The nature of phonological representations in young children by leader6


									   The nature of phonological
representations in young children
          Dr Julia Carroll and Joanne Myers

       Report on main findings

       This project has been funded by ESRC

During 2005 – 2006 we have worked with over 125 children aged 3 – 4 years old
from several Warwickshire nurseries and reception classes.

Children were very keen to work with us and completed a series of established tests
and activities that we have designed.

The activities concentrated on children’s language skills and asked them to think
about phonology – sounds which make up words.

This PowerPoint will give you a brief overview of the results of the activities forming
the main findings of our research. We hope that you will find this interesting and
Currently there is much interest in “phonics”; in schools, in the media and in

In education the focus is now on teaching children that words are made of
sounds (rather than teaching whole words) so that they can transfer this
knowledge to other words and to spelling.

A lot of teachers will agree that children find learning about sounds very difficult.
One possible reason behind this difficultly has been proposed in psychology.
It is thought that young children are not able to think about the sound units that
make up words. Therefore the concept that sounds make up words is abstract to
them when teaching begins.

When learning new words children need to “store” this information in their minds
– this is referred to as being “represented”. There are several ways of storing
words. For example cat could be stored in a large chunk as “cat” or could be
broken down into its phonemes - /k/ /a/ t/. These are two of several possible
Understanding how children represent words is important as it is thought that the
quality of representations predicts how well children learn to read.

It is believed that children with speech and language problems have weak
phonological representations. However not much is known about the nature of
representations in children who are developing typically.

We have therefore adapted several activities which ask children to compare,
remember and produce words. This is to allow us to see which sound units
children may be able to think about (initial or final sounds) and which sounds they
think sound the same in order to understand their classification system of sounds
within words. This will help us to understand the nature of representations of
words in typically developing children before and as they learn to read.

We also took basic measures of children’s expressive vocabulary and their
sound/letter knowledge.

This project is based on the way in which we all articulate - say sounds. There are
several elements (phonetic features) as to how we create a sound. In order to
understand some of the key terms of this research please look at these activities:
Say the sound /m/ and try to make it last for a few seconds.
Now say the sound /p/ and do the same. Can you make it last as long?
The path of airflow used to produce sounds like /p/ and /t/ is different and doesn’t
allow for a continuous sound whereas sounds like /m/ and /n/ can be said
Also say the sounds /p/, /t/ and /k/ and you will notice that your tongue and lips
move to different places to produce the sounds.
                             The theory
The different combinations of manner – airflow used to produce sounds and place
– where you position your mouth leads to a classification system where each
sound can be classified in this way. Some sounds share the same manner or
place of articulation. Please see the next slide.

This research was aiming to find out whether children use this same classification
system for storing - “representing” their information about sounds within words.

Very young children struggle to work with specific sounds within words (“what
sound is at the start of cat?”) but have shown that they can work with sections of
words. So it is possible that they are breaking words down into units of some kind
in their representations, but not right down into each specific sound.

A broader classification, being initially more efficient for the child, could be based
on the manner and place classification systems. This means that children may
think about words in terms of the way that they say them.
               The classification system

    Place   Bilabial   Labiodental   Dental   Alveolar   Postalveolar   Palatal   Velar   Glottal

Plosive     p b                               t d                                 k g

Nasal           m                               n

Fricative              f v           th ð     s z        sh   ʒ                    ng      h

Approxim                                        r                       j

Lateral                                         l
Forced Choice/ Sound Families
These activities asked children which words sounded a little bit the same
using friendly characters and their “favourite” words. The sounds within
the words here were being compared to each other. Children showed us
the words sounded “a little bit the same” by giving it to the character or
sounded different by putting the word in the bin.

                          “tough”        “bus”


Memory Confusions
This task used nonsense words as this is a purer measure of children’s
memory ability for different sounds. Children were given the nonsense
Christian names of animals and they had to remember them. The errors
that the children made were analysed. Children could make errors based
on the initial sounds of words or based on manner/place of articulation.

          “mern”                       “moab”                     “vit”
Children were shown pictures and had to quickly tell us what they were. However
beforehand they heard a word being spoken which did not match the picture. The
word was either matched on manner or place (or were unrelated) to investigate
which would have more of an effect on the time for the children to name the
picture. These words could have either helped or distracted the child.


Word Production
Here children were asked to listen to a word and then say as many words as they
could that sounded a little bit the same as it, using puppets. They could make up
nonsense words too.
Background measures
Both the nursery and reception classes scored average to slightly above
average in the expressive vocabulary tests.
As expected the nursery children showed low letter/sound knowledge, with
approximately half being unable to recall more than one.
As hoped the reception children showed good knowledge with half of the
children being able to name at least 24 letter/sounds.

General testing conditions
We were particularly pleased with the levels of concentration of most of the
children. The children were able to work with us and on a computer for up to 15
minutes. This is particularly commendable in the nursery children for whom this
would have been their first exposure to sitting down and completing a piece of
“work” with an adult.
Forced Choice/ Sound Families
Both tasks produced similar results.
Where the consonants within words were matched on manner of articulation
children were more likely to rate these words as sounding similar than when
matched on place or not at all.
This was more likely to happen when the match on manner was made at the
ends of words and not at the beginnings.

Children were more likely to rate words that started or ended with the same
sound as sounding the same as can be seen in the following graph.
                                 Sound similarities in Forced Choice
same" responses out of 2
 number of "sounds the











                                      Relationship between test and target
                                  ("favourite") word and where match occurs
Memory confusions
As adults we are more likely to mix up words that start with the same sound
(just think how many times you’ve mixed up the names Joshua, John and
However children work differently - they have shown us here that they do not
mix up initial sounds particularly.
The children were more likely to mix up words that shared manner at both the
start and end of words than was expected statistically.

Many children showed excellent concentration and memory skills, some
remembering all of the nonsense names over 24 hours period!
Unfortunately our results for priming were disappointing. The effect of words
matched to the target word on manner or place did not produce different effects
to the words that were not related to the target.

We do however plan to make a few changes to the task and hope to see some
results after working with more children.

Word Production
62% of the words that children produced here rhymed with the target word.
This means that when asking a child to produce words that sound the same –
they seem to translate this instruction into give words that rhyme. This may be
useful for teachers to know when introducing the idea of words sounding the

Children here performed very well considering some researchers assume that
children this young will not be able to perform this task. Most children spoke
confidently and rose to the challenge of thinking of words and even making
them up. This was a good exercise as it showed they were able to create
words that sounded the same.
As predicted by other research children are more likely to say that words
sharing manner (at the ends of words) sound similar compared to words sharing
place or words which are unrelated. Additionally, and probably because of this,
children make confusions based on the manner of sounds.
This therefore suggests that we have found evidence that children do represent
words based on the manner in which they are said.
This is relevant in that children could be mixing up sounds which share manner
when they are having to “sound out words” in spelling. So when sounding out
words “mum” and “sun” they could struggle to hear the difference between /m/
and /n/ as they are produced in the same manner. It is useful to be aware of
this possibility.
Some children are able to hear that words that start with the same sound and
end with the same sound (rhyming words) do sound similar. This is how older
children and adults think. It is encouraging to know that this ability is developing.
It could be as a result of educational activities in nursery and reception that
have encouraged this ability.
             How the information has
                   been used
The findings from this study have been presented to various conferences

-including “International Clinical Phonetics and Linguistics Association” (Croatia)
-“Child Language Seminar” (UK)
- “Society for the Scientific Study of Reading” (Canada).

The project will be written up and hopefully published in the following academic
 - Journal of Memory and Language
 - Journal of Child Language.
                     Where to next?
This study has formed the backdrop to a larger study we are about to embark

This will investigate the relationship between children’s phonological abilities
and how well children learn to read. It will concentrate on children with a variety
of speech and language difficulties/risks.

We will be working with children (aged 4 – 6 years)
- in speech and language therapy
- who are at a genetic risk of developing dyslexia
- from local schools who perform at the lower end on a test measuring their
phonological abilities.
                     Where to next?
We aim to work with the children and measure a large number of their
phonological abilities and the nature of their phonological representations (using
some of the same tasks presented here).
We will also test reading and spelling ability using standardised measurements.

We will follow up the children, being particularly interested in how their reading has
progressed over a 6 month period.

We will then be able to consider which elements of their phonology profile are
contributing to reading and which problematic areas may lead to problems also
developing with reading.

This study aims to be able to advise Speech and Language therapists which
children may go onto need some assistance in learning to read.
  Would you like to stay involved?
If you would like some more information on either project or would like to volunteer
for your child to take part in this upcoming research please contact either
       Dr Julia Carroll (024) 765 23613 or
         Miss Joanne Myers (024) 765 75527

   If you would like to receive a copy of the academic paper that will hopefully be
             published as a result of this work then please let us know.

         Thank you for your attention!

To top