
Parten, M. B., Social participation among preschool
children, Journal of Abnormal Psychology, 27 (1932),
pp. 24369
Bibliography
Pascual-Leone J., A mathematical model for the transition
role in Piagets developmental stages, Acta
Psychologica, 32 (1970), pp. 30145
Patrick, J., A Glasgow gang observed (London: Methuen
1973)
Patterson, C, J., Cosgrove, J. M. and OBrien, R. G., Non
verbal
indicants
of
comprehension
and
noncomprehension, Developmental Psychology, 16 (1)
(1980), pp. 3848
Patterson, G. R., Families: applications of social learning
to family life (Champaign: Research Press 1975)
Patterson, G. R. and Cobb, J. A., A dyadic analysis of
aggressive behaviours, in J. P. Hill (ed.), Minnesota
symposium on child psychology, vol. 5 (Minneapolis:
Univ. of Minnesota Press 1971)
Patterson, G. R., Littman, R. A. and Bricker, W., Assertive
behaviour in children: a step toward a theory of
aggression, Monographs of the Society for Research in
Child Development, 32 (serial no. 113) (1967)
Perera, K., Childrens writing and reading: analysing
classroom language (Oxford: Blackwell 1984)
Perlmutter, M. (ed.), Parentchild interaction and parent
child relations in child development, Minnesota
Symposium on Child Psychology, vol. 17 (Hillsdale, NJ:
Erlbaum 1984)
Perlmutter, M. (ed.), Cognitive perspectives on childrens
social and behavioural development, Minnesota
Symposium on Child Psychology, 1984, vol. 18
(Hillsdale, NJ: Lawrence Erlbaum Associates1984)
Perlmutter, M. (ed.), Intellectual development,
Minnesota Symposium on Child Psychology, 1985, vol.
19 (Hillsdale, NJ: Lawrence Erlbaum Associates 1985)
Peters, A. M., The units of language acquisition
Peters, R. S., Moral development: a plea for pluralism, in
R. S. Peters, Psychology and ethical development
(London: Allen & Unwin 1974)
Piaget, J., The childs conception of the world (London:
Piaget, J., The childs conception of physical causality
Piaget, J., The childs conception of number (New York:
Basic Books 1952)
Piaget, J., The construction of reality in the child (New
York: Basic Books 1954)
Piaget, J., The language and thought of the child (London:
Piaget, J., Play, dreams and imitation in childhood (New

York: Norton 1962)
Piaget, J., Six Psychological Studies (London: Univ. of
London Press 1968)
Piaget, J., The childs conception of time (London:
Piaget, J., Genetic Epistemology (New York, Columbia U.
Press 1970)
Piaget, J., The theory of stages on cognitive development,
in D. R. Green, M. P. Ford and G. B. Flanner (eds.),
Measurement and Piaget (New York: McGraw-Hill
1971a)
Piaget, J., Biology and Knowledge (Edinburgh: Edinburgh
Univ. Press 1971b)
Piaget, J., Structuralism (London: Routledge & Kegan Paul
1971c)
Piaget, J., The development of thought: equilibration of
cognitive structures (Oxford: Blackwell 1978)
Piaget, J. and Inhelder, B., The Psychology of the Child
Pollard, A., The social world of the primary school
(London: Holt, Rinehart & Winston 1985)
Pollock, L. A., Forgotten children (Cambridge: Cambridge
UP 1983)
Pool, D. L., Shweder, R. A. and Much, N. C., Culture as a
cognitive system: differential rule understandings in
children and other savages, in E. T. Higgins, D. N. Ruble
and W. W. Hartup (eds.), Social cognition and social
development: a sociocultural perspective (Cambridge:
Cambridge UP 1983)
Putallaz, M. and Gottman, J. M., Social skills and group
acceptance, in S. R. Asher and J. M. Gottman (eds.), The
development of childrens friendships (Cambridge:
Cambridge UP 1981)
Pylyshyn, Z. W., When is attribution of beliefs justified?,
Behavioural & Brain Sciences, 1 (1978), pp. 5923
Rachlin, H., Behaviour and learning (San Francisco: W. H.
Freeman 1976)
Radcliffe Richards J., The sceptical feminist
(Harmondsworth: Penguin 1982)
Radke-Yarrow, M., Zahn-Waxler, C. and Chapman, M.,
Childrens prosocial dispositions and behaviour, in E.
M. Hetherington (ed.), Socialisation, personality and
social development, vol. 4 of Handbook of Child
1983)
229

Raven, J., Parents, teachers and children: a study of an
educational home visiting scheme (London: Hodder &
Stoughton for the Scottish Council for Research in
Education 1980)
Ravenette, A. T., Grid techniques for children, Journal of
Child Psychology & Psychiatry, 16 (1975), pp. 7983
Resnick, L., A developmental theory of number
understanding, in H. P. Ginsburg (ed.), The
Development of mathematical thinking (New York:
Rest, J. R., Morality, in J. H. Flavell and E. M. Markman
(eds.), vol. 3 of Handbook of Child Psychology, series
ed. P. H. Mussen (New York: Wiley 1983)
Reynolds, P., Play, language and human evolution, in J.
Bruner, A. Jolly and K. Sylva (eds.), Play: its role in
development and evolution (Harmondsworth: Penguin
1976)
Reynolds, R. E. and Ortony, A., Some issues in the
measurement of childrens comprehension of
metaphorical language, Child Development, 51 (1980),
pp. 111019
Richards, M. P. M., Effects on development of medical
interventions and the separation of newborns from their
parents, in D. Schaffer and J. Dunn (eds.), The first year
of Life (New York: Wiley 1979)
Richards, M. P. M. and Light, P. H. (eds.), Children of social
worlds: development in a social context (Cambridge:
Polity Press, in press)
Richman, N., Stevenson, J. and Graham, P., Pre-school to
school: a behavioural study (London: Academic Press
1982)
Robinson, E. J., Metacognitive development, in S.
Meadows (ed.), Developing Thinking (London:
Methuen 1983)
Robinson, E. J. and Robinson, W. P., Development in the
understanding of the causes of success and failure in
verbal communication,Cognition, 5(1977), pp. 36378
Robinson, E. J. and Robinson, W. P., Development of
understanding about communication: message
inadequacy and its role in causing communication
failure, Genetic Psychology Monographs, 98 (1978),
pp. 23379
Robinson, E. J. and Robinson, W. P., Egocentrism in verbal
referential communication, in M. V. Cox (ed.), Are
Young Children Egocentric? (London: Batsford 1980)
Robinson, E. J. and Robinson, W. P., Ways of reacting to
communication failure in relation to the development of
230
childrens understanding about verbal communication,

European Journal of Social Psychology, 11 (1981), pp.
189208
Robinson, E. J. and Robinson, W. P., The advancement of
childrens verbal referential communication skills: the
role of metacognitive guidance, International Journal
of Behavioural Development, 5 (1982), pp. 32955
Robinson, E. J. and Robinson, W. P., Communication and
metacommunication: quality of childrens instructions
in relation to judgements about the adequacy of
instructions and the locus of responsibility for
communication failure, Journal of Experimental Child
Psychology, 36 (1983), pp. 30520
Robinson, E. J. and Robinson, W. P., Teaching children
about verbal referential communication, International
Journal of Behavioural Development (in press)
Robinson, W. P., Social psychology in classrooms, in G.
M. Stephenson and J. H. Davis (eds.), Progress in
Applied Social Psychology, vol. 2 (Chichester: Wiley
1984)
Rogers, C., The childs perception of other people, in H.
McGurk (ed.), Issues in Childhood Social Development
Rogers, C., A social psychology of schooling (London:
Rollins, B. C. and Thomas, D. L., Parental support, power
and control techniques in the socialisation of children,
in W. R. Burr, R. Hill, F. I. Nye and I. L. Reiss (eds.),
Contemporary theories about the family, vol. 1 (New
York: Free Press 1979)
Romaine, S., The language of children and adolescents: the
acquisition of communicative competence (Oxford:
Basil Blackwell 1984)
Rosa, A., Ochaita, E., Moreno, E., Fernandez, E. Carretero,
M. and Pozo, J. I., Cognitive development in blind
children: a challenge to Piagetian theory, Quarterly
newsletter of the Laboratory of Comparative Human
Cognition, 6 (1984), pp. 7581
Rosaldo, M. Z. and Lamphere, L. (eds.), Woman, Culture
and Society (Stanford: Stanford Univ. Press 1974)
Rosch, E. and Lloyd, B. B. (eds.), Cognition and
Categorisation (Hillsdale NJ: LEA 1978)
Rose, P., Parallel lives: five Victorian marriages
Rose, S. A. and Blank, M., The potency of context in
childrens cognition: an illustration through
Bibliography
conservation, Child Development, 45 (1974), pp. 499
502
Rosen, B., Moral dilemmas and their treatment, in B.
Munsey (ed.), Moral development, moral education and
Kohlberg (Birmingham, Alabama: Religious Education
Press 1980)
Rosenblatt, D., Developmental trends in infant play, in B.
Tizard and D. Harvey (eds.), The biology of play
(London: Heinemann Medical Books 1977)
Ross, L., The intuitive scientist formulation and its
developmental implications, in J. H. Flavell and L. Ross
Cambridge UP 1981)
Rotman, B., Jean Piaget: Psychologist of the Real
(Hassocks, Sussex: Harvester Press 1977)
Rotter, J., Chance, J. and Phares, E., Applications of a social
learning theory of personality (New York: Holt,
Rinehart & Winston 1972)
Rubin, J. Z., Provenzano, F. J. and Luria, Z., The eye of the
beholder: parents views on sex of newborns, American
Journal of Orthopsychiatry, 44 (1974), pp. 51219
Rubin, K. H., Fein, G. G. and Vandenberg, B., Play, in E.
M. Hetherington (ed.), vol. 4 of the Handbook of Child
1983 )
Rubin, K. H., Watson, K. S. and Jambor, T. W., Freeplay
behaviours in preschool and kindergarten children,
Rubin, Z., Childrens friendships (London: Fontana 1980)
Rumelhart, D., Toward an interactive model of reading, in
S. Dornic (ed.), Attention and performance, VI
Russell, J., The acquisition of knowledge (London:
Macmillan 1978)
Russell, J., Dyadic interaction in a logical reasoning
problem requiring inclusion ability,
Child
Development, 52 (1981a), pp. 13225
Russell, J., Why socio-cognitive conflict may be
impossible: the status of egocentric errors in the dyadic
performance of a spatial task, Educational Psychology,
1 (1981b), pp. 15969
Russell, J., Propositional attitudes, in M. Beveridge (ed.),
Children thinking through language (London: Edward
Arnold 1982)
Russell, J., Explaining mental life: some philosophical
issues in psychology (London: Macmillan 1984)
Rutter, M., Early sources of security and competence, in

J. S. Bruner and A. Garton (eds.), Human Growth and
Development (Oxford: Oxford UP 1978)
Rutter, M., Changing youth in a changing society
(Cambridge, Mass.: Harvard UP 1980a)
Rutter, M. (ed.), Scientific foundations of developmental
psychiatry (London: Heinemann Medical 1980b)
Rutter, M., Maternal deprivation reassessed (2nd ed.)
Rutter, M., School effects on pupil progress: research
findings and policy implications, Child Development,
54 (1983a), pp. 129
Rutter, M., Low level lead exposure: sources, effects and
implications, in M. Rutter and R. R. Jones (eds.), Lead
versus health (Chichester: Wiley 1983b)
Rutter, M., Statistical and personal interactions: facets and
perspectives, in D. Magnusson and V. L. Allen (eds.),
Human Development: an interactional perspective
(London: Academic Press 1983c)
Rutter, M., Family and school influences on behavioural
development, Journal of Child Psychology and
Psychiatry, 26 (3) (1985a), pp. 34968
Rutter, M., Family and school influences on cognitive
development, Journal of Child Psychology and
Psychiatry, 26 (5) (1985b), pp. 683704
Rutter, M., Psychopathology and development: links
between childhood and adult life, in M. Rutter and L.
Hersov (eds.), Child and adolescent psychiatry: modern
approaches (Oxford: Blackwell Scientific Publications
1985c)
Rutter, M. and Garmezy, N., Developmental
psychopathology, in E. M. Hetherington (ed.), vol. 4 of
the Handbook of Child Psychology, series ed. P. H.
Rutter, M. and Giller, H., Juvenile delinquency
Rutter, M. and Hersov, L. (eds.), Child and adolescent
psychiatry: modern approaches (Oxford: Blackwell
Scientific Publications 1985)
Rutter, M. and Jones, R. R. (eds.), Lead versus health:
sources and effects of low level lead exposure
Rutter, M. and Madge, N., Cycles of disadvantage
Rutter, M., Quinton, D. and Liddle, C., Parenting in two
generations: looking backwards and looking forwards,
231

in N. Madge (ed.), Families at risk (London: Heinemann
1983)
Rutter, M., Tizard, J., Yule, W., Graham, P. and Whitmore,
K., Research Report: Isle of Wight studies 19641974,
Psychological Medicine, 6 (1976), pp. 31332
Rutter, M. and Yule, W., Reading difficulties, in M. Rutter
and L. Hersov (eds.), Child psychiatry: modern
Saxton, M., Louisa May: a modern biography of Louisa
May Alcott (London: Andre Deutsch 1978)
Sayers, J., Biological politics: feminist and antifeminist
perspectives (London: Tavistock 1982)
Scardamalia, M., Information processing capacity and
the problem of horizontal dcalage: a demonstration
using combinatorial reasoning tasks, Child
Scardamalia, M., How children cope with the cognitive
demands of writing, in C. H. Frederiksen, M. F.
Whitemand and J. F. Dominic (eds.), Writing: the nature,
development and teaching of written communication
Scardamalia, M. and Bereiter, C., The development of
evaluative, diagnostic and remedial capabilities in
childrens composing, in M. Martlew (ed.), The
psychology of written language (Chichester: Johla
Wiley & Sons 1983)
Scarr, S. and Grajek, S., Similarities and differences
among siblings, in M. Lamb and B. Sutton-Smith (eds.),
Sibling relationships (Hillsdale, NJ: Erlbaum 1982)
Scarr, S. and Kidd, K. K., Developmental behaviour
genetics, in M. M. Haith and J. J. Campos (eds.), vol. 2
of the Handbook of Child Psychology, series ed. P. H.
Scarr, S. and McCartney, K., How people make their own
environments: a theory of genotype environment
effects, Child Development, 54 (1983), pp. 42435
Scarr, S. and Weinberg, R. A., The Minnesota Adoption
studies: genetic differences and malleability, Child
Schachter, F. F., Sibling deidentification and split-parent
identification: a family trend, in M. Lamb and B.
Sutton-Smith (eds.), Sibling relationships (Hillsdale,
NJ: Erlbaum 1982)
Schaffer, H. R., The childs entry into a social world
(London: Academic Press 1984)
Schaffer, H. R. and Emerson, P. E., The development of
social attachments in infancy, Monographs of the
232
Society for Research in Child Development, 29 (3) serial

no. 94 (1964)
Schank, R. C. and Abelson, R., Scripts, plans, goals and
understanding (Hillsdale, NJ: LEA 1977)
Schieffelin, B. B. and Cochran-Smith, M., Learning to
read culturally: literacy before schooling, in H.
Goetman, A. Oberg and F. Smith (eds.), Awakening to
literacy (London: Heinemann 1984)
Schwartzman, H., Transformations: the anthropology of
childrens play (New York: Plenum 1978)
Scribner, S. and Cole, M., The psychology of literacy
(Cambridge, Mass.: Harvard UP 1981)
Searle, J., Speech acts (Cambridge: Cambridge UP 1969)
Searle, J., A classification of illocutionary acts, Language
& Society, 5 (1975), pp. 123
Searle, J., Minds, brains and science: the 1984 Reith
Lectures (London: BBC Publications 1984)
Seitz, V., Rosenbaum, L. K. and Apfel, N. H., Effects of
family support intervention: a ten year follow-up, Child
Development, 56 (1985), pp. 37691
Selman, R. L., The child as a friendship philosopher, in S.
R. Asher and J. M. Gottman (eds.), The development of
childrens friendships (Cambridge: Cambridge UP
1981)
Selman, R. L. and Jaquette, D., Stability and oscillation in
inter-personal awareness: a clinical-developmental
analysis, in C. B. Keasey (ed.), Nebraska Symposium on
Motivation, vol. 25 (Lincoln, Nebraska: Univ. of
Nebraska Press 1977)
Serbin, L. A., Connor, J. M. and Citron, C. C.,
Environmental control of independent and dependent
behaviours in preschool girls and boys: a model for early
independence training, Sex Roles, 4 (1978), pp. 86775
Serbin, L. A., OLeary, K. D., Kent, R. N. and Tonick, I. J.,
A comparison of teacher response to the preacademic
and problem behaviour of boys and girls, Child
Development, 44 (1973), pp. 796804
Serbin, L. A., Tonick, I. J. and Sternglanz, S. H., Shaping
co-operative cross-sex play, Child Development, 48
(1977), pp. 9249
Shaffer, D., Brain Damage, in M. Rutter and L. Hersov
(eds.), Child & adolescent psychiatry: modern
Shaffer, D. R. and Brody G. H., Parental and peer
influences on moral development, in R. W. Henderson
(ed.), Parentchild interaction (New York: Academic
Press 1981)
Bibliography
Shantz, C. U., Social cognition, in J. Flavell and E.
1983)
Shatz, M., The relationship between cognitive processes
and the development of communication skills,
Nebraska Symposium on Motivation, 1977 (Lincoln,
Nebraska: Univ. of Nebraska Press 1978)
Shatz, M. and Gelman, R., The development of
communication skills, Monographs of the Society for
Research in Child Development, 38 (5), serial no. 152
(1973)
Shatz, M. and Gelman, R., Beyond syntax: the influence of
conversational constraints on speech modifications, in
C. E. Snow and C. A. Ferguson (eds.), Talking to children
Shepar, R. N., Recognition memory for words, sentences
and pictures, Journal of Verbal Learning & Verbal
Behaviour, 6 (1967), pp. 15663
Shields, M. M. and Duveen, G., The young childs image of
the personal and social world: some aspects of the
childs representation of persons, Paper presented at the
International Sociological Association Conference
Mexico City (August 1982)
Shultz, T. R., Play as arousal modulation, in B. SuttonSmith (ed.), Play and learning (New York: Gardner
Press 1979)
Shweder, R. A., Turiel, E. and Much, N. C., The moral
intuitions of the child, in J.H. Favell and L. Ross (eds.),
Social cognitive development (Cambridge: Cambridge
UP 1981)
Siegel, L. S., Development of the concept of seriation,
Siegel, L. S., The relationship of language & thought in the
preoperational child: a reconsideration of nonverbal
alternatives to Piagetian tasks, in L. S. Siegel and C. J.
Brainerd (eds.), Alternatives to Piaget (New York:
Siegel, L. S., McCabe, A. E., Brand, J. and Matthews, J.,
Evidence for the understanding of class inclusion in
preschool children: linguistic factors and training
effects, Child Development, 49 (1978), pp. 68893
Siegler, R. S., Developmental sequences within and
between concepts, Monographs of the Society for
Research in Child Development, 46 (1981), pp. 174
Siegler, R. S., Information processing approaches to
development, in W. Kessen (ed.), vol. 1 of the

Siegler, R. S., Mechanisms of cognitive growth: variation
and selection, in R. J. Sternberg (ed.), Mechanisms of
cognitive development (New York: W. H. Freeman 1984)
Siegler, R. S. and Robinson, M., The development of
numerical understandings, in H. W. Reese and L. P.
Lipsitt (eds.), Advances in Child Development and
Behaviour, vol. 16 (New York: Academic Press 1982)
Sinclair, A., Jarvella, R. J. and Levelt, W. J. M., The childs
conception of language (Berlin: Springer-Verlag 1978)
Sinclair, J. M. and Coulthard, R. M., Towards an analysis of
discourse: the English used by teachers and pupils
(Oxford: Oxford UP 1975)
Skinner, B. F., Walden Two (New York: Macmillan 1948)
Skuse, D., Extreme deprivation in early childhood I
Diverse outcomes for three siblings from an
extraordinary family II. Theoretical issues and a
comparative review, Journal of Child Psychology &
Psychiatry, 25 (1984), pp. 52342, 54372
Slobin, D. I., A case study of early language awareness, in
A. Sinclair, R. J. Jarvella and W. J. M. Levelt (eds.), The
childs conception of language (Berlin: Springer-Verlag
1978)
Sluckin, A., Growing up in the playground (London:
Sluckin, W., Herbert, M. and Sluckin, A., Maternal
bonding (Oxford: Blackwell 1983)
Smedslund, J., Analyzing the Primary Code: from
Empiricism to Apriorism, in D. R. Olson (ed.), The
social foundations of language and thought (New York:
W. W. Norton 1980)
Smith, F., Reading (Cambridge: Cambridge UP 1978)
Smith, F., Writing and the writer (London: Heinemann
1982)
Smith, N. V., The acquisition of phonology: a case study
Smith, P. K. (ed.), Play in animals and humans (Oxford:
Blackwell 1984)
Smith, P. K. and Connolly, K. J., The behavioural ecology
of the preschool (Cambridge: Cambridge UP 1981)
Smith, P. K. and Green, M., Aggressive behaviour in
English nurseries and playgroups. Sex differences and
response of adults, Child Development, 46 (1975), pp.
21114
233

Smith, P. K. and Simon, T., Object play, problemsolving
and creativity in children, in P. K. Smith (ed.), Play in
animals and humans (Oxford: Blackwell 1984)
Smith, P. K., Simon, T. and Emberton, R., Play, problem
solving and experimenter effects, British Journal of
Sroufe, L. A., Individual patterns of adaptation from
infancy to preschool, in M. Perlmutter (ed.), Minnesota
Symposium on Child Psychology, vol. 16 (Hillsdale NJ:
Erlbaum 1983)
Sroufe, L. A. and Waters, E., Attachment as an
organisational construct, Child Development, 48
(1977), pp. 118499
Standing, L., Learning 10,000 Pictures, Quarterly
Journal of Experimental Psychology, 25 (1973), pp.
20722
Steedman, C., The tidy house (London: Virago 1982)
Stent, G. S. (ed.), Morality as a biological phenomenon
(Berlin: Dahlem Conferenzen 1978)
Sternberg, R. J. (ed.), Handbook of human intelligence
Sternberg, R. J. (ed.), Mechanisms of cognitive
development (New York: W. H. Freeman 1984)
Sternberg, R. J., Mechanisms of cognitive development: a
componential approach, in R. J. Sternberg (ed.),
Mechanisms of cognitive development (New York:
Freeman 1984)
Stillwell, R. and Dunn, J., Continuities in sibling
relationships: patterns of aggression and friendliness,
Journal of Child Psychology & Psychiatry, 26 (4)
(1985), pp. 62738
Stipek, D. J. and Weisz, J. R., Perceived personal control
and academic achievement, Review of Educational
Research, 51 (1981), pp. 10137
Stockard, J. and Johnson, M. H., Sex roles: sex inequality
and sex role development (Englewood Cliffs: Prentice
Hall 1980)
Streeter, L., Language perception of 2 month-old infants
shows effects of both innate mechanisms and
experience, Nature, 259 (1976), pp. 3941
Stubbs, M., Language and literacy: the sociolinguistics of
reading and writing (London: Routledge & Kegan Paul
1980)
Stubbs, M., Language, schools and classrooms (London:
Methuen 1983)
Sulloway, F. J., Freud, biologist of the mind (London:
Fontana 1980)
234
Super, C. M., Clement, J., Vuori, L., Christiansen, N.,

Mora, J. O. and Herrera, M. G., Infant & caretaker
behaviour as mediators of nutritional and social
intervention in the barrios of Bogota, in T. M. Field, A.
M. Sostek, P. Vietze and P. H. Leiderman (eds.), Culture
and early interactions (Hillsdale, NJ: Erlbaum 1981)
Sutton, A., An introduction to Soviet developmental
psychology, in S. Meadows (ed.), Developing Thinking
Sutton-Smith, B. (ed.), Play and learning (New York:
Gardner Press 1979)
Sutton-Smith, B. and Kelly-Byrne, D., The idealisation of
play, in P. K. Smith (ed.), Play in animals and humans
Sylva, K., Play and learning, in B. Tizard and D. Harvey
(eds.), The biology of play (London: Heinemann 1977)
Sylva, K., Roy, C. and Painter, M., Childwatching in
playgroup and nursery school (London: Grant McIntyre
1980)
Tamburrini, J., Willig, J. and Butler, C., Childrens
conceptions of writing, in H. Cowie (ed.), The
development of childrens imaginative writing (London:
Croom Helm 1984)
Tanner, J. M., Foetus into Man: physical growth from
conception to maturity (London: Open Books 1978)
Taylor, E., The development of attention, in M. Rutter
(ed.), Scientific foundations of Developmental
Psychiatry (London: Heinemann 1980)
Taylor, E., Syndromes of overactivity and attention
deficit, in M. Rutter and L. Hersov (eds.), Child &
adolescent psychiatry: modern approaches (Oxford:
Blackwell 1985)
Taylor, M. C. and Hall, J., Psychological androgyny:
theories, methods & conclusions, Psychological
Bulletin, 92 (1982), pp. 34766
Teale, W. H., Reading to young children: its significance
for literacy development, in H. Goelman, A. Oberg and
F. Smith (eds.), Awakening to literacy (London:
Heinemann 1984)
Thomas, A. and Chess, S., Temperament and development
(New York: Brunner/Mazel 1977)
Thomas, K., Man and the natural world (Harmondsworth:
Penguin 1984)
Thomassen, A. J. W. M. and Teulings, H-L. H. M., The
development of handwriting, in M. Martlew (ed.), The
psychology of written language (Chichester: Wiley
1983)
Bibliography
Thwaite, A., Edmund Gosse (London: Secker & Warburg
1984)
Tinbergen, N., Study of instinct (Oxford: Clarendon Press
1951)
Tizard, B., Adoption: a second chance (London: Open
Books 1977)
Tizard, B. and Hughes, M., Young children learning
Tizard, B., Hughes, M., Pinkerton, G. and Carmichael, H.,
Adults cognitive demands at home and at nursery
school, Journal of Child Psychology and Psychiatry, 23
(2) (1982), pp. 10817
Tizard, B., Mortimore, J. and Burchell, B., Involving
parents in nursery and infant schools (London: Grant
McIntyre 1981)
Tizard, B., Philps, J. and Plewis, I., Play in preschool
centres, I and II, Journal of Child Psychology and
Psychiatry, 17 (1976), pp. 25174
Tizard, J., Schofield, W., Hewison, J., Collaboration
between teachers and parents in assisting childrens
reading, British Journal of Educational Psychology, 52
(1981), pp. 15
Tolstoy, L. (translated R. Edmonds), Anna Karenina
(Harmondsworth: Penguin Classics 1954)
Tough, J., Focus on meaning: a study of childrens use of
language (London: George Allen & Unwin 1973)
Tough, J., The development of meaning (London: George
Allen & Unwin 1977)
Trabasso, T., Representation, memory and reasons: how
do we make transitive inferences?, in A. D. Pick (ed.),
Minnesota symposium on child psychology, vol. 9
(Minneapolis: University of Minnesota 1975)
Trabasso, T. and Nicholas, D. W., Memory and inferences
in the comprehension of narratives, in F. Wilkening, J.
Becker and T. Trabasso (eds.), Information integration
by children (Hillsdale, NJ: Erlbaum 1980)
Trehub, S. E. and Rabinovitch, M. S., Auditory-linguistic
sensitivity in early infancy, Developmental Psychology,
6 (1972), pp. 747
Trevarthen, C., Communication and cooperation in early
infancy, in M. Bullowa (ed.), Before speech
Trevarthen, C., The foundations of intersubjectivity:
development of interpersonal and co-operative
understanding in infants, in D. R. Olson (ed.), The social
foundations of language and thought (New York: W. W.
Norton & Co. 1980)
Trevarthen, C., The primary motives for co-operative

understanding, in G. Butterworth and P. Light (eds.),
Social cognition: studies of the development of
understanding (Brighton: Harvester Press 1982)
Trivers, R., Social evolution (Menlo Park, California: The
Benjamin/Cummings Publishing Company 1985)
Trowell, J., Effects of obstetric management on the
motherchild relationship, in C. M. Parkes and J.
Stevenson-Hinde (eds.), The place of attachment in
human behaviour (London: Tavistock 1982)
Tucker, N., The child and the book (Cambridge: Cambridge
UP 1981)
Tulving, E., Episodic and semantic memory, in E.
Tulving and W. Donaldson (eds.), Organisation of
memory (New York: Academic Press 1972)
Turiel, E., Stage transition in moral development, in R.
Travers (ed.), Second handbook of research in teaching
(Chicago: Rand McNally 1973), pp. 73258
Turiel, E., Distinct conceptual and developmental
domains: social convention and morality, in C. B.
Keasey (ed.), Nebraska Symposium on Motivation 1977
(1978)
Turnbull, C., The mountain people (London: Pan Books
1974)
Vaillant, G. E., Adaptation to life (Waltham, Mass.: Little,
Brown 1977)
Valentine, E. R., Conceptual issues in psychology (London:
George Allen & Unwin 1982)
Vandell, D. L. and Mueller, E. C., Peer play and friendships
during the first two years, in H. C. Foot, A. J. Chapman
and J. R. Smith (eds.), Friendship and social relations in
children (Chichester: Wiley 1980)
Vuyk, R., Overview and critique of Piagets genetic
Epistemology 19651980, vols. 1 and 2 (London:
Vygotsky, L. S., Thought and language (Harvard: MIT
Press 1962)
Vygotsky, L. S., Mind in Society, M. Cole, V. John-Steiner,
S. Scribner and E. Souberman (eds.) (Cambridge, Mass.:
Harvard UP 1978)
Vygotsky, L. S., The prehistory of written language, in M.
Cole, V. John-Steiner, S. Scribner and E. Souberman
(eds.), Mind in Society: the development of higher
psychological processes (Cambridge, Mass.: Harvard
UP 1978a)
Wadsworth, M., Roots of delinquency (Oxford: Martin
Robertson 1979)
235

Wadsworth, M. and Wingfield, J., Preschool experience,
parenting styles and later verbal attainment scores in a
national longitudinal study, Journal of Early Childhood
Research (in press)
Wagner, D. A., Ontogeny in the study of culture and
cognition, in D. A. Wagner and H. W. Stevenson (eds.),
Cultural perspectives on child development (San
Francisco: W. H. Freeman 1982)
Walzer, M., Just and unjust wars (Harmondsworth:
Penguin 1984)
Ward, S., Wackman, D. B. and Wartella, E., How children
learn to buy: the development of consumer informationprocessing skills (Beverley Hills: Sage Publications
1977)
Wason, P. C. and Johnson-Laird, P. N., The Psychology of
Reasoning (London: Batsford 1972)
Watson, P., Twins: an investigation into the strange
coincidences in the lives of separated twins (London:
Hutchinson 1981)
Weikart, D., An economic analysis of the Ypsilanti Perry
Preschool Project (Ypsilanti: High/ Scope Educational
Research Foundation 1978)
Weiner, B. (ed.), Achievement and attribution theory
(Morristown, NJ: General Learning Press 1974)
Weiner, B., A theory of motivation for some classroom
experiences, Journal of Educational Psychology, 71
(1979), pp. 325
Weiner, B. and Kukla, A., An attributional analysis of
achievement behaviour, Journal of Personality and
Social Psychology, 15 (1970), pp. 120
Weinreich-Haste, H. (in press), Engagement and
commitment: the role of affect in moral reasoning and
moral responsibility, to be published (in German
initially) in W. Edelstein and G. Nunner (eds.) (Title to be
announced) (Suhrkamp Verlag 1986)
Wells, B. W. P., Personality and heredity (London:
Longmans 1980)
Wells, C. G., Describing childrens linguistic development
at home and at school, British Educational Research
Journal, 5 (1979), pp. 7598
Wells, C. G., Some antecedents of early educational
attainment, British Journal of Sociology of Education,
2 (1981a), pp. 181200
Wells, C. G., Pre-school literacy related activities and
success in school, in D. R. Olson (ed.), The nature and
consequences of literacy (Cambridge: Cambridge UP
1981b)
236
Wells, C. G. Language, learning and education (Bristol:

Centre for the study of language and communication
1982a)
Wells, C. G., Influences of the home on language
development, in A. Davies (ed.), Language and
learning in school and home (London: SSRC/
Heinemann 1982b)
Wells, C. G. Language development in the pre-school years
West, D. J., Delinquency: its roots, careers and prospects
West, D. J., Delinquency, in M. Rutter and L. Hersov
(eds.), Child and adolescent psychiatry: modern
Wetherford, M. J., and Cohen, L. B., Developmental
changes in infant visual preferences for novelty and
familiarity, Child Development, 44 (1973), pp. 41624
Whiting, B. B. and Whiting, J. W. M., Children of six
cultures (Cambridge, Mass.: Harvard UP 1975)
Whittaker, S., Memory development in 3 to 6 year olds:
learning how to learn (Unpublished PhD thesis,
University of St Andrews 1983)
Wilkinson, A. et al., Assessing language development
(Oxford: OUP 1980)
Willes, M. J., Children into pupils: a study of language in
early schooling (London: Routledge & Kegan Paul
1983)
Wilson, E. O., Sociobiology (Cambridge, Mass.: Belknap/
Harvard 1975)
Wilson, H. and Herbert, G. W., Parents and children in the
inner city (London: Routledge & Kegan Paul 1978)
Wilson, J., Philosophical difficulties and moral
development, in B. Munsey (ed.), Moral development,
moral education and Kohlberg (Birmingham, Alabama:
Religious Education Press 1980)
Wilson, R. S., The Louisville Twin Study: developmental
synchronies in behaviour, Child Development, 54
(1983) pp. 298316
Wimmer, H. Childrens understanding of stories:
assimilation by a general schema for actions of
coordination of temporal relations?, in F. Wilkening, J.
Becker and T. Trabasso (eds.), Information integration
by children (Hillsdale, NJ: Erlbaum 1980)
Winnicott, D. W., Collected Papers (London: Tavistock
1958)
Witkin, H. A., Cognitive styles in personal and cultural
adaptation (Worcester, Mass.: Clark Univ. Press 1977)
Bibliography
Witkin, H. A. and Goodenough, D. R., Cognitive Styles:
essence and origins
(New York: International
Universities Press 1981)
Witkin, H. A., Goodenough, D. R. and Oltman, P. K.,
Psychological differentiation: current status, Journal
of Personality & Social Psychology, 37 (1979), pp.
112745
Wolff, P. H., The natural history of crying and other
vocalisations in early infancy, in B. M. Foss (ed.),
Determinants of infant behaviour, vol. IV (London:
Tavistock/Methuen 1966)
Wollheim, R. Freud (London: Fontana 1971)
Wood, D. J., Teaching the young child: some relationships
between social interaction, language and thought, in D.
R. Olson (ed.), The social foundations of language and
thought (New York: Norton 1980)
Wood, D. J., Working with underfives (London: Grant
McIntyre 1981)
Wood, D. J., Bruner, J. S. and Ross, G., The role of tutoring
in problem solving, Journal of Child Psychology and
Psychiatry, 17 (2) (1976), pp. 89100
Yarrow, L. J. and Klein, R. P., Environmental discontinuity
associated with transition from foster to adoptive
homes, International Journal of Behavioural
Yarrow, M. R. and Waxler, C. Z., Dimensions and
correlates of prosocial behaviour in young children,
Yawkey, T. D. and Pellegrini, A. D., Childs play:

developmental and applied (Hillsdale, NJ: Erlbaum
1984)
Young, G. and Lewis, M., Effects of familiarity and
maternal attention on infant peer relations, MerrillPalmer Quarterly, 25 (1979), pp. 10519
Yule, W. and Rutter, M., Reading and other learning
difficulties, in M. Rutter and L. Hersov (eds.), Child and
adolescent psychiatry: modern approaches (Oxford:
Blackwell 1985)
Yussen, S. R. and Levy, V. M., Developmental changes in
predicting ones own span of shortterm memory,
Journal of Experimental Child Psychology, 19 (1975),
pp. 5028
Zahn-Waxler, C., Radke-Yarrow, M. and King, R., Child
rearing and childrens prosocial initiations towards
victims of distress, Child Development, 50 (1979), pp.
31930
Zajonc, R. B. and Marcus, G. B., Birth order and
intellectual development, Psychological Review, 82
(1975), pp. 7488
Zingg, R. M., Feral man and extreme cases of isolation,
American Journal of Psychology, 530 (1940), pp. 487
517
Zipes, J., Fairytales and the art of subversion (London:
Heinemann 1983)
237
Author index
Abelson, R. 111, 118

Adelson, J. 84, 202
Ainsworth, M. D. S. 144, 177, 178
Andres, D. 196
Applebee, A. 84
Archer, J. 162, 195
Aris, P. 13
Asher, S. R. 29, 191
Aslin, R. N. 51, 52, 53, 118, 119, 137
Atkins, E. 183
Austin, G. 44, 135
Baddeley, A. 40, 71
Bahrick, H. P. 71
Bahrick, P. O. 71
Baillargeon, P. O. 33, 34, 37, 139, 140
Bairstow, P. J. 54
Balfour, G. 99
Bandura, A. 149, 150, 163
Banks, M. S. 47, 48, 49, 50, 51, 77
Baring-Gould, C. 28
Baring-Gould, W. S. 28
Barker, P. 144
Barker, R. G. 191
Bartlett, F. C. 75
Bate, W. J. 179
Bates, E. 126, 135
Baumrind, D. 184, 185
Beilin, H. 37
Bereiter, C. 94, 95, 96
Berger, M. 144
Berlyne, D. E. 27
Berndt, T. J. 203
Bernstein, B. 127, 128, 129, 130, 131, 135
Bertenthal, B. J. 147
Berthoud-Papandropoulou, J. 138
Best, R. 29, 192, 197, 198, 199, 200
Bettelheim, B. 83
Beveridge, M. 193, 199
Biggs, J. B. 85, 87
Bisanz, J. 42, 43, 44
Bissex, G. L. 97
Blakemore, C. 58
238
Block, J. H. 146, 196

Bloom, L. 123
Blos, P. 201
Boden, M. A. 36, 37
Bone, M. R. 186
Borton, R. W. 62
Boulton, M. G. 176
Bower, T. G. R. 52, 62
Bowerman, M. 123
Bowlby, J. 177, 180, 181
Bradley, L. 78, 79, 88, 131, 137, 138
Braine, M. D. S. 38, 39
Brainerd, C. J. 20, 33, 98, 99, 141
Bransford, J. D. 41, 69, 76
Bremner, J. G. 64
Brierley, C. 193, 199
Brody, G. H. 183
Bromley, D. B. 110, 193
Bronfenbrenner, U. 20, 113, 179, 186, 203
Bronson, W. C. 152
Brooks-Gunn, J. 146, 147
Broughton, J. M. 62
Brown, A. L. 41, 44, 69, 71, 75, 85, 93, 94, 104, 105, 106,
107
Brown, G. 33, 37
Brown, G. W. 203
Brown, M. 185
Bruner, J. S. 44, 60, 61, 76, 107, 109, 121, 140, 175, 180
Bryant, B. 186
Bryant, P. E. 37, 38, 44, 62, 78, 79, 88, 99, 109, 131, 137,
138
Buhrmester, D. 189
Bullowa, M. 176
Burchell, B. 199
Burghardt, G. M. 26
Burnett, J. 13
Buss, A. H. 144
Butterworth, G. 49, 63, 64, 108, 177
Byers, J. A. 26
Calkins, L. 97
Callanan, M. A. 140
Camaras, L. A. 153
Author index
Campione, J. J. 47, 196
Candee, D. 171
Carey, S. 140
Carpenter, T. P. 101
Case, R. 40, 41, 43, 104
Cashdan, A. 25, 29, 30, 66, 70, 154, 186, 192, 194
Cavanaugh, J. C. 106
Chall, J. S. 78, 81, 131
Chapman, A. J. 191, 193
Chapman, M. 157
Chess, S. 144
Chi, M. T. H. 43, 75, 139
Chomsky, N. 123, 124
Cicirelli, V. G. 189
Clark, E. V. 64, 75, 119, 121, 123, 137, 139, 140
Clark, H. H. 119, 121, 123, 137
Clark, M. M. 78, 81
Clark, R. 136
Clark, R. W. 161
Clarke, A. D. B. 173, 181
Clarke, A. M. 173, 181
Clarke-Stewart, A. 186
Clay, M. 81, 91
Cobb, J. A. 191
Cochran-Smith, M. 81
Coffield, F. 185
Cohen, G. 42
Cohen, L. B. 72
Cole, M. 76, 141
Coleman, J. 146, 201, 202, 203
Coleman, M. 103
Collerson, J. 93
Collins, A. 83, 84
Condon, W. S. 118
Connolly, K. J. 30
Conroy, J. S. 101, 103
Cook, M. 110, 151
Corran, G. 102
Corrigan, R. 118
Coulthard, R. M. 132
Coveney, P. 96
Covitz, S. 126
Cox, B. 139
Cox, M. V. 37, 91, 95
Croll, P. 199
Crook, C. K. 53
Cross, T. 125
Crowder, R. G. 77, 78
Csiksentmihalyi, M. 24, 27
Curtiss, S. 173
Dale, P. S. 123
Dalton, K. 182
Davie, C. E. 98, 126, 131, 186
Davies, B. 153, 192, 199, 200
Dawkins, R. 155, 173
De Casper, A. J. 118
De Loache, J. S. 75, 106
De Mause, L. 13, 15
De Villiers, J. G. 119, 137
De Villiers, P. A. 119, 137
Delamont, S. 196, 199
Delia, J. 136
Dempster, F. N. 81
Desforges, C. 33, 37
Dickinson, J. 114, 115
Dodge, K. A. 153, 195
Doise, W. 105, 108
Dominic, J. F. 93
Donaldson, M. 40, 99
Douvan, E. 202
Downing, J. 78
Doyle, A.-B. 181
Dunn, J. F. 117, 144, 147, 148, 150, 157, 174, 187, 188, 189,
192, 196
Duveen, G. 109
Dweck, C. S. 150, 194, 195
Edmonds, R. 185
Edwards, C. P. 191
Ehrhardt, A. A. 54
Eimas, P. D. 118
Eisenberg-Berg, N. 164, 170
Elliot, A. J. 210
Elliot, E. S. 150, 194, 195
Ellis, A. W. 77, 78, 79
Emerson, P. E. 188
Emler, N. 105, 108, 114, 115
Erikson, E. 19, 145, 146, 201, 202
Ervin-Tripp, S. 136
Essen, J. 185
Etaugh, C. 196
Eysenck, H. J. 143, 144
Fagan, J. F. 50, 64, 72
Fagen, R. 24
Fein, G. G. 28, 29, 30, 186
Ferrara, R. A. 41, 69
Feshbach, N. D. 153
Feshbach, S. 153
Field, T. M. 175
Fifer, W. P. 118
239
Author index
Fine, G. A. 192, 200
Finkelstein, N. W. 150
Fischer, K. W. 147
Fivush, R. 198
Flapan, D. 110
Flavell, J. H. 37, 44, 73, 74, 75, 106, 108
Foot, H. C. 191, 193
Ford, M. E. 95
Forgas, J. P. 108
Francis, H. 77, 78, 79, 93
Fransella, F. 43, 139
Frederiksen, C. H. 93
Freeman, N. 37, 64, 91
Freud, A. 148
Freud, S. 25, 28, 145, 148, 161, 162, 170, 181, 201
Frith, U. 77
Fry, C. L. 119
Funnell, E. 81, 84, 107, 109, 112
Furman, W. 189
Furth, H. 113
Fuson, K. C. 98, 99
Gallistel, C. R. 98
Gallup, G. G. 147
Galton, M. 70, 199
Garbarino, J. 20, 185
Garmezy, N. 181
Garvey, C. 23, 25, 28, 29, 135
Gelman, R. 33, 34, 37, 98, 106, 109, 124, 139, 140, 188
Gewirth, A. 163
Gibson, E. J. 61, 64
Gibson, J. J. 61, 62, 63, 64, 65
Giddens, A. 127
Giles, H. 127, 133
Giller, H. 154, 204, 205, 208
Gilligan, C. 169, 170
Ginsberg, H. 138
Gleitman, L. R. 138
Gold, D. 181, 196
Goldman, J. 20
Goldman, R. 20
Goodenough, D. R. 70
Goodnow, J. J. 44
Goody, J. 76, 141
Gordon, J. C. B. 128, 129
Gottlieb, G. 16, 160
Gottman, J. M. 29, 191, 192
Gould, S. J. 24, 25, 26, 143
Grajek, S. 188
Gratch, G. 63
Graves, D. 97
240
Green, M. 154
Grice, H. P. 117, 128, 135
Grief, E. B. 170
Griffin, P. 76, 141, 203
Griffiths, M. 92, 93, 96
Groen, G. J. 101, 102
Gruendel, H. 112
Hagen, J. W. 70, 73
Haith, M. M. 47
Hale, G. H. 70
Hall, J. W. 98, 99
Halliday, M. A. K. 93, 119, 122, 135
Halverson, C. F. 144, 145
Hamlyn, D. W. 60
Hammersley, M. 199
Hardyment, C. 181
Hargreaves, D. 150, 199
Harlow, H. F. 16, 54, 160, 173
Harlow, M. K. 16, 54, 160, 173
Harris, P. L. 50, 61, 62, 64, 65, 77
Harter, S. 150, 202
Hartup, W. W. 147, 189, 191, 192, 193, 202
Harvey, P. G. 57
Hawkins, P. R. 128, 129
Hayward, C. 81, 83
Heath, S. B. 81, 125, 128, 133, 135, 145, 153
Heider, E. R. 139
Heider, F. 110
Held, D. 127
Herbert, G. W. 185
Herbert, M. 54
Hersov, L. 186
Hetherington, E. M. 181
Hewison, J. 81, 199
Hicks, L. 49
Hinde, R. A. 160, 176, 178, 179, 180, 188, 191
Hiscock, M. 59, 60
Hofer, M. A. 15, 16
Hoffman, M. L. 155, 157
Hoggart, R. 203
Houlbrooke, R. 13, 14
Howe, M. J. A. 85
Hughes, M. 84, 91, 92, 98, 102, 103, 107, 125, 128, 129,
130, 132, 186, 189
Humphreys, A. P. 27
Huston, A. 54, 192, 195, 196
Hutt, C. J. 25, 65, 66
Inhelder, B. 34, 37, 104, 106
Isbell, B. J. 108
Author index
Istomina, Z. M. 76, 107
Kurtines, W. 170
Jacklin, C. 154, 192

Jackson, B. 175
Jackson, S. 186
Jahoda, G. 113, 114
Jarman, C. 91
Jencks, C. 18
Jenkins, R. 203
Johns, M. A. 54
Johnson, N. S. 83
Johnson-Laird, P. N. 38, 44
Jones, O. H. M. 182
Jones, R. R. 57
Jorm, A. F. 78
Jusczyk, P. W. 51, 52
Labov, W. 117, 122, 128, 131

Lamb, M. E. 177, 178, 185, 188, 196
Lamphere, L. 195
Landau, M. 98, 103
Lane, H. 173
Lash, J. P. 121
Laslett, P. 13
Laszlo, J. I. 54
Leach, P. 113
Lesh, R. 98, 103
Levinson, D. J. 146, 202
Levy, V. M. 74
Lewis, M. 146, 147, 152, 191, 192
Lewontin, R. C. 26
Liberman, I. Y. 78
Lickona, T. 166
Liddle, C. 186
Lieven, E. V. M. 125
Light, P. H. 108, 112, 158, 177, 189, 193
Livesley, W. J. 110, 193
Lloyd, B. 75, 139, 162, 195
Lobban, G. 196
Lock, A. 175
Loftus, E. F. 75
Logan, P. 200, 201
Lord, A. B. 76
Luria, A. R. 76, 141
Kagan, J. 18, 28, 50, 70, 72, 145, 147

Kail, R. V. 42, 43, 44, 73, 74, 75, 107
Kamler, B. 91
Karmiloff-Smith, A. 104, 106, 136
Kaye, K. 174, 175, 176, 180
Kearsley, R. B. 72
Kelley, H. H. 110
Kelly-Byrne, D. 24
Kelso, J. 154
Kempe, C. H. 185
Kempe, R. S. 185
Kendrick, C. 117, 147, 148, 157, 174, 188, 189, 192
Kennell, J. 54, 175
Kessen, W. 209
Kidd, K. K. 16, 17, 57
Kilarr, G. 91
King, R. 198, 199
Kinsbourne, M. 34, 59, 60
Kitwood, T. 203
Klahr, D. 45, 100, 104
Klaus, M. 54, 175
Klausmeier, H. J. 37
Kobasigawa, A. 75
Koeske, R. D. 43, 139
Kogan, N. 70, 138
Kohlberg, L. 157, 164, 166, 167, 168, 169, 170, 171, 195
Kohn, M. 144, 154
Koluchova, J. 173
Konner, M. 57, 146, 151, 152
Kosslyn, S. M. 72
Kroll 93
Kuhn, T. S. 105
Kukla, A. 149
Kurdek, L. A. 157
Maccoby, E. E. 148, 149, 152, 154, 158, 171, 183, 187, 192,
195, 202
Macfarlane, J. A. 53, 54
McGlaughlin, A. 185
McGurk, H. 70
Machida, S. 139
McKee, L. 108
Mackie, D. 108
Maclean, R. 78
McRobbie, A. 203
McShane, J. 77, 120, 121
McTear, M. 136
MacWhinney, B. 135
Madge, N. 18, 19, 176, 185
Main, M. 177, 178
Maliphant, R. 205
Mandler, J. M. 83
Manning, M. 155
Maratsos, M. 123, 125, 210
Marcus, G. B. 188
Markman, E. M. 84, 94, 139
241
Author index
Martin, B. 203
Martin, J. A. 158, 171, 202
Martines, L. 195
Martlew, M. 93
Matthews, R. 78
Mayer, R. 97, 100
Mead, G. H. 146, 147
Meadows, S. 25, 29, 30, 34, 37, 66, 70, 154, 186, 192, 194,
198, 199
Medrich, E. A. 200, 201
Mehler, J. 173
Meltzoff, A. 62
Meyer-Bahlburg, H. 54
Michalson, L. 152
Midgley, M. 109, 160, 161
Millar, S. 25
Miller, P. 163
Miller, S. A. 37
Mills, M. 81, 84, 107, 109, 112
Money, J. 54
Moore, M. K. 62
Mortimore, J. 199
Morton, J. 77
Moser, J. M. 101
Moskowitz, D. S. 181
Moss, H. A. 145
Much, N. C. 170
Mueller, E. C. 189, 191, 192
Mugny, G. 105, 108
Munsey, B. 169
Mussen, P. H. 164, 170
Myers, N. A. 72
Nava, M. 203
Nelson, K. 111, 112, 118, 125
Nelson, K. E. 72
Newson, E. 96, 154, 182, 191, 197, 199, 200, 201
Newson, J. 96, 154, 182, 191, 197, 199, 200, 201
Nicholas, D. W. 83
Nicholls, J. G. 83, 150
Ninio, A. 107, 120, 175
Nottebohm, F. 120
Nowicki, S. 149
Nucci, L. P. 170
Oakley, A. 175
Ochs, E. 125, 135
OFaolain, J. 195
Olson, D. R. 76, 135, 136, 138, 141
Olweus, D. 18, 150, 153, 154
Opie, I. 29, 153, 200
242
Opie, P. 29, 153, 200

Ortony, A. 139
Osborn, A. F. 133, 149, 186
Painter, M. 65, 66
Paley, G. 29
Papousek, H. 53, 72
Parke, R. D. 152, 153, 191
Parkes, C. M. 177
Parkman, J. M. 101, 102
Parmelee, A. H. 55, 56, 58, 59
Parten, M. B. 25, 192
Pascual-Leone, J. 40, 41, 43
Patrick, J. 200
Patterson, G. R. 130, 153, 191
Perlmutter, M. 72, 106
Peters, R. S. 123, 169, 170
Petrie, P. 200, 201
Philps, J. 66
Piaget, J. 15, 20, 25, 28, 29, 33, 34, 35, 36, 37, 38, 43, 44,
45, 60, 61, 62, 63, 64, 69, 83, 98, 99, 101, 104, 105, 106,
108, 109, 110, 113, 136, 141, 163, 165, 166, 167, 169,
177, 191
Pisoni, D. B. 51, 52
Plewis, I. 66
Plomin, R. 144
Pollard, A. W. 154, 199
Pollock, L. A. 13, 15
Pool, D. L. 170
Powesland, P. F. 127, 133
Putallaz, M. 192
Quinton, D. 186
Rabinovitch, M. S. 118
Radke-Yarrow, M. 155, 157, 158, 171
Ramey, C. T. 150
Raven, J. 183
Ravenette, A. T. 43
Resnick, L. 101, 103
Reyholds, R. E. 139
Richards, M. P. M. 175
Riley, C. A. 38
Robinson, E. J. 75, 84, 105, 106, 107, 129, 131
Robinson, M. 98, 99, 101
Robinson, W. P. 84, 106, 107, 129, 131
Rogers, C. 110, 111, 150, 199
Rogoff, B. 72
Rollins, B. C. 158, 183, 185
Romaine, S. 125, 127, 129, 132, 133, 136
Romberg, T. A. 101
Author index
Roopnarine, J. L. 196
Rosaldo, M. Z. 195
Rosch, E. 75, 139
Rose, P. 179
Rosen, B. 169
Rosenblatt, D. 28
Rosenblum, L. A. 191
Ross, L. 108, 110, 111
Rotman, B. 37, 106
Rotter, J. 149
Roy, C. 65, 66
Rubin, J. Z. 196
Rubin, K. H. 28, 29, 30
Rubin, Z. 191, 192
Rumain, B. 38, 39
Rumelhart, D. 77
Russell, J. 40, 45, 105, 108
Rutter, M. 18, 19, 57, 79, 144, 154, 163, 176, 178, 181, 183,
185, 186, 191, 202, 203, 204, 205, 207, 208
Salapatek, P. 47, 48, 49, 50, 51, 77
Sander, L. W. 118
Saxton, M. 81
Sayers, J. 162
Scardamalia, M. 87, 88, 94, 95, 96
Scarr, S. 16, 17, 18, 57, 188
Schachter, F. F. 182, 188
Schaffer, H. R. 188
Schank, R. C. 84, 111, 118
Schieffelin, B. 81, 125, 135
Schofield, W. 81, 199
Schwartzmann, H. 30
Scott, M. S. 71
Scribner, S. 141
Searle, J. 45, 109, 117, 135
Serbin, L. A. 192, 196
Shaffer, D. 58
Shaffer, D. R. 183
Shantz, C. V. 108, 110, 111, 193, 194
Share, D. L. 78
Shatz, M. 107, 124, 135, 188
Shepard, R. N. 71
Sherrod, L. R. 177
Shields, M. M. 109
Shultz, T. R. 27
Shweder, R. A. 170
Siegel, L. S. 99, 101
Siegler, R. S. 34, 41, 43, 98, 99, 100, 101, 104
Sigman, M. D. 55, 56, 58, 59
Simon, B. 199
Simon, T. 65
Sinclair, J. M. 121, 132

Sipple, T. S. 37
Skuse, D. 34, 173
Slaby, R. G. 152, 153, 191
Slobin, D. J. 137
Sluckin, A. 24, 54, 153, 176, 192, 199, 200
Sluckin, W. 54, 176
Smedslund, J. 169
Smith, E. E. 83, 84
Smith, F. 77, 89, 90, 93
Smith, J. R. 191, 193
Smith, N. V. 120
Smith, P. K. 24, 27, 30, 65, 154
Spelke, E. 109
Sroufe, L. A. 178
Standing, L. 171
Steedman, C. 81, 96, 97
Stent, G. S. 160
Sternberg, R. J. 37, 41, 42, 44, 104
Stevenson-Hinde, J. 177
Stewart, M. A. 154
Stipek, D. J. 149
Streeter, L. 118
Stubbs, M. 78, 87, 130
Sulloway, F. J. 161
Sutton-Smith, B. 24, 188
Sylva, K. 29, 65, 66, 186
Tamburrini, J. 93
Tanner, J. M. 55, 195
Taylor, E. 69, 70, 71
Teale, W. H. 81
Teulings, H.-L. H. M. 88
Thomas, A. 144
Thomas, D. L. 158, 183, 185
Thomas, K. 108
Thomassen, A. J. W. M. 88
Thwaite, A. 179
Tinbergen, N. 173
Tizard, B. 18, 66, 84, 92, 98, 107, 112, 125, 129, 130, 132,
144, 181, 186, 189
Tizard, J. 81, 199
Torrance, N. G. 76, 138
Tough, J. 128
Trabasso, T. 38, 83
Trehub, S. E. 118
Trevarthen, C. 174, 176
Trivers, R. 160, 173, 187
Trowell, J. 175
Tucker, N. 81
Tulkin, S. 126
243
Author index
Tulving, E. 75
Turiel, E. 164, 166, 170
Turnbull, C. 160
Vaillant, G. E. 146
Valentine, E. R. 104
Valiant, G. 105, 108
Vandell, D. L. 189, 191, 192
Vandenberg, B. 28, 29, 30
Vuyk, R. 33, 37
Vygotsky, L. S. 89, 109, 125, 139, 180
Wadsworth, M. 183, 185, 206, 207
Waldrop, M. F. 144, 145
Walker, C. 149
Walkerdine, V. 102
Wallace, J. G. 45, 100, 104
Walzer, M. 155
Ward, S. 115
Wason, P. C. 38, 44
Waters, E. 178
Watson, P. 148
Watt, I. 141
Waxler, C. Z. 155, 157, 158, 171
Weaver, J. 198, 199
Wedge, P. 185
Weikart, D. 186, 208
Weinberg, R. A. 18
Weiner, B. 149, 194
Weinreich-Haste, H. 170, 171
Weisz, J. R. 149
Wellman, H. M. 74, 106
Wells, B. W. P. 152
244
Wells, C. G. 78, 81, 91, 92, 93, 96, 118, 121, 123, 124, 125,
126, 128, 129, 130, 131, 132, 133, 138, 150, 181
West, D. J. 185, 205, 206, 207
Weston, D. R. 177, 178
Wetherford, M. J. 72
Whiting, B. B. 158, 188, 191
Whiting, J. W. M. 158, 188, 191
Whittaker, S. 74, 107
Willes, M. J. 131, 198, 199
Wilson, E. O. 159
Wilson, H. 185
Wimmer, H. 83
Wingfield, J. 183
Winnicott, D. W. 175
Witkin, H. A. 70
Wolff, P. H. 118
Wollheim, R. 162
Wood, D. J. 109, 125, 132, 133, 194
Woods, P. 199
Wright, H. F. 191
Yarrow, M. R. 155, 157, 158, 171
Young, G. 192
Yule, W. 79
Yussen, S. R. 74
Zahn-Waxler, C. 155, 157, 158, 171
Zajonc, R. B. 188
Zelazo, P. R. 72
Zingg, R. M. 173
Zipes, J. 81
Subject index
accommodation in Piagetian theory 334, 356, 37,

435, 60, 71
accommodation of lens of eye 47, 49
action, as source of cognition 33, 60
activity: as temperament dimension 1445; sex differences
in 196, 197
adaptation 1617, 267; in cognition 33
addition strategies 101
adolescence 146, 2013
adoption 18, 186
affordances 61
aggression: and attributional judgements 153, 195;
between peers 152, 153, 155, 191, 195; between siblings
18991; biological bases of 152; continuity of 18, 153
4; definition 1501; development of 1525; ecology of
1525; in Freudian theory 1612; and relations with
adults 1535, 2058; sex differences in 16, 145, 154,
195, 197
altruism 155, 157, 158, 15960; see also pro-social
behaviour
ambiguity: and language development 1389; and
metacognitive development 1056
animistic thinking 109
anxiety 144, 194, 1989
approach-withdrawal as temperament dimension 144
arithmetic 59, 97103
arousal 27, 150, 175
articulation 119
Ashford, Daisy 96
assimilation 334, 356, 37, 435, 60
attachment 1778, 186
attention 43, 6971, 72, 144; attention problems 57, 70,
205; infants selective attention to faces 50,
173
attribution theory 11011, 1935; attributional judgements
11011, 14950, 153, 157, 1935, 1989
auditory brain 513
Austen, Jane 84, 97
automatization of cognitive processes 402, 435
autonomous morality 166, 1678
awareness: of cognition 34, 434, 45, 1045; see also
metacognition; of language 789, 1301, 1314, 1379;
of other people 1556, 1923, 195; see also empathy;

altruism; of sounds 7880, 120, 127, 1379
babbling 119
basic trust 145
behaviour problems 19, 57
behavioural recombination: in cognitive development 43
5, 85, 956, 103; in play 278, 656
Belloc, H. 81
bilingualism 119
binocular vision 4962
biological bases: of aggression 1512; of altruism 159; of
attachment 1758; of cognition 34; of development 15,
18; of language 117, 118, 119; of personality 1435; of
play 25; of social relationships 173, 174, 181
biological/social interaction 1518; and aggression 1512;
and language 11719; and moral behaviour 15961; and
neoteny 257; and personality 1435; and sex
differences 16, 196
bonding 1756, 1778
brain: brain damage 54, 57, 1756; evolutionary changes
and 17, 26; physical development of 4760
caretaker, sibling as 158, 188
categorization: and language labels 1212, 13940; in
memory 73, 75
catharsis, through play 28
causality 34
chess and memory 75
child abuse 185
child nature, concepts of 1314, 15, 96
class differences in language 12631
classroom: discourse 1324; interaction 1989 cognition:
as action 335; inequality between parent and child 180;
in infancy 605; information-processing model 405;
and logic 3840; and metacognition 435, 1048; and
perception 601; Piagetian model of 338; and social
interaction 1045, 1089
cognitive development 33115; mechanisms of 347, 40
5; and moral development 1634, 166, 16970, 1701;
Piagetian theory 338, 606, 104, 108; and play 29, 65
6; as social process 35, 1045, 1089
245
Subject index
c cognitive processes 405, 1048; see also arithmetic;
attention;
memory;
metacognition;
reading;
remembering; writing
cognitive
structures
335,
37,
40
cognitive style 701, 145
comfort strategies 1556
communication: in infancy 174, 176, 180; as language
function 117, 120, 122, 12445, 1356; within families
182,
1834,
1856
comparing quantities 99100
competence, sense of: and moral action 1701; and
personality development 113, 14950; and school
achievement 1935, 1989
competence-performance models 37, 45
competition 160, 197
complementarity in relationships 158, 179, 180, 1889,
191
complexity: as cause of infant pattern preference 50; in
language development 124, 128, 130; in recognition
memory 72
comprehension: and metacognition 1046; in reading 77;
of text 836
concept of childhood 1314, 96; cultural variations in, and
adultchild language 125, 129, 133
concepts 42, 1223, 13940
conditioning: of altruistic behaviour 158; and memory in
infants 72; and moral development 163; and Social
Learning Theory 163
confirmation and cognitive development 434, 109
conflict: and adolescence 202; between interests at school
199; between siblings 189, 190; cognitive 346, 37, 43
5, 1045, 106; and development of aggression 1525;
and moral development 164, 166, 168, 169; and
personality stages 1456, 1612
conscience 148, 162, 163, 1701; see also superego
conservation 33, 634, 99100, 105
consistency: in cognition 36, 37, 435; in language 117,
129, 132, 133, 137; in moral behaviour 163, 166, 170; in
personality 1426
continuity: in development 1819; and early experience
144, 145; of parenting over generations 1835
conversation 118, 1246, 128, 1367
co-operation: in cognitive development 1089; and moral
development 158, 160; with peers 158, 1912; with
siblings 158, 1889
counting 989
crisis theory: and adolescence 146, 202; of personality
development 145, 146, 1612
cross-model integration 612
246
cultural variations: in language use 125, 12634, 1401; in

literacy 789, 81, 92; in memory 756; in play 30, 200
1; in social understanding 11415
culture and evolution 1617
daycare 186
deaf childrens early language 119
delinquency 2048; and dysfunctional families 19, 185,
2058
dependency 145, 196, 200, 202
deprivation: maternal 181, 186; sensory 18, 34;
social 16, 153, 173, 189, 191, 195, 205, 2067
depth perception 61, 623
developmental behaviour genetics 16
digit span 72, 75
disadvantage 19, 57, 1756, 181, 183, 185, 186, 2058
discipline techniques 1315, 1845, 187; and delinquency
1535, 2058
discord in families 19, 1535, 181, 1856, 191, 2058
discourse: analysis 11718, 135; childrens competence in
discourse 1357; classroom discourse 1325
discrepancy: childrens interest in 147; and cognition 347,
435, 1046
distractibility 57, 69, 1445
divorce 181
Downs syndrome and motherchild interaction 182
drawings 37, 91
early experience 17, 1456, 1807, 2067
ecosystems in child development 201; and adolescence
2023; and delinquency 2058; homeschool interface
74, 789, 12635, 198200; in language 117, 1256,
12735; and literacy 789, 113, 1401; in memory 747;
parents 174, 1817; peer groups 1912, 1967, 2001
educational achievement: and delinquency 205, 207; and
self-concept 149, 150, 194; and social class differences
12630
ego 1612
egocentrism: in conversation 1367; and moral realism
1656; and peer relations 192, 193; and social empathy
and understanding 10910, 1559; and writing 93, 95, 97
electroencephalograms 589
Eliot, George 164
empathy:inearlybehaviour1557,174,176;as
moral principle 164, 16971; with parents 1834,
empathycont.
202; with peers 1923; with siblings 1889
epigenesiss 1519; in language development 117, 120,
1256
equifinality 17
equilibration 356, 37, 423, 166
ethical theory 1634, 169
Subject index
evolution and development 1617; of attachment 175, 177;
of play 257
executive strategies in cognition 40
experience: in brain development 5860; in cognitive
development 34; in sensory development 49, 53
extraversionintroversion 1434
eye: anatomy 479; functioning 4951; movements 501,
77
faces, infants preference for 50, 1734
family environment 1807; and adolescent relationships
2023; and delinquency 19, 2057; and language
development 1246, 12732, 1334; and literacy 789,
967; and personality 144, 145, 146, 1534, 157, 158;
and sibling relationships 157, 18790; see also parent
child relations; parenting
feedback: in cognition 435; in memory 74; in
metacognition 104, 105, 106
feeling and morality 1612, 16971
feminism, moral necessity of 195
feral children 16, 173
field dependence 701
flexibility: in genetic programming 17; in play 24, 656
framing: of children by parents 1745, 180, 185, 187;
play as 27
friends 199, 200; accounts of 193; choices of 189, 1967;
skills 1913
functions of language 1357
gender: development 192, 1958; and identification 148;
and morality 170; and self-categorization 147; see also
sex differences
genes: and aggression 19, 152; and altruism 15961; and
development 15, 16, 17, 19, 54; and sex differences 195;
and sibling differences 188
grammatical rules 1224, 137, 1389
Gurin, Mme 173
habituation 64, 72
hearing, development of 513
heteronomous morality 1656
historicism 17
history of childhood 1314
home language: and reading 789; and school behaviour
1314; variation in 125,12631
hormones and behaviour 54, 151, 196
hyperactivity and attention deficits 57, 70
id 161
identification 1489; and Oedipus complex 162; and sex
differences 148, 196, 197; and sibling relationships 188;
and Social Learning Theory 163
identity crisis 146, 202

imitation 148, 189
implicitness in language 1289
impulsivityreflectiveness 701, 1445
incidental memory 70
inconsistency in cognition 357, 435
infant search behaviour 624
infancy: cognition in 606; language in 11821; perception
in 4654; personality in 1447; play in 28; social
relationships in 1738, 1801
inference: in memory 75; schemas in logic 3840; in
understanding stories 83
information processing 405, 104
instincts, in Freudian theory and moral development 1612
institution rearing 186, 208
intelligence 18; and delinquency 19, 205; and lead
pollution 57; sensory evoked potential 589
intention: and aggression 151, 1523; childrens attribution
of 153, 1934, 195; imputed to infants 174; irrelevance
in logic 3940; and moral development 1636
interaction see social interaction
intersubjectivity 174, 176; see also empathy
intrusion errors in recall 41
Itard, Jean-Marc 173
jokes 1389
Keller, Helen 121
kinaesthetic development 54, 65
knowledge: about tasks 69, 73, 74, 847, 1048; of
language 77, 128, 129, 131, 1379, 1401; and memory
737; in metacognition 1048; organization of 36, 41,
42, 435; in reading 77, 84
knowledge-telling strategy in composition 94
language: awareness 77, 789, 1379; codes 12735;
development 113, 11726; ecology of 11718, 1237,
1306; functions 913, 1357; lack in feral children 173;
and literacy 779; localization in brain 59; in logic
problems 3940; and mathematics 1013; and social
interaction 1246, 12737
lateralization of brain 59
lead pollution 57, 70
learner characteristics 69, 73; and study skills 857
learning: and genes 17; and neoteny 26; and play 656;
skills 737, 837
life events and personality theory 1456
literacy: and cognition 83, 1401; composition of text 93
7; and memory 76, 1401; and reading processes 7782;
skills of 7787, 8897; and stories 81, 83; and study
skills 836; and writing 8693
247
Subject index
locus of control 149
locutionary force 135
logic 37, 3840
maternal deprivation 181
maternal instinct 1756
mathematics 33, 97103
maturation 34, 54
meaning: and categorization 75, 1213, 13940; in early
social interaction 174, 176; in language 1201; 1213,
125, 131, 135, 138; of text 778, 84, 88
measuring 99
memory: development 707, 107; and literacy 76, 1401;
mnemonic activities 737; working memory 405
mental space 405
metacognition 736, 789, 85, 1048, 189; metalanguage
789, 133, 1378; metamemory 736, 107
metaphor 122, 139
Middlemarch 164
modelling: of aggression 154; of altruism 158; of language
124, 125, 1301, 1323; of parenting 183, 185, 186, 189;
by parents 174; of reading 789, 81; of writing 913, 96,
1067
modulation of arousal: by parents 175; through play 27
monitoring: in cognitive development 435; of
comprehension of text 835; in language 129, 1379,
1401; of writing 88
moral development 1558, 15971
moral reasoning and judgement 160, 163, 16471
moral rules 1634, 16671
motherchild relations 12, 20, 174, 1756, 1778;
and personality development 1435, 1545
motherese 1246; use by siblings 174, 188
motivation: and moral action 1701; self-efficacy 14950,
1935, 1989; and study skills 857; and writing 91, 96
7
myelination: of neurons 55; of optic pathways 48
naming 1202
Nature 17, 1601
negative instances 44
neoteny 17, 256
nerve development 556
neuropathology and lead pollution 578
neuroticism 143, 144
neurotransmitters 58
noviceexpert differences 75, 1067
number 33, 98, 99, 100
nursery rhymes 28; and playground rhymes 192, 200
object identity 64
248
object permanence 624

object play 289
Oedipus complex 162
olfactory development 54
organic growth in cognitive development 34
parentchild relations 20, 17381; at adolescence 2023;
and altruism 158; and delinquency 19, 2057; and
heteronomous morality 162, 165; and identification 148;
and sex-roles 1967
parentinfant play 28; interaction 17381
parenting: age changes in 180; and aggression 1545;
changes after divorce 181; childs parent-rearing 16,
187; effects of different techniques 1813, 1834, 184
5, 2057; as framing 1745; history of 1314; and
language development 12431; and metacognition 107;
neoteny and 267; and personality 1456, 150; and prosocial behaviour 155, 157, 158, 160
pattern: pattern making and sorting behaviour 139;
preferences 50; recognition 77, 101
peers: at adolescence 2023; and aggression 152, 155, 191,
195; and experience with siblings 18991; and
metacognitive development 105; and moral
development 1579; play with 29, 192, 199, 2001; and
self-efficacy 150, 199; and sex stereo-typing 1967; and
social skills 1913
perceptual development 18, 4754, 77; cross-modal
integration 612; hearing 513, 118; movement 49, 53
4; and perceptual deprivation 18, 173; and reading 77;
seeing 4751, 77; taste, touch, smell 534
permanent objects 62
person concepts 10911; childrens perception of
person
conceptcont.
peers 191, 1923, 196; of teachers 1989
personality: physiological theories 1434, 1512; self
concept 857, 131, 133, 14650, 199, 2058; stage
models 1456, 1601; temperament 19, 1445, 153; see
also aggression; parenting; pro-social development;
self-efficacy
phenylketonuria (PKU) 57
phonemic awareness 7880, 88, 91, 11819, 120, 127,
1379
physical deprivation, effects of 18, 187
physiological basis for personality 1434, 1512
plasticity in development 1718
play: and cognitive development 656; ecology of 245,
2930, 2001; functions of 2730, 83; and sex
differences 192, 1967; theories of 237
playground: games 29, 200; social interaction in 192, 199,
200
pragmatics 128, 133, 1356
Subject index
pretend play 29
primacy effect in memory 72, 76
primary maternal preoccupation 175
pro-social behaviour 1558, 160, 1889
psychiatric disorder 19, 146, 184, 203, 2056
reading: language knowledge and 789, 131, 133, 138;
phonemic awareness and 7880, 88; processes 768;
strategies in 7984; and writing 88
recency effects in memory 72
reciprocity: dimension of relationships 179; parentchild
180; peers 1912; siblings 18991
redundancies in cognition 434
reference 120, 1212, 128, 13940; over- and underextension in 1223; referential communication 1056,
128, 129, 131
regularities in cognition 1056
regulation of cognition 435
rehearsal 735, 76
relationships: dimensions of 1789; and temperament 144;
see also complementarity; parent; peers; reciprocity;
siblings; social relationships
remembering 717; ecology of 767; and study skills 84
5; training of 737; see also memory
repertory grid analyses of concept organization 43, 139
representation: of knowledge 42, 60; in language 1201,
1212; need for social experience for 173; social 113
15; and symbolic play 29
resources for cognition 402, 44
retardation of development 26; see also neoteny revision of
text 956
role play 29, 189, 192
rough and tumble play 29, 192, 197, 200
Rousseau, Jean-Jacques 15
rules: cognitive 43; moral 1634, 169, 170; rule-governed
play 289
scaffolding 28, 1246, 1745
scanning pictures 702
school: childrens concepts of 131, 198200; and home 74,
789, 12635; language 1315; and personality
development 14950, 194, 197
scripts: in play 29; of school 198; in social cognition 111
13; and stories 83, 84, 113; in writing 93, 967
segmentation of speech sounds 77, 789, 123, 137, 140
self-categorization: by age 1478; by gender 1467, 196
8; and school 199
self-concept: in adolescence 205, 2068; development of
65, 14650; and school 857, 131, 133, 191; and study
skills 857
self-efficacy: development of 113, 150; and moral action

1701; and school 1935, 1989
self-monitoring: in cognition 44; in language 137; in
memory 745; in metacognition 104, 106; in reading 83
5
self-recognition experiments 147
sensitivity, auditory 52
sensori-motor cognition 37, 606
sensory evoked potential 589
separation from mother 18, 175, 1778, 181
sex differences: development of 54, 1958; in language
133; methodological problems of research on 16, 195,
196; in moral development 170; in peer interactions 192,
1967; in personality 16, 145, 148, 150, 154, 162; in
sibling interaction 189
sex roles: and identification 148; sources of 196
siblings 1478, 1567, 174, 18791
Sitwell, Edith 96
Skinner, B. F. 141
socialbiological interaction in development 1517;
and aggression 1512; language 11719; and moral
development 15961, 1612; and neoteny 257; and
personality 1435; and sex differences 16, 196
social class and language differences 12631
social cognition: and aggression 153; ideas of persons and
objects 109, 110, 111, 1923; ideas of social institutions
11115, 196, 1989; and morality 170; social interaction
as a source of cognitive development 35, 1045, 1089
social interaction: and aggression 1524; and altruism
1558; and cognition 35, 1045, 1089, 111; and
language 120, 1246, 1313, 136; and moral
development 160; parentchild 1739, 17987; with
peers 191201; and play 29; and self-concept 14650;
with siblings 18791; and temperament 1445
social interaction, lack of: and aggression 15395; and feral
children 16, 173; and poor social development 189, 191,
205, 2067
Social Learning Theory 162
social relationships: and delinquency 2058; dimensions of
1789; feral children 173; and neoteny 267; parent
child 17381, 2023; peers 191201; siblings 18791
social representation 11315
social routines and language use 125, 126, 133, 136
social sensitivity: and aggression 153; and morality 160,
163, 165, 170; and parentchild relationships 180, 183
5; and pro-social behaviour 1559; and scripts 112; and
siblings 189
social skills: and aggression 1534; and delinquency 205
6; and peers 1912; and siblings 1889
249
Subject index
social worlds: adolescents 2013; in Freudian theory 161
2; in moral reasoning 170; of school 198200; sex
differences in 192, 1967
socialization: and aggression 1534; of children by parents
181, 1836; and developmental change 180, 182; and
morality 158, 161, 163, 1656; of parents by children
187; and temperament 144; via books and stories 813
sociobiology and altruism 15961
sorting behaviour 139
sound-letter correspondences 7880, 91, 103
sound localization 52, 53
speech acts 11718, 120, 128, 1323, 1357
speech sounds: infants perception of 118; infants
production of 119; and reading 778, 7880
speed, cognitive development as increase in 41
spelling 77, 78, 7880, 88, 97
stage models 1920, 345, 37, 113; of moral development:
Kohlbergs 16670; Piagets 1646; neo-Piagetian 19,
113; of personality development 145, 1601; Piagets, of
cognitive development 19, 345, 37; of reading 7980
stories: childrens understanding of 813; childrens
writing of 967; as mode of socializing children 813; in
moral reasoning research 1656, 169; and reading 75,
78, 81, 84
story reading 78, 81
stress 24, 144, 2013, 2068; see also disadvantage
study skills 80, 847
subitizing 98
subtraction 101, 103
sucking 60, 174
Sullivan, Anne 121
superego 1612
supervision: of delinquents 205, 206, 207; of girls 196, 197,
200
syllogisms 3840
250
symbol manipulation 41
symbolic play 29
symbolic representation 37, 91, 1202
taste development 534
teachers: childrens accounts of teachers activities 198;
and childrens language 1315; of delinquent children
205, 207; interests in classroom 1989; judgements of
children 1315, 198; and sex typing 1967
television 196, 201, 205
temperament 1445; and aggression 19, 153
text: comprehension of 75, 835, 1056; meaning in 946
touch development 534
transfer: and memory 74; of metacognition 107
transitions between environments 21, 201, 202, 203, 207
transitive inference 378, 41
turntaking: in conversation 1357; in social interaction
174, 176
verbal communication 1056, 128, 129, 131
vigilance 70
vision 4751
visual cortex 18, 489
vocabulary 1212, 123, 127, 131, 133, 139, 140
vulnerability 144, 201, 2068
warmth/responsivity 144, 154, 158, 1834
withdrawal, as temperament dimension 144
word: coinage 140; meaning 138, 13940; order 1224;
production 119,1203
working memory 405
writig: basic skills 8891; childrens 967; composition
878, 935; functions 913; review 88, 956, 97
youth culture 203

Understanding Child Development

Uploaded by

Copyright:

Available Formats

Understanding Child Development

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Understanding Child Development

Uploaded by

Copyright:

Available Formats

Understanding Child Development

Psychology Editor Brian Foss