Anda di halaman 1dari 23

2

Deafness and Education International


Deafness Educ. Int. 9(1): 223 (2007)
Published online 28 November 2006 in Wiley InterScience
(www.interscience.wiley.com) DOI: 10.1002/dei.204

Writing Profiles of Deaf Children


Taught through British Sign
Language
DIANA BURMAN, TEREZINHA NUNES and DEBORAH EVANS,
University of Oxford, Department of Educational Studies, Oxford, UK
ABSTRACT
Congenitally, profoundly deaf children whose first language is British Sign Language
(BSL) and whose speech is largely unintelligible need to be literate to communicate
effectively in a hearing society. Both spelling and writing skills of such children can
be limited, to the extent that no currently available assessment method offers an adequate appraisal of their competence. Our aim was to create such an instrument to
aid assessment and to support teachers in setting objectives for their deaf students
writing development.
Writing samples describing the same four-picture story were collected from 29
congenitally, profoundly deaf 10-year-old users of BSL. Six experienced teachers of
the deaf ranked their writing productions in five levels; the correlations between their
ranks were high and significant. This indicates that the childrens texts were classified
reliably into categories, which could then be used for further descriptive analysis. The
texts in each category were analysed qualitatively to provide descriptive profiles for
each level. An indication of the concurrent validity of the profiles was obtained
through significant correlations with reading comprehension measures. Future research
should ascertain further the reliability and validity of this instrument and its usefulness
in setting goals for improving deaf childrens writing ability. Copyright 2006 John
Wiley & Sons, Ltd.
Key words: congenitally; profoundly deaf; pre-lingually; British Sign Language; writing assessments; English literacy
INTRODUCTION
Assessment is crucial to the design and evaluation of educational programmes
for the design of effective English literacy teaching for deaf children. A major
goal in effective literacy intervention programmes for deaf children is to identify their specific needs (Tur-Kaspa and Dromi, 2001). However, there are
Copyright 2006 John Wiley & Sons, Ltd.

Deafness Educ. Int. 9: 223 (2007)


DOI: 10.1002/dei

Writing profiles
presently no instruments to assist teachers in assessing the writing of literacy
learners who are congenitally and profoundly deaf and are taught primarily
through British Sign Language (BSL; see Powers and Gregory, 1998). Previous
attempts in achieving this goal have included the adaptation of instructions of
assessments designed for hearing children so that they can be presented to deaf
children (Tymms et al., 2003) and the development of measures appropriate
for deaf children who have made considerable progress in writing (Isaacson,
1996; Heefner and Shaw, 1996; Gormley and Sarachan-Deily, 1987). Our goal
was to develop a descriptive measure of the writing ability displayed by deaf
children who have not yet made great progress in writing in order to help define
aims for their instruction in writing English.
Deaf students writing performance and difficulties
Past research shows that prelingually, profoundly deaf children can learn to
speak but they do not usually attain similar levels of competence in the written
as in the oral form of the language they learned (e.g. Bishop, 1982; Gillam and
Johnston, 1992; Taeschner et al., 1988; Kelly and Whitehead, 1983). They
produce more mistakes when writing a story than when telling it orally; this
difference is observed even when the language sample is elicited through
experimental tasks (Geers and Moog, 1978).
In comparison to hearing peers, deaf students produce shorter sentences,
avoid complex syntactic structures, use a more restricted vocabulary, often omit
function words (e.g. articles and prepositions) as well as use more of these than
necessary on some occasions, and can also omit major constituents of the sentence (e.g. the verb to be or auxiliary verbs) and use inappropriate word order
(Volterra and Bates, 1989; McAffee et al., 1990). Such errors were observed
in the spontaneous writing of post-secondary students who had intelligible
speech and were, therefore, highly successful oral language learners. They had
the possibility of using oral language to plan their writing (for the significance
of oral inner speech in planning writing, see Mayer and Wells, 1996) but they
could not attain in writing the same level of performance in the use of grammar
and morphology that they attained in oral speech.
Studies about the writing of orally educated secondary school deaf students
provide a basis for teachers of adolescents and college students to design appropriate instruction (e.g. Taeschner et al., 1988; McAffee et al., 1990; Tur-Kaspa
and Dromi, 2001). In contrast, there is virtually no research on the writing
ability of primary school students educated in BSL. There is accumulating
evidence that profoundly deaf children exposed to a signed language learn sign
language as efficiently as hearing children learn an oral language (GoldinMeadow, 2003). There is also evidence that deaf children are able to think in
a visual-gestural code (Klima and Bellugi, 1979) and thus use sign language as
a form of internal speech. Swanwick (2002) has clearly documented the use of
BSL by deaf children in planning a piece of writing and has shown how translaCopyright 2006 John Wiley & Sons, Ltd.

Deafness Educ. Int. 9: 223 (2007)


DOI: 10.1002/dei

Burman et al.
tion is required from BSL to English for the childrens writing to be read by a
reader unfamiliar with BSL. Because BSL and English are different languages,
using BSL to plan a piece of written English may show positive transfer as well
as interference (Mayer and Wells, 1996). For example, positive transfer might
be shown in the use of core sentences that have subject, verb and object, but
BSL allows for variable word order and does not use auxiliary verbs for negative
and interrogative sentences. These differences could lead a child who uses BSL
to plan a piece of writing to produce texts that seem disconnected to a reader
unfamiliar with BSL: e.g. boy play where, a literal translation from BSL into
English, seems disconnected to a reader unfamiliar with BSL. In order to reach
the aim of improving the written communication of profoundly deaf children
educated primarily in BSL so that they can more effectively live and work
within the hearing world, we need good descriptions of the levels of writing
they attain during their primary school education. In this paper, we refer to
deaf children educated primarily in BSL simply as child BSL users and those
educated orally as oral deaf children.
There are significant differences between BSL and English, which must be
considered when primary school child BSL users are taught written English.
Some of these are illustrated here to help make the point that planning a piece
of English writing using BSL as the inner language may make the deaf childs
task rather difficult. The examples were chosen not to develop a theory of
written language acquisition but to identify some of the issues that arise when
the writing of deaf children is analysed. Many papers have analysed the positive
and negative transfers that can take place between signed languages and learning written English (e.g. Mayer and Wells, 1996; Mason, 1997; Mayer and
Wells, 1997; Wilbur, 2000; Johnston, 2002), to a large extent inspired by the
theory of linguistic interdependence (Cummins, 1991); others (e.g. Singleton
et al., 2004) have discussed this issue in the context of hearing bilingual learners exposed to two different oral languages or exposed to a signed and an oral
language, as in the case of hearing people born to deaf parents (Fabretti et al.,
1998). In view of the practical aim of this paper, these issues are not discussed
here.
An important difference between BSL and English, which arguably plays a
role in making English literacy acquisition difficult for child BSL users, is the
lack of one-to-one correspondence between signs and written words. BSL has
characteristics that are especially suited for deaf peoples information processing skills (Lane, 1984; Brennan, 1993; Sutton-Spence and Woll, 1999). For
example, BSL is a multi-channel visual language, drawing on deaf BSL users
superior ability to process simultaneous visual information (Emmorey et al.,
1993; Todman and Seedhouse, 1994; Zarfaty et al., 2004). For this reason, there
is not always a one-to-one correspondence between a word and a sign: for
example up until now is one sign, but English uses three words to communicate the same meaning. A child BSL user thus cannot use BSL elements that
is, signs as a match to written elements that is, English words when
Copyright 2006 John Wiley & Sons, Ltd.

Deafness Educ. Int. 9: 223 (2007)


DOI: 10.1002/dei

Writing profiles
trying to discover word order written in a story book. The ability to look at
sentences in books and to try to match spoken and written words is a prereading skill used by hearing pre-schoolers when pretending to read (Ferreiro
and Teberosky, 1983), but this is difficult for those child BSL users who do not
have good English language skills.
A second difference between BSL and English that operates at the word
level is that BSL does not contain many articles, function words (e.g. to, at),
and it does not use the verb to be. Therefore these words may be omitted in
child BSL users written English.
Beyond these differences at the word level, there is also a difference in how
morphemes and grammar are used. BSL does not use bound morphemes (e.g.
ed to mark the past tense or s to mark plurality); these are marked in BSL
by individual signs (e.g. a sign for yesterday indicates the past tense; plurality
is denoted by quantifiers or repetition of the sign). Such morphological differences, as argued by Mayer and Wells (1996), could have an impact on deaf
childrens written production. Research with orally educated students (Taeschner et al., 1988; Fabretti et al., 1998; McAffee et al., 1990; Tur-Kaspa and
Dromi, 2001) shows that omission of morphemes is common in their written
texts but there is little information on the writing of deaf students whose
primary language is a signed language.
The task faced by child BSL users when they are introduced to written
English is quite different from that faced by hearing and oral deaf children.
Deaf child BSL users may start learning literacy with a better knowledge of
spelling than the latter children, because of their exposure to fingerspelling.
Signed and written letters are different but their correspondences can be used
by deaf children in the identification of written words (Hirsh-Pasek, 1986,
1987). But spelling is only the skill involved in putting particular words on
paper; it is recognised by most researchers that writing goes well beyond spelling, to include semantic, syntactic and pragmatic considerations (see Kress,
1982; Moores, 1987; Halliday, 1989; Marschark, 1993; Mayer, 1999). Deaf
child BSL users start learning literacy with an inner speech that does not map
onto written English. The development of their writing skills is still not
described, in spite of the large number of theoretical papers written about the
possible consequences of the use of a signed language to plan writing in deaf
childrens literacy learning. Our aim is to start to address this gap in the literature, which we believe has serious consequences for the planning of deaf childrens literacy instruction.
The assessment of writing in English primary schools
The assessment of primary school childrens progress in literacy in England is
presently carried out according to guidelines proposed by the Qualifications
and Curriculum Authority (Department for Education and Employment, 1999).
Childrens writing is assessed by the teacher based on year-long observation
Copyright 2006 John Wiley & Sons, Ltd.

Deafness Educ. Int. 9: 223 (2007)


DOI: 10.1002/dei

Burman et al.
according to the criteria presented in Table 1. When hearing pupils are about
seven years of age, they are expected to reach Level 2 and when they are about
11 years of age they are expected to reach Level 4.
These writing levels have been used to assess deaf children who had a
cochlear implant before their fifth birthday and seemed to provide a basis for
describing their literacy progress (Watson, 2002) but we suggest that they are

Table 1: Writing assessment criteria proposed by the Qualifications and Curriculum


Authority (1999) and summary of results for 2004.
Attainment target: Writing
Level 1
Pupils writing communicates meaning through simple words and phrases. In their reading or
their writing, pupils begin to show awareness of how full stops are used. Letters are usually
clearly shaped and correctly orientated.
Level 2
Pupils writing communicates meaning in both narrative and non-narrative forms, using
appropriate and interesting vocabulary, and showing some awareness of the reader. Ideas are
developed in a sequence of sentences, sometimes demarcated by capital letters and full stops.
Simple, monosyllabic words are usually spelt correctly, and where there are inaccuracies the
alternative is phonetically plausible. In handwriting, letters are accurately formed and consistent in size.
Level 3
Pupils writing is often organised, imaginative and clear. The main features of different forms
of writing are used appropriately, beginning to be adapted to different readers. Sequences of
sentences extend ideas logically and words are chosen for variety and interest. The basic
grammatical structure of sentences is usually correct. Spelling is usually accurate, including
that of common, polysyllabic words. Punctuation to mark sentences full stops, capital
letters and question marks is used accurately. Handwriting is joined and legible.
Level 4
Pupils writing in a range of forms is lively and thoughtful. Ideas are often sustained and
developed in interesting ways and organised appropriately for the purpose of the reader.
Vocabulary choices are often adventurous and words are used for effect. Pupils are beginning
to use grammatically complex sentences, extending meaning. Spelling, including that of
polysyllabic words that conform to regular patterns, is generally accurate. Full stops, capital
letters and question marks are used correctly, and pupils are beginning to use punctuation
within the sentence. Handwriting style is fluent, joined and legible.
Summary of results for 2004
Pupils at the end of Key Stage 1 (7 years) are expected to attain Level 2. Percentages attaining Level 2 or above: all pupils: 81; boys: 76; girls: 87.
Pupils at the end of Key Stage 2 (11 years) are expected to attain Level 4. Percentages attaining Level 4 or above: all pupils: 63; boys: 56; girls: 71.
Sources: For level descriptors: http://www.ncaction.org.uk/subjects/english/levels.htm and for summary of
results: http://www.standards.dfes.gov.uk/performance/ap/AP2004Archive/?version=1

Copyright 2006 John Wiley & Sons, Ltd.

Deafness Educ. Int. 9: 223 (2007)


DOI: 10.1002/dei

Writing profiles
inadequate for most child BSL users in primary school for two reasons: they do
not take into account the positive transfer and interference that might exist
between BSL and written English, and they start at a level well in advance of
the primary school child BSL users writing. Most pupils in the population
concerned here would not meet the criteria for Level 1 and would be assessed
as W for many years during their learning process. This would render the
assessment useless for monitoring and planning purposes and could also result
in the implicit adoption of the medical deficit model, by placing emphasis on
what the deaf children fail to accomplish rather than by focusing on how their
writing develops (see Rodda and Eleweke, 2000, for a discussion).
Supplementary levels have been created for use with children who have
serious language or intellectual deficits, referred to as P Levels (see http://www.
qca.org.uk/printable.html?url=/8798_7668.html&title=English:
Writing;
accessed 22 August 2006), but these are well below the abilities shown by child
BSL users, who have language, communication and intellectual abilities well
above those for whom these criteria have been developed. An appropriate set
of criteria is urgently needed.

METHOD
Participants
Participants were 18 boys and 11 girls, congenitally deaf, BSL users, attending
Years 5 and 6, sampled from three total communication schools for the deaf in
the London area. Their mean age was 10 years 1 month (SD = 8.01 months).
This age range was selected to provide a greater variation of writing skills than
would be observed with younger children.
All participants had a profound hearing loss (between 100130 dB in the
better ear), except for one girl who was severely deaf (90 dB loss in the better
ear); two boys had cochlear implants and were attending special schools for
the deaf because they had not made progress in oral English. Five children were
born to deaf parents and learned BSL as their first language. Of those born to
hearing parents, seven children had at least one parent/carer able to communicate in sign. The parents of the remaining 14 participants used oral communication; eight used English and six had other first languages. Because of this
variety of linguistic backgrounds, BSL was the only language common to all
the children in their classrooms and was the predominant language of
instruction.
Design
The study was carried out in four phases. Firstly, writing samples were obtained
by presenting participants with pictures and asking them to write a story about
what happened in the picture. This is a familiar task for children taught in
Copyright 2006 John Wiley & Sons, Ltd.

Deafness Educ. Int. 9: 223 (2007)


DOI: 10.1002/dei

Burman et al.
BSL: telling a story from a picture sequence is often used in the teaching of
BSL and forms part of the assessment of proficiency in BSL by CACDP
(Council for the Advancement of Communication with Deaf People).
This method provides a standardised stimulus from which the children can
start their composition. Before the children start to write, the teacher leads a
brief discussion of the scenes, in order to familiarise the pupils with the situation. An alternative method would be to use free composition. We considered
the advantages of free composition but decided to discard this approach for
our purposes, because of the possible difficulty in understanding the meaning
of the childrens writing. When the written production is not entirely connected text and there are spelling mistakes, the meaning of the text might be
unclear and this makes it difficult for an assessor to evaluate the childs writing.
Fabretti et al. (1998) demonstrated the difficulty of such evaluations even
when participants are adults with a high mastery of the written language. A
third alternative approach would be to elicit writing from a signed story (see
Swanwick, 2002). This task might be akin to translation, which is arguably
different from a bilingual childs ability to perform in either language. Fabretti
et al. (1998) observed that writing a letter, the content of which was preestablished (declining an invitation), elicited significantly fewer non-standard
grammatical and morphological forms than writing a story previously told in
sign.
Our second step was to obtain independent ratings of the writing samples
by teachers of the deaf and assess the inter-judge reliability of their ratings.
The teachers were simply asked to classify the samples into five levels of skill.
Previous research indicated that, when the concepts to be used are fuzzy, one
should start with less precise ratings, up to five or seven categories, and then
use the categories in order to identify the criteria for inclusion in the different
levels of performance (Nunnally, 1978).
The third step was to examine the concurrent validity of the teachers
assessment. The concurrent validity was examined in two ways. Firstly, written
productions were obtained from the same children at the start and at the end
of the school year. Because our aim was to develop an assessment that can help
measure childrens progress, it was necessary that we should start from ratings
that can do so. Secondly, we gave the children a reading comprehension task
(Burman, 2004) and analysed the correlation between the writing levels and
reading comprehension. Reading comprehension and writing ability are known
to be correlated (Parodi, in press, reports a commonality of 51.8% between
factors affecting comprehension and production from the cognitive/textual
perspective; see also Shanahan, 1984, and Shanahan and Lomax, 1986); thus,
if a significant correlation is observed between the two assessments, there will
be some evidence for its validity.
The fourth and final step consisted of analysing the childrens productions
within each level in order to produce a description of levels of writing among
child BSL users.
Copyright 2006 John Wiley & Sons, Ltd.

Deafness Educ. Int. 9: 223 (2007)


DOI: 10.1002/dei

Writing profiles
Procedure
Writing samples were elicited by presenting the children with a four picture
sequence (see Figure 1) in poster size and in colour at the front of the class and
asking them to write a story. Each child had a copy of the pictures on the
workbook where their story was to be written.
The researcher (the first author, who is fluent in BSL) showed the children
the pictures and invited them to comment on what was happening in the pictures. They were told they could assign names to the people and that they could
request the spelling of words that they wanted to write but did not know how
to spell. The words they wrote from a finger-spelling or a card model were
underlined by the researcher to show that they had received spelling help. If
a card model was used, it was placed at a right-angle to the normal orientation
of writing, so that the child still had the task of recognising which way up to
place the card prior to copying. The children were allowed 25 minutes to write
their story.
Creating the writing levels
Six experienced teachers of the deaf individually and independently classified
the childrens anonymous productions. Two productions were obtained from

Figure 1: The Writing Task.

Copyright 2006 John Wiley & Sons, Ltd.

Deafness Educ. Int. 9: 223 (2007)


DOI: 10.1002/dei

10

Burman et al.
the same children, one at the start and the other at the end of the school year,
but the teachers were blind to testing occasion. Five levels, from least to most
proficient writing samples, were used: E represented the weakest and A the
strongest knowledge of written English.
RESULTS
The results are presented in three sections: the first describes the reliability of
the teachers ratings, the second the evidence for concurrent validity of their
ratings and the third presents the results of the qualitative analysis of the
writing levels.
Reliability
The inter-judge reliability of the assessments was analysed by computing the
Spearman inter-correlations between the teachers classifications. The correlations for the first occasion are presented in Table 2.
The correlations were all above 0.8 and significant at the 0.01 level. For
the second occasion, all correlations were also above 0.8 and significant at the
0.01 level. Thus the teachers were able to make reliable judgements, and their
classification could be used to produce a qualitative analysis of the different
levels of writing.
For the purpose of further analyses, discrepancies were resolved by averaging
across the classifications (using numbers, 1 to 5).
Validity of the teachers ratings
In order to obtain an indication of the validity of the teachers ratings, two
analyses were carried out. Firstly, the classifications of the childrens writing at
the start and end of the year were compared. An assessment designed for monitoring childrens progress is only valid if it is able to measure progress over the

Table 2: Inter-correlations between the ratings of the different teachers (N = 29).

Teacher 1
Teacher 2
Teacher 3
Teacher 4
Teacher 5
Teacher 6

Teacher 1

Teacher 2

Teacher 3

Teacher 4

Teacher 5

0.90**
0.89**
0.96**
0.92**
0.96**

0.87**
0.90**
0.85**
0.87**

0.90**
0.88**
0.90**

0.94**
0.96**

0.93**

**Correlation is significant at the 0.01 level (2-tailed)

Copyright 2006 John Wiley & Sons, Ltd.

Deafness Educ. Int. 9: 223 (2007)


DOI: 10.1002/dei

Writing profiles
course of one school year. Even if some children do not progress, the measure
should detect average progress. A t-test for correlated samples was carried out,
using values from 1 to 5 to represent the levels, with 5 representing Level A.
The mean score attained by the students at the start of the school year was
2.72 (SD = 1.39) and at the end of the year was 3.48 (SD = 1.33). The difference between the two testing occasions was significant (t = 4.48; df = 28; p <
0.001), demonstrating that the teachers ratings did detect progress in childrens writing ability, even though this progress would not be detected by the
levels used in the National Curriculum levels. Thus we felt confident in the
use of these ratings as the starting point for creating descriptions of writing
levels.
The Spearman correlation between the reading comprehension and the
writing levels, both administered at the beginning of the year, was 0.71, which
is significant at the 0.01 level. This provides some indication of the external
validity of the teachers ratings.
The writing levels
The previous analyses gave us confidence in the use of the teachers ratings as
the starting point for the qualitative analysis of the writing samples. In order
to illustrate the standard of writing in each level, every childs writing for
Picture 1, first assessment, is presented (Figures 2 through 6).
Even a superficial reading shows that most of these productions can only be
interpreted by attempting to rephrase the text in standard form. This method,
used by Fabretti et al. (1998), allowed them to arrive at a quantitative analysis
of adults writing, which identified the number of omissions, non-standard
forms (e.g. use of wrong tenses or agreement between noun, adjective and
article) and unnecessary insertions (e.g. use of an article before a verb). This
strategy could not be used with the writing samples in our study for two reasons.
First, there were productions whose meaning we could not identify. Second,
there were also many productions the meaning of which could be interpreted
but there would be many alternative ways of rephrasing the texts; identifying
different standard forms results in quantifying the non-standard forms differently. Thus we decided to analyse the samples qualitatively and provide a
description of the common features to create a profile of each level. When we
provide interpretations of the childrens writing, these are used in order to seek
their meaning, and not for quantification. Our interpretations are presented in
square brackets; the reader can then assess whether alternative interpretations
would better fit the context.
Levels were conceived in a way that progress attained in the lower levels
should also be manifested in higher levels. For example, the criterion uses
proportionally more words relevant to the story than words that are not relevant appears for the first time in Level C; this criterion should also be met by
texts classified in Levels B and A.
Copyright 2006 John Wiley & Sons, Ltd.

Deafness Educ. Int. 9: 223 (2007)


DOI: 10.1002/dei

11

12

Burman et al.

Figure 2: Level E Writing Samples.

Copyright 2006 John Wiley & Sons, Ltd.

Deafness Educ. Int. 9: 223 (2007)


DOI: 10.1002/dei

Writing profiles
Level E represents the weakest productions in this study (n = 7). These
samples show that the children were able to form letters correctly. Underlined words denote that a spelling was asked for in BSL. Most of the children
were able to write the correct letters when provided with the finger-spelling
for the words. However, not all children in Level E were able to do this: day
clothes dag was copied from written flash-cards. Children who request a
spelling for a sign understand that letters can be used to represent signs and
offer a means of communication, which is not the case for all children (see
the P levels referred to earlier on). It is not clear whether they realise that
letters represent the sounds in English or whether they think in terms of
direct connections between BSL signs and written words. However, the word
sequences observed in this level of writing reflect neither BSL nor English
syntax.
The writing profile for Level E was created by summarising the features of
the samples that had been classified in this level. At Level E the writing sample
demonstrates ability to:

write correctly oriented letters;


use spaces to create letter sequences that resemble words but are not always
identifiable words even in context (e.g. wihus; Ba);
memorise some finger-spelling configurations and their corresponding
written letter;
produce letter sequences for isolated words, which may or may not be relevant to the situation (e.g. door dag or written under the picture of the
car travelling to the coast);
occasionally, but not systematically, place a definite or indefinite article
before a remembered spelling, which may be correct or incorrect, and may
or may not be a noun (e.g. A play barch to two boy play [interpreted as
two boys play on the beach], a the simning [interpreted as the boy and
girl go swimming]).

Level D writing (n = 7) shows that the children were able to write some
words without help and to choose appropriate words to write under the pictures
but some obscure spellings are also observed. Their writing includes English
words not used in BSL, such as to and the. Word sequences were not wellformed sentences but it is possible to identify some noun phrases and subjectverb word order. This indicates the emergence of English syntax. The inclusion
of memorised words, for example, bags, father, holiday and summer suggests that the children understand that writing is a form of communication and
that the writing and the picture go together to create a meaningful story.
Occasional errors when the children wrote words that were finger-spelled were
observed.
Writing samples classified as Level D demonstrated the childrens ability
to:
Copyright 2006 John Wiley & Sons, Ltd.

Deafness Educ. Int. 9: 223 (2007)


DOI: 10.1002/dei

13

14

Burman et al.

Figure 3: Level D Writing Samples.

Copyright 2006 John Wiley & Sons, Ltd.

Deafness Educ. Int. 9: 223 (2007)


DOI: 10.1002/dei

Writing profiles

write some sequences showing BSL word order with emerging English
syntax marked by the presence of grammatical morphemes (e.g. father bag
put to the car shows non-standard English word order, with the object
before the verb, but also a preposition and an article);
form noun phrases and use subject-verb word order;
use pronouns (mostly I), conjunctions and some function words;
include some appropriate, although not systematic, punctuation;
begin to name people in order to create a story character.

Level C productions (n = 6) mainly demonstrate BSL syntax but also show


emerging English grammar in the use of phrases for example, we went to,
One day, The lady made and in the bag. The words went and to show a
grasp of English past tense and use of a preposition, neither of which is used
in BSL. All the samples include verbs and show subject-verb word order. This
level of writing can be interpreted as a text by a reader even if more than one
interpretation is possible and there any many non-standard forms.
In short, writing samples in Level C demonstrate ability to:

use a higher proportion of words relevant to the story than of unidentifiable


spellings or irrelevant words;
place words in a coherent order with greater awareness of English syntax
(e.g. Sard [characters name] want look to the biggest fish);
include verbs in most sentences and use subject-verb word order;
start to organise sequences in a developing story;
use some English idiomatic expressions (e.g. tidy up, far away) and
expressions typical of stories (e.g. One day);
correctly use optional elements in English syntax, for example, by placing
an adjective before a noun.

Level B productions (n = 5) essentially show the use of English words for


BSL signs but also indicate awareness of English syntax, through the use of
prepositions and pronouns, and a growing awareness of story development, by
giving names to and describing characters. Writing samples at this level differ
from those classified in Level C mostly by the use of a concatenation of sentences which begin to form a text.
Writing samples classified as Level B demonstrate the ability to:

produce a text where the vast majority of words are relevant to the story;
even if the spellings are not accurate, the words are recognisable in context
(e.g. gril for girl; very tried for very tired; hoilday for holiday; clot
for clothes; and divir for drive);
transcribe BSL into English words (e.g. in car will go to day hoilday with
friend boy called Edward);

Copyright 2006 John Wiley & Sons, Ltd.

Deafness Educ. Int. 9: 223 (2007)


DOI: 10.1002/dei

15

16

Burman et al.

Figure 4: Level C Writing Samples.

Copyright 2006 John Wiley & Sons, Ltd.

Deafness Educ. Int. 9: 223 (2007)


DOI: 10.1002/dei

Writing profiles

Figure 5: Level B Writing Samples.

Copyright 2006 John Wiley & Sons, Ltd.

Deafness Educ. Int. 9: 223 (2007)


DOI: 10.1002/dei

17

18

Burman et al.

include English words not used in BSL more systematically, not only
occasionally;
include English syntax through use of noun phrases and subject-verbobject sequences;
develop a story and create a character (e.g. Put bag in a car Daddy was
slepp in car was fast. Arrive in Spian that very hot in costa brava [note
that the child names the beach as Costa Brava in order to develop the
story]).

Level A productions (n = 4) show more systematic use of the indefinite and


definite articles, prepositions, conjunctions, inflections and the inclusion of the
verb to be. The first instances of subordinate clauses were observed in the
texts, with because as the conjunction.
Writing samples in Level A demonstrate the ability to:

use pronouns other than I and English grammar (pronouns followed by


verbs, even if agreement is not correct: e.g. he put, We goes);
include sufficient English syntax to produce fully standard sentences (e.g.
The boy was called Sam), though non-standard forms may appear also;
expand on the description of scenes (e.g. Mummy and two children have
a foods for family, when family feel very hunry in the car, then family have
a foods);
use of subordinate clauses;
use a higher proportion of accurate than wrong spellings and almost exclusively identifiable words even if some spellings are not accurate nor morphologically appropriate (e.g. The Sam goes downsalts [downstairs] and
put the Bag in the car Dada say hey we Ready for Lunch now .).

CONCLUSIONS AND DISCUSSION


Our aim was to develop a measure of child BSL users progress in writing
English, which could then be used to assess the childrens progress and help
define teaching aims for them. Our approach was to start with samples of
written productions, elicited by a set of pictures and a discussion of what they
showed, but without a specific story which the children were expected to write.
The children were free to produce the content of the text themselves.
As hypothesised in the design section, where we presented the rationale for
our choice of method, it proved useful to have a constrained production: on
many occasions, it was easier to interpret non-standard spellings and grammatical constructions because the context for the writing was known. This also
allowed us to realise that some of the children, particularly those whose productions were classified in Levels E and D, included in their writing spellings which
they remembered, correctly or incorrectly, but the relevance of the word to the
story was unclear to us.
Copyright 2006 John Wiley & Sons, Ltd.

Deafness Educ. Int. 9: 223 (2007)


DOI: 10.1002/dei

Writing profiles

Figure 6: Level A Writing Samples.

Copyright 2006 John Wiley & Sons, Ltd.

Deafness Educ. Int. 9: 223 (2007)


DOI: 10.1002/dei

19

20

Burman et al.
Writing samples of these two most rudimentary levels suggest that child
BSL users face difficulties not encountered by children who already speak
English: they have to learn that the language they are learning to write does
not represent their language of instruction in school.
Samples classified in the two lowest levels show little evidence of use of
English syntax. It is only from Level C onwards that the written sentences show
a consistent use of noun phrases and subject-verb constructions, thus going
beyond the use of unconnected words. The use of words that appear in English
but not in BSL such as articles, prepositions and the verb to be is rare in
the writing samples classified below Level C.
Consistent use of noun phrases and verbs, which can help identify units in
writing, is only present in writing samples classified in Levels B and A. Evidence of connections between clauses is only present in writing samples classified as Level A. However, the still only emergent use of English syntax
prevents these samples from meeting the criteria specified for Level 1 in the
National Curriculum, which requires children to be able to communicate using
simple words and sentences.
We suggest that the use of these profiles in the assessment and design of
specific instruction for deaf child BSL users can make a contribution to their
teaching and learning. Webster (2000) recognised the importance of literacy
for deaf individuals to function in a hearing society and expressed dismay at
the difficulty of locating effective routes for raising literacy achievements for
the deaf; he argued that the spectacular lack of success by most deaf children
(p. 131) in literacy should bring about more research to determine the effectiveness of different instructional methods. We suggest that, without proper
assessment tools, it is difficult to carry out meaningful research.
Mayer (1999) also addressed the low literacy levels of deaf individuals and
pointed out that their reading attainment is often at higher levels than their
writing, because in reading they can use compensatory strategies whereas these
are of much less consequence for writing. She further suggests that descriptions
that focus only on what might be seen as the deficiencies of the texts produced
by deaf students are less helpful than those that can shed some light on the
process of writing acquisition and composition. We believe that the descriptions offered in this paper are a first step towards understanding these processes
in child BSL users.
However, we are very aware of the limitations of this initial work and the
need for further research. First, our samples were small and restricted to primary
school age; it cannot be expected that they will represent all the variations
that can be observed before children reach Level 1 in the National Curriculum.
It is important to obtain larger samples of writing and analyse how well these
levels describe child BSL users performance. Second, we recognise that a
measure obtained under time constraints is necessarily a measure of performance, not a measure of competence (Goldstein and Bebko, 2003), and thus we
could underestimate the childrens writing skills. However, if the best possible
Copyright 2006 John Wiley & Sons, Ltd.

Deafness Educ. Int. 9: 223 (2007)


DOI: 10.1002/dei

Writing profiles
piece were to be obtained by increasing the amount of support that the children
receive, their difficulties could be easily underestimated. Finally, it is also necessary to evaluate through future research whether teachers can reliably use the
criteria provided by our writing profiles to assess their pupils progress and
design further instruction for them: this is, in the end, the proof of the pudding
the translation of assessment into practice that supports childrens further
learning.
ACKNOWLEDGEMENTS
We wish to thank the staff and pupils of Heathlands School for Deaf Children,
St Albans, and Grove House and Blanche Nevile Schools for the Deaf for
participating in this study. Without their generous collaboration, this work
would not have been possible.
REFERENCES
Bishop DVM. Comprehension of spoken, written and signed sentences in childhood language
disorders. Journal of Child Psychology and Psychiatry 1982; 23: 120.
Brennan M. The visual world of British Sign Language: An introduction. In: Brien D (ed.)
Dictionary of British Sign Language / English. London. Faber & Faber, 1993.
Burman DL. Using morphology to improve congenitally, profoundly deaf childrens spelling and
writing: a study with BSL users. Unpublished doctoral thesis, Oxford Brookes University,
2004.
Cummins J. Interdependency of first- and second-language proficiency in bilingual
children. In: Bialystok E (Ed.) Language Processing in Bilingual Children. Cambridge:
Cambridge University Press, 1991; pp. 7089.
Department for Education and Employment. The National Curriculum Handbook for Primary
Teachers in England. London. Qualifications & Curriculum Authority, 1999.
Emmorey K, Kosslyn S, and Bellugi U. Visual imagery and visual-spatial language: Enhanced
imagery abilities in deaf and hearing ASL signers. Cognition 1993; 46: 139181.
Fabretti D, Volterra V, and Pontecorvo C. Written language abilities of deaf Italians. Journal of
Deaf Studies and Deaf Education 1998; 3: 231244.
Ferreiro E, Teberosky A. Literacy Before Schooling. Exeter, New Hampshire. Heinemann Educational Books, 1983.
Geers AE, Moog JS. Syntactic maturity of spontaneous speech and elicited imitations of hearing
impaired children. Journal of Speech and Hearing Research 1978; 43: 380391.
Gillam R, Johnston T. Spoken and written language relationships in language/learning impaired
and normally achieving school-age children. Journal of Speech and Hearing Research 1992;
35: 13031315.
Goldin-Meadow S. The Resilience of Language. What gesture Creation in Deaf
Children Can Tell Us About How All Children Learn Language. New York. Psychology
Press, 2003.
Goldstein G, Bebko JM. The profile of multiple language proficiencies: A measure for evaluating
language samples of deaf children. Journal of Deaf Studies and Deaf Education 2003; 8:
452463.
Gormley K, Sarachan-Deily AB. Evaluating Hearing-Impaired Students Writing: A Practical
Approach 1987; The Volta Review 89, 157.

Copyright 2006 John Wiley & Sons, Ltd.

Deafness Educ. Int. 9: 223 (2007)


DOI: 10.1002/dei

21

22

Burman et al.
Halliday MAK. Spoken and Written Language. Oxford. Oxford University Press, 1989.
Heefner DL, Shaw PC. Assessing the written narratives of deaf students using the Six-Trait
Analytical Scale. Volta Review 1996; 98 (1), 147168.
Hirsh-Pasek K. Beyond the great debate: fingerspelling as an alternative route to word identification for deaf or dyslexic readers. The Reading Teacher 1986; 39: 340343.
Hirsh-Pasek K. The metalinguistics of fingerspelling: an alternate way to increase reading
vocabulary in congenitally deaf readers. Reading Research Quarterly 1987; 22: 455474.
Isaacson S. Simple ways to assess deaf or hard-of-hearing students writing skills. Volta Review
1996; 98 (1), 183199.
Johnston T. The representation of English using Auslan: Implications for deaf bilingualism and
English literacy. Australian Journal of Education of the Deaf 2002; 8: 2337.
Kelly JF, Whitehead RL. Integrated spoken and written English instruction for the hearing
impaired. Journal of Speech and Hearing Disorders 1983; 48: 415422.
Klima ES, Bellugi U. The Signs of Language. Cambridge (MA). Harvard University Press,
1979.
Kress G. Learning to Write. London. Routledge and Kegan Paul, 1982.
Lane H. When the Mind Hears. New York. Random House, 1984.
Marschark M. Psychological Development of Deaf Children. New York: Oxford
University Press, 1993.
Mason D. Response to Mayer and Wells: the answer should be affirmative. Journal of Deaf
Studies and Deaf Education 1997; 2: 277279.
Mayer C. Shaping at the point of utterance: An investigation of the composing processes of the
Deaf student writer. Journal of Deaf Studies and Deaf Education 1999; 4: 3749.
Mayer C, Wells G. Can the linguistic interdependence theory support a bilingual-bicultural
model of literacy education for deaf students? Journal of Deaf Studies and Deaf Education
1996; 1: 93107.
Mayer C, Wells G. The question remains: a rejoinder to Mason. Journal of Deaf Studies and
Deaf Education 1997; 2: 280292.
McAfee M, Kelly JF, and Samar VJ. Spoken and written English errors of postsecondary students
with severe hearing impairment. Journal of Speech and Hearing Disorders 1990; 55:
628634.
Moores D. Educating the Deaf Child: Psychology, Principles and Practice. Boston. HouhgtonMifflin, 1987.
Nunnally J. Psychometric Theory. New York. McGraw-Hill, 1978.
Powers S, Gregory S. The Educational Achievements of Deaf Children. Norwich. Department
for Education and Employment, Her Majestys Stationery Office, 1998.
Qualifications and Curriculum Authority (1998) National Literacy Strategy Framework for
Teaching. Available: www.qca.org.uk/printable.html?url=/8798_7668.html&title=English:
Writing Last accessed March 2006.
Rodda M, Eleweke CJ. Theories of literacy development in limited English proficiency
deaf people: A review. Deafness and Education International 2000; 2: 101113. DOI:
10.1002/dei.77.
Shanahan T. Nature of the reading-writing relation: An exploratory multivariate analysis.
Journal of Educational Psychology 1984; 76: 466467. DOI: 10.1002/dei.120.
Shanahan T, Lomax R. An analysis and comparison of theoretical models of the reading-writing
relationship. Journal of Educational Psychology 1986; 78: 116123.
Singleton JL, Morgan D, DiGello E, Wiles J, and Rivers R. Vocabulary use by low, moderate,
and high ASL-proficient writers compared to hearing ESL and monolingual speakers. Journal
of Deaf Studies and Deaf Education 2004; 9: 86103.
Sutton-Spence R, Woll B. The Linguistics of British Sign Language: An Introduction. Cambridge, UK. Cambridge University Press, 1999.

Copyright 2006 John Wiley & Sons, Ltd.

Deafness Educ. Int. 9: 223 (2007)


DOI: 10.1002/dei

Writing profiles
Swanwick R. Sign bilingual deaf childrens approaches to writing: individual strategies for bridging the gap between BSL and written English. Deafness and Education International 2002;
4: 6583. DOI: 10.1002/dei.120.
Taeschner T, Devescovi A, and Volterra V. Affixes and function words in the written language
of deaf children. Applied Psycholinguistics 1988; 9: 385401.
Todman J, Seedhouse E. Visual-action code processing by deaf and hearing children. Language
and Cognitive Processes 1994; 4: 129141.
Tur-Kaspa H, Dromi E. Grammatical deviations in the spoken and written language of Hebrewspeaking children with hearing impairments. Language, Speech and Hearing Services in
Schools 2001; 32: 7989.
Tymms P, Brien D, Merrell C, Collions J, Jones P. Young deaf Children and the prediction of
reading and mathematics. Journal of Early Childhood Research, 1: 197202. Handbook of
Childrens Literacy. In: Nunes T, Bryant P. Dordrecht. Kluwer Academic Publishers,
2003.
Volterra V, Bates E. Selective impairment of Italian grammatical knowledge in the congenitally
deaf: a case study. Cognitive Neuropsychology. 1989; 6: 273308.
Watson L. The literacy development of children with cochlear implants at age seven. Deafness
and Education International 2002; 4: 8498. DOI: 10.1002/dei.121.
Webster A. An international research review of literacy intervention strategies for children with
severe to profound deafness. Deafness and Education International 2000; 2: 128141.
Wilbur RB. The use of ASL to support the development of English and Literacy. Journal of Deaf
Studies and Deaf Education 2000; 5 (1), 81104. DOI: 10.1002/dei.84.
Zarfaty Y. Nunes T, and Bryant P. The performance of young deaf children in spatial and temporal number tasks. Journal of Deaf Studies and Deaf Education 2004; 9: 315326.

Address correspondence to: Dr Diana Burman, University of Oxford, Department


of Educational Studies, 15 Norham Gardens, Oxford OX2 6PY, UK. (E-mail:
diana.burman@edstud.ox.ac.uk)

Copyright 2006 John Wiley & Sons, Ltd.

Deafness Educ. Int. 9: 223 (2007)


DOI: 10.1002/dei

23