A psycholinguistic analysis of speech errors in PSC共12页文档.docx

A psycholinguistic analysis of speech errors in PSC共12页文档.docx

《A psycholinguistic analysis of speech errors in PSC共12页文档.docx》由会员分享，可在线阅读，更多相关《A psycholinguistic analysis of speech errors in PSC共12页文档.docx（6页珍藏版）》请在冰豆网上搜索。

ApsycholinguisticanalysisofspeecherrorsinPSC共12页文档

ApsycholinguisticanalysisofspeecherrorsinPSC

　　PSC（PutonghuaShuipingCeshi，TheTestForMandarinLevel）isastandardtestoforalmandarininChina，whichconcludetesters’levelonphoneticsound，vocabularyandsyntax.OneofthetesttypesinPSCiscalled“Speech”，inwhichtestersaregivenatopicbysortitionandthenrequiredtomakea3minutes’speechaboutthetopic.Usuallytesterswillhaveseveralminutestoprepareforthespeech.

　　Previousresearchesshowsthattesteesalwaysmakemanyspeecherrorsin“speech”test，whichexcitedourinteresttoconsiderindetailaboutthepsychologicalreasonsoftheseerrors.IwillusetheclassificationbyClarket.，i.e.pauses，repeats，falsestates，non-repeatfalses，corrections，interjections，stuttersandslipsofthetonguetoclassifytheerrorsbythetestees，andanalysistheerrorsthroughtheformulatingandimplementingoflinguisticplans.ThesignificanceofthisresearchistofindoutthefeatureandpsycholinguisticreasonsofspeecherrorsinChineseandhelpPSCtesteestoperformbetterinthetest.

　　2Methodology

　　2.1Participants

　　Ihavemadeexperimentswithfourteenparticipantsfromtwenty-threetotwenty-sixyears’old，includingonemaleandthirteenfemales.AllofthemareuniversitystudentsandhavebeenraisedinMainlandChina，elevenofthemaremonolingualspeakersofMandarin，threebilingualinCantoneseandMandarin.fifteenoftheparticipantshavebeentakenpartinPSCbefore，theirPutonghualevelsrangefrom2-Bto1-B（including2-B，2-Aand1-B）.

　　2.2Procedure

　　IimitatetheprocedureoftheSpeechTestpartofPSC，firstofallIshowthespeechtopictotheparticipant，therearetwotopicsforeachparticipant，i.e.“我喜欢的明星或其他知名人士（Myfavouritestarorotherfamouspeople）”and“谈谈科技发展与社会生活（Talkaboutthedevelopmentofscienceanddailylife）”，eachtopicwasgivenseparately，andIwillleaveaspaceoftimebetweenthesetwotopicsincasetheperformanceoffollowedspeechoftheparticipantwillbeaffectedbytheprecedingone.

　　3ResultsandFindings

　　Whendecidingwhetheranutteranceisspeecherrorornot，firstofallIwillrefertotheinterviewIdidafterthespeechofparticipants，forexample，intopiconeparticipantXHrepresented“炮制女朋友（thenameofamovie）”into“泡沫女友”，duiringtheinterviewshesaidit'sbecausesheneverheardtheparticularnameofthatmovie，soIdidn'tcategorizethiserroras"speecherror".Iwillalsoconsidertheeffectofdialectwhencountingspeecherrors，thebilingualparticipantswhosefirstlanguageareMandarinandCantoneseinourexperimentsoftenmixtheusageofretroflexconsonants（zh，ch，sh）anddentalconsonants（z，c，s），andIexcludedtheseerrorssuchas[tsu51.mi？

耷35]（"famous"）→[t？

拶u51.mi？

耷35].　　Theamountofspeecherrorsineachtypeis：

　　perseveration-3（T1：

0，T2：

3），anticipation-3（T1：

2，T2：

1），perseveration/anticipation-4（T1：

4，T2：

0），exchange-1（T1：

0，T2：

1），shift-0，substitution-13（T1：

7，T2：

6），addition-28（T1：

13，T2：

15），deletion-13（T1：

7，T2：

6），blend-3（T1：

1，T2：

2）.Seethebarchartoferrortypedistributionbelow：

　　3.1Typesofspeecherrors

　　InthisarticleIborrowedCarroll'stypelistofslipsofthetonguetocategorizespeecherrors，accordingtothepracticalsituation，Ialsoaddonetype（i.e.Anticipation/Perseveration）intotheformereighttypes，sothereareninetypesofspeecherrorsinmycategorization.

　　3.2Findings

　　A.Inbothofthetwotopics，theamountofSubstitution，AdditionandDeletionismuchmorethanothertypesoferrors，thesethreetypesoferrorsoccupies19.1%，41.2%and19.1%amongalltheninetypesoferrorsindividually.Shiftneverhappenedinourspeecherrordata.

　　B.InAdditionandDeletionerrors，mostoftheerrorsarelexicalerrors（92.9%，84.6%lexicalerrros），inSubstitutionerrors，theamountofphonologicalerrors（69.2%）ismuchmorethanoflexicalerrors（30.8%），whileinAnticipationerrorsandPerseverationerrors，plusAnticipation/PerseverationerrorsandBlenderrors，alltheerrorsarephonologicalerrors.

　　C.Oppositetothediscussionthatphonologicalerrorsarethemajortypesoferrordistribution（Nooteboom，1973&Levelt，1989&Wan，2000），inthisresearchIfindthatinbothofthetwotopics，phonologicalerrorsarelessthanlexicalerrors（phonological44.1%：

lexical55.9%inT1，phonological32.4%：

lexical67.6%）.

　　3.3speechspeedcomparison

　　InthestatisticsoftheresultsfromExperimentIandExperiment2，thereare4subjectswhospokefasterinExperiment2thanthatofExperimentI，andIwillcalculateandanalysistherestofdatacomefromtensubjects.TheaveragespeedofExperimentIspeechis4.07w/s，whileExperiment2is3.56w/s.TheresultindicatesthatthespeedofExperimentIspeechisfaster