文档下载 文档下载导航
设为首页 | 加入收藏
搜索 请输入内容:  
 导航当前位置: 文档下载 > 所有分类 > 工程科技 > 能源/化工 > Neural basis of semantic and syntactic interference in sentence comprehension 2
侵权投诉

Neural basis of semantic and syntactic interference in sentence comprehension 2

M.S.Nieuwlandetal./Brain&Language126(2013)151–158

Table2

Meanvaluesforrelevantlexicalandsententialvalues.ConditionAnimateCWInanimateCW

Lengthinletters6.3(1.8)7.1(2.2)

Log

frequency1.31(.64)1.28(.63)

Familiarity5.78(.81)5.73(.89)

Imageability5.58(.72)5.60(1.03)

Concreteness5.29(.72)5.21(1.07)

Relatedness(LSA-SSV).11(.12).12(.10)

Clozevalue(%)8.9(18.1)6.3(13.2)

Plausibility(1–7)4.80(1.13)4.87(1.28)

Grammaticality(after‘al/el’,%correct)93/89(12/17)92/94(14/12)

153

Note:Standarddeviationsaregiveninparentheses.Logfrequency,familiarity,imageabilityandconcretessweregeneratedfromDavisandPerea(2005).LSA-SSVsaresemanticsimilarityvaluesfromlatentsemanticanalysis,obtainedbycosinecomparisonofthecriticalwordvectorwiththesentencevector(sumofallword-vectorsineachsentence),usingtheGallito2.0Ósoftware(<doc.guandang.net).Clozevalueisthepercentageofresponsesinwhichtheanimate/inanimatewordswereusedtocompletesentencestruncatedafteral/el,respectively.Wealsocomputedpercentageofanimate/inanimateresponses(following‘al’M=91/9,SD=12/12,following‘el’M=7/93,SD=13/13),andsentenceconstraint(highestclozevaluefromallresponsespersentence;following‘al’M=35,SD=18,following‘el’M=36,SD=20).Plausibilityscores(1–7=implausible-plausible)aregivenforanimate/inanimatewordsfollowing‘al’and‘el’,respectively.Grammaticalityscoresarethepercentagecorrectresponsesfromthegrammaticalityjudgmentrating:whereasmarked-animateandunmarked-animatesentencesreceivedhigheroverallscores,nosigni cantinteractionbetweenanimacyandmarkingwasobtained(F<1).

2.3.Procedure

Participantsreadsentencesfromamonitor,presentedword-by-word(400ms,200msISI),self-pacedthroughbutton-press.Therewere54simpleyes/nocomprehensionquestions(halfrequiringa‘yes’button-pressresponse)distributedacrosssen-tencetypesthatprobedinformationotherthanabouttheobjectnoun.Usingfourlists,eachsentenceappearedinonlyonecondi-tionperlist,butinallconditionsequallyoftenacrosslists.Partic-ipantsthusread120experimentalsentences(30percondition)mixedwith30unproblematic llersinbreak-separatedsessions.Totaltime-on-taskwas40min.

2.4.Electroencephalogramrecording,dataprocessingandstatisticalanalysis

Fordetailsabouttheelectroencephalogramrecording,seeMar-tin,Nieuwland,andCarreiras(2012).Afteraverage-mastoidre-ref-erencingandICAocularartifactcorrection,waveformswereautomaticallyscreenedforartifacts(À150msto850msaftercrit-icalwordonset).Fourparticipantswereexcludedduetoexcessiveartifacts(trialloss>40%).Fortheremaining16participants,aver-ageERPs(baselinenormalized)werecomputedoverartifact-freetrialspercondition(averagepercentage=95%,range=95.2–95.6%acrossconditions).

First,usingaverageamplitudeperconditionacrossallEEGelec-trodesa2(objectmarking:correct,incorrect)Â2(animacy:ani-mate,inanimate)repeatedmeasuresanalysisofvariance(ANOVA)wasperformedinthe100–300,300–600and600–900mstimewindowsafterwordonset.Additionalanalyseswereperformedtoexplorethescalpdistributionoftheobservedeffectsandtorevealpotentialscalpdistributiondifferencesbetweenob-servedeffectsinsentenceswithanimateandinanimateobjects.Electrodesweregroupedintoquadrantsaccordingtoleft/rightofthemidlineandanterior/posteriortothecrossline(excludingmid-lineandcrosslineelectrodes).Foreachtimewindows,a2(objectmarking:correct,incorrect)Â2(animacy:animate,inani-mate)Â2(anteriority:anterior,posterior)Â2(hemisphere:left,right)ANOVAwasperformed.3.Results

CriticalwordselicitedinallconditionselicitedtheN1-P2-N400complexastypicallyseenforvisuallypresentedwords(seeFig.1),yetverydistinctERPeffectswereobservedasafunctionofmark-edness(Figs.2and3respectively):whereasunmarked-animateobjects(e.g.,‘elobispo’)elicitedanN400effectcomparedtomarked-animateobjects(e.g.,‘alobispo’)startingatabout300msafterwordonset(Fig.2),marked-inanimateobjects(e.g.,

‘alsuelo’)elicitedaP600effectcomparedtounmarked-inanimateobjects(e.g.,‘elsuelo’)thatstartedalreadyintheN400timewin-dow(Fig.3).

Noreliableeffectswereobtainedinthe100–300mstimewin-dow.Inthe300–600mstimewindow,therewasamarginallysig-ni cantobjectmarkingbyanimacyinteractioneffect(F1,15=3.64,p<.1)andafullysigni cantobjectmarkingbyanimacybyanteri-ority3-wayinteractioneffect(F1,15=13.25,p<.005).Follow-upofthelattereffectrevealedthattheobjectmarkingbyanimacyinter-actioneffectwasonlyreliableatposteriorelectrodes(F1,15=6.54,p<.05).ThisinteractionwasdrivenbymorenegativeERPsforun-marked-animateobjectscomparedtomarked-animateobjects(F1,15=4.92,p<.05,meandifference=1.00lV,SE=.45).Inaddi-tion,whileunmarked-animatesandmarked-inanimatesdidnotelicitarobustdifferentatposteriorchannels,theunmarked-inan-imatesdidelicitasigni cantlylargerN400thanthemarked-ani-mates(F1,15=10.84,p<.01,meandifference=1.06lV,SE=.32).Inthe600–900mstimewindow,therewasasigni cantobjectmarkingbyanimacyinteractioneffect(F1,15=5.38,p<.05)andamarginallysigni cantobjectmarkingbyanimacybyanteriority3-wayinteractioneffect(F1,15=3.43,p<.1;this3-wayinteractioneffectwasfullysigni cantwhenmidlineelectrodeswereincluded,F1,15=4.75,p<.05),signallingagainthattheobjectmarkingbyanimacyinteractioneffectwasmostpronouncedatposteriorelec-trodes(F1,15=6.51,p<.05),whilenotreliableatanteriorelec-trodes.Atposteriorelectrodes,noreliabledifferencewasfoundbetweenmarked-animateandunmarked-animateobject,whereasmarked-inanimateobjectselicitedmorepositiveERPsthanun-marked-inanimateobjects(F1,15=5.43,p<.05,meandiffer-ence=1.45lV,SE=.71).

Wealsoexaminedbrainresponsestothepresenceorabsenceofobjectmarking(‘al’and‘el’,respectively;thecorresponding gureisinthesupplementarymaterials).Nostatisticallysigni cantef-fectsofobjectmarkingwerefoundinanyofthesubsequent100mstimewindowsbetween100and500msafterarticleonset.

4.Discussion

WeusedERPstoexaminetheprocessingofcaseviolationsin-curredbyanimateandinanimatedirectobjectsinSpanish.Incor-rectlycase-markedinanimateobjects(e.g.,‘alsuelo’)elicitedacentral-posteriorP600effectcomparedtothesesameobjectswith-outmarking.However,animateobjectswithoutrequiredcasemarking(e.g.,‘elobispo’)elicitedanN400effectcomparedtothesameobjectswithcasemarking.Nooveralleffectofanimacywasobserved(althoughcorrectlyunmarkedinanimateobjectselicitedalargerN400thancorrectlymarkedanimateobjects),norwasthereareliableeffectofmarkednessonprecedingarticles.

上一页第2页

你可能喜欢

热门文档

相关文档

站点地图 | 文档上传 | 侵权投诉 | 手机版
本站所有资源均来自互联网,本站只负责收集和整理,均不承担任何法律责任,如有侵权等其它行为请联系我们.
文档下载 Copyright 2013 doc.guandang.net All Rights Reserved.  -  email