Center for the Study of Communication & the Deaf, Boston University Metalinguistic skills and vocabulary knowledge in ASL synonyms and antonyms in Deaf children Robert J. Hoffmeister, Phd & Rebecca McVey, EdM Research Question: Are there differences in metalinguistic skills in the error patterns of Deaf Children with Deaf Parents (DCDP) versus Deaf Children with Hearing Parents (DCHP)?

Methodology

Subjects: Results Ages

What This Study Hoped to Explore 4-5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 Total DCHP 9 17 10 12 25 20 8 18 20 12 26 15 26 22 19 259 Vocabulary knowledge is one window into furthering our understanding of metalinguistic DCDP 3 3 1 6 4 5 3 4 6 10 7 5 5 6 2 70 Synonyms awareness in language knowledge. Like any natural language that serves communicative needs, ASL Total 12 20 11 18 29 25 11 22 26 22 33 20 31 28 21 329 As the students increased in age, their knowledge of ASL synonyms improved. However, the average score for children has many synonyms and antonyms. To fully learn these forms, Deaf children require interaction with ! 329 Deaf students between 4 and 19 years of age. with hearing parents (DCHP) never passed the score achieved by even the 7-8 year old children with Deaf ! 70 DCDP ( 22%) and 359 DCHP (78%) fluent sign language users and the Deaf community, just as the acquisition of fluency in any language ! All subjects enrolled at two bilingual/bicultural schools for the parents (DCDP). While overall, subjects selected semantic foils most frequently, DCHP, particularly those below the requires interaction with its community of users. The task described in this study has been used to Deaf in the eastern United States. ! The median chronological age (CA) was 12.7. The subjects were median age of 12.7, were more likely to select foils that were both phonological and semantic in nature. Older DCDP examine the development of synonyms and antonyms in ASL in over 300 Deaf children at two divided into “younger” and “older” groups based on this median were likely to select errors that were phonologically related to the correct response, while older DCHP were age for the statistical analysis. bilingual-bicultural schools for the Deaf in the eastern United States. more likely to select nonsense foils. In fact, the older DCDP group never selected the nonsense foils. Synonyms Average Correct By Age Category ’ Testing Procedures: There is very little information available on the Deaf children s ASL vocabulary development and 15 ! Part of the the American Sign Language Assessment Instrument 14 13 metalinguisitic knowledge compared to research on children’s spoken English development. An (ASLAI), (Hoffmeister, Greenwald, Bahan, & Cole, 1989) 12 11 10 ! Receptive metalinguistic judgment task for ASL vocabulary 9 understanding of synonym and antonym knowledge in Deaf children has the potential to provide us 8 DCDP ! Synonym task has 15 multiple-choice questions (4 items) 7 DCHP 6 ! Antonym task has 14 multiple-choice questions (4 items) Number Correct 5 4 with greater insight into both extent and depth of their vocabulary knowledge and the related 3 ! For each question, students saw the stimulus item on video, then a 2 1 underlying metalinguisitc skills. The goal of this study is to examine metalinguistic judgments of fade, followed by a sequence of four response choices. 0 4-6 7-8 9-10 11-12 13-14 15-16 17-19

! Subjects must choose the item that best reflects the Age Categories synonyms and antonyms and the error patterns Deaf children make in ASL in order to determine synonym/antonym to the stimulus. vocabulary knowledge. ! An example of a question from the response booklet (used by children 8 and under) is below: Antonyms The antonyms task appeared to be more difficult than the synonyms task for both DCDP and DCHP, contrary to data Hearing Children’s Knowledge of Synonyms/Antonyms on hearing subjects (Stahl, 1999). As expected, the DCDP consistently outperformed the DCHP and, as with the Research confirms that synonym and antonym development is a complex metalinguistic task that synonyms task, the DCHP never had an average score higher than the average score of the 7-8 year old DCDP. In this requires understanding of the nuances of a language (Stahl, 1999; Charles, Reed & Derryberry,1994). test, the most frequent error type was semantic, followed by the foils that are both semantically and Yet, it is a skill most children with access to language generally develop naturally. It is related to phonologically related, as with the synonyms task. It seemed that when subjects were not sure of the answer they were metalinguistic ability and possibly theory of mind (Doherty and Perner, 1998). more prone to find something semantically related to the prompt, despite the fact that it often bordered on synonymy. Vocabulary and Deaf Children Antonym Average Correct by Age Category 14 13 Deaf children have been found to have typical conceptual development and often excel at visual- 12 11 10 9 8 spatial tasks compared to hearing peers. Many studies examining metacognitive or metalinguistic DCDP 7 6 DCHP 5

Number Correct skills in Deaf students have used English tasks, leading to erroneous conclusions about general 4 3 2 1 language knowledge. Research has demonstrated a link between proficiency in ASL (L1) and 0 4-6 7-8 9-10 11-12 13-14 15-16 17-19 English reading (L2) skills in Deaf children (Hoffmeister, 2000, 1995; Strong & Prinz, 1997). Age Categories This evidence supports Cummins’s Common Underlying Proficiency Hypothesis (CUP) which Overall Error Analysis states that interdependence of concepts, skills, and linguistic knowledge make transfer possible between one’s L1 and L2 (Cummins 1979, 2000). It has been shown that students with stronger In general across both of the synonym and antonym tasks, there was a significant Age X Parental Status ASL skills are also stronger writers (Singleton, Morgan, DiGello, Wiles, & Rivers, 2004), and Coding/Analysis: interaction (F=5.5:df=1/325:p<.02). This reflects that younger deaf children with deaf parents performed well on these vocabulary tasks, particularly when compared to their peers who have hearing parents. better readers (Mayberry, Waters, & Chamberlain, 2001). Mayberry et al. (2001) suggest that the 0=correct response SYNONYMS Younger Older Overall bilingual Deaf students are more prepared to encounter new linguistic structures in their second 1=phonological to prompt DCHP 44.77% 55.61% 53.54% language than those students who do not have a firm linguistic foundation in ASL. 2=phonological to correct response DCHP 65.64% 87.11% 75.66% 3=semantic ANTONYMS Younger Older Overall 4=both phonological and semantic Metalinguistic Skills in Deaf Children DCHP 39.70% 61.61% 47.08% 5=nonsense foil Deaf children are often tested using language tasks that have not been developed using natural DCDP 55.51% 71.82% 66.37% properties of signed languages, which have led to erroneous conclusions about their language 6= no response (x) or multiple

knowledge because these tasks do not test the children’s actual linguistic capabilities. Due to these The most important difference between groups is simply that DCDP subjects made less errors than their DCHP poorly designed tasks, previous studies on synonymy in Deaf children have falsely concluded that Each response frame was chosen to either represent the correct answer or act as a phonological, semantic or nonsense foil. peers, even when comparing younger DCDP to older DCHP on the synonyms test. DCDP were also less likely to Deaf children are delayed in their metalinguistic abilities and have have “impoverished select nonsense foils than their DCHP peers indicating that even when making errors, children of Deaf ” ” semantic representations which resulted from “impoverished educational experiences Using the numeric codes, the data accumulated from the 329 subjects was parents were using metalinguistic clues to guide their selection process while the children of Hearing that led to a lack of word knowledge (Strassman, Kretschmer, & Billsky, 1987, cited in Davey & analyzed to determine trends in patterns of incorrect response across the parents were not. Kind, 1990, p.229; Borman, Stoefen-Fisher, Taylor, Draper, Niederklein, 1988). These beliefs, based test based on subgroups related to age and parental hearing status. on English word acquisition, were erroneously extended to Deaf children’s general language knowledge. For too long Deaf children have been labeled as not having vocabulary or metalinguisitic knowledge and these findings have been used as a factor to support the concept of cognitive and language delay. Conclusion

In two multiple-choice tests of ASL vocabulary knowledge, errors that reflected semantically related choices were more common than foils that were phonologically related. Error analysis revealed that subjects were drawn to semantically related choices that reflected shades of meaning or metaphorical uses of the language. These semantic errors were more common than foil choices that were phonologically related in handshape, location and movement. However, Deaf children of Deaf parents did use the metalinguistic similarities in the phonologically related foils more than Deaf children of hearing parents, who were more likely to pick nonsense foils than their DCDP peers. Although DCDP demonstrated a stronger command of complex ASL vocabulary, the Special Thanks DCHP demonstrated demonstrated that they are able to rely on their knowledge of ASL to make decisions about the structure and content of ASL. DCHP were more variable in their application of metalinguistic strategies but tended to make errors based on phonology. This result suggests that many DCHP continue to approach the analysis of their L1 via a surface or phonological decision process

suggesting that more time is taken up on surface matters (analysis level proposed by Bialysok, 1991) which delays their access to the the deeper semantic information necessary to achieve fluency (control We would like to acknowledge the work and support of the many colleagues too many to name here, processes, Bialystok, 1991). who have contributed to our work:. The synonyms and antonyms task results reveal clear acquisition effects. As the students increased in age, the more their ASL scores improved. It is important to note that the DCHP never In addition, we would like to thank the students and staff at The Learning Center for Deaf Children, perform as well on this task as children with Deaf parents do at a young age. This confirms that Deaf children do have metalinguistic knowledge comparable to that of hearing children in the Framingham, MA, and the Scranton State School for the Deaf, Scranton, PA area of synonyms and antonyms when given access to a natural Signed Language from birth. The finding that DCDP consistently outperform compared to DCHP has been noted in many other areas of academic achievement including studies that have linked ASL ability to reading vocabulary and comprehension levels.. This continued lag in ASL knowledge supports findings by Mayberry & colleagues for their participation and collaboration. (2003, 2007) that a high percentage of DCHP do not reach proficiency in their L1 which will impact their acquiring an L2. Early signed language input needs to be investigated further in DCHP. We especially thank the Deaf students who have showed us that they are resilient, knowledgeable, The current study confirms that Deaf children are able to obtain metalinguistic knowledge comparable to that of hearing children in the area of synonyms and antonyms when provided and forthcoming. They have taught us a great deal about language learning and its impact on early access to ASL. Results challenge the claims that Deaf children have delayed vocabulary knowledge (Davey & Kind, 1990), or have deficient metalinguistic skills (Borman et al., 1988). achievement.