资源描述：
《语料库研究基本方法.ppt》由会员上传分享，免费在线阅读，更多相关内容在教育资源-天天文库。
1、语料库研究基本方法中国外语教育研究中心梁茂成主要内容语料库语言学的性质几个常用术语语料库研究的基本方法语料库语言学的性质1理性主义与经验主义Rationalism:IthinkthereforeIam.Empiricism:Mymindisa‘blankslate’.Seeingisbelieving.语料库语言学的性质1theWaxArgument:Heconsidersapieceofwax;hissensesinformhimthatithascertaincharacteristics,suchasshape,texture,size,color,smell
2、,andsoforth.Whenhebringsthewaxtowardsaflame,thesecharacteristicschangecompletely.However,itseemsthatitisstillthesamething:itisstillapieceofwax,eventhoughthedataofthesensesinformhimthatallofitscharacteristicsaredifferent.语料库语言学的性质theWaxArgument:Therefore,inordertoproperlygraspthenatureo
3、fthewax,hecannotusethesenses.Hemustusehismind.Descartesconcludes:“AndsosomethingwhichIthoughtIwasseeingwithmyeyesisinfactgraspedsolelybythefacultyofjudgmentwhichisinmymind.1语料库语言学的性质Empiricism:Empiricismemphasizesthoseaspectsofscientificknowledgethatarecloselyrelatedtoevidence,especial
4、lyasdiscoveredinexperiments.Itisafundamentalpartofthescientificmethodthatallhypothesesandtheoriesmustbetestedagainstobservationsofthenaturalworld,ratherthanrestingsolelyonreasoningandintuition.1语料库语言学的性质Scienceisconsideredtobemethodologicallyempiricalinnature.Corpuslinguisticsisempiric
5、alinnature.1语料库语言学的性质语言研究中的数据类型内省数据（introspectivedata）:rationalism实验数据（experimentaldata）:empiricism真实数据（anthenticdata）:empricism1语料库语言学的性质语料库语言学提倡真实数据我们不排斥其他数据类型1语料库语言学的性质即便在语料库语言学阵营之中Corpus-driven:minimumtheory-reliance.ExclusiverelianceoncorpusdataforalltheoriesCorpus-based:Relianceo
6、ncorpusdataforhypothesis-testingCorpus-referenced/informed:Occasionallyresortingtocorpusdataforillustrations1语料库语言学的性质我们坚决反对不顾语言事实的任何论断Nointrospectioncanclaimcredencewithoutverificationthroughreallanguagedata(Teubert2005).1几个常用术语2CorpusCorpuslinguistics几个常用术语Token,type,lemmaThelittlebo
7、ylookedattheotherboys.2几个常用术语Collocationisdefinedasasequenceofwordswhichco-occurmoreoftenthanwouldbeexpectedbychance.abigsmokerastrongsmokerahardsmokeraheavysmokerafurioussmoker2几个常用术语Itisquitepossible,infact,todescribeawomanashandsome.However,thisimpliesthatsheisnotbeautifulatallint
