名詞的/動詞的範疇一英語、日本語の動名詞構文からの考察

(1)

名詞的/動詞的範疇

一英語、日本語の動名詞構文からの考察

佐藤裕美

英語のジェランド(動名詞)や日本語の動名詞構文において共起する

動詞的性質と名詞的性質は、述語が統合する機能素性の特性に起因するこ

とを主張し、述語が範疇素性の指定を欠く場合はその範疇が機能素性によ

って決定される可能性を指摘する。英語の動名詞構文はその内部構造にお

いては動詞的な性質を呈するが、文中においては名詞句と同様の分布を示

す。本稿ではジェランドは非時制節(CP)であり、そのテンス素性により

前置詞が補文標識Cとして生ずることが可能であることを提案する。前

置詞+ジェランドはP+DPではなく、前置詞がその主要部となるCPと

分析することにより、ジェランドの内部構造と分布の双方に説明が与えら

れる。また、日本語の動名詞構文は形態的には名詞であるが動詞と同様に

節を形成し、さらにその節が格助詞を伴うなどの相反する範疇的な特徴を

表すが、これらに対しても、ジェランドと同様、名詞句、あるいは動詞

句/節としての特徴が、統合される機能素性により決定される分析を提示

する。さらに、名詞、動詞などの範疇ごとに関連する機能素性に区別が存

在するのではなく、異なる範躊間で共通する素性の存在を示唆する。

(2)

NominalandVerbalCategories

AStudyBasedonEnglishGerunds andJapaneseVerbalNouciClauses

HiromiSato

0.In血oducdon

Englishgerundsposelong‑standingproblemsofmixedcategorial

properties.Theirinternalstructureindicatesthattheybelongtothe verbalcategory,whiletheirexternaldistributionsuggeststheirnominal nature.ThisarticleproposesthattheV‑ingformiscategorially

underspecifiedandthepropertiesoftheprojectedstructurearedueto thoseoftheassociatedfunctionalfeatures.Themixedpropertiesof gerundswillthenbeattributedtothepropertiesoffunctionalfeatures characteristicoftheseconstructions.Otherstructureswithcategorially

ambiguouspropertiesarealsoattested.Theso‑called"verbalnouns"in Japaneseformastructurethatshowsbothnominalandclausal

properties.Clausesformedbyverbalnounsshownoverbal

morphologyandtheycanbefollowedbyacase‑particle.Inthisarticle thepropertiesoftheseclauseswillbasicallybeprovidedwiththesame accountastheanalysisproposedfarEnglishgerunds.

Insection1,IbeginbyreviewingtheprincipalpropertiesofEnglish

gerunds.ItwillbearguedthatgerundsareCPsinvolvingacertainT

featureandtheapparentDP‑likedistributionofgerundswillbe

(3)

reinterpretedasthedistributionofaCP.Section2addressestheissues concerningtheclausesformedbyJapaneseverbalnouns.Inthe

proposedanalysisthecase‑particlethatimmediatelyfollowsthese clausesisduetoafeatureofT,whichparallelsthefeatureproposedfor Englishgerundsinsection1.Section3statesthesummaryandthe conclusionofthediscussion.

1.EnglishGerundsasCPs

1.1TheNominalandVerbalPropertiesofGerunds

Gerundscanbeclassifiedintothreesubclassesbythemorphological formofthesubject:(a)thosewiththesubjectinpossessive‑case

(PASS‑ing},(b}thosewiththesubjectinaccusative‑case(ACC‑ing}, and{c)thosewithnoovertsubject{PRO‑ing).

(1)a.WerememberedJerry'sattendingtheclass.

b.Werememberedhimattendingtheclass.

c.WeirememberedPROSattendingtheclass.

Despitethedifferencesmanifestedinthesubjectforms,allthese threetypesofgerundssharepropertiesthatdistinguishthemfrom deverbalnounswiththe‑ingsuffix.Thoughtheexternaldistributionof

gerundsappearstoparallelthatofDPs,incontrasttodeverbal‑iszg nounstheirinternalstructureindicatesthattheyareclauses.

Iwillfirstreviewsomesalientpropertiesofgerunds,‑oftentaking POSS‑∫ ηggerundsasexamples.Nosystematicattemptwillbemadein thisarticletoderivethedifferenttypesofgerundsmentionedabove,

(4)

142

butabriefdiscussiononapossibleexplanationforthedifferenceswill bepresentedinsection1.2.

InVleWoftheobjectivecaseproperty,gerundsbelongtotheverbal categoryratherthantothenominalrenderingtheobjectDPinthe accusativeform.Thus,asshownin(2),gerundscontrastwitha deverbal‑ingnounwhoseobjectisrenderedintheof‑phrase.

(2)a.WerememberBill'sdrawingthepicture.

b.Werememberdrawingofthepicture.

Anotherdifferencebetweengerundsanddeverbal‑ingnounsisthat

theformerallowsadverbialbutnotadjectivialmodification,whereas thelattercanbemodifiedonlybyadjectives.

{3} a.WearecountingonJohn'scarefully/*carefulassemblingthe pars.

b.WearecountingonJohn's*carefully/carefulassemblingof theparts.

Thepossibilityoftheadverbialmodificationalsosuggeststhat

gerundiveconstituentsareverbalratherthannominalprojections.

ontheotherhand,theexternaldistributionofgerundsparallelsthat ofDPs,suggestingthenominalstatusofgerunds.Theexamplesin

(4)‐(5}belowindicatethatgerunds,likeDPsandunlike

finiteiinfinitivalCPs,cannotappearinthecomplementpositionof NorA.1

(5)

(4)

(5)

a.*John'spreferenceMeg'sstudyingFrench】

b.*John'sproof[thetheorem c.John'sclaim[thatMaryisguilty]

d.John'sattempt[PROtoprovehisinnocence]

a.*Weareaware[John'spassingthebarexam].

b.*Weareaware[John'seagerness].

c.Weareaware[thatJohnisgoingtotakethebarexam].

d.Johnisafraid[PROtotakethebarexam.

AnotherdistributionalcharacteristicofgerundsparallelingthatofDPs

isthatunlikeclauses,theycanbetheobjectofP.

{6) a.Johncountedon[Melt'swinningtheprizemoney].

b.Johncountedon[Meg'swinningoftheprizemoney].

c.*Johncountedon[thathewouldwintheprizemoney].

d.*Johncountedon[PROtowintheprizemoney].

InhisanalysistoaccountforthesemixedpropertiesofPOSS‑'ηg

gerunds,Abney(1987)proposesthatthesegerundsareDPsinwhich thenominalhead‑ingtakesaVPcomplementasillustratedin(7).

(7)

八

John'sD'

DへP

/¥

‑ingVP VDP

̲̲1

singtheMarseillaise

(6)

144

Abneyassumesthatthenominalizer‑ingtakesaVPandconvertsit intoanNP.Inhisproposalthenominalizer‑ingingerundsandtheone foundindeverbal‑ingnounsarethesameintermsoftheirbasic

properties,butthelatterdiffersfromtheformerinthatitattachestoV

ratherthantoV'sprojectionasshownin(8).

fig} ^DP

/¥

John'sD'_DNP

〈

NPP/¥

‑ingVoftheMarseillaise /¥D sing .l

Thestructurein(7)suggestsacategoryswitchingaffixationatthe

phrasallevel.Apartfromquestionsconcerningthisderivation,Abney's structureisproblematicbecausebothPOSS‑cng・anddeverba1‑ITIg nounsformDPs.oneproblemisthatalthoughbothP4SS‑inggerunds

anddeverbal‑'η8nominalsallowthesubjectingenitive‑case ,typical

determinerslikethe,a,andsomeareallowedonlyindeverbal‑ing nouns.

C9) (10)

[The/Mary'sreportingoftheevent]waspraised.

IrememberJahn's/*thesingingthesong.

Irememberthesingingofthesong.

Anotherproblemconcernsthenumberfeatureofnominals .As

(7)

observedintheexamplesinWasowandRoeper(1972),deverbal‑ing nounscanbepluralized,whilegerundsdonottoleratepluralization.

{11)a.SightingsofUFO'smakeMarynervous.

b.SightingUFO'smakesMarynervous.

c.*Sightings[.JFO'smakeMarynervous.

StillanotherproblemisthatsinceinAbney'sstructure‑ingattaches toVPandturnsitintoanominalcategory,noclausalpropertiesare

assumed.However,gerundsexhibitpropertiesthatsuggestthepresence ofT,henceindicatingtheclausalstatusofgerunds.Gueronand

Hoekstra(1995)observesthatanaspectualauxiliary,have,canappear ingerundsbutnotindeverbal‑ingnominals,suggestingthatTis

presentintheformerbutnotinthelatter.

(12}a.John'shavingpassedthisexamsurprisedus.

b.*John'shavingpassedofthisexamsurprisedus.

(13)a.WearesatisfiedwithJohnhavingfinishedthejobintime.

b.*WearesatisfiedwithJohn'shavingfinishedofthejobin

time.

GueronandHoekstra{1995}alsonotesthatwhiledeverbal‑ingnouns takenoastheirnegation,gerundstakethesententialnegationnot.

(14}a.John'snotreadingthebookwillnotsurpriseme.

b.Noreadingofthebookwillsurpriseme.

(8)

146

Thecontrastobservedin{14)immediatelyfollowsfromtheassumption thatnatneedstobelicensedbyT{cf.Zanuttini(1991}).

Furtherevidenceindicatingtheclausalstatusofgerundsisthatthe expletivethereand̀weather'it,whosedistributionsareisrestrictedto

thesubjectposition,canappearingerundsasdiscussedinReuland

(1983).

(15)a.Youmaycounton[therebeingaiotoftroubletonight].

b.Iwouldn'tcounton[itrainingtomorrow].

Thepresenceofthesenonargumentsstronglysuggeststhatgerundsare

clausalconstituentswhosesubjectpositionmustbefilled.Similarly,the followingexamplesshowthat̀raisingtosubject'ispossibleingerunds whileitisnotinDFs,alsoindicatingtheobligatorysubjectpositionin

gerunds.

(16)a.[John'sapPearingttobefriendly】easedourtension.

b.*[John'sappearancettobefriendly]easedourtension.

Theseobservationsleadtotheconclusionthatasthestructureof (PASS‑ing)gerunds,(7)isnottenable.Instead,Itentativelypropose thatgerundshaveaclausalstructureasexemplifiedin(17).Compare thisstructurewiththenominalstructureofadever17a1‑in.gnoungiven inX18).

(17) [TpJohn's垂T【vp'iv[vPreportingthenews]]

(9)

(18) [DPthe/John's[NPreportingofthenews]]

IassumethattheV‑ingformsinthestructuresin{17)and(1S)are

categoriallyunderspecified.IfaVingmergeswithnominalfunctional

features,itprojectstothenominalcategoryandisvaluedwithnominal featureslikenumber.Ontheotherhand,ifaVingmergeswithverbal

functionalfeatures,itprojectstotheverbalcategoryandisvaluedwith

verbalinflectionalfeatures.

IfgerundsareclauseswithT‑features,theirstructureshould eventuallybeextendedtoCP.Then,aquestionarisesconcerningtheir externaldistribution.Asseenabovetheexternaldistributionofgerunds appearstoparallelthatofDPsratherthanthatoftypicalCPs.In

particular,inthecomplementpositionofN,AandP,gerundsand typicalCPsshowcontrastivedistributionasobservedin(4)‐(6}above repeatedhereas{19)‐{21).

(19)a.*John'spreference[Meg'sstudyingFrench]

b.*John'sproofthetheorem]

c.John'sclaim[thatMaryisguilty]

d.John'sattempt[PROtoprovehisinnocence]

{20)a.*Weareaware[John'spassingthebarexam].

b.*Weareaware[John'seagerness.

c.VVeareaware[thatJohnisgoingtotakethebarexam].

d.Jahnisafraid[PROtotakethebarexam.

(21)a.Johncountedon[Meg'swinningtheprizemoney].

b.Johncountedon[Meg'swinningoftheprizemoney].

(10)

t48

c.*Johncountedon[thathewouldwintheprizemoney】.

d.*Johncountedon[PR4towintheprizemoney].

Theexamplesin(21}indicatethatlikeDPsandunlikeCPs,gerunds canappearafterapreposition.Also,theungrammaticalexamplesin (19)and(20)becomegrammaticalifanappropriateprepositionis

insertedbetweentheNIAandtheirargumentDPorgerund .Tothe extentthattheclausalanalysisofgerundsiscorrect,thissuggeststhat insentenceslike{21.a)aprepositioncantakeagerundiveclauseasits complement.onthebasisofthisobservation,Iproposethata

prepositionanditscomplementgerundformaCPratherthanaPP,and

theprepositionprecedingthegerundisaprepositionalcomplementizer.

Then,thestructuresinvolvingthegerundivecomplementofP,N,andA canbespecifiedasin(22).

(22)a・We【vPcounted[cP[con][TPJohn'sfindingoutthetruthj]]

(cf.21.a)

b.【DpJohn'spreference[cp[cof][TpMeg'sstudyingFrenchm

c.Weare[Apaware[cp[cof][TpJohn'spassingthebarexam】】】

Noncethatthestructuresofgerundsgivenin{22)areanaloguesof thestructureoffor‑tainfinitivesinthatCisaprepositional complementizertakinganon‑finiteTP,asshowninX23}.

(23} We[vPPrefer[cP【cfor】[TPJohntofindoutthetruth】]]

(11)

ByanalyzingtheprepositionsprecedinggerundstobetheheadofCP, thedistributionofgerundsisshowntoexactlyparallelthedistribution oftypicalCPs.

1.2SupportforaPrepositionalComplementizerin Gerunds

Anotherproblemthatstillremainstobeconsiderediswhetherthe structureproposedabovecanbemotivatedindependentlyofgerunds.

Inthefollowingexamples(takenfromtheBritishNationalCorpus), theprepositionsdespiteandwithtakeagerundivecomplement.

{24)a.Thisexpeditionbeganthismorningalmostanhourlaterthan Ihadplanned,[despite[myhavingcompletedmypacking wellbeforeeighto'clock]].

b.[Despite[hishavinggivenheratoughtimeonset]],Lynette isstillveryfondofWoodyAllen.

c,[With[Willishavingbeeninjuredinthesemi‑final]], Englandhadchosentoplayanextrabatsman,Larkins.

In(24.a‑‑c}thegerundscontainperfectivehave,suggestingthatthe gerundiveconstituentsareclauseswithT.Ifdespiteandwithare

nothingbutPintheseexamples,therelevantstructurewouldbe

problematicasclausalconstituentsareexcludedfromthecomplement positionofP.However,byanalyzingdespiteandwithas(prepositional}

compiementizersheadingadjunctCPs,thegeneralizationconcerning

thestructureofPPstillholds.

(12)

Ithasbeennotedalsothatprepositionssharecertainpropertieswith theclauseintroducingthatandfor.Kayne(1984}observescontrastsin theextractionoutofagerundivecomplementofP,asillustratedinthe followingexamples.

(25}a.MaryistheonewhoI'mcountingon[himmarrying̲].

b.*JohnistheonewhoI'mcountingon[ ̲marryingher].

Theexamplesin(25)clearlyindicatethatwhenagerundistheobject ofaprepositiontheextractionoftheobjectispossiblewhilethatofthe subjectisnot.Theobservedcontrastisanalogoustothecontrasts

knownas"that‑traceeffect"and"for‑traceeffect"observedinthe followingexamples.(cf.PesetskyandTorrego(2004))2

(26)a.Whodoyouthink[̲marriedher]?

b.*Whodoyouthink[thaLmarriedher】?

(27)a,Whodoyouwant[̲tomarryher]?

b.*Whodoyouwant[for ̲tomarryher]?

Ifthecontrastobservedin(25)hasthesamesourceas"that‑trace effect"and"for‑traceeffect"assurmisedinKayne{1984)andPesetsky andTorrego(2004},thisprovidessupportforourclaimthata

prepositiontakingagerundivecomplementisacomplernetizer.

AssumingthatC‑featurescorrelatewithTfeatures,theappearanceof prepositionalcomplementizersshouldbeattributedtotheTfeatures specifictogerunds.AprototypicalpropertyofPisthatittakesa

(13)

nominalcomplement.Then,ifthenonfiniteTingerundshasanominal property,thiscanbetherationalefortheoccurrenceofaprepositional complementizer.

Theassociationofnominalpropertieswithnonfiniteclausesisnotan ad‑hocsolutionspecificallyforEnglishgerunds.Theassumptionof nominalpropertiesismotivatedinotherLanguagesaswell.Raposo {1987)claimsthatRomanceinfinitivesarenominal.Comparethe followingFrenchexamplesinwhichtheinfinitivesarethe complementsofV,AandN.

(28}a.Jeandesirechanter.

Johndesiressing‑inf.

b.Jeanestdesireux*(de)chanter.

Johnisdesirousdesing‑inf.

c.ledesir*{de)chanter thedesiredesing‑inf.

RaposoarguesthatinFrenchtheappearanceoftheprepositiondewith infinitivesinthecomplementpositionofAandNisrelatedtocase.If thecontrastsobservedintheaboveexamplesareduetoavailabilityof casefortheinfinitivecomplements,itnaturallyfollowsthatthese infinitivesarenominalinnature.

FurthersupportfortheclaimthatRomanceinfinitivesarenominal canbefoundinthefactthatinseveralRomancelanguagesinfinitives canbeprecededbyadefinitearticle.Theexamplebelowistakenfrom Rizzi(1982).

(14)

152

{29} 1'esserlasituazionesuscettibledemiglioramento thebe‑ingthesituationcapableofimprovement

ThesenominalpropertiesofRomanceinfinitivescannaturallybe attributedtoinfinitivalTwithanominalfeature.IfEnglishgerundsare associatedwithanominalfeaturethatcanbemergedwitha

prepositionalcomplementizer,anominalfeaturemustbepresentin gerundiveT.Thisnominalfeatureshouldbetheonethatlicensesthe inflectionalmorphologyonV.Thisfeaturewillhenceforthbereferred towiththenotation[+NTH.

SofarIhavenotbeenconcernedwiththedifferencesamongthe threetypesofgerunds,i.e.,POSS‑ing,ACC‑ingandPRO‑ing.

Developingasubstantialanalysistoaccountforthesedifferencesin detailisbeyondthescopeofthisarticle,soIwillsimplysketcha

possiblescenario.lntheproposedanalysis,Tofgerundsisabundleof featuresconsistingatleastof卜finite,+NT】.Supposethatthecaseof thegenitivesubjectisthemorphologicalreflexoftheagreementwith

【+NT】aloneorwithbothofthesefeaturescompositionally.Inthe

formercase,POSS・inggerundsw川result.When[‑finite ,+NTl

complexisinvolved,thesubjectwillbevaluedwith[accusative]ifitis overt.Ifthesubjectisnon‑overt,itisvaluedwith[null.Thus,ACC‑

ingandPR4‑inggerundswouldbederived.3

Wehavethusfarestablishedthestructureofgerundsasillustrated in(30).

{30) WecountedCcP[coneCTPJohn's[T‐finite+NT]takingcare

(15)

ofourchildren]].

Itremainsunexplained,however,whyaprepositional

complementizerdoesnotappearinallgerundiveconstructions.

Prepositionalcomplementizersareapparentlypresentonlywhen gerundsareinthecomplementpositionofN,A(asin(31})andV subcategorized.forPP(asin(30}above),andwhentheyappearin.

adjunctsasin(32).

(31)a.John'sinterests[inMary'sexpandingherbusiness b.Johniscertain[ofMary'sknowingtheanswer].

(32)Marylookedhappy[despiteJohn'shavinglosthisjob].

Inthefollowingexamples,ontheotherhand,aprepositional complementizerisnotovertlypresent.

(33}a.WeresentJohn'sdeceivingourchildren].

b.[Jill'swinkingatTom]angeredAlice.

Aplausiblesuggestionisthatthesesentencescontainanon‑overt complementizer.Thesuggestedstructureisakintothatofcontrol

infinitiveswithouttheovertcomplementizerfor.Thesuggestionis

well‑foundedinviewofthefactthatgerundscanappearinthesubject

positionoffiniteclauses.Thus,withanullcomplementizeramore

explicitstructureofthesentencesin(33)willbeshownasin(34}:

(16)

154

(34)a.Weresent【cpCnull[㎜John'sdeceivingourchildren】].

b.【CPCnuil[TPJi11'swinkingatTom】1angeredAlice.

Thecomplementizeringerundsmustbeinconcordwiththefeature

l+NT】.WhenagerundisthecomplementofN,A,orVwhichselects thecategoryPoralexicallyspecificP,aprepositionoccursasanovert Cinthegerund.Whentheselectionisnotlexicallyspecified,the

prepositionofappearsasdefaultasin(31.b}.Insentences(34)and (31.a)theprepositionsonandin,whichareduetotheselectionbythe higherVandN,becomethecomplementizerintherespective

sentences.Asforovertprepositionalcomplementizersinadjunct positionsasin(32),theirpresenceissanctionedbytheirindependent semanticspecifications.

Tosummarizethediscussionthusfar,Ihavearguedthatgerundsare CPswithaprepositionalcomplementizer.TheV‑ingformis

underspecifiedintermsofitscategory,andcanprojecteitheranominal

oraverbalcategorydependingonthefunctionalfeaturesitmerges

with.ThenonfiniteTingerundshasanominalpropertyintheformof

thefeature[+NT1,whichisresponsibleforlicensingthe‑IYIginflection andaprepositionasacomplementizer.Thecomplementizerinagerund canbenon‑overt,butwhenrequiredbytheselectionalpropertyofthe higherheadorbysemanticfeatures,anovertprepositional

complementizermustbepresent.

Thenextsectionaddressesquestionsconcerningtheso‑called

"verbalnouns"inJapanese .Asinsinuatedbytheterminology, constructionsinvolvingthesewordsposeproblemsofmixedcategorial

(17)

properties.Itwillbeshownthatproblemsassociatedwiththe"verbal nouns"canbeaccountedforonthebasisofcategoryunderspecification andtheassociatedfunctionalfeatures.

2.NominalandClausalPropertiesofVerbalNounClauses inJapanese

2.1VerbalNounsHeadingVerbalProjections

Japanesehasaclassoflexicalitemsoftencalled"verbalnouns."

ManyverbalnounsderivedfromSino‑Japanesecompoundsoriginated fromChineseverbs.4Examplesaregivenin(35}.

(35)a.benkyoo b.ryokoo c.kenkyuu d.soodan

"studying"

"traveling"

"research"

"consultation"

ThoughmanyJapaneseverbalnounshavetheiroriginsin correspondingChineseverbs,theyhavedistinctnominal

characteristics.Asshownin(36.a)and(36.b}verbalnounscanco‑

occurwithdemonstrativessuchaskono"this"andsono"that."

Furthermore,asshownin(36.b)(36.c}thegenitivecasemarker‑no attachestotheargumentsofaverbalnounindicatingthatthehead responsibleforthecase‑markingisnominal.

(36)a.Anotorihiki‑wamondai‑gaaru.

thatdeal‑topproblem‑nombe

(18)

156

"Thereisaprobleminthatdeal

."

b.[Ano[syusyoo‑nohatugen]]‑gamondai‑ninat‑ta . thatpremier‑genremark‑nomproblem‑datbecome‑past

"Thatremarkbythepre

mierbecameaproblem."

c.[Jun‑noigirisu‑e‑noryuugaku]‑waasobi‑gamokuteki

Jun‑genEngland‑to‑genstudy‑abroad‑topplay‑nompurpose dat‑ta.

be‑past

"Jun'sstudyinginE

nglandhasitspurposeinhavingfun."

Whenverbalnounsarefollowedbytheso‑calledlightverbsuru{or

shita,shiteiru,thepasttenseandtheprogressiveform,respectively),a verbalnoun‑‑suru(VN‑suru}compoundfunctionsasV.InVN‑suru

compoundsthelightverbisdevoidofsemanticcontentwithnotheta‑

rolesofitsown.Itfunctionsonlyasthecarrieroftheverbal morphology,similartoEnglishpleonasticdo.

(37)a.Jun‑gabenkyoo‑suru/sita.

Jun‑nomstudy‑do/did

"Junwillstudy/studied"

b.Jun‑gaAkiko‑todeeto‑suru/sita.

Jun‑nomAkiko‑withdate‑do/did

"Junwillgoo

ut/wentoutwithAkiko"

c.tun‑gaAkiko‑totorihiki‑suru/sita.

Jun‑nomAkiko‑withdeal‑do/did

"Junhas/hadadeal

withAkiko"

(19)

Whatisinterestingaboutverbalnounsinrelationtothediscussionin theprecedingsectionisthatincertaincontextsverbalnounscan

functionasVwithoutlightverbsuruandtheyprojectaclausal

structure.

AsShibananiandKageyama{1988)discussesverbalnounscanbe followedbyexpressionswithtemporalmeaningssuchas‑noato

"after"

,‑fzomae"before",‑nosai"instance"and‑noon"time."Inthe followingexamplesverbalnounsshouldbeanalyzedastheheadofNP astheirargumentsaremarkedbygenitivecase‑particle‑‑no,a characteristicofthenominalcategory.

(38)a.[Bengosトnoyougisya‑to‑nomendan]‑noato,

[attorney‑gensuspect‑with‑geninterviews‑noafter, torisirabe‑gahajimat‑ta

interrogation‑nomstart‑past

"Aftertheattorneymetwiththesuspect

,theinterrogation started"

b.{Ybtoo‑nokokkai‑e‑nohouan‑noteisyutu]‑no

[rulingpartygentheDiet‑to‑genbill‑gensubmission]‑no mae,kanryoutati‑watetuyadesouan‑okaitei‑ta

beforebureaucrats‑topall‑nightdraft‑accwriting‑past

"BeforetherulingpartysubmittedthebilltotheDietsession , thebureaucratsstayedupailnighttowritethedraft."

c.[Bushdaitooryoo‑nohakone‑e‑notootyaku]‑noori,Koizumi [Bushpresident‑genhakone‑to‑genarrivals‑notimeKoizumi syusyoo‑gamizukarademukae‑ta

(20)

158

primeminister‑nomhimselfgo‑meet‑past

"WhenPresidentBusharrivedinHakone

,PrimeMinister Koizumihimselfwentouttogreethim."

Inadditiontotakingthegenitive‑casemarking,verbalnounscan demonstrateanarrayofothercaseparticlesthatarenormallyobserved

withverbalandadjectivalpredicates.

(39)a.[Bengosi‑gayougisya‑tomendan]‑noato,

[attorney‑nomsuspect‑withinterview]‑noafter, torisirabe‑gahajimat‑ta

interrogation‑nomstart‑past

"Aftertheattorneymet

withthesuspect,theinterrogation started"

b.〔Ybtoo‑gakokkai。nihouan‑oteisyutu】‑nomae,

[rulingparty‑nomtheDiet‑tobill‑accsubmission]‑nobefore kanryoutati‑watetuyadesouan‑okaitei‑ta.

bureaucrats‑topall‑nightdraft‑accwriting‑past

"Beforetherulingpartysubmittedthebillt

otheDietsession, thebureaucratsstayedupallnighttowritethedraft."

c.[Bushdaitooryoo‑gahakone‑nitootyaku]‑noori,Koizumi

[Bushpresident‑nomhakone‑toarrival]‑notimeKoizumi syusyoo‑gamizukarademukae‑ta.

primeminister‑nomhimselfgo‑meet‑past

"WhenPresidentBusharri

vedinHakone,PrimeMinister Koizumihimselfwentouttogreethim.

(21)

Ifthedifferencesinthecase‑particlesindicatethatverbainouns

projecteitheranounphraseoraclause,thedifferencesshouldalsobe

reflectedinadjectival/adverbialmodificationofverbalnouns.In(40.a) and(40.c)below,theverbalnounteisyutùsubmission'ISmodifiedby

anadverb,andbyanadjectivein(40.b)and(40.d).

(40)a.【Ybtoo‑gahouan‑okokkai‑niawatadasikuteisyutu]

rulingparty‑nombill‑acctheDiet‑tohurriedlysubmission]

nomae,kanryoutati‑watetuyadesouan‑okaitei‑ta.

‑nobeforebureaucrats‑topail‑nightdraft‑accwriting‑past

"BeforetherulingpartyhurriedlysubmittedthebilltotheDiet

session,thebureaucratsstayedupallnighttowritethedraft."

b.*[Yotoo‑gahouan‑okokkai‑niawatadasiiteisyutu]

[rulingparty‑nombill‑acctheDiet‑tohurriedsubmission]

nomae,kanryoutati‑watetuyadesouan‑okaitei‑ta.

‑nobeforebureaucrats‑topall‑nightdraft‑accwriting‑past

c.*[Yotoo‑nohouan‑nokokkai‑e‑noawatadasiku

[rulingparty‑genbill‑gentheDiet‑to‑‑genhurriedly teisyutu]nomae,kanryoutati‑watetuyadesouan‑okaitei‑ta.

submission]=nobeforebureaucrats‑topali‑nightdraft‑acc writing‑past

d.[Yotoo‑nohouan‑nokokkai‑e‑noawatadasii

rulingparty‑gentheDietWto‑genbill‑genhurried

teisyutu]nomae,kanryoutati‑watetuyadesouan‑okaitei‑ta.

submission]‑nobeforebureaucrats‑topall‑nightdraft‑acc wrltlng‑past

(22)

Aspredictedwhentheargumentsofaverbalnounaremarkedbythe case‑particle‑no,whichgenerallyappearinnominalphrase,only adjectivalmodificationispossible,andwhentheseargumentsare markedbycase‑particlesassociatedwithclauses,onlyadverbial modificationisacceptable.

ApartfromtheobservationsmadebyShibataniandKageyama

(1988}andTsujimura(1996}thatverbalnounswithoutthesightverb canprojectclausesintemporaladjuncts,Ipointoutthatverbalnoun clauses(VN‑clauses)arenotrestrictedtoappearintemporaladjuncts.

Intheexamplesin{41)VN‑clausesappearinadjunctsexpressing

reasonancircumstance.

(41}a.[Bushdaitouryoo‑gaKyoto‑nisyukuhaku]notame, [Bushpresident‑nomKyoto‑tostays‑noreason, genjyunakeibitaisei‑gasik‑are‑ta

strictguardsystem‑nomlay‑pass‑past

̀̀SincePresidentBush

wasgoingtostay/wasstayingin Kyoto,astrictsecuritysystemwasorganized."

b.[Jun‑gaLondon‑nisyuttyou]‑notugo‑de

Jun‑nomLondon‑tobusinesstrips‑nocircumstance‑by kaigi‑gatoriyameninat‑ta.

meeting‑nomcancelledbecome‑past

"Underthecircumstance

sinwhichJunisgoingtobe/is/was inLondononbusiness,themeetingwascancelled."

c.[Ryokoudairiten‑gahoteru‑tosoodan]‑noue,betu‑no [travelagency‑nomhotel‑withdiscussing]‑notop,other

(23)

eya‑gatoals‑are‑ta.

room‑genarrange‑pass‑past

"Astheresultofthediscussionbetweenthetravelagency andthehotel,otherroomswereprepared."

Theexamplesin(42)showthatVN‑clausescanfunctionas pronominalmodifiers.s

(42)a.[[Sanbansen‑nitootyaku]noressya】‑wal7:30hatu

[[No.3track‑atarriving]gentrain]‑top17:30departing Tokyo‑yukide‑su.

Tokyo‑boundcopular‑honorific.

"Thetrainthatisarriving/arrivedatTrackNo

.3isboundfor Tokyo,departingat17:30."

b.[[Taro‑gaitaria‑desyukuhaku]nohoteru]‑waizen‑wa

[Taro‑nomItaly‑instaying]genhotel‑toppast‑top kizoku‑nosiroat‑ta.

noble‑gencastlecopular‑past

"ThehotelinwhichTaroisgoingtostay/isstaying/stayedin

Italywasonceanobleman'scastle."

Intheexamplesin(43)theVN‑5配r配clausescorrespondingtotheVN‑

clausesin{42}formrelativeclauses.

{43)a.[[Sanbansen‑nitootyaku‑sita]ressya]‑wa17:30hatuTokyo‑

[[No.3track‑atarrive‑‑did]train]‑top17:30departingTokyo一

(24)

X62

yukide‑su.

boundcopular‑honoriflc.

"ThetrainthatarrivedatTrackNo

.3isboundforTokyo, departingat17:30.

b.1[Taro‑gaitaria‑desyukuhaku‑suru】hoteru]‑waizen‑wa

[Taro‑nomItaly‑instay‑do]hotel‑toppast‑top kizoku‑nosirodat‑ta.

noble‑gencastlecopular‑past

"ThehotelinwhichTaroisgoingtosta

yinItalywasoncea nobleman'scastle."

AssumingthattheVN‑clausesin(42)haveastructurethatparallelsthe structureoftheVN‑sururelativeclausesin(43),thestructureofthe bracketednounphrasesin(42)willbeillustratedasin(44).

(44) lNplcpOpi[Tp.'Dpi̲VN】 ηoNi]

Thewordsatòafter','"α8̀before',ort̀occasionltime',SQl

̀instance'

,tamèreason',uètop'arenounsastheyrevealproperties characteristicofN:(i)theycanbemodifiedbydemonstrativessuchas konòthis'andanòthat',(ii)theycanfollowanounwiththegenitive‑

caseparticle‑no,(iii)theycanbefollowedbyapostpositionsuchasde andni.ThesecharacteristicsareshownintheFollowingexamples.

(45)a.Konoato‑de,Jun‑gakuru‑hazu‑da.

thisafter‑inJun‑nomcome‑likely‑pres

(25)

"Afterthis

,Junislikelytocome."

b.Anoon‑ni‑wasyusyoo‑mohinans‑are‑ta.

thatoccasion‑in‑toppremier‑toocriticize‑pass‑past

"lnthatoccasion

,thepremierwascriticizedasweil."

{46)a.Syokuji‑nomae‑nioinorisuru.

Meal‑genbefore‑inpray

"(We)praybeforemeals ."

b.Jilin‑nosai‑ni‑waerebeetaa‑watsuka‑a‑nai.

Earthquake‑geninstance‑in‑topelevator‑topuse‑possible‑not

"lntheeventofearthquake

,elevatorscannotbeused."

Havingestablishedthatthewordsthatappearafterthesequenceofa VN‑clauseandtheparticle‑noarenouns,thestructureofadverbial VN‑clausesisillustratedasin{47}below.b

(47) [Np【cpDPs呵(XP)VNl一 ηoN】(P)

2.2DerivingtheNominalPropertiesofVN‑Clauses

Letusnowturntotheparticle‑no,whichhasthusfarbeenleft unidentified.Japanesehasseveralhomophonousparticles‑nowith

differentfunctions.Inthestructuresdiscussedinthisarticle,aplausible functionof‑noseemstobeeitheranominalizerorthegenitivecase

particle.Iwillarguebelowthat‑nothatmediatesbetweentheVN‑

clauseandthefollowingNshouldbeanalyzedasthegenitivecase particle.

Iwillbeginthediscussionbyconsideringwhethertheparticle‑no

(26)

164

immediatelyfollowingVN‑clausescanbeanominalizer.The

nominalizingparticle‑no(glossedasNer)attachestoaclauseasshown inexample(4$).

{48) [[Jun‑gakaisya‑oyame‑ta]‑no]‑oAkiko‑wa

[[Jun‑nomcompany‑accquit‑past]‑Ner]‑aceAkiko‑top

sitteiru.

knows

"AkikoknowsthatJunquitthe

company."

Onecharacteristicofthenominalizer‑noisthatitcanbereplacedby thenounkotòfact/thing'.

(49) [[Jun‑gakaisya‑oyame‑ta]‑koto]‑oAkiko‑wasitteiru.

[[Jun‑nomcompany‑accquit‑past]fact]‑accAliko‑topknows

"Akikoknows(thefact)thatJunquitthecompany

."

Theparticle‑nofollowingVN‑clauses,however,cannotbereplacedby kotoasillustratedin(54).

(50)a.[Koizumisyusyoo‑gagaimudaijin‑o

[Koizumiprimeminister‑nomforeignminister‑acc kootetu]‑nosai,kare‑nosijiritu‑gaagat‑ta.

replacing‑nooccasionhe‑genapprovalrating‑nomrise‑past

"WhenPrimeMinisterKoizumire

placedtheForeign

Minister,hisapprovalratingwentup."

(27)

b.*[Koizumisyusyoo‑gagaimudaijin‑o

[Koizumiprimeminister‑nomforeignminister‑acc kootetu】kotosai,kare‑nosijiritu‑gaagat‑ta.

replacing]factoccasionhe‑genapprovalratingnomrise‑

past

Thefactthat‑nocannotprecedeanounwhenaVN‑clause

modifyingthenounisreplacedbyacorrespondingVN‑suruclause indicatesthatnounsmodifiedbyVN‑clausesdonotrequirethe

precedingclausestobenominalized.Infact,asshownin{51)neither thenominalizer‑nonorkotoispossibleafterVN‑suruclausesinsuch contextsas:

(51)a.[Koizumisyusyoo‑gagaimudaijin‑okootetu

[Koizumiprimeminister‑nomforeignminister‑accreplace

‑sita]sai ,kare‑nosijiritu‑gaagat‑ta.

‑did]occasionhe‑genapprovalratingnomrise‑past

"WhenPrimeMinisterKoizumireplacedtheForeign Minister,hisapprovalratingwentup."

b.*[Koizumisyusyoo‑gagaimudaijin‑okootetu

[Koizumiprimeminister‑nomforeignminister‑accreplace

‑sita]‑nosai ,kare‑nosijiritu‑gaagat‑ta.

‑did‑Neroccasionhe‑genapprovalrating‑nomrise‑past c.*IKoizumisyusyoo‑gagaimudaijin‑okootetu

[Koizumiprimeminister‑nomforeignminister‑accreplace

‑sitarkotosai ,kare‑nosijiritu‑gaagat‑ta.

(28)

166

一did]factoccasionhe‑genapprovalrating‑nomrise‑past

d.[Sanbansen‑nitootyaku‑suru]ressya‑waTokyoyuki

[[No.3truck‑inarrive‑do]trains‑topTokyobound desu.

‑copular‑honorsc

"ThetrainarrivinginTruckNo

.3isboundforTokyo."

e.*[Sanbansen‑nitootyaku‑sure}no/kotoressya‑waTokyo [[No.3truck‑inarrive‑do]Ner/fact]train]‑topTokyo

yukidesu.

boundcopular‑honorific

Stillanotherargumentagainstanalyzingtheparticle‑noassociated withVN‑clausesasanominalizingparticlecomesfromthefactthatthe sequenceofVN‑clausefollowedbytheparticle‑nocannotbefollowed byacaseparticleorthetopicmarker‑wa.Whenthenorninalizing

particle‑noattachestoaclause,theresultingnominalizedconstituent needstobeattachedbyacase‑particle,thetopicmarker,orthelikeas alreadyseeninexample(48)above.Example(4S)contrastswith(52) belowwherethenominaiizedconstituentiswithoutacase‑particle.

(52) *[[Jun‑gakaisya ‑oyame‑ta]‑no]Akiko‑wasitteiru . [[Jun‑nomcompany‑accquit‑past]‑Ner]Aliko‑topknows

"AkikoknowsthatJunquitthecompany ."

AfairlystrongcasehasbeenmadetoruleoutthepossibilityofWno beinganominalizingparticleafterVN‑clauses.Weareleftwiththe

(29)

otherpossibilitymentionedabove.Iftheparticle‑noafterVN‑clauses isthegenitive‑caseparticle,aquestionarisesastowhytheparticleis obligatoryafterVN‑clauseswhileitisexcludedafterVN‑suruclauses inthesamecontexts.

Thefollowingexamplesillustratethatacaseparticlecanattachtoa clauseonlywhenitisfirstassociatedwithanN‑elementlike‑noand koto.7

(53)a.*[Jun‑gasyatyoo‑nisyuunin‑suru]‑gasudeni

[Jun‑nompresident‑toassumption‑doh‑nomalready ketteis‑are‑ta.

decide‑pass‑past

"Jun'stakingupthepostasthecompanypresidentwas alreadydecided."

b.[Jun‑gasyatyoo‑nisyuuninsuru‑surukoto/no]‑gasudeni [Jun‑nompresident‑toassumption‑dofact/Ner]‑nomalready ketteis‑are‑ta.

decide‑passive‑past

Asobservedin(51.b)and(51.c),inthecontextswherebothVN‑

clausesandVN‑suruclausesarepossible,thelattercannottakea

nominalizer.Therefore,case‑particlesincludingthegenitive‑particle cannotappearimmediatelyafterVN‑suruclauses.

Ifacuse‑particlecannotattachtoanun‑nominalizedclause,this leadstoanotherquestionastowhythegenitive‑caseparticlecanattach toVN‑clauses,whichexhibitclausalproperties.Theproblemposed

(30)

168

hereforVN‑clausesseemsanalogoustotheproblemofgerundsbeing

precededbyapreposition.Inbothcases,averbalprojectionis associatedwithanitemwhichisnormallyassociatedonlywiththe

nominalcategory.Inthefollowingitwillbesuggestedthattheproblem ofthegenitive‑caseparticleattachedtoVN‑clausescanbegiven

principallythesameexplanationastheaccountproposedforgerunds, VerbalnounsinVN‑clausesdonotdisplayverbalmorphology,and theseclausesdonotdenoteindependenttemporalWalues.Thus,like Englishgerunds,VN‑clausesarenon‑finiteclauses.Supposethatthe non‑finiteTinVN‑clausescontainsafeaturewithanominalproperty, anequivalentofthefeatureNTproposedforgerunds.IfinJapanesethe presenceofNTandthefeatureslicensingverbalmorphologyare

complementary,theabsenceofverbalmorphologyinVN‑clauses wouldbeexplained.Thepresenceofthegenitive‑nocanalsobe

attributedtothenominalpropertyofNT.SinceinVN‑suruclausesNT isnotpresent,thegenitive‑caseparticlecannotbelicensed.The structureinvolvingaVN‑clausecannowberepresentedasin(54).

(54) [TPDPs呵[vP(XP)VN1【T‑finite,NT】1‑genN

AssumingfurtherthatinJapaneseacase‑particleprojectsitsown projection,KP,thestructurein(54)canberenderedevenmoresimilar

tothatofgerunds.

(55) [Kp[Tp[vp(XP)VNHT‑finite,NTIHK‑gen]]N

(31)

In(55}thegenitive‑caseparticletakesTPofaVN‑clauseasthe complement,whichappearscomparabletoaprepositional complementizertakingagerundiveTPasthecomplement.

Iwillendthediscussionbysuggestingthepossibilityofthegenitive‑

caseparticle‑noappearingasaninstanceofacomplementizerafter VN‑clauses.Theexamplesin(56)belowindicatethatthe

complementizertoyuucanappearafterVN‑clausesinplaceofthe

genitivecase‑particle‑no.8

(56)a.【[Minsyutoo‑gasenkyo‑detaihai}toyuu]ori,

[democraticparty‑‑nomelection‑inbig‑loss]Comp]time too‑gabunnretu‑sita.

party‑nombreak‑up‑did

"WhentheDemocraticPartyexperiencedahugelossinthe

election,thepartywassplit."

b.[[[Minsyutoo‑gasenkyo‑detaihai]toyuu]kekka]

[【【democraticparty‑nomelection‑inbig‑loss】Comp]resulq

‑wayosousi‑nak‑katta .

‑topexpect‑not‑past

"TheresultthattheDemocraticPartyexperiencedahuge

lossintheelectionwasnotexpected."

Unlikethegenitive‑caseparticle‑no,thecomplementizertoyuucan alsoappearafterVN‑suruclausesasshowninthefollowingexamples.

(57)a.[【Minsyutoo‑gasenkyo‑detaihai‑suru)toyuu】

{{democraticparty‑nomelection‑inbig‑loss‑do]Comp]

(32)

17Q

or1,too‑gaunnretu‑seta.

time...

b.[[[Minsyutoo‑gasenkyo‑detaihai‑suru

[[[democraticparty‑nomelection‑inbig‑loss‑do]

toyuukekka‑wayosousi‑nak‑katta.

Comp]result]‑topexpect‑not‑past

Since,asshownintheseexamples,toyuucanappearaftereitherVN‑

clausesorVN‑suruclauses,itsoccurrencedoesnotdependonNT.The structureofVN‑clausesandVN‑suruclausesin(56)and(57)canbe illustratedasfollows:

(58)a.[cp[Tp[vp(XP)VN][T‑finite,NT]】[c'oy殿]]

b・【cP[TP[vP(XP)VN‑suru][T諭nite】1「ctoyuu]]

Totheextentthatanalyzingtoyuuasacomplementizeriscornet,the sentencesin(56)confirmtheclausalstatusofVN‑clauses.Itmaybe suggestedfurtherthatthegenitive‑caseparticle‑nocanappearafter VN‑clausesasaninstanceofacomplementizerthatconformstoNT.I willnotfurtherexplorethispossibilityonthisoccasion,butleaveitfor futureinquiry.

3.Conclusion

V‑ingformsinEnglishandverbalnounsinJapanesesharethe

propertyofprojectingthenominalorverbal(clausal)category.Ihave

proposedthatthesealternativepossibilitiesareduetothepropertiesof functionalfeatureswhichtheymergewith.Thenominalproperties

(33)

observedingerundsandVN‑clausesareshowntobeattributabietoa non‑finiteT‑featurewithanominalproperty.Duetothisfeature, gerundscanbecomethecomplementofP,whichappearsasa

prepositionalcomplementizer,andthegenitive‑caseparticlecanfollow VN‑clauses,Theanalysispresentedinthearticlesuggeststhat

classifyingfunctionalfeaturesintothedichotomyofN‑relatedandV relatedfeaturesmaynotbeadequate.Certainflexibilityallowedforthe choiceoffeaturesassociatedwithawordcouldresultinco‑occurrence ofthepropertiesnormallyassociatedwithdistinctcategories.

Notes

1

2

3

AsthecomplementofNandA,gerunds,likeDPs,needtobeprecededbya preposition.

(i)a.

b.

C.

11)a...(

b.

C.

John'spreferenceof[Meg'sstudyingFrench John'spreferenceof[MegstudyingFrench]

John'sproofofthetheorem

WeareawareofJohn'stakingthebarexam].

Weareawareof[Johntakingthebarexam].

Weareawareof[John'seagerness].

PesetskyandTorrego(2004)arguesthattheclauseintroducingthatandforarenot CbutaparticularrealizationofTmovedtoC.TheyextendthisanalysistoPs takinggerundivecomplementand,arguethatthecategoryPisakindofT,which occupiesapositionwithincertainDPs(gerunds}thatparallelstheposition occupiedbyTinCP.ThisarticledoesnottakePasTitself.Iwillarguethatin

gerundsPsareprepositionalcomplementizersthatcorrelatewithTthatlicenses thegerundiveinflectiononV.

TheobservationpresentedinAbney(1987)thatPOSS‑inggerundsshowmore nominalcharacteristicsthanACC‑inggerundscouldbeattributedtotherelative

prominenceof[+NT】inPOSS‑ingreflectedinthesubjectcase.Forexample,

(34)

172

4

5

6

PASS‑ingfaresbetterintheembeddedsubjectposition.

(i)a.IbelievethatJohn'ssmokingstogieswouldbotheryou.

b.?ibelievethat.lohnsmokingstogieswouldbotheryou.(Abney1987:172) Sententialadverbsalsoproducecontrastbetweenthetwo.Whilesententialadverbs arepossibleinACC‑ing,theydonotproduceequallyfelicitoussentenceinPOｺSS‑

ing.Forexample,

!

(ii}a.Johnprobablybeingaspy,Billthoughtitwisetoavoidhim.

b.?*John'sprobablybeingaspy,Billthoughtitwisetoavoidhim.(Abney 1987:180)

AsTsujimura{1996:138)describesverbalnounsinJapanesealsoincludeloan wordsaswell.asdeverbalnounsofJapaneseorigin.

a.Englishloanwords

b.deverbalnouns (3apaneseorigin}

kisu

deeto

doraibu

torihiki

toriatukai

ほロつ

tor1Slmarl ukekotae

"kiss"

"date"

"drive

"trading"

"handling"

"pickingup"

cc,,answer

Inadjunctsofreason/circumsatance,verbalnounscanprojecttonominal

projectionsasinadjunctsoftime.Incontrast,VN‑NPcannotappearasa

prenominalmodifier.

Jun‑noLondon‑e‑nosyuttyou)‑notsugou‑dekaigi‑gatoriyameninat‑ta.

Jun‑genLondon‑to‑genbusinesstrip]‑gencircumstance‑bymeetingcancelled

become‑past

"DuetoJun'sbusinesstriptoLondon

,themeetingwascancelled"

*[Sanbansen‑e‑notootyaku】noressya ‑wal7:30hatuTokyo‑yukide‑su .

[No.3truck‑in‑genarrivinggentrain‑top17:34departingTokyo‑boundcopular‑

honorific.

TherearealsowordsthatcanappeardirectlyafteraVN‑clausewithouttheparticle一 ὴ,.

(i}[Keiji‑gayougisya‑ojinmon]‑tyuu,sinhannin‑gatukamat‑ta.

(35)

7

S

detective‑nomsuspect‑accinterrogation]‑while,realculprit‑nombe‑arrested

"Whilethedetectivewasquestioningthesuspect

,therealculpritwas arrested."

(ii)[pro.daigaku‑osotugyoo}sidaiJun;‑wakagyoo‑otui‑da.

[pro;university‑accgraduation]‑assoonasJun‑topfamilybusiness‑acc

succeed‑past

"Assoonashegraduatedfromtheuniversity

,Jungsucceededthefamily

business."

(iii)[pro.daigaku‑osotugyoo]‑go,Jung‑wakagyoo‑otui‑da.

[propuniversity‑accgraduation]‑afterJun.‑topfamilybusiness‑osucceed‑past

"Afterhegraduatedfromtheuniversity

,Junsucceededthefamilybusiness"

Whilethenounsdiscussedinthetext,sar,or','ame,etc.areamongthenative

Japanesevocabulary,thenounsliketytctt̀while',sidaìassoonas'and80

̀after'areofChineseorigin

.Thewordsinthelatterclassare ,boundforms,and

thustheycanonlyappearaffixedtotheimmediatelyprecedingword.Therefore,

‑tr.ocannotmediatebetweenthesenounsandVNclauses.

Interrogativeclausescanbefollowedbythetopic‑markerandcase‑particles.

(i)[Dare‑nlsekinin‑gaaruka]‑wasaibansyo‑gahandan‑suru‑daroo.

Ewho‑toresponsibility‑nombeQ]‑topcourt‑nomdecision‑do‑future

"Thecourt‑willdecidewhoisresponsible

."

{ii}[[Dare‑nisekinin‑gaaruka]‑nohandan]‑wamuzukasii.

[who‑toresponsibility‑nombeQ]‑gendecision‑topdifficult

"ltisdifficulttodecidewhoisresponsible

."

{iii)[Dare‑nisekinin‑gaaruka]‑osaibankan‑gakimeru.

[who‑toresiponsib董ebeQ】‑acccourt‑nomdec{de

"Thecourtwilldecidewhoisresponsible

."

Nakau(197:0arguesthattnyuLiisthenouncomplementizer,whichisinvolvedin

theembeddingofasentenceasacomplementtoanounasopposedto'oandyθoη'

involvedinpredicatecomplementation.SincethedistributionofVN‑clausesis

restrictedtothepositionimmediatelyprecedinganoun,theycannotbefollowed

(36)

174

byacomplementizerbelongingtothelattertype.Examplescontaininbthese

complementizersareasfollows.

(i}[[Kinoko‑wagan‑oyoboo‑nikooks‑gaaruTP]toyuucF]syutyouNP]‑wa

C[CMushroom‑topcancer‑prevention‑toeffect‑nomhave]C]claim]‑top settokuryoku‑gaaru.

pursuasiveness‑have

"Theclaimthatmushroomsareeffectiveincancerpreventionispers

uasive."

(ii)[[Kinoko‑wagan‑yoboo‑nikooka‑gaaruTp]tocp】iw‑are‑ru.

[mushroom‑topcancer‑prevention‑toeffect‑nomhave]C]say‑pass‑pres

"ltissaidthatmushroomsareeffectiveincancerprevention

."

(iii)【【Kinoko‑wagan‑yoboo‑nikooka‑gaanITplyoonicplomow‑are‑teiru.

[[mushroom‑topcancer‑prevention‑toeffect‑nomhave]C]think‑pass‑

probress.‑pres

"Mushroomsarebelievedtobeeffectiveincancerprevention

."

References

Abney,S.(1987}"EnglishNounPhraseinitsSententialAspect,"Ph.D.dissertation,

MIT,Cambridge,Mass.

Emonds,」.(1985)A乙lnifiedTheoりrofSy'吻Ct'cCategories,Folis,Dordrecht,Holland.

Gueron,J.andT.Hoekstra.(1995)"TheTemporalInterpretationofPredication,"inA.

Cardinaletti,andM.T.Guasti(eds.),5yη'ωr伽4∫ 召配 αη"csVol.28∫ 脚〃Clauses,

77‑107,AcademicPress,SanDiego,Calif.

Grimshaw,J.(2005)Wθrd∫ αη43∫r部 σ̀̀re,CSLI,Stanford,Calif.

Johnson,K.(1988}"ClausalGerunds,theECP,andGovernment,"LinguisticInquiry,

Vol.19,5$3‑609.

Kayne,R.S.(1984)Cθ ηη8σ84'1e∬andBinaり2」Bra,tc・h'ng,Foris,Dordrecht,Holland.

Kayne,R.S.(2000)、Pαrω ηθ∫8雌 ω14Universals,OxfordStudiesinComparativeSyntax,

OxfordUniversityPress,Oxford,U.K.

Nakau,M.(1973)Serrto'π'calCo'npleme'1'ntゴ̀,η 加Jqρ α〃63ε,Kaitakusha,Tokyo.

Pesetsky,D.andE.Torrego.(2004}"Tense,Case,andtheNatureofSyntactic

(37)

Categories,"inJ.GueronandJ.Lecarme(eds.)TheSyntaxofTine,495‑537,The

MITPress,Cambridge,Mass.

Raposo,E.{1987)"RomanceInfinitivalClausesandCaseTheory,"inC.NeidelandR.

Nunez‑Cedeno(eds.),Stud'es'ηRo'nanceLanguages,237‑249,Foris,Dordrecht,

Holland.

Reuland,E.(1983)̀℃overning‑ing,"L加8配'、S'isInquiry,VoLl4.10レ】36.

Rizzi,L.(1982)LssuesinItalianSyntax,Foris,Dordrecht,Hoiland.

5hibatani,M.andT.Kageyama.{1988)"ReadjustmentandCompoundFormation,"in

W.Poser,(ed.),Papers ,from伽3εcoη41η'ern.a'1012Cl'Ẁ,rkshoPθ η ノ卯 αη召∫召

Syntax,193‑235,CSLI,Stanford,Calf.

Tsujimura,N.(1996)、AηIn'rodczc"̀ノ η ωJapanese乙加guis'ICS,BlackwellPublishers,

Oxford,U.K.

Wasow,T.andT.Roeper(1972)"OntheSu切ectofGerunds,"F̀胴da'inncifLarzgcrnge

S,44‑61.

Zanuttini,R.(1991)"SyntacticPropertiesofSententialNegation,"Ph.D.dissertation,

UniversityofPennsylvania,Philadelphia,Penn.

名詞的/動 詞的範疇 一 英語 、 日本語 の動名詞構 文か らの考察