AnErrorAnalysisofJapanesescientists'
researcharticles
LeighMcDowell(B.App.Sci,M.Sci.Ed.)
DepartmentofLinguistics
FacultyofHumanSciences
MacquarieUniversity
April,2016
ThisthesisispresentedasapartialfulfilmentoftherequirementsfortheMastersof
ResearchinAppliedLinguisticsatMacquarieUniversity,Sydney,Australia.
TableofContents
Abstract.............................................................................................................................................vi
StatementofCandidate..............................................................................................................vii
Acknowledgements....................................................................................................................viii
Chapter1:Introduction................................................................................................................1
Chapter2:LiteratureReview......................................................................................................4
2.1EnglishasaGlobalLanguage..........................................................................................................5
2.1.1ThedevelopmentofEnglishLanguageTeaching............................................................................7
2.1.2EnglishforResearchPublicationPurposes.......................................................................................9
2.1.3JapanesescientistsandEnglishforResearchPublicationPurposes....................................10
2.2ErrorAnalysis...................................................................................................................................13
2.2.1ThedevelopmentofErrorAnalysis....................................................................................................13
2.2.2ErrorAnalysismethodology..................................................................................................................14
2.2.3LimitationsofErrorAnalysis.................................................................................................................17
2.2.4ErrorAnalysisandEnglishLanguageTeachingandresearch................................................18
2.3SystemicFunctionalLinguistics.................................................................................................20
2.3.1Rankscaleandsystemnetworks.........................................................................................................21
2.3.2SystemicFunctionalLinguisticsandErrorAnalysis...................................................................25
2.4CorpusLinguistics...........................................................................................................................27
2.4.1CorpusLinguisticsandacademicdiscourse....................................................................................27
2.4.1CorpusLinguisticsandErrorAnalysis..............................................................................................28
2.5ChapterSummary............................................................................................................................29
Chapter3:Methodology.............................................................................................................30
3.1IntegratedFrameworkofAnalysis............................................................................................31
3.2CorpusDesign...................................................................................................................................36
3.3ElaboratedErrorAnalysisProcedure......................................................................................39
3.3.1Recognitionandreconstructionoferrors........................................................................................39
3.3.2Classificationandquantificationoferrorsinnominalgroups................................................42
3.3.2.1Classificationoferrorsbylocationanderrortype...............................................................................42
3.3.2.2Errorswitharticlesandplural-s..................................................................................................................43
3.3.2.3Errorswithpreposition-of..............................................................................................................................46
3.4ChapterSummary............................................................................................................................48
ii
Chapter4:Findings......................................................................................................................49
4.2OverviewofErrorsinNominalGroups....................................................................................54
4.2.1Errorsincomplexandsimplenominalgroups..............................................................................57
4.2.2ErrorsinPre-modifier,Head,andPost-Modifier..........................................................................58
4.3ErrorPatternsinNominalGroups............................................................................................62
4.3.1Errorswitharticlesandplural-s.........................................................................................................63
4.3.1.1Singularreferenterrorswitharticlesandplural-s..............................................................................71
4.3.1.2Pluralreferenterrorswitharticlesandplural-s...................................................................................74
4.3.1.3Massreferenterrorswitharticlesandplural-s.....................................................................................77
4.3.1.4Summaryoferrorswitharticlesandplural-s........................................................................................79
4.3.2Errorswithpreposition-of.....................................................................................................................81
4.3.2.1Additionerrorswithpreposition-of...........................................................................................................86
4.3.2.2Selectionerrorswithpreposition-of.........................................................................................................87
4.3.2.3Omissionerrorswithpreposition-of..........................................................................................................90
4.3.2.4Orderingerrorswithpreposition-of..........................................................................................................91
4.3.2.5Summaryoferrorswith-of.............................................................................................................................93
4.4ChapterSummary...........................................................................................................................94
Chapter5:Conclusion.................................................................................................................96
References.......................................................................................................................................99
Appendices...................................................................................................................................107
AppendixA:EthicsApprovalLetter...............................................................................................107
AppendixB:ParticipantConsentForm........................................................................................109
AppendixC:ParticipantQuestionnaireandResponses..........................................................111
AppendixD:1,000-wordExcerpt....................................................................................................116
AppendixE:NominalGroupDenotationandQuantification................................................118
AppendixF:ThesisWordCount......................................................................................................120
iii
FiguresandTables
Figure2.1Distributionin2012offoreignstudentsintertiaryeducationbycountryof
destination...................................................................................................................................................8
Figure2.2Algorithmforthefirststageoferroranalysis................................................................15
Figure2.3CanonicalhierarchyofSystemicFunctionalLinguistics............................................21
Figure2.4SystemnetworkmappingtheEnglishpersonalpronounsystem.........................24
Figure3.5Systemnetworkforarticlesandplural-s........................................................................35
Table2.1Top20countriesbasedonpaperspublishedinThomsonReuters-indexed
journalsfrom1999-2009...................................................................................................................11
Table2.2Errortypesandlinguisticlevels............................................................................................16
Table2.3Rankscaleatlexicogrammarstratum.................................................................................22
Table3.4Pre-modificationpatternofcomplexnominalgroups.................................................33
Table3.5OverviewofCorpusofJapaneseScientificDiscourse(COJSD).................................38
Table3.6OverviewofNominalGroupSubcorpus(NGS)................................................................41
Table3.7Errorswitharticlesandplural-s...........................................................................................44
Table3.8Matrixforanalysisoferrorswitharticlesandplural-s..............................................46
Table4.9OverviewoferrorfrequenciesintheCorpusofJapaneseScientificDiscourse
(COJSD).......................................................................................................................................................51
Table4.10OverviewoferrorfrequenciesintheNominalGroupSubcorpus(NGS)...........55
Table4.11ErrorsinPre-modifier,Head,andPost-modifieraspercentageofnominal
grouperrors.............................................................................................................................................60
Table4.12Errorswitharticlesandplural-s........................................................................................65
iv
Table4.13Percentagedistributionoferrortypesinerrorswitharticlesandplural-s....68
Table4.14Relativefrequenciesofarticleandplural-serrors.....................................................70
Table4.15Errortypeswithinarticleandplural-serrorswithsingularreferents.............72
Table4.16Errortypeswithinarticleandplural-serrorswithpluralreferents.................75
Table4.17Errortypeswithinarticleandplural-serrorswithmassreferents...................78
Table4.18Errorswithpreposition-of....................................................................................................82
Table4.19Proportionoferrortypesinerrorswith-ofconstructions.....................................85
Table4.20Selectionerrorswithpreposition-of................................................................................89
v
Abstract
Forscientistsinthetwenty-firstcentury,amasteryofwrittenEnglishforResearch
PublicationPurposes(ERPP)isfundamentaltoprofessionalsuccess.FormanyJapanese
scientists,thisisasourceoffrustrationgiventheiruseofEnglishasaforeignlanguageis
subjecttoidiosyncrasies,whichmaybeperceivedbyeditors,reviewers,andreadersas
errorsdetractingfromtheimpactoftheirresearch.Asabasisforbuildingdata-driven
needs-specificpedagogicalsupport,thisstudyinvestigatesthemajorerrorpatternsin
Japanesescientists'writtenEnglish.Participantsinthestudyare13Japanesescientists
workinginthefieldofmaterialsscience.Theprimarydataaretheparticipants'scientific
researcharticlemanuscripts(i.e.,theresearcharticlesbeforepublication).An
elaboratedcorpus-assistedErrorAnalysis(EA)methodologyisemployed,investigating
errorpatternsthroughthelensofSystemicFunctionalLinguistics(SFL).Inshort,theEA
involvesthecollectionoftexts,identificationoferrorsinthetexts,andtheclassification,
linguisticdescription,andquantificationofthoseerrors,withtheseclassificationsand
descriptionselaboratedthroughtheSFLtheory.Forpurposesofscope,theinvestigation
focusesonerrorsinnominalgroups—akeyfeatureofscientificEnglish.Atotalof654
nominalgrouperrorsareidentifiedandanalysedaccordingtotheelaboratedEA
methodology.Fromtheanalyses,twomajorerrorpatternsemerge:errorswitharticles
andplural-s,anderrorswithpreposition-of.ResultshighlightthedifficultiesJapanese
scientistsfacewithboththepre-andpost-modificationofcomplexnominalgroups.
Particularly,theomissionoftheindefinitearticleawithsingularreferents,andthemisselectionofembedded-ofprepositionalphrasesinpost-modificationarerevealedas
dominanterrorsinthisstudy.Detaileddescriptionsofthetwomajorerrorspatterns
andtheirimplicationsalongwithrecommendationsforpedagogicalinterventionare
presented.
vi
StatementofCandidate
Icertifythattheworkinthisthesisentitled“AnErrorAnalysisofJapanese
scientists'researcharticles”hasnotpreviouslybeensubmittedforadegreenorhasit
beensubmittedaspartofrequirementsforadegreetoanyotheruniversityor
institutionotherthanMacquarieUniversity.
Ialsocertifythatthethesisisanoriginalpieceofresearchandithasbeenwritten
byme.AnyhelpandassistancethatIhavereceivedinmyresearchworkandthe
preparationofthethesisitselfhavebeenappropriatelyacknowledged.
Inaddition,Icertifythatallinformationsourcesandliteratureusedareindicated
inthethesis.
TheresearchpresentedinthisthesiswasapprovedbyMacquarieUniversity
EthicsReviewCommittee,referencenumber:ReferenceNo:5201500154onthe18th
May,2015.
________________________________
LeighMcDowell
StudentID:43828507
Date:26thApril,2016
vii
Noteverythingthatcanbecountedcounts.
Noteverythingthatcountscanbecounted.
MaximascribedtoWilliamBruceCameron,1963
Acknowledgements
Thisstudyisconcernedwiththecountingoferrors.Itisnottoassertthat
countingerrorsisallthatcounts.Whatreallycountsisthelearning,whichcannotbe
counted.Forguidingmesograciouslythroughthelearning,Ioffermysinceregratitude
tomysupervisor,DrCassiLiardet.ForsupportingmeallthewayfromJapantoSydney
andback,Ithankmyfamily.IwouldalsoliketothankProf.JimMartinforincludingme
inhisspeciallectureseriesattheUniversityofSydney;LianQianandthemembersof
theMacquarieUniversityLanguageTeachersandLearnersGroupforincludingmein
theirresearchcommunity;andMacquarieUniversityprofessorsandstaffStephen
Moore,AnnabelleLukin,PhilChappell,RobynBishop,andKylieDrake,fortheirkind
assistance.Lastbutnotleast,IexpressgratitudetotheJapanesescientistswho
participatedinthisstudy.Myhopeisthatthelearningfromthisthesismayhelpthem
alongtheirpaths.
viii
Chapter1:Introduction
Chapter1:Introduction
Thisstudyexaminesthesecondlanguage(L2)Englisherrorsinthescientific
writingofJapanesescientists.1Specifically,thisthesispresentsanelaboratedError
AnalysisframeworkfortheinvestigationofL2errors,withtheaimofrevealingthemost
prevalentandfrequenterrorpatternsinacorpusofscientificresearcharticleswritten
byJapanesescientists.Thisintroductorychapteroutlinesthemotivationsofthestudy,
beginningwithanoverviewoftheresearchcontext.
Englishisfirmlyestablishedinthetwenty-firstcenturyasagloballanguageand
isthelinguafrancainmanyinstitutionalanddisciplinarysettings(Crystal,2003;
Kirkpatrick,2007).Onedisciplinewherethisisreadilyapparentisthesciences.InJapan,
despitebeingaforeignlanguage,Englishistheprimaryprofessionallanguagefor
Japanesescientists,whoarethesecondlargestproducersofscientificpublications
behindtheUnitedStatesintherecentdecadebetween1999–2009(ThomsonReuters,
2009).Japanesescientistshavetraditionallyfavouredpublishingtheirworkin
internationalEnglish-mediumjournalsduetothehigherimpactfactorsrelativeto
Japanese-mediumjournals(Yamazaki,1995).However,thesubsequentlinguistic
demandsofpublishinginaforeignlanguagecanbeasourceoffrustrationleadingmany
Japanesescientiststoseekspecialisedlanguagesupport.
ThepersonalmotivationsforthisstudystemfrommyworkinEnglishLanguage
Teaching(ELT)inJapan,whereforthelastsixyearsIhavetaughtEnglishtograduate
studentsinthematerialssciencedepartmentofanationaluniversity.Inadditionto
teaching,IprovideEnglishlanguagesupportforPhDstudentsandprofessorsintheir
effortstopublishtheirscientificresearcharticlesininternationaljournals.Formanyof
theseearly-careerscientistsandestablished-careerprofessorsalike,thelevelsof
sophisticationandaccuracydemandedintechnicalpublicationspresentsignificant
challenges.TheJapanesescientistsIworkwithareconsciousthatwhenwritingin
Englishtheyarepronetousinglanguagethatmaybeabstruseorperceivedaserrors,
detractingfromthequalityoftheirworkandthemessagestheywishtoconveytotheir
peers.ThisconcernovererrorspromptsJapanesescientists,andoftentheirreviewers
1Inpractice,theseJapanesescientistsuseEnglishasaforeignoradditionallanguage(i.e.,EAL).
L2,asitisusedinthisthesis,encompassesthesedistinctions.
1
Chapter1:Introduction
andeditors,torequireextensiveproofreadingoftheirmanuscriptspriortopublication
(Wiley&Tanimoto,2012).Inmyrolesupportingearly-andestablished-career
materialsscientists,Ihaveproofreadmanymanuscripts,andoverthecourseofthese
activities,noticedpatternsoferrorsoccurringregularlyacrossalltexts.Theseactivities
motivatethisinvestigationtofirstly,elucidatethemostprevalenterrorsinorderto,
secondly,developaneeds-specific,focusedpedagogytohelpJapanesescientistsachieve
theprofessionalstandardsdemandedinscientificpublications.
Thetheoreticalmotivationsforthisstudyderivefromthreelinguistic
approaches:ErrorAnalysis(EA),SystemicFunctionalLinguistics(SFL),andCorpus
Linguistics(CL).Thefirstofthese,EA,wasdevelopedinthe1960s–70sasa
methodologyfortheinvestigationofL2errors,withtheoreticalrootsinearlySecond
LanguageAcquisition(SLA)research(Corder,1981,pp.35-36).EAhasmetwith
criticismsregardingitslimitationsandinrecentdecadeshasbeenmostlyabandonedby
SLAresearchers;however,ithasfoundwiderpracticalapplicationsinELTasahands-on
anddata-drivenapproachtotheinvestigationofarelevantandeverydayprofessional
concern(i.e.,errors;James,1998,p.x).PrimaryamongEA'slimitationsisitsrelianceon
traditionalsurfacegrammarforthelinguisticdescriptionandunderstandingoferrors
(Hamilton,2015;Schleppegrell,2002).Inordertoovercomethislimitation,thisstudy
integratesaSystemicFunctionalLinguistics(SFL)approachwithintheEAframework.
SFLoffersacomprehensiveviewoflanguageasitfunctionsincontextasameaningmakingresource(Eggins,2004,p.3).Inadditiontoenrichingtraditionalgrammarwith
delicategrammaticaldescriptions,SFLenablesanalysesalongfunctionallines,and
providesuniqueinsightintothesystem-structurerelationshipsofgrammar(Martin,
2013;Matthiessen&Halliday,2009).ByintegratingSFLwithintheEAframework,this
studybenefitsfromthedecadesoftheoreticaladvancesmadeinSFLsinceEAwas
developed.
WhileSFLprovidesthebasisforfine-grainedlinguisticanalyses,oneconsistent
criticismisitslimitationtoasmallnumberoftexts(Butler,1999).Toovercomethis
limitation,theEAframeworkinthisstudyisfurtherelaboratedbytheintegrationofa
CorpusLinguistics(CL)approach.CLenhanceslinguisticdescriptionsbyprovidingan
empiricalbasisforanalysesthroughcollectionsofreallanguagetexts,orcorpora
(Meyer,2002).Corporaaredesignedtofocusonaspecificpopulation(e.g.,Japanese
2
Chapter1:Introduction
scientists)oraspecifictexttypeorgenre(e.g.,researcharticles),andinthisway
providerepresentativelanguagesamples.Corpussoftwaretools,forexamplefrequency
andkeywordanalyses,canbeemployedfortheefficientanalysesoflargebodiesoftexts
(Evison,2010).Additionally,acorpus-assistedapproachmayemploymanualanalyses
oftextstodeducepatternsnotreadilyidentifiedbysoftwaretools(Park,2012).Inthis
study,aspecialisedcorpusisdevelopedfromthescientificresearcharticlesofJapanese
scientists(i.e.,theCorpusofJapaneseScientificDiscourse),andacorpus-assisted
approachisintegratedwithintheEAframework,enablingfrequencyanalysisandthe
identificationoferrorpatternsacrossmultipletexts.
Thisthesisisorganisedinfivechapters.Followingfromthisintroduction,
Chapter2presentsabackgroundtothetheoreticalandanalyticalframeworksofthe
studywithareviewofthemostrelevantliteratureintheareasofEA,SFL,andCL.
Chapter3describestheresearchdesignandmethodologyincludinganoverviewofthe
integratedframeworkofanalysisandelaboratedEAprocedure.Chapter4presentsthe
findingsoftheinvestigationoferrorsinthescientificwritingofJapanesescientists,
highlightingthemajorerrorpatternstoemergefromthedata.Theresultsand
discussionsarecombinedheretofacilitateanintegratedreading.Finally,Chapter5
concludestheinvestigationwithasummaryofthemajorresultsandcontributions,
alongwiththelimitationsanddirectionsforfutureinvestigation.
3
Chapter2:LiteratureReview
Chapter2:LiteratureReview
ThischapterpresentsabackgroundtotheinvestigationofL2Englisherrorsin
thescientificwritingofJapanesescientists.Thebackgroundcomprisesacomprehensive
reviewoftheliteraturemostrelevanttotheinvestigationandtheknowledgeintegralin
understandingthetheoreticalandanalyticalframeworks.Thechapterispresentedin
fourmainsections.Section2.1illustratesthepredominanceofEnglishasaglobal
language,andthesubsequentdevelopmentofthefieldofEnglishLanguageTeaching
(ELT),particularlytheemergenceofEnglishforResearchPublicationPurposes(ERPP).
Section2.2tracesthedevelopmentofErrorAnalysis(EA),amethodologyemployedin
theinvestigationofL2errors,fromitstheoreticalbeginningstoitspracticalapplication.
Section2.3presentsanoverviewoftheSystemicFunctionalLinguistics(SFL)approach
tolanguageanalyses,andtheresourcesthattheapproachoffersintheinvestigationof
L2errors.Finally,Section2.4reviewsthecomplimentaryapproachofCorpusLinguistics
(CL),identifyingitscontributiontounderstandingacademicdiscourseandinvestigating
errors.
4
Chapter2:LiteratureReview
2.1EnglishasaGlobalLanguage
ThedominanceofEnglishasagloballanguagehasbeenwelldocumented.As
earlyas1997,Englishhasbeendistinctlyrankedfirstamongtheworld'stenmost
influentiallanguages(Weber,1997).2In2003,Crystal(2003)estimatedthenumberof
Englishspeakersthroughouttheworldtobe1.5billion,3whichgiventheworld's
populationhadreachedsixbillion,indicatedthatoneinfourpeopleintheworldwere
abletouseEnglish"toausefullevel"(Crystal,2003,pp.68-69).Asof2016,anonline
dailyreportshowsEnglishcurrentlyaccountingfor53.6%ofwebpagesontheInternet,
withthenextnearestlanguages,RussianandGerman,at6.4%and5.7%,respectively
(W3Techs,2016).Furthermore,inastudyofthreemassivegloballanguagenetworks,
Ronen,Gonçalves,Hu,Vespignani,PinkerandHidalgo(2014)identifythecentralityof
Englishasagloballanguagehub.4Thisanalysismappedmillionsofonlineandprinted
instancesoflanguage,revealingthat,"theworld’slanguagesexhibitahierarchical
structuredominatedbyacentralhub,English"(p.E5622).
TheoriginsofthedominanceofEnglishcanbetracedbacktoBritishcolonialism
andmorerecently,itsexpansionintonewterritories,especiallybytheUnitedStates
(Phillipson,1992).ThesubsequentriseandsuccessesofAmericancapitalism,political
andmilitarypower,alongwiththeproliferationofitsmedia,sports,andtechnologyhas
furtheredthespreadofEnglishtoallcornersoftheglobe(Blake,Lass&Romaine,1992,
p.6;Pennycook,2014).ThespreadofEnglishthroughouttheworldhasledtoitsuseas
thelinguafrancaofmanyprofessions.Forexampleindiplomacy,theLeagueofNations
wasoneofthefirstmajorinternationalorganisationstoadoptEnglishtoconductits
proceedings(Crystal,2003,p.86).Englishhasbeenadoptedasthedefactolingua
2Languageswererankedviaapointsystemwithpointsassignedtoeachlanguageafter
weighingsixfactors:(a)numberofprimaryspeakers,(b)numberofsecondaryspeakers,(c)
numberandpopulationofcountrieswhereused,(d)numberofmajorfieldsusingthelanguage
internationally,(e)economicpowerofcountriesusingthelanguages,and(f)socio-literary
prestige.Thetoptenlanguagesinorderofrankingwere:(a)English,(b)French,(c)Spanish,(d)
Russian,(e)Arabic,(f)Chinese,(g)German,(h)Japanese,(i)Portuguese,and(j)Hindi(Weber,
1997).
3Approximately750millionFirst-andSecond-Languagespeakers,andanother750million
speakersofEnglishasaForeignLanguage
4Controlledforincomeandnumberofspeakers,thethreegloballanguagenetworkswere(a)
booktranslations,(b)multiplelanguageeditionsofWikipedia,and(c)multilingualtweetson
Twitter.
5
Chapter2:LiteratureReview
francaoftheAssociationofSouth-EastAsianNations(ASEAN)andisnowtheprimary
workinglanguageofEuropeanUnionorganisations,theUnitedNations,and
InternationalOlympicCommittee(Ammon,2006,p.321;Hagen,2016;Kirkpatrick,
2007,p.169).Similarlyinbusiness,multinationalcompaniesareincreasinglyinstituting
Englishas"thecommoncorporatelanguage"(Neely,2012,p.116).5Furthermore,in
marinetransportandcivilaviation,EnglishisthebasisofSeaspeakandAirspeak,
respectively,twocontrollednaturallanguagesdesignedforsimpleandeffective
internationalcommunication(Glover,Johnson,Strevens,&Weeks,1988;Robertson,
1988).OtherprofessionsinwhichlinguafrancaEnglishiswidelyemployedincludelaw,
medicine,nursing,pharmacology,engineering,computersandIT,tourismand
hospitality(Paltridge&Starfield,2013,pp.v-vi).
OneprofessionwhereEnglishdistinctlyoccupiesacentralroleisinthesciences.
BuildingontheearlysuccessesofBritishscientists,theriseoftheUSinthetwentiethcenturyboostedEnglishtotheforefrontofscientificlanguages,withthebacklash
againsttheGermanlanguageaftertheworldwarscementingitsposition(Ammon,
2001;Ferguson,2007;Gordin,2015).Inspecificscientificdisciplines,Englishhaslong
beenthepredominatelanguageofcommunication;forexample,asearlyas1997,98%of
GermanphysicistsclaimedEnglishastheirdefactoworkinglanguage(Graddol,1997).6
Moreover,in1995,Englishaccountedfor87.2%ofjournalpublicationsinthenatural
sciencesand82.5%inthesocialsciences(Ammon,2003,p.244).Morerecently,more
than95%ofnaturalsciencejournalsand90%ofsocialsciencejournalsarepublished
all,orinpart,inEnglish(Flowerdew,2013a,p.301).
Inthisway,Englishisfirmlyestablishedinthetwenty-firstcenturyasaglobal
languageandlinguafrancainmanydisciplinaryandinstitutionalsettings.This
widespreaduseofEnglishhasnecessitatedfurtherdevelopmentsinELTand
specifically,inresearchingthewaysinwhichEnglishisusedwithinvariousanddistinct
discoursecommunities.7
5Neely's(2012)examplesincludecompaniessuchas,Airbus,Daimler-Chrysler,FastRetailing,
Nokia,Renault,Samsung,Technicolor,Microsoft-Beijing,andRakuten(p.116).
6CloselyfollowedbyGermanchemists(83%),biologists(81%),andpsychologists(81%).
7Swales(1990)definesdiscoursecommunitiesassociorhetoricalnetworkswithcommongoals,
inwhichspecificlanguage,orgenres,areusedtocommunicateandrealisethosegoals(p.9).
6
Chapter2:LiteratureReview
2.1.1ThedevelopmentofEnglishLanguageTeaching.
Thepre-eminenceofEnglishasaprofessionalandscientificlinguafrancahasled
tothedevelopmentofasub-fieldofAppliedLinguisticsandspecialisedbranchofELT
knownasEnglishforSpecificPurposes(ESP).Dudley-EvansandStJohn(1998)define
ESPintermsof"absolute"and"variable"characteristics,assertingthatESPisdesigned
tomeetspecificlearnerneeds(pp.4-5).8ESPprepareslearnerstoentertargetdiscourse
communitiesbyfamiliarisingthemwiththedistinctlanguagepracticesthatdefinethose
communities(Basturkmen,2006,p.88).ThecentralfeatureofESPisafocusonthe
specialisedlanguage-learningneedsofpurpose-orientedlearners,whichistypically
identifiedthroughneedsanalyses(e.g.,Hutchinson&Waters,1987;Long,2006).
Increasingly,thesespecialisedneeds,andthechallengesforESPprofessionalstomeet
them,arearisingin"noncentercountries"whereEnglishisusedasaforeignor
additionallanguage(i.e.,EAL;Belcher,2006,p.150).9
AlongsidethegrowthinEnglishasagloballanguagehascomeincreasedmobility
ineducationandstudyabroad.Overthelast40years,thenumberofstudentsstudying
abroadfortertiaryeducationhasincreasedmorethanfive-foldfrom750,000in1975to
morethan4.5millionin2012(OECD,2014).Significantly,themajorityofthesestudents
arechoosingtostudyinEnglish-speakingcountries,particularlytheUSAandUK.As
illustratedinFigure2.1,thecollectivepercentageofstudentsstudyinginEnglishspeakingcountriesin2012is44%.
8AccordingtoDudley-EvansandStJohns'(1998)absolutecharacteristicsofESP,(a)ESPis
definedtomeetspecificneedsofthelearners;(b)ESPmakesuseofunderlyingmethodology
andactivitiesofthedisciplineitserves;(c)ESPiscentredonthelanguageappropriatetothese
activitiesintermsofgrammar,lexis,register,studyskills,discourseandgenre.Accordingtothe
variablecharacteristicsofESP,(a)ESPmayberelatedtoordesignedforspecificdisciplines;(b)
ESPmayuseadifferentmethodologyfromthatofGeneralEnglish;(c)ESPislikelytobe
designedforadultlearners,eitheratatertiarylevelinstitutionorinaprofessionalwork
situation;(d)ESPisgenerallydesignedforintermediateoradvancedstudents.
9"Noncentercountries"correlateswiththe"outer"and"expanding"circlesinKachru'sthree-
circlesmodelofEnglishasitisusedthroughouttheworld(seeforexample,Kachru,1992).
7
Chapter2:LiteratureReview
Othernon-OECD
countries17%
OtherOECD
countries8%
UnitedStates
16%
Belgium1%
Korea1%
Netherlands1%
Switzerland1%
SouthAfrica2%
NewZealand2%
Austria2%
Italy2%
China2%
UnitedKingdom
13%
Spain2%
Japan3%
Russian
Federation4%
Canada5%
Australia6%
Germany6%
France6%
Figure2.1Distributionin2012offoreignstudentsintertiaryeducationbycountryof
destination
Note.AdaptedfromOECD(2014,p.367).Inthischart,theEnglish-speakingcountriesaretheUS,
UK,Australia,Canada,NewZealand,andSouthAfrica.
Thisincreaseinstudentmobilityhasresultedinmillionsofstudentsreceivingtertiary
educationinEnglish,givingEnglish"acentralplaceintheprocessofconstructing,
disseminating,andlegitimizingknowledge"(Canagarajah,2002,p.6).
CoincidingwiththeboominEnglish-mediumeducation,EnglishforAcademic
Purposes(EAP)hasemergedfromthelargerfieldofESPasabranchofApplied
Linguisticsinitsownright.IncontrastwithESP,wherethefocusisonlearnersin
specificoccupations,EAPfocuseswhollyonacademiccontexts(Jordan,1997,p.3).
AccordingtoHyland&Hamp-Lyons(2002),
8
Chapter2:LiteratureReview
Themodern-dayfieldofEAPaddressestheteachingofEnglishintheacademyat
allageandproficiencylevels,anditdrawsonarangeofinterdisciplinary
influencesforitsresearchmethods,theoriesandpractices.Itseekstoprovide
insightsintothestructuresandmeaningsofacademictexts,intothedemands
placedbyacademiccontextsoncommunicativebehaviours,andintothe
pedagogicpracticesbywhichthesebehaviourscanbedeveloped.(p.3)
Inthisway,EAPfocusesonthediverseacademicgenreswiththeirspecialisedlinguistic
patternsdistinctfromeverydaysocialinteractions.Forexample,researchinEAPshows
thatacademicdiscourseischaracterisedasmorelexicallydenseandauthoritative,with
informationpresentedasstatic,synopticentitiesincause-and-effectrelationalnetworks
(Hyland,2009,p.7;Schleppegrell,2004,p.43-45).ThemasteryofthisformofEnglishis
essentialtothosewhodesiretoparticipateeffectivelyinacademicdiscourse
communities,andhasgivenrisetonotonlyEAP,butalsoanemergingsub-fieldknown
asEnglishforResearchPublicationPurposes(ERPP).
2.1.2EnglishforResearchPublicationPurposes.
InconjunctionwiththespreadofEnglishasagloballanguage,thetwenty-first
centuryhasseenariseinscholarlypublication.Asof2010,therewereanestimated5.5
millionscholars,2,000publishers,and17,500researchorhighereducationinstitutions
worldwidecontributingtoacademicwritingforpublication(Lillis&Curry,2010,p.1).
ThenumberofrefereedacademicjournalslistedinUlrich'sPeriodicalsDirectoryrose
from14,694in2001to25,864in2009,and67%oftheseacademicperiodicalswere
publishedinpartorfullyinEnglish(Mabe,2003,p.192;Lillis&Curry,2010,p.10).
Flowerdew(2013b)attributesthisincreaseinpublicationsto(a)moreuniversities,
morefaculty,moreresearchstudents;(b)universitiescompetinggloballytoproduce
more;(c)publicationbecomingarequirementfordegrees;and(d)Englishasalingua
franca.ThelastpointparticularlyleadsLillisandCurry(2010)tonote,"theevergrowingstatusofthejournalandjournalarticlesisparalleledbytheever-growinguse
ofEnglishasthemediumofsucharticles"(p.9).
ThisburgeoningenterpriseofresearchpublicationinEnglishandthesupport
fromELTprofessionalsthatitdemandshasledtoafurtherspecialisedfieldofinquiryin
AppliedLinguistics:ERPP.CargillandBurgess(2008)identifyERPPasabranchofEAP,
9
Chapter2:LiteratureReview
"addressingtheconcernsofprofessionalresearchersandpost-graduatestudentswho
needtopublishinpeer-reviewedinternationaljournals"(pp.75-76).Social
constructivismandsituatedlearningarethemaintheoreticalapproachesbeingapplied
inthisresearch,wherethefocusisonwriters'academicliteracypracticesand
experiences(Flowerdew,2013a,pp.314-315).ThiscontraststoworkinEAP,which
tendstofocusmoreonthetextualfeaturesofacademicEnglish(Curry&Lillis,2004,p.
664).ResearchinERPPhasfoundthatlanguageisnotalwaysthemajorimpedimentfor
publicationacceptance(e.g.,Belcher,2007;Rozycki&Johnson,2013),leadingERPP
practitionerstofocusspecialisedsupportinEnglishbeyondthediscourselevel,inareas
suchasstudydesign,literaturereview,andacculturationintoacademiccommunities.
2.1.3JapanesescientistsandEnglishforResearchPublicationPurposes.
WhileremainingontheperipheryforEnglishmediumcommunication,Japanhas
establisheditselfascentralinthefieldsofscienceandtechnology.Forexample,Japan
consistentlyranksamongthetopcountriesintermsofpatentsgrantedworldwide
(WIPO,2014),andasTable2.1illustrates,between1999-2009,wasthesecond-largest
producerofpublishedscientificpapersfollowingtheUS.
10
Chapter2:LiteratureReview
Table2.1Top20countriesbasedonpaperspublishedinThomsonReuters-indexed
journalsfrom1999-2009
Rank Field
Papers
1
USA
2,974,344
2
Japan
788,650
3
Germany
766,162
4
England
682,018
5
China
649,689
6
France
548,046
7
Canada
424,562
8
Italy
403,588
9
Spain
305,430
10
Australia
276,622
Note.Source:ThomsonReuters(2009)ScienceWatch
AcomparisonofJapanese-andEnglish-mediumjournalsrevealsthattheh5-index10of
thetop-100Japanese-mediumjournalsrangesfrom5–12citations,muchlowerthanthe
109–377rangeforthetop-100English-mediumjournals.11Forthisreason,Japanese
scientistshavetraditionallypreferredtosubmittheirpaperstoEnglish-medium
journals,andinparticular,tothehigh-impactjournals(Yamazaki,1995).Inthisway,for
manyJapanesescientists,ERPPisanessentialpartoftheirprofessionallife.
10Theh5-indexmeasuresthenumberofpublications,byanindividualororganisation,andthe
numberofcitationsperpublication.
11AsindexedbyGoogleScholar,2016April5th,
https://scholar.google.com.au/citations?view_op=top_venues&hl=en
11
Chapter2:LiteratureReview
ThechallengesthatariseforJapanesescientistsusingEnglishasbotha
professionallanguageandaforeignonehavemotivatedvariousERPPstudiesinrecent
decades.Forexample,Yamazaki(1995)examinestherefereeingsystemsofthejournals
mostpreferredbyJapanesescientists,reportingarejectionrateofoneintwopapers
(i.e.,49%;p.126).Similarly,Gosden(2003)examinesthelinguisticexchangesbetween
Japanesescientists,journaleditorsandreferees,andhighlightsinteractionaldeficiencies
inJapaneseresearchers'manuscriptsasamajorconcern.Inanotherstudy,Willeyand
Tanimoto(2012;2013)identifythatJapanesescientistsregularlyseekproofreadingof
theirEnglishmanuscriptsbeforesubmittingthemtointernationaljournalsfor
publication;andininvestigatingthestrategiesthatJapanesescientistsemployto
overcomethedifficultiesinmasteringscientificdiscourseinEnglish,Okamura(2006)
foundthatinadditiontoreadingacademictextswithintheirfield,directattentionto
masteringEnglishlanguageseemsto"payoffinthelongrun"forJapanesescientists(p.
77).Oneareawherethis"directattention"maybeapplicableiswithintheanalytical
frameworkofErrorAnalysis(EA),abranchofAppliedLinguisticsdedicatedto
identifyingdiscreteareasofweaknesswithinaparticulargenreordiscourse.
12
Chapter2:LiteratureReview
2.2ErrorAnalysis
EAisamethodologydevelopedearlywithinthefieldofSecondLanguage
Acquisition(SLA)tohelpresearchersrefinetheirtheoreticalfocusontheconstructthat
cametobeknownasinterlanguage(IL).IListhelearner'slanguageatagivenpointin
theirlearning;itissystematic,withrulesthatmayapproximatebutdifferfromthe
targetlanguage(Selinker,1972,1992).IL'ssignificanceinSLAhasbeenestablishedin
thecharacterisationofILasthe"disciplinarybeginnings"ofSLA(Ortega,2013,p.2).
Duringthesebeginnings,thesearchforacomprehensivetheoryofsecondlanguage
acquisitionwastheresearchthemebringingthefieldtogether,anditwasfromthis
investigationthattheEAmethodologywasborn.
2.2.1ThedevelopmentofErrorAnalysis.
ThedevelopmentofEAinthe1960–70swasconcurrentwithashiftinELTfrom
whatCandlin(inthePrefaceofRichards,1974)called"teacherascontroller"towardsa
more"learner-centredview"(p.iii).Basedonthehypothesisthaterrorsinlearners'L2
utterancesaresystematic,EAwasdevelopedforSLAresearcherstouncoverthe
systemsunderlyingL2errors(Corder,1967,p.163).WithinfluencefromChomsky,a
distinctionwasmadebetween"errorsofcompetence"and"errorsofperformance"
(Corder,1967,p.166-167).Errorsofcompetencearesystematic,notreadilyselfcorrected,and"provideevidenceofthesystemofthelanguagethat[thelearner]isusing
ataparticularpointinthecourse"(Corder,1967,p.167).Incontrast,errorsof
performancearerandom,readilyself-correctedandcanbedistinguishedasmistakes,
whichareofnoconcerninEA.ReflectingonthedevelopmentofEA,James(1998)
synthesisedtwodecadesofEAresearchtoderivethefollowingdefinitionforerrors:
Errorscannotbeself-correcteduntilfurtherrelevant(tothaterror)input
(implicitorexplicit)hasbeenprovidedandconvertedintointakebythelearner.
Inotherwords,errorsrequirefurtherrelevantlearningtotakeplacebeforethey
canbeself-corrected.(p.83)12
12Oneearlierdistinctiondenoteserrorsatthelevelofpragmaticsasinfelicities(Austin,1962).
PragmaticinfelicitiesareoutsidethescopeofEAandthisreview.
13
Chapter2:LiteratureReview
AsSLAresearchmovedawayfromcontrastiveanalysis,13EAwassubsequently
developedtofocuswhollyontheL2andthe"built-insyllabus"ofthelearner,reflecting
asCorder(1967)noted,amove"awayfromapreoccupationwithteachingtowardsa
studyoflearning"(p.163).
2.2.2ErrorAnalysismethodology.
TheEAmethodologywasfirstintroducedbyCorder(1967,1971)and
subsequentlyelaboratedbyothers(e.g.,Abbott,1980;Dulay,Burt&Krashen,1982;
James,1998;Levelt,1978;Richards,1974).Afterdataelicitation(i.e.,thecollectionofL2
texts),mostEAanalysescanbemappedalongthefollowingthreestages:(a)recognition
andreconstructionoferrors,(b)descriptionoferrors,and(c)explanationoferrors.In
thefirststage,errorsareidentified(i.e.,recognised)asidiosyncraticorunsuccessful
languageandrevised(i.e.,reconstructed)intowell-formedsentencesbasedonplausible
interpretation.14AnalgorithmproposedbyCorder(1971)forthisfirststageis
reproducedinFigure2.2.
13PriortothedevelopmentofEA,anothermethodologyknownascontrastiveanalysis(CA)was
thedominantparadigminthesearchtoexplainsecondlanguageacquisition.Basedonthe
languagetransferhypothesis(i.e.,linguisticsystemssimilarintheL1andL2areeasiertolearn,
andconversely,thosedifferentaremoredifficulttolearn),CAwasusedtopredictL2
acquisition.TheseL1-basedpredictionsweresubsequentlyusedtoinformlanguagesyllabiand
ELTpractices.However,thefailureofCAtoproduceaccurateandinformativepredictions,along
withitslinktothediscreditedbehaviourismtheory,ledSLAresearcherstoabandontheL1
contrast.
14PlausibleinterpretationoferrorscanbetakendirectlyfromtheL2,orback-translationfrom
theL1(Corder,1981,pp.21-24).
14
Chapter2:LiteratureReview
Figure2.2Algorithmforthefirststageoferroranalysis
Note.ThisalgorithmwasoriginallypublishedinCorder(1971);howevertheversionreproducedhereistakenfromare-printoftheoriginalarticleinCorder(1981,p.23).
InthisalgorithmandgenerallyinEA,thesentenceistheunitofanalysis.Everysentence
isregardedasidiosyncraticuntilshowntobeotherwise(Corder,1981,p.21).
ForthesecondstageofEA,thedescriptionoferrors,avarietyoferror
classificationsystemshavebeendeveloped.Theinitialclassificationisintermsofthe
linguisticlevel(e.g.,phonological,grammatical,lexico-semantic)andthewaysinwhich
theerrorsoccur(e.g.,omission,addition,etc.),asillustratedinTable2.2.
15
Chapter2:LiteratureReview
Table2.2Errortypesandlinguisticlevels
Graphological
Grammatical
Lexico-semantic
Phonological
Omission
Addition
Selection
Ordering
Note.ReproducedfromCorder(1981,p.36),thisframeworkservesonlyasvisual
representation;itisnotacompleteEAmodel.
ThefirstcolumnofTable2.2identifiesfourerrortypes.Dulay,etal.(1982)validated
theseerrortypesasthefourprincipalwaysinwhichILandL1diverge,inwhatthey
termedthesurfacestructuretaxonomy.James(1998)laterelaboratedthefourerror
types,re-labellingandaddingafifth,inwhathecalledthetargetmodificationtaxonomy:
(a)omission,(b)over-inclusion,(c)mis-selection,(d)mis-ordering,and(e)blends(p.
111).Thisinitiallevelofclassificationissimplythe"startingpointforsystematic
analysis",asfurtherlinguisticdescriptionisappliedtoindicatewhatisbeingomitted,
added,mis-selected,ormis-ordered(Corder,1981,p.36).Forexample,atthe
grammaticallevel,errorsarefurtheranalysed"intermsofsystems,suchastense,
number,mood,gender,case,andsoon"(Corder,1981,p.37).15
ThethirdstageoftheEAmethodologyistheexplanationoferrors.Thiswasthe
theoreticalraisond'êtreofthemethod.Inotherwords,itwastheexplanationoferrors,
ormorespecificallytheexplanationofthesystemsunderlyingtheerrors,thatmotivated
SLAresearcherstopursuethismethodtowardsanoverallunderstandingofIL.In
additiontoanearlierdistinctionbetweeninterlingualandintralingualerrors,16one
enduringframeworkfortheexplanationofthepossibleprocessesunderlyingILerrors
15Anumberofothererrortaxonomieshavebeendeveloped.Forexample,BurtandKiparsky's
(1972)errortaxonomydistinguishesinterlingual,intralingual,localandglobalerrors.
16FromCA,interlingualerrorsderivefromL1transfertotheL2,whileintralingualerrorsare
theresultofsystemswithintheL1(Richards,1971).
16
Chapter2:LiteratureReview
isRichards'(1971):(a)overgeneralisation,(b)ignoranceofrulerestriction,(c)
incompleteapplicationofrules,(d)falsehypothesis.Thetheoreticalunderstandings
thattheseexplanationsfosteredwerecentraltothedevelopmentofthefieldofSLA.
However,theapproachwasnotwithoutcriticismandlimitations.
2.2.3LimitationsofErrorAnalysis.
WhileEAwasintegraltoadvancingSLAtheoryintheearlystagesofthefield,it
wassoonmetwithcontentionandlimitations.OneearlycriticismoftheEAapproachto
understandingtheprocessesofSLAisavoidance(Schachter,1974;Schachter&CelceMurcia,1977).Learners'avoidanceofL2formscannotbeaccountedforbyEAsinceitis
notpresentinthedata.AsJames(1998)notes,thefirsterrortypeinthetarget
modificationtaxonomy,omission,couldactuallybe"non-acquisition"(p.107).Another
criticismofEAisthecomparativefallacyhypothesis(Bley-Vroman,1983,p.15).This
hypothesisassertsthatILisacompletelanguagewithitsownsetofrules,andnot
simplyadefectiveversionoftheL2.Fromthisperspective,EAisrejectedonthegrounds
ofitscomparisonwiththetargetlanguage(seealsothetargetdeviationhypothesis;
Klein,1998).Furthermore,whileCorder'smethodologysetthestandardforEA,
technicaldifficultieswithitsdescriptivebasiscreatedissuesofvalidity(Corder,1981,p.
24).Eventheveryname"erroranalysis"itselfbecamecontentious.17
Despitetheseshort-comings,EAhasalwayshadamore"practicaltask"in
guidinglanguagepedagogyandlearningintheclassroom(Strevens,1969,p.6).As
Corder(1981)explains:
Itisnowgenerallyrecognizedthatthatbranchofappliedlinguisticactivitywhich
isusuallycallederroranalysishastwofunctions.Thefirstisatheoreticalone
andthesecondisapracticalone.[...]Thepracticalaspectoferroranalysisisits
functioninguidingtheremedialactionwemusttaketocorrectanunsatisfactory
stateofaffairsforlearnerorteacher.(p.45)
Inthedecadesfollowingitsdevelopmentinformingtheoreticalanddescriptive
linguistics,thispracticalapplicationofEAhasbeenutilisedwidelyinELT.
17Inordertoavoidthe"target-languagebased"associations,Corder(1971)actuallypreferred
theterm,idiosyncraticdialectanalysis(p.61).Hammarberg(1974)movedtore-brandthe
approachasperformanceanalysis,takingintoaccounterrorsandnon-errors.
17
Chapter2:LiteratureReview
2.2.4ErrorAnalysisandEnglishLanguageTeachingandresearch.
WhileEAwasinitiallyestablishedasamethodfordevelopingSLAtheory,ithas
alwaysappealedtoELTpractitionersasa"methodologyfordealingwithdata"(Cook,
1993,p.22).Inotherwords,EAoffersteachersa"handson"approachtothe
investigationoferrors,arelevantprofessionalandeverydayconcern(James,1998,p.x).
ThisaspectofEAledtoitswidespreadapplicationinELT.ForexampleinTaiwan,in
ordertomeasurethepre-andpost-treatmenteffectsofacomputer-assisted-instruction
programonlearners'grammarskills,Chen(2006)performedanEAofL1Taiwanese
writtenessays.TheEAprocedurefollowedfourstagesandclassifiedtheerrorsinto15
categories.18Whiletheresultsindicatednosignificantdifferencebetweenthe
experimentalandcontrolgroups,theEAenabledtheidentificationofthemostsalient
errorcategoriesforpedagogicalintervention.
InIran,Khodabandeh(2007)usedanEAapproachintheinvestigationofPersian
cross-linguisticinfluenceonnewspaperheadlinetranslations.Inthisstudy,errorswere
investigatedatthreelinguisticlevels:grammatical,lexico-semantic,anddiscoursal.An
extensiveframeworkoferrorclassificationwasadoptedfromearlierwork(i.e.,Burtand
Kirparsky,1972;Keshavarz1993)andadaptedtofiteachofthethreelevelslinguistic
levels.Theresultsshowederrorsoccurredmostfrequentlyatthegrammaticallevel,
followedbythediscoursal-thenlexico-semanticlevels.Resultsweresubsequentlyused
infeedbackonsyllabusdesignandteaching.AlsoinIran,Falhasiri,Tavakoli,Hasiriand
Mohammadzadeh(2011)examinedtheeffectsofcorrectivefeedback(explicitand
implicit)onerrors,usinganEAmethodologytouncoverthemostfrequentlyoccurring
grammaticalandlexicalerrorsinL1Farsilearners'writtencompositions,whichwere
subsequentlytargetedforcorrectivefeedbackasatreatment.Pre-andpost-treatment
errorfrequenciesindicatedthatmosterrorswereintralingual,andthatexplicit
feedbackforinterlingualerrorswasthemosteffectiveformofcorrectivefeedback.
Moreover,inMalaysia,inastudyofcollocationerrorsinL1Malaysianlearners'
writtenessays,Hong,Rahim,HuaandSalehuddin(2011)employedacorpus-basedEA
18Chen's(2006)fourEAstagesare(a)datacollection,(b)identificationoferrors,(c)
classificationoferrorsintoerrortypes,and(d)statementoferrorfrequency.The15error
categoriesareerrorsin(a)nouns,(b)articles,(c)pronouns,(d)verbs,(e)prepositions,(f)
adjectives,(g)adverbs,(h)conjunction,(i)sentencefragments,(j)syntax,(k)lexicon,(l)
punctuation,(m)spelling,(n)capitalization,(o)subjectomission.
18
Chapter2:LiteratureReview
frameworkinfivestagestoexaminethetypesandsourcesofcollocationerrors,
specificallyverb-nouncollocationunits.19Withinthesecollocationunits,resultsindicate
thatpreposition-relatederrorsaremostfrequent,andintralingualerrorsaredominant.
Theresearchersusedthefindingstogenerateaseriesofrecommendationsforlanguage
pedagogytargetedatcollocationlearningforMalaysians.
AlthoughEAremainsaneffectivetoolforELTresearch,EAfindingscanonlybe
asvalidandeffectiveastheanalyticalframeworksandsystemsoflinguisticdescriptions
utilised.Hongetal.'s(2011)corpus-basedapproachhighlightsonepotentialareafor
enrichingthetraditionalEAframework(seeSection2.4).Onefurtherareaof
developmentwithintheEAliteratureinrecentyearsistheadoptionofaSystemic
FunctionalLinguisticapproachintheexaminationanddescriptionoferrors.Thislatter
approachisoutlinedinthefollowingsection.
19Hongetal.'sfivestagesofEAare(a)datageneration,(b)identificationoferrors,(c)error
classification,(d)errorquantification,and(e)analysisoferrorsources.
19
Chapter2:LiteratureReview
2.3SystemicFunctionalLinguistics
SystemicFunctionalLinguistics(SFL)isacomprehensivetheoryoflanguageasa
socio-semioticsystem(i.e.,languageasitfunctionsincontextasaresourceformeaningmaking;Matthiessen&Halliday,2009,pp.8-16).ByintegratingSFLdescriptionsintoan
EAanalyticalframework,researchintoerrorsbenefitsfromthedecadesoftheoretical
advancesmadeinSFLsinceEAwasdeveloped.Inadditiontofurtheringtraditional
grammaticaldescriptions,suchanintegrationenablesanalysisalongfunctionallines,
andprovidesinsightintotheparadigmaticsystemsofchoiceunderlyinggrammatical
structures.
SFLviewslanguageasastratifiedmodelacrossthreelevelsofexpression(i.e.,
phonologyandgraphology),lexicogrammar,anddiscoursesemantics,asillustratedin
Figure2.3.
20
Chapter2:LiteratureReview
Figure2.3CanonicalhierarchyofSystemicFunctionalLinguistics
Note.AdaptedfromMartin(2014,p.7).
Accordingtothismodel,sounds(i.e.,phonology)orletters(i.e.,graphology)combineto
formwordsorganisedbysystemsintostructures(i.e.,lexicogrammar),whichinturn
construemeanings(i.e.,semantics).Intheinternalstratumoflexicogrammar,grammar
andlexisareviewedasacontinuum(Matthiessen&Halliday,2009,p.10).Thisview
enableserroranalyststointegrategrammaticaldescriptionsoferrorswithlexicalones.
2.3.1Rankscaleandsystemnetworks.
ThelexicogrammaticalstratumcanbefurtherdividedandanalysedinSFLviathe
rankscaleandsystemnetworks.TheSFLrankscaleordersgrammaticalunitsintoa
hierarchywheregenerallyeachunitconsistsofunitsfromthenextrankdown
(Matthiessen&Halliday,2009,p.71).20Table2.3illustratestherankscalewithinthe
stratumoflexicogrammar.
20Embeddingofunitsfromthesamerankoraboveisalsocommon.Forexample,aprepositional
phrasemayfunctionwithinanominalgroup(e.g.,theethylenegaspressureeffectforthis
photoreaction).
21
Chapter2:LiteratureReview
Table2.3Rankscaleatlexicogrammarstratum
Rank
Example
Clause
/Thresholdvoltageinstabilitywasinvestigatedfor4H-SiCMOSFETs
withphosphorus-dopedandNO-annealedgateoxides/
Group/phrase
[Thresholdvoltageinstability][wasinvestigated][for4H-SiC
MOSFETswithphosphorus-dopedandNO-annealedgateoxides]
Word
[{Threshold}{voltage}{instability}]
Morpheme
{<Thresh><hold>}{<volt><age>}{<in><stabil><ity>}
Note.Inthedenotationofrankelements,theforwardslash(i.e.,/)signifiesthedivisionofa
clause,thesquarebrackets(i.e.,[and])signifythedivisionofagrouporphrase,thebrace(i.e.,{
and})signifythedivisionofaword,andthearrows(i.e.,<and>)signifythedivisionofa
morpheme(Thompson,1996,p.23).
Thehighest-rankingunitofanalysisinlexicogrammaristheclause,whichconsistsof
fiveclassesofgroups/phrases:21(a)nominalgroup,(b)verbalgroup,(c)adverbial
group,(d)conjugationgroup,and(e)prepositionalphrase(Halliday&Matthiessen,
2014,pp.362-363).Thegroup/phraserankconsistsofwords,andwordsarerealisedby
morphemes.Eachunitintherankscalecarriesdifferenttypesoflinguisticpatterns,or
structuralorganisations,forexampletheclauseunitisorganisedwithparticipants,
processes,andcircumstances,whilethegroupunitsisanexpansionofaheadword
(Eggins,2004,p.126).ByapplyingtherankscaleinEA,analystscanexaminetheways
inwhicherrorpatternsemergewithintheconstituentparts.Particularly,thenominal
groupisidentifiedinSFLasadefiningfeatureintheacademicandscientificregistersfor
21InSFLterms,thedistinctionbetweenagroupandphraseisthatagroupisanexpansionofa
word,whereasaphraseisacontractionoftheclause(HallidayandMatthiessen,2014,p.437).
22
Chapter2:LiteratureReview
itsroleindenselycompactingmeaningswithintheclause(e.g.,Halliday,1985,p.73;
Halliday&Martin,1993;Martin&Veel,1998).22
AnotherkeyfeatureofSFListhemappingofsystem-structurerelationships,
referredtoassystemnetworks.Systemnetworksareavisualrepresentationofthe
choiceswithinlanguageandtheirstructuralrealisations.Inotherwords,thestructures,
orgrammaticalforms,aretherealisationsofthechoiceswithinagrammaticalsystem.
Figure2.4demonstratesasystemnetworkmappingthepersonalpronounsystemfor
English.
22ThecentralityofthenominalgroupinacademicdiscoursehasbeenwelldocumentedinSFL
research.Forexample,Fang,SchleppegrellandCox(2006)demonstratehownominalgroups
expandtheclausetoincludemoreinformationinacademictexts.McCabeandGallagher(2008)
revealthecentralroleofthenominalgroupintheideationalmetafunctionofmeaningexpressioninundergraduateacademicwriting,andWhittaker,Llinares,andMcCabe(2011)
highlightthemanagementofnominalgroupsasanessentialskillindevelopingsuccessfulL2
academicwriting.
23
Chapter2:LiteratureReview
Figure2.4SystemnetworkmappingtheEnglishpersonalpronounsystem
Note.Thissystemnetwork,adaptedfromMartin(2013,p.23),operatesatthewordrankwithin
thelexicogrammarstratum.
FollowingSFLconventions(e.g.,Martin,2013;pp.13-18;Matthiessen&Halliday,2009,
pp.98-99),theentryconditionforthesystemnetworkisgivenonthefarleft-handside
(i.e.,personalpronoun),andtheboxonthefarright-handsideliststhenetwork's
structuralrealisations(e.g.,I,me,he,etc.).Thesystems,illustratedbyright-pointing
horizontalarrows,indicatethechoicesavailablewithinthenetwork.Thesystemnames
aregivenincapitalletters(e.g.,PERSON),andthesystemoptions,orfeatures,aregiven
inlowercase(e.g.,first,third).Thecurlybracket,orbrace,representssimultaneous
systems,indicatingthatthereisaconcurrentchoiceinthesystems,andthediagonal
arrowspointtotheformsthataretherealisationsofthesystems(e.g.,Iistherealisation
24
Chapter2:LiteratureReview
ofthechoicesoffirstinthePERSONsystem,singularintheNUMBERsystem,neuterin
theGENDERsystem,andnominativeintheCASEsystem).23
Inthisway,systemnetworksfacilitateaparadigmaticviewoflanguagesuchthat
grammaticalforms,andsubsequentlytheerrorsthatmanifestinthoseforms,canbe
examinedthroughtheirunderlyingsystems.
2.3.2SystemicFunctionalLinguisticsandErrorAnalysis.
Inrecentyears,thisSFLapproachhasbeenappliedintheexaminationoferrors
inL2texts.Forexample,Kim(2010)performedanSFL-inspiredEA,examiningKorean
students'translationsbyanalysingclausesaccordingtometafunctions(e.g.,withinthe
experientialmetafunction,errorswhereclassifiedaccordingtoParticipant,Process,or
Circumstance).Errors,definedas"problematicpartsintermsofaccuracyand
appropriateness"weremarkedandtalliedtoidentifyerrorpatternsinindividual
students'translationsfromL1KoreantoL2Englishandrevealareasforfocussed
improvement.Kim(2010)concludedthatitwaspossibletoclassifyerrorsbasedon
meaningusinganSFLapproachandthatthisknowledgecouldbeusedtoproducemore
systematicfeedbackonerrorsandcompetenceintranslationskills.
Similarly,Hamilton(2015)exploredthepotentialcontributionofSFLwithina
traditionalEAframework,examiningapilot-corpusoffirst-yearFrenchuniversity
students'essaysontwolevels:thetraditionalmorphosyntacticlevel,andtheSFLlevel
ofmetafunction.24Hamilton's(2015)findingsindicatethatwithinthefirst-levelof
grammaticalerrors,mosterrorsarelocatedinthenominalgroup,andinparticular,
manifestinarticle/determinererrors.Inthesecondlevelofanalysis,theresultsshow
thaterrorsoccurmostfrequentlywithinparticipantelementsintheideational
metafunction,whilewithinthetextualmetafunction,almostthree-quartersoftheerrors
23ThesuperscriptcapitalsIandTareabbreviationsofifandthen,respectively.Theysignalthat
ifafeaturewithsuperscriptIischosen(e.g.,neuterI),thenthereisonlyonecorresponding
choiceinanothersystem(e.g.,nominativeT)(i.e.,ifneuter,thennominative).
24Halliday'sparadigmaticviewofgrammarledhimtothediscoveryofthreesystemicclustersin
language,knowninSFLasmetafunctions,whichmapthethreedistinctwaysinwhichmeaning
inlanguageisconstrued.Thethreemetafunctionsare:ideational,interpersonal,andtextual
(Hasan,2014,p.14).
25
Chapter2:LiteratureReview
occurredwithintheRhemeelementoftheclause.25Hamilton(2015)concludedthatan
SFLintegratedapproachtoEAfostersfreshinsightintotheunderstandingofL2errors,
buildingonthetraditionalEAframework.
Furthermore,inacomparisonofL2Englishusers'andnative-Englishspeakers'
deploymentofgrammaticalresourcesinscientificlaboratoryreports,Schleppegrell
(2002)examinedlabreportsfromtwoangles:theclause-levellinguisticresources
deployedinmeaning-making,andthesentence-levelerrors.Errorsaredefinedasclause
formationsdepartingfromnative-speakernorms,andfoundtobedistractingtothe
instructorsevaluatingthelabreports.Findingsindicatethaterrorsoccurredmost
frequentlyinarticleusage,pluralandcount/massnounmarkingsystems,and
prepositionchoice,andhighlighttheneedtoassistL2writersindrawingonthe
appropriatelinguisticresourcestopresenttheirintendedmeanings(p.140).
ThissectionhaspresentedanoverviewoftheSFLstratifiedmodeloflanguage,
itsparadigmaticperspective,andnotableapplicationsinEAinrecentyears.Another
importantandoftencomplementaryapproachtolanguageanalysisisCorpusLinguistics
(CL).
25InSFL,theThemeisafunctionaltermforthetopicofaclause,andtheRhemeistheremaining
part(i.e.,nottheTheme),whichisgenerallyaround75%.Giventhisratio,Hamilton's(2015)
findingdoesnotdesignateanyspecialstatusforerrorstotheRheme.
26
Chapter2:LiteratureReview
2.4CorpusLinguistics
CLisasystematicapproachtotheanalysesoflanguageincollectionsof"real"
textsorcorpora(Meyer,2002,pp.xi-xii).Corporaprovideabasisfortheempirical
analysesoflanguage;forexample,frequencyanalysescanrevealtheprevalenceof
linguisticitemsacrosslargebodiesoftexts,andconcordancelinescanelucidatethe
contextsinwhichthoseitemsoccur(Evison,2010,pp.123-135).Thisdata-driven
approachallowsgeneralisationstobemadeacrosswholelanguages,andenablesthe
precisecharacterisationofspecificregisterswithinalanguage(e.g.,writtenscientific
discourse;Hunston,2002).
CLresearcherscandesigncorporatorepresentaspecialisedtext-typeorgenre
(e.g.,thescientificresearcharticle).Similarly,corporamayfocusonaspecific
populationoflanguageuserstocomparetheirlanguageusewithothers.Forexample,
learnercorporacanbeacollectionofL2textsproducedbyasampleorpopulationof
languagelearners,andmaybecomparedwithexpertorlargegeneralreference
corpora.26Inthisway,theCLapproachfacilitatesthequantitativedistinctionoflearner
constructionsfromexpertornativeones,enablingresearcherstopinpointtargetareas
forneeds-specific,data-drivenlanguagepedagogy(Johns&King,1991).Oneareawhere
theCLapproachtoexaminingrepresentativetextshasrevealedgreatinsightsiswithin
thedomainofacademicdiscourse.
2.4.1CorpusLinguisticsandacademicdiscourse.
CLresearchhascontributedsignificantlytocurrentunderstandingsofacademic
discourse,andsubsequentadvancesinthefieldofEAP.Forexample,theAcademic
WordList(AWL),nowwellknownandappliedinEAP,identifiesacoregroupofsubtechnicalwordsusedfrequentlyandonlyinacademicdiscourse(Coxhead,2000).27
Similarly,CLresearchhasprovidedempiricalevidenceforthegrammaticalfeaturesof
academicdiscourse,namelytheextensiveuseofdenselypackednominalgroups(Biber,
26Generalreferencecorpora,suchastheCorpusofContemporaryAmericanEnglish(COCA)and
BritishNationalCorpus(BCN)comprisingmillionsofwordsinthousandsoflanguage
instantiation(i.e.,texts)areoftenusedinCLtorepresenttheEnglishlanguageasitisgenerally
spokenorwritten.
27TheAWLisalistof570frequentlyoccurringwordscentraltoacademicEnglish,asrevealed
througha3.5million-wordspecialisedcorpus(Coxhead,2000).
27
Chapter2:LiteratureReview
1988,pp.28-29).Furthermore,accordingtoBiber,Johansson,Leech,Conrad,and
Finegan(1999),aroundsixtypercentofallcontentwordsinacademicdiscourseare
nouns(p.15).Othercorpusresearchershaveshownthatthisextensiveuseofnominal
groupsleadstosubsequentincreasesintheuseofdeterminersandprepositions(Aarts
&Granger,1998;Hunston,2002).Forexample,CLanalyseshaverevealedcommon
collocations,ormulti-wordunits,co-occurringinacademictexts,suchasthenominal
patternthe*of*(e.g.,thesetofrules)(Biberetal.,1999;Biber,2009).Inadditionto
mappingthelinguisticfeaturesmostprevalentinacademicEnglish,CLprovidesastrong
platformforinvestigatingerrors.
2.4.1CorpusLinguisticsandErrorAnalysis.
Thedata-drivenCLapproachhasbeenemployedintheinvestigationoferrorsin
recentyears.Forexample,Granger(2003)developedalarge-scalelearnercorpus,the
InternationalCorpusofLearnerEnglish(ICLE),witherror-taggingenablingquantitative
EAonamuchlargerscalethanpreviouslypossible.Similarly,Gressang(2010)
developedanounphraseerrortaggingsystemtoinvestigateerrorsinarticles,
determiners,andpronounsinalearnercorpusofessaysproducedbyL2English
learnersfromarticle-lessL1backgrounds(i.e.,Korean,Chinese,andTaiwanese).Her
studyrevealsasignificantunder-useoftheindefinitearticlea/an.Crompton(2005)also
developedalearnercorpustocontrastL1Malayspeakers'useofEnglishrelativisations
withthoseprevalentintwoexpertwrittenacademicEnglishcorpora,highlightingthe
over-useofwhere-constructionsinthelearnercorpusandunder-useofstandardforms
ofrelativisationcommoninwrittenacademicEnglish.Inanotherlearnercorpus
examination,Flowerdew(2006)compiledacorpusofargumentativeessayswrittenby
L1CantoneselearnersofEnglishtoinvestigatetheiruseofsignallingnouns.Basedon
frequencydatafromthecorpus,hederivedanerrortaxonomyofthevariouserrortypes
withsignallingnouns.Thesecorpusinvestigationsprovideauniqueinsightintohow
largecollectionsofauthenticdatacanrevealerrorpatternsrepresentativeofa
particularpopulation.
28
Chapter2:LiteratureReview
2.5ChapterSummary
ThischapterhasprovidedabackgroundreviewfortheinvestigationofL2errors
inthescientificwritingofJapanesescientists.Thisreviewhasmappedthe
predominanceofEnglishasagloballanguageandsubsequentdevelopmentsinELT,
situatingthisresearchwithinthefieldsofESP,EAP,andtheemergingsub-fieldofERPP.
Athoroughoverviewoftheanalyticalandtheoreticalframeworksoftheinvestigation
waspresented,byfirstfocusingontheEAmethodologythenonSFL,andfinallyonCL.
Thefollowingchapterwillelaborateonthesetheoreticalframeworkstodefinethe
methodologyemployedinthisexaminationofJapanesescientists'L2Englishscientific
writing.
29
Chapter3:Methodology
Chapter3:Methodology
ThisstudyemploysanelaboratedErrorAnalysis(EA)methodologyinthe
investigationoferrorsinthewritingofJapanesescientistsusingEnglishforResearch
PublicationPurposes(ERPP).Thiscorpus-assistedEAfocusesonerrorpatternsin
nominalgroupsthroughthelensofSystemicFunctionalLinguistics(SFL).Theresearch
designandmethodologyfortheinvestigationaredescribedinthreemainsections.The
firstsection(Section3.1)providesanoverviewofthetheoreticalandanalytical
framework;thesecondsection(Section3.2)describesthecorpusdesign,includingarich
descriptionoftheJapanesescientistsparticipatinginthestudyandthemethodofdata
collectionandpreparation;thethirdandfinalsectionofthechapter(Section3.3)
outlinestheelaboratedEAprocedure.
30
Chapter3:Methodology
3.1IntegratedFrameworkofAnalysis.
Thetheoreticalandanalyticalframeworkintegratedinthisstudyderivesfrom
threesources:EA,SFL,andCorpusLinguistics(CL).Thissectionprovidesanoverviewof
thesethreebases,beginningwithEA.
AsdiscussedintheLiteratureReview(Section2.2.2),EAresearchtypically
involvesathree-stageanalysis(Corder,1981,pp.21-25):
(1) Errorrecognitionandreconstruction
(2) Errordescription
(3) Explanationoferrors
Thepresentstudyfollowsasimilartrajectorytowardsidentifyingandexaminingerrors
intwostages:
(1) Errorrecognitionandreconstruction
(2) Errorclassificationandquantification
Thefirststagecorrespondswithproofreading,inwhicherrorsareidentifiedand
corrected(i.e.,recognisedandreconstructed).Inthesecondstage,errorsareclassified
andquantifiedtoelucidatethemostfrequentlyoccurringerrorpatternsinthedata.
Classificationisinitiallyintermsofthefourerrortypes:omission,addition,selection,
andordering(Dulay,Burt&Krashen,1982).Thisinitiallevelofclassificationisfurther
augmentedwithlinguisticdescriptionindicatingwhatisbeingomitted,added,misselected,ormis-ordered.Forthepurposesofthisstudy,thelinguisticdescriptionsdraw
fromanSFLmodeloflanguage.
SFLprovidesalensthroughwhicherrorscanbebetterunderstood.Asdiscussed
intheLiteratureReview(Section2.3),SFLviewslanguageasastratifiedmodelalong
planesofexpression,lexicogrammar,andsemantics(Martin,2014,pp.6-7).Generallyin
EA,thefocusofanalysisisatthe"grammatical"and"lexico-sematic"levels(Corder,
1981,pp.36-37).InSFL,thiscorrespondswiththestratumoflexicogrammar,which
placesgrammarandlexisonacontinuum;inotherwords,twopolesonasinglecline
(Halliday&Matthiessen,2014,p.24).Withinthelexicogrammarstratum,errorscanbe
examinedaccordingtowheretheyarelocatedwithintheclauseviatherankscale.The
31
Chapter3:Methodology
rankscaledividestheclauseintofourgroupclassesandonephraseclass:nominal
group,verbalgroup,adverbialgroup,conjugationgroup,andprepositionalphrase
(Halliday&Matthiessen,2014.pp.362-363).Forpurposesofscope,thepresentstudy
examineserrorpatternswithinthegroup-phraserank,specifically,withinnominal
groups.
InEnglish,andparticularlyinacademicandscientificregisters,muchofthe
meaninginlanguageispackedintonominalgroups.Inotherwords,intechnicalEnglish
theexpansionormodificationofnominalgroupsisanimportantlinguisticresourcefor
extendingtheclausewithdenselypackedmeaning(Halliday,1985,p.73;Halliday&
Martin,1993;Martin&Veel,1998).AllnominalgroupscontainaHead.28Insimple
nominalgroups,theHeadmayberealisedbyapronoun(e.g.,this),apropernoun(e.g.,
Mitsubishi),anumeral(e.g.,two),andincertaincasesanadjective(e.g.,complete).29In
complexnominalgroups,theHeadmayberealisedbyacommonnoun,whichis
modified.Typically,complexnominalgroupsarepre-modifiedbyaDeictic,30which
specifiesasubsetofareferent.Asmoremeaningispackagedintothegroup,additional
elementsmaybeaddedfollowingthepre-determinedpatternillustratedinTable3.4.
28FollowingSFLconvention,initialcapitallettersareusedtodistinguishfunctionaltermsfrom
traditionalgrammaticalclasses.
29AnEpithetmayactasHeadwhenthenominalgroupfunctionsintheclauseasAttribute
(Halliday&Matthiessen,2014,p.391).Forexample,completeinthefollowing:Aftersputtering,
thesol-gelprocessiscomplete.
30Deicticsintraditionalgrammarareoftensynonymouswithdeterminers;forexample,eachin
eachlayerwasprepared,orainalayerwasprepared.SimilarlyinSFL,theDeictichasa"pointing
out"function(Halliday&Matthiessen,2014,p.60).
32
Chapter3:Methodology
Table3.4Pre-modificationpatternofcomplexnominalgroups
Functional
Deictic
Numerative Epithet
Classifier
Head
nounor
noun
element:
Wordclass
determiner numeral
adjective
typicallyrealising
adjective
thefunction:
Example:
these
two
optically-
substance
syntheses active
Note.AdaptedfromHallidayandMatthiessen(2014,p.379).
AllelementsbeforetheHead(i.e.,Deictic,Numerative,Epithet,andClassifier)constitute
thenominalgroup'sPre-modifier.Tofurtherelaboratemeaninginnominalgroups,premodificationiscoupledwithpost-modification.Post-modificationofnominalgroupsis
typicallyrealisedbyembeddedprepositionalphrasesandrankshiftedclauses.
Sentences1and2demonstratepost-modification,withcomplexnominalgroups
denotedinsquarebracketsandthePost-modifiershighlightedinitalics.
1.
[Met-hemecoordination]contributesto[thestabilityofthestructure]and
[theabilityofelectrontransferincytcfamilyproteins].[05EST]31
2.
[IntegerphotonicspinobeyingBose-Einsteinstatics]confines[pluralspins
withthesamequantumstates].[02ESC]31
AsevidentinSentences1and2,thePost-modifierscompriseasubstantialportionofthe
nominalgroupinwhicherrorsmayoccur.Therefore,inthisframeworkofanalysisthe
locationoferrorsinthePre-modifier,Head,andPost-modifierprovidesanentrypoint
intotheclassificationoferrorsoccurringinnominalgroups.
Inadditiontotheerror'slocationwithinthegroup,theerrorclassificationcanbe
furtherelucidatedbySFLsystemnetworks.Systemnetworksprovideanefficientwayto
notonlyelucidateandvisualisegrammar(Section2.3.1)butalsotoanalysetheerrors
withinthosegrammaticalsystems.Figure3.5presentsthesystemnetworkmappedfor
31[05EST]and[02EST]arelabelsforcorpusextracts.SeeSection3.2fordetails.
33
Chapter3:Methodology
andappliedinthisstudyfortheanalysisofcomplexnominalgrouperrorswitharticles
andplural-s.
34
Chapter3:Methodology
specific
DETERMINATION
non-specific
complexnominal
group
singular
count
NUMBER
COUNTABILITY
plural
mass
thething
thethings
themass
athing
things
mass
Figure3.5Systemnetworkforarticlesandplural-s.
Thecomplexnominalgroupistheentryconditionforthissystemnetwork;inother
words,simplenominalgroupsdonotenterthesystem.Thenetwork'sstructural
realisationsaregivenintheboxonthefarright-handside.Thenetworkcomprisesthree
systems:DETERMINATION,COUNTABILITYandNUMBER.Thebinarychoiceswithin
thetwosystemsofDETERMINATIONandCOUNTABILITY—andsubsequentlywithin
theNUMBERsystemsinceitisdependentontheCOUNTABILITYsystem—are
simultaneousandcontingentoneachother.Therefore,errorsmis-construingreferent
specificity(i.e.,thevs.aand∅)areintricatelyboundwitherrorsmis-construing
countabilityandnumber(i.e.,singular,pluralormass),andarethereforeanalysed
simultaneouslyunderthisanalyticalframework.Inthisway,systems-structure
relationshipsvisualisedthroughsystemnetworksaffordauniqueviewinto
lexicogrammaranderrors.Tobroadenthatview,thethirdtheoreticalbasis,CL,is
employed.
CLenablestheidentificationoferrorpatternsacrossmultipletextswithina
samplepopulation.Suchempiricalinvestigationsmayrevealerrorpatternsdistinctto
certaindemographics;forexample,frequencyanalysescanrevealthemostprevalent
errorstypesandtheirhierarchicalordersinaspecialisedcorpusrepresentingatarget
population.Tothisend,thepresentcorpusisdevelopedfromacollectionofscientific
35
Chapter3:Methodology
researcharticleswrittenbyJapanesescientists.Fromthiscorpus,aspecialisedsubcorpusoferrorsiscompiledfromtheerrorsoccurringinnominalgroups.Adetailed
descriptionofthecorpusdesignfollowsinSection3.2.
3.2CorpusDesign
ThepresentstudyexaminesaspecialisedcorpusofJapanesescientists’English
writingforresearchpublicationpurposes:theCorpusofJapaneseScientificDiscourse
(COJSD).32Thecorpusisa46,263-wordcollectionofresearcharticlemanuscripts
writtenbythirteenJapanesescientistsworkinginthematerialssciencedepartmentofa
nationaluniversityinJapan.Thefieldofmaterialsscienceincludesworkinphysics,
chemistry,bioscienceanddeviceengineering.Toholdafacultypositioninthe
department,participantsareexpectedtohavetop-authoredatleastfivepapers
publishedininternationaljournalsandco-authoredfiveotherpapers.33Forpurposesof
thisstudy,onlythosefacultymemberswhoholdfacultypositionswererecruited,and
thereforetheseparticipantsareconsideredandhereinreferredtoasEstablishedCareer
Scientists(ECS).34EightofthethirteenparticipantsareAssistantProfessors,fourare
AssociateProfessors,andoneisaFullProfessor.35ECSarechoseninthisstudyover
early-careerscientistsbecausetheirgreaterwritingexperienceincreasesthelikelihood
oftheirerrorsbeingsystematic,ratherthanindividualmistakes.
AllthirteenparticipantsaremaleJapanesenationalslivinginJapan.Accordingto
theresultsofanonlinequestionnaireadministeredtotheparticipants,36atthetimeof
datacollectiontheseECSwereactivelywritingresearcharticlesinEnglishfor
publicationatleastonceayear,andsomemorethan2-3timesperyear(e.g.,Journalof
AmericanChemicalSociety,Angewandte,PLOSOne,etc.).Throughouttheseprocesses,
halfoftheparticipantshadbeeninformedthattheirmanuscriptsrequiredproofreading
32Acronympronouncedko-jay-es-dee.
33Withinthedepartment,internationaljournalsareconsideredasthosepublishedinEnglish.
ThisisincontrastwithjournalspublishedinJapanese,whichareintendedforadomestic
audience.
34Incontrastwithearlycareerscientistsorapprenticescholars(e.g.,PhDcandidates),whoare
typicallywritingtheirfirstpapers.
35IntheJapaneseacademiccontext,FullProfessoristhehighestacademicrank;Associate
ProfessorisonerankdownfromFullProfessor;followedbyAssistantProfessors.
36ThequestionnaireandresponsesarepresentedinAppendixC.
36
Chapter3:Methodology
forEnglishgrammarpriortobeingacceptedforpublication.Allthirteenparticipants
self-requestedEnglishproofreadingfromtheresearcherandagreedtotheir
manuscriptsbeinganalysedforthepurposesofthisstudy.
Theprimarydatainthisstudyaretheparticipants'scientificresearcharticle
manuscripts.Theresearcharticle(RA)isselectedasthe"pre-eminentgenre"of
academicandscientificdiscourse(Hyland,2010,p.117)andinthiswaythemost
representativeofscientificwriting.AmanuscriptisthecompleteRApriortosubmission
forpublicationandtherefore,atthispointintherevisionprocess,issolelythevoiceof
theauthorswithoutinputfromreviewersoreditors.WhilesomeoftheRAswerecowrittenwithotherresearchgroupmembers,inallcasestheparticipantisthetopauthorandprimaryauthorofthemanuscript.Onemanuscriptwascollectedfromeach
ofthethirteenparticipants,andthenreviewedforbasicgrammaticalcorrections,which
weremarkedupusingtheTrackChangestoolsinMicrosoftWord.OnecopywithmarkupwasreturnedtotheECS,andanotherwasfurtheranalysedusingthemethodology
mappedinSection3.3below.
Fortheprotectionofparticipants'anonymity,authors'names,affiliations,
acknowledgments,andreferencesareremovedfromthecorpus.37Otheridentifying
informationisreplacedwithgenericheadings(e.g.,Title,Keywords,etc.).Figurecaptions
andtabletitlesareleftintact,asthesecontainerrorsformingpartofthedata.For
labelling,eachtextisassignedanumberfromonetothirteencombinedwiththe
abbreviationforEstablishedCareerScientistsgiveninsquarebrackets(e.g.,[01ECS]).
Thislabellingisappliedwhenevercorpusexcerptsarepresented.Anoverviewofthe
COJSDwithwordandsentencecountsispresentedinTable3.5.
37Theseelementsdonotcontributeinanywaytothedata(e.g.,word-counts,etc.).
37
Chapter3:Methodology
Table3.5OverviewofCorpusofJapaneseScientificDiscourse(COJSD)
Text
Totalwords
Totalsentences
01ECS
3,458
190
02ECS
4,909
245
03ECS
2,881
154
04ECS
4,497
282
05ECS
4,517
238
06ECS
1,851
96
07ECS
2,376
160
08ECS
5,123
218
09ECS
4,137
315
10ECS
2,517
90
11ECS
1,679
92
12ECS
4,144
218
13ECS
4,174
211
46,263
2,509
3,559
193
Total
Average
Note.UAMCorpusTool(O'Donnell,2015)usedforwordandsentencecounts.
ThelengthoftheRAmanuscriptsvariesfrom1,679to5,123words,withanaverage
wordcountof3,559,andthetotalnumberofsentencesrangesfrom90to315,withan
averageof193.Toaccountfortherangeintextlength,resultsoftheanalysesare
normalisedtooccurrencesper1000wordsorpresentedasrelativepercentages.These
analysesaredetailedinthefollowingsection.
38
Chapter3:Methodology
3.3ElaboratedErrorAnalysisProcedure
ThissectiondescribestheelaboratedEAprocedureanddemonstrateshowthe
resourcesofSFLandCLhavebeenintegratedtoprovideamorein-depthandempirical
understandingofthemostcommonnominalgrouperrorsoccurringinJapanese
scientists’writing.Inthefirststage,errorsareidentifiedandreconstructedthroughthe
initialproofreadingprocess(Section3.3.1),andinthesecondstage,errorsareclassified
andquantified(Section3.3.2).
3.3.1Recognitionandreconstructionoferrors.
Oncethemanuscriptsarecollectedandanonymised,thetextsintheCOJSDare
proofreadforgrammaticalerrors.Errorsarerecognisedasunsuccessfuloridiosyncratic
languageinterruptingreadingflowandcomprehension.38FollowingCorder(1981),
theseerrorsarereconstructedintowell-formedsentencesbasedon"plausible
interpretation"elucidatedfromthecontext(pp.21-23).Anexampleofthisprocesscan
beseeninExcerpts3-4,whereExcerpt3representstheoriginaltextpriorto
proofreading,andExcerpt4illustratesthereconstructionofidentifiederrors(e.g.,the
acidgeneration,whichalsogiveriseto,etc.)
Beforeerrorrecognitionandreconstruction.
3.
Therefore,multiplestepsareinvolveduntiltheacidgeneration,whichalso
giverisetomanydecomposedfragmentsremainedinasystemforthese
classesofPAGs.[06ECS]
Aftererrorrecognitionandreconstruction.
4.
Therefore,multiplestepsareinvolveduntiltheacidgenerationisachieved,
whichalsogiveingrisetomanydecomposedfragmentsremainedingin
athesystemfortheseclassesofPAGs.[06ECS]
Duringthisproofreadingstage,thetextmark-upisannotatedusingtheTrackChanges
toolsinMSWord.Errorsarehighlightedinred,withstrikethroughandunderline
indicatingthereconstructions.
38ThisoperationalisationisasynthesisofpreviousEAresearch;especially,Corder(1981)and
James(1998)
39
Chapter3:Methodology
Oncetheerrorshavebeenidentifiedandreconstructed,theoverallfrequencyof
errorsinthecorpusasawholeisquantified.However,asdiscussedinSection3.1,a
completeandthoroughanalysisofalltheerrorsrecognisedandreconstructedinthe
COJSDisbeyondthescopeofthisthesis.Therefore,torefinethefocusofthestudy,a
subcorpusofnominalgroupswitherrorsisextracted.AnoverviewoftheNominalGroup
Subcorpus(NGS)ispresentedinTable3.6.
40
Chapter3:Methodology
Table3.6OverviewofNominalGroupSubcorpus(NGS)
Text
Approximate
Nominalgroups
nominalgroups
containing
pertext
errors
inCOJSD
inNGS
01ECS
562
54
02ECS
627
89
03ECS
495
78
04ECS
911
79
05ECS
731
10
06ECS
300
29
07ECS
380
20
08ECS
858
51
09ECS
831
50
10ECS
384
30
11ECS
278
32
12ECS
747
8
13ECS
776
68
7,879
598
606
46
Total
Average
Note.ThelabellingoftextsintheNGSremainsconsistentwiththewhole-corpuslabels(e.g.,
[01ECS]).Theapproximatenumberofnominalgroupspertext(i.e.,secondcolumn)isbasedon
thenumberofnominalgroupsina1,000-wordsamplefromeachtext(seeAppendixE).
41
Chapter3:Methodology
TheNGScomprisesallidentifiednominalgroupswitherrors(i.e.598),whichis7.58%
ofthetotalnumberofnominalgroupsintheCOJSD.Thesenominalgroupswitherrors
formthebasisfortheclassificationandquantificationoferrors,asdiscussedinthe
followingsections.
3.3.2Classificationandquantificationoferrorsinnominalgroups.
AftertheNGSiscompiledandtheoverallfrequencyoferrorsinnominalgroups
quantified,thenominalgrouperrorscanbefurtheranalysedaccordingtolocation(i.e.,
wheretheerroroccurswithinthenominalgroup)andtheerrortypes.Thefollowing
threesectionsoutlinethesesub-stagesintheclassificationoferrorsintheNGS.
3.3.2.1Classificationoferrorsbylocationanderrortype.
AsoutlinedintheIntegratedFrameworkofAnalysis(Section3.1)nominal
groupscanbesimpleorcomplex,withcomplexnominalgroupscontainingaHead,
whichmaybepre-andpost-modified.39Errorsincomplexnominalgroupscanbe
analysedaccordingtotheirlocationintheHead,Pre-,orPost-modifierasillustratedin
Excerpts5–7.
ErrorinPre-modifier.
5.
ASsimilarphenomenon[01ECS]
ErrorinHead.
6.
Thecompletecompletionofauto-activation[08ECS]
ErrorinPost-modifier.
7.
theethylenegaspressureeffecttoforthisphotoreactionin
methylcyclohexaneusingpressure-resistantreactionvessel[03ECS]
Theseerrorscanbefurtherclassifiedaccordingtothefourerrortypes:omission,
addition,selectionandordering(seeSection2.2.2andSection3.1).Forexample,in
Excerpt5,theerroroccursinthePre-modifier(i.e.,asimilarphenomenon)andcanbe
39Amongallerrorsinnominalgroupsidentifiedinthisstudy,onlyoneerrorwasfoundina
simplenominalgroup.ThisresultispresentedanddiscussedintheFindings,Section4.2.1.
42
Chapter3:Methodology
classifiedasanomissionerror,wheretheindefinitearticleaisomitted.InExcerpt6,
theerroroccursintheHead(i.e.,completecompletion)andcanbeclassifiedasa
selectionerror,wheretheadjectiveformcompleteismis-selectedforthenounform
completion.InExcerpt7,theerroroccursinthePost-modifier(i.e.,toforthis
photoreaction...)andcanbeclassifiedasaselectionerror,wheretheprepositiontois
mis-selectedfortheprepositionfor.
Oncetheerrorshavebeenclassifiedaccordingtoerrorlocation(i.e.,Pre-
modifier,Head,orPost-modifier)anderrortype(i.e.,omission,addition,selection,or
addition),theseoccurrencesareexaminedforfrequencytodeterminewhicherror
patternsemergeasthemostprevalentacrosstheNGS.Inparticular,theanalysisreveals
thatthemajorityoferrorsinnominalgroupsinvolvepre-modifyingarticlesandtheuse
ofplural-s.ThesecondmostprominenterrorpatternfoundwithintheCOJSDinvolves
thepost-modifying-of.Thefollowingtwosectionsdetailhowthesetwomajorerror
patternsarefurtheranalysed.
3.3.2.2Errorswitharticlesandplural-s.
Errorswitharticlesandplural-sareclassifiedbyerrortypeasoutlinedinTable
3.7.
43
Chapter3:Methodology
Table3.7Errorswitharticlesandplural-s
Excerpt
Error
anincreaseinhighorderoligomers
Reconstruction
Errortype
∅
a
Omission
bothendsofthe6-system[06ECS]
∅
the
Omission
thepolymerizationofmonomerswith
∅
plural-s
Omission
a
∅
Addition
the
∅
Addition
plural-s
∅
Addition
a
the
Selection
the
a
Selection
[05ECS]
theformationof1c[06ECS]
ayellowish,turbidPF8T2particles
[02ECS]
theparticlescattering[02ECS]
chemicalbiologyies,pharmaceuticals,
medicinalandagriculturalareas
[04ECS]
thediffractionfromathepentacene
film[01ECS]
theasurface[01ECS]
Note.∅indicatestheabsenceofanarticleorplural-s.
AsillustratedinTable3.7,articlesandplural-smaybeerroneouslyomitted,added,or
mis-selected.Notethatafourtherrortypeispossible(i.e.,orderingerrors),howeverno
suchoccurrenceswereidentifiedintheCOJSD.40
AsdiscussedintheIntegratedFrameworkofAnalysis(Section3.1),errorswith
articlesandplural-scanbeanalysedaccordingtothesystemnetworkinwhichthey
occur(Figure3.5).Inordertoelucidatepatterningwithintheerrorswitharticlesand
plural-s(i.e.,exactlywhatisbeingerroneouslyomitted,added,ormis-selected),these
40ThispointiselaboratedintheFindings(Section4.3.1).
44
Chapter3:Methodology
errorsareanalysedaccordingtothefivefeaturesofthesystemnetwork(i.e.,specific,
non-specific,singular,plural,andmass),asillustratedinTable3.8.
45
Chapter3:Methodology
Table3.8Matrixforanalysisoferrorswitharticlesandplural-s
Singular
Specific
1.
2.
3.
4.
Non-specific
theoptimizationofthewhole
proteinstructure
themuchhighercatalyticactivityof
thematureenzymethanthatof
zymogen
theYASARAstructuremolecular
modelingsoftwarepackage
theauto-oxidationofthecysteine
thiolgroupontheenzymes
1.
2.
thelargestbindingenergies
1.
2.
3.
4.
averysmallamount
acombinationofPAC-1andthe
maturecaspase-3
acobalt-chelatingaffinitycolumn
adecreaseintheconcentrationof2ME
Plural
1.
higherconcentrations
freeze-and-thawcycles
Mass
1.
2.
3.
thebindingofsyntheticmolecules
adockingsimulationforthebinding
ofN-[4-(2,3-diphenyl-3,4dihydropyrazol-5yl)phenyl]acetamidetoan
uncleavablemutantprocaspase-3
reportedbyClarkandco-workers
themodulationofproteinfunctions
Note.Excerptswithinthismatrixrepresentallarticleandplural-somissionerrorsfromone
text:[08ECS].
ThecolumnandrowheadingsofTable3.8derivefromthefeaturesofthesystem
network(Figure3.5)andclassifythereconstructedreferent(i.e.,nottheerroneous
referent).Therefore,thesecondcolumn(i.e.Specific)classifieserrorswithspecific
referents,andthethirdcolumn(Non-specific)classifieserrorswithnon-specific
referents.Thethreerowsenablesimultaneousclassificationforsingular,plural,and
massreferents.Thenumberingoferrorsineachcellsimplyfacilitatesquantification
withinthegiventext(i.e.,08ECS).TheresultsofthisanalysisarepresentedinSection
4.3.1. 3.3.2.3Errorswithpreposition-of.
Thesecondmajorerrorpatterntoemergefromthedatainvolvesthepreposition
-ofineitherthePost-orPre-modifiersofcomplexnominalgroups,asillustratedin
Excerpts8and9.
46
Chapter3:Methodology
Post-modifiererrorwithpreposition-of
8.
asuddenleapofinreactionconversion[03ECS]
Pre-modifiererrorwithpreposition-of
9.
thedepeakofde[03ECS]
Whiletheerrorswithpreposition-oftypicallyoccurinthePost-modifier,asshownin
Excerpt8,theyalsoinvolvepre-modificationinthereconstruction,asillustratedin
Excerpt9.Therefore,althoughtheerrorinExcerpt9occursinthePost-modifier,itis
classifiedasaPre-modifiererrorbecausethereconstruction(i.e.,thedepeak)iswithin
thePre-modifier.
Theerrorswithpreposition-ofarefurtheranalysedaccordingtothefourerror
types:omission,addition,selection,andordering,asillustratedinExcerpts10–13.
Omissionerrorwithpreposition-of
10.
knowledgeandunderstandingofanefficientgenerationandswitchingof
opticallyactivesubstances[02ECS]
Additionerrorwithpreposition-of
11.
theincidentX-rayangleofx-ray[01ECS]
Selectionerrorwithpreposition-of
12.
Onepossiblereasonofforthedecreaseindomainsize[01ECS]
Orderingerrorwithpreposition-of
13.
therefractiveindexofthecosolventsat589nm(nD)ofthecosolvents
[02ECS]
Theclassificationoferrorswithpreposition-ofcompletestheanalysesofthesecond
majorerrorpatternemergingfromthedata.Theresultsofthisanalysisarepresentedin
Section4.3.2.
47
Chapter3:Methodology
3.4ChapterSummary
Thischapterhaspresentedtheresearchdesignandmethodologyforthis
investigationintoJapanesescientists'writingforresearchpublicationpurposes.The
analyticalframeworkunderpinningthestudywasoutlined,highlightinghowthethree
theoreticalbases—EA,SFL,andCL—arecombinedtoprovideanintegratedframework
ofanalysis.ThischapterhasfurtherdescribedthedevelopmentoftheCOJSD,acorpusof
13JapaneseScientists’researcharticlemanuscriptssubmittedforproofreading(i.e.,
errorreconstruction),andthespecialisedNGS.TheelaboratedEAprocedurewas
outlinedintwostages:errorrecognitionandreconstruction,anderrorclassificationand
quantification.Withintheerrorclassificationandquantificationstage,theanalysiswas
furtherelaboratedaccordingtothelocationoftheerrorswithinthenominalgroups(i.e.,
Head,Pre-modifier,Post-modifier),theerrortypes(i.e.,omission,addition,selection,
andordering)andthetwomostprominenterrorpatternsidentifiedintheanalysis(i.e.,
errorswitharticlesandplural-s,anderrorswithpreposition-of).Theresultsofthese
analysesarepresented,andtheirfindingsdiscussedinChapter4.
48
Chapter4:Findings
Chapter4:Findings
ThischapterdetailstheinvestigationoferrorsintheEnglishwritingofJapanese
scientists.Thechaptercommenceswithanoverviewofthecorpusprovidingthebasis
foranalysesandafocusonerrorsinnominalgroupsbyexaminingnominalgrouptypes,
formsofnominalgroupmodification,andlocationoferrorswithinthenominalgroup.
Thechapterthenelaboratestheinvestigationsoftwomajorerrorpatternsemerging
fromthedata:errorswitharticlesandplural-s,anderrorswithpreposition-of.
Throughoutthechapter,theresultsanddiscussionsareintegratedtofacilitatea
comprehensiveoverviewoftheinvestigation.
49
Chapter4:Findings
4.1OverviewofErrorsintheCorpus
ThisexaminationoferrorpatternsintheL2writingofJapanesescientists41
commenceswithabroadoverviewofthekeyfeaturesoftheCorpusofJapanese
ScientificDiscourse(COJSD)andtheerrorswithinthecorpus.TheCOJSDcomprises13
textswithatotalof46,263wordsin2,509sentences.AsdescribedintheMethodology
(Section3.3),thethirteentextscollatedintheCOJSDareproofreadforgrammatical
errors;thesereconstructedtextsthenserveasthedataforthepresentanalysis.Table
4.9identifiesthenumberofwords,sentences,andimportantlytheprevalenceofthe
errorsineachtextintheCOJSD.
41AsnotedinChapter1,L2encompassesEnglishasaforeignandadditionallanguageinthis
thesis.
50
Chapter4:Findings
Table4.9OverviewoferrorfrequenciesintheCorpusofJapaneseScientificDiscourse
(COJSD)
Percentage
Sentences
Text
Words
Sentences
with
pertext
pertext
errors
witherrors
errors
Total
per1,000
(%)
errors
words
01ECS
3,458
190
68
35.79
130
38
02ECS
4,909
245
80
32.65
179
36
03ECS
2,881
154
77
50.00
171
59
04ECS
4,497
282
96
34.04
138
31
05ECS
4,517
238
34
14.29
39
9
06ECS
1,851
96
40
41.67
63
34
07ECS
2,376
160
30
18.75
40
17
08ECS
5,123
218
76
34.86
109
21
09ECS
4,137
315
70
22.22
96
23
10ECS
2,517
90
36
40.00
89
35
11ECS
1,679
92
51
55.43
80
48
12ECS
4,144
218
27
12.39
34
8
13ECS
4,174
211
117
55.45
200
48
46,263
2,509
802
1,368
3,559
193
62
105
31
Total
Average
Total
sentences
51
31.96%
Chapter4:Findings
WithintheCOJSD,atotalof1,368errorswererecognisedandreconstructedin
802sentences,withanaverageof31errorsper1,000words.42Thenumberoferrors
pertextspansfrom8to59per1,000words,reflectingtherangeintheparticipants'
writtenEnglishproficiency.Intermsofsentenceswitherrors,onaveragearound32%
ofallsentencesinthedatacontainederrors,oraroundoneineverythreesentences
(Table4.9).ThesentencesinExcerpts1to4,extractedfromtheCOJSD,illustratethe
variationwithwhichdifferenterrortypesoccurwithinthevariousclauseelements.43
1.
Forhigh-yieldingreductionofazides,manyconditionshavebeenappeared,
e.g.zincmetal/aceticacid,hydrogenolysis,Staudingerreactions,andother
reducingagentsshowninScheme35.[04ECS]
2.
Moleculardynamics(MD)simulatesimulationssuggestedthreepossible
bindingpaths.[08ECS]
3.
Thismeansthatthemoststableconformerinthediastereoselective
photoreactionofcyclohexenonecanbestabilizedefficientlybythe
clusteringeffectinthegroundstate,andmayaffordhighlydevalues.
[03ECS]
4.
Theweakcorrelationbetweenstructuraldissymmetryandthe
luminescencedissymmetryfactorimpliesaweakcontributescontribution
ofthestatic-couplingmechanismtothecircularlypolarizeddissymmetryof
thelanthanide(III)f-ftransition.[10ECS]
ThefirstfeatureofExcerpts1–4isthatwhiletheyachievetheacademicregister
ofscientificdiscoursewithsophisticatedlinguisticresources,theyallcontainerrors
(highlightedinred)interruptingthereadingflowandcomprehension.InExcerpt1,
thereisanadditionerroroftheauxiliaryverbbeintheverbalgroup.InExcerpt2,there
isaselectionerrorinthenominalgroup,wheretheverb-formsimulateismis-selected
forthenoun-formsimulation.InExcerpt3,twoerrorsoccurwithinasinglenominal
group:themis-selectionoftheadverb-formhighlyforadjectiveformhigh,andthe
omissionofplural-s.Finally,inExcerpt4therearefourerrors:(1)theadditionofthe
42AsamplefromaCOJSDtextdemonstratingtheaverage31errorsper1,000wordsis
presentedinAppendixD.
43Extractsarelabelledwithtextnumbers(e.g.,04ECS)asdescribedintheCorpusDesign
(Section3.2).
52
Chapter4:Findings
definitearticletheforanon-specificreferentinonenominalgroup(i.e.,the
luminescence),(2)theomissionofboththeindefinitearticleaand(3)thedefinitearticle
theinanothernominalgroup(i.e.,aweakcontribution,andthestatic-coupling
mechanism),and(4)themis-selectionoftheverbformcontributesforthenounform
contribution.
AnotherfeatureofExcerpts1–4andacrosstheCOJSDistheoccurrenceof
multipleerrorswithinsinglesentences.Excerpts1and2containsingleerrors,while
Excerpts3and4containmultipleerrors(i.e.twoandfourerrors,respectively)typical
oftheCOJSD.AccordingtothetotalsinTable4.9,amongthesentenceswitherrorsthere
areanaverageof1.7errorspersentence.
OnefurtherfeatureofthesentenceswitherrorsinExcerpts1–4andthroughout
theCOJSDisthatwhileerrorsmayoccuracrossdifferentclauseunits,(i.e.,inverbal
groups,nominalgroups,adverbialgroups,conjugationgroups,andprepositional
phrases)theytendtooccurmorefrequentlyinnominalgroups.Ashighlightedinthe
LiteratureReview(Sections2.3.1and2.4.1)andMethodology(Section3.1),the
packagingofmeaningintonominalgroupsisacentralfeatureofacademicandscientific
discourse.Asnominalgroupsareelaborated,andbecomedenser,thereisgreater
potentialforerrorstooccur.Giventhisroleofnominalgroupsandtheirsignificancefor
meaning-makingwithinacademicandscientificdiscourse,44thenominalgroupwas
selectedasthefocusforinvestigationinthisstudy.Thefollowingsectionsexaminethe
differentpatternsoferrorsfoundwithinnominalgroupsintheCOJSD.
44Seeforexample:Biber,1988,pp.28-29;Fang,Schleppegrell&Cox,2006;Halliday,1985,p.
73;Halliday,2004,pp.75-79;Halliday&Martin,1993;Martin&Veel,1998;McCabeand
Gallagher,2008;Whittaker,Llinares,andMcCabe,2011.
53
Chapter4:Findings
4.2OverviewofErrorsinNominalGroups
Whilenominalgroupsareakeylinguisticresourceformeaning-makingin
academicregisters,theyarealsoamajorsourceoferrorsintheCOJSD.Thissection
providesanoverviewoferrorsinnominalgroups,thetypeofnominalgroups
containingerrors(Section4.2.1),andthelocationoferrorswithinthegroup(Section
4.2.2).
Tobeginthisexamination,nominalgroupswitherrorsarecollatedintoasub-
corpus(seeSection3.3.1).WithintheNominalGroupSubcorpus(NGS),atotalof654
errorsareidentified,comprisingalmosthalf(47.81%)ofthetotalerrorsintheCOJSD,as
illustratedinTable4.10.
54
Chapter4:Findings
Table4.10OverviewoferrorfrequenciesintheNominalGroupSubcorpus(NGS)
Nominal
groups
containing
errors
Text
Nominal
groups
containing
errorsas
percentageof
approximate
nominal
groupsin
COJSD
(%)
Totalnumber
oferrorsin
nominal
groups
Nominalgroup
errorsas
percentageof
totalerrorsin
COJSD
(%)
Nominal
grouperrors
per1,000
words
01ECS
54
9.60
72
55.38
21
02ECS
89
14.20
98
54.75
20
03ECS
78
15.77
79
46.20
27
04ECS
79
8.67
83
60.14
18
05ECS
10
1.37
12
30.77
3
06ECS
29
9.66
29
46.03
16
07ECS
20
5.27
22
55.00
9
08ECS
51
5.95
54
49.54
11
09ECS
50
6.02
54
56.25
13
10ECS
30
7.82
40
44.94
16
11ECS
32
11.50
34
42.50
20
12ECS
8
1.07
8
23.53
2
13ECS
68
8.77
69
34.50
17
598
654
Total
Average
46
7.59%
50
55
47.81%
15
Chapter4:Findings
Notably,thishighproportionoferrorsinnominalgroupsispredictablegiven
thattechnicalmeaningisfrequentlyrealisedinnominalgroups,andprepositional
phrasesalsoconsistmostlyofembeddednominalgroups.Illustratingthissecondpoint,
Excerpt5demonstratesoneprepositionalphraseconsistinglargelyofoneembedded
nominalgroup(i.e.,thevariousmobilityenhancingfabricationtechniques).
Prepositionalphrasewithembeddednominalgroup
5.
Amongthevariousmobilityenhancingfabricationtechniques[01ECS]
WithintheNGS,654errorswerereconstructedin598nominalgroups,meaning
thatwhilemostnominalgroupswitherrorscontainsingleerrors,instancesofmultiple
errorsalsooccur.Excerpts6–10provideanoverviewoftheerrorsinnominalgroups
andtheirfeatures.
6.
awiderangeofthermoelectricSWNT-basedmaterials[11ECS]
7.
anattractivetargettoforsyntheticchemists[04ECS]
8.
naturalmarinenaturalproducts[05ECS]
9.
thegateoxideformation[13ECS]
10.
athepotentialforitsuseapplicationastheanelementarycarriers[10ECS]
TypicaloferrorsintheNGS,thenominalgroupsinExcerpts6–9containsingle
errors(i.e.,omission,selection,ordering,andadditionerrors).InExcerpt6,the
indefinitearticleaisomitted,andinExcerpt7,theprepositiontoismis-selectedforfor
(i.e.,selectionerror).InExcerpt8,thereisanorderingerroroftheadjectivenatural,and
inExcerpt9thedefinitearticletheismistakenlyadded.Excerpt10,istheonlynominal
groupwithmultipleerrorsinthissample,containingthreeselectionerrors:(1)themisselectionoftheindefinitearticleaforthedefinitearticlethe,(2)themis-selectionofits
useforapplication,and(3)themis-selectionofthedefinitearticletheandplural-sfor
thesingularindefinitearticlean(i.e.,mis-construalofanindefinitesingularreferentasa
definitepluralreferent).
Notably,allofthenominalgroupsinExcerpts6-10arecomplexnominalgroups,
asopposedtosimplenominalgroups,whicharevoidofmodification.Thisfeatureand
theerrorpatternsincomplexandsimplenominalgroupsarediscussedindetailin
followingsection.
56
Chapter4:Findings
4.2.1Errorsincomplexandsimplenominalgroups.
Nominalgroupsareclassifiedascomplexorsimple(seeSection3.1).Inshort,
complexnominalgroupscontrastwithsimplenominalgroupsintermsofmodification.
Insimplenominalgroups,typicallyrealisedbypronounsandpropernouns,theHead
standsalone,whereasincomplexgroups,ittakessomeformofmodification.Excerpts
11and12illustratethisdistinction.
Simplenominalgroup
11.
Phe7[12ECS]
Complexnominalgroup
12.
ThehydrophobicpackingofPhe7,Leu27,Tyr34,Phe44,andIle83[12ECS]
ThesimplenominalgroupinExcerpt11expressesthenameofachemicalcompound
withoutmodification.Incontrast,thecomplexnominalgroupinExcerpt12containsfive
simplenominalgroupsembeddedinasinglerankshiftedprepositionalphrase(i.e.,of
Phe7,Leu27,Tyr34,Phe44,Ile83).Rankshiftedprepositionalphrasesandclausesare
analysedaspartofthecomplexnominalgroupinthisstudy.
Theseembeddedelementscontributetotheelaborationthatisafeatureof
nominalgroupsinacademicandtechnicaldiscourses,andinturnprovidefurther
possibilityforerror.Notably,amongthe598nominalgroupswitherrorsintheNGS,
597werecomplex;inotherwords,onlyoneerrorwasidentifiedwithinasimple
nominalgroup.Thatsimplenominalgroup(i.e.,PAG1a)ispresentedwithitserror(i.e.,
theadditionofthedefinitearticlethe)inExcerpt13.
Simplenominalgrouperror
13.
Themethoxygroupof1a(OMe)wasdeprotectedtoahydroxylgroupby
BBr3andasubsequentesterificationwithmesylchloride(MsCl)gavethe
PAG1a.[06ECS]
57
Chapter4:Findings
PAG1aisthenameofaspecificchemicalinagroupofchemicalsknownasphotoacid
generators(PAG).Thisnamingmakesitinherentlyspecificanddistinguishesitfrom
otherPAG,inthisway,nullifyingthenecessityofanarticle.45
Thisresultthat597ofthe598nominalgroupswitherrorsarecomplexnominal
groups(i.e.,99.83%),suggeststhatpedagogyaimedatimprovingaccuracyinJapanese
scientists'L2Englishshouldfocusontheformsofmodificationthatmakethecomplex
nominalgroupbothacentralfeatureofscientificwritingandaprimesourceoferror.
Theseformsofmodificationarefurtherinvestigatedinthefollowingsection.
4.2.2ErrorsinPre-modifier,Head,andPost-Modifier.
Asmoremeaningispackedintonominalgroupsintechnicalregisters,complex
nominalgroupsareincreasinglymodified.ThismodificationmayprecedetheHead,in
whichcaseitisreferredtoasthePre-modifier,oritmayfollowasthePost-modifier
(Matthiessen,Teruya,&Lam,2010,p.24).46Additionally,theHeaditselfmaybe
modified(e.g.,addingplural-s).Pre-modificationoccursinalargelyfixedsequence,
formalisedinfunctionaltermsasfollows:Deictic,Numerative,Epithet,Classifier,Head
(Halliday&Matthiessen,2014,p.379).47Importantly,eachelementtotheleftofthe
Headmodifiestheonetotheright,asillustratedinExcerpt14.
Pre-modificationofcomplexnominalgroup
14.
adefinitively-controlledabsoluteoptically-activesubstancesynthesis
[02ECS]
InExcerpt14,theHead(i.e.,synthesis)ismodifiedbytheClassifier(i.e.,substance);both
aresubsequentlymodifiedbythreeEpithets(i.e.,definitively-controlled,absolute,
optically-active).ThereisnoNumerativeinExcerpt14;thefinalelementontheleftis
theDeictic(i.e.,an),indicatingthelevelofspecificityoftheHead.Asmoremeaningis
packedintothenominalgroup,Classifiers,Epithets,andDeicticscanbeaddedfollowing
45Similarly,vitaminCisonespecificvitamininthegroupofvitamins,anddoesnotrequirethe
additionalspecificityofthedefinitearticle,unlessitismodified(e.g.,thevitaminContheshelves,
orthecommonlytakenvitaminC).
46ThePost-modifierisalsolabelledfunctionallyasQualifier.
47Deicticsaretypicallyrealisedbydeterminers,Numerativesbynumerals,Epithetsby
adjectives,Classifiersbynounsoradjectives,andHeadsbynouns(Table3.4).
58
Chapter4:Findings
thispatternofleftmodifyingright.Notably,Excerpt14isfreefromerror,butitis
readilyapparenthowthisformofmodificationpresentschallengesforJapanese
scientistsneedingtonominalisescientificmeaninginthisway.
Incontrast,post-modificationofcomplexnominalgroupsisrealisedby
rankshiftedclausesandembeddedprepositionalphrases,asdescribedintheIntegrated
FrameworkofAnalysis(Section3.1),andillustratedinExcerpts15and16.
Post-modificationofcomplexnominalgroups
15.
theE1-mechanismstabilizingthecarbocationintermediate[06ECS]
16.
theirusefromtheviewpointofacidgeneration[06ECS]
Excerpts15and16demonstratepost-modificationthrougharankshiftedclause(i.e.,
stabilizingthecarbocationintermediate)andanembeddedprepositionalphrase(i.e.,
fromtheviewpointofacidgeneration),respectively.
Therefore,errorsincomplexnominalgroupsmayoccurinthreelocations:the
Pre-modifier,Head,andPost-modifier,asshowninExcerpts17–19,orinmorethanone
oftheselocations,asshowninExcerpt20.
Pre-modifiererror
17.
itsaminoacidsequencewiththatofhorsecytc[12ECS]
Headerror
18.
opticallyactivesubstances[02ECS]
Post-modifiererror
19.
Detailedproceduretoforcalculateingtheband-edgeprofile[01ECS]
ErrorsinPre-andPost-modifier
20.
athepotentialforitsuseapplicationastheanelementarycarriers[10ECS]
Table4.11presentstheprevalenceoferrorsinthesethreelocations,expressedasa
percentageofnominalgrouperrorsineachtext,andoverallintheNGS.
59
Chapter4:Findings
Table4.11ErrorsinPre-modifier,Head,andPost-modifieraspercentageofnominal
grouperrors
Text
Errorsin
Errorsin
Errorsin
Pre-modifier
Head
Post-modifier
(%)
(%)
(%)
01ECS
40.28
5.56
54.17
02ECS
53.06
23.47
23.47
03ECS
78.48
13.92
7.59
04ECS
74.70
10.84
14.46
05ECS
75.00
16.67
8.33
06ECS
55.17
3.45
41.38
07ECS
40.91
18.18
40.91
08ECS
48.15
20.37
31.48
09ECS
53.70
20.37
25.93
10ECS
60.00
7.50
32.50
11ECS
64.71
23.53
11.76
12ECS
50.00
12.50
37.50
13ECS
60.87
14.49
24.64
NGS
59.02%
14.98%
25.99%
Notably,errorsinthePre-modifieraretwiceasfrequentaserrorsinthepost-
modifier,andfourtimesasfrequentaserrorsintheHead,highlightingthesignificant
obstaclepre-modificationposesforJapanesescientists.However,furtheranalysisis
requiredtoelucidateindetailthevariouserrorswithinthethreelocationsofthe
60
Chapter4:Findings
nominalgroup.Thefollowingsectionexaminesthetwomostsalienterrorpatterns
occurringinnominalgroups.
61
Chapter4:Findings
4.3ErrorPatternsinNominalGroups
FromtheanalysisoferrorsinthePre-modifier,Head,andPost-modifier,two
majorerrorpatternsemerged:errorsrelatingtoarticlesandplural-s,anderrors
relatingtothepreposition-of.Thefollowingsectionsinvestigatethesetwomajorerror
patterns,beginningwiththemostsalient,errorswitharticlesandplural-s.
62
Chapter4:Findings
4.3.1Errorswitharticlesandplural-s.
ErrorswitharticlestheandaoccurfrequentlyinboththePre-andPost-
modifiersofcomplexnominalgroups,asillustratedinExcerpts21–26.48
ErrorswitharticlesinPre-modifier
21.
athecoordinationgeometryaroundEuIII[10ECS]
22.
agreaterchiropticalamplitudes[02ECS]
23.
amorestableconformer[03ECS]
ErrorswitharticlesinPost-modifier
24.
wettabilityofathegateinsulatorsurface[01ECS]
25.
Theamountofthedimerproducedbythetreatmentwithethanol[05ECS]
26.
Thefurthereliminationofaproton[06ECS]
Furthermore,thispatternoferrorswitharticlescanbeextendedtotheHeadwhen
articlesaremis-selectedforplural-sinnon-specificpluralreferents,asshownin
Excerpts27–29.
ErrorswitharticlesinPre-modifierandplural-sinHead
27.
theclusters[03ECS]
28.
thepreviousstudyies[13ECS]
29.
ahighlevelsofcircularpolarization[10ECS]
Thisapparentintegratednatureoferrorswitharticlesandplural-smotivatesfurther
investigationintothesystemsunderlyingthegrammar.AsdiscussedinChapter3
(Figure3.1),thechoicesintheDETERMINATIONsystemoccursimultaneouslywith
choicesintheCOUNTABILITYsystem,andsubsequentlyintheNUMBERsystem.In
otherwords,thechoiceofdefiniteorindefinitearticleisconcurrentwithandcontingent
onthechoiceinsingular,plural,ormass,andviceversa,indicatingthaterrorswith
articlesareinextricablyboundwithplural-serrors,andthuscanbeanalysedasone
48Theindefinitearticleaanditsphoneticequivalentanaretreatedasthesamearticleinthis
investigation.
63
Chapter4:Findings
pattern.Thefrequencyofthearticlesandplural-serrorpatternispresentedinTable
4.12,illustratingtheprevalenceofthiserrorpatternthroughoutallthirteentexts.
64
Chapter4:Findings
Table4.12Errorswitharticlesandplural-s
Errorswitharticles
Text
Errorswith
andplural-sas
Errorswith
articlesand
percentageof
Errorsin
articlesand
plural-s
errorsinnominal
nominal
plural-s
per1,000
groups
groups
pertext
words
(%)
01ECS
72
44
13
61.11
02ECS
98
67
14
68.37
03ECS
79
55
19
69.62
04ECS
83
63
14
75.90
05ECS
12
6
1
50.00
06ECS
29
19
10
65.52
07ECS
22
16
7
72.73
08ECS
54
31
6
57.41
09ECS
54
34
8
62.96
10ECS
40
28
11
67.50
11ECS
34
26
15
76.47
12ECS
8
4
1
50.00
13ECS
69
49
12
71.01
654
442
131
50
34
10
NGS
Average
67.43%
Note.NGSreferstotheNominalGroupSubcorpus,whichincludesonlythosenominalgroups
witherrors.
65
Chapter4:Findings
Theerrorswitharticlesandplural-spatternaccountsfortwointhreeerrorsin
theNGS(i.e.,67.58%),andoccursonaveragetentimesper1,000words.Intermsof
totalerrors,thisoneerrorpatternconstitutesaround32%ofallerrorsintheCOJSD.
Comparingthisresultwithpriorresearchrequirescautiongiventhatmuchofthat
researchmeasurespercentageaccuracyundercontrolledconditions(e.g.,gapfill
exercises)ratherthanfrequencyinfreeproduction,49howeverwithintheELTanderror
analysisliteratureseveralstudiesarecomparable.Forexample,withL1Spanish
university-levellearners,DottiandO'Donnell(2014)foundthatarticleerrorsaccount
foraround10%ofallerrors.ForL1Frenchuniversitystudents,Hamilton(2015)
reportsthatfromatotalof906errors,articlesanddeterminererrorsaccountedfor
11.04%oftotalerrors.AlsoinEurope,Klages-Kubitzki(1995)reportsthat49.39%ofall
errorsproducedbyL1GermanuniversitylecturersinEnglishdepartmentsinGermany
werewitharticles,particularlythedefinitearticle.Schleppegrell(2002)reports
"count/massnounandarticle"errorsaccountforupto37.87%ofthetotalsurface
grammarerrorsproducedbythreeESLstudents(p.137).50InAsia,Chen(2006)reports
asurprisinglylowerrateofarticleerrorsof4.23%forTaiwaneselearners(i.e.,141in
3,332;p.89),whileHongetal.(2011)reportacombinedtotalof19.20%forarticlesand
plural-serrorsinthewrittenessaysofL1Malaysianstudents(i.e.,12.91%and6.29%,
respectively;p.39).
Theprevalenceofthismajorerrorpatterndemandscloserinvestigationto
elucidatethewaysinwhicherrorswitharticlesandplural-sunfold.Aspreviously
discussed(Sections2.2.2,3.1,and3.3.2),thisEAclassifieserrorsintofourerrortypes:
omission,addition,selection,andordering.Excerpts30–40demonstratethefourerror
typeswithintheerrorswitharticlesandplural-spattern.
Omissionerrors
30.
AphotographoftheMChmeta-molecule[07ECS]
31.
newfunctionaldevicessuchasa“one-waymirror”[07ECS]
49Forexample,inastudyof60universitystudentsinJapan,Goto-Butler(2002)reportsan
overalllevelofaccuracyfor"nontargetlikearticleuse"of34.16%.
50DetailsoftheESLstudents'languagebackgroundsarenotgiven.
66
Chapter4:Findings
32.
thefunctionofproteinscomposedofseveralprotomerunitssuchas
caspase-3[08ECS]
Additionerrors
33.
thedirectionalanisotropy[07ECS]
34.
adispersion-typefeaturesatthesefrequencies[07ECS]
35.
thetotallengthofthemeta-moleculeinmillimeters.[07ECS]
Selectionerrors
36.
athewaytowardtherealizationofanartificialgaugefield[07ECS]
37.
thearesultofbindingtotheprotomerinterfacecavity[08ECS]
38.
thefunctionofaproteinscomposedofseveralprotomerunitssuchas
caspase-3[08ECS]
39.
atypicalreactionsforproducingamatureenzyme[08ECS]
Orderingerror
40.
theveryweakthecatalyticactivity[i]
NotethatExcerpt40isaninventedexample(i.e.,[i]),createdtodemonstratean
orderingerror.IntheNGS,noorderingerrorsarefoundwithinthiserrorpattern,
indicatingthattheorderingofarticlesandplural-spresentsnodifficultyforJapanese
scientists.Thissuggeststhatpedagogyaimedataddressingerrorswitharticlesand
plural-scanbefocusedonthethreeothererrortypes:omission,addition,andselection.
Thepercentagedistributionofomission,addition,andselectionerrorswitharticlesand
plural-sisgiveninTable4.13,highlightingthedominanceofomissionerrors.
67
Chapter4:Findings
Table4.13Percentagedistributionoferrortypesinerrorswitharticlesandplural-s
Text
Omissionerrors
Additionerrors
Selectionerrors
(%)
(%)
(%)
01ECS
63.64
18.18
18.18
02ECS
83.58
11.94
4.48
03ECS
74.55
9.09
16.36
04ECS
92.06
6.35
1.59
05ECS
83.33
16.67
0.00
06ECS
84.21
10.53
5.26
07ECS
56.25
37.50
6.25
08ECS
51.61
25.81
22.58
09ECS
70.59
23.53
5.88
10ECS
50.00
28.57
21.43
11ECS
73.08
23.08
3.85
12ECS
50.00
50.00
0.00
13ECS
81.63
12.24
6.12
Total
74.21%
16.29%
9.50%
Notably,omissionerrorsaremorethanfourtimesasfrequentasadditionerrors,
andeighttimesmorefrequentthanselectionerrors,makingitthedominanterrortype
intheerrorswitharticlesandplural-spattern.Thisresultcorrelateswithprevious
studiesshowingthatL2learnerswitharticle-lessL1s(e.g.,Chinese,Taiwanese,Korean)
frequentlyomitarticlesintheirwriting(Gressang,2010;Robertson,2000;White,2003).
Similarly,RozyckiandJohnson(2013)foundthedroppingofarticlestobeafeatureof
"non-canonical"grammarinresearcharticleswrittenbynon-nativeengineers,and
Master(1997)hasshownthatevenadvancedlearnershavedifficultieswitharticles.
68
Chapter4:Findings
Thisfindingpromptsacloserinvestigationintothevariousmanifestationsof
errorswithinthesethreeerrortypestoidentifywhatisbeingomitted,addedandmisselected.Inthisinvestigation,the442errorswitharticlesandplural-swereanalysed
accordingtothesystemnetworkinFigure3.5.Inshort,eacherrorisanalysedaccording
tothefivefeaturesofthesystemnetwork(i.e.,singular,plural,mass,specific,andnonspecific).Theresultsofthisanalysis,presentedinTable4.14,illustratethedominanceof
errorswithsingularreferents,andatrendtowardserrorswithnon-specificreferents.
69
Chapter4:Findings
Table4.14Relativefrequenciesofarticleandplural-serrors
Singular
Specific
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
Non-specific
theoptimizationofthewhole
proteinstructure[08ECS]
thePauliexclusionprinciple[02ECS]
theYASARAstructuremolecular
modelingsoftwarepackage[08ECS]
theauto-oxidationofthecysteine
thiolgroupontheenzymes[08ECS]
thegas-liquidinterface[09ECS]
thecriticalpointinparticular
[03ECS]
theexcitedstate[03ECS]
thequenchingreaction[04ECS]
thekealiininefamily[04ECS]
theoxidizedmonomericM61APA
cytc551[05ECS]
theE1-mechanism[06ECS]
thelowerhalfofthebandgap
[13ECS]
athesubstratesurface[01ECS]
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
27.66%
Plural
2.
3.
4.
5.
thelargestbindingenergies[08ECS]
thesynthesesofbioactivemolecules
[04ECS]
thedesiredphotoproducts[03ECS]
theanti-productsdecreased[03ECS]
9.07%
4.
5.
6.
6.
7.
8.
9.
Mass
3.
4.
5.
themodulationofproteinfunctions
[08ECS]
thechlorophyllinChlorellacells
[09ECS]
theflexibility[11ECS]
1.
2.
3.
7.26%
4.
averysmallamount[08ECS]
acombinationofPAC-1andthe
maturecaspase-3[08ECS]
acobalt-chelatingaffinitycolumn
[08ECS]
adecreaseintheconcentrationof2ME[08ECS]
abacksidealignmenttechnique
[09ECS]
acouplingpartner[03ECS]
amoderateconversion[03ECS]
an11-stepconversion[04ECS]
ahighlydense-functionalized
cyclopentanecore[04ECS]
AnAuelectrode[05ECS]
anincreaseinhighorderoligomers
[05ECS]
Thefurthereliminationofaproton
[06ECS]
thealargedeviation[03ECS]
30.16%
higherconcentrations[08ECS]
freeze-and-thawcycles[08ECS]
lowspatialfrequencyregions
[01ECS]
complicatedstructures[03ECS]
Highlystrainedcyclobutane
skeletons[03ECS]
thecorrespondingacylazides
[04ECS]
thechemists[04ECS]
16.55%
Theamountofthedimerproduced
bythetreatmentwithethanol
[05ECS]
anultimatesensitivitytoUVlight
[06ECS]
Moststudiesonthedirectional
anisotropy[07ECS]
evidencesofthedirectobservation
oftheMCheffectsinthemetamolecule[07ECS]
9.30%
Note.Columnandrowheadingsrefertothereferentaftererrorreconstruction.
ThesamplesofexcerptspresentedinTable4.14areweightedtoillustratetheir
relativefrequencies.Thisillustrationdemonstratesthatarticleandplural-serrorsoccur
mostfrequentlywithsingularreferents,followedbypluralandmassreferents.
Furthermore,atrendtowardserrorsinnon-specificreferentsappears.Thefollowing
70
Chapter4:Findings
threesectionselaboratetheerrorswithsingular,plural,andmassreferents,
commencingwiththemostprominent,singularreferents.
4.3.1.1Singularreferenterrorswitharticlesandplural-s.
Errorswithsingularreferentsaccountforacombined57.82%oftheerrorswith
articlesandplural-serrorpattern(Table4.14).GenerallyinEnglish,singularnouns
occurthreetimesmorefrequentlythanpluralnouns(Taylor,2012,pp.154-155).
Moreover,Englishhasmanymorecountnounsthanmassnouns(Fieder,Nickels,&
Biedermann,2014,p.14).Therefore,theerrorfrequenciespresentedinTable4.14may
simplybeareflectionofthenaturallyoccurringfrequenciesofnounsinEnglish,with
singularnounsthemajority.
WithintheDETERMINATIONsystem(i.e.,specificandnon-specificreferents),
errorswithsingularreferentsaredistributedevenlybetweenspecificandnon-specific
referents(i.e.,27.66%and30.16%,respectively),withjustaslighttrendtowardserrors
withnon-specificreferents.However,whenconsideringthenaturallyoccurring
frequenciesofspecificandnon-specificreferentsinEnglish,thetrendtowardserrors
withnon-specificreferentsismorepronounced.Inlargecorporarepresentinggeneral
English(e.g.,theBrownCorpusandtheBritishNationalCorpus),thedefinitearticle
occursaround2.6timesmorefrequentlythantheindefinitearticle.51Incontrasttothis
trendinlargercorporaandthegreaterpotentialforerrorinthemorefrequently
occurringdefinitearticle,errorsoccurmorefrequentlywiththeindefinitearticleinthe
NGS.ThisfindingsuggeststheindefinitearticlepresentsgreaterdifficultiesforJapanese
scientiststhanthedefinitearticleandshouldbeconsideredcarefullyinpedagogyaimed
ataddressingthiserrorpattern.
Theerrorswithsingularreferents,bothspecificandnon-specific,occurinthe
followingthreeways:thearticleis(1)omitted,(2)mis-selectedor(3)inappropriately
added(i.e.,omission,selection,andadditionerrors,respectively).Withintheerrorswith
singularreferents,omissionerrorsarethemajority(85.16%),followedbyselection,
thenadditionerrors,asoutlinedinTable4.15.
51Intheone-millionwordBrownCorpusofwrittenAmericanEnglish,thedefinitearticleoccurs
69,971times,whiletheindefinitearticleoccurs23,225times(Francis&Kucera,1961).
Similarly,intheBritishNationalCorpusthedefinitearticleoccurs5,973,437times,compared
withtheindefinitearticleoccurring2,136,923times(BNC,2007).Inbothcorpora,thisisa2.6
timesincreaseinthedefinitearticle.
71
Chapter4:Findings
Table4.15Errortypeswithinarticleandplural-serrorswithsingularreferents
Errortype
Total
Proportion
Average
Average
frequency
(%)
frequency
frequency
pertext
per1,000
inNGS
words
Omission
218
85.16
16.77
4.71
Selection
35
13.67
2.69
0.76
Addition
3
1.17
0.23
0.06
19.69
5.53
Total
256
100.00%
Theseresultsindicatethattheomissionofarticleswithsingularreferentsisapersistent
errorintheL2writingofJapanesescientists,occurringaround4–5timesper1,000
words.Toillustratetheeffectsoftheseerrorsonreadingflowandcomprehension,
Excerpts41–44demonstrateomissionerrorswithspecificandnon-specificsingular
referents,beforeandaftererrorreconstruction.
Omissionofarticleerrorswithsingularreferentsbeforeerrorreconstruction
41.
AsshowninFig.1,thegapelectrodeswerefabricatedonthermally
oxidizedSichip.[ECS09]
42.
Toelucidatethismechanism,weinvestigatedrelationshipbetween
conductivityandthermoelectricpropertiesoftpp-dopedSWNTfilmsat
variousdopantconcentrations.[11ECS]
Omissionofarticleerrorswithsingularreferentsaftererrorreconstruction
43.
AsshowninFig.1,thegapelectrodeswerefabricatedonathermally
oxidizedSichip.[ECS09]
44.
Toelucidatethismechanism,weinvestigatedtherelationshipbetween
conductivityandthermoelectricpropertiesoftpp-dopedSWNTfilmsat
variousdopantconcentrations.[11ECS]
Whiletheimpactofeacharticleomissionerrormaynotbecriticaltointelligibility,the
prevalenceandsubsequentaccumulationoftheseerrorsoverawholetextdistracts
fromthecontentandreflectsnegativelyontheaccuracyofthewriting(and
72
Chapter4:Findings
subsequently,theresearch).Giventhattheseomissionerrorswithsingularreferents
occuronaverage16-17timespertext(Table4.15),thefrequentdisruptionstothe
readingflowandcomprehensionthattheseerrorscausemaybeconsequential.
Thenegativeimpactoftheomissionerrorsisfurthercompoundedbythe
selectionerrors.Theselectionerrorsinsingularreferentsaccountfor13.67%ofthe
errorswithsingularreferents(Table4.15)andcombinewiththeomissionerrorsto
interruptthereadingflowandcomprehension,asdemonstratedinExcerpts45-48,
beforeandaftererrorreconstruction.
Selectionofarticleerrorswithsingularreferentsbeforeerrorreconstruction
45.
Powerfactorsaretheusefulmeasureofthermoelectricproperties.[11ECS]
46.
Sincethesplittingofnarrowf-femissionlinesishighlysensitivetoa
coordinationgeometryaroundEuIII,theStarksplittingpatternoftheEuIII
complexesisagoodopticalfingerprintforidentificationoftheir
coordinationstructuresinsolution.[10ECS]
Selectionofarticleerrorswithsingularreferentsaftererrorreconstruction
47.
Powerfactorsaretheausefulmeasureofthermoelectricproperties.
[11ECS]
48.
Sincethesplittingofnarrowf-femissionlinesishighlysensitivetotheathe
coordinationgeometryaroundEuIII,theStarksplittingpatternoftheEuIII
complexesisagoodopticalfingerprintforidentificationoftheir
coordinationstructuresinsolution.[10ECS]
Thethirderrortypewithinsingularreferents,additionerrors,occursonlythree
timesthroughouttheNGS.ThosethreeoccurrencesarepresentedinExcerpts49-51.
Additionofarticleerrorswithsingularreferents
49.
Forthepurposeofclarity,Figure1c,1g,1k,and1pshowtheeitheroneof
thecrystalstructures.[10ECS]
50.
Theweakcorrelationbetweenthestructuraldissymmetryandthe
luminescencedissymmetryfactorimpliesaweakcontributionofthestatic-
73
Chapter4:Findings
couplingmechanismtothecircularlypolarizeddissymmetryofthe
lanthanide(III)f-ftransition.[10ECS]
51.
Themicroplasmasources,whichsupplyreactivespecies,arelocatedonthe
backoftheeachmicrowell.[09ECS]
Interestingly,allthreeoccurrencesinvolvetheadditionofthedefinitearticle,rather
thantheindefinitearticle;however,theverylowfrequencyandirregularityofthese
additionerrorswithsingularreferentssuggeststhattheyaremorelikelytoberandom
mistakesthansystematicerrors.
4.3.1.2Pluralreferenterrorswitharticlesandplural-s.
Afterthesingularreferenterrors,articleandplural-serrorswithpluralreferents
constitutethesecondmostfrequentlyoccurringerrorsinthiserrorpattern,accounting
for25.62%ofthearticlesandplural-serrors(Table4.14).Inpluralreferenterrors,
non-specificreferenterrorsarealmosttwiceasfrequentasspecificreferenterrorsin
thedata(i.e.,16.55%and9.07%,respectively;Table4.14).Aspreviouslydiscussed
(Section4.3.1.1),thistrendismorepronouncedwhenconsideringthatnon-specific
referentsactuallyoccurlessfrequentlythanspecificreferentsinEnglish.
Similartosingularreferenterrors,pluralreferenterrorsaredominatedby
omissionerrors,asillustratedinTable4.16.
74
Chapter4:Findings
Table4.16Errortypeswithinarticleandplural-serrorswithpluralreferents
Errortype
Frequency
Proportion
Average
Average
(%)
frequency
frequency
pertext
per1,000
words
Omission
87
76.99
6.69
1.88
Selection
3
2.65
0.23
0.06
Addition
23
20.35
1.77
0.50
100.00%
8.69
2.44
Total
113
Sinceerrorsoccuralmosttwiceasfrequentlywithnon-specificreferentscomparedwith
specificreferents(i.e.,16.55%and9.07%,respectively;Table4.14),thetendencyofthe
Japanesescientistsinthisstudyistoomitplural-s,asdemonstratedinExcerpt52.
Omissionofplural-serrorinnon-specificpluralreferent
52.
Differenteffectsfromsurfacetreatment[01ECS]
Incontrast,omissionerrorsinspecificpluralreferentsmayoccurwitheitherthe
definitearticleortheplural-sasdemonstratedinExcerpt53and54,respectively.
Omissionofdefinitearticleandplural-serrorsinspecificpluralreferents
53.
anapparentweakeningofthebrightspotsintheFOMimageofh-PF8T2
afterther-CPirradiation[02ECS]
54.
themechanismsunderlyingplasmamedicine[09ECS]
Combinedtogether,theseomissionerrorswithspecificandnon-specificreferentsoccur
onaveragetwotimesper1,000wordsintheCOJSD.
Aftertheomissionerrors,additionerrorsaccountfor20.35%oftheerrorsin
pluralreferents(Table4.16).Inall23oftheseadditionerrorsinthedata,theerrors
occurwithnon-specificpluralreferents,asillustratedinExcerpts55and56.
Additionofarticleerrorsinpluralreferents
55.
allkindsofthesubstratesusedinthiswork[01ECS]
75
Chapter4:Findings
56.
agreaterchiropticalamplitudes[02ECS]
Notably,thedefinitearticleiserroneouslyadded19times,whiletheindefinitearticleis
addedonly4times,furtherhighlightingtheJapanesescientists'infrequentuseofthe
indefinitearticle.
Thethirdtypeofpluralreferentserror,selection,israre,onlyoccurringthree
timesintheNGS(Table4.16).ThesethreeselectionerrorsarepresentedinExcerpts57–
59.
Selectionofarticleandplural-serrorsinpluralreferents
57.
areducingreagentsand/orastabilizer[08ECS]
58.
thefunctionofaproteinscomposedofseveralprotomerunitssuchas
caspase-3[08ECS]
59.
thepreviousstudyiesstudies[13ECS]
Inallthreeoccurrences,theerroristhemis-selectionofasingularreferentforaplural
one.Giventhatthiserroroccursonlythreetimesinthetotal442errorswitharticles
andplural-s,andgiventwoofthoseoccurrencesarewithintheonetext(i.e.,[08ECS]),
thiserrorisnotconsideredanendemicerrorpatternforJapanesescientists.
Comparedwiththesingularreferenterrors,errorswithpluralreferentstendto
producedifferenteffectsonreadingflowandcomprehension.Toillustratethispoint,
Excerpt60demonstratesthemostfrequent,omissionerrorwithanon-specificreferent,
beforeandaftererrorreconstruction.
Errorswitharticlesandplural-sinpluralreferentsbeforeerrorreconstruction
60.
Highlystrainedcyclobutaneskeletonshowsimportantpotentialasthe
precursorofnotonlynaturalproductsbutalsovariousuniquecompounds,
andmuchefforthasbeendedicatedtothisreactioninvariousfields.
[03ECS]
Errorswitharticlesandplural-sinpluralreferentsaftererrorreconstruction
61.
Highlystrainedcyclobutaneskeletonsshowsimportantpotentialasthe
precursorofnotonlynaturalproductsbutalsovariousuniquecompounds,
76
Chapter4:Findings
andmuchefforthasbeendedicatedtothisreactioninvariousfields.
[03ECS]
Theimpactoftheseerrorsonreadingcomprehensionandtheaccuracyofthecontent
cannotbeconsideredminimal.Withanaveragefrequencyof8–9timespertext,these
errorsdisruptthereadingflowandpotentiallydetractfromthequalityoftheJapanese
scientists'work,especiallywhencombinedwiththesingularreferenterrorsdiscussed
above,andthemassreferenterrorsexaminedinthefollowingsection.
4.3.1.3Massreferenterrorswitharticlesandplural-s.
Thefinaldescriptorintheinvestigationoferrorswitharticlesandplural-serror
pattern,errorswithmassreferents,accountsfor16.56%oftheerrorswitharticlesand
plural-s(Table4.14).Aspreviouslydiscussed(Section4.3.1.1),thisrelativelylow
frequencyoferrorswithmassreferentsreflectstherelativelylowfrequencyofmass
nounsinEnglishcomparedwithsingularandpluralnouns.Furthermore,thisresultmay
beanartefactofthelackofindefinitearticleandplural-swithmassnouns.Inother
words,giventhereisnoindefinitearticleorplural-swithmassreferents,thereisless
potentialforthemostcommonerrortype,omission,tooccur.Indeed,Table4.17shows
that,comparedwithomissionerrors,additionerrorsaccountfortwicethenumberof
errorsinmassreferents.
77
Chapter4:Findings
Table4.17Errortypeswithinarticleandplural-serrorswithmassreferents
Errortype
Frequency
Proportion
Average
Average
(%)
frequency
frequency
pertext
per1,000
words
Omission
23
31.51
1.77
0.50
Selection
4
5.48
0.31
0.09
Addition
46
63.01
3.54
0.99
Total
73
100.00%
5.62
1.58
Errorswithmassreferentsoccuronaverage1.58timesforevery1,000wordsin
theCOJSD.Excerpts62and63demonstratethemostprominenterrortypewithmass
referents,addition,highlightingthenuancedchoiceintheCOUNTABILITYsystem(i.e.,
thechoicetoconstrueanounasacountormassnoun).
Additionerrorswitharticlesandplural-sinmassreferents
62.
chemicalbiologyies,pharmaceuticals,medicinalandagriculturalareas
[04ECS]
63.
anevidenceofanoptofluidiceffect[02ECS]
WhileitmaybepossibletoconstruethemassnounsinExcerpts62and63ascount
nouns(e.g.,biologies,anevidence),thatchoiceisatypical52andratherthanenrichingthe
meaningincontext,functionsonlyasadisruptiontoreadingflowandcomprehension.
Notably,nounsthatcanbeconstruedascountormass(e.g.,difficulty,influence,
temperature,condition,etc.)areprevalentintheCOJSD,butinthesecasesthetendency
oftheJapanesescientistsinthisstudyistoomitthedefinitearticle,asillustratedin
Excerpts64and65,ratherthanmis-construethemascountnouns.
Omissionofdefinitearticleerrorsinmassreferents
64.
thedifficultyofreactioncontrol[04EST_43]
52IntheBritishNationalCorpus,biologyoccurs1,074times,whilebiologiesoccursonlyonce.
Similarly,evidenceoccurs20,995timeswhileanevidenceoccursonly11times(BNC,2007).
78
Chapter4:Findings
65.
theinfluenceofthesurfacetreatmentscorrectly[01EST_22]
Thesecondmostcommonmassreferenterror,omissionerrors,accountfor
31.51%ofthemassreferenterrors(Table4.17).Notably,theomissionofthedefinite
article,showninExcerpts64and65,istheonlyformofomissionthatmayoccurwith
massreferents,asthereisnoarticleinnon-specificmassreferents.Theomissionofthe
definitearticlewithspecificmassreferentsoccursin9ofthe13textsintheCOJSDfora
totalof23times(Table4.17).Excerpts66and67illustratetheeffectsofbothan
omissionerrorandthemoreprevalentplural-sadditionerrorformassreferents,before
andaftererrorreconstruction.
Additionofplural-serrorandomissionofdefinitearticleerrorinmassreferents
beforeerrorreconstruction
66.
Thisfindingsuggeststhatweshouldpayattentionstotheeffectof
concomitantreagentsinreactionmediaevaluatingtheabilitiesofsmall
organicmoleculesinmodulationofproteinfunctions.[08ECS]
Additionofplural-serrorandomissionofdefinitearticleerrorinmassreferents
aftererrorreconstruction
67.
Thisfindingsuggeststhatweshouldpayattentionstotheeffectof
concomitantreagentsinreactionmediaevaluatingtheabilitiesofsmall
organicmoleculesinthemodulationofproteinfunctions.[08ECS]
Theoveralleffectoftheseerrorsisonceagainnotcriticalforintelligibility,but
interruptsreadingflowandcomprehension,negativelyimpactingthelevelof
sophisticationandaccuracycentraltoacademicandscientificregisters.
4.3.1.4Summaryoferrorswitharticlesandplural-s.
Thisinvestigationoferrorswitharticlesandplural-sindicatestheseerrors
compriseamajorerrorpatternaccountingfortwointhreeoftheerrorsintheNGS,and
occurringonaverage34timespertext,or10timesper1,000words(Table4.12).The
prevalenceofthiserrorpatterncanbeexpectedgiventheabsenceofanarticlesystem
intheparticipants'L1(Butler,2002,p.453),theabstractnatureofthe
DETERMINATIONandCOUNTABILITYsystems(Barner,Inagaki,&Li,2009),andthe
79
Chapter4:Findings
relativelyhighoccurrenceofarticlesinEnglish(Berry,1991).Theseerrorswere
classifiedaccordingtothefourerrortypeswithintheEAanalyticalframework(i.e.,
omission,addition,selection,ordering),withfindingsrevealingomissionerrorsasthe
predominanterrortype,followedbyadditionandselectionerrors(i.e.,74.21%,16.29%,
and9.50%,respectively;Table4.13).Theseerrortypeswerefurtheranalysedaccording
totheDETERMINATION,COUNTABILITYandNUMBERsystems.Inotherwords,errors
werefurtherclassifiedsimultaneouslyaccordingtowhethertheyoccurinspecificor
non-specificreferents,andsingular,pluralormassreferents.Findingsindicatethatthe
distributionoferrorsintheDETERMINATIONsystemtendstowardserrorswithnonspecificreferents.Furthermore,intheCOUNTABILITYandNUMBERsystems,errors
occurtwiceasfrequentlyinsingularreferentsthanplural(i.e.,57.82%,25.62%,
respectively),andmorethanthreetimesmorefrequentlythaninmassreferents(i.e.,
16.56%;Table4.14).Therefore,theJapanesescientistsinthisstudyprimarilyomit
articlesinsingularnon-specificreferents,asillustratedinExcerpt68.
Predominantarticleerrorwithindefinitesingularreferent
68.
AsshowninFig.1,thegapelectrodeswerefabricatedonathermally
oxidizedSichip.[09ECS]
ThetypeoferrordemonstratedinExcerpt68occursonaverage16–17timespertextin
theCOJSD,andwhileitmaynotbecriticaltointelligibility,theregulardisruptiontothe
readingflowandcomprehensionmaynegativelyimpacttheaccuracyofthewritingand
subsequentaccuracyofthecontent.Theintegrationofthethreesystems(i.e.,
DETERMINATION,COUNTABILITY,andNUMBER)andtheerrorsoccurringwithinthose
systemssuggeststheneedforJapanesescientiststounderstandthewholegrammatical
system.Inotherwords,giventhatthechoicesineachofthethreesystemsare
concurrentwithandcontingentoneachother,anunderstandingofthewhole
grammaticalsystemnetworkisnecessarytoreduceanyoftheerrorsinthismajorerror
pattern.Needs-specificpedagogyaimedatimprovingaccuracyintheL2Englishof
Japanesescientistsmusttakeaccountofthisintegration.
80
Chapter4:Findings
4.3.2Errorswithpreposition-of
Thissectionexaminesthesecondmajorerrorpatterntoemergefromthe
analysisoferrorsintheNGS,namelyerrorsinvolvingthepost-modifying-of.
ThroughouttheNGS,errorswith-ofappearpredominantlyinthePost-modifiersof
complexnominalgroups,butalsomayinvolvepre-modification.Inotherwords,
althoughtheerrororiginatesinthepost-modifier,thereconstructionoftheerrormay
relocateittothepre-modifier.Table4.18outlinesthefrequencyoftheerrorswith
preposition-ofinrelationtothetotalnumberoferrorsintheNGS,highlightingthe
prevalenceoftheseerrorsthroughoutall13textsintheCOJSD.
81
Chapter4:Findings
Table4.18Errorswithpreposition-of
Errorsin
nominalgroups
72
Text
01ECS
Errorswith-of
pertext
18
Errorswith-of
per1,000words
5
Errorswith-ofas
percentageof
errorsinnominal
groups
(%)
25.00
02ECS
98
11
2
11.22
03ECS
79
3
1
3.80
04ECS
83
5
1
6.02
05ECS
12
4
1
33.33
06ECS
29
4
2
13.79
07ECS
22
3
1
13.64
08ECS
54
8
2
14.81
09ECS
54
9
4
16.67
10ECS
40
4
1
10.00
11ECS
34
3
1
8.82
12ECS
8
1
0
12.50
13ECS
69
5
3
7.25
654
78
24
50
6
2
NGS
Average
11.93%
Note.NGSreferstotheNominalGroupSubcorpus.
AsillustratedinTable4.18,theerrorswith-ofpatterncomprisesaround12%of
alltheerrorsintheNGS,andoccursonaveragetwiceper1,000wordsinCOJSD.While
theseerrorsarelessfrequentthanthoseinvolvingarticlesandplural-s,they
82
Chapter4:Findings
nonethelessconstitutethesecondmostfrequentlyoccurringerrorpatternintheNGS,
requiringacloserexaminationtoidentifythespecificobstaclestheypresenttoJapanese
scientistsstrivingtoproducesophisticatedandaccurategrammaticalEnglishfor
researchandpublicationpurposes.
AsoutlinedintheMethodology(Section3.3.2.3),theerrorswith-ofpatterncan
beanalysedaccordingtothefourerrortypesintheEAanalyticalframework(i.e.
omission,addition,selection,andordering).Excerpts69-72illustrateeachofthefour
errortypesinvolvingthepost-modifying-of.
Omissionerrorwithpreposition-of
69.
knowledgeandunderstandingofanefficientgenerationandswitchingof
opticallyactivesubstances[02ECS]
Additionerrorwithpreposition-of
70.
theincidentX-rayangleofx-ray[01ECS]
Selectionerrorwithpreposition-of
71.
Onepossiblereasonofforthedecreaseindomainsize[01ECS]
Orderingerrorwithpreposition-of
72.
therefractiveindexofthecosolventsat589nm(nD)ofthecosolvents
[02ECS]
Excerpt69demonstratesanomissionerrorwith-ofinthePost-modifier,wherethe
omissionmis-construesthenoununderstandingastheverb-form,causinginterruption
toreadingflowandcomprehension.Inotherwords,inthecaseofnominalisationssuch
asunderstanding,thenominalformismorphologicallyidenticaltoitsprocessform(e.g.,
understandingsomethingisimportantvs.theunderstandingofsomethingisimportant)
andisonlydistinguishablethroughpre-andpost-modifyingelements(i.e.,apremodifyingthe-andpost-modifying-of).53Therefore,inthesecircumstances,the
inclusionofthepost-modifying-ofisparticularlyimportanttodesignatethenominal
53Thisparticulartypeofnominalisationisknownasgerundivenominalisation(Heyvaert,2008;
Lees,1960).
83
Chapter4:Findings
grammaticalfunction.Excerpt70containsanadditionerrorinwhich-ofis
inappropriatelyusedtoembedtheclassifyingmeaningofx-rayinthePost-modifier(i.e.,
angleofx-ray)ratherthaninpre-modifyingposition(i.e.,x-rayangle;detailedinSection
4.3.2.3).IntheselectionerrorinExcerpt71,-ofismis-selectedforfor(i.e.,reasonofthe
decreaseratherthanreasonforthedecrease),andinExcerpt72,the-ofconstructionis
mis-ordered(i.e.,initsoriginalerrorform,itismodifyingthecircumstance,at589,
ratherthantheHeadofthenominalgroup,index),againinterruptingthereadingflow
andcomprehension.
Acloserlookatthefrequenciesoftheseerrorsinvolving-ofrevealsthataddition
andselectionerrorsarethemostcommonerrortypes,asillustratedinTable4.19(i.e.,a
combined82.05%ofall-oferrors).
84
Chapter4:Findings
Table4.19Proportionoferrortypesinerrorswith-ofconstructions
Text
Omissionerrors
Additionerrors
Selectionerrors
Orderingerrors
(%)
(%)
(%)
(%)
01ECS
5.56
22.22
72.22
0.00
02ECS
54.55
0.00
36.36
9.09
03ECS
33.33
33.33
33.33
0.00
04ECS
20.00
80.00
0.00
0.00
05ECS
0.00
100.00
0.00
0.00
06ECS
0.00
100.00
0.00
0.00
07ECS
33.33
0.00
66.67
0.00
08ECS
12.50
50.00
37.50
0.00
09ECS
11.11
66.67
22.22
0.00
10ECS
0.00
100.00
0.00
0.00
11ECS
0.00
0.00
66.67
33.33
12ECS
0.00
100.00
0.00
0.00
13ECS
0.00
60.00
40.00
0.00
15.38
44.87
37.18
2.56
Total
Tofurtherunderstandthevariouserrorsinvolving-of,thefollowingsectionsexamine
eachofthefourerrortypesindetail(i.e.,omission,addition,selection,andordering),
beginningwiththemostfrequentlyoccurringadditionerrors.
85
Chapter4:Findings
4.3.2.1Additionerrorswithpreposition-of
Additionerrorsaccountforalmosthalfoftheerrorswith-ofpattern(i.e.,
44.87%;Table4.19).Withintheseadditionerrors,therearetwowaysinwhich-ofis
inappropriatelyadded,asillustratedinExcerpts73and74.
Additionerrorswith-of
73.
affectionofthewholeofproteinstructure[08ECS]
74.
One2x2cm2plasma-on-chipof2x2cm2[09ECS]
Excerpt73demonstratesastraightforwardadditionerrorsimplyinvolvingthe
unnecessaryadditionof-of.ThiserroroccursatotalofsixtimesinfiveoftheCOJSD
texts.Incontrast,theadditionerrorillustratedinExcerpt74ismorecommon
throughoutthecorpus,occurring27timesin10ofthe13texts,andinvolvingthemore
complexadditionofanembeddedof-prepositionalphraseinthePost-modifier.While
theadditionisinthePost-modifier,thiserrorisclassifiedasaPre-modifiererror
becausethereconstructionrequiresthere-configurationofthemeaningintothePremodifierasanEpithet.54Theprevalenceandmorecomplexnatureofthiserrorwarrant
furtherexplanation.
InExcerpt74,whilethestructuringofthemeaningasaQualifierinthePost-
modifierisgrammaticallypossibleandacceptable(i.e.,oneplasma-on-chipof2x2cm2),
thereisanimportantnuancedchoiceinthetextualmeaningtoconsider.Notably,the
unmarkedfocusofanominalgroupfallsonthelastelementofthegroup,whichisoften
theHeadbutmovestotheQualifierwhenpost-modificationispresent(Halliday&
Matthiessen,2014,p.387;Halliday&Webster,2009,p.231).Therefore,comparedwith
packagingmeaninginthePre-modifierasaClassifierorEpithet,locatingthemeaningin
thepost-modifyingQualifierhasthe"greaterpotentialfornews"(Halliday&
Matthiessen,2014,p.388).Thisfocusonthelastelementofthegrouphasimportant
implicationsinthegiven-newtextualflowofmeaninginEnglish.55Ingiven-new
prosodies,thePost-modifierconveysinformationasnew,whereasthesameinformation
54Theconventionthroughoutthethesisistoclassifyerrorsbasedontheirlocationinthe
nominalgroupaftererrorreconstruction.
55SeeHallidayandWebster(2009)foranoverviewofthegiven-newprosodiesofEnglish.
86
Chapter4:Findings
locatedinthePre-modifierconstruesitasgiven.AsHallidayandMatthiessen(2014)
summarise:
Ifapropertyorclassrelatingtoathinghasbeenintroducedinthediscourse
precedinganominalgroup,itislikelytobepresentedasanEpithetorClassifier
oftheThingratherthanas(partof)aQualifier.(p.388)
Excerpt75furtherillustratesthisprosody.
Additionerrorwith-ofreconfiguredasClassifier
75.
Conversely,theglumvaluesof1-5iPrarerathersensitivetosolventsince
their1-5iPrcoordinationgeometriesof1-5iPrrearrangeupondissolutionin
solvent.[10ECS]
InExcerpt75,theembedded-ofprepositionalphraseinthePost-modifier(i.e.,of1-5iPr)
ismoreappropriatelyreconstructedasapre-modifyingClassifier(i.e.,1-5iPr
coordinationgeometries)tobecomethe'given'informationthesecondtimeitisused.
Theprevalenceofthisadditionerrorwith-ofprepositionalphrasesinthePost-
modifieronceagainhighlightsthedifficultiesJapanesescientistsfacewithcomplex
nominalgroupmodification,specificallywithpre-modification.Itfurtherrevealsthe
Japanesescientists'tendencytooveruse-ofconstructionsinthePost-modifier.This
over-relianceonthepreposition-ofisfurthercompoundedbytheselectionerrors
describedinthefollowingsection.
4.3.2.2Selectionerrorswithpreposition-of
Selectionerrorswithpreposition-ofarethesecondmostfrequenterrortypein
thiserrorpattern,comprising37.18%ofthetotalerrorswith-ofpattern(Table4.19).
Excerpts76and77illustratetwoselectionerrorswith-of.
Selectionerrorswithpreposition-of
76.
activecomponentsofforfullyflexiblethermoelectricmodules[11ECS]
77.
halfofasthatinavacuumof3.0x108msec-1[02ECS]
InExcerpt76,ofismis-selectedforfor,andinExcerpt77,asismis-selectedforof.The
tendencyoftheJapanesescientistsinthisstudyistoover-selectof,asinExcerpt76,
87
Chapter4:Findings
ratherthanunder-selectit,asinExcerpt77.Notably,ofismis-selectedforanother
preposition(i.e.,ofisusedwhenanotherprepositionismoreappropriate)
in26ofthe28selectionerrors,whileitisunder-selectedonlytwice.Table4.20presents
thefrequenciesofthevariousselectionerrorswith-ofintheCOJSD,highlightingthe
mis-selectionofofforinasthemostfrequent.
88
Chapter4:Findings
Table4.20Selectionerrorswithpreposition-of
Frequency
Mis-
acrossthe
Excerpt
selection
NGS
78.
offorin
12
offorfor
9
offorwith
2
offorfrom
1
offoron
1
theon-stateVgsofat10V[13ECS]
offorat
1
halfofasthatinavacuumof3.0x108msec-1
asforof
1
toforof
1
thedecreaseofinthenumberofenergy
barriersatdomainboundariesintheTFT
channel[01ECS]
79.
Thereasonofforthisinvestigation[08ECS]
80.
temperaturedependenceofwithVthinstability
[13ECS]
81.
SuchadrasticinfluenceoffromtheOTS
treatment[01ECS]
82.
theauto-oxidationofthecysteinethiolgroup
ofontheenzymes[08ECS]
83.
84.
[02ECS]
85.
aswappingofsignsinchiropticalpolarization
tooftheparticles[02ECS]
TOTAL
28
Note.NGSreferstotheNominalGroupSubcorpus.
Inshort,theresultsforselectionerrorsfurtherhighlighttheJapanesescientists'
over-relianceonthepreposition-of.Corroboratingthisfinding,awordfrequency
89
Chapter4:Findings
analysis56revealsthatofoccurs1942timesinthe46,263-wordCOJSD,indicatinga
coverageof4.20%.ComparedwiththecoverageintheBrownandBNCcorpora(i.e.,
3.09%),thisisanincreaseofmorethan1%.57Giventheaveragelengthofatextinthe
COJSDis3,559words,a1%increaseintheuseofofresultsinanadditional35
occurrencespertext,thusgreatlyincreasingthepotentialforerror.Theidentificationof
thisover-usecouldbeafirststeptowardsbothconsciousness-raisingandinstructional
interventiontoaddressthiserrorpattern.Inotherwords,simplyhighlightingthis
overusetoJapanesescientistsmaymotivateandhelpthemunderstandtheneedto
explorealternativeformsofcomplexnominalgroupmodification,especiallypremodification.
4.3.2.3Omissionerrorswithpreposition-of
Omissionerrorswith-ofoccurin7ofthe13COJSDtextsandcomprise15.38%of
all-oferrors(Table4.19).Therearestraightforwardcasesofomissioninwhichthe-of
prepositionissimplyomitted,suchasthatillustratedinExcerpt86.
Straightforward-ofomissionerror
86.
theparticipationofGlul24and/orArg164inthefixationofPAC-1[08ECS]
However,theomissionerrorswith-ofaretypicallymorecomplex.Excerpt87illustrates
amorecommon-ofomissionerror.
Complex-ofomissionerror
87.
theequilibriumamongtheconformationofvariouspro-chiralsubstrates'
conformation[03ECS]
Inomitting-of,theerrorinExcerpt87inappropriatelydeploysthevariouspro-chiral
substratesasthepossessiveoftheHeadconformation(i.e.,thevariouspro-chiral
substrates'conformation).Whilethispossessiveformisagrammaticalpossibility,with
anunderstandingofthecontext,itisapparentthatthefocusoftheequilibriumisinfact
56UsingAntConc3.4.3(Anthony,2014)
57IntheBrownCorpus(written),ofoccurs30,971timespermillionwords(i.e.,coverageof
3.09%)(Francis&Kucera,1961).Similarly,inthe100millionwordspokenandwrittenBritish
NationalCorpus,ofoccurs3,009,801times(BritishNationalCorpus,2007).
90
Chapter4:Findings
ontheconformation,andnotthevariousprochiralsubstrates.Thecomplexnatureofthis
errorpromptsfurtherelaboration.
AsdiscussedinSection4.3.2.1,the"unmarked"focusofthenominalgroupfalls
onthefinalelementofthegroup.58However,thereisanimportantexceptioninvolvingof,knownfunctionallyastheFocus(Martin,Matthiessen&Painter,2010,pp.169-171;
originallytermedtheFacetfunctionbyHallidayandMatthiessen,2004,pp.333-335).
TheFocusfunction,realisedbyanembeddednominalgroupwith-ofasLinker,marks
thefocusofmeaningatthefrontofthenominalgroup,asillustratedinExcerpt88.59
Focus 88. the
Linker Deictic
conformation of
various
Classifier
Thing
pro-chiral substrates [03ECS]
Thedeploymentofthislinguisticresourcehasadistinctfunctioninthetextualflowof
meaningindiscourse.Inotherwords,thepositioningofelementsinthenominalgroup
canchangetheintendedmeaning.WithoutanunderstandingoftheFocusfunction,
Japanesescientistsaremorelikelytomisplacethefocusofnominalgroups.Therefore,
pedagogyaimedatimprovingaccuracyinnominalgroupmodificationforJapanese
scientistsmusttakeaccountofthisfunction.Giventherelatednatureofthe-ofaddition
errors(Section4.3.2.1),thispedagogymaycontrasttheFocusfunctionwiththegivennewtextualprosodiesofnominalgroups.
4.3.2.4Orderingerrorswithpreposition-of
Onlytwoorderingerrorswith-ofareidentifiedintheNGS,makingthistheleast
frequenterrortypeintheerrorswith-ofpattern.Notably,theissueisonceagainwith
thechoiceinlocationofthequalifyingmeaning,asillustratedinExcerpt89.
58Forexample,themeaningin(a)and(b)isexperientiallythesame:(a)theroomtemperature
samples,and(b)thesamplesatroomtemperature.Howevertextually,in(b)thepost-modifying
elementatroomtemperatureislikelytobecarriedforthasthe"newsworthy"elementinthe
given-newtextualflowofthediscourse.
59AdaptedfromMartin(2013,p.65),wholiststhefollowingasexamples:thefrontofstage,a
collectionofsongs,achapterofthebook,etc.
91
Chapter4:Findings
Orderingerrorwith-of
89.
therefractiveindexofthecosolventsat589nm(nD)ofthecosolvents
[02ECS]
InExcerpt89,theerroneousembedded-ofprepositionphrase(i.e.,ofthecosolvents)at
theendofthegroupcausesaninterruptiontoreadingflowandisthusmore
appropriatelyrelocateddirectlyfollowingtheHead,index.Whilethisorderingerror
typeisinfrequentinthedata,forpedagogicalpurposesitcanbetreatedaspartofthe
errorswith-ofpattern,andinparticularaspartofthelargerissueoflocatingmeaning
withinthenominalgrouptomostappropriatelymaintainthegiven-newprosody,as
discussedinSection4.3.2.1.
92
Chapter4:Findings
4.3.2.5Summaryoferrorswith-of
Duetoitscommoncolligationwithnounsandnominalisations(Hunston,2002),
thepreposition-ofisoneofthemostfrequentlyoccurringwordsinEnglish,accounting
foraround3%ofallwrittenandspokendiscourse(BNC,2007).Aspreviously
demonstratedwiththeerrorswitharticlesandplural-s(Section4.3.1),thishigh
occurrenceaffordsgreaterpotentialforerror.Errorswith-ofconstitutethesecond
mostprevalenterrorpatterntoemergefromtheNGS,accountingforoneineveryten
errorsintheNGS(Table4.18).Errorswith-ofareprimarilylocatedinthePost-modifier
ofthenominalgroup,thoughtheyalsoinvolvepre-modification.Additionandselection
errorsarethemostcommonerrortypesoccurringwithinthispattern,accountingfor
82.05%oftheerrors(i.e.,44.87%and37.18%,respectively;Table4.19).Theseaddition
andselectionerrorshighlighttheparticipants'over-relianceonof;particularly,their
tendencytoemployembeddedof-prepositionalphrasesinthePost-modifierwhenthe
meaningwouldbemoreappropriatelyconfiguredasEpithetsorClassifiersinthePremodifier.
Thefindingsfromthisinvestigationsuggestthatpedagogyaimedatimproving
accuracyintheL2EnglishwritingofJapanesescientistsshouldconsiderthisoverrelianceonofandthevariouswaysitmanifestsaserrors.Instructionmayhighlightthe
subtlebutimportantdifferencebetweenplacingmeaningsinthePre-orPost-modifier;
namely,thelastpartofthenominalgrouphasthegreaterpotentialfornews,but-of
constructionscanshiftthefocusofthenominalgrouptothefront.
93
Chapter4:Findings
4.4ChapterSummary
Thischapterhasoutlinedtheanalysisoferrorsinthe46,263-wordCorpusof
JapaneseScientificDiscourse(COJSD).Atotalof1,368errorswererecognisedand
reconstructed,correspondingwithanaverageof31errorsper1,000words,oronein
everythreesentences.Aroundhalfoftheerrorswerefoundtooccurwithinnominal
groups(i.e.,598errors)andtheseerrorswereexaminedcloselyasthefocusofthis
investigation(i.e.,theNGS).Analysesoftheerrorsinnominalgroupsrevealsthat
complexnominalgroups,andthevariousformsofmodificationthatareafeatureof
complexnominalgroups,accountfor99.83%oftheerrorsintheNGS.Incontrast,
simplenominalgroups,whichhavenomodification,presentnodifficultiesforthe
Japanesescientistsinthisstudy.Thevariouserrorswithincomplexnominalgroups
wereclassifiedaccordingtomodificationandlocationwithinthegroup(i.e.,their
locationinthePre-modifier,Head,orPost-modifier).Resultsindicatethaterrorsinthe
Pre-modifieroccurtwiceasfrequentlyaserrorsinthePost-modifier,andfourtimes
morefrequentlythanerrorsintheHead,highlightingthedifficultiesofnominalgroup
pre-modificationforJapanesescientists.Fromtheanalysisoferrorsinthemodifying
elementsofcomplexnominalgroups,twomajorerrorpatternsemergedfromthedata:
errorswitharticlesandplural-s,anderrorswithpreposition-of.
ErrorswitharticlesoccurintheDeicticinbothPre-andPost-modifiers.This
errorpatternalsoextendstotheHeadtoincludeerrorswithplural-s.Errorswith
articlesandplural-swereclassifiedaccordingtothefourerrortypes:omission,
addition,selection,andordering.ThisanalysisrevealsthattheJapanesescientistsinthis
studyprimarilyomitarticlesandplural-s.Furtheranalysisemployingthesystem
networkmappingthegrammaticalchoicesofarticlesandplural-sdemonstratesthatthe
Japanesescientistsinthestudymakeerrorsmostfrequentlywithsingularreferents,
andinparticularwithnon-specificsingularreferents.
Thisinvestigationoferrorswitharticlesandplural-shighlightstheintegrated
natureofthechoicesintheunderlyinggrammaticalsystemsandtheerrorsthatoccur
withinthosesystems(e.g.,ahighlevelsofcircularpolarization[10ECS]).Inotherwords,
thechoiceofdefiniteorindefinitearticleisconcurrentwithandcontingentonthe
choiceinsingular,plural,ormass.Itisthereforenecessaryforneeds-basedpedagogy
94
Chapter4:Findings
aimedataddressingtheseerrorstoaccountforthisintegration.Forexample,thesystem
networkpresentedinFigure3.5inthisstudycouldbeutilisedtoillustratethose
grammaticalchoicestoJapanesescientists,andsamplesfromthisstudycouldbe
presentedtoillustratethewaystheerrorsinthesystemsunfold.
ThesecondmajorerrorpatternthatemergedfromtheNGSdatawaserrors
involvingthepreposition-of.Errorswith-ofareprevalentthroughouttheCOJSD,
occurringinall13textsandaccountingforaround12%ofthetotalerrorsintheNGS.
Analysisoftheseerrorsaccordingtothefourerrortypes(i.e.,omission,addition,
selection,andordering)revealsadditionandselectionerrorsoccurmostfrequentlyin
theNGS.Overall,theJapanesescientiststendtoover-relyon-ofconstructionsto
construemeaninginnominalgroups,particularlythroughtheuseofembedded-of
prepositionalphrasestopackagequalifyingmeaninginthePost-modifier,whenthat
meaningismoreappropriatelyconfiguredasanEpithetorClassifierinthePre-modifier.
Theimplicationsofthesefindingshighlighttheneedtopreservethegiven-new
prosodiesinnominalgroupstoeffectivelyfocusontheappropriatenominalgroup
element.Furthermore,theresultsindicatethattheFocusfunctionisanarearequiring
consideration(e.g.theresultsofthisstudy,incomparisonwiththisstudy'sresults).
Needs-basedpedagogyaimedataddressingtheuseof-ofmayequipJapanesescientists
withabetterunderstandingofnominalgrouppre-modificationandthevarious
linguisticresourcesavailabletomoreaccuratelyandcoherentlypackagemeaningin
nominalgroups.
Thetwomajorerrorspatternscombined(i.e.,errorswitharticlesandplural-s,
anderrorswithpreposition-of)accountfor79.36%(Table4.12andTable4.18)ofall
errorsintheNGS.GiventhattheNGScomprises47.81%ofthetotalnumberoferrorsin
theCOJSD,twoineveryfiveerrorsoccurringintheCOJSDhavebeenelaboratedinthis
investigation,revealingthemostfrequentandprevalentpatternsrequiringinstructional
intervention.Furtherinvestigationisnecessary,however,toprovideafulleraccountof
theobstaclesJapanesescientistsfacewhenpreparingmanuscriptsforpublication.
95
Chapter5:Conclusion
Chapter5:Conclusion
Thisstudyhasprovidedanintroductoryexaminationofthesecondlanguage(L2)
EnglisherrorsinthescientificwritingofJapanesescientists,focussingonerrorsin
nominalgroups.Significantly,twomajorerrorpatternsareidentified:errorswith
articlesandplural-s,anderrorswithpreposition-of.Togetherthesetwoerrorpatterns
accountfortwoineveryfiveerrorsinthe46,263-wordcorpusanalysed.This
concludingchaptersummarisesthemajorachievementsofthestudyalongwithits
limitations,andoutlinesrecommendationsforfutureresearch.
ThisstudycontributestothewiderunderstandingofL2errorsbyprovidingan
elaboratedframeworkforErrorAnalysis(EA).ThiselaboratedEAframeworkintegrates
thedelicategrammaticaldescriptionsofSystemicFunctionalLinguistics(SFL)andthe
empiricalbasisofCorpusLinguistics(CL)withinthetraditionalEAframeworkof
analysis.Byintegratingtheseapproaches,thepresentinvestigationandfutureresearch
intoL2errorsbenefitsfromthedecadesoftheoreticalandanalyticaladvancesmadein
thefieldsofSFLandCLsinceEAwasdevelopedinthe1960s-70s.
TheelaboratedEAframeworkresultedintheidentificationoftwomajorerror
patternsintheCorpusofJapaneseScientificDiscourse(COJSD):errorswitharticlesand
plural-s,anderrorswithpreposition-of.Thesefindingsinformamorefocused
pedagogysupportingtheneedsofJapanesescientistsstrivingtoachievethelevelsof
accuracydemandedinscientificpublications.Moreover,theexcerptsanderrorsamples
providedinthisinvestigationserveasfield-specificanddata-drivenresourcestoinform
thatspecialisedpedagogy.
WhilethisstudyhasestablishedanelaboratedEAframeworkandidentifiedtwo
majorL2errorpatternsinthewritingofJapanesescientists,duetoitslimitedscope,the
currentthesisservesasanintroductoryinvestigationoferrorsintheL2Englishof
Japanesescientists.Amongthelimitationsofthisstudyisthesamplesizeofthirteen
texts.Thissamplesizelimitspotentialtoinfertheresultsbeyondthestudy's
participantstowardsthewiderpopulationofJapanesescientists.Similarly,thesample
ofJapanesescientistsparticipatinginthisstudyislimitedtoestablished-career
materialsscientists.Consequently,thefindingscannotbecorroboratedwitherrorsin
96
Chapter5:Conclusion
theL2writingofotherJapanesescientists,especiallyearly-careerscientistssuchasPhD
studentswhoaretypicallywritingtheirfirstpapersinEnglishandmaybeproneto
differenterrorpatterns.
Furtheramongthelimitationsofthisstudyistherestrictedfocusonerrorsin
nominalgroups.Whilenominalgroupsareacentralfeatureofscientificwritinganda
majorsourceoftheerrorsinthisstudy,afullanalysisoferrorsbothwithinandbeyond
theclauseremains;particularly,theremaininggroupsandphraseunits(i.e.,verbal
groups,adverbialgroups,conjunctiongroups,andprepositionalphrases).Furthermore,
forpurposesofscope,thisexaminationoferrorsinnominalgroupsisrestrictedtothe
twomostprevalenterrorpatterns,leavingadditionalpatternstoinvestigatefurther.
Finally,theanalysisoferrorsisonlyaseffectiveastheinitialidentificationand
classification.Theexaminationoferrorsinthisstudyislimitedtooneanalyst.Whilethis
approachfacilitatesconsistency,itsubsequentlylimitsthereliabilityofthefindings.Due
totheselimitations,thisinvestigationmaybestserveasapilotstudyforalarger
examinationofL2errorsinthescientificwritingofJapanesescientists.
Buildingfromthisintroductorystudy,acompleteunderstandingofL2errorsin
thescientificwritingofJapanesescientistsrequiresfurtherinvestigation.Specifically,in
ordertoinferresultstothewiderpopulationofJapanesescientists,futureresearchwill
requirealargersamplesizeapproximatinganormalpopulation.Furthermore,in
additiontoestablished-careerJapanesescientists,suchasampleshouldincludeearlycareerscientists.Theexaminationofalargerandmorediverserangeofparticipanttexts
mayestablishwhethertheerrorpatternsidentifiedinthisstudyareconsistentacross
andwithinthepopulation.Subsequently,thisincreaseindatarequiresmoreefficient
analyses,particularly,corpussoftwaretoolsforelectronicannotationandautomated
analyses.
Additionally,futureworkinvestigatingL2errorsinthescientificwritingof
Japanesescientistsrequiresanexaminationoferrorsbeyondnominalgroupstoinclude
thewholeclause.Suchanexaminationincludesnotonlytheremaininggroupand
phraseunits(i.e.,verbalgroups,prepositionalphrases,conjunctiongroups,adverbial
groups),butalsopotentiallyfunctionalelementswithinandabovetheclause(e.g.,
errorsinParticipants,Processes,Circumstances;anderrorsinThemeandRheme).
Theseanalysesmayalsobenefitfromintelligibilitystudiesmeasuringtheimpactof
97
Chapter5:Conclusion
errorsoncomprehension,andobligatoryoccasionsanalysestocorroboratefrequency
data.
Inconclusion,thepresentinvestigationofL2errorshascontributedsignificantly
totheanalysisoferrorsinJapaneseScientists'researchwritingthroughthe
developmentofanelaboratedEAframeworkandtheidentificationofthetwomost
frequenterrorpatternsinnominalgroups:errorswitharticlesandplural-s,anderrors
withpreposition-of.Thisstudyservesasaninitialsteptowardsalargerendgoal,which
isthedevelopmentandimplementationofneeds-based,field-specific,anddata-driven
pedagogytohelpJapanesescientistsachievethelevelsofsophisticationandaccuracy
requiredforsuccessfulinternationalcommunicationandpublicationinscientific
English.Towardsthisend,thisinvestigationofL2errorsinthescientificwritingof
Japanesescientistswillbefollowedbyamorein-depthandextensiveexaminationofa
largerpopulationofJapanesescientists'researcharticles.
98
References
References
Aarts,J.,&Granger,S.(1998).Tagsequencesinlearnercorpora.InS.Granger(Ed.),
LearnerEnglishoncomputer(pp.pp.132–141).London:Longman.
Abbott,G.(1980).Towardsamorerigorousanalysisofforeignlanguageerrors.
InternationalReviewofAppliedLinguisticsinLanguageTeaching,18(1-4),121-134.
Ammon,U.(2001).ThedominanceofEnglishasalanguageofscience:Effectsonother
languagesandlanguagecommunities(Vol.84).Berlin:MoutondeGruyter.
Ammon,U.(2003).TheinternationalstandingoftheGermanlanguage.InJ.Maurais&
M.Morris(Eds.),LanguagesinaGlobalisingWorld(pp.231-249).Cambridge:
CambridgeUniversityPress.
Ammon,U.(2006).LanguageconflictsintheEuropeanUnion.InternationalJournalof
AppliedLinguistics,16(3),319-338.
Anthony,L.(2014).AntConc(Version3.4.3)[ComputerSoftware].Tokyo:Waseda
University.Retrievedfromhttp://www.laurenceanthony.net/
Austin,J.L.(1962).Howtodothingswithwords.FairLawn,NJ:OxfordUniversityPress.
Barner,D.,Inagaki,S.,&Li,P.(2009).Language,thought,andrealnouns.Cognition,
111(3),329-344.
Basturkmen,H.(2006).IdeasandoptionsinEnglishforspecificpurposes.Mahwah,NJ:
LawrenceErlbaumAssociates.
Belcher,D.D.(2006).Englishforspecificpurposes:Teachingtoperceivedneedsand
imaginedfuturesinworldsofwork,study,andeverydaylife.TESOLQuarterly,40(1),
133-156.
Belcher,D.D.(2007).SeekingacceptanceinanEnglish-onlyresearchworld.Journalof
SecondLanguageWriting,16(1),1-22.
Berry,R.(1991).Re-articulatingthearticles.ELTjournal,45(3),252-259.
Biber,D.(1988).Variationacrossspeechandwriting.Cambridge:CambridgeUniversity
Press.
Biber,D.(2009).Acorpus-drivenapproachtoformulaiclanguageinEnglish.
InternationalJournalofCorpusLinguistics,14(3),275-311.
Biber,D.,Johansson,S.,Leech,G.,Conrad,S.,&Finegan,E.(1999).Longmangrammarof
spokenandwrittenEnglish(Vol.2).London:Longman.
Blake,N.F.,Lass,R.,&Romaine,S.(1992).TheCambridgehistoryoftheEnglishlanguage
(Vol.6).Cambridge:CambridgeUniversityPress.
Bley-Vroman,R.(1983).Thecomparativefallacyininterlanguagestudies:Thecaseof
systematicity.LanguageLearning,33(1),1-17.
99
References
BNC.(2007).TheBritishNationalCorpus,version3(BNCXMLedition).Oxford
UniversityComputingServicesonbehalfoftheBNCConsortium.Retrievedfrom
http://www.natcorp.ox.ac.uk/
Burt,M.K.,&Kiparsky,C.(1972).Thegooficon:ArepairmanualforEnglish.Rowley,MA:
NewburyHousePublishers,Inc.
Butler,C.S.(1999).Somepossiblecontributionsofcorpuslinguisticstothefunctionallexematicmodel.InM.J.FeuGuijarro&S.MolinaPlaza(Eds.),Estudiosfuncionales
sobreléxico,sintaxisytraducción:UnhomenajeaLeocadioMartínMingorance
(Functionalstudiesonlexis,syntaxandtranslation:AtributetoLeacadioMartin
Mingorance(pp.19-38).Cuenca:UniversidaddeCastilla-LaMancha.
Butler,Y.G.(2002).Secondlanguagelearners'theoriesontheuseofEnglisharticles.
StudiesinSecondLanguageAcquisition,24(03),451-480.
Canagarajah,A.S.(2002).Ageopoliticsofacademicwriting.Pittsburgh,PA:Universityof
PittsburghPress.
Cargill,M.,&Burgess,S.(2008).Introductiontothespecialissue:Englishforresearch
publicationpurposes.JournalofEnglishforAcademicPurposes,7(2),75-76.
Chen,L.L.(2006).TheeffectoftheuseofL1inamultimediatutorialongrammar
learning:AnerroranalysisofTaiwanesebeginningEFLlearners’Englishessays.
AsianEFLJournal,8(2),76-110.
Cook,V.J.(1993).Linguisticsandsecondlanguageacquisition.London:Macmillan.
Corder,S.P.(1967).Thesignificanceoflearner'serrors.InternationalReviewofApplied
LinguisticsinLanguageTeaching,5(1-4),161-170.
Corder,S.P.(1971).Idiosyncraticdialectsanderroranalysis.InternationalReviewof
AppliedLinguisticsinLanguageTeaching,9(2),147-160.
Corder,S.P.(1981).Erroranalysisandinterlanguage.Oxford:OxfordUniversityPress.
Coxhead,A.(2000).Anewacademicwordlist.TESOLQuarterly,34(2),213-238.
Crompton,P.(2005).‘Where’,‘InWhich’,and‘InThat’:ACorpus-BasedApproachto
ErrorAnalysis.RELCJournal,36(2),157-176.
Crystal,D.(2003).Englishasagloballanguage(2nded.).Cambridge:Cambridge
UniversityPress.
Curry,M.J.,&Lillis,T.(2004).Multilingualscholarsandtheimperativetopublishin
English:Negotiatinginterests,demands,andrewards.TESOLQuarterly,38(4),663688.
Dotti,F.,&O’Donnell,M.(2014).AddressingarticleerrorsinSpanishlearnersofEnglish
usingalearnercorpus.Paperpresentedatthe2014TSLLConference,IowaState
University,September2014.
http://www.uam.es/proyectosinv/treacle/Publications/Dotti-ODonnellArticlePaper.pdf
100
References
Dudley-Evans,T.,&StJohn,M.J.(1998).DevelopmentsinEnglishforspecificpurposes:A
multi-disciplinaryapproach.Cambridge:CambridgeUniversityPress.
Dulay,H.,Burt,M.,&Krashen,S.(1982).Languagetwo.NewYork,NY:OxfordUniversity
Press.
Eggins,S.(2004).Introductiontosystemicfunctionallinguistics(2nded.).London:
Continuum.
Evison,J.(2010).Whatarethebasicsofanalysingacorpus?InA.O’Keeffe&M.J.
McCarthy(Eds.),TheRoutledgeHandbookofCorpusLinguistics(pp.122-135).
London:Routledge.
Falhasiri,M.,Tavakoli,M.,Hasiri,F.,&Mohammadzadeh,A.(2011).Theeffectivenessof
explicitandimplicitcorrectivefeedbackoninterlingualandintralingualerrors:A
caseoferroranalysisofstudents’compositions.EnglishLanguageTeaching,4(3),
251-264.
Fang,Z.,Schleppegrell,M.J.,&Cox,B.E.(2006).Understandingthelanguagedemandsof
schooling:Nounsinacademicregisters.JournalofLiteracyResearch,38(3),247-273.
Ferguson,G.(2007).TheglobalspreadofEnglish,scientificcommunicationandESP:
Questionsofequity,accessanddomainloss.Ibérica,13,7-38.
Fieder,N.,Nickels,L.,&Biedermann,B.(2014).Representationandprocessingofmass
andcountnouns:areview.FrontiersinPsychology,5.
Flowerdew,J.(2006).Useofsignallingnounsinalearnercorpus.InternationalJournal
ofCorpusLinguistics,11(3),345-362.
Flowerdew,J.(2013a).Englishforresearchpublicationpurposes.InB.Paltridge&S.
Starfield(Eds.),TheHandbookofEnglishforSpecificPurposes(pp.301-321).
Chichester:JohnWiley&Sons,Ltd.
Flowerdew,J.(2013b).SomethoughtsonEnglishforresearchpublicationpurposes
(ERPP)andrelatedissues.LanguageTeaching,1-13.
Francis,W.N.,&Kucera,H.(1961).BrownCorpus.Retrievedfrom
https://the.sketchengine.co.uk/open/
Glover,A.,Johnson,E.,Strevens,P.,&Weeks,F.(1988).SeaSpeaktrainingmanual.
Oxford:PergamonPress.
Gordin,M.D.(2015).Scientificbabel:Howsciencewasdonebeforeandafterglobal
English.Chicago:UniversityofChicagoPress.
Gosden,H.(2003).‘Whynotgiveusthefullstory?’:functionsofreferees’commentsin
peerreviewsofscientificresearchpapers.JournalofEnglishforAcademicPurposes,
2(2),87-101.
Graddol,D.(1997).ThefutureofEnglish?:Aguidetoforecastingthepopularityofthe
Englishlanguageinthe21stcentury.London:BritishCouncil.
101
References
Granger,S.(2003).TheinternationalcorpusoflearnerEnglish:Anewresourcefor
foreignlanguagelearningandteachingandsecondlanguageacquisitionresearch.
TESOLQuarterly,37(3),538-546.
Gressang,J.E.(2010).Afrequencyanderroranalysisoftheuseofdeterminers,the
relationshipsbetweennounphrases,andthestructureofdiscourseinEnglishessaysby
nativeEnglishwritersandnativeChinese,Taiwanese,andKoreanlearnersofEnglishas
aSecondlanguage.(DoctorofPhilosophy),UniversityofIowa.Retrievedfrom
http://ir.uiowa.edu/etd/507
Hagen,A.S.(2016).Englishspreadingworldwide.Retrievedfrom
http://ndla.no/en/node/9166
Halliday,M.A.K.(1985).Spokenandwrittenlanguage.Geelong,Victoria:Deakin
UniversityPress(republishedbyOxfordUniversityPress1989).
Halliday,M.A.K.(2004).Thelanguageofscience.London:Continuum.
Halliday,M.A.K.,&Martin,J.R.(1993).Writingscience:Literacyanddiscursivepower.
London:TheFalmerPress.
Halliday,M.A.K.,&Matthiessen,C.M.I.M.(2004).Anintroductiontofunctional
grammar(revised3rded.).London:EdwardArnold.
Halliday,M.A.K.,&Matthiessen,C.M.I.M.(2014).Halliday'sintroductiontofunctional
grammar.Oxford:Routledge.
Halliday,M.A.K.,&Webster,J.J.(2009).StudiesinEnglishlanguage(Vol.7).London:
Continuum.
Hamilton,C.E.(2015).Thecontributionofsystemicfunctionalgrammartotheerror
analysisframework.TESOLInternationalJournal,10(1),11-28.
Hammarberg,B.(1974).Theinsufficiencyoferroranalysis.InternationalReviewof
AppliedLinguisticsinLanguageTeaching,12(1-4),185-192.
Hasan,R.(2014).Towardsaparadigmaticdescriptionofcontext:systems,
metafunctions,andsemantics.FunctionalLinguistics,1(1),1-54.
Heyvaert,L.(2008).Ontheconstructionalsemanticsofgerundivenominalizations.Folia
Linguistica,42(1-2),39-82.
Hong,A.L.,Rahim,H.A.,Hua,T.K.,&Salehuddin,K.(2011).CollocationsinMalaysian
EnglishLearners'Writing:ACorpus-basedErrorAnalysis.3L:TheSoutheastAsian
JournalofEnglishLanguageStudies,17(SpecialIssue),31-44.
Hunston,S.(2002).Corporainappliedlinguistics.Cambridge:CambridgeUniversity
Press.
Hutchinson,T.,&Waters,A.(1987).Englishforspecificpurposes.Cambridge:Cambridge
UniversityPress.
Hyland,K.(2009).Academicdiscourse:Englishinaglobalcontext.London:Continuum
InternationalPublishingGroup.
102
References
Hyland,K.(2010).Constructingproximity:Relatingtoreadersinpopularand
professionalscience.JournalofEnglishforAcademicPurposes,9(2),116-127.
Hyland,K.,&Hamp-Lyons,L.(2002).EAP:Issuesanddirections.JournalofEnglishfor
AcademicPurposes,1(1),1-12.
James,C.(1998).Errorsinlanguagelearninganduse:Exploringerroranalysis.Harlow:
Longman.
Johns,T.,&King,P.(1991).Classroomconcordancing.EnglishLanguageResearch
Journal,4.
Jordan,R.(1997).Englishforacademicpurposes:Aguideandresourcebookforteachers.
Cambridge:CambridgeUniversityPress.
Kachru,B.B.(Ed.)(1992).Theothertongue:Englishacrosscultures(2nded.).Urbana.
IL:UniversityofIllinoisPress.
Keshavarz,M.H.(1999).Contrastiveanalysisanderroranalysis.Tehran:Rahnama.
Khodabandeh,F.(2007).Analysisofstudents'errors:Thecaseofheadlines.TheAsian
ESPJournal,3(1),6-21.
Kim,M.(2010).Translationerroranalysis:Asystemicfunctionalgrammarapproach.In
C.Coffin,T.Lillis,&K.O'Halloran(Eds.),AppliedLinguisticsMethods:AReader(pp.
84-94).London:Routledge.
Kirkpatrick,A.(2007).WorldEnglishes:Implicationsforinternationalcommunication
andEnglishlanguageteaching.Cambridge:CambridgeUniversityPress.
Klages-Kubitzki,M.(1995).ArticleusageinEnglish:acomputer-basedself-teaching
programmeonthebasisofafunctionaltheoryofreference.FrankfurtamMain:Peter
Lang.
Klein,W.(1998).Thecontributionofsecondlanguageacquisitionresearch.Language
Learning,48(4),527-549.
Lees,R.B.(1960).ThegrammarofEnglishnominalisation(Vol.12).Bloomington,IN:
IndianaUniversityResearchCenterinAnthropology,Folklore,andLinguistics.
Levelt,W.J.M.(1978).Skilltheoryandlanguageteaching.StudiesinSecondLanguage
Acquisition,1(1),53-70.
Lillis,T.M.,&Curry,M.J.(2010).Academicwritinginglobalcontext:Thepoliticsand
practicesofpublishinginEnglish.London:Routledge.
Long,M.H.(Ed.)(2006).Secondlanguageneedsanalysis.Cambridge:Cambridge
UniversityPress.
Mabe,M.(2003).Thegrowthandnumberofjournals.Serials,16(2),191-197.
Martin,J.R.(2013).Systemicfunctionalgrammar:Anextstepintothetheory:Axial
relations.Beijing:HigherEducationPress.
103
References
Martin,J.R.(2014).Evolvingsystemicfunctionallinguistics:Beyondtheclause.
FunctionalLinguistics,1(1),1-24.
Martin,J.R.,Matthiessen,C.M.,&Painter,C.(2010).Deployingfunctionalgrammar(2nd
ed.).Beijing:CommercialPress.
Martin,J.R.,&Veel,R.(Eds.).(1998).Readingscience:Criticalandfunctionalperspectives
ondiscoursesofscience.London:Routledge.
Master,P.(1997).TheEnglisharticlesystem:Acquisition,function,andpedagogy.
System,25(2),215-232.
Matthiessen,C.M.,&Halliday,M.A.(2009).Systemicfunctionalgrammar:afirststep
intothetheory.Beijing:HigherEducationPress.
Matthiessen,C.M.I.M.,Teruya,K.,&Lam,M.(2010).Keytermsinsystemicfunctional
linguistics.London:Continuum.
McCabe,A.,&Gallagher,C.(2008).Theroleofthenominalgroupinundergraduate
academicwriting.InC.Jones&E.Ventola(Eds.),Newdevelopmentsinthestudyof
ideationalmeaning:Fromlanguagetomultimodality(pp.189-208).London:Equinox.
Meyer,C.F.(2002).Englishcorpuslinguistics:Anintroduction.Cambridge:Cambridge
UniversityPress.
Neeley,T.(2012).GlobalbusinessspeaksEnglish:Whyyouneedalanguagestrategy
now.Harvardbusinessreview,90,116-124.
O'Donnell,M.(2015).UAMCorpusTool3.2[Computersoftware].Retrievedfrom
http://www.wagsoft.com/CorpusTool/download.html
OECD.(2014).EducationataGlance2014:OECDIndicators.Retrievedfrom
http://www.oecd.org/education/eag.htm.
Okamura,A.(2006).HowdoJapaneseresearcherscopewithlanguagedifficultiesand
succeedinscientificdiscourseinEnglish?:InterviewswithJapaneseresearcharticle
writers.TheEconomicJournalofTakasakiCityUniversityofEconomics,48(3),61-78.
Ortega,L.(2013).SLAforthe21stcentury:Disciplinaryprogress,transdisciplinary
relevance,andthebi/multilingualturn.LanguageLearning,63(s1),1-24.
Paltridge,B.,&Starfield,S.(2013).ThehandbookofEnglishforspecificpurposes.
Chichester:JohnWiley&Sons.
Park,K.(2012).Learner–corpusinteraction:Alocusofmicrogenesisincorpus-assisted
L2writing.AppliedLinguistics,33(4),361-385.
Pennycook,A.(2014).TheculturalpoliticsofEnglishasaninternationallanguage.
Oxford:Routledge.
Phillipson,R.(1992).LinguisticImperialism.Oxford:OxfordUniversityPress.
Richards,J.C.(1971).Anon-contrastiveapproachtoerroranalysis.EnglishLanguage
Teaching,25,204-219.
104
References
Richards,J.C.(1974).Erroranalysis:Perspectivesonsecondlanguageacquisition.
London:Longman.
Robertson,D.(2000).VariabilityintheuseoftheEnglisharticlesystembyChinese
learnersofEnglish.SecondLanguageResearch,16(2),135-172.
Robertson,F.A.(1988).Airspeak:Radiotelephonycommunicationforpilots.Hemel
Hempstead:PrenticeHall.
Ronen,S.,Gonçalves,B.,Hu,K.Z.,Vespignani,A.,Pinker,S.,&Hidalgo,C.A.(2014).Links
thatspeak:Thegloballanguagenetworkanditsassociationwithglobalfame.
ProceedingsoftheNationalAcademyofSciences,111(52),E5616-E5622.
Rozycki,W.,&Johnson,N.H.(2013).Non-canonicalgrammarinBestPaperaward
winnersinengineering.EnglishforSpecificPurposes,32(3),157-169.
Schachter,J.(1974).Anerrorinerroranalysis.LanguageLearning,24(2),205-214.
Schachter,J.,&Celce-Murcia,M.(1977).Somereservationsconcerningerroranalysis.
TESOLQuarterly,11(4),441-451.
Schleppegrell,M.J.(2002).ChallengesofthescienceregisterforESLstudents:Errors
andmeaning-making.InM.J.Schleppegrell&M.C.Colombi(Eds.),Developing
advancedliteracyinfirstandsecondlanguages:Meaningwithpower(pp.119-142).
Mahwah,N.J.:LawrenceErlbaumAssociates
Schleppegrell,M.J.(2004).Thelanguageofschooling:Afunctionallinguisticsperspective.
Mahwah,NJ:LawrenceErlbaumAssociates.
Selinker,L.(1972).Interlanguage.InternationalReviewofAppliedLinguisticsin
LanguageTeaching,10,209-231.
Selinker,L.(1992).Rediscoveringinterlanguage.London:Longman.
Strevens,P.(1969).Twowaysoflookingaterror-analysis.PaperpresentedattheGAL
Meeting,Stuttgart.
Swales,J.(1990).Genreanalysis:Englishinacademicandresearchsettings.Cambridge:
CambridgeUniversityPress.
Taylor,J.R.(2012).Thementalcorpus:Howlanguageisrepresentedinthemind.Oxford:
OxfordUniversityPress.
Thompson,G.(1996).Introducingfunctionalgrammar.London:Arnold.
Thomson-Reuters.(2009).ScienceWatch.Retrievedfrom
http://archive.sciencewatch.com/dr/cou/2009/09decALL/
W3Techs.(2016)Usageofcontentlanguagesforwebsites.Retrievedfrom
http://w3techs.com/technologies/overview/content_language/all
Weber,G.(1997).Theworld's10mostinfluentiallanguages.LanguageToday,2(3),1218.
105
References
White,L.(2003).FossilizationinsteadystateL2grammars:Persistentproblemswith
inflectionalmorphology.Bilingualism:Languageandcognition,6(02),129-141.
Whittaker,R.,Llinares,A.,&McCabe,A.(2011).WrittendiscoursedevelopmentinCLIL
atsecondaryschool.LanguageTeachingResearch,15(3),343-362.
Willey,I.,&Tanimoto,K.(2012).“ConvenienceEditing”inaction:ComparingEnglish
teachers’andmedicalprofessionals’revisionsofamedicalabstract.Englishfor
SpecificPurposes,31(4),249-260.
Willey,I.,&Tanimoto,K.(2013).“Convenienceeditors”aslegitimateparticipantsinthe
practiceofscientificediting:Aninterviewstudy.JournalofEnglishforAcademic
Purposes,12(1),23-32.
WIPO.(2014).Worldintellectualpropertyindicators:Totalpatentgrantsbycountryof
origin:2000-2014.Retrievedfrom
http://www.wipo.int/ipstats/en/statistics/country_profile/profile.jsp?code=JP
Yamazaki,S.(1995).Refereeingsystemof29lifesciencejournalspreferredbyJapanese
scientists.Scientometrics,33(1),123-129.
106
Appendices
Appendices
AppendixA:EthicsApprovalLetter
Office of the Deputy Vice-Chancellor
(Research)
Research Office
Research Hub, Building C5C East
Macquarie University
NSW 2109 Australia
T: +61 (2) 9850 4459
http://www.research.mq.edu.au/
ABN 90 952 801 237
CRICOS Provider No 00002J
18 May 2015
Dr Cassi Liardet
Macquarie University
NSW 2109
Dear Dr Liardet,
Reference No: 5201500154
Title: Japanese scientists' research articles: An error analysis
Thank you for submitting the above application for ethical and scientific review. Your
application was considered by the Macquarie University Human Research Ethics
Committee (HREC (Human Sciences & Humanities)) at its meeting on 27 March 2015 at
which further information was requested to be reviewed by the Ethics Secretariat.
The requested information was received with correspondence on 8 and 14 May 2015.
I am pleased to advise that ethical and scientific approval has been granted for this project
to be conducted at:
Macquarie University
This research meets the requirements set out in the National Statement on Ethical Conduct
in Human Research (2007 – Updated March 2014) (the National Statement).
This letter constitutes ethical and scientific approval only.
Standard Conditions of Approval:
1. Continuing compliance with the requirements of the National Statement, which is
available at the following website:
http://www.nhmrc.gov.au/book/national-statement-ethical-conduct-human-research
2. This approval is valid for five (5) years, subject to the submission of annual reports.
Please submit your reports on the anniversary of the approval for this protocol.
3. All adverse events, including events which might affect the continued ethical and
scientific acceptability of the project, must be reported to the HREC within 72 hours.
4. Proposed changes to the protocol must be submitted to the Committee for approval
before implementation.
107
Appendices
It is the responsibility of the Chief investigator to retain a copy of all documentation related
to this project and to forward a copy of this approval letter to all personnel listed on the
project.
Should you have any queries regarding your project, please contact the Ethics Secretariat on
9850 4194 or by email [email protected]
The HREC (Human Sciences and Humanities) Terms of Reference and Standard Operating
Procedures are available from the Research Office website at:
http://www.research.mq.edu.au/for/researchers/how_to_obtain_ethics_approval/human
_research_ethics
The HREC (Human Sciences and Humanities) wishes you every success in your research.
Yours sincerely
Dr Karolyn White
Director, Research Ethics & Integrity,
Chair, Human Research Ethics Committee (Human Sciences and Humanities)
This HREC is constituted and operates in accordance with the National Health and Medical
Research Council's (NHMRC) National Statement on Ethical Conduct in Human Research
(2007) and the CPMP/ICH Note for Guidance on Good Clinical Practice.
108
Appendices
AppendixB:ParticipantConsentForm
DEPARTMENT OF
LINGUISTICS
Faculty of Human Sciences
Macquarie University
NSW 2109 Australia
T: +61 (2) 9850 6701
F: +61 (2) 9850 9199
[email protected]
CRICOS Provider No 00002J
Chief Investigator: Dr. Cassi Liardét, Lecturer, Linguistics
Co-Investigator: Leigh McDowell, Student, Macquarie University
Participant Information and Consent Form
Name of Project: Japanese scientists' research articles: A needs analysis
Dear Japanese scientist,
You are invited to participate in a study of Japanese scientists' research articles. The purpose of the
study is to collect English manuscripts to build a corpus titled the Corpus of Japanese scientific
discourse. This corpus will be analysed for patterns of lexico-gramatical errors (e.g. the misuse of
"consisted with" for "consisted of"). By understanding these error patters better, we hope to provide
better education and support for Japanese scientists who need to publish their research in
international journals.
The study is being conducted by Cassi Liardét (Chief Investigator) of the Department of Linguistics,
Faculty of Human Sciences, Macquarie University, and Leigh McDowell, a Higher Degree Research
student at Macquarie University. This project is designed for Leigh McDowell's master's degree
research. Contact details of the investigators are as follows.
Dr. Cassi Liardét (Chief Investigator)
[email protected]
Bldg C5A, Rm. 537
Macquarie University, NSW 2109
Ph: +61 (0)2 9850 6704
Leigh McDowell (Co Investigator)
[email protected]
Department of Linguistics, Faculty of Human Sciences
Macquarie University, NSW 2109
If you decide to participate, you will be asked to send by email an electronic copy of the English
manuscript of one of your recent publication drafts. You will also be asked to complete a brief
questionnaire to provide a few details (e.g. your position at NAIST, name of the journal to which you
submitted the manuscript, etc.). This is a one-off process which should not take any more than 20
minutes in total.
Local contact details if you have any questions:
630-0192
8916-5
0743-72-6172
[email protected]
109
Appendices
DEPARTMENT OF
LINGUISTICS
Faculty of Human Sciences
Macquarie University
NSW 2109 Australia
T: +61 (2) 9850 6701
F: +61 (2) 9850 9199
[email protected]
CRICOS Provider No 00002J
Any information or personal details gathered in the course of the study are confidential, except as
required by law. No individual will be identified in any publication of the results. Only the chief
investigator (Cassi Liardét) and her co investigator (Leigh McDowell) will have access to your
name. If you choose to participate, you will be assigned a number to protect your anonymity both
in the collected corpus and in any related publications. A short report of the results of the study will
be made available to you via email on an annual basis.
Participation in this study is entirely voluntary: you are not obliged to participate and if you decide to
participate, you are free to withdraw at any time without having to give a reason and without
consequence. Note: If you decide not to participate in the current study, your manuscript will still be
reviewed by the researchers but it will not be included in the corpus for any research analysis.
I, __________________________ (participant’s name) have read and understand the information
above and any questions I have asked have been answered to my satisfaction. I agree to participate
in this research, knowing that I can withdraw from further participation in the research at any time
without consequence. I have been given a copy of this form to keep.
Participant’s Name:
(Block letters)
Participant’s Signature:______________________________ Date:
Investigator’s Name:
(Block letters)
Investigator’s Signature: _________________________ ___ Date:
The ethical aspects of this study have been approved by the Macquarie University Human
Research Ethics Committee. If you have any complaints or reservations about any ethical
aspect of your participation in this research, you may contact the Committee through the
Director, Research Ethics (telephone (02) 9850 7854; email [email protected]). Any
complaint you make will be treated in confidence and investigated, and you will be informed
of the outcome.
PARTICIPANT'S COPY
Local contact details if you have any questions:
630-0192
8916-5
0743-72-6172
[email protected]
110
Appendices
AppendixC:ParticipantQuestionnaireandResponses
Q1.IneedtopublishmyresearchininternationaljournalsinEnglish.
#
Answer
Response
%
11
100%
0
0%
0
0%
0
0%
0
0%
11
100%
1
Strongly
Agree
2
Agree
Neither
3
Agreenor
Disagree
4
Disagree
5
Strongly
Disagree
Total
Q2.IneedtopublishmyresearchinJapanesejournalsinJapanese.
#
Answer
Response
%
0
0%
7
64%
3
27%
1
9%
0
0%
11
100%
1
Strongly
Agree
2
Agree
Neither
3
Agreenor
Disagree
4
Disagree
5
Strongly
Disagree
Total
111
Appendices
Q3.HowoftendoyouwriteresearcharticlesinEnglishforpublication?
#
Answer
Response
%
4
36%
5
45%
2
18%
0
0%
0
0%
11
100%
Morethan21
3Timesa
Year
2
2-3Timesa
Year
3
OnceaYear
4
Lessthan
OnceaYear
5
Never
Total
Q4.HowoftendoyouwriteresearcharticlesinJapaneseforpublication?
#
Answer
Response
%
0
0%
1
9%
5
45%
5
45%
0
0%
11
100%
Morethan21
3Timesa
Year
2
2-3Timesa
Year
3
OnceaYear
4
Lessthan
OnceaYear
5
Never
Total
112
Appendices
Q5.WritingresearcharticlesinEnglishisdifficultforme.
#
Answer
Response
%
2
18%
9
82%
0
0%
0
0%
0
0%
11
100%
1
Strongly
Agree
2
Agree
Neither
3
Agreenor
Disagree
4
Disagree
5
Strongly
Disagree
Total
Q6.MyEnglishmanuscriptsneedproofreadingbeforesubmissiontojournals.
#
Answer
Response
%
8
73%
3
27%
0
0%
0
0%
0
0%
11
100%
1
Strongly
Agree
2
Agree
Neither
3
Agreenor
Disagree
4
Disagree
5
Strongly
Disagree
Total
113
Appendices
Q7.Whatisyourcurrentposition?
#
Answer
1
Professor
Response
%
0
0%
3
27%
8
73%
0
0%
0
0%
11
100%
2
Associate
Professor
3
Assistant
Profosser
4
PhD
Candidate
5
Other
Total
Q8.WhichsectionsarethemostdifficultforyoutowriteinEnglish?Youcanchoose
morethanoneifneeded.
#
Answer
1
Title
2
Abstract
3
Introduction
4
Method
5
Results
6
Discussion
7
Conclusion
8
References
Response
%
3
27%
0
0%
9
82%
1
9%
0
0%
5
45%
2
18%
0
0%
0
0%
0
0%
9
Supporting
information
10
Other
114
Appendices
Q9.HaveyoueverbeentoldbyajournaleditororreviewertohaveyourEnglish
manuscriptproofread?
#
Answer
1
Yes
2
No
Response
%
6
55%
5
45%
11
100%
Total
115
Appendices
AppendixD:1,000-wordExcerpt
Inorganicandmedicinalchemistry,OrganicorganicAzidesazides(R-N3)havebeenwellstudied
amongorganicchemistryandmedicinalchemistrysincePeterGrießsynthesizedthefirstorganic
azide,phenylazide.AfterthedevelopmentofCurtiusrearrangementusinghydrogenazide(HN3),
which produces isocyanates from the corresponding acyl azides, many organic azide-utilizing
reactions were produced. Despite their explosive and toxic properties, organic azides are
attractive not only industrially, but also agriculturally, and pharmaceutically. For these reasons,
organic azide chemistry has developed extensively. Interestingly, natural products possessing
azidegroupshavenotbeenisolatedtodate,whilethosehaving1,2,3-triazineand1,2,3-triazene
structureswerehavebeenfoundinnature.
Structurally, azides consist of three nitrogen atoms in linear form (not straight, but
slightly bent; a calculated angles of N1-N2-N3 is 172.7° and of R-N1-N2N3 is 115.2° in methyl
azide)(Schemex).Organicazidesshowdifferentchemicalreactivities:theN1atomcanworkasa
nucleophile, and the N3 position nitrogen atom shows electrophilic reactivity (Scheme x). The
specific efficiency of organic azides is its character as 1,3-dipolar and this provides [3+2]
cycloadditions with unsaturated bonds to give triazolines, triazoles and tetrazoles. Recently, the
cyclizationreactionsoforganicazideswithalkynes(Huisgenreaction)havebeenafocusedinthe
areaofchemicalbiologyareaandextensivereportshavebeenpublished(Meldal-Sharplessclick
reactionorCopper-CatalyzedAzide-AlkyneCycloaddition—CuAAC).
Organic azides consist with of amines (N1) and the excellent leaving group diazonium
cation (N2N3) producing nitrogen gas. Actually, the bond length of N1–N2 (1.237 Å) is
computationally longer than that of N2–N3 (1.156 Å) in methyl azide. Thus, azides can easily
evolve nitrogen gas in many reactions (Scheme x). Especially, heating conditions or
photoirradiations produce nitrenes from organic azides, which are highly reactive and give
aziridinationsandC-Haminations(eq.x).
Inthisreview,wedescribethemorerecentapplicationsoforganicazidesinthesynthesis
ofnaturalproducts.Thereactionstepsinvolvingtheuseofthisfunctionalgroupaswellasthea
descriptionofthemethodsofdecomposition/reductionarelistedaccordingtotheasfollowsing
order: (1) preparation of organic azides; (2) C-H insertion reaction by nitrenes; (3) Curtius
rearrangement;(4)Schmidtreaction;(5)[3+n]cycloadditionreaction;(6)Staudinger/aza-Wittig
reactionand(7)organicazidesasmaskedaminofunctionalgroup.
Before starting reviewing the more recent synthetic applications of organic azides, we
showtheirgeneralpreparationmethods.Generalproceduresaresummarizedinequations4–10
inSchemex.Toprepareorganicazides,useofvolatile,toxicandhighlyexplosivehydrogenazide
should strongly be completely avoided. SN2 azidation of alkyl halides and acyl halides using
nucleophilic azides, mostly sodium azide (NaN3), is the most general method (Scheme x). For
transformation of alcohols and carboxylic acids, Shioiri reagent (DPPA—diphenylphosphoryl
azide) is often used. To replace hydroxy groups, Mitsunobu reaction conditions (PPh3 and
azodicarboxylates DEAD/DIAD) can introduce azide groups with DPPA or zinc azide, and later
direct conversion without Mitsunobu conditions. Recently, Kitamura and co-workers
demonstrated that the safer and more stable azido compound ADMP (2-azido-1,3dimethylimidazolinium hexafluorophosphate) was a good azidation reagent. Azides can be
prepared from primary amines by way of diazoniumion formation followed by nucleophilic
azidations. Aryl azides are usually synthesized with these procedures. From primary amines,
diazotransfer reaction can also deliver organic azides. However, use of unstable and explosive
trifluoromethanesulfonyl azide (TfN3) was a problem. Recently, safer diazotransfer reagents,
ADMPandGoddart-Borgerreagent(1),werehaveappearedandcanbeusedinsteadofTfN3.
1,4-Additionofazidesareisproblematicbecauseaformal[3+2]reactioncouldoccurasa
side reaction. Miller et al. reported successful reaction conditions to perform preferential 1,4additionofazidestoα,β-unsaturatedcarbonylcompounds,whichwasrecentlydemonstratedina
modelstudyinthetotalsynthesisofcortistatinA4andJ5byYamashitaandHirama’sgroup.
1,4-Azidation was has also been reported in the total syntheses of marine bisindole
alkaloids hamacanthin A (11), B (12) and the antipodode of cis-dihydrohamacanthin B (13) by
116
Appendices
Kawasaki et al. (Scheme x). Treatment of 2-methoxyindoline derivative 6 with methanesulfonic
acid generated eniminium intermediate 7, and then 1,4-azidation was proceeded with TMSN3 to
give a desired azide compound 9 in 56% along with its epimer 8. The azide group in 9 was
reducedbytheStaudingerreactionfollowedbyprotectiontoafford10,whichwasconvertedto
naturalproducts11,12andantipode13.
Notonlynucleophilicazidation,butalsothedirectintroductionofazidesbyelectrophilic
azidations of carbanions have been reported, which are performed with sulfonyl azides, usually
TrisN3(2,4,6-trimethylbenzenesulfonylazide).Inthereactionwithenolates,theprotonsourcefor
thequenchingreactioniscriticalandEvansandco-workersrevealedthataceticacidwasthebest.
Kozmin et al. reported total syntheses and biological activities of streptolydigin 17 and its
analoguesusingthisstereoselectiveelectrophilicazidationreaction(Schemex).
Lovelyandco-workersreportedthetotalsynthesisofsomecomponentsofthekealiinine
family (Scheme x). In their synthesis, the 2-aminoimidazole moieties were successfully
constructed in the late stages of syntheses by way of lithiation-electrophilic azidation of the
correspondingimidazoles18–20.ThefollowinghydrogenolysisgavekealiininesA24,B25andC
26.HisgroupalsoachievedconcisetotalsynthesisofrelatednaamineG27andnaamidineH28
withthesamestrategy.
Veryrecently,Fukuyamaandco-workersreportedtotalsynthesisofpentacyclicalkaloid
lyconadinA34andB35(Schemex).Aftersomeinvestigationsinmodifiedsyntheses,theyfound
that nucleophilic azidation was the best way to construct the enone 33, which was essential for
pyridonecoresynthesis.Lithiationofvinylbromide29followedbytrappingwithTris-N3–acetic
acid quenching gave vinyl azide 30. Transformation of 2-methylvinyl azide moiety of 30 into
enone was performed by heating in the presence of acid. The protonation of the azide triggered
nitrogen evolution (Schmidt reaction) to give unsaturated imine 32. The following hydrolysis
affordeddesiredenone33.
In 1992, Magnus et al. discovered allylic azidation of triisopropylsilyl enol ethers with
TMSN3 in the presence of hypervalent iodines. Recently, White et al. utilized this reaction to
introduceaminogroupsinthetotalsynthesisofhuperzineA40(Schemex).
FigureD1,000-wordexcerptillustratingerrorrecognitionandreconstruction
Note.ExcerpttakenfromIntroductionsectionof[04ECS].
117
Appendices
AppendixE:NominalGroupDenotationandQuantification
[Met-hemecoordination]contributesto[thestabilityofthestructure]and[theabilityof
electrontransferincytcfamilyproteins].Although[theopticalabsorptionspectra]and
[redoxpotentials]weresimilarbetween[monomericanddimericWTPAcytc551],
[heme-ligatingHisandMet]originatedfrom[differentprotomersinthedimer],similar
to[thecaseofdimericHTcytc55].In[thecaseofdimerichorsecytc],[Met-heme
coordination]wasperturbedand[ahydroxideion]wascoordinatedto[thehemeiron].
[ThedifferenceinthehemecoordinationstructurebetweendimericPAcytc551and
dimerichorsecytc]maybedueto[thedifferencesinthestabilityoftheMet-heme
coordinationbond]and[therigidityoftheloopcontainingtheheme-ligatingMet].
Accordingto[DSCmeasurements],[ΔHforthedissociationofdimerichorsecytcto
monomers]exhibited[alarge,negativevalue],whereas[theΔHvaluesforthe
dissociationofdimericPAcytc551anddimericHTcytc552]were[~0]and[+14kcal/mol],
respectively.[Theseresults]showthat[thecoordinationofMettotheheme]contributes
to[stabilizationofthedimer]enthalpically.
Note.ThisexcerptisoneparagraphtakenfromtheDiscussionsectionof[05EST].Nominal
groupsaredenotedinsquarebracketshighlightedinredwithPostmodifiersinitalics.Total
words:182;Wordsinnominalgroups:148(81%);Totalnominalgroups:25(⇒137nominal
groupsper1,000words).
118
Appendices
TableEApproximatenumberofnominalgroupsper1,000wordsandpertextinthe
CorpusofJapaneseScientificDiscourse(COJSD)
Approximate
Approximate
nominalgroups nominalgroups
Text
per1,000words
pertext
01ECS
157
562
02ECS
118
627
03ECS
160
495
04ECS
183
911
05ECS
154
731
06ECS
158
300
07ECS
150
380
08ECS
161
858
09ECS
192
831
10ECS
151
384
11ECS
161
278
12ECS
173
747
13ECS
176
776
2,094
7,879
161
606
Total
Average
Note.Theapproximatenumberofnominalgroupsper1,000words(i.e.,secondcolumn)isbased
onthemanualanalysisof1,000wordssamplesfromtheDiscussionandConclusionsectionsof
eachtextintheCOJSD.Theapproximatenumberofnominalgroupspertext(i.e.,thirdcolumn)
isanextrapolationfromthe1,000wordsamples.
119
Appendices
AppendixF:ThesisWordCount
TableFWordsperchapter
Words
Introduction
954
LiteratureReview
5,600
Method
3,560
Findings
9,869
Conclusion
870
Total
20,853
Note.Wordcountsincludetablesandfigures.Footnotesarenotincluded.
120