ThePragmaticsofIndirectReportsandSlurringAlessandroCaponeAbstractAccor - PDF document

ThePragmaticsofIndirectReportsandSlurringAlessandroCaponeAbstractAccordingtoVolosinov()thereisatensionbetweentwoindirectdiscoursepractices;oneinwhichthereportedmessage’sintegrityispreservedandtheboundariesbetweenthemainmessageandtheembeddedreportedmessageareformallymarkedandoneinwhichsuchboundariesaredissolvedasthereportingcontextallowsthereportingspeakertointrudetoagreaterextentandtransformthemessagebystylisticinterpolations.Thistensionisclearlyresolved,inthecontextofmypaperonindirectreports,throughtherecognitionofpragmaticprincipleswhichassigndefaultinterpretations(accordingtowhichtheboundariesbetweenthereportingmessageandthereportedmessageareclearlyvisibleandthereportedspeaker’svoiceprevailsatleastwithintheembeddedmessage),whileallowingcontexttocreateprioritieswhichoverridethedefaultinterpretationsandmaketheotherwisecostlyviolationsofthepragmaticprinciplesworthwhilethankstothefacilitationandsubordinationoftheinformationowtotheexigenciesoftheembeddingcontext(Ofcourse,thistensionisclearlyinstantiatedinlanguage(itisnotonlyatheoreticalproblem).Asarefereepointsout,wearefocusingonacaseinwhichtwopracticesareintension.Theresolutionofatensionbetweentwodifferent,possiblyopposite,practicesclearlydependsonpracticalconsiderationsleadingthelanguageuserstopreferonetotheother.Deviationfromapracticethatconformstoidealprinciplesofusemustalwaysinvolveacostthatneedstobeoffsetbypracticaladvantages.Oneoftheseadvantagescouldbethefacilitationoftherecognitionofareferent.Anotherpossibleadvantagecouldbe,ashappensinmanycases,thesimultaneousutteranceofaspeechreportandacriticism). A.Capone(UniversityofMessina/Palermo,SanFrancescoP10798051Barcellona,Italye-mail:alessandro.capone@istruzione.itA.Caponeetal.(eds.),PerspectivesonLinguisticPragmatics,PerspectivesinPragmatics,Philosophy&Psychology2,DOI:10.1007/978-3-319-01014-4_6,SpringerInternationalPublishingSwitzerland2013 1IntroductionThepracticeofindirectreportinginvolvesamixtureofseriousandnon-serioususe,asthispractice,ontheonehand,involvestransformationsinthesenseofGoffman(ontheotherhanditinvolvesusinglanguageinthecontextofaseriousactivity,suchasdescribingwhatanotherpersonsaid.Thepracticeofindirectreportingissensitivetocontextualinformationand,thus,itgoeswithoutsayingthatthericherthecuesandcluesallowingspeakerstointerprettransfor-mations(seeDascalandWeizman),themorecomplexarethetransforma-tionsinvolvedintheindirectreports.Andthemorecomplexthetransformationsare,thegreatertheneedforadecouplingprinciplealongthelinesofClarkandGerrig(Speakersintendtheiraddresseestorecognizedifferentaspectsoftheirquota-tionsasdepictive,supportive,andannotative.Mutatismutandis,wecanapplytheDecouplingPrincipletoindirectreports:Speakersintendtheiraddresseestorecognizedifferentvoicesbelongingtotheindirectreportand,inparticular,toseparatevoicesattributingthemtotheoriginalsource,thecurrentspeaker(theindirectreporter)orsomeotherpersoninvolvedincontext.Theyalsointendaddresseestorecognizesupportiveandannotativeaspects.Tomaketheconsiderationsabovelesscryptic,Inotethatsupportiveaspectsarethosewhichinonewayortheotherallowthespeakertomaketheindirectreport.Forexample,thereportermayuseEnglishtoreportaLatinutterance.ThisuseofEnglishisclearlysupportiveandNOTdepictive(ofcourse,hearersshouldhavepragmaticwaystodecouplesuchaspects).Annotativeaspectsarethosewhicharenoted,withoutservingaprincipalpurposeinthepracticeofreporting(forexampleIcannotethattheoriginalspeakerwasgigglingwhileusingacertainword).Depictiveaspectsconcernthewordsactuallyproffered.Ihavenowalreadydepartedtosomeextentfromthestandardpracticetoconsiderindirectanddirectreportsneatlydifferentiated.ClarkandGerrigthem-selvesconsiderthetwopracticestobeneatlyseparated,becausequotationprev-alentlymakesuseofdepictiveaspectsoflanguageusewhileindirectreportsmakeuseofdescriptions.Presumably,usingClarkandGerrig’sterminology,thereareotherreasonsforkeepingthetwopracticesdistinct.ClarkandGerrig(p.771)notethatquotationinvolvesbothseriousandnon-seriouslanguageuse.ItinvolvesseriouslanguageuseinthatthequoteditemissyntacticallyanNP;itinvolvesnon-seriouslanguageuseinthatthequoteditemissyntacticallyasentence(S)and,thus,depictiveelementsprevailiftheitemisconsideredasentence.Presumably, Forexample,shiftsfromserioustonon-seriousordepictiveuses.Presumably,ClarkandGerrigseemtoacceptthatanNPispresuppositional,thusexpressesanextensionalobject;asentenceembeddedinaverbofpropositionalattitudeorinaquotativestructurecanexpressanon-extensionalobject.SimplecasesthatcanillustratewhatClarkandGerrighaveinmindcouldbethefollowing:‘‘Iwantthatcar’’(or‘‘Iwantthecar’’).HereitisplausiblethattheNPfollowing‘want’identiesanextensionalobjectandnotONLYanA.Capone byClarkandGerrig’sstandards,indirectreportsshouldonlyinvolveserioususesoflanguage,sinceonlyNPsareinvolvedhere,ratherthansentencesintendedintheirdepictivesense.However,weallknowthatindirectreportingisveryoftenapolyphonicpracticewherethehearer’smaintaskistoseparatevoicesattributingthemtodifferentactors.EvenifwestaywithinClarkandGerrig’sterminology,itisuniversallyrecognizedthattherearewhatareoftencalled‘mixedquotations’,thatistosaycasesofindirectreportsinwhichsomesegmentsarequoted.Mixedquotationsareconsideredrelativelyrarecases—whilethepointofmydiscussionisthattheyshouldbeconsideredasprototypicalcasesofindirectreportsandthatindirectreportsingeneralshouldbemodeledaftermixedquotations(seeCaponeIhavealreadysaidthatindirectreportsareinterpretedincontext.Here,how-ever,theterm‘context’isambiguous,because,strictlyspeakingatleasttwotypesofcontextshouldberelevanttotheinterpretationofindirectreports:thecontextofutterance(oftheoriginalspeaker)andthecontextofutterance(oftheindirectreporter).Thereisoftenaninterestinginterplaybetweenthetwo.Weshouldnotefromthebeginningthatchronologicalconsiderationsareimportantinrankingthetwocontextsandthatthecontextofutterance(ofthereportingspeaker)isthedeparturepointfromwhichinterpretationstarts.Itisoftenuseful,therefore,tobearinmindwhatthepurposeoftheindirectreportisormightbe.Indirectreportsarecasesinwhichyoutransmitknowledgeofwhatanotherpersonsaidandwhatanotherpersonsaidistheonlywayoroneofthewaysinwhichyoucangainknowledgeaboutacertainsituationorevent .Thesituationisclearlydifferentfromthatofperception,wheretheonlymediatingelementsaretheperceptionsystemandcertainaprioriprinciplesofknowledge.Inindirectreports,thesituation istransformedtwotimes:oncebytheoriginalspeakerandthenby (Footnote2continued)intentionalone.TherearecaseswherewemightobjecttotheequationofanNPwithanextensionalobject,asinthecase‘‘Johnwantstosellhiscello’’.Here‘Hiscello’couldeitherescapethescopeofthemodal‘wants’oritcouldstillbeunderitsscope.Despitethesecon-troversialcriteria,therearesyntacticpositionscorrelatingwithextensional/non-extensional,suchas‘wantNP’,especiallyiftheNPisdenite.Anyway,IquiteagreewitharefereethatthecriteriabyClarkandGerrigarenotuncontroversial.Thesituationdescribedbyanindirectreportisusuallyanutterancebyanoriginalspeakerwho,inhisspeechact,describedorbroughtaboutasituation(inthecaseofanon-assertivespeechact).Asituationisastateoreventwithpossibleparticipantsinit.WhenIsaythatasituationistransformedthroughanindirectreport,ImeanthatthereporterusesNPsthatarenotneutral,butmayexpresshis/herpointofviewand,inparticular,acriticalattitude(forexampleifthereportermakesuseofepithets).Thesituationmaybetransformedinanotherway,asthereportermayavoidusingNPsactuallyusedbytheoriginalspeaker,butmayusedifferentNPstomakesurethattheHearercanidentifythereferentinquestion.Iusedtheterm‘transformations’butIcouldhaveusedtheterm‘modications’.However,‘transformations’referstoanoperationeffectedlinguisticallyandpossiblyinasystematicway.Infact,itmightbepredictableandthereforesystematicthatifthehearercannotidentifythereferentthroughanNPusedbytheoriginalspeaker,thereportermustuseadifferentNP,onethatallowsthehearertoidentifythereferent.Transformationsarerathersystematicpractices.WhenyouareconfrontedwithanindirectreportThePragmaticsofIndirectReportsandSlurring155 thereportingspeaker.So,thetaskofthehearerisclearlyaninferentialtask;howtodeletepossibletransformationsandhowtoget(back)to withouttheinter-ferenceofpossibletransformations.Thisisclearlyaninferentialtaskrequiringpragmatics.Now,iftheheareroftheindirectreportisinterestedintheindirectreportmainlybecauseitallowshertohaveaccessto ,thereportingspeakerknowsthisandmayverywelltakethisintoaccountinhertreatmentoftheinformationconcerningtheoriginalutterance.SowemaygrantthatatleastpartofthetransformationsmaybeshapedbythedesiretomeettheinterestsofHinknowingabout .OthertransformationsmaybeindependentoftheinterestsofHormayconictwithit.Justtomentionacase,considerthereporterwhosaid:‘JohnsaidthatthebusforOxfordisontheleftwhenyougetoutoftheairport’.Itiscrucial,inthisinterpretationprocess,thattheperspectivebethesame.Andthatmustbe:passengergettingoutoftheairport.Iftheperspectiveadoptedintheindirectreportwasdifferentfromthatadoptedintheoriginalutterance,confusionwouldensue.Thusweexcludethattheperspectivecouldbe:relativeswaitingforthepassengeroutoftheairport.If,forsomereason,theindirectreportertrans-formedtheutterancewithouttakingintoconsiderationthehearer’sinterests,anuninterpretableutterancewouldresult(ortobemorecorrectanutterancepro-vidingmisleadinginformationwouldresult).Asupshotofthis,thepurposeoftheindirectreportmustfeatureprominentlyamongthefactorstotakeintoaccountintheinterpretationaswellasintheproductionofindirectreports.Letusconsider,provisionally,thebasicstructuralelementsthatgointoanindirectreport.Context1(originalspeaker;originalHearer)Context2(reportingspeaker;reportingspeaker’sHearer)DecouplingPrincipleSeparatetheoriginalspeaker’sfromthereportedspeaker’svoice.Establishwhichportionsofthetexthaveadirectlypictorialfunction.Separatethosepartswhichhaveasupportiveoranannotativefunction.Purpose1(originalspeaker)Purpose2(reportingspeaker)Purpose3(addressee). (Footnote3continued)thatmakesuseofepithets(thatbastard),youmaybeprettysurethatthespeakerisusinglanguageinacriticalwayandthusasystematiceffectontheheareristhedesiretoknowthedifferencebetweentheNPusedbythereporterandtheoneusedbytheoriginalspeaker.Thesetransfor-mationsaresystematicalsointhesensethatitmightbepossibletospotthemandtogobacktotheoriginalutteranceviareectiveprocesses.Perhapsthebestexampleofthepictorialfunctionisthefollowing:JohnhasSEENMaryintheBATHROOM.Therearecases,liketheoneabove,inwhichlanguageisusedtoexpresstheform(boldface,forexample)ofanutterance.Inthiscasewehaveavisualdimension,butsometimeswehaveanauraldimension,asinthecaseofaspeakerwhoimitatesthevoiceofanotherspeaker(imitation)(orthestyle).Normally,however,by‘pictorial’ClarkandGerrigmeanthataspeakerdepictstheactualwordsemployedinacertainutterance.A.Capone Pointofview1(Originalspeaker)Pointofview2(reportingspeaker)Pointofview3(addressee).Nowthatthestructuralcomponentsofthepracticeofindirectreportsareinplace,wecanexpectthatatheoryofindirectreportscouldbebuiltobjectivelyonthisbasis,perhapsonthebasisofrecursiveoperationsthattakeintoaccountthebasiccomponents.IndirectreportsaslanguagegamesInmypreviousworkonindirectreportsIhavefocusedonreportsaslanguagegames(Capone).Languagegamesareactivitiesproducedthroughspeechinconformitytosocialrulesdeterminingwhatcancountaswhat(inthespeechsituation).Alanguagegameisastripofsocialactivity(ofsociallife)wherelanguage(speechacts)playanimportantroleintheexecutionanddeco-dication(andinterpretation)oftheactivity.Alanguagegameisaformoflife,theindividualbeingablethroughittoparticipateinasocialformoflife(beingintegratedinasocialdimensionandcoordinatedthroughactionwithothermembersofthegroup).Now,whiletheremaybedifferencesbetweenGoffman’sterminologyasusedintheprevioussectionandtheterminologyoflanguagegames,itisalsoclearthatthereissubstantialoverlap.Goffmanpresumablysawthecontinuumofsocialpracticesassegmented[or‘framed’(Goffman1974Eachsegmentwastoberecognizableastherehadtobeboundariesbetweenoutsideandinsideactivities.AnexampleofGoffmaniananalysisthatiswellknownisthatofthelecture(Goffman).Thelectureisaboundedactivity,whichhasitsownrules.Participantsknowwellandinadvancehowtobehaveinthissegmentedarea,theyknowthatthereislittlespaceforinterruptions,theyknowthatlectureshaveaforthcomingsegmentreservedforquestionsandanswers(bythelecturer).Clearly,thelectureisalsoalanguagegame,becauseitisstructured,ithasrules,itispartofsocietalactivities,itissufcientlydifferentiatedfromotherlanguagegames.So,substantially,Goffman’stheoryofframesandformsoftalkmustcoincideinbroadlineswithatheoryoflanguagegames—oratleastitmustbepossibletoexploreinterconnectionsandoverlappingterritory.Butwhyshouldwewanttodealwithindirectreports—activitiesconnedtosmallsegmentsofinteraction—intermsoflanguagegames,whichareusuallyactivitiesthatunfoldforsometimeandoccuratsomeplacewhichissubstantiallyinvolvedinthelanguagegameandevenservestocharacterizeit(forexample,courtprocedures)?Andnowmyansweristhatevenifindirectreportsarenotnormallyreallyextendedintimeaslanguageactivities,theyinvolveembeddingssuchasthosedescribedintheIntroduction,andwhichwemayillustratesche-maticallyasinthefollowing:IndirectreportC(indirectreporter) InconformitywithWittgenstein’sPhilosophicalInvestigations.ThePragmaticsofIndirectReportsandSlurring157 OriginalspeakerC(originalspeaker)AddresseeC(addressee).Althoughthelanguagegameistemporallylimited,ifweconsiderthedimen-sionofthelinguisticactivitythatunfoldsinatemporalsuccession,thetemporalembeddingsobtainedbyreconstructingtheoriginalspeaker’ssituationarepotentiallymanifoldandcomplex.Complexityisintroducedwhenweseecon-nectionwithotherlanguagegamessuchasthefollowing.Considerachildgame,whichalmosteveryonepracticedinchildhoodoradolescence:Thereare,say,20boys(orgirls)inaroom.Eachwhisperstothenextpersonintheline(orcircle)whatwaswhisperedintohishearpreviously.Theaimofthegameistoshowthat,although,ideally,theinitialandthelastutterancehavetobethesame,theinitialutteranceissotransformedthatthelastutterancecanhardlybeheardtobearanymeaningfulrelationtoit.Thismightbeagamepointingtoapracticewhichisquitestandardinsocietyandisbasedonreliablemethodsfortransmittingandpreservinginformationduringthetransmissionprocess.Thepreviousgamedramaticallyillustratestheproblemsinherentinthegame‘reportinginformation’or‘reportinganutterance’.Consideranothergamesuchasthedumb-show.Adumb-showwasoneofourfavoritegamesinchildhood.Wepracticedit,Ipresume,asaformofpreparationinviewofmoreseriousorimportantsocietallanguagegames.Inadumb-showyoumustdepictinformationbyavoidingwords.Youusuallyusegestures,evenifyoucanpointtowordswhichhappentobewrittenonablackboardoronaposter.Now,sincedepictingoccurssoheavilyindirectreportingand,alsoinindirectreports,thisisclearlyacaseinwhichweconsider‘depicting’animportantpartoflanguagegames,acomponentwhichmaybesharedbydifferentlanguagegames.Anotherlanguagegamewhichiscrucialfortheunderstandingofindirectreportsisatheatricalperformance.Inatheatricalperformanceweusuallypretend,wearenotusinglanguageinaseriousway.Anactordoesnottalkforherself,butonbehalfofacharacter.Thisismoreorlesswhathappensindirectreports,butalsowhathappensinindirectreports,ifweconsiderthemaspolyphonicactivities(seealsotheproblematiccaseofmixedquotationinindirectreports).Anothercaseoflanguagegamethatisdeeplyrootedinsocietyistestimonyincourt.Hereitmightbeimportanttobeabletoreportwhatanotherpersonsaidonacertainoccasion.Thismaywellbeanextremecase,wherethereislittlefreedomfortransformationsandwhereoneneedstoseparateone’svoicefromthatoftheoriginalspeakerbyformalmarkings.Thispracticedivergesfromthedailyprac-tice,toagreatextent.Hereareportermaybeaskedbytheprosecutortoreectonthewordsused,tomakeanefforttoseparateherownvoicefromthatoftheoriginalspeaker.Thispracticemaywellinvolveameta-representationalcompo-nent,asoneisinterestedinthemeaningsaswell,intheconnotationsaswellasinthedenotations.Thereportermaywellbeturnedintoananalystofherownspeech.A.Capone (Doyouexactlymeanthat?).Thisisclearlyacasewherereportingiscon-sideredameta-linguisticactivity.So,althoughitistruethatindirectreportsaresmallsegmentsoftalkorsmallstripsofsocialbehavior,theyneverthelesshavemanyfeaturesincommonwithotherstripsofbehaviorwhichwearelessreluctanttocall‘languagegames’.Thismaybeenoughtoseethattheconnectionbetweenlanguagegamesandindirectreportsiswelljustied.Dascaletal.considerthatthenotionoflanguagegamebyWittgensteininvolvesashiftfromphenomenalismtophysicalism,languagegamesbeingprimarilyintendedtocreatesocialreality.Canthelanguagegameofindirectreportingbesointended?IfwefollowTannen(),indirectreportscan,indeed,beconsideredasactionsservingtoconstructsocialreality.Anindirectreportcanhaveeffectsondeliberation,onaction,inthatitcanpresentapieceofinformationthatcanbeintegratedintotheargumentativestructureofpracticalreasonings.Seeninthislight,anindirectreportcanbecomea‘formoflife’(WittgensteinAnotherfeatureoflanguagegames,accordingtoDascaletal.()isthattheyarecooperative[theyareconstructedjointlybydifferentspeakers(orthespeakerandthehearer)].Canthisbeacharacteristicofindirectreports(suchaslanguagegames)?MyreplyinCapone()wasthattherecognitionoftheroleoftheHearer(oraddressee)intheamountoftransformationsrequiredinthepracticeofindirectreportingamountstoarecognitionofthecooperativenatureofindirectreports.Indirectreports—likeotherlanguagegames—involveanaltruisticstancetowardstheaddressee,whichisinstantiatedinimportantlinguisticchoicesthatcanbeseenastransformations.InCapone()Ispecicallydiscussedindirectreportsaslanguagegames,inthelightofconsiderationsbyDascaletal.()onlanguagegames.HereIcannotexpandthatdiscussion,butIconnemyselftoextrapolatingthemostimportantpoints.Dascaletal.considerpolyphonyaspeciclanguagegame—now,whilesurelyindirectreportsareinterestingalsoforotherfeatures,suchasrepresentationalones,itisclearthatpolyphonyisalanguagegamethatisembeddedinthepracticeofindirectreporting.Thegamealsoconsistsinthewaycluesandcuesareutilizedtoseparatethevoicesoftheparticipants.Thelanguagegame‘polyphony’aimsattheintegrationofdifferentvoices(expressingdifferentpointsofview).Integrationdoesnotmeansummation,butaninteractionbetweentwopointsofviewsuchthatoneis,often,acommentaryontheother.Oneoftheproblemsweencounterinthedescriptionofindirectreportsis,infact,thatanapparentlysingleutterancecontainsdifferentvoices/pointsofview(thusitispolyphonic),apparentlymakingitdifcultforthehearertoseparatethem.However,theproblemisnotonlyhowtoseparatepointsofview,buthowtoseetheinteractionbetweenthem.Inanindirectreport,wedonotonlyhaveaneutralpresentationofpointsofview,butnormallythepointofviewoftheindirectreporteristhemainlterthroughwhichwehearothervoices.Thus,ithappenscharacter-isticallythattheremaybearelationshipofcriticismorotherwiseafliationbetweenthepointofviewoftheindirectreporterandthepointofviewofthereportedspeaker.Polyphony,asalanguagegame,thereforedoesnotconsistinamereThePragmaticsofIndirectReportsandSlurring159 summationofvoices,butintheintegrationofthemwithinarelationshipofcom-mentarybetweenthevoices(onevoicebeingacommentontheother).Polyphonyinvolvescommentaryinthatthereportingspeakercanpresentthereportedspeaker’svoiceinaderogatoryway(hemayshout,speaksardonically,expressrage,etc.whenexpressingthereportedspeaker’svoice).Wemayverywellspeakofaconcertofvoices,whicharehowever,regimentedbythereportingspeakerandbytheinfer-encesofthehearer.Theheareriscapableofusinginferencestodifferentiatevoices,butalsotonoticeifsomeelementisaddedillegitimatelybythereportingspeaker.So,thegameisnotonlyonewhichhasthereportingandthereportedspeakerasitsmainparticipants,butonewherethehearerisanimportantjudge,whocanaddthingsnotsaidorsubtractunnecessaryelements.DavidsononindirectreportsInthispaper,Iamnotafterthelogicalformofindirectreports.Iammainlyafterapragmatictreatmentbasedonthenotionofthelanguagegame.However,IwillbrieymentionDonaldDavidson’s()treatmentofthelogicalformofindirectreportsbecauseitisthetreatmentthatbestaccordswithmyviewofindirectreportsaslanguagegames.AccordingtoDavidsonasentencesuchas:1.JohnsaidthatMaryisinParisistobeaccountedfor,truth-conditionally,bythefollowinglogicalform:Johnsaidthat.MaryisinParis.Inotherwords,Davidsonasksustoconsideraproposalaccordingtowhichthecomplementizer‘that’disappearsfromlogicalform,beingreplacedbythepro-nominal‘that’.Aproposofthis,Davidsonbrieymentionshistoricalconsider-ationsonthedevelopmentofthecomplementizer‘that’fromthepronominal‘that’.Now,IamawarethatthereisastrandofresearchthatbuildsonDavidson’sproposal(sometimesaimingtoameliorateit,sometimesaimingtodestroyit;seeRumtt).ButasinthispaperIammainlyinterestedinthelanguagegame‘indirectreport’andinthepragmaticsofindirectreporting,Iwillskipsuchdis-cussions.IwillneverthelessrehearsesomeconsiderationsbyDavidson,whicharenowverypopularinphilosophy:Wetriedtobringtheavoroftheanalysistowhichwehavereturnedbyrewordingourfavoritesentenceas‘‘Galileoutteredasentencethatmeantinhismouthwhat‘Theearthmoves’meansnowinmine’’.Weshouldnotthinkillofthisverboseversionof‘‘Galileosaidthattheearthmoves’’becauseofapparentreferencetoameaning(‘‘Whattheearthmovesmeans’’),thisexpressionisnottreatedasasingularterminthetheory.Weareindeedaskedtomakesenseofthejudgmentofsynonymybetweenutterances,butnotasfoundationsofatheoryofLanguage,merelyasanunanalyzedpartofthecontentofthefamiliaridiomofindirectdiscourse.Theideathatunderliesourawkwardparaphraseisthatofsamesaying:whenIsaythatGalileosaidthattheearthmoves,IrepresentGalileoandmyselfassamesayers.(Davidson,140).Now,byextrapolatingthisexcerpt,IwanttoemphasizethatforDavidsonitwasclearthatoratioobliquaisadiscourseinvolvingmultiplevoices.ThemouthsA.Capone utteringthewordsbelongtodifferentpersons.Thesentencesactuallyuttered,asfarasDavidsonisaware,mayverywellbedifferentprovidedthatthetwoutterancesaresemanticallyequivalent,thatistosaytheirimportsaretruth-con-ditionallyequivalent.Therearetwovoices,twopointsofviewinvolved,andanindirectreportis,obviously,atransformationoftheoriginalutterance.Baldwin,273)claimsthatonedefectstandardlyattributedtoDavidon’sformulationofthetheoryisthatitseemstoimplythatthereisonemoreutterancebesidestheutterance‘Theearthmoves’.This,whichfromaphilosophicalpointofview,countsasadefect(whichcouldberemediedanyway,ifwefollowthediscussioninBaldwin),isnotnecessarilyadefectfromalinguisticpointofviewasitmakesusseethatthecaseofindirectreports(anditslogic)dependsonthetensionbetweenthereportedspeaker’svoiceandthereporter’svoice.Itisnosurprisethattheremaybetwoutterances,whosecontentisfundamentallythesame,althoughpartsofit,thosepartswhichdonotcountfortheprovisionofanextensionalsemantictheoryofindirectreports,neednotbethesame.TheconsiderationsbyDavidsononp.143arenotequallyfamous,butinmyopiniontheyleadtoaviewofindirectreportsaslanguagegames,inthestudyofwhichpragmaticsisprevalentlyoratleastsubstantiallyinvolved:Wewoulddobetter,incopingwiththissubject,totalkofinscriptionsandutterancesandspeechacts,andavoidreferencetosentences.Forwhatanutteranceof‘‘Galileosaidthat’’doesisannounceafurtherutterance.Likeanyutterance,thisrstmaybeseriousorsilly,assertiveorplayful,butifitistrue,itmustbefollowedbyanutterancesynonymouswithsomeother.Thesecondutterance,theintroducedact,mayalsobetrueorfalse,doneinthemodeofassertionorplay.Butifitisasannounced,itmustserveatleastthepurposeofconveyingthecontentofwhatsomeonesaid.(Davidson,143).AtthispointwenoticethatDavidsonhastouchedonadeepissue—thecontentofindirectreportsmaybedeterminedpragmatically.So,itispossiblethattheutterancexfollowing‘‘Galileosaidthat’’maybesynonymouswithanutterancewhichisnottruth-conditionallyequivalenttox,butcanbemadepragmaticallyequivalenttox,saythroughpragmaticintrusion.(Inotherwords,weshouldconsidertheexplicaturesastruth-conditionallyequivalent).Ingeneral,theexcerptaboveraisedtheimportantquestionthatthepurposeandthespeechactcommu-nicatedbytheindirectreportmayprominentlygurewhenwetrytoestablishwhetherthereportingutteranceandtheoriginalutterancematchincontent.Idiscussed,however,brieythisnotioninCapone().Forthesakeofthisdiscussion,itisimportanttopointoutthatDavidsonthinkswemustseparatetruth-conditionalcontentandpragmaticcontent.EvenifDavidsondoesnotmovetowardsaradicalpragmaticviewofindirectreports,itisclearthatthenotionofpragmaticequivalenceiswhatisatstakewhenwesaythattheoriginalutteranceandthereportingutterancematchincontent.Suppose,forexample,thattheori-ginalutteranceis:2.Marioisreallybraveandthereporter,whetheraccuratelyornot,transforms(2)byuttering(3)(withaviewthat(3)and(2)matchincontent).ThePragmaticsofIndirectReportsandSlurring161 3.JohnsaidthatMarioisalion.Shouldwesaythattheindirectreportmatchesincontenttheoriginalutterance?Forsomepurposes,wemaybepersuadedtoanswerpositively,evenifthelocu-tionaryforcesoftheseutterances(clearly)donotmatch.Itmaybearguedthat(2)and(3)cannotbetakentoconveythesamecontent,asmetaphoricalmeaningisinsomesensenon-conventional(asreferee2says,inordertoguaranteesamenessofcontentthemetaphorwouldhavetobeconventional.Butthenitwouldbewellonitswaytobeingalexicalsenseoftheexpression).Iamnotpersuadedbytheideathattwoutterancesmatchincontentonlyiftheconventionalmeaningsmatch,butofcourseIagreewithreferee2thatametaphoricalexpressionconveys(usually)muchmorethantheconventionalexpressionitwasusedtoreplace.Thereareeffectsintermsofpoetry,force,rhetoricwhicharenotexpressedbyanon-metaphoricalexpression.Butgrantingsomedifferences,thinknowofthefol-lowinglanguagegame.Wecanreportthoughtsbyusingcertaincards,oneachofwhichacertainwordisprinted.Wedonothavecardsforeveryword.Sowemustdowhatwecantoexpressourthoughts,andourreadersmustaccepttheapproximationswhichwecanuse.Nowsupposewehaveacardfor‘lion’butnotonefor‘brave’.Couldweengageinthelanguagegameofreportingthespeechact,nevertheless?Theanswer,inthecontextofthislanguagegame,withitsobviouslimitations,ispositive.Toreport‘JohnsaidthatMarioisalion’iscertainlybetterthannothingandourreaderswillhavetoputupwiththelimitationsofourlanguagegame.However,eveninadifferentcontext,areportingspeakermaywanttomodifysomewhattheoriginalutterance,toconveysomethingwhichJohndidnotsaybutprobablywantedtosay(orwouldhavesaidindifferentcircum-stances).Perhapsthereportingspeakerisjudgingthathisindirectreportismorefaithfultothespeaker’sintentionsthantheoriginalspeaker’swords.Perhapsthereportingspeakerisrelyingonclueswhicharenotavailabletothehearer(orreader)andisreconstructingthespeaker’sintentionstothebestofhisownabil-ities.Afterall,arewenotallowedtoinferandvoicesomeone’sintentions,evenifthatpersonwasnotcapableoffullyexpressingthem?Now,thisargument,clearly,hastakenmesomewayfromtheconsiderationsbyreferee2.Inendingthissection,IwanttoremindreadersthattheinitialDavidsonianformulationofindirectdiscoursewascriticizedbecauseitwasimmunetointen-tionality(Baldwin1982,272)andwasthuslaterreplacedbyabetteranalysiswhichwascompletelyextensional(Baldwin,273):Galileosaidxiff(y)[GalileoutteredyandSameincontent(x,y)].Capone()andindirectreportsInCapone()Iadvancedanumberofideasonhowtocaptureconstraintsonreplacementsofco-referentialNPsinthecontextofindirectreporting(and,inparticular,inthecomplementthat-clause).Theexplanationmaybeparallel,butnotidenticalwiththeoneIgaveontheissueofbeliefreportsinCapone(Suchanexplanationrestsontheideathatreplacementsofco-referentialNPsshouldnotalterthespeechactwhichtheindirectreportaimstoreport(ordescribe)A.Capone andthattheoriginalspeakerwouldliketoseeherselfreportedinsuchawaythatitdoesnotattributeheroffenses,impoliteness,rudeness,obscenity,andalsoslurring.Inotherwords,reportingmustbedoneinawaythatthevoiceofthereporterisseparatedfromthevoiceofthereportedspeakeror,ifthisseparationisnotpossible,insuchawaythattheoriginalspeaker’svoiceisprevalent.Whyshouldthereportedspeaker’sandNOTthereportingspeaker’svoicebeprevalent?Iassumethatitisamatterofrelevance.Sincewearedealingwiththeverb‘say’,wearehappytoprimarilyexpresstheoriginalspeaker’svoiceandthenthereportingspeaker’svoice,butonlyifthisispossible.InowsuccinctlysumupthemainpointsofCapone(Thepracticeofindirectreportsrestsonthefollowingprinciples:ParaphrasisPrincipleThethat-clauseembeddedintheverb‘say’isaparaphrasisofwhatYsaidthatmeetsthefollowingconstraint:shouldYhearwhatXsaidhe(Y)hadsaid,hewouldnottakeissuewithit,butwouldapproveofitasafairparaphrasisofhisoriginalutterance.ThefollowingisaprecisicationofthepreviousPrinciple,whichremediessomeofitsdefects,asitdoesnotonlytakecontentintoaccount,butalsomakesreferencetoform.Paraphrasis/FormPrincipleThethat-clauseembeddedintheverb‘say’isaparaphrasisofwhatYsaid,andmeetsthefollowingconstraints:shouldYhearwhatXsaidhe(Y)hadsaid,hewouldnottakeissuewithit,astocontent,butwouldapproveofitasafairparaphrasisofhisoriginalutterance.Furthermore,hewouldnotobjecttovocal-izingtheassertionmadeoutofthewordsfollowingthecomplementizer‘that’onaccountofitsform/style.InmypaperIalsodiscussedpossibleobjectionstotheParaphrasis/Formprinciple.Sincethisdiscussionwillbeampliedinthepresentpaper,Ipresentsomeoftheoriginaldiscussioninthissection.Dependingonthecontext,Ineedn’tbebeholdentotheoriginalspeaker’s‘approval’ofmyparaphasisasfair,norneedIavoidmannersofspeechwhichtheoriginalspeakerwouldshyawayfrom.Insuchcontexts,ifJohnsaidofapersonxthathewillbecomingtotheparty,myreporttothateffectistruewhetherIrefertopersonxpolitely,asJohnwouldapproveof,orimpolitely,as(letusimagine)myhearerwouldapproveof.Johnmay,uponhearingmyreport,demur:‘‘Well,Idon’tknowwhyyou’dcallxajerkbut,yes,Ididsayhewascomingtotheparty’’.TheParaphrasisPrincipleandtheauthor’sotherremarksareintendedtoruleout ThispositionissomewhatreminiscentofSeymour’s()treatmentofindirectreports,inwhichreferencetoatranslationofthereportedsentenceisexplicitlyincorporatedinthesemanticsofindirectreports.ThePragmaticsofIndirectReportsandSlurring163 contextsofindirectreportingthatseemtoallowthistypeoflicensewiththeoriginalspeaker’swords.AsIsaidinCapone(),IamquiteopentothepossibilitythatinsuitablecontextsoneshouldbeabletoreplaceanNPwithacoreferentialexpressioninthethat-clauseofanindirectreport.However,Isticktotheproposalthat,intheabsenceofabundantcontextualcluesandcuesallowingustoseparatetheoriginalspeaker’svoicefromthatofthereporter,thedefaultinterpretationoftheutteranceconformstotheparaphrasisrulesstatedabove.SomeconsiderationsonWielandonindirectreportsWieland()considersthatmosttheoriesonindirectreportsconcludethatthepracticeofindirectreportingmustbestudiedessentiallyfromapragmaticpointofview.Wieland,however,refusestoacceptthatonecannotsaysomethingsystematicandofgeneralimportaboutthepracticeofindirectreporting.Sheisadamantinconsideringthecaseofindirectreportingdistinctfromthecaseofquotationandthecaseofbeliefreports.Now,ifsuchpropositionsareaccepted,itgoeswithoutsayingthatindirectreportsallowacertainamountofsubstitution(ofNPshavingidenticalreferents)andthusitisnottobetakenforgrantedthattheyarecharacterizedbyopacity.Sincetheyarenotexpressionsofbelief,theattitudeoftheoriginalspeakerneednotinterferewithsubstitutionofNPshavingidenticalreference.Now,Idonotwanttodisputethesepropositions,asthereisobviouslysometruthinthem.Butitispossiblethattheinferentialstepfromthesepropo-sitionstothelackofopacityexhibited(accordingtoWieland)byindirectreportsisnotnecessaryorneeded;inotherwords,itmaydistractusfromsomeobviousconnectionsbetweenatheoryofquotationandatheoryofindirectreports.Andthemostobviouslinkbetweenthetwotheoriesisthatinbothcasesweneedtoestablishwhichvoicesbelongtothevarioussegmentsmakinguptheutterance.Indirectreporting(asmadeclearbyCappelenandLepore)involvesmixedquotation,atleastinsomecases.Sotheonlywaytomakethetwoissuesseparatenowistoinsistonquotationasbeingcharacterizedstrictlybyopacityandindirectreportsasnotbeingcharacterizedbyopacity(orinbeingcharacterizedlessstrictlybyit).However,ifwegrantthatindirectreportscancontainquotedsegments,itislessclearthatopacityandlackofopacitycanbeusedtodistinguishthetwocases.Inmyarticleonquotation(Capone2013Iinsistedthatinvertedcommasneednotalwaysbeusedtosignalthequotativefunction,astheyareoftenabsentintheorallanguage.Ratherweneedpragmaticwaysofsignalingthatcertainsegmentsarebeingquoted.Butifthisisthecase,thenitgoeswithoutsayingthatimplicitlymanysegmentsofindirectreportscancomeoutasbeingquoted,atleastthrough OneofthemosttypicalcontextsallowingsubstitutionsofcoextensiveNPsisonewheretheNPusedintheoriginalspeechactwouldnotallowthehearer(oftheindirectreport)toidentifythereferent,andthusthereporterdeemsitnecessarytouseanNPwhichdoesindeedallowthehearertorecognizethereferent(SeealsoCapone).OntheroleplayedbycontextininferentialprocessesseeCapone(SeealsoSaka()foradiscussionofquotationinphilosophyoflanguage.A.Capone somepragmaticmeans.ThesedifferencesofopinionbetweenWielandandmyselfdonotpreventmefromseeingtheimportanceofherotherconsiderationsonindirectreports.Anditisonthesecrucialconsiderations—whichIshouldsayarebothimportantandcontroversial—thatIwanttoconcentratenow.SomehowdepartingfrommyconsiderationsinCapone(),Wielandarguesthatinsomecontexts,whenthereportingspeakerhasapurposewhichservestoadvancethecommunicationprocess—ratherthanimpedingitthroughtheuseofanNPwhosesemanticimportisnotknowntothehearer—itislicittointer-substituteco-referentialterms.Considerthisco-referentialsubstitution:4.A:Myfavouritepatatasbravas.B:Asaidthatherfavoriteisthethirditemonyourmenu.Wielandsays:Inthiscase,theterm‘patatasbravas’issubstitutedwithadenitedescriptionwithavaluethatcanonlybedeterminedinthereportingcontext.Itwouldbeimplausibletosuggestthattheoriginalspeakermeantanythinglike‘thethirditemonyourmenu’intheoriginalcontextofutterance.Nevertheless,ordinaryreportingpracticestakeadvantageofthissortofinter-substitution(WielandAndIagreethatinreportingtheoriginalutterancebytransforminganNPinthiswayallowingthehearertogettothereferentinaquickerway,aspeakerhasapracticalpurpose.Thispracticalpurposedoesnotcompletelytransformtheori-ginalutterance,inwaysthatmightgiverisetocomplaintsbytheoriginalspeaker.Furthermore,thisisclearlyacaseinwhichtheNPusedtotransformtheoriginalNPisquiteneutral;andmostimportantly,byusingit,ahearercanhaveaccesstothethoughtentertainedbytheoriginalspeaker(insayingwhateverhesaid),astheNPwhichwasusedasareplacementwilleventually,albeitnotimmediatelynowduringtheindirectreport,butoncethereporthasbeenheardinitsentirety,allowthehearertoreconstructtheitemthatismomentarilymissing.Iproposetouseatechnicaltermforitemssuchas‘thethirditemonthemenu’—thesearesortofpro-forms,butunlikepronominals,whichpointtoobjects,theyarequotativepro-,astheypointtolocutionarysegmentsofthetalk.(Obviouslytheyrefertotypes,ratherthantokens).ThereareotherinterestingtransformationswhichWielanddrawsourattentionto.Considerthefollowing,fromherpaper:5.A:Iwenttothetacostandandboughtasoda.B:Asaidthatshewenttothetacostand.B’sutteranceisclearlyobtainedbyconjunctionelimination.Nowapparently,thisisthecaseofaninnocuous,eveninnocenttransformation.However,therearedoubtsthatthistransformationcanbeeffectedwithoutconsequenceswhencon-junctionisinvolvedinanexplicature,asinthefamousexamplesbyCarston).So,supposethatChurchillsaid(6) ThisisinlinewithWieland(ThePragmaticsofIndirectReportsandSlurring165 6.TheGermansraidedLondonandwefoughtthemback7.TheGermansraidedLondonbutwefoughtthemback.Theremaybeexplicaturesorconventionalimplicatures(seePottsattachedtoacertainconjunction(andasaconsequence,wefoughtthemback).Thuseliminatingaconjunctfromanindirectreportinsuchcases,givesustheimpressionthatpartoftheoriginalmeaningislost.Soif(7)isreportedas(8)8.ChurchillsaidthatwefoughttheGermansbackWehavepartiallyreportedtheutterance.Itisapartialreport.Couldapartialreportbefelicitous?Therearecontextsinwhichitmightandcontextsinwhichitmightnotbefelicitous.So,itisnotstraightforwardthatconjunctioneliminationisanoperationthatcanbeusedalwaysfelicitouslyinindirectreporting.Considernowmodierelimination.Itmightbethoughtthatmodierelimi-nationisaninnocuouslogicaloperationinindirectreports,simplybecauseitissupportedbylogical/semanticalentailments:IfNP[VPADVBVNP],thenitmustbethecasethatNP[VPVNP].So,ifImetabeautifulwomanattheparty,itmustbethecasethatImetawomanattheparty.AndifJohnsays:9.Imetabeautifulwomanatthepartyitcouldbeclaimedthatonecouldreportfelicitously:(9)Johnsaidthathemetawomanattheparty.ButnowsupposethatonadifferentoccasionJohnsaidofthesamewoman,unawarethatshewasthatwoman:10.Thatwomanishorrible.Nowwecouldconjoin(9)with(10),sinceafterallJohnwastalkingaboutthesamewomanandobtain:11.Johnsaidhemetawoman,whowashorrible,attheparty. Acaseforthepotentialinfelicityofpartialindirectreports.MrsSavattawastheheadmistressinahighschoolinItaly.Atameetingwiththeteachers,shesaid‘SupposeIsaythatMrBuccheriisanidiot.’’Ofcourse,shesaidthatinacontext,andhercontextwasprovidedinpartbyherpreviousutterances.Therewasarhetoricalrelationshipbetweenthisutteranceandthepreviousones—shewaspresumablyusingthisutteranceaspartofa(complex)argument.However,theteacherwasoffendedbythisandalonglegalquarrelfollowed.Thesecretaryofthemeetingreportedjustthisutterancebutcompletelyomittedthepreviousutterances,thusmakingitappearasiftheheadmistresswascompletelymad.Grantingthattherewassomethingamissinthislinguisticcontribution,however,therewassomethingcompletelyamissinthereportofherutterance,becauseitwasapartialreport.Cuttinganutteranceandreportingjustpartofitcanmakethingsappearinthewronglight,asthefunctionofanutteranceinasequenceofspeech(inparticulartherhetoricalconnections)seemstohavebeenlostandtheimmediateresultisthatthespeakercanbepresentedasaderangedperson.A.Capone SotheproblemIseeinmodiereliminationisthatitwillallowustoconjoinareportofwhatJohnsaidonsomeoccasionwithareportofwhathesaidonanotheroccasionwhichcontradictedwhathesaidbefore.Thecontradictionpassesunnoticed,ifwesimplysupporttheviewthatmodiereliminationisafeasibleoperationinindirectreporting.IshouldnoticethatWielandaddsalittlelaterthat‘‘SomemodiÞereliminationsmodiÞerintroductionsaltertheoriginalutteranceinapragmaticallyinfelic-itouswayandsomedonot.Thesearegovernedbypragmaticconstraintsonrelevanceandnotsemanticrules.’’Iquiteagreewiththeseconsiderations,evenifIwouldtakesidewithamoregeneralpositioninwhichpartialindirectreportsarealwayslessinformativethanexhaustiveindirectreportsandthustheyrequireacontextthatjustiestheextracognitiveeffortrequiredinthelogicaloperationofthereporting(sincereducinginvolvesanextralogicaloperation).ThismaywellbeinlinewiththegeneralpositionbySperberandWilson()accordingtowhichRelevanceisabalanceofpositiverewards(effects)andcognitiveefforts.AnotherimportantconsiderationbyWielandisthatthelogicaloperationinferencecanbeincorporatedintoindirectreports.Shefelicitouslycallsthiscase:inferentialindirectreport.Anexampleofthispracticemightbethefollowing(alwaysfromWieland12.A:Ididn’tfailanystudents.B:ProfessorAsaidMaryannepassedherexam.Wielandsays:‘‘JustaslongasBknowsthatMaryanneisoneofA’sstudents,thenBcanfelicitouslyreportA’sutteranceinthisway.Thefactthattheinter-substitutabilityofco-referentialtermsandparaphraseonthebasisofinferencearenotonlypossiblebutcommonplacesuggeststhatanindirectreportdoesnotfunctiontoreplicatetheoriginalutterance,anditdoesnotevenfunctiontoconveycontentthatisidenticaltotheoriginalutterance,butratheritspragmaticfunctionistoconveywhateverisrelevantabouttheoriginalutterancetothereporterandaudiencegivennewfactsaboutthereportingcontext’’.Nowthereissomethingweirdaboutthiscase.SupposeProfessorBisuniversallyknownaspassingonlyverygoodstudents(hefailsthosewhoarepassableforotherprofessors).Then,givenwhatisknownaboutProfessorB’sbeliefs,itcouldbeclaimedthatProfessorBsaidthatMaryannewasaverygoodstudent.ThensupposeitiswellknownthatprofessorBbelievesthatallhisgoodstudentswillbecomeUniversityProfessors.ThenitwillbeheldthatProfessorBsaidthatMaryannewillbecomeaUniversityProfessor.ButitisnotclearthatProfessorBsaidallthesethings.Now,whilein TheproblemismoreorlessofthesametypeasnotedbyIgorDouveninconnectionwiththepragmaticsofbelief.Oneshouldnotmakeinferencesthatarelikelytodeceiveone’sfutureself.Now,whileclearlytheinferencesDouvenhasinmindarepragmatic,herewehavealogicaloperationofmodierelimination.Buttheresultissimilarasone’sfutureselfmaybemisledbybeingallowedtomakeotherlogicaloperations(suchasconjunction).Bytheway,Iamnotthinkingthatmodiereliminationandconjunctionarerelatedthings.However,thereisadangerinusingbothoperations,sometimes.ThePragmaticsofIndirectReportsandSlurring167 myownexamples,theproblemmightderivefromidentifyingthewordssaidwiththebeliefsnormallyassociatedwiththosewords,inWieland’scasetheproblemisevenworse,becauseprofessorBissaidtohavesaidsomethingwithoutevenbelievingit,asheneverhadanybeliefsaboutMaryanne(supposetheexaminationwascarriedoutonpapersmarkedbyacode,tomakethemanonymous).Theremightbeinterminablediscussionsonpointssuchasthese—anditisgoodthatthesediscussionsshouldbeundertaken.Myintuitionisthatweareatapointinwhichitisnoteasytodistinguishbetweenlegitimatecasesofindirectreportsandcasesthatareparasiticonthem.Itispossiblethatthismightbealooseusage.Butevenifalooseusage,itisstillanindirectreport,andthusWielanddoeswelltopointoutthatinferencemayplayanelementinreporting.(Giventhatitmayplayaroleinestablishingthetruthofareport,IproposethatwegivegreatconsiderationtoWieland’scase).ThecasejustdiscussedremindsmeofcasesinwhichpragmaticinferentialaugmentationsarebannedbyIgorDouven’s()thePragmaticsofbeliefand,inparticular,byhisEpistemicHygienics.IgorDouvenproposesthatwhenwestoreabelief(intheformofanassertionorasentenceorathought),weavoidstoringittogetherwithinferentialaugmenta-tionswhichmayleaduslatertoremembersomethingwhichwasnotthecase.ThisiscalledEpistemicHygienics.AvividexamplewhichcomesfromthatpaperisthereferencetoGettier’sproblem.SupposeIknowthatp.Then,evenifIcaninfer‘porq’from‘p’,itwillnotdotostoreinmemory‘porq’ifthatisgoingtocreatetroublelater,leadingmetobelievesomethingthatisfalseorunjustied.WemayrememberthatwhatcreateshavocinGettier’sproblemistheshiftfrom‘p’to‘porq’.Keepinginmemory‘porq’whenonebelieves‘p’maypossiblycreatetrouble,asthatmayleadtoanapparentlyjustiedbeliefwhichhappenstobetrue.ThePrinciplewhichwillavoidusmanyproblemsinthefutureisthefollowing:EpistemicHygienics(EH):Donotacceptsentencesthatcouldmisleadyourfutureselves.OtherinterestingexamplesbyDouvenarethefollowing: Arefereemakesanimportantconsiderationandsaysthatvirtuallyitcouldbepossibleforanysentencetomisleadone’sfutureself(soweneedastoryaboutwhatitisthatmakesasentenceacandidatetomislead).Well,considerthesentence:‘Johnwenttothecinema’.Imayutterithavinginmindthereferent‘TheApollo’,butifImemorizethesentencewithoutassociatingthereferent‘TheApollo’totheNP‘thecinema’,Imayendupinthefutureusingthesentencetoreferto‘TheOdeon’.Afterallcouldnotmistakesofthissorthappen?Toavoidtheover-generationofentailments,wewouldprobablyhavetokeepinmindthatweneedtomemorizenotonlyabstractsentences,butsentencesutteredincontext,hencecompletethoughts.Pragmaticintrusionisagoodwaytoavoidtheover-generationofpossibleentailments.Havingdoneso,westillhavetoavoidthoseentailmentswhicharelikelytomisleadourfutureselves.Ofcourse,IshouldnotethatforDouventheproblemisnotanentailment‘perse’,butthefactthatwhenwecommitthingstomemory,wecouldkeeptheentailmentsseparatefromthesentencesthatgeneratedthemandwecouldevenendup,inextremecases,admittedly,forgettingthesentenceswhichgeneratedthoseentailments,whileretainingtheentailmentsinquestion.Andthisisquitebad,becausewewillenduprememberingthingswhicharelikelytomisleadusandhavenegativeconsequencesonaction.A.Capone 13.Peggy’scarisblue;14.Peggy’scarisbluish.Now,itisclearthatifPeggy’scarisblue,itisalsobluish(bluebeingastrongergradationofbluish).However,ifonecommitstomemory‘Peggy’scarisbluish’whenonebelievesthatitisblue,onewillcommittomemoryapieceofinfor-mationwhichmaypossiblymisleadone’sfutureself(Supposethat,inacoupleofdays,thesamepersonisaskedbyMarytosaythecolorofPeggy’scar;hesaysthatitisbluish;thenMaryisnotabletoidentifyPeggy’scarintheofce’sgarage.Sometroublehasensued).Douvencomparesmemorizingorcommittingtomemorytowritingnotes(e.g.Turnoffthegas)whichwillbeofusetoourfutureselves.Ifmemoriesarelikenotes,weshouldavoidwritingnotesthatmisleadourfutureselves.IgorDouven’spaperisofgreatimportancetoepistemologybutalsotoprag-matics.Heshowsthatpragmaticsandepistemologyareintimatelyconnected.WhileIgorDouven’sstorycanbeinterpretedinthelightofmoregeneralprin-ciplesofcognition(amemorythatismisleadingobviouslyisacaseinwhichabelievedassumptionismorecostlythanbenecialintermsofcognitiveeffects;positivecognitiveeffectsbeingthosewhichputmeintouchwithreality,notthosewhichdrivemeawayfromit),Icannotdothisinthispaper.Now,toreturntoWieland’scase.HowcanwedealwithitintermsofthepragmaticsofbeliefbyIgorDouven?Ifweaccept:EpistemicHygienics(EH):Donotacceptsentencesthatcouldmisleadyourfutureselves,itisclearthatcreatingindirectreportsbyresortingtoinferentialstepsthatcanmisleadourfutureselvesisillicit.So,ifonthebasisof(12a),Imaketheindirectreport(in(12b),Iwillbeentitledinthefuturetoexpectthat,onmeetingMaryanne,ProfessorAwillrecognizeherandsay‘Hello’toher.Butthismayneverhappen,ifhepassedheronlybymarkingananonymouspaper.Norshouldweexpectthat,beingreallyimpressedbyherpaper,onseeingapaperbyMaryanneintheJournalofPhilosophy,hewillbeabletoconnectthispapertohispastpositiveexperience(commenting‘‘Oh,this Arefereesaidthat12bdoesnotentitleanyonetoexpectrecognition.And,ofcourseIagreethatProfessorAisnotexpectedtorecognizeMarianne.But,given12b,isnotahearerledtobelievesomehowthatProfessorAhassomeoneinmind(possiblyjustthenameandthethoughtthatapersonwiththatnamehaspassedtheexam)?However,minimal,thisthoughtseemstohavebeenconveyed.Ofcourse,onecouldadjustthecontextsomehow.SupposeeveryoneknowsthatProfessorAdoesnotlookatthenamesonthepapers,hejustcoversthem(oraskshissecretarytodothat)withcoloredsellotape.Thenhemarksthepapersandgivesthemtohissecretarywhoassignsmarkstoindividualstudents.ThishabitissoremarkablethatprofessorAhasbecomefamousforthis.Then,inthis(heavilycontrived)context,theutterance‘ProfessorAsaidthatMariannepassedtheexam’couldbeinterpretedas‘ProfessorApassedMarianne’spaper’.But,evenwithallthiscontextualadjustment,wehaveafeelingthat‘ProfessorApassedMarianne’spaper’and‘ProfessorAsaidthatMariannepassedherexam’areverydifferentutterances,asthelatterimpliessomehowthatProfessorAsaidsomethingofMarianne,thathehadherorhernameinmind,atsomepoint.ThePragmaticsofIndirectReportsandSlurring169 isanotherpaperbyMaryanne).Butallthismakessense,ifweareawarethatthereissomethingstrangeinthepracticeallowingustogofromtherststepof(12)toitssecondstep.ThelastcasediscussedbyWielandthatisofconsiderableinterest(presumablybasedonsomecasesImyselfpointedoutinCapone(),askindlynotedbyWieland)iswhetherweshouldconsidertheliteralorthemetaphorical/indirect/ironiclevelasthebasiclevelofcontentofanindirectreport.Wielandseemstooptfortheviewthatthecontentofanindirectreportshouldbeconstitutedbyinter-pretedandnotbyliteralsegmentsofspeech.Thusanutteranceof(15)15.Maryisalionessshouldbereportedas:16.JohnsaidthatMaryisbrave.However,Inoticethatitisnotcasesofmetaphorsthatareparticularlythorny,becauseherebyreportingtheliterallevelofmeaning,oneallowsthehearerneverthelesstocomputetheindirectornotliterallevelofmeaning.Themostproblematiccasesarethoseofirony,becausethecontextoftheoriginalutteranceismissing(ormaybemissing)andthusthehearercannotmovefromtheliteraltotheironic(orechoic)meaning.Thusthetransitionfrom(17)to(18)isnoteasy:17.Thetalkwasverygood.18.Hesaidthatthetalkwasreallybadandhedidn’tlikeitmuch.ItappearsthatWielandisuncontroversiallymovingtowardsaviewofindirectreportsinwhichthecontentoftheindirectreportisonlytheintendedmeaning,ratherthanthe(possiblyunintended)literalmeaningoftheoriginalutterance.Now,ifsuchaviewisaccepted,indirectreportscouldNOTbeusedasCappelenandLepore()doastestsforliteralmeaningorminimalsemantics.Myimpressionisthatincontextwemustsettlewhetheranindirectreportisaliteraloranon-literalreport.Thereisevidenceinfavorofbothviews.Giventhefactthatitispossibletousedirectquotation,whenwewanttomentionthewordsused,theuseofanindirectreportforthesamepurposewouldultimatelyobtainthesameeffectsofaquotation,butwithgreatprocessingefforts(asonewillultimatelycomparethequotativeconstructionwiththeindirectreport).However,giventhatindirectreportsareoftenmixedwithquotativesegmentsandgiventhatquotationis(asIclaimedinCapone)aradicallypragmaticoperation,itispossibleintheorythatanindirectreportmightoverlapwithaquotativestructure(seealsoBurton-Roberts2006)—whichiswhathappensinthemostthornyexamplesbyCappelenandLepore.Iwillstopthediscussionhere,asIdonotwantittoslideintoadiscussionofSemanticMinimalism.Inthispaper,Iammainlyinterestedin Ofcourse,aliteralmeaningcanbeintended,inwhichcaseIthinkWielandwouldhavetoacceptthatitisthecontentofanindirectreport.Iftheliteralmeaningisnotintended,inthesensethatitissupersededbynon-literalelementswhicharespeaker-meant,thenWielandwillnotacceptitaspartofthecontentoftheindirectreport.A.Capone thepolyphonicstructureofindirectreportsanditisthisaimIhaveinmindthroughout.Theoverlapbetweenquotationandindirectreportsamplyatteststothispolyphonicstructure.AcasenotdiscussedbyWieland,whichcertainlytsthetypologyofexamplessheproposes,isthatofhowtoreportanungrammaticalutterance.Surelyweshouldaskourselveswhethercorrectinganungrammaticaloriginalstatementbyproposinganindirectreportfromwhichtheerrorhasbeenremoved(abidingbythePrincipleofCharity)resultsinalteringdrasticallywhattheoriginalspeakersaidandinsuchawaythats/hewouldnotapproveoftheindirectreport.Andcanindirectreportingwithcorrectionresultinopacity,inthattheindirectreportpurgestheoriginalspeaker’sthoughtofsomethingthatwasessentialtothethought?Inotherwords,wewanttoestablishwhetheropacityonlyrestsontheimpossibilityofintersubstitutingco-referentialNPsorwhetheritalsorestsontheimpossibilityofintersubstitutingcoreferentialsentencesoneofwhichissyntac-ticallyincorrect.Paradoxically,thecaseisnotofimportancefortheilliteratespeaker,whoattacheslittleimportancetogrammarandwhomayevenbeunawareofthesubstitution.However,considerwhathappenswhentheoriginalspeakerisagrammarianandtheoriginalutteranceisreportedthroughanindirectreportwhosegrammarexhibitsanelementwithwhichtheoriginalspeakermaytakeissue.Fidelitytothegrammaroftheoriginalstatementmaywelldependonthecontext.Ifweareinacontextinwhichwehavetoassignmarksdependingonthegrammaticalcorrectnessofwhattheoriginalspeakersaid(supposewearemarkingstudents’papers),evenslightlyimprovingthegrammarofheroriginalsentenceinanindirectreportmaybeconsideredunacceptable.Inthiscasemixedquotationmaybedeemednecessary.IndirectreportsandquotationWhilescholarsaregenerallyadamantthatthereisaclear-cutdistinctionbetweenquotationandindirectreports,thispaperis,infact,blurringthesetwopractices.Andtheresultofblurringthetwopracticestsinwiththeideathatopacityisaphenomenontobefoundbothinquotationsandinindirectreports.Infact,theDavidsoniantreatmentofindirectreportsalsoinvolvedtheblurringofquotationandindirectreports,asthecomplementizer‘that’forDavidsonwasademonstrativepronominalandthethingwhichfollowedthedemonstrativepro-nominalcouldbeeasilyassimilatedtoaquotation(whichexplainedwheretheopacitycamefrom)(SeeBaldwin’simportantconsiderations,whichagreewiththis).Currentscholarstrytokeepapartindirectreportsandquotation—andperhapstheirpracticeiscorrectuptoapoint.However,doingsoinarigidmanner Baldwin(,273)writes:‘‘Davidsonarguesagainstsuchquotationaltheoriesandtherebyimpliesthathisparatactictheoryisnotaquotationalone.Buthetreatsquotationasabbreviatedspellingout,andif,moresensibly,onetreatsquotationmarksasademonstrativedevice,andonetreatsthesymbolswithinthequotationmarksasadisplayofthatwhichisreferredtobythedemonstrative,thenthedifferencebetweenparatacticandquotationaltheoriesbecomesonelargelyofnotation.’’ThePragmaticsofIndirectReportsandSlurring171 wouldpreventusfromunderstandingwhereopacitycomesfrominsuchcases.Evenifwegrantthatindirectreportsarenotalwaysopaque,wesurelymustconcedethattheyarepreponderantlyopaque.Andeveniftheywereonlysome-timesopaque,wewouldstillhavethetroubleofexplainingwheretheopacitycomesfrom.Andofcourse,theopacityofindirectreportscomesfromthefactthatquotationandindirectreportsaresimilartosomeextent,asinvariablyprovenbythepracticeofmixedquotation(inindirectreports).Iwanttobelievethatmixedquotationisnotjustaquirk,somethingthatoccurssometimes,butissomethingthatoccursfrequently,sinceIhaveaccepted(Capone)thatquotationbothintheoralandinthewrittenlanguagecandispensewithquotationmarksandcanresorttopragmaticmarking.Giventhatanysegmentofanindirectdiscoursecouldbemarkedpragmaticallyasbeingmixedquoted,itisclearthattheanalogiesbetweenquotationandindirectreportsarequitestriking.SupposethatweacceptwhatIsaidinCapone()oncancellabilityofexplicatures(namelythatexplicaturesareNOTcancellable).Thenifwehavepragmaticcluesleadingustointerpretalinguisticitemasenvelopedininvertedcommas,thequotationalinterpretationcannotbecancelled,butwillamounttoapragmaticintrusionintotruth-conditionalmeaning.Fromthisitfollowsthatitwillnotdotoreplacethatlinguisticitemwithanothercoextensiveitem,because,otherwise,thespeaker’scommitmenttohavingutteredthatthoughtwillevaporate.Butthisisexactlywhatopacityamountsto.Wecannotreplacealinguisticitemwithacoextensiveone,withoutexpressingadifferentthought.However,wehaveobtainedopacitythroughsomepragmaticmeans.Itisnotexactlysemanticopacitywearewritingabout(tobemoreprecise).ButnowIwanttopursuethislineofreasoningfurther.Considertaboowords,usuallyrelatingtosexualorgans,etc.Scholarshaveinsistedthat,despitethefactthataspeakertakesgreatpainstodistanceherselffromtheuseofatabooword,thankstoquotation,shecannotreallymanagetodoso,andforsomestrangereason,stilltobeexplainedadequately,thetaboowordisassignedtohervoiceaswell.So,considerthefollowingexample:19.Marysaidthat‘(WhereTstandsforataboowordinsertedwithinasententialframeRegardlessoftheframingdeviceofquotation,theresponsibilityforthetaboowordisassignedequallytoMaryandthe(direct)reporter.Now,wewouldexpectthemattertobedifferentinindirectreports.Giventhat‘that’isnotademon-strativepronominal(astheDavidsoniananalysishasit),butonlyacomplemen-tizer,thethat-clauseshouldcomefromtheperspectiveoftheindirectreporter.Thuswecouldexpect,iftherewasarealdifferencebetweendirectquotationandindirectreports,thatonlythereporterwouldberesponsibleforthetaboowordinthefollowingutterancetype:A.Capone 20.MarysaidthatButthisexpectationisnotbornout.WeequallyattributetheTwordtothereporterandtotheoriginalspeaker.Andwepossiblyattributethegaffetotheoriginalspeakertothesameextentastothereporter.Sothingsstandexactlyinthesameway,asfarasobscenitiesandothertaboowordsareconcerned.Now,giventhatwearewillingtogivesimilaranalysesoftheindirectreportsandofthedirectreportsinthesecases,itisclearthatneitherquotationmarksnorthecomplementizercanpreventresponsibilityfrombeingassignedtothereporter.Thetwodifferentfunctionsofthecomplementizerandofquotationmarkswouldleadustoexpectthatquotationmarkscouldbemoreprotectiveforthereporter,butthisisnotthecase.Thepresenceofthecomplementizerinindirectreportswouldleadustoexpectthatthecomplementizercouldbemoreprotectivefortheoriginalspeaker,butthisisnotthecase.Andwhynot?Thetruthisthatifquo-tationandconcealedmixedquotationinindirectreportsaretriggeredandinter-pretedpragmatically,thenwehaveapragmaticmachinerycapableofexplainingwhytheresponsibilityofacertainsegmentoftalkisassignedtotheoriginalspeaker,orbothtotheoriginalspeakerandthereporter.Now,atthispoint,wecangoonusingthemachineryofindirectreportsfordirectquotationsaswell.ParaphrasisPrincipleThethat-clauseembeddedintheverb‘say’isaparaphrasisofwhatYsaidthatmeetsthefollowingconstraint:shouldYhearwhatXsaidhe(Y)hadsaid,he Thereasonforthisisthatthereportercouldhavechosenadifferentword(amoredescriptivestrategy),buthedidnotdoso(thusheisguiltyandresponsibleforthewordsreported).Theoriginalspeakerisattributedthewordsbydefault,becausethestatementisaboutwhathesaid,andthepartialresponsibilityofthereportercannoteradicatethetopicalityoftheindirectreport,asthatisABOUTtheutteranceprofferedbytheoriginalspeaker.Iftheoriginalspeakerdidnotutterthosewords,whyshouldwereportthem,allowingthehearertopossiblyattributethewordstotheoriginalspeaker?Areportthatfocusesonslurringortaboowordsincasetheoriginalspeakerdidnotusethosewordsisinfelicitousandirrelevant,since,donotforget,thepurposeofanindirectreportistofocusonwhattheoriginalspeakersaid,NOTonwhathedidnotsay.Referee2saysthatwecaneasilyimagineareporterindirectlyreportinganoriginalutterancethatcontainsataboowordusingaeuphemisminsteadorpointingoutheavilythatthetaboowordattachesexplicitlytotheoriginalspeakerinsomeway.Thetendencyofreporterstousedescriptiveeuphemismsfortaboowords,e.g.‘TheS-word’for‘Shit’orthe‘F-word’for‘fuck’suggestsweattributethetaboowordtoreportersmorethantotheoriginalspeaker.Ithinktheseconsiderationsareilluminating.Icertainlyagreethatinreportingaspeakerissensitivetocertainrules(aprohibitionagainstusingorevenmentioningtaboowords).Butthefactthatthereporterisdissociatinghimselffromtheuseofthosewordsclearlyisanindicationthatintheworldsharedbythereporterandthereportedspeakerusingthosewordsisprohibited.Andifmerelymentioningthosewordsisprohibited,wecaneasilyimaginethatthereisanevenstrongerprohibitionagainstUSINGthosewords(andpresumablythosewordswereusedandnotonlymentionedbytheoriginalspeaker).SoIagreewiththerefereeonlyuptoapoint,asItakethefactthatthereporterisdistancinghimselffromtheT-wordstorevealajudgmentagainsttheuseofthosewords(intheoriginalutterance)intherstplace.ThePragmaticsofIndirectReportsandSlurring173 wouldnottakeissuewithit,butwouldapproveofitasafairparaphrasisofheroriginalutterance.Paraphrasis/FormPrincipleThethat-clauseembeddedintheverb‘say’isaparaphrasisofwhatYsaid,andmeetsthefollowingconstraints:shouldYhearwhatXsaidhe(Y)hadsaid,hewouldnottakeissuewithit,astocontent,butwouldapproveofitasafairparaphrasisofhisoriginalutterance.Furthermore,hewouldnotobjecttovocal-izingtheassertionmadeoutofthewordsfollowingthecomplementizer‘that’onaccountofitsform/style.Accordingtothesetwoprinciples,wecanexplainwhytheresponsibilityfortheobscenityisassignedtotheoriginalspeakerbothinthecaseofindirectreportsandinthecaseofdirectquotation.Ofcoursethecaseofindirectreportsowseasilyanddirectlyfromtheprinciplesabove.Inthecaseofquotation,weneedaD-tour.Itispragmaticsthatassignstheobscenitytotheoriginalspeaker,bymarkingasegmentasbeingquoted,sincethepointofthequotationistoassignherthosewords.Itfollowsthatifthepragmaticsofquotationisok,theoriginalspeakerwouldapproveoftheutterancethatisbeingattributedtoher.Now,whyisitthatthereporter(boththedirectandtheindirectreporter)isguiltyofobscenity?Whyisitthatthequotationmarksdonotprotecther?Andtheanswerisobvious.Thereportercouldhaveavoidedreportingthelocutionandcouldhavefoundwaysofexpressingthecontentinsuchawaythatthecontentaswelltheobscenitycouldbeperceived,withoutdepictingtheobscenitybutbydescribingit.Inthisway,shewouldhavedissociatedherself(hervoice)fromthevoicingoftheobscenity.Now,intheindirectreport,theoriginalspeakerisguiltyoftheobscenitytoagreaterextentbecauseasegmentoftheindirectreportisbeingmixed-quotedthroughthepragmaticmachinery.Nevertheless,thereporterisresponsiblefortheobscenity—eveniftoasmallerextent—becauseshecouldhavereportedthecontentbydescribingtheobscenityratherthanbydepictingit.Sinceshepreferreddepictingtodescribing,hemustbedeemedguiltyofnotsparingthehearertheembarrassmentofhearingtheobscenity.5.1MichelSeymour()onindirectdiscourseandquotationMyapproachtoindirectreportsisreminiscentoftheideasexpressedbySeymour(),whichisauniqueand,inmyview,importantpaperonthecloseconnectionsbetweenindirectreportsandquotation.Seymourisambivalentbetweenquotationproperandadomesticatedviewofquotationinwhichthequotedsentencedescribesanactofsayinginthedirectsense,buttranslatesitaccordingtotheconceptualschemeofthereporter’stranslationalmanual.Now,ifIamcorrect,Seymourallowsamixtureofelementswhichreectthequotedperson’svoiceandelementswhichreectthereporter’sconceptualtranslationmanual.So,ifthe(English)reporterreports‘ShesaidthatMarywenttoRome’,it Seealsoreferee2’sconsiderations,voicedinfootnote16.A.Capone ispossiblethattheoriginalspeakerused‘Maria’and‘Roma’inherutterance,butthesearetranslatedas‘Mary’and‘Rome’.Thebasicstructureandcontentofthequoteditemisthesame,butcertaininterpolationsweremade.NowIbelievethatthegreatmeritofSeymour’sanalysisliesinmakingusseethatindirectreportsare(normally)ablendofquotationandpureindirectreports.Pureindirectreportsonlyrepresentaschematicsummarymadebythereporterofwhatthereportedspeakersaid.Theblendedreportcouplesthissummarywithaquotationstructure,orcouplesaquotationstructurewithauseofthesamesentenceasifitwasnotquoted.Inmyviewpureindirectreportsdonotexistorrepresentanabstractideal,whilethequotativeapproachtoindirectreports(theapproachaccordingtowhichthereareimplicitquotationmarksprovidedpragmaticallyinsidethethat-clause)iswhatIacceptfully,providedthatweacceptthatspeakersandhearersrelyonapragmaticmachineryallowingthemtodistinguishvoicesintheindirectreport.So,doesSeymourdenitelyabandontheDavidsoniananalysis?Clearlyhedoesnot,sinceheblendsasententialapproach(onethatconsidersthesentencereportedasifinquotationmarks)withaparaphraseapproach(inwhichcontent(regardlessofthewordsused)isofparamountimportance),and,furthermore,heacceptsthatindirectreportsrestonasemantictheorybasedontheconceptoftruth,ofsy-stematicityandrecursiveness.ThefactthatSeymour’s(aswellasmyview)isablendofthepraraphaseandsententialtheoriesdoesnotpreventthetheoryfrombeingbasedontruth,sincebothparaphraseandquotationarestructureswhichcanbeevaluatedtruth-conditionally.Thetheoryisclearlysystematic—beingbasedonanabstractlinguisticsystemthatworksthroughcompositionality.Anditisrecursive,sinceitispossibletoapplythesamesemanticrulesrecursively(JohnsaidthatMarysaidthatRobertsaidthatMyviews,however,divergefromMichelSeymourinatleastanimportantrespect.Myanalysisof‘quotation’doesnotinvolve/presuppose(likehis)aviewbasedonnamesandisclearlybasedonamoredevelopedviewofquotation,saytheonebasedonRecanati(2010)andtheoneIdevelopedinCapone(),whichisradicalinclaimingthatpragmaticsonlyisinvolvedindecidingwhatthethingquotedis(alexeme,aphoneticform,awrittenform,somethingsomebodysaid,etc.).TheotherimportantdifferenceisthatIdonotattachspecialimportancetotheambiguity(whethersemanticorinterpretative,butIassumeitmakessensetoclaimitisinterpretative)betweenasenseof‘Xsaidthat’thatisthatofindirectreportingthecontentofwhatanotherpersonsaidandanothersensewhichamountstoaspecialinterpretationofquotation:inreporting‘Xsaidthatp’,oneisbasicallysayingthatthereisapropositionp,suchthatXsaid‘p’andthecontentof‘p’isgiven(translated)bythesentenceutteredbyX.IgorDouven’spointofview.Reactingtomypaper,IgorDouven(personalcommunication)writesthefollowing:Iwaswonderingwhethertheparaphraseprinciplesdonotgivetoomuchweighttothespeaker’sapproval.Couldn’taspeakerhaveulteriormotivesfordisapprovingsomeThePragmaticsofIndirectReportsandSlurring175 paraphrase,evenifanimpartialthirdpartywouldapproveofit?Perhapsthespeakerregretswhatheorshesaid.Orthespeakerhasafalsememoryaboutwhathe/shesaidandisperfectlyhonest(thoughmistaken,asseenfromanimpartialstandpoint)indisagreeingwiththeparaphrase.Iwasalsowonderingwhetheritwouldbeworthtryingtoadoptinsteadoftheparaphraseprinciplesaprinciplelikethefollowing,whichwouldconnecttothecurrentdebateaboutcontextualisminepistemology:‘Ssaidthatp’istrueiffbyanassertionofthatsentencethehearercomestoknowwhatSsaid.Asvariousepistemologistshaveargued,thestandardsforknowledgemayvarywithcontext.Insomecontexts,notmuchevidenceisneededtogainknowledge;inothercontexts,alotofevidenceisneeded;andofcoursethereareallsortsofintermediatecases.ThismightexplainwhyinsomecontextswethinkalooseparaphraseofwhatsomeonesaidisOK,whileinotherswefeelthatthespeakershouldstayveryclosetotheoriginalspeaker’swording.Ok.Considerthecaseinwhichdisapprovalcomesbecausethespeakerregretshavingsaidwhatshesaid.Wearenotworriedaboutthiscase,becauseifthespeakerishonestenough,hemustacceptthathispriorselfwouldhaveapproved(regardlessofthereservationsbyhiscurrentself)thesentencereportedintheindirectreport,ifitreectedtheformandthethoughtheexpressedintheoriginalutterance.Thefactthataspeakerdisapproveswhathesaidbecauseshehasafalsememoryofwhatshesaiddoesnotworryuseither—aswemayconneourselvestothecaseinwhichtheoriginalspeakerrememberswellwhatshesaid.ConsideringthesecondpartofDouven’scomments,Iamsympathetictowardsacontextualistviewofthematter.PresumablyDouvenconnects‘Xsaidp’withknowledgeofwhatXsaidonthepartofthehearer.Transformingtheissueofindirectreportsintoanepistemicissueamountstobringingincontextualism.AccordingtoContextualists(e.g.KeithDeRose)thetruthofaknowledgeclaimmaydependontheamountofevidencerequiredtoassessit.Insomecontexts,weneedagreateramountofevidenceforthetruthof‘Xknowsthatp’.Inothercontexts,weneedaninferioramountofevidence.Inhighstakecontexts,theevidenceneededissuperiorthantheoneneededinlowstakescontexts.Analogously,inhighstakecontexts,wecouldsaythattheParaphrasePrincipleisadheredtomorestrictlythaninlowstakescontexts.ButthisisnottheonlycaseinwhichweneedtodepartsomehowfromtheParaphraseprinciple.Ihavealreadydiscussedthecaseinwhichaspeakermaybeinterestedinlettingtheheareridentifyareferentandthusmayuseamodeofpresentationofthereferencedistinctfromtheoneusedorapprovablebytheoriginalspeaker.Thissituationisnotlinkedtocontextualisminatheoryofknowledge,asthemodeofpresentationisdifferentregardlessofwhetherweareinahighstakeoralowstakesituation.Presumably,however,IgorDouvenwouldwanttosaythatweareinalowstakesituationsandthisexplainswhythereporterisinclinedtomodifythemodeofpresentationusedbytheoriginalspeaker.NowtherearecogentreasonstobesympathetictoIgorDouven’streatment,evenifamodicationofhiswayofputtingthingsisrequired.Iproposetomodifyhisassertion:A.Capone ‘Ssaidthatp’istrueiffbyanassertionofthatsentencethehearercomestoknowwhatSsaid.Iprefer,instead:anassertionof‘Ssaidthatp’isfelicitousiffbyanassertionofthatsentencethehearercomestoknowwhatSsaid.ShouldtheParaphraseprinciplebeabandonedthen?Perhapsareformulationisneededthatlinksittohighstakescontexts.Alternatively,onecouldoptforthepositionthatassertionsof‘Xsaidthatp’whichdepartfromtheParaphrasePrincipleareparasiticorlooseuses.ThiswouldgivegreaterlegitimacytotheParaphrasePrinciplewhileadmittingthatinsomecontextswemaydepartfromitsomehow.SlurringIftheconsiderationsaboveontaboowordsrelatingtothesexualspherearecorrect,wewouldexpectananalogytoworkbetweentaboowordsingeneralandslurring.Slurring—totakeupideasbyLeporeandAnderson(2013)amountstousingwordsthatarederogatoryandoffendvastcategoriesofpeople(usuallyminorities)suchasJews,Chinese(inUSA),blackpeople,homosexuals,etc.Ourproblemisnotslurringperse,butwhateffectsdoesslurringhaveonquotationandonindirectreports.LeporeandAndersonmainlydealwithindirectreports—whichuseplugssuchastheverb‘say’—butitisclearthatindirectreportsandquotationsworkinaparallelwaywhenslurringisembeddedinthequotationorindirectreportstructure.LeporeandAndersonrejecttheviewthatslurringpersistsinindirectreports(inthatthereporterisbeingassignedresponsibilityfortheslur-ring,ratherthantheoriginalspeaker)becauseofaconventionalimplicature(Williamson2007)orbecauseofapresupposition(seeWilliamsonfordis-cussion).Presuppositionsusuallydonotescapeverbsofsaying,whicharecalled‘plugs’becausetheytendtoblockpresuppositions(seealsoLevinson).Butthenslursbehaveunlikepresuppositionsbecausetheycansurviveembeddinginplugs(eveniftheyoftensurviveembeddinginnegation,if-clauses,etc.likemostpresuppositions).OfcourseLeporeandAndersondonotconsiderapragmaticviewofpresupposition(alongthelinesofSimons),accordingtowhich,atleastinseveralcases,presuppositionsareprojectedthroughconversationalim-plicatures(butthen,inthiscasetheyarenotpresuppositionsbutconversationalimplicatures).WeknowhowLeporeandAndersonwouldreplytoapossibleobjectionbySimons.Ifthepersistenceoftheslurringisduetoaconversationalimplicature,rstofallweshouldaccountfortheimplicaturethroughapragmaticstory.Second,theimplicaturewouldhavetobecancellable,atleastinsomecontexts.Andyetweseethattheimplicaturecanhardlybecancelled,althoughitmaybemitigatedtosomeextentsayinscienticcontextsinwhichthewritermakesitabsolutelyclearthatherpurposeindealingwiththeprohibitedwordisscientic.Ifonlymitigationisobtainedthroughcontextualvariation,thenitishardlythecaseofaconversationalimplicature.Thecaseagainstconventionalimplicatureismorethorny.Asusual,weareinterestedincasesofplugs,suchas:ThePragmaticsofIndirectReportsandSlurring177 21.JohnsaidthatMaryisobstinatebutbrave(however,Idonotpersonallyseeanycontrastbetweenbeingobstinateandbeingbrave).Plugsdonotmaketheconventionalimplicaturedisappearcompletely,asthespeakerof(21)presumablyacceptsthatforsomeoneitmustbetruethatthereisacontrastbetweenbeingobstinateandbeingbrave.However,theydemoteitfromtheepistemiccommitmentsofthespeaker.Thus,LeporeandAndersonarejus-tiedinholdingthatslurringcannotbeamatterofconventionalimplicature.Oneofthepropertiesthatcharacterisesslurringisitspersistencedespiteself-correction.Thisshowsthattheentailmentsoftheslurringwordcannotbeun-said;andinthisrespecttheslurringwordsaredifferentfromotherwords,whichallowself-corrections.Icouldsay‘Maryistall.Oh,sorry,Imeantshort,Igotconfused’.Thesecorrectionsareputupwithintheorallanguage(lessinwrittentexts).Butwithslurringwords,thisisnotthecase,andnoreplacementorcorrectioncanrepairtheslurringwhichwascausedbyusingaslurringword.Consider,infact,thefollowing:22.Lookatwhatthatnegroisdoing—oh,Imeanthatblackgentleman.Arepairliketheonein(22)seemstomakethingsworse,becauseittendstoaddanironicinterpretation.LeporeandAndersondiscussatlengththeword‘Negro’—buttheydonotdiscuss—notevenenpassant—thatinthepasttheword‘Negro’seemedtobeacceptableorusableinAmericanEnglish.Considerforinstancethe‘Ihaveadream’speechbyM.L.King.Iwasmyselfperplexedbysuchuses.Wouldtheycountasusesinvolvingcamaraderieamongblacksoraretheyechoicusestobewrappedininvertedcommas?(Itispossibletooscillatebetweenthetwoviews).Itisdifculttoanswerthisquestioninthecontextofthispaper,asitinvolvesdiachronicconsiderationstoo.However,ifthereisatleastonesuchcontextinwhichtheslurringword,wrappedupbyquotationmarks,doesnotcountasslurring,onecouldoptforaconversationalimplicature.So,theonlycardsonthetablearethefollowing:aconversationalimplicatureandaruleofuse.TheruleofuseviewhasbeenadvocatedbyLeporeandAnderson.Theyclaimthatthereisaprohibitionagainstusingslurringwords.Ofcourse,thisprohibitionworksforthegroupsoutsidethepotentiallyslurredgroups.So,thereisnoprohibitionformembersoftheslurredgroupsagainstusingaslurringword.Thiscouldexplainwellwhythecontextualvariationhassuchpowerfultransformativeeffectsontheslurringpotentialofthework‘negro’or‘queer’.Theconversationalimplicatureviewwouldnolongerbeneeded—orcouldcountasanalternativeviewhavingmoreorlessthesameexplanatorypower.Butwhatwouldtheconversationalimplicatureviewamountto?Withoutgoingintodetails,itwouldhavetosaythat Kennedy(,19)writesabouttheword‘negro’:‘nigger’isanugly,evil,irredeemableword.Hecitessomeoneconsideringtheword‘‘thenuclearbombofracialepithets’’(p.61).A.Capone certainwordsareslursinordinarycontextswherethespeakerspeaksforherself(andnodirectreportorquotativestructureisinvolved),andtheyareslurspre-sumablybecausethereisasocietalruleagainsttheuseofthesewords.Thenitwouldhavetoexplain,onthebasisofthisgeneralprohibition,whyinvertedcommasorindirectreportsdonotrescindtheresponsibilityoftheindirectreporterfromthatoftheoriginalspeakerwhopresumablyisresponsibleforslurring.ButnowtheconversationalimplicatureviewisparasiticontheruleofuseadvocatedbyLeporeandAnderson.So,itwouldbesimplertoholdthattheruleofusebasedonasocietalProhibitionworksbothfortheoriginalspeakerandthereporter.Butifitwasaruleofuse,howcanweexplainthefactthatquotationmarksdonotrescindtheresponsibilityofthereporterfromthatoftheoriginalspeaker?Afterall,itiscommonlyheldthatquotationinvolvesmentioning(atleastinsemantictextbookssuchasLyons).Ifitinvolvesmentioning,whyshouldaruleofusebeapplicabletothereporter?ClearlyindirectreportsdonotposeaseriousthreattoLeporeandAndersonbecauseitmightbeclaimedbytheoriststhatthecomp-lemetizer‘that’neednotworklikeademonstrativepronominalandtheindirectreportercanbeconsideredasonewhousesthewordsinthethat-clause,atleastpartially.WhatIhavesaidbeforeabouttheparallelconsiderationsonquotationandindirectreportsdiscourageusfromthisPyrrhicvictory,socheaplyobtained.Iclaimedthatinindirectreportstoothehearerisfacedwiththethornytaskofseparatingtheoriginalspeaker’sfromthereporter’svoice.Thus,itisnotimpossible,especiallyinthepresenceofappropriateclues,toconsidertheslurringwordsoftheindirectreportasbeingembeddedininvertedcommas(inthiscasetheoriginalspeakerwouldhavetoacceptresponsibilityfortheslurring).Sotheproblemraisedbyquotationisnottrivial.TheruleofuseadvocatedbyLeporeandAndersondoesnotseemtoworkwell,rstofallbecausequotationstruc-turesaswellasindirectreportsintendedashavingaquotativestructuredonotallowustopassthetheorybasedonaruleofuse(aprohibition),astheoriginalspeakercouldbeassignedmajorresponsibilityforutteringtheslurringwords.Second,weneedtonotethatcontrarytoLeporeandAnderson,whoclaimthatindirectreportscontainingslurringwordsassigngreaterresponsibilityfortheslurringtothereporterthantotheoriginalspeaker,Iclaimthat,ifanything,apragmatictheoryliketheonevoicedinCapone(2010a)makesitthecasethattheoriginalspeakerhasresponsibilitytoo.So,weneedapragmaticmachineryliketheoneexpressedin: LeporeandAnderson()writethat‘‘Indirectreportsandotherattitudinalinscriptionsfailtoattributeslurringtowhomevertheyreportsincetheoffenseofthereporter‘‘screensoff’’,sotospeak,theoffenseofwhoeverisbeingreported.Thispositionisinteresting,butneedlesstosay,itwouldneedgreaterjustication.ThePragmaticsofIndirectReportsandSlurring179 ParaphrasisPrincipleThethat-clauseembeddedintheverb‘say’isaparaphrasisofwhatYsaidthatmeetsthefollowingconstraint:shouldYhearwhatXsaidhe(Y)hadsaid,hewouldnottakeissuewithit,butwouldapproveofitasafairparaphrasisofheroriginalutterance.Paraphrasis/FormPrincipleThethat-clauseembeddedintheverb‘say’isaparaphrasisofwhatYsaid,andmeetsthefollowingconstraints:shouldYhearwhatXsaidhe(Y)hadsaid,hewouldnottakeissuewithit,astocontent,butwouldapproveofitasafairparaphrasisofhisoriginalutterance.Furthermore,hewouldnotobjecttovocal-izingtheassertionmadeoutofthewordsfollowingthecomplementizer‘that’onaccountofitsform/style.Now,theseprincipleswouldallowustoassigntheoriginalspeakertheprin-cipalresponsibilityfortheslurring,takingforgrantedorpresupposingLeporeandAnderson’sruleofuse(orprohibition).Thereportingspeaker,givensuchause,isguiltyofnothavingusedanalternativewordoradescription,ratherthanasegmentwhichhasdepictiveproperties.Giventhatshehasnotavoidedtheslur-ringword,whensheobviouslycoulddoso,sheherselfbecomesresponsiblefortheslurring.Butnowwehaveexplainedwhythepragmaticexplanation,despitebeingparasiticonLeporeandAnderson’sruleofuse,doesmoreworkthantheoriginalexplanationbyLeporeandAnderson.Thus,itcouldberecommendedbyModiedOccam’sRazor,becauseevenifLeporeandAnderson’sviewappearstobesimpler,itcannotexplainwhattheconversationalimplicatureview—whichismorecomplex—doesexplain.Objection.Whyshouldthereporterhavetousesomeformofsubstitutionoftheslurringinquestion,ifafterallthedevicesofquotingandofmix-quotinginindirectreportsallowhertoavoidresponsibility,sinceafterallquotingdoesnotamounttousingacertainexpression.Thereplyissimple.Itistruethatthereporterisnotusingtheslurringinquestionand,therefore,cannotbeaccusedofhavingaslurringword.However,indepictingtheslurring,ratherthandescribingitbyasuitabletransformationandbysomedescriptivephrasealludingtothe ThispositionissomewhatreminiscentofSeymour’s()treatmentofindirectreports,inwhichreferencetoatranslationofthereportedsentenceisexplicitlyincorporatedinthesemanticsofindirectreports.Itisdifculttosuggestwhichalternativeismoreneutralthanaslurringterm.Baugh(franklypointsoutthatAmericansnditdifculttondatermwhichisnotinsultingorlessinsultingthat‘negro’.Theterm‘black’usedtobeoffensiveinthepast,butnolongeris.Theterm‘coloured’usedtobeacceptableinthepastbutisnowoffensive.Presumablytheleastconnotativeis‘AfricanAmerican’atermwhichtheReverendJesseJacksonmanagedtointroduceintoAmerican’spubliclife.However,asDuBois()stresses,ifhatredanddespisetargetacertainsocialgroup,thenitwillsurvivedespitethefactthatnewnamesreplacedtheoldonesassociatedwithnegativeconnotations.However,Iwanttopointoutthattheterm‘AfricanAmerican’isdestinedtobesuccessfulbecauseitavoidsallreferencetothecoloroftheskin.So,ideologically,itismuchbetterthanmanyothernames.ItvoicesthedesireNOTtobeclassiedbycolorandarejectionoftheoldstereotypethatpeopleshouldbeclassiedbycolor.A.Capone slurringcharacteroftheoriginalphrase,thereporterissignallingsomecomplicitysincesheisnotdistancingherselffromthetrespasser(theoriginalspeaker).Sinceusingdepictiveelementsinvolvestakingtheshortestrouteinthedescriptionprocess,whenthereisanalternativeroutewhichbyembarkingonatransforma-tioninvolvesgreaterprocessingefforts(andproductionefforts),itisclearthattheavoidanceofgreaterprocessingcostsistakenasasignofcomplicity,whilethemorecostlytransformationistaken(orwouldbetaken)asawayofsignallingthatoneisdistancingoneselffromtheoffensivesegmentoftalk.Wecouldconsider‘complicity’alanguagegame,inwhichtwovoicesblendincasetheysharethesamepointofview.Whileinthenormalcaseinwhichtwospeakershavedifferentpointsofview,theytendtodifferentiatetheirvoices,inthecaseofcomplicitytwovoicesarepresentedasundifferentiated.Indirectreportsareprototypicalcasesinwhichanutterancegivesexpressiontotwovoices,theoriginalspeakerandthereporter.Thus,itgoeswithoutsayingthatanindirectreportshouldpresenttwoslotsincasetheoriginalspeaker’svoiceandthereporter’svoicearedifferentiatedandonlyoneslotincasethetwovoicesblend(beingundifferentiated).Thepresenceofjustoneslot,insteadoftwoslotsclearlyexhibitsthecomplicitybetweenthetwovoices.Ofcourse,readersmayask,howcanwehavetwoorjustoneslotforvoicesinindirectreports?Isthisasemanticorapragmaticmatter?Thenaturalansweristhattheslotsareprovidedpragmaticallyandshouldbeconsiderednotasdistinctsyntacticpositionsbutasportionsoftextwherewecanrevealimplicitquotationmarks.Itistherichstructureofcuesandclueswhichwillpointtowardstwoor,rather,oneslotcapableofexpressingpointofview.2ConclusionInthischapter,Ihavedeepenedmyviewthatindirectreportsarecasesoflanguagegames.Ihavemainlyexploredtheanalogiesbetweenquotationandindirectreports,andIhavemaintainedthatsuchanalogiesallowaparallelpragmatictreatment.Intheend,IhaveconcentratedonslurringandIhaveexplainedwhybothtaboowordsandslurringwordscannotbeembeddedinquotationstructureswithoutlosingtheiranti-socialstatus.Itisclearthatslurringtooinvolvesthetaskofseparatingvoicesandofacceptingtheessentiallypolyphonicstructureofdis-course.Essentiallytheproblem,inourcase,ishowitcomesaboutthatwhensomeonereportsaslurringexpression,thereareinfactatleasttwopeople—andnotjustone—doingtheslurring.Thisisacomplicatedbutinterestingquestion,whichputstothetestboththetheoryofquotationandthatofindirectreports,throwinglightonparallelproblemsaboutpolyphonyandthewayitissupportedbyconversationalimplicature.ThePragmaticsofIndirectReportsandSlurring181 Baldwin,T.(1982).PriorandDavidsononindirectreports.PhilosophicalStudies,42Baugh,J.(1991).Thepoliticizationofchangingtermsofself-referenceamongAmericanslaveAmericanSpeech,66(2),133–146.Burton-Roberts,N.(2006).Cancellationandintention.NewcastleUniversity,SchoolofEnglish,Mn.PublishedintheNewcastleworkingpapersinlinguistics.InE.Romero&B.SoriaExplicitcommunication:RobynCarstonÕspragmatics(2010)(pp.138–155).London:Capone,A.(2008).Beliefreportsandpragmaticintrusion:Thecaseofnullappositives.ofPragmatics,40,1019–1040.Capone,A.(2009).Areexplicaturescancellable?Towardatheoryofthespeaker’sintentionality.InterculturalPragmatics,6(1),55–83.Capone,A.(2010a).Onthesocialpracticeofindirectreports(furtheradvancesinthetheoryofJournalofPragmatics,42,377–391.Capone,A.(2010b).Whatcanmodularityofmindtellusaboutthesemantics/pragmaticsdebate.AustralianJournalofLinguistics,30(4),497–520.Capone,A.(2012).Indirectreportsaslanguagegames.PragmaticsandCognitionCapone,A.(2013).Thepragmaticsofquotation.PragmaticsandSociety(3),259–284Cappelen,H.,&Lepore,E.(2005a).Insensitivesemantics:Adefenceofsemanticminimalismandspeechactpluralism.Oxford:Blackwell.Cappelen,H.,&Lepore,E.(2005b).Varietiesofquotationrevisited.BelgianJournalofPhilosophy,17,51–76.Carston,R.(2002).Thoughtsandutterances.Thepragmaticsofexplicitcommunication.Oxford:Clark,H.H.,&Gerrig,R.J.(1990).Quotationsasdemonstrations.Language,66(4),764–805.Dascal,M.,&Weizman,E.(1987).Contextualexploitationofinterpretationcluesintextunderstanding:Anintegratedmodel.InJ.Verschueren&M.Bertuccelli-Papi(Eds.),pragmaticperspective(pp.31–46).Amsterdam:JohnBenjamins.Dascal,M.,Hintikka,J.,&Lorenz,K.(1996).Gamesinlanguage.InM.Dascal,D.Gerhardus&K.Lorenz(Eds.),(vol.II,pp.1371–1392).MoutonDeGruyter.Davidson,D.(1968).Onsayingthat.Synthese,19,130–146.DeRose,K.(2009).Thecaseforcontextualism.Oxford:OUP.Douven,I.(2010).Thepragmaticsofbelief.JournalofPragmatics,42,35–47.DuBois,W.E.B.(1928).Thename‘Negro’.TheCrisis,35,96–97.Goffman,E.(1974).Frameanalysis.NewYork:Harper&Row.Goffman,E.(1981).Formsoftalk.Philadelphia:UniversityofPennsylvaniaPress.Kennedy,R.(2002).ÔNiggerÕ:Thestrangecareerofatroublesomeword.NewYork:Pantheon.Lepore,E.,Anderson,L.(2013).Slurringwords.(1),25–48.Levinson,S.(1983)..Cambridge:CUP.Lyons,J.(1977).Semantics,1-2.Cambridge:CUP.Potts,C.(2005).Thelogicofconventionalimplicatures.Cambridge:CUP.Recanati,F.(2010).Truth-conditionalpragmatics.Oxford:OxfordUniversityPress.Rumtt,I.(1993).Contentandcontext:Theparatactictheoryrevisitedandrevised.(407),429–454.Saka,P.(1998).Quotationandtheuse-mentiondistinction.Mind,107(425),113–135.Seymour,M.(1994).Indirectdiscourseandquotation.PhilosophicalStudies,74(1),1–38.Simons,M.(2013).MandySimons,ontheconversationalbasisofsomepresuppositions.InA.Capone,F.LoPiparo&M.Carapezza(Eds.),PerspectivesonpragmaticsandphilosophyDordrecht:Springer.A.Capone Sperber,D.,Wilson,D.(1986).Relevancetheory.Oxford:Blackwell.ReprintedwithpostfaceinTannen,D.(1989).Talkingvoices.Repetition,dialogue,andimageryinconversational.Cambridge:CUP.Vološinov,V.N.(1971).Reportedspeech.InL.Matejka&K.Pomorsko(Eds.),ReadingsinRussianpoetics(pp.149–175).Cambridge,Mass.:MITPress.Wieland,N.(2010).Minimalpropositionsandrealworldutterances.PhilosophicalStudies,148Wieland,N.(2013).Indirectreportsandpragmatics.InA.Capone,F.Piparo&M.CarapezzaPerspectivesonpragmaticsandPhilosophy.Dordrecht:Springer.Williamson,T.(2007).Reference,inferenceandthesemanticsofpejoratives.InJ.Almog&P.Leonardi(Eds.),FestschriftforDavidKaplan.Oxford:OUP.Wittgenstein,L.(1953).Philosophicalinvestigations.Oxford:Blackwell.ThePragmaticsofIndirectReportsandSlurring183