Can bilingual children turn one language off? Evidence from perceptual switching

Download 195.95 Kb.
View original pdf
Size195.95 Kb.
  1   2

Can bilingual children turn one language off?
Evidence from perceptual switching
Leher Singh a, Carolyn Quam b
Department of Psychology, National University of Singapore, Singapore 117570, Singapore b
Departments of Speech, Language, and Hearing Sciences and Psychology, University of Arizona, Tucson, AZ 85721, USA
a rt i cl e info Article history:
Received 5 November Revised 11 March 2016
Lexical tone
Language development
Novel word learning
Mandarin Chinese
Childhood abstract Bilinguals have the sole option of conversing in one language in spite of knowing two languages. The question of how bilinguals alternate between their two languages, activating and deactivating one language, is not well understood. In the current study, we investigated the development of this process by researching bilingual children’s abilities to selectively integrate lexical tone based on its relevance in the language being used. In particular,
the current study sought to determine the effects of global conversation-level cues versus local (within-word phonotactic)
cues on children’s tone integration in newly learned words.
Words were taught to children via a conversational narrative,
and word recognition was investigated using the intermodal preferential-looking paradigm. Children were tested on recognition of words with stimuli that were either matched or mismatched intone in both English and Mandarin conversations. Results demonstrated that 3- to 4-year-olds did not adapt their interpretation of lexical tone changes to the language being spoken. In contrast- to 5-year-olds were able to do so when supported by informative within-word cues. Results suggest that preschool children are capable of selectively activating a single language given word-internal cues to language 2016 Elsevier Inc. All rights reserved. 0022-0965/
Ó 2016 Elsevier Inc. All rights reserved.

Corresponding author.
E-mail address:
(L. Singh).
Journal of Experimental Child Psychology 147 (2016) Contents lists available at
Journal of Experimental Child
Psychology journal homepage

One of the most significant challenges facing bilingual learners is the mastery of two systems that are linguistically distinct yet conceptually linked. This challenge is potentially complicated when a bilingual learner’s native languages maintain conflicting rules. One area of potential conflict lies in the language-specific use of phonetic variation languages often share sources of phonetic information
(e.g., voice onset time, aspiration, pitch movements) yet vary in how they partition these sources of variation to form native phonetic categories. For example, tone languages incorporate systematic variation in pitch movements to distinguish word meanings. However, non-tone languages also incorporate systematic pitch variation in away that does not distinguish word meanings but instead maybe used to differentiate emotions, questions and statements, or stress and focus (
Gussenhoven, 2004
). As a result, a bilingual learner attempting to master atone language and a non-tone language needs to interpret pitch movements in a language-selective manner based on the language being spoken in a given context. The effectiveness with which bilingual children go back and forth in terms of how they interpret phonological variation across their languages, activating and deactivating sensitivity to common cues as befits the language context, remains unclear. This process is known as perceptual switching. Perceptual switching has been investigated quite extensively in bilingual adults. However,
the extent to which children can engage in perceptual switching (and modifiers of their ability to do so) remains unclear, yet this ability represents a crucial step in the journey toward bilingual pro- ciency. The focus of the current study was to determine whether perceptual switching is observable in children and to identify possible cues that influence perceptual switching during childhood.
There have been several investigations of perceptual switching in adults. For example, many languages use variation along the voice onset time (VOT) continuum to contrast consonants. However,
the location of the boundary between phonetic categories differs across languages such as in English and Spanish. Adult bilinguals of English and Spanish, therefore, need to impose a different VOT boundary on particular sounds when listening to English versus Spanish. Ina study with adult bilinguals,
Elman, Diehl, and Buchwald (reported that Spanish–English bilinguals could selectively shift their judgments of the same sounds (b vs. p) based on whether they were listening to English and Spanish carrier sentences. Participants capacities for perceptual switching were mediated by their relative proficiency in each of their languages balanced bilinguals with similar proficiency levels across their languages were better able to shift their judgments in response to the language context than unbalanced bilinguals. Such effects of language context in perceptual switching have also been reported in speech production (
Flege & Eefting, 1987; Hazan & Boulakia, 1993
) and in other phonetic categories in Spanish (
Garcia-Sierra, Diehl, & Champlin, 2009; Garcia-Sierra, Ramirez-Esparza, Silva-
Pereyra, Siard, & Champlin, 2012
; but see
Caramazza, Yeni-Komshian, & Zurif, 1974; Caramazza,
Yeni-Komshian, Zurif, & Carbone, Perceptual switching in adults is not limited to conditions where perception is cued by language context. Rather, switching can be enabled by the availability of informative within-word cues to language identity. Ina recent study by
Gonzales and Lotto (2013)
, instead of varying the broader language context, the authors manipulated within-word phonetic information to cue a particular language.
They found that when the word [b/p]afri contained a Spanish-like r, Spanish–English bilinguals—but not English monolinguals—interpreted a preceding ambiguous initial [b/p] sound as a Spanish phoneme. When the target word contained an English-like r, bilingual Spanish–English speakers and monolingual English speakers interpreted the same ambiguous sound as an English phoneme. Just as subtle differences between phonetic segments can distinguish phonetic categories in two languages, languages also differ—and potentially conflict—in how they use suprasegmental cues. A classic example is the family of languages called tone languages. When processing atone language, learners must integrate particular pitch contours, lexical tones, into their representations of words in order to successfully contrast word meanings. By contrast, in a non-tone language such as English, pitch information does not determine the lexical identity of a word, although it cues other relevant information such as the speaker’s emotions, the stressed syllables of words, and phrase boundaries. Bilingual learners of atone language and a non-tone language must accurately interpret pitch cues with reference to the target language. Ina study with bilingual adults learning atone language and a non-tone language
L. Singh, C. Quam / Journal of Experimental Child Psychology 147 (2016) 111–125

English and Mandarin),
Quam and Creel (taught adults a series of novel words in English and
Mandarin. Participants sensitivity to tone when retrieving these words was investigated. Participants were cued to the language identity of novel words by context (the language of the conversation) or within-word cues. Quam and Creel’s findings demonstrated that in adult Chinese–English bilinguals,
within-word cues enabled a language-specific interpretation of tone, whereas language context cues did not when presented in isolation.
The vast majority of studies on perceptual switching have focused on adult bilingual processing of ambiguous phonemes. There has been a minimal focus on the developmental origins of this ability.
Developmental investigations of phonological sensitivity in word learning in bilingual children have focused predominantly on sensitivity to phonetic variation within one of their languages (e.g.,
Byers-Heinlein, Fennell, & Werker, 2013; Fennell & Byers-Heinlein, 2014; Fennell, Byers-Heinlein, &
Werker, 2007; Mattock, Polka, Rvachew, & Krehm, 2010
). There has been much less focus on how bilinguals alter their phonological sensitivities as they switchback and forth between their languages.
However, the ability to recalibrate phonetic perception in response to language input is a crucial component of acquiring bilingual proficiency. Although adults demonstrate a capacity to recalibrate in response to language context, the extent to which children can do the same remains unknown.
Furthermore, it remains unknown whether children harness the same set of cues to the target language when switching between languages as adults or whether they use a different set of cues.
There has been one previous study to investigate perceptual switching during infancy. Ina study with infants from 7 to 11 months of age,
Singh and Foong (familiarized bilingual Mandarin–Eng- lish infants with two words and then measured infants abilities to recognize the familiarized words in sentences. Infants were tested on their capacity to interpret tone variation as lexical in Mandarin and to interpret pitch variation as non-lexical in English. Each infant was tested in two sessions an English session and a Mandarin session. In one session, infants were exposed to two English words. They were then presented with passages containing familiarized words as well as unfamiliar passages. One of the words matched in pitch between familiarization and test, and the other word was mismatched in pitch. A second session, conducted in Mandarin, was otherwise identical to the English session except that one familiarized word was presented in a form that matched in pitch across familiarization and test. In contrast, the other word changed in pitch, corresponding to a change in lexical tone, between familiarization and test. At 7.5 months of age, consistent with results of a similar study with monolingual English-learning infants (Singh, White, & Morgan, 2008
), bilingual infants only recognized English words when they were pitch matched with words from the familiarization phase. In Mandarin, they only recognized tone-matched words at 7.5 months and did not equate pitch and tone variants in English and Mandarin contexts, respectively. However, at 11 months, infants displayed language-specific integration of pitch variation, recognizing pitch-matched and mismatched English words. In addition,
they only recognized Mandarin words when they were matched intone across familiarization and test.
Singh and Foong’s (study hints at perceptual switching in infants, there are two important considerations when interpreting their results. First, there was no evidence to suggest that infants had attached meaning to familiarized words. The degree of phonological precision attached to word representations has often differed for infant word segmentation tasks versus word learning tasks in toddlers that involve establishing the meanings of words (e.g.,
Jusczyk & Aslin, 1995
, Experiment 3
Swingley & Aslin, 2000
). For example, although
Singh and Foong (found that bilingual Mandarin English learners did not differentiate English words based on pitch characteristics at 11 months of age,
Singh, Hui, Chan, and Golinkoff (found that at 18 months Mandarin–English bilingual infants did differentiate English words by pitch contour when associating word forms with meaning.
Examples of task dependence in prior investigations of infants phonological representations are not uncommon (see also
Stager & Werker, 1997
). In particular, tasks that require auditory sensitivity to sound change versus those that require word–object mapping often yield different findings. The former often give the appearance of relative strength on the part of infants, and the latter often demonstrate a learning cost introduced by weightier task demands (see
Quam & Creel, 2015
, fora discussion. The process of mapping phonetic variation onto meaning arguably carries a greater cognitive load than recognition of word forms or discrimination of sounds (
Curtin, Byers-Heinlein, & Werker,
2011; Werker & Curtin, L. Singh, C. Quam / Journal of Experimental Child Psychology 147 (2016) 111–125 113

A second difference between
Singh and Foong’s (study and prior investigations of perceptual switching in adults is that traditionally participants are presented with the same type of variation across language contexts. In Singh and Foong’s study, the pitch variation built into the English contexts was acoustically distinct from the tone variation incorporated into the Mandarin contexts. Perceptual switching usually involves a language-selective interpretation of the same phonetic information. For these reasons, it remains an open question as to whether language-selective sensitivity to tone when children need to reinterpret similar tone changes across languages indifferent ways.
In the current study, we sought evidence of perceptual switching in bilingual preschool children by investigating whether lexical tone was processed in a language-selective manner. In particular, we investigated influences of discourse-level and word-internal cues on bilingual children’s integration of tone during word learning in each language. Across two experiments, we taught children novel words in English and Chinese narrative contexts and tested their recognition of these words when they were matched intone or mismatched intone. Of primary interest was whether children would interpret tone variation as lexical in Chinese and as non-lexical in English. In Experiment 1, target words were phonologically plausible in both languages such that the only cue to the relevance of lexical tone changes was the language used in the narrative context. In Experiment 2, target words in the English context were phonologically legal in English but not in Mandarin. Target words in the Mandarin sentences were identical to those used in Experiment 1 (i.e., phonologically legal in both languages. As such, children had the benefit of two sources of information (language context and word-internal cues for one of the languages) when interpreting lexical tone changes. In both experiments, children were tested at 3 to 4 years of age and at 4 to 5 years of age.
Experiment In Experiment 1, children were presented with two videos where they were taught the meanings of new words in English and Mandarin. Videos consisted of conversations in which two puppets discussed two novel objects, one of which was explicitly labeled (target) and one of which was discussed but never labeled (distractor). In both languages, children were tested on their ability to recognize the target when it matched versus mismatched the tone of the training stimulus via a preferential looking paradigm. Target words were phonotactically legal in both English and Mandarin and were equally biased toward English and Mandarin phonology (
Quam & Creel, 2012
The sample of participants comprised 34 bilingual preschool children in two age groups 17 year- olds (range = 3;1 [years;months] to 3;11) and 17 4-year-olds (range = 4;2 to 5;0). All participants were attending bilingual immersion preschools and were judged to have native proficiency in Mandarin
Chinese and English. All participants had daily exposure to Mandarin and English and were judged by bilingual experimenters to be equally proficient in both languages. Following a short min conversation with each child, experimenters rated the child on his or her proficiency in English and Mandarin on a scale from 1 to 5. Each experimenter was a native speaker of English and Mandarin. A score of on this scale indicated that children could understand conversational speech in English and Mandarin,
could accurately respond in full sentences and maintain the conversation successfully, and spoke with a native language accent. Only children who received 5 on this scale for both English and Mandarin were tested further. An additional 6 children were tested but not included due to failure to complete testing in both languages (n = 3), equipment failure (nor zero attention during all the test trials fora trial type (n = 1).
All speech stimuli were recorded by a bilingual female native speaker of English and Mandarin who was asked to produce the words in infant-directed speech. The speaker was selected on account of having a local accent that would be familiar to the children. She was from the same geographical
L. Singh, C. Quam / Journal of Experimental Child Psychology 147 (2016) 111–125

origin and had the same language background (native speaker of English and Mandarin) as the sample.
The novel words biufa and fipu were adopted from
Quam and Creel (and were designed to be phonologically equally biased toward English and Mandarin phonology. Each word contained a
Mandarin tone on the first syllable and a neutral tone on the second syllable. Tones 2 and 4 were chosen because they are reportedly highly distinguishable to native and nonnative speakers of
Mandarin (
Halle, Chang, & Best, 2004
). Tone 2 is marked by arising contour. Tone 4 is marked by a falling pitch contour (see
Fig. fora depiction of pitch contours of syllables assigned Tones 2 and Across participants, the pairings of words and tones were counterbalanced.
Acoustic analyses were conducted on the first syllable of the target words to determine mean pitch onset and offset (summarized in
Table 1
). To ensure that the two tones used in this experiment—Tones
2 and were accurately pronounced, 10 adults who were native speakers of Mandarin were asked to complete atone identification task. The first syllable of the target word was excised from carrier sentences and randomly concatenated into an audio file consisting of each target word presented in citation form in each tone. The adults were presented with all tokens in citation form and were instructed to rate them with Mandarin tone numbers, 2, 3, or 4. All stimuli were rated with accuracy. To ensure that there was no language-specific variation in the realization of tones across English and Mandarin on the part of the speaker, target words were exactly the same tokens between the English and Mandarin videos. Specifically, they were spliced from the Mandarin context and inserted into the English video.
Apparatus and procedure
The experiment was conducted in a quiet, dimly lit room. Stimuli were presented on a Macintosh computer. The experimenter sat on a chair beside the child, with the child facing the center of the
Time (ms)
Tone Tone Pitch (Hz Fig. 1. Sample pitch contours of Mandarin Tones 2 and Table Acoustic analyses of target words.
Training trials
Test trials
Fundamental frequency (Hz)
Word onset
Word offset
Word onset
Word offset (rising)
English gree
223.7 350.6 245 330.8
280.5 414 253.6 Mandarin biu
232.01 356.56 212.6 298.8

272.8 405.9 252.7 328.1 4 (falling)
English gree
356.87 206.45 378.4 222.9
395.15 237.31 390.3 Mandarin biu
426.12 271.38 378.2 225

429.29 274.93 435.2 L. Singh, C. Quam / Journal of Experimental Child Psychology 147 (2016) 111–125 115

computer screen. Another experimenter sat behind the laptop and video-recorded the child’s eye movements. Auditory stimuli were presented over speakers at conversational level (65 db. The participant watched two videos (English and Mandarin, each consisting of conversations between two puppets. Each video consisted of a conversational narrative between two puppets where a word
Fig. 2. Sequence of events during the experimental session.
L. Singh, C. Quam / Journal of Experimental Child Psychology 147 (2016) 111–125

was introduced, followed by a test phase where recognition of the learned words was investigated.
The participant watched the entire narrative and was tested on recognition of familiarized words during the test phase in one language before viewing the same sequence of events in the other language.
Each video featured the same bilingual protagonist named Lily, who appeared to be Chinese (see
Fig. for the trial sequence. In the Mandarin video, Lily initiated a conversation with another Chinese puppet, Hui Xian, who also had long straight hair and wore Chinese clothing. During the conversation,
Lily talked about two objects. One object was explicitly labeled (target) and one was discussed for an equal duration but never labeled (distractor) (fora sample of the conversational narrative, see
Fig. Toys were animated by Lily during the conversation and appeared in the lower half of the screen at the center.
During the English video, Lily introduced two different objects to a Caucasian puppet, Elizabeth.
Elizabeth wore Western clothing and had blonde hair. The dialogues in the English and Chinese videos were translations of each other, although the background scene changed, as did the target and distrac- tor objects, to sustain attention across both videos and reduce interference between word–object mappings. During the conversation, each target label was produced 10 times and each object appeared on the screen for 45 s such that the participant was equally familiar with the labeled and unlabeled objects prior to entering the test phase. Target words were labeled as a biufa or a fipu. One label was selected to name the target object in the English video, and the other target word was selected to name the target object in the Mandarin video. The child watched an English video and a Mandarin video in succession. After each conversation, a test phase was initiated. The test phase was presented in English following the English conversation and in Mandarin following the Mandarin conversation.
Conversations and test phases were blocked such that the participant watched the English conversation and English test block followed by the Mandarin conversation and Mandarin test block or vice versa. For each language, the structure of the test phase was identical except for the language in which auditory stimuli were presented. During the test phase, the labeled object served as the target and the

Share with your friends:
  1   2

The database is protected by copyright © 2017
send message

    Main page
mental health
health sciences
gandhi university
Rajiv gandhi
Chapter introduction
multiple choice
research methods
south africa
language acquisition
Relationship between
qualitative research
literature review
Curriculum vitae
early childhood
relationship between
Masaryk university
nervous system
Course title
young people
Multiple choice
bangalore karnataka
state university
Original article
academic performance
essay plans
social psychology
psychology chapter
Front matter
United states
Research proposal
sciences bangalore
Mental health
compassion publications
workplace bullying
publications sorted
comparative study
chapter outline
mental illness
Course outline
decision making
sciences karnataka
working memory
Literature review
clinical psychology
college students
systematic review
problem solving
research proposal
human rights
Learning objectives
karnataka proforma