Professional Documents
Culture Documents
Ó The Author 2006. Published by Oxford University Press. All rights reserved. doi:10.1093/deafed/enj032
For Permissions, please email: journals.permissions@oxfordjournals.org Advance Access publication on February 9, 2006
Deaf Children’s Interactions 239
delay in developing sign language proficiency. In However, we do not have substantial data on the
some cases they do not attain the level in sign lan- language development of deaf children for both SLN
guage of deaf children from deaf parents (Marschark and spoken Dutch. For the Dutch situation, only the
et al., 2002; Singleton, Supalla, Litchfield, & Schley, longitudinal study by Van den Bogaerde (2000) on
1998). very young deaf and hearing children (1.0–3.0 years)
Critical factors in the early acquisition of spoken of deaf mothers is relevant. Over a period of 2 years,
language by deaf children are age at first intervention, she followed six children of deaf mothers. Three of the
early speech-reading ability, and possible thresholds children were deaf and 3 were hearing. The study by
of unaided hearing (Dodd, McIntosh, & Woodhouse, Van den Bogaerde provides data on the lexical richness
1998; Gilbertson & Kamhi, 1995; Limbrick, and syntactic complexity of Dutch deaf children’s
Dutch language area, apart from some general Van den Bogaerde has compared her results with
observations as can be found in, for example, Knoors the results of case studies in other sign languages. The
et al. (2003), there is hardly any research on this results of the studies of Hoffmeister (1978), Kantor
topic. (1994), and Richmond-Welty and Siple (1999) show
With respect to linguistic complexity, Lederberg similar developmental patterns of lexical and syntacti-
(2003) found that some deaf children of hearing pa- cal complexity in this age group.
rents show lexicons that are comparable in size to With respect to language dominance, we know
those of hearing peers, whereas others remain in the from studies on hearing bilingual children that, at an
slow word-learning phase for quite some time. They early age, children almost exclusively follow the strat-
slowly add new spoken words to their vocabulary and egies or choices their parents and/or teachers implic-
natural and accessible language for deaf children, we 1. hearing loss at least 80 dB on the best ear,
had the following expectations. We expected ‘‘linguis- 2. prelingual hearing loss (i.e., before the age of 3),
tic complexity’’ to increase over time—in the early 3. normal performance intelligence,
years more for signs than for spoken words—and to 4. no other severe disabilities.
be different in conversations with the deaf versus
At the start of the investigation in September
the hearing adults. We expected young children to
1998, 4 of these 6 children were placed in a group
be dominant in signs and to use significantly more
for 4-year-olds (pupils A1–A4) and 2 in a group for
signs with the deaf than with the hearing adults.
3-year-olds (pupils B1 and B2). The 6 children be-
We expected a clear pattern of ‘‘language accommoda-
longed to the first generation of pupils at Viataal to
tion’’ to emerge over time (‘‘one person one language
Method
Measures
Participants
The language development data were collected by
Six children participated in the study for 3 consecutive conversations between children and adults. Twice a
years. All parents agreed to the inclusion of their child year, semistructured conversations between the indi-
in the investigation. School professionals carried out vidual children and a deaf adult were organized,
the selection procedure using the following criteria: followed by conversations with a hearing adult. The
interactions were recorded on video. Each conversa- the longest utterances is regarded as a more valid
tion was filmed for 20–30 min. indicator of the syntactic competence of a person.
From each video-recorded conversation, deaf and Vermeer (1986), for instance, opted for the MLU10;
hearing researchers selected a 5-min section repre- the Childes manual mentions the MLU5 and MLU50
senting the most animated conversational fragment. (MacWhinney, 1991). In calculating the MLU10, we
This section was saved on CD-ROM and then linguis- distinguished three types of utterances: (a) signs only,
tically transcribed in full. The transcription system (b) mixed utterances, and (c) spoken words only. We
was partly derived from MacWhinney (1991; the excluded nonlinguistic utterances. A deaf and a hear-
Childes manual) and partly from Johnson and Rash, ing researcher coded all utterances of all transcripts
and Aarssen (working documents applying Childes independently. Their interrater reliability was perfect
effects for three independent variables (significance Right after the recording of the exchanges, the
level is .05). The independent variables were language adults—who knew the children well—were asked to
used (Dutch vs. SLN), time (development over 3 indicate the representativeness of the conversation in
years), and language mode (deaf vs. hearing adult in- a written questionnaire. All conversations were judged
teraction partner). In the analyses, the two recordings to be representative of the communicative behavior of
per year with the respective adults were taken together the participating children.
in order to obtain more stable measures as well as to At the start of the project, when the children were
avoid missing observations. Occasionally, a recording younger, the conversations between the children and
was lacking, for practical reasons (illness of the child, adults were mainly about the materials available in the
for instance). room. As the children grew older, they started to tell
level of proficiency in sign language is not delayed. interaction effects were observed: language mode by
However, this interpretation is tentative, as—although language and language mode by time.
the parents of all six children have tried to organize The time effect means that, in the course of 3
consistent exposure to SLN for their young deaf years, the children’s lexical richness increased signifi-
child—only child A2 is a native signer from a deaf cantly, F(2, 10) ¼ 14.952, p ¼ .008. The absence of
family. a main effect for language used signifies that there was
no difference in the Guiraud index between spoken
Linguistic Complexity Dutch and SLN. After 3 years, the indices for the
two languages met more closely than at the start of
This subsection presents the results for both lexical the investigation. There was a main effect for language
richness and MLU. mode, F(1, 5) ¼ 20.537, p ¼ .006, which implies that
the context in which the conversation took place, that
Lexical richness. The results of lexical richness are is, communicating with a deaf versus a hearing adult,
given in Table 3. Figure 1 gives a graphical represen- did indeed make a difference. The two significant
tation of the mean scores. interaction effects add relevant information. The
The data were analyzed for possible effects of lan- language mode by language used effect, F(1, 5) ¼
guage used (SLN vs. spoken Dutch), time (develop- 17.944, p ¼ .008, points to the lower scores for spoken
ment over 3 years), and language mode (deaf vs. Dutch in interaction with the deaf adult. Figure 1
hearing adults). Two main effects were found for shows that over all 3 years the Guiraud scores are
lexical richness, namely for time and language mode.
The effect for language used was not significant. Two
systematically lower. In the conversation with the was an interaction effect found. This outcome seems
hearing adult the Guiraud scores for the two languages to indicate that mixing is a crucial communicative in-
are far more similar. The language mode by time strument for the deaf children, and mixed utterances
effect, F(2, 10) ¼ 6.661, p ¼ .023, reflects that the represent the children’s language development the
Guiraud scores for spoken words in the third year best.
became so high that there is no longer a mean differ-
ence between the conversations with the deaf and the
Language Dominance
hearing adult.
In Table 5, the results of the 6 children are given for
Mean length of utterance. In Table 4 and Figure 2, the language dominance.
MLU10 results are presented. Table 4 and Figure 2 All mean values in Table 5 are positive, pointing
make clear that all conversations contained signed and out that sign language is by far the dominant vehicle
mixed utterances. In many cases, however, children of expression. A value of 1 indicates a situation where
did not produce any monolingual utterance in spoken the number of signs is about twice as large as the num-
Dutch. Their MLU10 is set to 0, which explains the ber of spoken words. A value of 2 indicates that the
score less than 1 for MLU10 spoken Dutch. The mean
value over all children for spoken Dutch never exceeds
the value of 1 in Table 4 and Figure 2.
For the MLU10 in signs, no significant effects
were found, neither for the main effects of time and
language mode nor for the interaction effect of time by
language mode. The values seem to increase over time,
but the effect was not strong enough to be statistically
significant for the 6 children investigated. For the
MLU10 in spoken words, one effect is significant, that
is, language mode, F(1, 5) ¼ 7.105, p ¼ .045. It means
that monolingual utterances in spoken Dutch are more
exceptional in the conversations with the deaf adult.
Unlike the results for MLU10 in spoken words
and signs, the mixed category shows a significant de-
Figure 2 Mean values for MLU10. H ¼ hearing adult;
velopment over time, F(2, 10) ¼ 15.148, p ¼ .001. D ¼ deaf adult; signs ¼ SLN; mixed ¼ mixed utterances
No main effect for language mode was found, nor SLN/Dutch; words ¼ spoken Dutch.
246 Journal of Deaf Studies and Deaf Education 11:2 Spring 2006
Table 5 Mean values for language dominance Analyses of main and interaction effects only
Mean values (SD) yielded a significant main effect for time: The partic-
Research year ipation scores of the children as a group increased
Interaction with 1 2 3 significantly over the years of the study, F(2, 10) ¼
4.615, p ¼ .038. The participation scores of the chil-
Hearing adult 1.45 (0.87) 1.21 (0.43) 1.06 (0.71)
Deaf adult 1.81 (0.89) 2.05 (1.01) 1.41 (0.75) dren as a group show a similar increase, both in the
Note. The values represent mean values of the logit (ln) of the number of
conversations with the deaf and the hearing adults.
signs versus the number of spoken words (s/w); standard deviations are There is no extra accommodation effect in one of
given in parentheses.
the two language modes.
language, spoken Dutch, does not seem to show any sign language development and in utterance length is
progress. The lexical richness of the hearing children’s demonstrated between the ages of 2;0 and 3;0. Van den
spoken Dutch does not show a marked increase, Bogaerde found MLU10s for signs varying from 1.1 to
whereas their sign scores are lower than the scores 1.5 at the age of 1;6 years for deaf children of deaf
for spoken Dutch (2.76, 2.70, 2.99). parents. At the age of 3;0 years she found MLU10s
If we compare these results with the scores of the ranging from 3.4 to 3.7. All the MLU10s for signs of
children in our project, it appears that a difference of the children in our study are smaller than the scores
2–3 years can be observed to the advantage of the of the 3-year-old children of the study of Van den
children of Van den Bogaerde’s study (2000). In any Bogaerde. She did not calculate MLU10s for deaf
case, this holds for sign language. For spoken Dutch, children’s spoken Dutch because of the few utterances
of our study. Perhaps the hearing adult’s positive atti- parents follow SLN courses and try to use SLN or
tude towards sign language in this study and her ob- sign-supported Dutch with their deaf children. In this
vious sign language skills explain the difference with way, in the families, there is no clear-cut language in-
Fortgens’ outcomes, where the hearing adult was put pattern where languages are related to specific
a strong proponent of monolingualism in Dutch. Deaf persons. Moreover, in another research project at the
children are probably also aware of the fact that a hear- same school, it appeared that (hearing) parents who
ing adult may understand both signs and spoken had opted for a bilingual curriculum were found to
words, whereas, for a deaf native signer, speech read- shift completely to Dutch as soon as their children
ing is much more difficult. socialemotionally and cognitively performed well.
Their ultimate aim was to try and integrate their
measured by the tests could be improved, it is equally Blamey, P. J., Sarant, J. Z., Paatsch, L. E., Barry, J. G., Bow, C. P.,
important to state that the growth of sign language was Wales, R. J., et al. (2001). Relationships among speech
perception, production, language, hearing loss, and age in
mixed with continuous Dutch language development. children with impaired hearing. Journal of Speech, Lan-
This suggests that signing does not hinder the chil- guage, and Hearing Research, 44, 264–285.
dren becoming more proficient in spoken Dutch, or, Bol, G. (2005, November). Methodological issues in SLI research.
Paper presented at the EMLAR workshop (Experi-
probably, that signing even furthers the development
mental Methods in Language Acquisition Research), The
of Dutch. At the most, bilingually raised children Netherlands, Utrecht.
need more time to arrive at expected levels in both Bos, H. F. (1994). An auxiliary verb in Sign Language of the
languages. Netherlands. In I. Ahlgren, B. Bergman, & M. Brennan
(Eds.), Perspectives on sign language structure. Papers from
It is reassuring in this respect that the children’s
Gillis, S., & Verlinden, A. (1988). Nouns and early verbs in Maassen, B., & Povel, D.-J. (1985). The effect of segmental and
early lexical development: Effects of input frequency? suprasegmental correction on the intelligibility of deaf
Antwerp Papers of Linguistics, 54. Antwerp: Antwerp speech. Journal of the Acoustical Society of America, 78,
University Press. 877–886.
Harding, E., & Riley, Ph. (1986). The bilingual family. A hand- MacWhinney, B. (1991). The Childes Project: Tools for analyzing
book for parents. Cambridge, UK: Cambridge University talk. Hillsdale, NJ: Erlbaum.
Press. Marschark, M., Lang, H. G., & Albertini, J. A. (2002). Educat-
Herman, R., Holmes, S., & Woll, B. (1999). Assessing British ing deaf students. From research to practice. New York: Oxford
Sign Language development: Receptive Skills Test. Coleford, University Press.
UK: Forest Bookshop. Mogford, K. (1988). Oral language acquisition in the prelingu-
Hoffmeister, R. J. (1978). The development of demonstrative istically deaf. In D. Bishop & K. Mogford (Eds.), Language
pronouns, locatives, and personal pronouns in the acquisi- development in exceptional circumstances (pp. 110–131).
Singleton, J. L., Supalla, S., Litchfield, S., & Schley, S. (1998). Vermeer, A. (1986). Tempo en structuur van tweede-taalverwerving
From sign to word: Considering modality constraints in bij Turkse en Marokkaanse kinderen [Tempo and structure of
ASL/English Bilingual Education. Topics in Language second language acquisition by Turkish and Moroccan chil-
Disorders, 18(4), 16–29. dren]. Tilburg, The Netherlands: Katholieke Universiteit
Spencer, P. E. (2004). Individual differences in language perfor- Brabant.
mance after cochlear implantation at one to three years of Vermeer, A. (2000). Coming to grips with lexical richness
age: Child, family, and linguistic factors. Journal of Deaf in spontaneous speech data. Language Testing, 17(1),
Studies and Deaf Education, 9, 395–412. 65–83.
Van den Bogaerde, B. (2000). Input and interaction in deaf fam- Wodlinger-Cohen, R. (1991). The manual representation of
ilies. Utrecht: Netherlands Graduate School of Linguistics speech by deaf children, their mothers, and their teachers.
LOT. In P. Siple & S. D. Fischer (Eds.), Theoretical issues in
Van der Stelt, J. M. (1993). Finally a word. A sensori-motor sign language research: Vol. 2. Psychology (pp. 149–169).