word learning in the switch task

Word Learning in the Switch Task: The Case of Phonologically Unassimilable Non‐Na?ve Words Lillian May & Janet F. Werker University of Bri.sh Columbia Introduc?on •A narrowing occurs by 20 mo for “what may be a word.” •Given sufficient social/referenFal/perceptual cues, infants 12‐18 mo, but not 20‐26 mo, learn non‐linguisFc sFmuli as object labels 1‐4 . •Digital noises, gestures, pictograms. •The mechanism behind this narrowing is not known. •Is it language vs. non‐language sounds? •Or, are there naFve language constraints? Phonological Category Assimila?on Hypothesis •We propose that “what may be a word” narrows according to a Phonological Category AssimilaFon 5,6 word‐ learning constraint. •For older word‐learning infants (20mo+), the sounds within “possible words” must all assimilate into naFve language phonological categories. “Possible Words”? Methods Par?cipants •14 mo (n=13, Age range= 13.75‐ 14.5 mo). •20 mo (n=20, Age range= 19.5‐ 20.5 mo). •Infants were all monolingual English‐learning (>80%). S?muli •CV words containing click consonants, from the language Nǀuu 7 . •Click consonants have been found to be non‐ assimilable for English language listeners (Best, McRoberts, & Sithole, 1988). •[!a] and [ǁu] Procedure •ReferenFal 8 Switch Word‐Learning Task 9 . •Parents filled out MacArthur‐Bates Vocabulary Short‐Form Vocabulary Checklist (14 mo: Level I, 20 mo: Level II‐Form A). Results •Overall, “Switch” > “Same”, F= 8.72, p<.01. •At 20 mo, an interacFon was found between test trial and vocabulary, F= 9.506, p<.01. Discussion •14 mo infants treat words containing non‐assimilable sounds as potenFal object labels. •By 20 mo, only infants with smaller vocabularies accept non‐assimilable sounds in labels. •Vocabulary data support a Phonological Category AssimilaFon word‐learning constraint that emerges once infants have acquired a sufficiently sized lexicon. References: 1. Woodward & Hoyne, 1999; 2. Namy & Waxman, 1998; 3. Namy, 2001; 4. Hollich et al., 2000; 5. Best, McRoberts, & Sithole, 1988; 6. Best & McRoberts, 2001; 7. Miller, 2009; 8. Fennell & Waxman, in press; 9. Werker, Cohen, et al., 1998. Contact: [email protected] /t/ /p/ [t] [p] [!] [ǁ] Training (x4): Habitua?on: Pair 1: Pair 2: Test: Same: Switch: [!a] [‖u] [!a] [‖u] [kiqy] 8.598 8.729 11.325 10.709 0 2 4 6 8 10 12 14 14 Months 20 Months Mean Looking Time (sec) Age (Months) Infant Looking Time Same Switch p=.04 p=.07 8.255 9.203 12.876 8.542 0 2 4 6 8 10 12 14 16 Low High Mean Looking Time (sec) Vocabulary Group (Md Split) 20 month‐olds: Effect of Produc?ve Vocabulary Same Switch p=.004 ns ✓ ✗ What may be a Word: Developmental Narrowing Assimilable sounds Non‐language sounds What may be a word? 20+ mo Non‐assimilable sounds What may be a word? What may be a word? 12‐18 mo ? (click consonants) © Julia Leibowich

Upload: others

Post on 05-Oct-2021

2 views

Category:

Documents

0 download

Report

Download

Embed Size (px):

TRANSCRIPT

Page 1: Word Learning in the Switch Task

WordLearningintheSwitchTask:TheCaseofPhonologicallyUnassimilableNon‐Na?veWords

LillianMay&JanetF.WerkerUniversity of Bri.sh Columbia

Introduc?on• Anarrowingoccursby20mofor“whatmaybeaword.”• Givensufficientsocial/referenFal/perceptualcues,infants12‐18mo,butnot20‐26mo,learnnon‐linguisFcsFmuliasobjectlabels1‐4.

• Digitalnoises,gestures,pictograms.• Themechanismbehindthisnarrowingisnotknown.

• Isitlanguagevs.non‐languagesounds?• Or,aretherenaFvelanguageconstraints?

PhonologicalCategoryAssimila?onHypothesis

• Weproposethat“whatmaybeaword”narrowsaccordingtoaPhonologicalCategoryAssimilaFon5,6word‐learningconstraint.

• Forolderword‐learninginfants(20mo+),thesoundswithin“possiblewords”mustallassimilateintonaFvelanguagephonologicalcategories.

“PossibleWords”?

MethodsPar?cipants• 14mo(n=13,Agerange=13.75‐14.5mo).• 20mo(n=20,Agerange=19.5‐20.5mo).• InfantswereallmonolingualEnglish‐learning(>80%).

S?muli• CVwordscontainingclickconsonants,fromthelanguageNǀuu7.

• Clickconsonantshavebeenfoundtobenon‐assimilableforEnglishlanguagelisteners(Best,McRoberts,&Sithole,1988).

• [!a]and[ǁu]

Procedure• ReferenFal8SwitchWord‐LearningTask9.• ParentsfilledoutMacArthur‐BatesVocabularyShort‐FormVocabularyChecklist(14mo:LevelI,20mo:LevelII‐FormA).

Results

• Overall,“Switch”>“Same”,F=8.72,p<.01.• At20mo,aninteracFonwasfoundbetweentesttrialandvocabulary,F=9.506,p<.01.

Discussion• 14moinfantstreatwordscontainingnon‐assimilablesoundsaspotenFalobjectlabels.• By20mo,onlyinfantswithsmallervocabulariesacceptnon‐assimilablesoundsinlabels.• VocabularydatasupportaPhonologicalCategoryAssimilaFonword‐learningconstraintthatemergesonceinfantshaveacquiredasufficientlysizedlexicon.

References:1.Woodward&Hoyne,1999;2.Namy&Waxman,1998;3.Namy,2001;4.Hollichetal.,2000;5.Best,McRoberts,&Sithole,1988;6.Best&McRoberts,2001;7.Miller,2009;8.Fennell&Waxman,inpress;9.Werker,Cohen,etal.,1998.

Contact:[email protected]

/t//p/

[t][p] [!] [ǁ]Training(x4):

Habitua?on:

Pair1:

Pair2:

Test:

Same:

Switch:

[!a]

[‖u]

[!a]