üsgmi Jacqueline Leybaert & Jésus Alegria The Role öf Cued Speech in Language Development of Deaf Children Despite normal intelligence and normal potential 1996; Kuht &r Meltzoff, 1988; MacKai for learning, children bom profoundly deaf gener­ Kennedy, Spieker, & Stem, 1983). ally exhibit lags across all activities involving pho­ Lipreading constitutes the primai nological representations based on speech: speech deaf children to gain information abc perception and speech production, oral language nological. structure of spoken langv development, metaphonological abilities, immedi­ 1976). Although lipreading provides ate ordered memory for linguistic stimuli, reading; about some phonological contrasts (e arid spelling. In addition, their pattern of hemi­ articulation), it does not afford the p spheric specialization for language processing is others, like nasality and voicing (Erber generally atypical. The most likely explanation of den, Prosek, Montgomery, Scherr, &r j these findings lies in deaf children's reduced access Through lipreading deaf children havi to oral language through lipreading. to phonetically underspecified infor It is now widely recognized that lip movements they develop underspecified represer involved in the production of speech are automat­ respect, to heard-andLspoken" languaj ically processed by hearing persons in normal con­ ders deaf children's acquisition of oral ditions of listening. The fact that'visual speech in­ of all cognitive activities that rely upc formation influences thé automatic processing of ical representations. auditory information. (McGurk •&. MacDonald, To help deaf children perceive 1976) indicates that the visual speech information about the phonological structure of is dealt with by structures in the brain common-to guage through the visual channel, diff those involved in the processing of the auditory sig­ have been elaborated. One of these sy nal (Calvert et al., 1997). Hearing people thus de­ speech (CS; Cornett, 1967). Reviewin. velop phonological representations through access search on the effect of CS allows on to lipreading as well as through acoustic informa­ whether the development of a phono! tion. The basis for. the development of such amodal, depends on the delivery of accuratt perceptual representations of speech seems to oc­ about phonological contrasts, indepe cur during the first weeks of life (Bumham & Dodd, modality. More specifically, if phone 262 - Language and Language Development sentations can be elaborated on the basis of a well- specified visual input, then the development of all Gued Arnericart English Consonant Handsijapé ': abilities relying on such representations should be improved. Finally, a review of the research on CS also permits us to examine the question of the im­ pact of modality: does, the development of a lin­ Id/ (Joes A/ cals IW he guistic competence from.a visual input rather than ipt Paul M. vote Is/ saw Izhl measure Mnl there Id rate from an auditory input (with the same phonologi­ &/ zinc cal content of the' input) entail differences in the cognitive processes? Previous work has already reviewed the data on 0 ¥ ¥ the effect of exposure to CS on language acquisition M be An/my M Lee (ol no m feöl /iv/ was •m and the development of cognitive-architecture (AU /hw/ where m tickle /sh/shy egria, Leybaert, Charlier, & Hage, 1992; Leybaert, 1998; Leybaert, Alegria, Hage, & Charlier, 1998). Speech production has not been noticed to improve relative to that of deaf children using, other lan­ ¥ & guage systems (Ryalls, Auger, & Hage, 1994), but Igt go Ingl ring V jump • Jy/. your important advantàgesiiave been noted in receptive AdV üirough /cW chime language and in the degree to which language is organized neurologically. The chapter will thus be 'I Cued American English Vowel Placements focused on the following issues: how is the infor­ mation provided by the lips and by the hands in­ tegrated, and what are the possibilities for auto­ m matic systems of cueing? How are rhyming, WWWthroa t ' remembering and reading developed'by deaf chil­ tee/'mouu sere lue/'chin yo *u tool'ihroa cooA k dren using CS? Are the neural substrates involved teil her /aw/ saw HI his lei Ned ? in speech perception and in cued speech percep­ lal apple tion the same or different? Can CS provide useful information for cochlear implant users?. 1 -: sfdeforwarhd side-down {bw\ rfsri 3 to A incJM to shta el thto» » Cued Speech /oh/go A)h/up lw/1-ni hoist /ah/-/i/ my V /ah/ father lei-Hi sails /atv/co/cow, Cued speech, developed by Orin Comett in 1966, u^ii.uiMgm^èi^Ji^^^mfflnaammi --'-•ta •< and adapted to more than 40 languages and major L «Ml ttojuy U&SB.II*, dialects (Cornett, 1994), is neither a sign language Figure 19-1. Handshape and locations of American nor a manually coded system' that uses signs from Cued Speech © 2001, Language Matters, Inc. Re^ a sign language in a spoken-language word order.. printed by permission. Instead, CS is a mode of communication for.visu- ally conveying traditionally spoken languages at the phonemic level (i.e.-, the same .linguistic level con? handshape (like /p/, /d/, and /zh/) or at the same veyed via speech to hearing individuals). In CS, the location. Conversely, phonemes that have similar speaker complements lip gestures of speech with lipshape are coded with different handshape (like manual cues. A cue is made up of two parameters: /p/, /b/, and ha/) and hand location (like N and handshape arid hand location around, the mouth. lei). Information given by the cues and information The American English form of CS uses eight hand- given by lipreading is thus complementary. Each shapes corresponding to groups of consonants and time a speaker pronounces a consonant-vowel .four hand locations to convey vowels.and di- (CV) syllable, a cue (a particular handshape at a phtongs (see figure 19-1); Phonemes that are dis­ specific position) is produced simultaneously. For tinguishable by lipreading are coded by a same example, vyhen saying the words "bell" and "bowl," i Cued Speech in Language Development 263 two différent hand locations would be used to dis­ about one fourth of what is said even in dyadic tinguish between thé two vowels; when saying the conversations (Liben, 1978). Large improvement of words "bat" and "pat," two different handshapes deaf children's speech reception skills has been would be used to code the inidal consonant. Syl­ demonstrated when cues are added to lipreading labic structures other than CV are produced with, both for English- and French-speaking children additional .cues. For example, a vowel syllable is (Alegria, Charlier, ö Mattys, 1999; Nicholls & represented by the.neutral handshape at the hand Ling, 1982; Périer, Charlier, Hage, & Alegria, placement corresponding to that vowel. Syllables 1988). Nicholls and Ling (1982) found that the including consonant clusters, or codas, are coded speech reception scores of profoundly deaf children using the handshape corresponding to the addi­ taught at schoolwith CS for at least .3 years in­ tional consonant at the neutral position. creased from about 30% for both syllables and The' handshapes and hand locations used in CS; words in the lipreading alone condition to more ! unlike those of fingerspelling, are not, by them- than 80% in the lipreading- plus-cues condition. i selves, interprétable as language, Instead, the visual Périer et al. (1988) showed that the advantage on ! information provided by lipreading is also neces­ sentence comprehension provided by the addition sary. The integration of labial and manual informa­ of cues was greater in children whose parents in­ tion points to a single, unambiguous, phonological tensively used CS to communicate with them at percept that deaf children could not have achieved home at an early age than in those children who from either source alone. Deaf children .ate thus in a . benefited from CS later, and only at school, usually situation in which they can interpret the oral input- from the age of 6. This differential benefit displayed as a reliable visual language in which the.gestures by the early and late-CS users may be explained in (i.e., the combination of lip movements and man­ two ways: early CS-users might be more familiar ual cues) are now.entirely specified, both at thesyl* with words presented in CS, and/or they might labic and at the phonemic levels. For each syllable have a more efficient phonological processor, (and for each phoneme), there corresponds one which dépends of the quality of the mental repre­ (and only one) combination of labial and manual sentations of the phonemes. information, and vice versa, a characteristic diat in a study by Alegria et al. (1999), early CS makes CS entirely functional for speech perception. users displayed a larger improvement related to the Two aspects of CS design are. worth comment- addition of. cues both for word perception and for ' ing oft. First, the arbitrary decision to code the vow­ pseudo-word perception. Because pseudo-words els by hand locations and the consonants by hand were unfamiliar for both groups of subjects, these placements seems ecologically valid. Indeed, vow- results support the idea that experience with CS [•'. els have a longer duration on the acoustic level, enhances the efficiency of the processing of pho­ which corresponds to the relatively long time re- nological information in early users. , quired to pass from one location to another (see ! below). In contrast, consonants are relatively short Automatic Generation of Cued Speech 5 events, and it is possible to get rapidly from one ?. handshape to another. It is noteworthy that CS ap- Given the good results provided by the use. of CS |l pears to honor this linguistically motivated distinc- on the reception of speech by deaf children, various f- tion. Second, the possibility to transmit informa- systems of automatic generation of CS have been *• tion about a consonant and a vowel in one single elaborated: the Autocuer, developed in the late If gesture allows'a rapid r.ate of information transmis- 1970s (Cornett, Beadles, & Wilson, 1977; Duch­ | • sion.