Adults cognitive demands at home and at nursery

J. OiUdPsychot. Psychiat. Vol. 23, No. 2, pp. 105-116, 1982.
Primed in Great Britain.
0021-9630/82/020105-12 $03.00/0
Pergamon Press Ltd.
© 1982 Association for Child Psyehology and Psychiatry.
ADULTS' COGNITIVE DEMANDS AT HOME AND AT
NURSERY SCHOOL
BARBARATIZARD, MARTIN HUGHES, GILLPINKERTON
and HELEN CARMICHAEL
Thomas Coram Research Unit, University of London
INTRODUCTION
OUR MAIN concern in this paper is with what we have termed ' 'cognitive demands".
These are, essentially, verbal requests which adults make of young children and
which require them to use particular cognitive skills. For example, the demand
"What's that called?" typically requires the child to produce the name for an
object; the demand "What did you do at nursery today?" requires the child to
recall earlier events and provide a description of them; the demand "Why did you
go upstairs?" requires her to provide an explanation or justification for her actions.
Such cognitive demands are usually expressed in the form of "wh-questions"
(Brown, 1968; Ervin-Tripp, 1970), but they may occasionally take the form of a
request or command (e.g. " Give me the red one " ).
Our interest in cognitive demands arises from their current popularity in nursery
education. In recent years the use of cognitive demands has become a characteristic
feature of the speech of nursery teachers and playgroup workers (Wood et al., 1980),
and such demands are an integral part of the compensatory pre-school programmes
developed by Blank (1973) and Tough (1977). In these contexts, cognitive demands
are used both as means of assessing the child's skills and as a device for stimulating
her thinking. Thus, by asking "What happened when you mixed the paints?", the
teacher is hoping to assess the extent to which the child can recall, describe and
interpret perceptions, and also to foster the development of these skills. This
approach to compensatory education is based on a number of beliefs; for example,
that verbal cognitive skills are developed primarily by the parent's use of cognitive
demands, that working class parents make these demands of their children less often
than middle class parents and that this deficiency can be made good by the judicious
use of cognitive demands by the teacher.
The empirical foundation of these beliefs is decidedly shaky. There is certainly
evidence of a social class difference in children's verbal cognitive skills, both in l.Q.
tests and in a test designed to assess their ability to answer a range of cognitive
demands (Blank et al., 1978). There is no evidence that these skills are developed in
young children as a result of having cognitive demands put to them by their parents.
Requests for reprints to: Dr. B. Tizard, Thomas Coram Research Unit, University of London,
Institute of Education, 41 Brunswick Square, London WCIN lAZ, U.K.
Accepted manuscript received 25 March 1981
105
106
BARBARA TIZARD, MARTIN HUGHES, GILL PINKERTON AND HELEN CARMICHAEL
or that parents of different social classes make such demands to a different extent.
There is, of course, evidence that enrolment in special language programmes will
enhance children's verbal test scores (Tizard, 1975), but no evidence that the
widespread adoption of a questioning style by the nursery teacher in her day to day
interactions with children is an effective educational technique.
Indeed, the possible deficiencies of cognitive demands as an educational tool were
pointed out by Nathan Isaacs forty years ago, when, in opposition to Piaget, he
argued "It is a complete error to equate the situation in which we ask children
'why' questions, with that in which they ask us. . . . When such questions are put
to the child, he is not involved, often not interested. He has experienced no shock
or stimulus or puzzlement . . . they are sprung at him in the mere form of words,
when he is worlds away from the state which creates such questions in him" (1930).
Some support for Isaacs' claims comes from the fmdings of Wood et al. (1980), who
reported that ". . . if the adult maintains the dialogue largely through questions,
children's answers tend to be terse and even monosyllabic". Our own informal
observations in nursery school also suggested that teachers' cognitive dernands, far
from stimulating curiosity, often met with a minimal response or no response at all.
We suspected that this was not because the child was unable to answer the question
or express herself in words, but because she was not sufficiently interested to do so.
Demands to which the teacher already knew the answer—which we refer to as
"testing" demands—seemed particularly uninteresting to the child. In contrast, we
suspected that cognitive demands which arose more naturally out of everyday
conversations—such as those between a child and her mother—might receive a more
favourable response from the child.
A study in which we recorded the spontaneous conversations of four-year-old
girls enabled us to examine evidence bearing on these issues^. Half of the girls were
working class and half middle class, and all were recorded talking both at home with
their mothers and at nursery school with their teachers. We looked first of all at
whether cognitive demands were used differently by nursery teachers, by middle
class mothers and by working class mothers. Because we had already found that there
was more than four times as much talk per hour at home as at school (Tizard et al.,
1980), we looked separately at both \he proportion of cognitive demands in the adult's
total talk and at the overall hourly rate of such demands. The first measure provides
information about the adult's characteristic style of talking, while the second measure
tells us about the overall input to the child. We assessed the effect of the adult's
cognitive demands on the child by looking at how often such demands were
answered, at the adequacy of the child's answers and at the association between the
adult's demands and the child's own questions. Finally, we looked at the relation
between cognitive demands and the children's l.Q. For each analysis we looked
separately at the overall number of cognitive demands, at whether the demands were
"testing" or "non-testing" and at the relative complexity of the demands involved.
METHOD
Subjects
The subjects were 30 girls, aged between 3 years 9 months and 4 years 3 months (mean age 3 years
11 months, S.D. 1.8 months). Half the girls were working class: that is, their fathers were manual
workers and their mothers had left school at the minimum school-leaving age with no educational
ADULTS' COGNITIVE DEMANDS AT HOME AND AT NURSERY SCHOOL
107
qualifications. The other 15 girls were middle class: that is, their fathers were in professional or
managerial positions and their mothers either had had tertiary education or had qualified for it. The
other criteria were that the children should attend nursery school, spend their afternoons at home with
their mothers, have no more than two sibs and that the language used at home should be English.
Three of the girls were only children. A third of the mothers in both social classes worked part time,
in the morning or evening. Of the working class fathers, two-thirds were skilled manual workers,
e.g. electricians and fitters; the rest were semi-skilled or unskilled.
The working class children in our study could not in any reasonable sense be considered "deprived";
they lived in small, two-parent families, the majority in Council housing, and appeared to be well cared
for, much loved and plentifully supplied with toys. Nursery schooling in Britain, although free, is not
compulsory and the children were therefore not a random selection of four-year-olds. However, when
we asked the mothers why they had sent their children to nursery school, none of the working class
mothers gave "educational" reasons—they usually answered that the child needed company, that she
got bored at home and that the school was just down the road. At any rate, the children were probably
typical of the majority of working class children who attend half-day nursery school. They were
drawn from nine different schools—in each school, children from both social classes were selected. In
schools where more than two children from a particular social class fulfilled our criteria, the study
children were selected by tossing a coin. All the parents of the children who fulfilled our criteria
agreed to take part in the project, i.e. we were not studying a self-selected group. Two very timid
children could not be persuaded to wear the special dress required and were replaced by children from
the same schools who met the criteria.
Schools
All nine schools were run by the local authorities for children aged three and four years. The children
were in classes of 20-25 children staffed by a trained teacher and assistant, usually helped by one or
two students. Most of the morning was spent in free play with a wide variety of equipment. The staff
role was primarily to suggest, and sometimes to demonstrate, ways of using play material, and to
help the children's langauge by informal conversation. Some of the schools also had a group story or
music session for the whole class, but we excluded these sessions from our study.
Recording method
The children wore a sleeveless
transmitter had a range of about
garden. An observer followed the
conversations (c.f. Hughes et at.,
dress, fitted with a tiny microphone and radio-transmitter. The
100 yards, so that the child could move freely about house and
children fairly closely in order to record a detailed context of the
1979). Using these methods, almost all the talk was intelligible
{Tizard et at., 1980).
Recording times
Since our pilot study had suggested that the first day of observation tends to be atypical (Hughes
et al., 1979), we observed in the homes for two consecutive afternoons from 1 pm-3.30 pm, but used
only the data from the second day. In the schools, we observed for three consecutive mornings from
about 9 am-11.30 am and used only the data from the second and third days. It was necessary to
record for an extra day at school because of the smaller number of adult-child conversations there
(Tizard etal., 1980).
Coding cognitive demands
These were essentially requests for the child to use certain defined cognitive skills (see below).
Cognitive demands usually took the form of direct questions (e.g. "What colour is that?") although a
small number were expressed as imperatives (e.g. "Find me the blue one"). We excluded yes/no
questions (e.g. "Is that blue?"), tag questions (e.g. "That's a blue one isn't it?"), questions which
the child could answer any way she wished (e.g. "What colour Smartie would you like?"), questions
which were primarily expressions of affect ("Why do you keep annoying me?") and exact repetitions
of previous cognitive demands. Each demand was assigned to one of the following codes, which were
developed from the work of Blank (1973), Blank et al. (1978) and Tough (1976). It is not claimed that
these codes cover all cognitive processes: they do, however, include important thinking skills and
represent the kinds of questions included in pre-school educational programmes.
108 BARBARA TIZARD, MARTIN HUGHES, GILL PINKERTON AND HELEN GARMICHAEL
Type A: labelling.
(1) Asking the child to identify, name or point to an object, person or place present (e.g. "What's
that?").
Type B: attributes and simple descriptions.
(2) Asking the child to identify or use attribute concepts, such as colour, shape, size, texture, etc.
(e.g. "Show me a long one"); includes comparisons (e.g. "Which of these is bigger?").
(3) Asking the child to match shapes, etc. (e.g. while doing ajigsaw, "Which piece fits here?").
(4) Asking for description of ongoing activity (e.g. "What are you doing?"), or an immediately
past activity (e.g. "What have you been doing?"), or an immediately future activity (e.g.
"What are we going to do with these sleeves?", where the child is expected to answer "Push
them u p " or similar); also, description of pictures (e.g. "Tell me about your painting").
Type C: recall and narration.
(5) Asking for recall of an object or an event (e.g. "Where did you leave your coat yesterday?")
or a description of a past event, and asking "When" questions.
(6) Asking for narration (e.g. "Tell me the story of the Three Bears").
Type D: explanations and generalisations.
(7) Asking for an interpretation, as distinct from a description of a picture of story (e.g. "What does
the picture tell you?").
(8) Asking "What do you mean?" not rhetorically, i.e. waiting for a response.
(9) Asking "How do you know?" not rhetorically, i.e. waiting for a response.
(10) Asking the child for similarities and differences and for class-names (e.g. "What do we call
Smarties and jelly babies?").
(11) Asking the child to explain a specific process (e.g. "How did you make the mask?"), to
hypothesize actions or consequences of actions (e.g. "What will happen if we mix these
together?"), to explain causal relatonships (e.g. "Why doesn't that work?"), to solve problems
(e.g. "What could we do to join these together?") and to project into another's feelings or
reactions (e.g. "What do you think she's feeling now?").
(12) Asking the child to justify her actions, opinions and wishes (e.g. "Why did you go upstairs?",
'' Why do you want that ?").
(13) Asking for general knowledge (e.g. "Where do we go if we have a bad tooth?") and for
purposes of objects (e.g. "What are knives for?"), and definitions (e.g. "What is a pilot?").
Type E: 3R knowledge.
(14) Asking for 3 R knowledge (e. g. " How many apples ? ", " What does that letter say ?").
Coding "testing"demands
These were a sub-set of cognitive demands to which, from the context, the adult clearly knew the
answer. If the issue was in doubt, the code was not used.
Children's responses to cognitive demands
The following response codes were used, based on those of Blank et at. (1978).
No response.
(a) Child says nothing or changes subject.
(b) Child refuses to respond (e.g. "Tell me about your painting". "No I won't").
Inadequate response.
(c) Child says she doesn't know.
(d) Child gives focused but incorrect answer (e.g. "What colour is that?". "Blue", when the
object is red).
(e) Simple assertion (e.g. "Because it does"), or repeats question in statement form.
(f) Appeals to authority (e.g. "Because Daddy said so").
(g) Appeals to wishes or feelings without further explanation (e.g. "Because I want to").
(h) Only implicitly or obliquely answers the question (e.g. Q_. "How did she make it?" A. "She
put all hers on"),
(i) Cives associated information, but doesn't answer question (e.g. Q. "What do tigers look like?"
A. "They bite you"),
(j) Irrelevant response, with no apparent relation to the matter of the question (e.g. Q. "What
other shapes are there besides squares?" A. "All sorts of things I can do").
ADULTS' COGNITIVE DEMANDS AT HOME AND AT NURSERY SCHOOL
109
Adequate response. A response which is explicit and focused on the central issue of the question. The use
of this code did not imply that there was only one "correct" response to a demand. A wide range of
answers could be considered adequate, provided that they were explicit and focused on the question.
High standards were not expected. The following were typical of answers considered adequate.
(1) C. "I can't do the rest of it" M. "Why's that?" C. "Cause it's been hard work".
(2) Teacher. "How did you make it?" (a bag) C. "Well, I got this (points) and then put this on
(points) and then Miss tied it".
The following answers were, by contrast, considered too implicit or unfocused on the intention of the
question to be adequate.
(1) C. " I don't like this (TV programme)" M. "Why?" C. "Him and his guns, he hurt me".
(2) C. "My (toy) kettle's broken" M. "How did it break?" C. "You see, it won't stick on the wall".
(3) Teacher. "Were you born in America? Or were you born in England?" C. " I was born in my
mummy's tummy".
Reliability of coding
The agreement between four coders, analysing about fifty conversations at home and fifty at school,
on whether a turn included a cognitive demand was 91 per cent, and on the type of cognitive demand,
86 per cent. The agreement between two coders on the type of child's response was 84 per cent.
Measures of I. Q.
After the observations had been made an independent psychologist, who had no knowledge of the
children or their families, tested the children with the Stanford Binet I.Q. Eorm LM.
RESULTS
Frequency of all cognitive demands
Table 1 shows that cognitive demands were, as we expected, used proportionately
more often by nursery teachers than by mothers of either social class. However,
because there was more than four times as much adult-child talk at home than at
school (Tizard et ai, 1980) the hourly rate of cognitive demands was also calculated.
This shows that, in absolute terms, cognitive demands were in fact used more
frequently by mothers of both social classes than by teachers. The range in the
number of cognitive demands put to children was very large in both social classes.
At school, the hourly rate of cognitive demands varied between 1 and 16 and at home,
between 4 and 48. There were no significant social class differences in the hourly
rate, the range, or the proportion of cognitive demands addressed to children at
home or at school.
TABLE 1. FREQUENCY OF COGNITIVE DEMANDS
School
Cognitive demands per 100 turns
of talk
Cognitive demands per hour
"Testing" demands per 100 turns
"Testing" demands per hour
Home
we
MC
we
9.3
8.5
3.3
7.6
4.0
3.6
7.9
4.7
4.5
14.8
1.0
5.3
MC
Effect
4.5 Home vs school
Social class
Interaction
Home vs school
19.6 Social class
Interaction
Home vs school
2.3 Social class
Interaction
Home vs school
9.6 Social class
Interaction
F
P
26.6
0.1
t.l
23.2
< 0.001
NS
NS
< 0.001
1.1
1.3
NS
11.5
<0.01
1.6
0.2
5.7
2.2
1.5
NS
NS
NS
<0.05
NS
NS
110 BARBARA TIZARD, MARTIN HUGHES, GILL PINKERTON AND HELEN GARMICHAEL
Frequency of' 'testing'' cognitive demands
Table 1 shows that the distribution of "testing" demands—i.e. demands to which
the adult knew the answer—was very similar to the overall distribution of all cognitive
demands. That is, testing demands were used proportionately more often by teachers
than by mothers: however, in absolute terms, testing demands occurred more
frequently at home than at school. There were no significant social class differences
in the use of testing demands at home and at school.
Types of cognitive demand: as proportions of all cognitive demands
Table 2 shows the proportion of each type of cognitive demand in the total
demands made by mothers and teachers. There were only two significant differences.
Mothers of both social classes used proportionately more demands for recall (Type
C) than did teachers: this reflects our earlier findings that home conversations were
more often about topics outside the here-and-now than were school conversations
(Tizard et al., 1980). Secondly, working class girls both at home and at school were
more often asked for attribute names (e.g. "What colour is that?") and for simple
descriptions (Type B) than were middle class girls.
TABLE 2. PROPORTIONAL USE OF EACH TYPE OF COGNITIVE DEMAND AS A PERCENTAGE OF ALL DEMANDS
School
we
Type A Labels
34
Home
MC w e
25
25
Type B Attributes, descriptions
28
17
16
TypeC
13
23
24
20
27
24
Recall, narration
Type D Explanations, generalisations
TypeE
"3Rs"
6
9
10
MC
Effect
22 Home vs school
Class
9 Home vs school
Class
27 Home vs school
Class
19 Home vs school
Class
23 Home vs school
Class
F
0.5
0.4
7.0
14.4
P
NS
<0.1
<0.01
< 0.001
6.7
2.7
<0.02
0.7
NS
0.6
1.5
NS
NS
1.2
NS
NS
There were no significant interactions.
Although there were no significant home-school or social class differences in the
proportion of demands which asked for explanations and generalisations (Type D
demands), there were significant differences in the kind of explanation asked for.
(Home vs school, deviance 40.6, d.f. 2, P< 0.001; social class, deviance 10.2, d.f. 2,
P<. 0.005). At home, much the most frequent demand of this kind was to ask the
child to justify her motives or intentions (e.g. "Why are you doing that?"). This
type of demand made up 51% of the Type D demands made by working class
mothers and 36% of those made by middle class mothers, but only 17% of Type D
demands made by teachers. Teachers were more likely to ask the children to explain
how they made something, or (usually in the context of informal story reading) to
interpret a picture or to predict what would happen. One type of demand was used
almost exclusively by middle class mothers—asking the child to refiect about the
nature and basis of her communication, e.g. "How do you know?" and "What do
you mean?".
ADULTS' COGNITIVE DEMANDS AT HOME AND AT NURSERY SCHOOL
111
Types of cognitive demand: hourly rate
Table 3 shows the hourly rate of each of the five main types of cognitive demand.
Apart from Type B demands (attributes and simple descriptions), there were significantly more demands of all types at home than at school. There were no significant
social class differences either at home or at school. Type E demands (for " 3 R "
knowledge) were rare at school, but figured prominantly in the demands of middle
class mothers, a good proportion coming in the course of structured games (e.g.
card games, a number puzzle, a reading set).
TABLE 3, HOURLY RATE OF EACH TYPE OF COGNITIVE DEMAND
School
Home
we
MC w e
MC
2,5
2,3
4,0 Home vs school
Sociad class
Interaction
Home
vs school
1,5
Social class
Interaction
5,3 Home vs school
Social class
Interaction
4,0 Home vs school
Social class
Interaction
Type A Labels
3,5
TypeB
Attributes and simple description 2,1
1,3
2,1
TypeC
Recall
1,4
3,9
0,9
Type D Explanations and generalisations 1,6
TypcE
"3Rs"
0,5
2,3
0,6
3,3
2,0
Effect
FP
4,1
0,1
0,3
0,1
3,2
0,1
20,4
1,4
0,4
6,0
1,2
0,0
<0,05
NS
NS
NS
NS
NS
< 0,001
NS
NS
<0,05
NS
NS
4,8 Too few at school for analysis
Children's responses to cognitive demands
A relatively large proportion of cognitive demands were simply not answered by
children of both social classes: this happened significandy more often at school
than at home (37% compared with 26%: see Table 4). The proportion of "nonresponses" did not differ significantly across the various types of demand, including
testing demands, although 3R demands were answered more often than other
demands. This was almost certainly due to the large number of 3R demands occurring
during 3R games, particularly at home: these games usually captured the child's
interest and generated high response rates, "Testing" demands were answered as
often as other demands. Type D demands (e.g. for explanations and justifications)
received fewer adequate and more inadequate answers than other types of demand.
Many of these Type D demands were clearly too difficult for children of this age. No
child, for example, gave an adequate definition ofa word, or answered "How do you
know?", and only one or two children could give class names, explain the purpose
TABLE 4, TYPE OF RESPONSE (AS % OF ALL RESPONSES)
All demands combined
School
MC
we
Home
MC
we
1 No response
38
37
27
26
2 Inadequate response
22
20
26
24
3 Adequate response
40
44
47
50
Analysis of deviance
Home vs school
Social class
Interaction
= 25 ,9
= 2 ,4
= 0 ,0
/>< 0,001
NS
112 BARBARA TIZARD, MARTIN HUGHES, GILL PINKERTON AND HELEN CARMICHAEL
of objects, or answer "What do you mean?". This last question was usually answered
by a repetition—e.g. Adult. "What do you mean, liquidy milk?" Child, "Liquidy
milk"—while "How do you know?" questions were usually met by an assertion
"Cause I know".
Mean I. Q. s and correlation with I. Q. s
The mean Binet I.Q. of the working class girls was 106,4, S,D, 13,2; of the middle
class girls, 122.3, S.D, 11.3. Correlations were calculated between the child's LQ.
and the frequency and proportion of all cognitive demands, of "testing"
demands and of each type of cognitive demand, both within and across
social class groups, at home and at school. All correlations were very low and not
significant, with two exceptions. At school there was a significant positive correlation,
for working class girls only, between LQ, and the frequency of Type D demands
(the most difficult demands) (r= +0,54, P< 0,05), There was also a significant
negative correlation in the case of middle class girls at school only, between the
child's I.Q. and the frequency of Type A demands (labelling) (r = -0.57, P< 0,05).
This suggests that the teachers were to some extent "tailoring" their demands to the
children's intellectual level. Correlations between the number and proportion of
cognitive demands and of both adequate and inadequate answers given by the
children were also low and not significant.
Correlations between measures at home and at school
The correlation between all the measures at home and at school were very low
and not significant. This was true of each social class separately and for the whole
group of children.
Correlation between adults' cognitive demands and child's questions
The correlation between the hourly rates of the adults' cognitive demands and the
children's questions were very low at school, but somewhat higher at home (working
class homes, r = 0.32, middle class homes, r = 0,45, PK 0.05, one tail).
DISCUSSION
As we expected, the talk of nursery teachers contained a much higher proportion
of cognitive demands and "testing" demands than the talk of either working class
or middle class mothers. However, because there was so much more adult-child
talk at home, the children received about twice as many cognitive demands and
testing demands in an hour at home as in the same period at school. There was
therefore no question of the children requiring cognitive demands at school because
they did not get them at home—and this was just as true of the working class children
as of the middle class children.
Although all types of demands were made more often by mothers than by teachers,
mothers use proportionately more of some types of demands than teachers, and less
of others. Because talk at home arose from the demands of living together, mothers
were much more likely to question the child about motivations and purposes and to
ask her to recall the past. On the other hand, because staff-child talk at nursery
school arose primarily from the teachers' educational intentions, they were more
ADULTS' COGNITIVE DEMANDS AT HOME AND AT NURSERY SGHOOL
113
likely to ask the child for attribute names (e.g, colours, shape and size), to explain
how she had made something and to interpret a picture or story. That is, teachers
tended to orient children towards play materials; mothers, towards people and their
behaviour, past and present.
It may be objected that for this reason cognitive demands at home often serve a
different and less educational function th2in those at school. Demands for justification,
for example, which were especially frequent in working class homes ("Why are you
doing that?") may well have the force of a challenge or correction. Nevertheless,
it is important to note that such demands potentially have a cognitive as well as a
regulative function. If the child is to come up with an adequate reply (e.g. "Because
you said I could if I took my shoes off") she will have to satisfy both the cognitive
and the disciplinary aspects of the demand.
In fact, both at home and at school, less than half the cognitive demands were
adequately answered. This was partly because some of the demands were too difficult
for children of this age to answer adequately—as the difficulty increased, the
children's responses tended to be assertions ("because it is"), or very unfocused or
implicit answers. In addition, a substantial minority of demands were not answered
at all—more so at school (36%) than at home (26%). This failure to respond seemed
to indicate a lack of interest on the child's part in the adult's questions, since failure
to respond was not related to the difficulty of the question. Our expectation that
"testing" demands would be answered less often than other demands was not
confirmed. This seemed to be because some testing demands, especially those
occurring in the context of a structured game, did in fact interest the children.
Structured games, e.g. card and board games and " 3 R " games, were played more
often at home than at school. They usually maintained the child's interest and
tended to produce high rates of both demands and responses. Whilst many such
games call only for labelling, it would be possible to develop games which, whilst
still enjoyable, made more complex demands, TV programmes such as Sesame
Street have included games of this kind.
One argument often put forward to justify a questioning style is that it will
encourage children to ask questions themselves. Whilst our findings in the homes
offered some support to this belief—there were moderate correlations between the
rate of mothers' cognitive demands and children's questions—the situation in the
school was very different. There the teachers' talk contained a high proportion of
cognitive demands, but the children asked very few questions indeed (Tizard
et al., in preparation). Similar findings emerged from Wood et al.'s study (1980) of 24
playgroup workers and nursery teachers: those staff who asked the most questions
received the least elaborate answers from the children and were themselves asked
the fewest questions by the children.
These findings may perhaps be understood as a reflection of the underlying
patterns of social control in the two settings. Socio-linguists such as Mishler
1975a, b) and Corsaro (1979) have pointed out that questions are a very effective
way of controlling conversations and that the extensive use of questions by teachers
reflects the authority relationship between teacher and child. Our fmdings were
certainly compatible with such an interpretation. The teachers' control was hardly
ever challenged by the children, whilst at home disputes over control were common.
At school over 75% of adult-child conversations were sustained by the teachers'
114 BARBARA TIZARD, MARTIN HUGHES, GILL PINKERTON AND HELEN GARMICHAEL
questions, whereas at home mothers and children played a nearly equal role in
questioning each other and sustaining conversations (Tizard et al., 1980).
The resulting contrast in conversational styles is illustrated by the following two
conversations involving the same working class child Joyce. The first took place at
the clay table at school.
T. What's that going to be, Joyce?
C. No reply.
T. How are you making it?
C. Rolling it.
T. You're rolling it are you? Isn't that lovely. Oh what's happening to it when
you roll it?
C. Getting bigger.
T. Getting bigger. Is it getting fatter?
C. Yeah.
T. Is it, or is it getting longer?
C. Longer.
T. Longer. Are my hands bigger than your hands?
C. My hands litde.
T. Your hands are little, yes.
The teacher's attempts to develop size concepts evoked a minimal response from
Joyce, who seemed to be more interested in making a clay model than in talking
about it. At home, however, Joyce struggled to express herself as she discussed a
recent family trip to the seaside.
C. Mum, some people don't have something to eat at the seaside.
M. What do they do then? Go without?
C. Mm,
M. Ithinkyou'dhave to have something to eat. , . .
C, I still hungry,
M. When?
C. When we was at the seaside, wasn 't I?
M. We weren't. We had sandwiches, we had apples.
C. But we, but when we was there we were still hungry wasn't we ?
M. No, you had breakfast didn't you?
C. But we were thirsty when we got there.
M. Yeah, I suppose so, yeah we were.
C, What happened ? We wasn't thirsty or hungry,
M. Why weren't we? What happened?
C. Well, all that thirsty went away, (Conversation continues.)
It is easy to see that the teacher might well have drawn a false conclusion about
the child's language skills from the conversation at school.
A similar effect of change of setting on children's speech style has been described
by Cole et al. (1978). They compared the dialogue between adults and pre-school
children in two settings: a trip to a locad supermarket and a discussion about the trip
on their return to the classroom. In the supermarket the children talked freely, even
excitedly, about the things that interested them and answered the adults' questions
readily. Back in the classroom, however, their replies to the adults' questions
about the trip were minimal.
ADULTS'COGNITIVE DEMANDS AT HOME AND AT NURSERY SCHOOL
115
Provided the child's interest is aroused, the use of cognitive demands may well
stimulate her thinking. However, our study threw no light on the long-term role of
cognitive demands in intellectual development. All the mothers, of both social
classes, used a variety of types of cognitive demand and we found no evidence of any
simple relationship between the number and type of the mothers' cognitive demands
and the child's I.Q,, There were also few significant social class differences in the
frequency of use of those demands and certainly nothing that would explain the
significant social class difference in I.Q. However, there was a social class difference
in the frequency with which mothers used some of the most difficult demands, e.g.
asking the children to reflect on what they had said ("How do you know?", "What
do you mean?"), but such questions were asked rarely and very few of the children
were able to answer them. Failure to find a correlation between cognitive demands
and I.Q. may have been because the relative infrequency of cognitive demands,
especially the more difficult ones, made them unreliable as a measure. Whatever the
reason, our findings certainly indicate that the association between cognitive
demands and intellectual development is by no means as simple or as strong as is
often suggested. From an educational point of view, our study suggests that
questioning children about their play, a practice currently much used by nursery
school teachers, is an ineffective way of eliciting responses, questions and spontaneous
talk from the child. This appears to be both because of the child's lack of interest
in such conversations and because the social distance between staff and children
inhibits spontaneous conversation.
Our findings also suggest that most children at nursery school seem likely to
receive plenty of cognitive demands at home; teachers could, perhaps, more appropriately concentrate on other aspects of their education. Where this is not the case,
two very different approaches seem more likely to evoke responses from the child
than questioning her about her play—the use of structured games, or a reduction
in the social distance between staff and children, perhaps by the use of smaller
groups and the development of more intimate relationships.
SUMMARY
A study was made of the "cognitive demands" made of children during spontaneous conversations with their teachers at nursery school and their mothers at
home. Social class differences were small compared with home/school differences.
Teachers' talk contained a higher proportion of cognitive demands and "testing"
demands than did mothers' talk; however, the hourly rate of these demands was
higher in the mothers' talk. Teachers used more demands for attributes, simple
descriptions and interpretations, whilst mothers used more demands for recall and
explanation of motives and purposes. More demands were left unanswered at
school than at home. The implications for nursery education are discussed.
Acknowledgements—We would like to acknowledge the co-operation of the teachers, parents and children
in the study, and the statistical help of Ian Plewis and Charles Owen, We are also grateful to Marion
Blank for much useful discussion and to Vera West for help in coding.
116 BARBARA TIZARD, MARTIN HUGHES, GILL PINKERTON AND HELEN CARMIGHAEL
REFERENCES
BLANK, M . (1973) Teaching Learning in the Preschool: A Dialogue Approach. Charles E. Merrill, Columbus,
Ohio.
BLANK, M . , ROSE, S. A. and BERLIN, L . J . (1978) The Language of Learning: The Preschool Years.
Grune & Stratton, New York.
BROWN, R . (1968) The development of Wh-questions in child speech, y. Verb. Learning Verb. Behav. 7,
279-290.
COLE, M . , DORE, J . , HALL, W . and DOWLEY, G . (1978) Situation and task in young children's talk.
Discourse Processes 1, 119-176.
CORSARO. W. A. (1979) Sociolinguistic patterns in adult-child interaction. In Developmental Pragmatics
(Edited by OCHS, E. and SHIEFFELIN, B. B.). Academic Press, New York.
ERVIN-TRIPP, S. (1970) Discourse agreement: how children answer questions. In Cognition and the
Development of Language (Edited hy HAYES,J. R . ) . Wiley, New York.
HUGHES, M . , CARMICHAEL, H . , PINKERTON, G . and TIZARD, B . (1979) Recording children's
conversations at home and at nursery school: a technique and some methodologicEd considerations.
/ . Child Psychol. Psychiat. 20, 225-232.
ISAACS, N . (1930) Children's " W h y " questions. In Intellectual Growth in Young Children (Edited by
ISAACS, S.). George Routledge & Sons, London.
MiSHLER, E. (1975a) Studies in dialogue and discourse: and exponential law of successive questioning.
Lang. Soc. 4, 31-52.
MiSHLER, E. (1975b) Studies in dialogue and discourse—II. Types of discourse initiated by and
sustained through questioning.y. Psycholing. Res. 4,99-121.
TIZARD, B . (1975) Early Childhood Education: A Review and Discussion of Current Research in Britain.
N.F.E.R., Slough.
TIZARD, B . , CARMICHAEL, H . , HUGHES, M . and PINKERTON, G . (1980) Four year olds talking to
mothers and teachers. In Language and Language Disorders in Childhood, Book Suppl. 2 fory. Child
Psychol. Psychiat. (Edited by HERSOV, L . A., BERGER, M . and NiCHOL, A. R.), Ch. 3, pp. 49-76.
Pergamon Press, Oxford.
TOUGH, J. (1976) The Development of Meaning: A Study of Children's Use of Language. Allen and Unwin,
London.
TOUGH, J. (1977) Talking and Learning: A Guide to Fostering Communication Skills in Nursery and Infant
Schools. Ward Lock Educational, London.
WOOD, D . , MCMAHON, L and CRANSTOUN, Y . (1980) Working With Under Fives. Grant Mclntyre,
London.