Memorie Fonologica

  • View
    219

  • Download
    0

Embed Size (px)

Text of Memorie Fonologica

  • 8/14/2019 Memorie Fonologica

    1/12

    174

    Perrine Willems

    Jacqueline Leybaert

    Originales

    Universit libre de BruxellesLaboratoire de PsychologieexprimentaleBrussels (Belgium)

    Correspondence:Jacqueline LeybaertUniversit libre de BruxellesLaboratoire de Psychologieexprimentale, CP 191

    50, avenue F.D. RooseveltB-1050 BrusselsBelgiumE-mail: leybaert@ulb.ac.be

    Phonological Short term memory in deaf

    children fitted with a cochlear implant:effects of phonological similarity, wordlenght and lipreading cues

    Copyright 2009 AELFA yGrupo Ars XXI de Comunicacin, S.L.

    ISSN: 0214-4603

    Revista de Logopedia, Foniatra y Audiologa2009, Vol. 29, No. 3, 174-185

    40

    Abstract

    Phonological short-term memory (pSTM), or the abil-ity to hold information in mind for a few seconds, isinvestigated in deaf children fitted with a cochlearimplant (CI children) before the age of 3 years, in theframework of Baddeleys model. Results show that,compared to their age-matched hearing controls, CIchildren are delayed in the development of theirpSTM capacity, and exhibit reduced effect of phono-logical similarity (PSE) and word length (WLE). How-ever, when CI children are matched for pSTM capac-ity with younger NH children, the differenceregarding PSE and WLE disappear. The CI children donot produce more order errors than NH children.

    Taken together, the results indicate normal resourcesof functioningof pSTM. The reasons for the shorterpSTM span in CI children are discussed.

    Key words: Cochlear implant, auditory perception, phonological short-termmemory, phonological similarity effect, word length effect, memory span.

    La memoria fonolgica a corto plazo en niossordos con implante coclear: efectos de lasimilitud fonolgica, la longitud de la palabray la lectura labial

    En este trabajo se investig, en el marco del modelode Baddeley, la memoria fonolgica a corto plazo(MfCP), o habilidad para mantener informacin en lamente durante unos segundos, en nios sordos quehan recibido un implante coclear (IC) antes de los3 aos. Los resultados muestran que, comparados con

    un grupo de nios oyentes de igual edad, los nioscon IC presentan un retraso en el desarrollo de su

    capacidad de MfCP, y muestran un efecto reducido desimilitud fonolgica y de longitud de la palabra. Sinembargo, cuando se empareja a los nios con IC connios oyentes jvenes por su capacidad de MfCP, des-aparece la diferencia en similitud fonolgica y longi-tud de la palabra. Los nios con IC no producen mserrores de orden que los nios oyentes. Tomados con-juntamente, estos resultados indican unos recursosnormales de funcionamientode la MfCP. Se discutenlas razones del lapso ms bajo de MfCP de los nioscon IC.

    Palabras clave: Implante coclear, percepcin auditiva, memoria fonolgica acorto plazo, efecto de similitud fonolgica, efecto de longitud de la palabra,memoria retentiva.

    Introduction

    Working memory refers to the capacity to main-tain and manipulate information during an ongoingtask. Working memory seems an important factor inexplaining changes in language skills in hearingadults and children (Baddeley, Gathercole andPapagno, 1998; Cowan, 1996), even in very youngchildren aged 13 to 20 months. Apparent changes invocabulary development may emerge out of incre-mental changes in capacity such as working mem-ory and experience with individual words (Mills, Con-boy and Paton, 2006). Given the relationshipbetween language development and working mem-ory capacity, the question of the development of thiscapacity in deaf children fitted with a cochlearimplant is a very important one.

    Phonological short term memory (pSTM) is thepart of the working memory allocated to the short-term maintenance of linguistic information. In the

  • 8/14/2019 Memorie Fonologica

    2/12

    41

    RevLogop

    Fon

    Audiol2009,

    Vo

    l.29,

    No.

    3,

    174-185

    PHONOLOGICAL SHORT TERM MEMORY IN DEAF CHILDREN FITTED WITH A COCHLEAR IMPLANT: EFFECTS OF

    PHONOLOGICAL SIMILARITY, WORD LENGHT AND LIPREADING CUES

    PERRINE WILLEMS

    JACQUELINE LEYBAERT

    175

    case of speech, pSTM mechanisms have been best

    described by the phonological loop model of Badde-ley et al. (1984). This model includes two compo-nents: a phonological buffer (or store) that holdsmemory traces for approximately 2 seconds, and asubvocal rehearsal process used to refresh the mem-ory traces.

    According to Jacquemot and Scott (2006),phonological short-term memory (pSTM) is related tospeech processing in two ways. First, the perform-ance of the phonological buffer which store verbalinformation transiently, is affected by the phonolog-ical nature of speech stimuli. Normally-hearingadults and children show the phonological similar-

    ity effect(PSE): stimuli that are phonologically dis-similar (non-rhyming words) are recalled better thanstimuli phonologically similar (rhyming words). In thecase of rhyming words, the candidate representa-tions for a particular serial recall position are similaron the vowel. Maintaining a sequence of such rep-resentations in correct order is more difficult thanmaintaining a sequence of dissimilar representations.The locus for the phonological similarity effect isthe phonological store, a component sensitive tonative phonological properties that are processed bythe speech perception mechanisms. For instance, therecall performance of French subjects worsens with

    sequences of stimuli that differ in stress location,whereas Spanish subjects perform well. Unlike Span-ish, stress location in French is predictable andFrench speakers need not code stress in their phono-logical representations (Dupoux, Peperkamp andSebastian-Galls, 2001; Thorn and Gathercole, 2001).This suggests that native phonological propertiesinfluence recall performance and that the code usedto store the stimuli is phonological in nature.

    Second, the pSTM also interacts closely with thespeech production system. The pSTM performance ishighly affected by the length of the stimuli, in termsof duration of articulation (Baddeley, Thompson andBuchanan, 1975). The word length effect(WLE) refersto the fact that words that take more time to artic-ulate and subvocally rehearse (long words) provokea reduced span compared to words that are shorterto articulate and subvocally rehearse. The WLE isabolished under articulatory suppression, suggestingthat this effect is linked to the subvocal rehearsalcomponent of pSTM (Baddeley, et al., 1984). ThepSTM performance depends on how quickly speechstimuli can be produced subvocally (i.e. internally

    produced without any spoken output). For instance,

    digit span are larger in languages whose digits arefast to pronounce, an effect even observed in bilin-gual subjects (Murray and Jones, 2002).

    While the pSTM can been conceived as involvingprocesses overlapping with both speech perceptionand speech production, it is more appropriate toconceive it as encompassing processes involved inthe perception and production of any language, be itsigned, cued, or spoken. As a matter of fact, the PSE,the WLE and the articulatory suppression effect havebeen observed in participants congenitally deafbelonging to different linguistic backgrounds (SignLanguage, oralism, Cued Speech). Studies have sug-

    gested that in sign languages, as in spoken lan-guages, information is stored in a phonological code(hand configuration, movement, place of articulationand hand orientation are the phonological primitivesin sign languages). Tested with sequences of signedstimuli, deaf signers show a PSE: recall performancesis lower for sequences composed of similar signsthan for those composed of dissimilar signs (Wilsonand Emmorey, 1997). As in spoken language, deafsigners performance is lower when signed stimuli tobe remembered are long rather than short to articu-late (Wilson and Emmorey, 1998). This sign lengtheffect (SLE) is similar to the WLE found in hearing

    subjects: under manual articulatory suppression theSLE is abolished suggesting that it originates fromthe rehearsal process. These results suggest that thesigned pSTM consists of a buffer that stores informa-tion using the phonological structure of the sign lan-guage, and a submanual rehearsal process that seemsto operate like a subvocal rehearsal process describedin hearing subjects (Wilson, 2001).

    In deaf participants educated orally or with CuedSpeech(1), the pSTM consists of a buffer that storesinformation using the phonological structure of spo-ken or cued languages, and a submanual rehearsalprocess sensitive to the WLE as described in hearingsubjects. The performance of deaf youngsters is sen-sitive to the PSE and the WLE: better recall isobtained for non-rhyming words than for rhymingwords, and also better for short words than for longwords (Campbell and Wright, 1990; Conrad, 1979,Hanson, 1990; Leybaert and Charlier, 1996; Leybaertand Lechat, 2001; Lichtenstein, 1998). Interestingly,

    (1) Cued Speech (CS) is a system of manual cues which disambiguate speech reading.Children raised early and intensively with CS show a rate of language developmentsimilar to that of age-matched hearing children (see Leybaert et al., 1998).

  • 8/14/2019 Memorie Fonologica

    3/12

    42

    RevLogop

    Fon

    Audiol2009,

    Vo

    l.29,

    No.

    3,

    174-185

    PHONOLOGICAL SHORT TERM MEMORY IN DEAF CHILDREN FITTED WITH A COCHLEAR IMPLANT: EFFECTS OF

    PHONOLOGICAL SIMILARITY, WORD LENGHT AND LIPREADING CUES

    PERRINE WILLEMS

    JACQUELINE LEYBAERT

    176

    the performance is also better for written syllables

    that are dissimilar on the lips than for syllables whichare similar on the lips, suggesting that the phonolog-ical properties conveyed by speechreading influencerecall performan