模板word.docx

模板word.docx

《模板word.docx》由会员分享，可在线阅读，更多相关《模板word.docx（18页珍藏版）》请在冰豆网上搜索。

模板word

Word-finalConsonantandClusterAcquisitioninIndianEnglish（es）*

CarolineR.Wiltshire

UniversityofFlorida

1.Introduction

Intheacquisitionofasecondlanguage,numerousfactorsinteractsimultaneously,includingthepatternsofthefirstlanguage（L1）,thepatternsofthetargetlanguage（L2）,universalsoflanguageacquisition,andtheamountandtypeofexposuretotheL2.EnglishislearnedasasecondlanguageinIndia,wherethefirstlanguagesoftheEnglishlearnersdifferwidelyinthephonotacticsofconsonantsandconsonantclusters.BecauseofthediversityofL1phonotacticsandtherelativeunityoftheL2target,variationsinIndianEnglishcanbeusedtotesttheeffectoftheL1phonotacticsontheacquisitionoftheL2,i.e.,theroleoftransfer.OTaccountshaveproposedameansformodelingtwootherimportantfactorsinL2acquisition,inadditiontotransfer:

markedness/universals,andinputfrequency（e.g.,Hancin-BhattandBhatt1997,Broselow,Chen,andWang1998,Hancin-Bhatt2000,BroselowandXu2004,Hansen2004,PengandAnn2004,Barlow2005）.ThevariationsinL2IndianEnglishword-finalconsonantandclusteracquisitionletustestthesuccessoftheOTapproach.

Inthisstudy,datafromspeakersoffivedifferentIndianL1swererecordedspeakingIndianEnglish,andtheirproductionofword-finalconsonantsandconsonantclusterswasexamined.TheirproductionsofthevoicingoffinalobstruentsdependedheavilyontransferfromtheirL1ortheemergenceoftheunmarked,withvoicelessobstruentsprevailingforspeakersoftheL1thatdidnotallowanyword-finalobstruents.Theirproductionoffinalconsonantclustersalsorevealedtheeffectsofmarkednessandtransfer,but,inaddition,suggeststhatL2speakersofEnglishtreatfinalclustersendingin/-s/asspecial,justlikeL1speakers.ApplyingtheGraduateLearningAlgorithm（Boersma1997,BoersmaandHayes2001）helpstoillustratethatthespecialtreatmentof/s/inclusterscannotresultfromfrequencyalone,supportingtheclaimthatCsclustersshouldbetreatedasspecialinL2aswellasL1phonology（Yildiz2005）.

2.DataCollection

DatawasgatheredinHyderabad,India,fromtwenty-fiveproficientspeakersofIndianEnglish.TheparticipantshadsimilarlevelsofeducationinEnglish,mosthavingbeguntolearnEnglishinnurseryoratthebeginningofschool（ages3-5）,andhavingattendedEnglishmediumschool.Theparticipantsweregenerallyintheirearlytomid-twenties,andwereattendingcollegeoruniversityinHyderabad,wherethemediumofinstructionwasEnglish.Thetwenty-fivespeakerscamefromfiveL1groups,fivespeakersineachgroup（seeAppendixAforfurtherdetailsonthespeakersandthelocationoftheirL1languages）.（对participants的介绍）

ThefiveL1sbelongtotwodistinctlanguagefamiliesspokeninIndia.GujaratiandHindi,spokeninnorthernIndia,arebothIndo-Aryanlanguages,distantlyrelatedtoEnglish,andwithsimilarphonotacticsforword-finalconsonantsandconsonantclusters（Mistry1997,Ohala1999）.Theotherthree,Angami,Ao,andMizo,arespokeninthenorth-eastpartofIndiainthestatesofNagalandandMizoram;theselanguagesarefromtheTibeto-Burmanlanguagefamilyandrequiremuchsimplersyllablestructure.AngamiallowsonlyCVsyllables,withnofinalconsonantsorclusters（Ravindran1974）.MizoandAoeachallowonlyasingleconsonantword-finally,limitedtoeitherasonorantoravoicelessobstruent（Gurubasave-Gowda1972,Chhangte1986,Lalrindiki1992,Coupe2003）.IndianEnglish,thetargetlanguage,allowsthesameconsonantsandclustersword-finallyasothervarietiesofEnglishsuchas*Iwouldliketothankalltheparticipantswhoprovidedthedata,aswellasPriyankooSarmah,RahulBalusu,K.Ashtamurthy,andK.G.VijayakrishnanofCIEFLfortheirhelpingatheringthedata,toTonyHung,BertusvanRooy,andtheGeorgeMasonUniv.SpeechAccentArchiveforwordlistsandsentencesfordatacollection,andtoEllenBroselowandZhengXuforthescriptsthatgottheGLAanalysisstarted.IalsothankparticipantsofBUCLD30fortheirhelpfulcomments,andapologizefornotyetactingonalloftheinsightfulsuggestionsoffered.2BritishorAmerican（CIEFL1972,Pandey1981）.Thephonotacticsofword-finalconsonantsinthesedifferentlanguagescanbesummarizedasinTable1:

L1sAngamiMizoAoGujaratiHindiIndianEnglishallowsfinalCnoyesyesyesyesyesallowsfinalvoicedObsnononoyesyesyesallowsfinalCCnononoyesyesyes

Table1:

Word-finalConsonantPhonotacticsofL1sandIndianEnglish（对L1的介绍，此部分可放入Introduction中）

Speakerswererecordedreadingaword-listandsentencestoelicitthegreatestlevelofaccuracyinproduction.Thewordlistconsistedof95wordsinisolation;thesentencesincluded98isolatedsentencesand29shortdialogues;theseweredesignedforexaminingthesegmentalinventoriesandphonotacticsoftheIndianEnglish.Therewasalsooneshortparagraphpassageforexaminingsegmentalandprosodiccharacteristicsinconnectedspeech（fromtheGeorgeMasonUniversitySpeechAccentArchive）.TheinterviewerwasaspeakerofIndianEnglish,totrytoreduceaccomodation.ThestimuliwererecordedonaDATrecorderandtransferredviaaCSLcomputerforanalysis（samplingrate=44.1kHz）.Transcriptionsweremadebyonephonetically-trainedresearcherofisolatedwords,keywordsfromisolatedsentencesanddialogues,andtheshortpassage.Oftheword-finalconsonantsexamined,18weresingleconsonants,includingvoicedandvoicelessobstruentsandvoicedsonorants,and28wereclustersgroupedintoeighttypes（seeTable2）,including6typesofCCand2typesofCCCclusters.TheCCconsistedofnasal-stop,lateral-stop,lateral-nasal,fricative-stop,stop-/s/,andstop-stop.TheCCCwereclassedasCCplus/s/vs.CCplusstop.

Word-finalConsonantsT

ypes（and#wordsif>1）Tokens（perL1group）

VoicelessStopsp

（2）,t5,ÿ/t（3）,tS,k

（2）45

VoicedStopsb

（2）,ê（3）,dZ,g（3）45

VoicelessFricativesf,s（4）,S30

VoicedFricatives1v

（2）,z（4）30

Sonorantsl,m,n,N20

Totals18types170tokens

Word-finalClustersTypes（and#wordsif>1）Tokens（perL1group）

Nasal-stopmp,nt/÷ÿ

（2）,ndZ,nd/÷ê

（2）30

Lateral-stoplt

（2）,ld,lp20

Lateral-nasallm5

FricativeStopst

（2）,sk,St20

Stop-sps,t5s,ÿs,bz,dz,gz30

Stop-Stoppt

（2）,kt

（2）20

CC-slts,mps,nts,fts,sks,sts,kts35

CC-Stoplpt,kst10

Totals28types170tokens

Table2:

Word-finalconsonantsandclusters,TypesandTokens,examinedforeachL1group

1Iexaminedthewords“raise,wise,please,organize”forword-final/z/.Thepluralmarkersisproducedas[s]inallcontextsbymanyIndianEnglishspeakersandthereforewasavoidedhere.However,astheproductionofthepluralasavoicelessfricativemaybeattributabletospelling,threeofthefourwordsusedheremayalsobeintentionallyproducedwiththevoicelessfricative/s/ratherthanbeingcasesofdevoicingofanintendedtarget/z/.Thereremains,however,alargedifferencebetweenthebehavioroftheTibeto-BurmanL1speakersandtheIndo-Aryan（seeTable3）,sothatatleastsomespeakersaretreatingtheseasfinal/z/.

3

3.Results

Theproductionsofword-finalconsonantsandclustersshowedcleardifferencesbasedontheL1

phonotactics;theseresultswillbedescribedfirstforsingleconsonants（3.1）andthenclusters（3.2）.

3.1Singleconsonants

Sonorantandvoicelessobstruent:

Allspeakers,fromallL1s,producedsonorantsandvoiceless

obstruentsconsistentlyandcorrectly.Theresultsforsonorantsarenotlistedindetailherebecausethey

areuniformlyaccurate,butnotethatevenspeakersofAngami,whichallowsnocodas,haveacquired

word-finalsonorantsatthispoint.

VoicedObstruents:

ThespeakersofTibeto-Burmanlanguages,whichallownocodas（Angami）or

obstruentcodasonlyifvoiceless（Ao,Mizo）devoicedword-finalconsonantsintheirEnglishfarmore

thandidspeakersofGujaratiandHindi,whichdoallowvoicedobstruentsincodapositions.Although

theGujaraiandHindispeakersdodevoiceoccasionally,theiroverallaveragesareconsistentwithEdge’s

（1991,citedinPengandLong2004）findingsof2-16%devoicingfornativeEnglishspeakersperforming

similartasks.SpeakersofallL1groupstendedtodevoicefricativesmorethanstopsinword-final

position.

AngamiAoMizoGujaratiHindi

Voicedstops（45）5894345

voiceless39373610

deleted10000

VoicedFricatives（30）0201517

voiceless2928301313

deleted10020

%voiceless:

stops86.782.2802.30

%voiceless:

frics96.793.310043.343.3

%voiceless:

overall90.786.78818.9217.3

Table3:

Word-finalVoicedObstruents:

numbersproducedvoicelessordeletedinEnglish,byL1groups

3.2Clusters

SpeakersoftheL1swhichdonotallowclusters（Angami,Ao,Mizo）deletemoreoftenthanspeakers

ofL1sthatdoallowclusters（Gujarati,Hindi）.Furthermore,notalltypesofclustersaresubjecttothe

sameratesofdeletion.Deletionswererarefornasal-stopandstop-fricativeclustersforallL1s,andmore

commonforfricative-stop.

Cluster（#oftokens）AngamiAoMizoGujaratiHindi

Nasal-stop（30）21420

Lateral-stop（20）22411

Lateral-nasal（5）12320

FricativeStop（20）73921

Stop-s（30）00220

Stop-Stop（20）991321

CC-s（35）181520125

CC-Stop（10）33720

Totals

%altered（outof170）

42

24.7%

38

20.6%

62

36.5%

25

14.7%

8

4.7%

Table4:

Word-finalClusters:

numbersproducedwithC-deletioninEnglish,byL1groups

4

4.Analysis

FormulatingtheanalysisasarankinginOptimalityTheoryallowsustoseethattheresultscanbe

attributedpartlytotransferandpartlytotheeffectofuniversals.As