Upload File

phonetic vs. phonological considerations in inter...

  • Home
  • Documents
  • Phonetic vs. Phonological Considerations in Inter ...d-scholarship.pitt.edu/26718/1/tse_h_2016_lsa_poster...1 7 X 30 = 510 tokens relevant for analysis 6 X 510 = 3060 vowel tokens
1
Phonetic vs. Phonological Considerations in Inter-Generational Vowel Change in Toronto Heritage Cantonese 8) Next Steps 7) Discussion 5) Methodology 6) Results 2) Question: What about Toronto HCAN speakers? 3) Data HLVC Project (Nagy 2011) •Soc iolinguistic Interviews (~ 1 hour long) •Ethnic Orientation Questionnaire •Word L ist (based on picture naming task) 4) Speaker Information STEP 1: F1 and F2 measurements of 6 vowels (/y/, /u/, /i/, /ɛ /, /ɔ /, /a/) ac ross two phonetic contexts (open vs. closed syllable) using PRAAT (Boersma & Weenik 2015) • /u/ vs. /y/ contrast maintained for all speakers • Sex effect found for /y/ only (backing among females) • Age effect found for both vowels (backing of both among youngest speakers) • à does notshow phonetic assimilation between CAN /u/ and TOENG /u/ • Non-significant factors: GEN, EOQ, Syllable Type, Tone • Supports findings from Chang et al (2011) • à HL bilinguals maintain language-internal distinctions Short-Term Goals •Measure all 8 monophthongs •Larger speaker sample size and more vowel tokens Long-Term Goals •Compare HCAN and TOENG vowel spac e using CinC (Contact in the City) data (Hoffman & Walker 2010) •Compare Toronto and HK CAN vowel spac e Boe rsma, Paul.; and David W e enink. 2 0 15 . Praa t: d o ing p ho netics by co mp u ter [ Co mputer prog ra m] . http://w ww.praat.org/. Chang, Charle s B.; Y ao Y ao.; Erin F. Hayne s.; and Russe ll Rhode s. 20 11. Productionofphonetic andphonological contrast by heritagespeakers of Mandarin. Th e J o u rna l of the Acou stica l Society o f America1 29.3 96 4– 3 9 8 0 . doi:1 0.11 21 /1 .35 69 736 . Cui, Naomi.; Minyi Zhu.; Vina Law.; Holman Tse .; and Naomi Nagy. 2 0 1 4. Exploring automated formant analysis forcomparative variationist study of HeritageCantoneseand English. Presentationat C ha nge a n d Variation in Canada(CVC8) . Queen’s University, K ingston, ON, Canada. http://d-scholarship.pitt.edu/26261/. Fabricius, Anne H.; Dominic W att.; and Danie l Ezra Johnson. 2 0 09. A comparison of three speaker- intrinsic vowel formant frequency normalization algorithms for sociophonetics. La ng ua ge Variatio n an d Change 2 1.4 13 –43 5. doi:1 0.1 01 7/S0 954 39 450 99 901 60. Hof f man, Michol F.; and Jame s A. Walke r. 20 10 . Ethnolects andthe city: Ethnic orientationand linguistic variation inTorontoEnglish. La ngu ag e Va ria tio n and Ch an ge 2 2.3 7–6 7. doi:1 0 .10 17/S0 95 439 45 099 902 38 . Johnson,DanielEzra. 2 00 9. Getting off the GoldVarbStandard: Introducing Rbrul f or Mixed-Ef fects Variable Rule Analysis. Lan gu ag e an d Linguistics Compass 3.3 59 –38 3. Nagy, Naomi. 2011. A Multilingual Corpus toExploreVariationin LanguageContact Situations. Ra sseg na Ita lian a d i Ling uistica Applicata 4 3.6 5–8 4. Thomas, Erik.; and Tyle r Ke ndall. 2 0 07. NORM: Th e vo w el no rma lization and plottingsuite. http://ncslaap.lib.ncsu.edu/tools/norm/. 9) References 1) Introduction CAN Background ENG Background Demographic Characteristics Male Female Totals GEN 1 CAN dominant Variable Born & raised in HK, lived in GTA for > 20 years 4 (ages: 46-62) 5 (ages: 50-82) =9 GEN 2 Proficient enough for sociolinguistic interview ENG dominant Born & raised in GTA (or lived in GTA since age 3) 3 (ages: 21-44) 5 (ages: 16-21) =8 =7 = 10 N = 17 Vowel Open Syllable Closed Syllable Speaker Totals /y/ 10 5 =15 /u/ 5 10 =15 =15 =15 N = 30 17 X 30 = 510 tokens relevant for analysis 6 X 510 = 3060 vowel tokens measured 10) Acknowledgements STEP 2: Watt & Fabricius Modified Method (Fabricius,Watt, and Johnson 2009) for normalization of 3060 tokens using NORM STEP 3: Mixed Effects Modeling using R-brul (Johnson 2009) Dependent Variable F2 Independent Variables Fixed Effects Social: GEN, Sex, Age, EOQ Linguistic: Syllable Type, Tone Random Effects Speaker, Word English-Mandarin Bilingual Speakers compared (Chang et al. 2011) HK Hong Kong GTA Greater Toronto Area HL Heritage Language (acquired as child) L2 Second Language (acquired as adult ) (H)CAN (Heritage) Cantonese (TO)ENG (TO)ENG: Toronto English GEN 1 Generation 1 speakers GEN 2 Generation 2 speakers EOQ Ethnic Orientation Questionnaire Scores Poster # 13 Language Contact in the Mind and in the Community: Insights from Bilingual Phonetics and Phonology Linguistic Society of America Annual Meeting, January 8, 2016, Washington, DC Holman Tse ([email protected]), Dept. of Linguistics, Univ. of Pittsburgh, Pittsburgh, PA 15260 Youngest speakers most retracted Youngest and female speakers most retracted HERITAGE LANGUAGE VARIATIONAND CHANGE I N TORONTO HTTP://PROJECTS.CHASS.UTORONTO.CA/NG N/HLVC u u y L2 Mandarin Distinct but with ENG phonetic influence y HL Mandarin Phonological considerations inhibit fronting u TOENG GEN 2 HL speakers (Cuiet al.2014) u y HCAN??? ??? ??? 多謝晒! Naomi Nagy, Scott Kiesling, ShelomeGooden, HLVC Cantonese RA's: Abigail Chan, Karen Chan, Vivian Chow, Naomi Cui, Tiffany Chung, Radu Craioveanu, Joyce Fok, Vina Law, Samuel Lo, Grace Lui, Rita Pang, Andrew Peters, Mario So Gao, Josephine Tong, Sarah Truong, Elaine Wang, Ka-man Wong, Alfred Wu, Olivia Yu, Minyi Zhu u y ( Thomas and K endall 2 007) Age (p < 0.01)** Coefficient +1 2.391 Sex (p < 0.01)** Coeff. N Mean (Hz) M 56.148 105 1808 F -56.148 150 1682 Age (p < 0.01)** Coefficient +1 3.237 Cantonese is the 2 nd most widely spoken language in Toronto y u Distinc t for all speakers (p = 8.77 X 10 -22 )***

Upload: others

Post on 13-May-2021

2 views

Category:

Documents


0 download

Report

TRANSCRIPT

Page 1: Phonetic vs. Phonological Considerations in Inter ...d-scholarship.pitt.edu/26718/1/tse_h_2016_lsa_poster...1 7 X 30 = 510 tokens relevant for analysis 6 X 510 = 3060 vowel tokens

Phoneticvs.Phonological Considerations inInter-Generational VowelChangeinTorontoHeritage Cantonese

8) Next Steps

7) Discussion

5) Methodology

6) Results

2) Question: What about Toronto HCAN speakers?

3) DataHLVCProject (Nagy2011)•SociolinguisticInterviews (~1hourlong)•EthnicOrientationQuestionnaire•WordList(basedonpicturenamingtask)

4) Speaker Information

STEP 1:F1andF2measurementsof6vowels(/y/,/u/, /i/, /ɛ/,/ɔ/, /a/)acrosstwophoneticcontexts(openvs.closedsyllable)usingPRAAT (Boersma &Weenik 2015)

• /u/vs. /y/contrast maintained for allspeakers• Sexeffect foundfor /y/ only(backing among females)• Age effectfound for both vowels(backing of bothamong youngest speakers)• à doesnotshow phoneticassimilation between CAN/u/and TOENG /u/• Non-significant factors: GEN, EOQ, Syllable Type, Tone• Supports findings from Chang etal (2011)• à HL bilingualsmaintain language-internal distinctions

Short-TermGoals•Measureall8monophthongs•Largerspeakersamplesize andmorevoweltokens

Long-TermGoals•CompareHCANandTOENGvowelspaceusingCinC (Contactin theCity)data(Hoffman&Walker2010)•CompareTorontoandHKCAN vowelspace

Boersma, Paul.;and David Weenink.2015.Praat:doingphoneticsbycomputer[Computerprogram].http://www.praat.org/.Chang, CharlesB.;Yao Yao.; ErinF.Haynes.;and Russell Rhodes.2011.ProductionofphoneticandphonologicalcontrastbyheritagespeakersofMandarin. TheJournalof theAcousticalSocietyofAmerica129.3964–3980.doi:10.1121/1.3569736.Cui,Naomi.;Minyi Zhu.;VinaLaw.;Holman Tse.;and NaomiNagy.2014.Exploringautomatedformantanalysisforcomparativevariationist studyof HeritageCantoneseandEnglish. Presentationat ChangeandVariationinCanada(CVC8).Queen’sUniversity,Kingston,ON,Canada.http://d-scholarship.pitt.edu/26261/.Fabricius, AnneH.;DominicWatt.;and DanielEzraJohnson.2009.Acomparisonof threespeaker- intrinsicvowelformantfrequencynormalizationalgorithmsforsociophonetics. LanguageVariationandChange21.413–435.doi:10.1017/S0954394509990160.Hoffman,Michol F.;andJamesA.Walker.2010.Ethnolects andthecity:EthnicorientationandlinguisticvariationinTorontoEnglish.LanguageVariationandChange22.37–67.doi:10.1017/S0954394509990238.Johnson,DanielEzra.2009.Gettingoff theGoldVarbStandard:IntroducingRbrul forMixed-EffectsVariableRuleAnalysis. LanguageandLinguisticsCompass 3.359–383.Nagy, Naomi.2011.AMultilingualCorpustoExploreVariationinLanguageContactSituations.Rassegna Italianadi LinguisticaApplicata 43.65–84.Thomas, Erik.; and Tyler Kendall.2007.NORM:Thevowelnormalizationandplottingsuite.http://ncslaap.lib.ncsu.edu/tools/norm/.

9) References

1) Introduction

CANBackground

ENGBackground

DemographicCharacteristics

Male Female Totals

GEN 1 CANdominant Variable Born &raised inHK,livedinGTA for> 20years

4(ages: 46-62)

5(ages: 50-82)

=9

GEN 2 Proficientenough forsociolinguisticinterview

ENG dominant Born &raised inGTA (or livedinGTAsinceage 3)

3(ages: 21-44)

5(ages: 16-21)

=8

=7 = 10 N=17

Vowel OpenSyllable

ClosedSyllable

SpeakerTotals

/y/ 10 5 =15/u/ 5 10 =15

=15 =15 N=3017X30=510tokensrelevantfor analysis6X510=3060voweltokensmeasured

10) Acknowledgements

STEP 2:Watt& FabriciusModifiedMethod (Fabricius,Watt,andJohnson2009)fornormalizationof3060tokensusingNORM

STEP 3:MixedEffectsModeling usingR-brul (Johnson2009)

Dependent Variable

F2

IndependentVariables

FixedEffectsSocial: GEN, Sex, Age, EOQLinguistic: Syllable Type, Tone

Random EffectsSpeaker, Word

English-Mandarin Bilingual Speakerscompared (Changetal. 2011)

HK Hong KongGTA Greater Toronto AreaHL Heritage Language (acquired aschild)L2 Second Language (acquired asadult)

(H)CAN (Heritage) Cantonese(TO)ENG (TO)ENG: Toronto EnglishGEN 1 Generation 1speakersGEN 2 Generation 2speakersEOQ Ethnic Orientation Questionnaire Scores

Poster # 13

Language Contact in the Mind and in the Community: Insights from Bilingual Phonetics and PhonologyLinguistic Society of America Annual Meeting, January 8, 2016, Washington, DC

Holman Tse ([email protected]), Dept. of Linguistics, Univ. of Pittsburgh, Pittsburgh, PA 15260

Youngest speakers most retractedYoungest and femalespeakers mostretracted

HERITAGE LANGUAGE VARIATIONAND CHANGE IN TORONTO

HTTP://PROJECTS.CHASS.UTORONTO.CA/NGN/HLVC

uuyL2Mandarin

DistinctbutwithENGphonetic influence

yHLMandarin

Phonological considerationsinhibit fronting

uTOENG

GEN2HLspeakers (Cuietal.2014)

uyHCAN???

??? ???

多謝晒!

Naomi Nagy, Scott Kiesling, ShelomeGooden, HLVC Cantonese RA's: Abigail Chan, Karen Chan, Viv ian Chow, Naomi Cui, Tiffany Chung, Radu Craioveanu, Joyce Fok, Vina Law, Samuel Lo, Grace Lui, Rita Pang, Andrew Peters, Mario So Gao, Josephine Tong, Sarah Truong, Elaine Wang, Ka-man Wong, Alfred Wu, Oliv ia Yu, MinyiZhu

uy

(ThomasandKendall2007)

Age (p<0.01)**Coefficient

+1 2.391

Sex(p <0.01)**Coeff. N Mean (Hz)

M 56.148 105 1808F -56.148 150 1682

Age (p<0.01)**Coefficient

+1 3.237

Cantoneseisthe2ndmostwidelyspokenlanguageinToronto

y u Distinct for all speakers(p=8.77X10-22)***

Tokens Warpack Everblight

PDF Tokens| Medals and the Like - Schafluetzel/Tokens/Tokens Master.pdf · Tokens, Medals and the Like Our Strategy ... Badge for D.O.K.K Convention. ... Entered 12-9-46. Chattanooga

Game Contents - MyMiniFactory...3 jungle hexes 2 volcano hexes 1 river (gold) hex 92 colony tokens: - 24 discovery tokens (jungle) - 22 gold tokens - 22 ore tokens - 24 gem tokens

Pondemonium Tokens

BIROn - Birkbeck Institutional Research Onlineeprints.bbk.ac.uk/26718/1/26718.pdf · 2 VALENTIN FÉRAY AND AMARPREET RATTAN A rather general formula for c ; is due to A. Goupil and

Two Towers Tokens

Simulia Used Tokens

Tokens expressionsin C++

Tokens modernos

WFG Tokens

Conformance Checking of Processes Based on Monitoring Real ...julien/Julien_at_Nijmegen/DWFTT09_files/ronny_DWFTT.… · = 2 = 0 = 0 missing tokens remaining tokens consumed tokens

SRM Space Tokens

Canada Bank Tokens

Excerpt from "Tokens"

Jugula Tokens Uk

Security Tokens - A Primer - Early-stage Venture Capital...5.1 Debt Tokens 14 5.2 Hybrid/Convertible Tokens 16 ... jurisdictional financial regulations. These 2nd Generation tokens

TOKENS Saloni rajput

Tokens of Trust

2 Categories of Tokens · 2 Categories of Tokens Utility Tokens Digital coins used to transfer value or gain access to decentralized networks. Security Digital coins that Tokens represent

The JSON Data Interchange Format - VRML set of tokens includes six structural tokens, strings, numbers, and three literal name tokens. The six structural tokens: [ U+005B left square

UNDERSTANDING DIGITAL TOKENS

WozuKasusrektionauszeichnenbeiPräpositionen? · Abbildung1:Verteilung der Kasus über allen 1.164.766 Tokens der TUEBA. Kasuslose Tokens Kasuslose Tokens („-“)dominierenstark.DerhäufigsteKasusistDativ(Dat),gefolgtvonNominativ

JSON Web Tokens

Texas Trade Tokens

Trustee Tokens

Gold Coin Tokens

Platforms and Tokens

at43 tokens

Hard Times Tokens Part Two Store Cards. The concept of store card tokens goes back to the 1790s in England when British Trade tokens (aka Conder tokens)

iGO white-paper V3 V3... · 2019. 10. 15. · IGOS IGOS 1 IGOS = 0,1$ 100.000.000 IGOS TOKENS 80.000.000 IGOS TOKENS 5.000.000 IGOS TOKENS 75.000.000 IGOS TOKENS 7th October 2019,

Skeleton Tokens

Security Tokens

Advanced Tokens Manager

333PP Tokens

Sand Tokens Released