Vous êtes sur la page 1sur 19

Vocabulary Use by Low, Moderate, and High

ASL-Proficient Writers Compared to Hearing


ESL and Monolingual Speakers
Jenny L. Singleton
Dianne Morgan
Elizabeth DiGello
Jill Wiles
Rachel Rivers
University of Illinois at Urbana-Champaign

The written English vocabulary of 72 deaf elementary school from findings one might expect based on previous studies
students of various proficiency levels in American Sign with deaf writers and their vocabulary use. The writing of the
Language (ASL) was compared with the performance of 60 deaf writers also differed from the writing of hearing ESL
hearing English-as-a-second-language (ESL) speakers and speakers. Implications for deaf education and literacy in-
61 hearing monolingual speakers of English, all of similar struction are discussed, with special attention to the fact that
age. Students were asked to retell The Tortoise and the ASL-proficient, deaf second-language learners of English
Hare story (previously viewed on video) in a writing activity. may be approaching English vocabulary acquisition in ways
Writing samples were later scored for total number of words, that are different from hearing ESL learners.
use of words known to be highly frequent in childrens
writing, redundancy in writing, and use of English function Researchers have documented the strong relationship
words. All deaf writers showed significantly lower use of between vocabulary knowledge and literacy develop-
function words as compared to their hearing peers. Low-
ASL-proficient students demonstrated a highly formulaic
ment. According to Nation (1990, p. 93), in everyday
writing style, drawing mostly on high-frequency words and spoken English, 99% of what is said can be captured by
repetitive use of a limited range of function words. The the first 2000 most common words (see frequent-word
moderate- and high-ASL-proficient deaf students writing lists such as Dale & Chall, 1948; Hillerich, 1978;
was not formulaic and incorporated novel, low-frequency
vocabulary to communicate their thoughts. The moderate- Kucera & Frances, 1967; West, 1953). Furthermore,
and high-ASL students performance revealed a departure incidental learning of vocabulary (e.g., the learning of
This study was based in part on the undergraduate honors theses of Jill new vocabulary by listening to conversation, exposure
Wiles, who now works in Chicago, Illinois, and Rachel Rivers, adjunct to novel vocabulary through reading) accounts for
faculty at University of Hawaii, Department of Educational Technology.
Dianne Morgan and Elizabeth DiGello are both doctoral students at a large proportion of first-language vocabulary growth
University of Illinois. We would like to thank Brenda Schick, Christie by school-age children (Nagy, Herman, & Anderson,
Yoshinaga-Itano, Sharon Litchfield, and the anonymous JDSDE
reviewers for helpful feedback during the preparation of this article. 1985). First-language learners add from three to seven
We thank the following individuals for their valuable assistance with data new words per day to their vocabularies. Nagy and
collection and coding: Jennifer Alms, Ellen Altermatt, Benita Cervantes,
Robert Dramin, Susan Dramin-Weiss, Jessica Fulton, Lisa Jucha, Herman (1987) suggest that even one incidental en-
Jennifer Klodnicki, Johanna Lewis, Sharon Litchfield, Susan Maller,
counter with a novel word may be enough to learn it,
Sarika Mehta, Eriko Miyao, Cindy Neuroth-Gimbrone, Sara Schley,
Melissa Sonka, Samuel Supalla, Wendy Troop-Gordon, Steven Vaupel, while other researchers contend that 6 to 12 encounters
Tina Wix. Correspondence should be addressed to Dr. Jenny Singleton,
may be a more realistic estimate (Jenkins & Dixon,
Department of Educational Psychology, 1310 S. Sixth Street, 226
Education, Champaign, IL 61820 (e-mail: singletn@uiuc.edu). 1983). Children learn new words from other sources

Journal of Deaf Studies and Deaf Education vol. 9 no. 1 Oxford University Press 2004; all rights reserved. DOI: 10.1093/deafed/enh011
Deaf Writers Vocabulary Use 87

such as television, computers, glossaries, and dictio- 17-year-old students. Silverman-Dresner and Guilfoyle
naries (Nation, 1990). Teachers can also provide explicit provided no information about participant inclusion
vocabulary instruction in the classroom context. criteria or communication modes used by the deaf
As a result of their hearing loss, profoundly deaf students and their teachers. We can only surmise from
children have great difficulty acquiring English the date of research that the deaf students were either
vocabulary through the same incidental learning orally educated or beginning to use Total Communica-
processes as hearing children. Because they do not tion; we have no way of knowing the students access to
overhear conversation and have limited early literacy American Sign Language (ASL) outside of the
experiences in English, deaf children struggle to classroom context. What we can take from this study
develop their English vocabularies at age-appropriate is some evidence to support the notion that deaf
levels. Many researchers have documented depressed children historically have struggled with English
English vocabulary and reading comprehension scores vocabulary, especially function words. However, since
among deaf children when compared to their hearing the method of Silverman-Dresner and Guilfoyles
peers (see King & Quigley, 1985; Marschark, 1997; vocabulary assessment required reading four definitions
Karchmer & Mitchell, 2003; Wilbur, 2000, for re- per word, we also must question the language
views). King and Quigley (1985) reported that deaf comprehension demands of the task and ask whether
students usually have their lowest performance on the same results would have been obtained if the deaf
vocabulary or word meaning subtests when compared students were presented the definition choices in ASL.
to their performance on other standardized test In short, DeVilliers and Pomerantz (1992) are
domains such as spelling, mathematics, and language likely correct in their suggestion that many hearing-
mechanics (e.g., punctuation, capitalization). LaSasso impaired students are caught in a vicious circle: their
and Davey (1987) found that vocabulary scores impoverished vocabularies limit their reading compre-
correlated strongly with reading comprehension scores hension and poor reading strategies and skills limit
for the deaf students they studied. Paul and Gustafson their ability to acquire adequate vocabulary knowledge
(1991) also found similar correlations. from context (p. 428).
In 1972, Silverman-Dresner and Guilfoyle (re- Paul (1998) includes writing deficits as part of that
searchers connected with Lexington School for the vicious circle. He states that there is a reciprocal
Deaf in New York) authored a monograph entitled relationship between conversational and written forms
Vocabulary Norms for Deaf Children. This list of of English, and deaf children are unable to benefit fully
7300 words, based initially on Dale and Challs list of from this relationship. Based on their review of the past
3000 vocabulary words (Dale & Chall, 1948), was 50 years of research on deaf childrens writing,
validated on a national sample of 26,414 deaf students Marschark, Lang, and Albertini (2002) report that
between the ages of 8 and 17. Each deaf student was
deaf students generally produce shorter and less
presented with two lists of 100 vocabulary terms
structurally variable sentences than hearing peers,
randomly selected from the pool of 7300 words and
when they produce complete sentences at all. Al-
was asked to select the correct definition (which
though they are generally as competent as hear-
appeared alongside three distractor definitions) for
ing peers in the use of punctuation and in
each word. Silverman-Dresner and Guilfoyle reported
spelling, deaf children tend to use stock words
accuracy levels of respondents according to age and
and phrases repeatedly in a text, and they use
word type (using a criterion of 62.5% of children in the
more articles and nouns and fewer adverbs and
age group identifying the correct definition). Deaf
conjunctions. Words are frequently omitted, and
students aged 89 correctly identified only 18 words, 6
sentences generally are less syntactically complete
of which were color terms, and none of which were
and less well interconnected in compositions than
function words (words serving a grammatical function
those of hearing peers. (p. 172)
such as prepositions). In fact, of the 65 function words
from the pool, only half of them were correctly Most studies have grouped together deaf children
identified by more than 62.5% of the 16- to of varying sign language skill (but with similar hearing
88 Journal of Deaf Studies and Deaf Education 9:1 Winter 2004

loss levels) and have found that profoundly deaf Yoshinaga-Itano, Snyder, and Mayberry (1996b) found
children perform poorly on English vocabulary that the 49 deaf students (ages 1014) included in their
measures. The unique contribution of the current study produced shorter essays than the hearing
study is that we have categorized deaf children comparison group. However, these two groups did
according to their proficiency level in ASL, which not differ in the number of propositions they ex-
previous studies have not done. In the present study, pressed, indicating a similarity in conceptual expres-
we assess English vocabulary use by deaf children on an sion; hearing students included more words within
elicited writing task. The study includes children in the each proposition, yielding a greater overall total
emergent stages of writing development (ages 612), number of written words. Hearing students also
whereas most prior writing and vocabulary research introduced topics and then elaborated upon them
with deaf children has focused upon older writers (ages while deaf students tended to introduce many new
1018). Finally, we provide a context for the findings topic sentences without subordinated, elaborative
by offering parallel data from two comparison groups: information. Wilbur (2000) discussed similar findings
hearing monolingual speakers of English and hearing and argued that deaf students have difficulty with
English-as-a-second-language (ESL) speakers. pushing old information into the background and
These two comparison groups (monolingual and bringing new information into the foreground.
ESL) are important for several reasons. First, our
vocabulary measure is based upon word-frequency
Syntactic structure
norms derived from hearing childrens writing samples
(Hillerich, 1978). The writing elicitation task we Deaf childrens writing typically contains simple
designed needed to be validated in terms of the patterns sentences, showing limited use of compound and
of hearing childrens use of high-frequency words in complex sentences involving conjoined phrases or
English. However, a number of researchers have subordinated clauses (Paul, 1998). Deaf writers are
expressed concern over which norms (hearing vs. deaf more likely to make errors by adding unnecessary
reference samples) are appropriate for use with deaf words, omitting necessary words, and inappropriately
populations (see Maller, 2003, for a review of assessment substituting words. They have great difficulty with
issues). Given these concerns and our interest in verb inflections, use of plurals, and passive voice.
exploring the nature of deaf childrens vocabulary use Wilbur (1977, 2000) suggested that deaf writers have
(e.g., accessibility, selection, redundancy), we collected more syntactic errors on the intersentential level,
parallel data from hearing ESL speakers, a population rather than intrasentential level, in that adjacent
that also may have difficulty with English vocabulary. As sentences within longer passages of text do not always
we interpret our findings with deaf writers with connect to or have reference to each other.
differing ASL proficiency levels, it will be useful to
contextualize the results in terms of vocabulary use Lexical/semantic structure
patterns established by native and nonnative speakers of
English. For example, a pattern noted in high-ASL- The content of deaf childrens written texts has been
proficient deaf childrens writing may be characterized characterized as bland or stilted, having a limited
as unique, unlike the patterns of hearing native and vocabulary, relying heavily on same-item repetition,
nonnative speakers of English, or alternatively, the filled with content words (words that convey semantic
patterns may be found to mirror vocabulary use of information) and lacking in function words (words that
hearing ESL speakers, but not monolingual speakers. convey grammatical information; Gormley & Sara-
chan-Deily, 1987; Maxwell & Falick, 1992; Quigley &
Deaf Childrens Written English Skills Kretschmer, 1982; Quigley & Paul, 1984; Wilbur, 2000).
Deaf childrens access to the distributional fre-
Length of text
quency patterns of English vocabulary (either spoken
Deaf children tend to write fewer words than their or in print) is undoubtedly disadvantaged. Marschark
hearing peers on writing activities. For example, et al. (2002) proposed that frequency in print is
Deaf Writers Vocabulary Use 89

generally not a good indicator of vocabulary knowledge The judges could not sort the essays of these two
for deaf students. Function words, in particular, are groups in a reliable fashion.
highly frequent in English, but difficult to master for While it was not a writing task, Mayberry, Lock,
deaf children. Walter (1978) administered a cloze and Kazmi (2002) designed a grammatical judgment
procedure test to 199 deaf students and 277 hearing task involving the presentation of printed English
students between the ages of 10 and 14 to assess sentences to deaf native ASL signers and hearing ESL
English content-word (nouns, verbs, adjectives, and learners. They found that both groups performed
adverbs) vocabulary knowledge as a function of word- similarly. By contrast, a group of deaf adults who did
frequency level. Based on his findings, he asserted that not learn ASL until after age nine performed at chance
a 10-year-old deaf child reading a book with controlled level on the same grammatical judgment task. Charrow
vocabulary drawn from the first 2000 most frequently and Fletcher (1974) examined TOEFL and SAT test
used words in English could be expected to understand performance and found similarities between hearing
only about 30% of the words. This figure is con- ESL speakers and deaf children born to deaf parents
siderably lower than that of the hearing 10-year-olds in English structure and writing ability, but poorer
who demonstrated vocabulary knowledge at the 85% performance among the deaf students in vocabulary
level of mastery. and reading comprehension.
In contrast, in studies examining deaf students
higher-level semantic structure and cohesive use of
Examining the Relationship of ASL Proficiency
language, better performance has been noted as
to English Literacy Development
compared to their syntactic or lexical knowledge
(Gormley & Sarachan-Deily, 1987; Yoshinaga-Itano, Wilbur (2000) contends that if a deaf child has a firm
Snyder, & Mayberry, 1996a). For example, Yoshinaga- grasp of ASL when he or she enters the reading task,
Itano and Downey (1996) found that deaf children the situation becomes more similar to that of any
showed a significant increase in semantic structure learner of ESL. In contrast, Mayer and Wells (1996)
(inference, sequencing, and story grammar) between argue that in the case of the deaf student, ASL can
the ages of 7 and 18. While Marschark, Mouradian, develop the cognitive power that would support broad
and Halas (1994) found that deaf children demon- cognitive and conceptual transfers between ASL and
strated reduced grammatical complexity and less English. However . . . the possibility of any linguistic
figurative language than hearing peers, they did find transfer or interdependence is unlikely (p. 105). As
that deaf children used a coherent goal-action- ASL has no written form, Mayer and Wells maintain
outcome sequence in their written stories, similar to that there is no logical opportunity to transfer
hearing peers. literacy skills from their first language (ASL) to
literacy skills of their second language (English). While
the theoretical problems raised by Mayer and Wells
Deaf/ESL Comparisons
are compelling, the problem remains that the research
Only a few studies have compared the writing of deaf literature lacks empirical studies that investigate
students to that produced by hearing ESL learners (see whether particular aspects of first-language ASL
Paul, 1998, for a review), with most concluding that knowledge facilitate certain aspects of second-language
both groups show similar performance patterns in the English literacy.
omission of function words (e.g., prepositions, articles, Most previous studies examining deaf childrens
conjunctions) and the order of difficulty of acquisition development of English literacy have not made a clear
of syntactic structures. Langston and Maxwell (1988) distinction between ASL-proficient children and less-
asked a set of judges (all experienced in working with proficient signing deaf children, so a meta-analysis of
deaf students) to compare writing samples collected the literature may not even be feasible. Some empirical
from deaf students (ages 1519) and age-matched studies have identified the number of deaf children
hearing ESL students in an anonymous sorting task. born to deaf parents included in their sample, but
90 Journal of Deaf Studies and Deaf Education 9:1 Winter 2004

typically there are not enough of these children in the a straightforward process (Mayer & Wells, 1996;
sample to warrant separate analyses. Even large-scale Singleton et al., 1998; Supalla et al., 2001; Wilbur,
evaluations of American deaf childrens achievement 2000). According to Mayer and Wells (1996), the
performance (such as within the Stanford Achievement linguistic conditions facing deaf children do not match
Test, 9th edition) have not identified deaf-of-deaf the conditions assumed by Cummins linguistic in-
(and presumably ASL-proficient) as a distinct subgroup terdependence model (Cummins, 1991). Cummins has
(Traxler, 2000). Israelite, Ewoldt, and Hoffmeister proposed that first-language literacy skills can transfer
(1992) reviewed several studies that point to the to second-language literacy.
superior performance of deaf children born to deaf While several researchers have asserted that ASL
parents on English literacy measures; however, these knowledge brings about greater world knowledge,
conclusions have been criticized because of small broad conceptual transfers, and enhanced creative
samples in the studies cited and the fact that many of vocabulary, we are not aware of specific proposals
the studies are outdated and in some cases unpublished. characterizing the process of how ASL proficiency can
Based on these concerns, Marschark et al. (2002) and facilitate English vocabulary development, especially
Paul (1998) argue that the purported deaf of deaf given the lack of an ASL writing system.1 Neverthe-
advantage is a premature conclusion that is still often less, based on what we have learned from previous
cited in the research literature. studies, English function words would remain a chal-
However, with the advent of new instruments lenge for the ASL-proficient learner of English because
developed to assess ASL proficiency (see Singleton & the syntactic elements conveyed by function words
Supalla, 2003, for a review) researchers now have the (prepositions, articles, conjunctions) in English are
opportunity to classify deaf students on the basis of conveyed through different syntactic devices in ASL
individual ASL proficiency rather than parent hearing (grammatical facial expression, body posture, spatial
status. Indeed, a number of recent studies have found modifications to signs; Mayer & Wells, 1996; Paul,
that higher ASL proficiency (regardless of parent 1998; Wilbur, 2000). However, English content words
hearing status) is positively correlated with higher (e.g., nouns, verbs, and adjectives) may be more
performance on various measures (mostly reading accessible than function words to an ASL-proficient
comprehension) of English proficiency (Hoffmeister, signer because of the potentially higher number of
2000; Mayberry et al., 2002; Padden & Ramsey, 2000; sign-word equivalencies among content words.
Singleton, Supalla, Litchfield, & Schley, 1998; Strong In summary, deaf children, as a group, have been
& Prinz, 1997, 2000), especially among older deaf shown to perform significantly below their hearing
students. peers across many measures of written English skill
Other researchers have come to similar conclusions (see Albertini & Schley, 2003, for a review). In
and have advocated the advantages of interactive ASL particular, their vocabularies are limited, as is their
instruction to teach English literacy in the classroom knowledge of English syntactic and semantic rules.
(Israelite et al., 1992; Lane, Hoffmeister, & Bahan, What is not known is whether ASL proficiency may
1996; Marschark et al., 2002; Musselman, 2000; Paul, contribute to the successful development of certain
1998; Supalla, Wix, & McKee, 2001; Wilbur, 2000). English vocabulary forms. Furthermore, we need
According to Paul (1998), world knowledge acquired to understand better the comparative performance of
through ASL can help deaf children expand their hearing ESL learners and deaf English learners (of
conceptual framework and build a more creative varying ASL skills) on the types of words they acquire
vocabulary. Maxwell and Falick (1992) also reiterate and use and the breadth of their vocabularies. With
the role of world knowledge, communicative experi- this understanding, we can improve our theorizing
ences, and genre expectations on the development of and interpretation of potential transfer effects from
cohesiveness in writing. proficient ASL to English and explore the possible
Nevertheless, many researchers acknowledge that application of ESL pedagogical theory to ASL-based
mapping ASL proficiency to English literacy is not deaf-education contexts.
Deaf Writers Vocabulary Use 91

Aims of the Present Study overlap in ASL and English; i.e., both groups are
expected to be equally disadvantaged in function-word
In this study, we focused on several key aspects of low-,
development. One interesting issue was whether we
moderate-, and high-ASL-proficient deaf childrens
would observe any differences among the ASL groups
written English vocabulary. We examined (a) text
on particular English function words with common
length (total number of words written); (b) use of
ASL sign equivalents (e.g., ON, MY). While high-
words known to be highly frequent in childrens
ASL-proficient children would be expected to show
writing; (c) redundancy in their writing; and (d) use of
better performance on these forms (as compared
function words in English. Further, we compared the
to those English function words without ASL sign
written vocabulary performance of our deaf writers
equivalents), we were uncertain a priori whether low-
(who were at three ASL skill levels) to that of hearing ASL-skilled signers would have mastered enough ASL
monolingual speakers of English and hearing speakers to know the equivalent signsif they did, then we
of English as a second language. Based on the research would expect similar beneficial transfer on those forms.
reviewed thus far, we expected the following patterns to Finally, with a solid language base, we expected that
emerge in our deaf childrens writing samples when these high-ASL writers would take the writing stance
compared to hearing writers: of trying to tell a story, which would result
 Deaf children will write shorter stories than in a writing product with diverse propositions and
hearing children. vocabulary use (no repetitive, formulaic sentences).
 Deaf children will produce a lower proportion of With respect to comparisons between deaf and
function words than hearing children. hearing ESL learners, we expected to find differences
 Deaf children may not produce words from a most between these groups in function-word use. Consis-
frequent word (MFW) list at a proportion similar to tent with the prior research, deaf writers are expected
hearing children (as a result of the overrepresen- to include fewer function words and in their writing,
tation of function words on the list and/or as auditory access to these forms offers ESL learners
disadvantaged access to word-frequency patterns a distinct advantage over deaf learners in their
of English). acquisition and use. With the rationale that they have
 Deaf children will show greater redundancy in more restricted vocabularies in their second language
their writing (e.g., higher repetition of words or use and tend to use vocabulary words that they encounter
of stock phrases). with high frequency, we also expected ESL learners to
produce a higher proportion of words from the MFW
However, given that we classified our sample of list. However, the issue with deaf writers and high-
deaf writers by ASL proficiency level (low-, moderate-, frequency word lists is more complicated. Outcomes
and high-ASL skill categories), we also expected to in which they have a lower proportion of MFWs in
obtain a different pattern of results across our groups their writing may be because (a) they are disadvan-
of deaf writers. Our hypothesis was that on some taged in their access to the distributional frequency
measures, such as novel-word use and repetitiveness, patterns and have difficulty accessing highly frequent
the high-ASL students would show a pattern of per- vocabulary, or (b) they have more creative vocabularies
formance superior (i.e., greater off-list word use and are writing a relatively higher proportion of off-
and less repetitiveness) to low-ASL students. This list words (thus reducing their on-list proportion). By
expectation was based on the assumption that high- examining the deaf writers within their ASL pro-
ASL-proficient students have a broader vocabulary ficiency groups, and contextualizing their results
base in ASL and that this semantic knowledge assists within ESL and monolingual patterns, we may be
them in their acquisition and use of less common able to enhance our understanding of these compli-
words in English. No difference was expected between cated issues.
low- and high-ASL-proficient children in function- Given that ESL learners also have a solid native
word use, as function words have much less direct language base and would be trying to tell a story with
92 Journal of Deaf Studies and Deaf Education 9:1 Winter 2004

diverse propositions, we expected no significant differ- Table 1 Stem and leaf display of deaf students grade
ences between high-ASL-signers, ESL learners, and level by ASL-proficiency level and school environment
monolinguals on their proportion of unique words (our ASL-PA Residential Self-contained
measure of repetitiveness or redundancy). We expected proficiency school classroom
level environment environment
low-ASL signers to have a lower proportion of unique
Low ASL (n 20) 1 22222222
words (i.e., greater repetitiveness), consistent with
33 3333
prior research on deaf writers. 4 55
66
Method Moderate ASL (n 20) 1111
222
Participants 3333
5555
The participants in this study were divided into five
66666
groups for comparative analyses. Within each group
High ASL (n 32) 111
there was adequate representation of participants across
22222
grade levels, spanning first through sixth grade2 (see 333333
Tables 1, 2 for a stem and leaf display of the distribution 444444
of hearing and deaf students included in the study). We 555555
obtained parental consent for all participants included 666666

in the study.
going home on the bus instead of retelling the story
Deaf students with different levels of ASL proficiency. As presented in the elicitation task (the tortoise and hare
part of an earlier research project conducted by story), and the other student was an extreme outlier. As
Singleton and Supalla (1998), writing samples were a third grader, he wrote an essay with over 230 words,
collected from 74 deaf students with varying levels of almost double the average total words produced by
ASL proficiency. In order to obtain a diverse sample of hearing monolingual or ESL students, even at the
ASL skills, the deaf students were sampled from highest grade level. While this deaf student was indeed
a residential school program and a self-contained an interesting individual case study of outstanding
classroom in public school using Total Communication
writing ability, clearly he was not representative of any
(English-based signing co-produced with spoken
student in this study, deaf or hearing. Therefore, 72
English). Selection criteria for participant inclusion
deaf students were included in the final analyses.
in the original research project included severe-to-
Each deaf student was administered the American
profound hearing loss and no known disabilities that
Sign Language Proficiency Assessment (ASL-PA;
would interfere with learning. A Nonverbal IQ (Test of
Maller, Singleton, Supalla, & Wix, 1999) and assigned
Nonverbal Intelligence; TONI-2; Brown, Sherbenou,
a proficiency score and rating of high-, moderate-, or
& Johnsen, 1990) was administered to rule out global
low-ASL proficiency (high, n 5 32; moderate, n 5 20;
cognitive deficits. Scores for the deaf students tested
ranged from 84 to 143 and were normally distributed low, n 5 20). In previous analyses, Maller et al.
with an average score of 110. School officials reported determined that neither age nor grade correlated with
that no deaf students included in the sample came from ASL-proficiency score when used with a population
homes with extreme poverty conditions (which could between the ages of 6 and 12. In other words, a first
potentially interfere with learning aptitude). Forty- grader could be rated as high-ASL proficient just as
eight deaf students had hearing parents and 26 deaf likely as a sixth grader could be rated as low-ASL
students were born into families with deaf parents. proficient (as illustrated in Table 1). A high ASL-PA
Parent sign ability was not formally assessed. Two deaf score was assigned to students who demonstrated a full
students of hearing parents were dropped from the range of ASL grammatical features in their spontane-
original sample of 74. One student wrote an essay about ous signing (see Maller et al. 1999, for details regarding
Deaf Writers Vocabulary Use 93

Table 2 Stem and leaf display of hearing students grade public elementary schools in the Midwestern U.S. Half
level distribution of the students came from a school population that was
Hearing ESL 11111111111111111 predominantly middle to upper socio-economic class,
(n 60) 222222222222222 and the other half attended a school that was pre-
333333333 dominantly lower socio-economic class. Our purpose in
4444444
selecting students from these two schools was to ensure
55555555555
Hearing 1111111111111111 a range of student backgrounds and diversity in literacy
monolingual 22222222222 experiences. As the control or baseline group, the
(n 61) 3333333333333 selection had to capture the natural range of perfor-
444444444 mance levels with respect to written literacy. Struggling
555555555555
writers can be found at all grade levels among the
hearing monolingual population. All students were
identified by school staff members as being native,
ASL assessment procedures and psychometric prop-
monolingual speakers of English and having no known
erties of the ASL-PA instrument).
disabilities that would detract from their performance
on the writing task. No students were dropped from the
Hearing ESL speakers. Hearing English-as-a-second- original sample of 66 monolingual students.
language students (n 5 60) were selected from a Mid-
western U.S. elementary school with a special magnet Participant sample decisions. The sample of deaf
bilingual program that supports native language students was one of opportunity; these writing samples
education and literacy experience. The ESL students, were obtained from another research project data
who represent 14 different language backgrounds, were set (Singleton & Supalla, 1998). We knew the sample
selected according to the following criteria: they (a) were included signers of low-, moderate-, and high-ASL
learning ESL; (b) had lived and attended school in the proficiency. The hearing ESL and monolingual
United States for at least one year, but not more than samples were collected to mirror the distribution
two and one-half years; (c) had emergent-to-proficient in the deaf sample in order to establish baseline
literacy skills in their native language (according to their vocabulary information, necessary for interpreting the
teacher); (d) were expected by their teachers to be able potential differences among the deaf children of
to complete the writing task in both their native differing ASL proficiencies. Another design approach
language and in English; and (e) had no known language would have been to conduct a cognitive assessment on
or learning disability. We originally sampled 61 ESL all participants and match the deaf and hearing
students, yet one student was dropped from the analysis participants on cognitive performance rather than
grade or age as deaf children are notoriously behind
because he did not write any words in the writing
their grade level hearing peers in their academic per-
booklet used for this task. We included all 1st5th grade
formance. However, this alternate approach was not
ESL students who met the selection criteria. We did not
feasible for the present study for at least four reasons.
obtain family SES information for this group, and we
First, according to Maller (2003), there is considerable
are assuming some diversity in the population as the
debate about the appropriateness of particular assess-
students came from many different family income and
ment instruments when used with both deaf and
education backgrounds (the districtwide magnet pro-
hearing children for leveling purposes (i.e., a test
gram buses ESL students to the campus rather than
used with hearing children may underestimate a deaf
draw from the surrounding school neighborhood, and
childs cognitive skills). Second, Maller also questions
parents include Ph.D. students at a nearby university as
whether cognitive scores will reliably predict deaf
well as agricultural workers).
childrens performance on spoken-language measures.
Third, other issues regarding language accessibility
Hearing monolingual speakers of English. A group of and testing would arise for the cognitive assessment of
66 hearing students were randomly selected from two the hearing ESL speakers. Finally, the public schools
94 Journal of Deaf Studies and Deaf Education 9:1 Winter 2004

available to us were not amenable to us conducting Fourth- through sixth-grade students received six-
cognitive assessments of the hearing children. The page picture-prompt packets. Most students did not
school personnel and parents of the participants agreed simply describe the picture on the page; rather, the
only to us collecting a writing sample from the hearing samples indicated that they retold the story without
participants. being anchored to the picture prompt. This interpre-
tation is based on the students inclusion of written
Procedure information on the page that was not included in the
picture and also on the fact that childrens stories
Each student was shown a three-minute silent cartoon
flowed through multiple pages, with events extending
of the classic tale The Tortoise and the Hare. In this
past the picture that correlated with their text.
video, the characters do not have any verbal exchange,
Occasionally, students would ask for help with
but there is a considerable amount of posturing and
spelling a word. Because the purpose of this task was not
nonverbal signals exchanged between the tortoise and
centered on spelling accuracy, the experimenter was
the hare. The video begins with the start of the race,
allowed to spell or fingerspell a word upon a students
then the hare zooms ahead of the tortoise and, confident
request (but no more than twice at a normal pace).
that he is far ahead of the tortoise, the hare stops to
We did not want children to have their writing flow
take a nap under a tree. Meanwhile, the tortoise slowly
disrupted by a spelling challenge. If the child continued
catches up and passes the sleeping hare. The hare wakes
to struggle, we encouraged them either to skip the word
up and tries desperately to catch up and win the race,
and keep going, draw a picture, or make a good guess
but the tortoise stretches out his neck at the finish line
and move on. We wanted each child to feel positive
and wins the race. The hare reacts to his loss with shock
about the writing task. Children were told that they
and the crowd cheers for the tortoise.
could withdraw at any time and that this writing sample
The video served as the stimulus for the writing
would not be counted as part of their schoolwork or
activity for all participants. All students watched the
evaluated as such. There was no strict timeframe for the
video in groups of five to eight students and were asked
task administration, but after 45 minutes the children
to write the tortoise and the hare story without helping
were told to wrap it up as best they could. Most children
each other. The deaf students were asked by a deaf
finished within 20 to 30 minutes. No child withdrew
native ASL signer to retell The Tortoise and the
from the study and most seemed to enjoy the writing
Hare story in sign. Approximately one week later, the
activity. All children were given a colorful pencil for
deaf students viewed the videotape again in groups of 5
their participation in the project. Participating schools
to 8 and were asked by a hearing native ASL signer to
were given Visual Dictionaries in appreciation for their
retell the story in written English.
cooperation in the project.
In order to match the procedure that was used with
the deaf students, the ESL students were also asked to Coding Procedures
write the story twice: first they wrote it in their native
A straightforward total word measure was used for the
language and then one week later, after viewing the
present study. This is a simple count of every word
videotape again, they wrote the story in English, their
written, including misspelled but interpretable words
second language. The ESL speakers native language
(such as trtle for turtle). Note that uninterpretable
writing samples have not been translated or coded for
words were not counted in the total-word measure.
this project. The hearing monolinguals were only
Total word count serves as the denominator for the
administered the writing task once, following the same
proportional data reported in the vocabulary analyses.
video elicitation procedure described above.
The focus of the current study was an investigation
Writing booklet. First- through third-grade students into vocabulary use in the writing samples we collected
received four blank pages for writing their stories. Each from deaf and hearing students. The vocabulary
page included one still image taken from the video as analysis included the following measures: total words,
a picture prompt on the top half of the writing page. frequent words, unique words, and function words.
Deaf Writers Vocabulary Use 95

The first stage of coding the written samples was function-word category includes elements that convey
conducted by a team of five coders, all of whom were grammatical information such as pronouns, conjunc-
extensively trained. In a second coding process, 25% of tions, articles, auxiliaries, demonstratives, quantifiers,
the samples coded by this team were randomly selected and prepositions (Tartter, 1986). We also included ad-
to check for errors. Total-word, frequent-word, and verbs in this category and the words no, not, and yes.
unique-word scoring revealed very few coding errors. Previous research has suggested that deaf children have
Function-word scoring indicated a higher rate of great difficulty with function words in their English
coding errors. Two of the authors subsequently writing (Albertini & Schley, 2003). Because of the
reviewed all samples to correct all function-word overrepresentation of function words on Hillerichs
scoring. After this step, only a very small number of MFW list (approximately 65%), we also conducted an
disagreements emerged and were corrected before final analysis of the data that considers function-word cate-
analyses were conducted. gory membership as opposed to frequency list status.

Frequent words. We began with a 99-word list of


most frequently used words in childrens writing Results
(Hillerich, 1978, as cited in Luckner & Isaacson, 1990).
In the analysis of variance, language group was a fixed
We added 6 highly common words that were specific to
factor (low ASL, moderate ASL, high ASL, ESL,
this tortoise and hare task (race, rabbit, shhhh, sleep,
monolingual). The four dependent variables were
turtle, win). All writing samples were examined and
number of total words (TW), proportion of most
words that appeared on the MFW list were tallied (see
frequent words (MFW), proportion of unique words
Appendix A for the list of the 105 words). For each
(UW), and proportion of function words (FW). Grade
participant, we determined the proportion (out of total
level was entered as the covariate for this analysis. This
words written) of frequent words and nonfrequent
statistical procedure controls for the grade level of the
words (i.e., words that were off the MFW list).
child and permits us to investigate differences that may
be due to language group background rather than age.
Unique words. In order to quantify the redundancy For each vocabulary measure, the overall ANOVA was
or repetitiveness of vocabulary use, we counted the num- significant (TW, F 5 12.123, p , .0001; MFW, F 5
ber of unique words used (i.e., five repetitions of the same 5.532, p , .0001; UW, F 5 4.704, p , .01; FW, F 5
word still count as one in this analysis). Distinct 47.884, p , .0001). A summary of estimated marginal
conjugations of verb forms were counted as unique means for each vocabulary measure is presented in
words. Plural forms of nouns were not counted as distinct Table 3. Estimated marginal means reflect an adjust-
forms, rather they were counted as a multiple of the ment for the covariate (grade level). We first discuss the
singular form. The proportion is reported as the number results for each vocabulary measure, and subsequently
of unique words divided by the total number of words in provide a summary of the performance of each language
the sample. This proportion is also known as a type token group on this writing task. For all interpretations of
ratio. This analysis enabled us to identify children who statistical significance, we used an alpha-level of .05.
wrote a large number of words but had a small creative
vocabulary (i.e., those who used the same words
Total Words
repetitively). We found that all participants used the
same character-identifying labels hare and tortoise Pairwise comparisons of the estimated marginal means
(or more commonly, rabbit and turtle), repeatedly. Thus, (Bonferroni test) indicated that the hearing mono-
what we were looking for was redundancy beyond the lingual group (M 5 83.98) did not write significantly
expected use of story character names. more words than the ESL (M 5 83.33) group ( p 5
.914). The low-ASL group (M 5 55.11) did not write
Function words. We recognized that MFW lists typi- significantly more words than the moderate-ASL (M 5
cally contain a high number of function words. The 45.92) and high-ASL (M 5 44.93) groups. However,
96 Journal of Deaf Studies and Deaf Education 9:1 Winter 2004

Table 3 Summary of vocabulary measure performance by language group


Number Proportion

Group of total words of words on MFW list of unique words of function words of unique function words
Monolingual
(n 66) 83.98 .66 .57 .585 .28
ESL
(n 60) 83.33 .70 .52 .56 .255
Low ASL
(n 20) 55.11 .77 .46 .43 .17
Moderate ASL
(n 20) 45.92 .675 .57 .28 .17
High ASL
(n 32) 44.93 .65 .57 .29 .16
Note. All proportions are calculated with total word as the denominator. The estimated marginal means have been adjusted to take the covariate (grade
level) into account.

the three deaf groups wrote significantly fewer words linguals) that the writers have generated a higher
than the two hearing groups on the total word measure, proportion of creative off-list vocabulary. The compar-
a result consistent with previous research. ative pattern of the low ASL to high ASL reveals that
these two groups of deaf writers have very different
Most Frequent Words patterns with respect to their use of less common
English words. The striking fact that low ASL writers
In the monolingual group, 66% of their words were on
had the highest proportion of MFW-list vocabulary
the MFW list, which sets the baseline for use of highly
(and the lowest proportion of off-list creative vocabu-
frequent or common vocabulary. Therefore, 34% of
lary) of all groups supports the interpretation that high-
a native English speakers vocabulary would be inter-
ASL proficiency affords a different entry point into
preted as off the MFW list or an estimate of their
less-common English vocabulary acquisition.
creative vocabulary use (see Luckner & Isaacson, 1990, for
a similar interpretation). Thus, an MFW proportion
Unique Words
greater than 66% indicates a reliance on highly common
words and the use of less creative vocabulary. In- This measure involved counting a distinct word only
terestingly, the low-ASL group (77%) had the highest once (regardless of repetitions of the same word in the
proportion of words that appeared on the MFW list, passage), or a type token ratio. A high proportion
significantly higher than all other groups (all p values , on this measure was interpreted as an indicator of
.02). By contrast, the high-ASL (65%) and moderate- vocabulary breadth. A low proportion would indicate
ASL (67.5%) groups did not differ significantly from the high redundancy or formulaic writing. Of the mono-
monolingual group on this measure. The ESL group lingual groups vocabulary, 57% was unique (i.e., not
(70%), as we had predicted, had a significantly higher repeated again in the passage)we considered this
proportion of MFW than the monolingual group ( p , the target level of performance. The high-ASL (57%)
.02). and moderate-ASL (57%) groups obtained the same
It is interesting to note that the high-ASL group unique word proportion and did not differ significantly
produced more vocabulary that would be considered from the monolingual group. The high- and moderate-
less common, diverging away from the MFW list, ASL and monolingual groups performed significantly
than the ESL group and the low-ASL group. Recall our higher than the low-ASL group, which had the lowest
two earlier predictions, especially with respect to deaf proportion of unique words (46%). In sum, the low-
writers. A lower MFW proportion could mean dif- ASL group had the greatest redundancy (using many
ficulty mastering common core vocabulary or it could of the same words repeatedly) in their written text. The
mean (similar to the interpretation we afford mono- ESL group (52%) was very close to outperforming
Deaf Writers Vocabulary Use 97

the low-ASL group (46%), but did not reach the .05 signers. They may use more function words overall in
alpha level of significance ( p 5 .06) used for these their writing than high-ASL signers, but they are using
analyses. the same few function words repeatedly. They do not
demonstrate the same level of variability in function-
Function Words
word use that we observe in the writing samples of
Regardless of frequency list status, over half of the monolinguals and ESL speakers.
monolingual (58.5%) and ESL (56%) students total
vocabularies for this task could be characterized as Summary: Language Group Vocabulary Profiles
function words, conveying grammatical information in Hearing monolinguals. Monolingual speakers of Eng-
English. The monolingual and ESL groups did not lish had little difficulty performing this writing task.
differ significantly from each other on the function- While there was a range of performance across grades,
word proportion measure ( p 5 .189) and produced we controlled for grade level in all analyses. These data
significantly higher proportions than all three ASL serve as a baseline for target writing performance on
groups. The moderate-ASL (28%) and high-ASL the Tortoise and Hare story-retelling task. The
(29%) groups were not significantly different from each hearing monolingual students had an average of 83.98
other in their proportion of function words used in their total words. Out of this total, 66% were words found
writing samples. While the low-ASL (43%) group used on the MFW list, 57% of the words were unique, and
a significantly higher proportion of function words in 58% of the words were considered function words.
their writing than moderate- and high-ASL groups ( p Appendix B provides a representative writing sample
values ,.001), their proportion was still significantly of a 3rd-grade hearing monolingual student.
lower than the ESL and monolingual groups.
A more detailed examination of the relative use of ESL speakers. The ESL students (M 5 83.33)
common function words (i.e., those on the MFW List) performed similarly to the monolingual group (M 5
and novel function words (i.e., those off the MFW 83.98) in the total word measure and proportional use
List) by each group did not yield any support for of function words (56%). However, the ESL students
group differences. In general, between 76 and 87% of tended to use a higher proportion (70%) of frequent
the function words used by all students, regardless words (common words appearing on the MFW list)
of language group, came from the MFW List. One and demonstrated fewer (52%) unique words (i.e.,
interesting finding was that high-ASL-proficient (as were more redundant in their writing) than their
compared to low-ASL-proficient) signers had a signif- monolingual counterparts. In sum, the ESL students
icantly greater likelihood of using function words on the could be characterized as sticking with highly frequent
MFW list that had a common ASL sign equivalent (e.g., words and having a tendency to be more repetitive than
ON, AFTER)3. monolinguals. However, as a group, the ESL students
All three ASL groups were not significantly do appear to have adequate access to function words in
different from each other in their limited variation of English. Appendix B includes a representative writing
function words. The proportion of unique function sample from a 3rd grade hearing ESL student.
words for each group was as follows: low ASL (17%),
moderate ASL (17%), and high ASL (16%). The three Low-ASL group. The low-ASL group (M 5 55.11)
ASL groups had significantly fewer unique function wrote significantly fewer words than their hearing
words compared to the monolingual (28%) and ESL counterparts, but did not differ significantly from the
(26%) speakers (all pairwise comparisons had p values moderate- and high-ASL groups on the total word
less than .0001, and the two hearing groups were not measure. The low-ASL group used a significantly
significantly different from each other, p 5 .055). higher proportion of function words (43%) in their
In summary, these results enhance our understand- writing than the moderate-ASL (28%) and high-ASL
ing of the higher proportion of function-word use (29%) groups, which was most likely due to high
observed in the writing samples of low-ASL-proficient repetitiveness of a limited number of forms. Of all
98 Journal of Deaf Studies and Deaf Education 9:1 Winter 2004

groups, the low-ASL group demonstrated the highest their thoughts. While their vocabulary does not
proportion of words that could be found on the MFW necessarily reflect the distributional frequency patterns
list (77%). The low-ASL group also had the lowest level most common to hearing children (i.e., their writing
of uniqueness in their writing (46%), significantly tends not to include the high proportion of function
worse than all other groups except the ESL group, from words typical of hearing childrens writing), they
which it did not significantly differ. While both low- produce words that are novel and creative. This writing
ASL and ESL groups had a higher proportion of style is well illustrated in Appendix B by the
common words than the monolingual group, the key representative sample of a 3rd-grade high-ASL-pro-
difference is that the low-ASL group had a much higher ficient deaf student. The reader of this passage gets
proportion of common words and was significantly a clear sense that this student has retold the story in
more repetitive with a limited range of function words. proper chronological sequence and incorporates crea-
Based on this finding, we can argue that perhaps the tive vocabulary that characterizes the inner thoughts of
ESLwriters were more comfortable with using common the main characters. While there are noticeably few
vocabulary, but at least they included a broad range of
function words used in this story, this student attempts
these common forms in their writing. On the other
to use semantically rich vocabulary such as tiptoe, ignore,
hand, the low-ASL writers appeared more formulaic,
puzzle, and worry.
using a limited number of common forms with high
repetition in their text. Appendix B includes a repre-
Discussion
sentative writing sample by a 3rd-grade low-ASL-
proficient deaf student. The formulaic writing style This study compared three groups of deaf students
is well illustrated in this example. This student includes (low-, moderate-, and high-ASL proficient) and two
function words (such as I, am, can, will), but they are groups of hearing students (monolingual English
used repeatedly as stock phrases in a formulaic fashion. speakers and ESL speakers) on four vocabulary
measures as assessed through analysis of their writing
Moderate- and high-ASL groups. The moderate- and samples. The low-ASL group exhibited more of the
high-ASL groups were not significantly different from patterns that would have been expected based on
each other on any vocabulary measures in this study, previous research on deaf childrens writing and
thus the reported results below generally characterize vocabulary use. Their writing was redundant, that is,
both groups. On total words, the moderate- and high-
they used stock vocabulary in a repetitive fashion.
ASL groups performed at the same level as the low-
While we documented higher proportion of function-
ASL group, and all three deaf groups performed
word use, we found that the low-ASL-proficient
significantly below the monolingual and ESL groups. In
students were simply using a low number of function
terms of words on the MFW list, the moderate- and
words with high repetitiveness. We should also note
high-ASL groups were not significantly different from
that in this analysis, we were simply counting the
the hearing monolingual speakers and, interestingly,
presence of a function-word attempt. We have not
exhibited a different pattern from hearing ESL and the
low-ASL groups. The moderate- and high-ASL groups formally evaluated the accuracy of its usage.
demonstrated a higher proportion of unique words in It is important to consider the fact that a majority
their writing at a level that surpassed the low-ASL and of the low-ASL students in this sample attend public
hearing ESL groups. Moreover, they were not signif- school in a self-contained classroom for deaf students.
icantly different from monolingual speakers on this The predominant communication mode in the class-
unique-words measure. The moderate- and high-ASL room was Total Communication (English-based sign-
groups had the lowest proportion of function words in ing simultaneously presented with spoken English).
their writing, (significantly below low-ASL, ESL, and Thus, their daily classroom experience may have
monolingual groups). placed a greater focus on certain function words and
In summary, the moderate- and high-ASL groups this may explain the higher frequency and redundancy
used diverse and original vocabulary to communicate of such forms in their written stories. Alternatively,
Deaf Writers Vocabulary Use 99

a fragile linguistic system and tendency to be formulaic Research Implications for Deaf Education
may also account for this pattern we observed. Wilbur
How do these vocabulary findings translate into
(2000) suggested that teaching techniques used with
recommendations for educational practice? The pri-
deaf children artificially hold students to a stilted and
mary point that parents, educators, and language
choppy style of writing that is based on teaching them
specialists can take from this study is that deaf children
grammatical structures in isolation without giving
with moderate- to high-ASL skills are using English
them a clear understanding of communicative function
vocabulary in ways that appear different from the
in context.
patterns observed with profoundly deaf children with
Nevertheless, function words are used by the low-
low-ASL skills. In this study, high-ASL-proficient
ASL group at a frequency rate that is still far lower than
children appeared to tap into their broad semantic
the ESL students in this sample. This result, we
knowledge and generated novel vocabulary in their
maintain, highlights the inaccessibility of the distribu-
English writing, whereas low-ASL-proficient children
tional patterns of function words in a deaf childs
produced writing that reflected formulaic reproduction
everyday experience with English even when his or her
of a limited range of forms. For all deaf children, access
school context places a strong emphasis on speaking and
to the distributional frequencies of English vocabulary
signing English.
is most likely reduced, but proficiency in ASL may
The high-ASL groups performance on this writing
provide a new entry point into the learning and use of
task revealed a departure from the classic findings of
English vocabulary. This is an important point that is
previous research with deaf children. We found that
while these deaf writers still struggled with function perhaps overlooked in educational practice.
words and generally wrote fewer words overall than One pedagogical approach might be to develop an
their hearing counterparts, they demonstrated evidence MFW list for second-language learning of English that
that their expressive vocabularies in English are creative would be appropriate for deaf students fluent in ASL.
(nonformulaic) and diverse. The high-ASL group did With such a list, educators could use ASL to facilitate
not rely as much on MFWs; rather, they generated the development of easy-to-learn vocabulary for deaf
nonfrequent (off the MFW list) words at rates similar to students. Also, based on the idea that there may be
that of monolingual speakers. Moreover, they out- some advantage to learning English function words
performed ESL speakers on the proportion of non- that have an ASL sign equivalent, educators could ex-
frequent words in their writing. A preliminary analysis plore new ASL-based approaches to teaching English
of patterns of using English vocabulary which have ASL function words.
sign equivalents suggested some possible learning What needs further study is how ASL knowledge
advantage. Further analyses may reveal particular could potentially shape a deaf childs English vocabu-
ASL vocabulary that transfers well in terms of English lary learning process. Continued study of this ASL
vocabulary development. advantage for vocabulary breadth and diversity is
These findings suggest that in this writing task, deaf important for developing techniques and strategies
children who are highly proficient in ASL use English for connecting sign to print.
vocabulary in ways that are different from their low- This study also reinforces previous research which
ASL-proficient and hearing ESL peers. It appeared that suggests that deaf children exhibit difficulties with
when writing in English, the high-ASL students drew function words in English, and they also lack some of
upon their semantic understandings in ASL and the elaboration skills that we observe in hearing writers
generated propositions that included novel and meaning- (compare the level of elaboration in the high-ASL to
ful (although mostly content word) vocabulary. While the hearing monolingual passage in Appendix A).
function-word use and grammatical accuracy are still Educators may want to explore instructional strategies
pervasive problems, it is clear that these students are for English function words that acknowledge the deaf
communicating their ideas through text and not trying to childs experiential deprivation and barrier to in-
generate formulaic sentences as if filling in a workbook. cidental learning of these forms. Wilbur (2000) argues
100 Journal of Deaf Studies and Deaf Education 9:1 Winter 2004

that in deaf education, instruction of English structure children are indeed thinking and creating. Therefore, as
often focuses on isolated sentences and thus limits educators, the onus is upon us to harness those novel
a deaf childs input to connected discourse. ASL- thoughts that might be expressed so fluently in ASL and
proficient children are quite capable of generating develop instructional techniques that can connect this
elaborative stories and cohesive ties in ASL, thus creativity to their developing literacy skills in English.
future studies could explore ASL-based ways to
connect these known concepts into their English
Appendix A: List of 105 Most Frequent Words
writing to enhance both text length and elaboration.
Used for Coding Writing Samples
Finally, this study suggests that educators should
use some caution when considering ESL teaching Content Words
strategies with deaf children. Neither the low- nor
VERBS: come, came, get, got, go, going, went, make,
high-ASL groups exhibited vocabulary use patterns
play, put, said, saw, see, sleep, win, write
mirroring the ESL students, a result that is some-
NOUNS: day, friend, home, house, mother, rabbit,
what inconsistent with Paul (1998) and Langston and
race, school, time, turtle, one, two
Maxwells (1988) suggestion that the writing of deaf
ADJECTIVES: good, little, dear
students is similar to that of ESL speakers. It is clear
OTHER: shhh
that hearing gives one an undeniable advantage in
terms of exposure to the probabilistic patterns of
Function Words
vocabulary in English. The ESL students generated
a higher proportion of common words (from the MFW ARTICLES: a, an, the
list) and greater redundancy (fewer unique words) in PREPOSITIONS: about, after, as, at, back, by, down,
their writing as compared to monolingual speakers. for, from, in, of, on, out, over, up, to, with, like
However, their text length and proportion of function PRONOMINALS/DEMONSTRATIVES: he, she,
words used was similar to monolinguals. Clearly, these me, you, us, we, they, I, it, his, her, him, my, your,
ESL children exhibited some weaknesses in their their, our, them, this, that
writing, yet their high-proportioned use of function CONJUNCTIONS/CONNECTIVES: and, because,
words or attempts clearly demonstrate an important but, if, or, then, so, well
contrast with deaf writers. Again, this is critical when ADVERBIALS: here, how, just, now, there, very,
we compare teaching strategies traditionally used with what, when
deaf students to ESL teaching strategies, which may AUXILIARIES: can, could, will, would
be based upon different assumptions about access to COPULA: (TO BE): am, are, is, was, were
function words. Our intention is not to devalue some VERB AUXILIARIES: had, has, have, do, did
of the potential strengths of ESL pedagogy in deaf QUANTIFIERS: all, some
education; we only seek to illustrate that deaf learners NEGATION: not
(especially those with high- vs. low-ASL skill) are
different than hearing ESL learners and that appro- Notes:
priate adaptations should be investigated before All forms of the verb have and do were counted as
teachers of the deaf adopt an ESL approach. function words because they are not simple verbs; in
In closing, we wish to emphasize the importance other words, they also can function as a verbal auxiliary
of writing stories that have something to say. Deaf form (She has learned how to play soccer; He did
students who generate repetitive and formulaic senten- not know how to play soccer).
ces are not demonstrating that they are true writers. Six words (in italics) were added to Hillerichs
While the ASL-proficient students lacked important original list of 99 words. These words were related to
grammatical elements in their stories, their writing the story and were highly frequent in the writing
demonstrated original and creative expression. These samples.
Deaf Writers Vocabulary Use 101

Appendix B: Representative Writing Samples from 3rd-grade Students in Each Group


Deaf, Low ASL Deaf, High ASL
I rabbit Turtle and Rabbit Race Try
I turle Who win turtle
I can fast rabbit Rabbit sleep tiptoe Turtle and Wake Rabbit
I can fast win turle Miss Race Laugh
I can win chase Turtle walk slow
I can win chase rabbit Rabbit Sleep
I can win chase turle Turtle Puzzle
I am can win bed Worry
I am can win rabbit bed Igore
I am can win turle bed elft turtle
I am can win under rabbit bed yes Not Fair Win
I am can win rabbit and turle bed yes Told Wirn
I am can win fight Smat Learn
I am can will win rabbit turle

Hearing ESL Hearing Monolingual


One day rabbit and turtle was race. The rabbit and the turtle were at the starting line.
The rabbit can run fast then turtle. After that they were running.
The rabbit think that turtle is far away from rabbit. Rabbit was far away from the turtle.
So rabbit sleepy. So the rabbit went to sleep next to a tree.
Soon turtle was came to place that rabbit sleep. While the rabbit was sleeping the turtle walked
Turtle was going very quite. past the rabbit.
Soon turtle came to finish line. After that the rabbit woke up and the turtle was
Rabbit was late than turtle. near the finishing line so rabbit was running again.
Rabbit fall down. The rabbit was close to the finishing line. The turtle
When the rabbit was trying to stand the stretched his neck out and touched the finishing line.
tutle was on the peoples hand. The rabbit was surprised that had won
The turtle was happy.
Rabbit was surprise and angry. The end
Note. Misspellings are as they appeared in original writing sample.

Notes 2. The deaf students attended grades first through sixth,


but the hearing student sample only went up to fifth grade
1. One exception is that proposed by Supalla, Wix, &
students. This discrepancy was due to the fact that we were
McKee (2001) who have implemented an instructional in-
tervention that involves deaf students learning and use of an unable to collect writing samples from hearing middle school
intermediary coding system for ASL writing, subsequently students (6th grade).
making a conceptual link to an English gloss writing system, 3. A native signer of ASL reviewed the MFW list and
and then developing the English print form of the word. In determined whether there was a common ASL equivalent sign
a sense, it is a bridging strategy to permit the child to write for each of the function words. Signs invented within Manually
signs, then use this form to look up the English form in an Coded English were not counted as equivalent. Only the
ASL/English dictionary that they helped to create. Eventually, uninflected English form (e.g., COME) was counted, forms
the sign-writing is no longer needed once the children develop with inflections were not counted as having a common ASL sign
a large print-English vocabulary. The effectiveness of this equivalent (e.g., CAME). In general, the determination was
intervention is still being investigated. a highly conservative estimate of word-sign equivalents.
102 Journal of Deaf Studies and Deaf Education 9:1 Winter 2004

References Maller, S. J. (2003). Intellectual assessment of deaf people: A


critical review of core concepts and issues. In M. Marschark
Albertini, J. A., & Schley, S. (2003). Writing: Characteristics, & P. Spencer (Eds.), The Oxford handbook of deaf studies,
instruction, and assessment. In M. Marschark & P. Spencer language, and education (pp. 451463). New York: Oxford
(Eds.), The Oxford handbook of deaf studies, language, and University Press.
education. New York: Oxford University Press. Maller, S. J., Singleton, J. L., Supalla, S. J., & Wix, T. (1999).
Brown, L., Sherbenou, R. J., & Johnsen, S. K. (1990). TONI-2: The development and psychometric properties of the
Test of Nonverbal Intelligence, 2nd ed. Austin, TX: Pro-Ed. American Sign Language Proficiency Assessment (ASL-
Charrow, V., & Fletcher, J. (1974). English as a second language PA). Journal of Deaf Studies & Deaf Education, 4, 249269.
of deaf children. Developmental Psychology, 10, 463470. Marschark, M. (1997). Raising and educating a deaf child. New
Cummins, J. (1991). Interdependence in first- and second- York: Oxford University Press.
language proficiency in bilingual children. In E. Bialystok Marschark, M., Lang, H. G., & Albertini, J. A. (2002). Educating
(Ed.), Language processing in bilingual children. Cam- deaf students: From research to practice. New York: Oxford
bridge: Cambridge University Press. University Press.
Dale, E., & Chall, J. S. (1948). A formula for predicting Marschark, M., Mouradian, V., & Halas, M. (1994). Discourse
readability. Educational Research Bulletin, 27, 1120. rules in the language processing of deaf and hearing
DeVilliers, J., & Pomerantz, S. (1992). Hearing-impaired children. Journal of Experimental Child Psychology, 57, 89
students learning new words from written context. Applied 107.
Psycholinguistics, 13, 409431. Maxwell, M., & Falick, T. G. (1992). Cohesion and quality in
Gormley, K., & Sarachan-Deily, A. B. (1987). Evaluating deaf and hearing childrens written English. Sign Language
hearing-impaired students writing: A practical approach. Studies, 77, 345372.
Volta Review, 89, 157170. Mayberry, R., Lock, E., & Kazmi, H. (2002). Linguistic ability
Hillerich, R. L. (1978). A writing vocabulary of elementary and early language exposure. Nature, 417, 38.
children. Springfield, IL: Charles C. Thomas Publishers. Mayer, C., & Wells, G. (1996). Can the linguistic interdepen-
Hoffmeister, R. (2000). A piece of the puzzle: ASL and reading dence theory support a bilingual-bicultural model of literacy
comprehension in Deaf children. In C. Chamberlain, J. P. education for deaf students? Journal of Deaf Studies & Deaf
Morford, & R. Mayberry (Eds.), Language acquisition by eye Education, 1, 93107.
(pp. 143163). Mahwah, NJ: Lawrence Erlbaum. Musselman, C. (2000). How do children who cant hear learn to
Israelite, N., Ewoldt, C., & Hoffmeister, B. (1992). Bilingual/ read an alphabetic script? A review of the literature on
bicultural education for deaf and hard of hearing students: A reading and deafness. Journal of Deaf Studies & Deaf
review of the literature on the effective use of native sign Education, 5, 931.
language on the acquisition of a majority language by hearing Nagy, W., & Herman, P. (1987). Breadth and depth of vocabulary
impaired students. Toronto, Canada: MGS Publication knowledge: Implications for acquisition and instruction. In
Services. M. McKeown & M. Curtis (Eds.), The nature of vocabulary
Jenkins, J. R., & Dixon, R. (1983). Vocabulary learning. acquisition (pp. 1936). Mahwah, NJ: Lawrence Erlbaum.
Contemporary Educational Psychology, 8, 237260. Nagy, W., Herman, P., & Anderson, R. (1985). Learning words
Karchmer, M. A., & Mitchell, R. E. (2003). Demographic and from context. Reading Research Quarterly, 19, 304330.
achievement characteristics of deaf and hard-of-hearing Nation, I. S. P. (1990). Teaching and learning vocabulary. New
students. In M. Marschark & P. Spencer (Eds.), The Oxford York: Newbury House Publishers.
handbook of deaf studies, language, and education (pp. 2137). Padden, C., & Ramsey, C. (2000). American Sign Language and
New York: Oxford University Press. reading ability in deaf children. In C. Chamberlain, J. P.
King, C. M., & Quigley, S. P. (1985). Reading and deafness. San Morford, & R. Mayberry (Eds.), Language acquisition by eye
Diego, CA: College Hill Press. (pp. 165189). Mahwah, NJ: Lawrence Erlbaum.
Kucera, H., & Frances, W. N. (1967). A computational analysis of Paul, P. (1998). Literacy and deafness: The development of reading,
present-day American English. Providence, RI: Brown writing and literate thought. Boston: Allyn and Bacon.
University Press. Paul, P., & Gustafson, G. (1991). Hearing-impaired students
Lane, H., Hoffmeister, R., & Bahan, B. (1996). A journey into the comprehension of high-frequency multi-meaning words.
Deaf world. San Diego, CA: Dawn Sign Press. Remedial and Special Education, 12(4), 5262.
Langston, C. A., & Maxwell, M. M. (1988). Holistic judgment of Quigley, S., & Kretschmer, R. (1982). The education of deaf
texts by deaf and ESL students. Sign Language Studies, 60, children: Issues, theory, and practice. Austin, TX: Pro-Ed.
295312. Quigley, S., & Paul, P. (1984). Language and deafness. San Diego,
LaSasso, C., & Davey, B. (1987). The relationship between CA: College Hill Press.
lexical knowledge and reading comprehension for prelin- Silverman-Dresner, T., & Guilfoyle, G. R. (1972). Lexington
gually, profoundly hearing-impaired students. Volta Review, School for the Deaf Education Series: Vocabulary norms for
89, 211220. deaf children (Vol. 7). Washington, DC: Alexander Graham
Luckner, J. L., & Isaacson, S. (1990). A method of assessing the Bell Association.
written language of hearing-impaired students. Journal of Singleton, J., & Supalla, S. (1998). The effects of ASL fluency
Communication Disorders, 23(3), 219233. upon deaf childrens cognitive, linguistic, and social de-
Deaf Writers Vocabulary Use 103

velopment. Unpublished technical report. University of deaf and hard-of-hearing students. Journal of Deaf Studies
Illinois, Urbana-Champain. J. Singleton. & Deaf Education, 5, 337348.
Singleton, J., & Supalla, S. (2003). Assessing childrens Walter, G. (1978). Lexical abilities of hearing and hearing-
proficiency in natural signed languages. In M. Marschark impaired children. American Annals of the Deaf, 123, 976
& P. Spencer (Eds.), The Oxford handbook of deaf studies, 982.
language, and education. New York: Oxford University West, M. (1953). A general service list of English words. London:
Press. Longman, Green & Company.
Singleton, J. L., Supalla, S., Litchfield, S., & Schley, S. (1998). Wilbur, R. B. (1977). An explanation of deaf childrens difficulty
From sign to word: Considering modality constraints in with certain syntactic structures in English. Volta Review,
ASL/English bilingual education. Topics in Language 79, 8592.
Disorders, 18(4), 1629.
Wilbur, R. B. (2000). The use of ASL to support the
Strong, M., & Prinz, P. (1997). A study of the relationship
development of English and literacy. Journal of Deaf Studies
between American Sign Language and English literacy.
& Deaf Education, 5, 81104.
Journal of Deaf Studies & Deaf Education, 2, 3746.
Yoshinaga-Itano, C., & Downey, D. (1996). The effect of hearing
Strong, M., & Prinz, P. (2000). Is American Sign Language skill
loss on the development of metacognitive strategies in
related to English literacy? In C. Chamberlain, J. P.
written language. Volta Review, 98, 97143.
Morford, & R. Mayberry (Eds.), Language acquisition by
Yoshinaga-Itano, C., Snyder, L., & Mayberry, R. (1996a). Can
eye (pp. 131141). Mahwah, NJ: Lawrence Erlbaum.
Supalla, S., Wix, T., & McKee, C. (2001). Print as a primary lexical/semantic skills differentiate deaf or hard-of-hearing
source of English for deaf learners. In J. Nicol (Ed.), One readers and nonreaders? Volta Review, 98, 3961.
mind, two languages: Bilingual language processing (pp. 177 Yoshinaga-Itano, C., Snyder, L., & Mayberry, R. (1996b). How
190). Malden, MA: Blackwell Publishers. deaf and normally hearing students convey meaning within
Tartter, V. C. (1986). Language processes. New York: Holt, and between written sentences. Volta Review, 98, 938.
Rinehart & Winston.
Traxler, C. B. (2000). The Stanford Achievement Test, 9th Received October 2, 2002; revisions received August 5, 2003;
Edition: National norming and performance standards for accepted August 6, 2003