Contrastive Tense and Aspect Data

ContrastivetenseandaspectdataDianaSantosINESCR.AlvesRedol,[e-mail]1991ForewordThisreportoriginatedinapaperIwastowritewithLauriCarlsononthesemanticsoftenseandaspect,andwhichneverbecameready,sincetheformalizationpartcouldnotbemadegoodenoughintherequiredschedule(rsttrimesterof1991).Intheoriginalpaper,Itriedtodescribethedataandproposeatheoryofmyown.NowIdecidedtoselectonlythedescriptionoftheproblemsandthegeneralphilosophyofthetreatmentIwasproposingatthattime,andpublishitasanINESCinternalreport.ThistextisthussimplyalinguisticpaperonthedierencesbetweentheEnglishandPor-tuguesetenseandaspectsystems,conceivedtoexplainthelattertonon-Portuguese-speakingreaders,andtoshowthatthedierencesmaybringsignicantinsightforthestudyofeachlanguage.Asfarasmyownworkisconcerned,thisisanhistoricalpaper,wheretheoriginsofmytheoryofaspectualclassescanbefound,withdetailedjusticationandexamples.However,atthetimeofmywritingthisintroductorynote(October1991),Ialreadyhaveamoresophisticatedtheoryofaspectualclassicationandaratherstrongopiniononhowthedatapresentedhereshouldbereformulated.Nevertheless,ifthedataaremainlycorrect,theirpresentationhere,nomatterhowbiasedbyanearlyview,shouldendurealongerlifethanendinginthewastebasket,Ithink.Notonlythebarefactscanbecited,butacomparisonofdierenttheoriestryingtoexplainthesefactswouldbepossible.Ithereforehopetocitethisreportoftenfromnowon,ifonlytodemonstratehowmynewviewsaremoreadequate.Awordofcautionshouldbeadded:thispaperdoesnotintendtocoveralldierencesbetweenEnglishandPortuguese,justafew.Itisalsonotverydetailedonthesimilarities,sinceitspurposewastohighlightproblemsfortranslation(orformodellingthesamephenomenonwiththesamemodelintwodierentlanguages).ThusitisnottobereadbythelaymanwhowantstolearnPortuguese.Iwouldsayitrequiressolidknowledgeofthetwolanguages,orthatthereaderisfamiliarwiththetenseandaspectliteratureanditsproblems.1Contents1.1Introduction.......................................31.2Generalremarks....................................31.3Aspectualclassication.................................41.3.1Acquisitions...................................41.3.2Intervalstates,statesandactivities......................51.4DierencesbetweenEnglishandPortuguese.....................71.4.1ImperfeitoversusPerfeito...........................81.4.2Presentperfectversuspastsimple......................91.4.3ThePortuguese\presentperfect.......................91.4.4PastperfectandMaisQuePerfeito......................111.4.5Progressiveforms................................111.4.6Perfeitoprogressivo..............................131.4.7Aspectualizers.................................131.4.8Temporalprepositions.............................141.4.9Temporalconnectives.............................151.4.10Dierentassignmentoflexicalaspectualclass................162Tenseandaspectdata1.1IntroductionThispaperreportsatheoryoftenseandaspect(henceforthT&A)whosedevelopmentisstillataninitialstage,andthusmanydetailsarenottotallyworkedoutnorallproblemssolved.Itseemsthatthereportingofthisendeavourmayhoweveralreadybeusefulgiventhatourapproachincorporatesasetofnewideaswhichweproceedtosummarize.Therstpointismethodological:weuseasdata(sourceofproblemsandconrmationofourideasaboutthelanguage)bilingualexamples.Thisleadsononehandtoalessintuition-drivendiscussionofT&Aphenomena,providingsetsofunarguableexamplestostartwith;also,itmakestheanalysislessbiasedbytheparticularitiesofaspeciclanguage,andalsohopefullybroaderincoverage.Inthispaper,weargueforauniedtheoryoftenseandaspect,wherebyuniedweun-derstandthatthemeaningofeveryandeachT&Amechanismworksessentiallythesameway,thatis,syntacticcategoriesdonotimplyatthelevelofsemanticsafundamentaldistinction.Whilethisideaisnotoriginal(seeforinstanceMoens(1987)),wepresentthesketchofaformaltranslationintorst-orderpredicatelogicthatobeysthiscondition.Thedepartingpoint,aspectualclasses,arerepresentedbymeansofprimitivestateswithaparticulartemporalprole.Then,aT&Amechanismisrepresentedasasetofassertionsaboutitsargumentplusasetofassertionsaboutthemechanismwiththeargument.Finally,wesuggesttheintroductionoftwonewaspectualclasses,’intervalstates’and’ac-quisitions’,inadditiontotheVendlerianclassication.WeseeaspectualclassesasprimitivesregardingsententialT&Acalculus,thoughtheyareinthemselvesdecomposable(asmentionedabove,weproposeasemanticdescriptionintermsoftemporalproleandbasicstates)inalex-icalsemanticsperspective.Still,weagreewithVerkuyl(1989)thatitisrathersomepropertiesofthoseclassesthatareactuallyusedintheaspecttheory,namely,thereferencetodierentkindsoftemporalobjects.1.2GeneralremarksWeproceedwithsomeremarksthatweconsiderrelevanttoafullunderstandingandassessmentofthiswork:Weconsideraspectualclass(verbaspect,Aktionsart,whateveryouwanttocallit)tobeameaningfulpropertyofVERBS.Thatis,averbcarriesinitsbasicmeaningsomeindicationofaspect.Theinteractionofothermechanismsmaychange(transform,restrict,addto)theinitialaspectualclass,andthisexplainswhyseveralsentenceswiththesamemainverbcanbelongtod