The prospective as nonveridical: polarity items, speaker commitment, and projected truth

Transcription

1 101 Theprospectiveasnonveridical:polarityitems,speaker commitment,andprojectedtruth AnastasiaGiannakidou 1 UniversityofChicago 1.Introduction:NPIsinprospectivecontexts Frans Zwarts and I wrote two papers in 1999 on the connection between temporal structure and nonveridicality Aspectual properties of temporal connectives, and Temporal,Aspectualoperatorsandnonveridicality. Inthosepapers,weuncoveredalink betweentemporalstructureandnegativepolarity,inthefactthatnpis(suchasenglish any, and its Dutch and Modern Greek counterparts ook maar iets, kanenas) appear in prospective, i.e. future oriented clauses including will and before clauses, but not in retrospective,i.e.,pastclauses.thatgreatproject,whichwashonoredwithafellowship to me from the Royal Dutch Academy of Sciences, was alas not completed, as life and workedmovedmefromgroningentochicago;butitremaineddeartomyheart,andleft mewithanabidingcuriosityabouttemporalreasoning,truthandpolarity.iworkedout detailsinmymorerecentworkonthesubjunctive,modalsandthefuture,andthispaper isareporttofransofhowfarreachingourinitialideaswere. Iwilltakethefollowingpassagefromour1999apaperasthepointofdeparture. (1) ProspectiveisNonveridical thesis In a linear model of time, retrospective past is deterministic, and in this sense, veridical.prospectivefuture,ontheotherhand,embodiesanotionofprojectedbut not actual truth, hence it is nondeterministic, and thus nonveridical: we do not 1 IamthrilledtopresentthispaperontheoccasionofFrans retirement ahumble,andmost certainlyinadequate,thankyoufortheopportunitieshegaveme,andforhisfaithinmeandin thisprogram.bygivingmeanphdpositioningroningen,fransliterallychangedthecourseofmy life,honoredmewithhisintellectualaswellaspracticalsupport and,aboveall,hisfriendship through the years. I will never forget the warmth that Frans and his wife Sharon have been offeringsogenerouslytomeformorethan20yearsnow,andforthis,iamdeeplygratefulto bothofthem.frans,ihopethatnowthatyouwillberetired,wemaytakeupsomeoftheseinitial questionstogetheragain.ihavebeenwaitingforalongtime!

2 102 know whether the expected events will take place. Hence, in our view, future behaves more like a modality than a real tense. (Giannakidou & Zwarts 1999a: 109). Theintriguingpremiseisthesuggesteddifferencebetweenactualandprojected truth,anditisthisdifferencethatiwilladdresshere.theinitialempiricalmotivationfor (1)wastoexplainwhyNPIsappearwithprospectiveandfutureorientedoperators,but not under past. I illustrate below in English, Dutch, and Greek. The respective future markersarethemodalverbswill,zullen,andthegreekparticletha: (2) a Atthedinnertonight,Nicholaswilleatanything. b *Atthedinnerlastnight,Nicholasateanything. (3) a *Dekinderenvertrokkenzodrazeookmaarietsontdekten. thechildrenleft.3sgassoonastheynpithing discovered.3sg Thechildrenleftassoonastheydiscoveredanything. b Dekinderenzullenvertrekkenzodrazeookmaarietsontdekken. thechildrenwillleave.3plassoonastheynpithingdiscover.3pl Thechildrenwillleaveassoonastheydiscoveranything. (4) WhatwillNicholaseat? a ONicholasthafai kamiamakaronada. thenicholaswilleat.pnp.3sgnpipastadish Nicholaswilleatapastadish. b *ONicholasefagekamiamakaronada. thenicholasate.3sgnpipastadish Nicholasateapastadish. Importantforfutureorientationisnotjusttheparticleorthemodalverb,butalsothe verbalformilabelherepnp,forperfectivenonpast(wecomebacktothisinsection3). WhentheseNPIdatawerefirstdiscoveredinthemid90s,thesharedwisdomaboutNPIs wasthattheyappearinnegativeormeredownwardentailing(de)contexts andfrans s earlier work (most notably, his 1986 dissertation and his 1996 paper) were among the pioneersofthatthesis.however,whenconfrontedwithdataliketheabove,oneneedsto say that the sensitivity of NPIs goes beyond mere polarity (negationaffirmation)or DE. NPIsmustalsobesensitivetotemporalstructure;somehow,thepastisbadforthem,but thefutureisgood.whyisthat?thisdifferencecannotbetiedtonegationandde.

3 103 FransandIsuggestedthatthistemporalsensitivityofNPIsisduetoveridicality:the past is veridical, but FUT p is nonveridical. Therefore the veridicality judgment, i.e. the judgmentaboutthetruthofsentences,isapparentlyrelevantforthelicensingofpolarity items,inatleastthesethreelanguages.sincethen,similardatahavebeenproducedfor EastAsianlanguages(Lee1999,Lin1996),Salishlanguages(kudeterminer,Matthewson 1998),Navajo(FernaldandPerkins2006),Albanian(Xherija 2013)andotherlanguages, see Giannakidou 2011 for an overview). And although any licenses a free choice implicatureinthesecontexts(giannakidou2001,2011),ingreek,dutch,andtheother languages we are talking about, the NPIs appearing in future clauses do not have free choice readings. At the same time, free choice items (FCIs) also appear in the future context,asillustratedbelowforgreek: (5) a ONicholasthafai otidhipote. thenicholaswilleat.pnp.3sgfcithing Nicholaswilleatsomething. b *ONicholasefageotidhipote. thenicholasate.3sgnpifcithing Nicholasatesomething. Giannakidou1998,2001showsthatFCIsarealsopolarityitems,inthesensethattheir distribution is limited to more or less the same contexts as NPIs, as seen also above. Therefore,the overarching generalization seems to be that future clauses are good environments for polarity items generally (NPIs and FCIs) while past positive sentences systematicallyblockthem.thegiannakidouandzwartsthesisoffersanexplanationofthis factbysayingthat(a)thepastisveridicalandthefuturenonveridical,and(b)veridicality andnonveridicalityarekeyfactorsinunderstandingthedistributionofnpis. Iwillproceedwithdiscussingthenotionofveridicalandnonveridicalwhenapplied totemporaldomainsinthenextsections.butfornow,asanotherpieceoffoundational informationonwhythisgeneralizationisimportant,considerthatnpistendtoappearin othercontextsthat,likethefuture,arenonveridical,andhaveprospectiveorientation,i.e. theymakereferencetotimes(andpossiblyevents)afterthespeechtime.thesecontexts are the scope of modal verbs, deontic as well as epistemic, but I will give here mostly deonticexamplesbecauseoftheirclearprospectiveorientation: (6) ONicholasbori namilisi mekanenan/opjondhipotefititi. thenicholasmaysubjtalk.pnp.3sgtonpi/fci student Johnmaytalktoanystudent.

4 104 (7) ONicholasprepeinamilisimekanenanfititi. 2 thenicholasmaysubjtalkpnp.3sgtonpistudent Nicholasmusttalktoanystudent. (8) Faekanena/opjodhipoteglyko! Eatany cookie! Deonticmodalitiesandtheimperativehaveprospectiveorientation,andtheyseemtobe goodfornpisandfcis.hencethegeneralizationis,correctly,aboutprospectivityandnot, strictlyspeaking,thefuture.also,igivebelowexamplesfromearlymoderndutch(from Hoeksema2010)illustratingtheNPIenig,whichjustlikeGreeklacksFCIuse: (9) Modal Menmoesttochwel enigeaanwijzinghebben. oneoughtprtprtsomecluehave Oneshouldhavesomeclue. (10)Subjunctive Enwiegeensteenenkanaandragenstorte[..]eenigegiftindeofferbus. Andwhonostonescantocarrythrow.SUBJsomegiftintheoffertorybox Andwhocannotcarrybricks,shoulddonatesomegiftintheoffertorybox Hereagainweobserveprospectiveorientationwithadeonticmodalandasubjunctive. Besides the future, Greek possesses a number of other prospective particles, and whichalsotypicallylicensenpis:thesubjunctiveparticlena whichwesawalreadyasa complementtothemodalverbs,butcanalsooccurinmainclauses(giannakidou2009), theoptativeas,theconditionalparticlean,andthetemporalconnectiveprin before.as shownhere,npisappearinallthesecases: 2 FCIsarenotplausiblewiththeuniversalmodalcontextsbecausetheytriggeranexhaustivityinferencethat seemstobeatoddswiththeuniversalquantifier(giannakidouandquer2013);any,naturally,alsoisodd, becauseofitsfreechoiceimplicature.iwilllargelyshyawayfromtheinterpretationofnpisandfcisinthis paperbecausethefocusisontheprospectivenessquestion.

5 105 (11)WhatshouldNicholasbringfordinner? a Na(subj.) ferikanenaglyko. Itwillbeagoodideatobringadessert. b As(optative) ferikanenaglyko. Lethimbringadessert. These are neutral, soft invitations or suggestions for Nicholas to bring dessert, some dessertorother,thespeakerremainsagnosticastowhattheactualdessertwillbe.(fci interpretationisagainodd,henceanyisimpossible). (12)Prinerthi{kanenas/opjiosdhipote}, prepinakatharisoumetodomatio. beforecome.pnp.3sgnpi/fciperson, mustsubjclean.pnp.3pltheroom Beforeanyonecomes,wemustcleanuptheroom. (13)Anerthi{kanenas/opjiosdhipote} prepinakatharisoumetodomatio. beforecome.pnp.3sgnpi/fciperson, mustsubjclean.pnp.3pltheroom Beforeanyguescomes,wemustcleanuptheroom. Notice again the use of the PNP, crucial to the prospective orientation.it becomes obviousthatwehavearealgeneralizationhere:prospectiveorientationcreatesagood licensingenvironmentforpolarityitemsofvariouskinds,and,asgiannakidouandzwarts suggested,thereasonforthisisthatprospectivecontextsarenonveridical. Now,puttingtogetherthefactthatthesepolarityitemsalsoappearinnegativeand downwardentailingcontexts,wecansummarizethewholetheoryofnpisinthediagram below.thisschemasaysthatallpolarityitemlicensersarenonveridical.theinnermost circles are the negative domain, what Zwarts calls strong negations; DE is minimal negation.crucially,negativecontextsarepropersubsetsofthenonveridical.buttheset ofnonveridicaloperatorsincludesnonnegativefunctionstoo.theprospectiveoperators we are talking about belong to this class and as we saw, they are overwhelming licensersofpolarityitemscrosslinguistically.nonveridicalitythusallowsustoexplainwhy NPIs appear with seemingly unrelated licensers e.g. negation, the future and other modalities withoutclaimingthatnegationsomehowbecomesmodal.

6 106 (14)TheGiannakidouZwartsTheoryofpolarityitems Havingclarifiedthesefoundationalpremises, wecannowproceedtoaddressthetwo important points in the Prospectiveness and Nonveridicality thesis (1): (a) the past is deterministicandthereforeveridical,and(b)thatprospectiveexpressions(includingthe future) embody a notion of projected but not actual truth and are therefore non nonveridical.iwillillustratebystudyingthegreekparticles. 2.(Non)Veridicalityjudgment:subjunctive,optativeandcommitmentweakening Considerfirstbare,i.e.unmodalized,nonnegated,sentencesinthesimplepastorpresent (whichisthepresentprogressiveinenglish): (15)a Nicholasbroughtdessert. b Nicholasiswashingthedishes. Thesearesentencesaboutactualeventsthathappenedinthepastorarehappeningright now; inthepresent,thespeakeractuallymaybewitnessingtheeventunfolding.so,when one judges the sentences describing these facts, one judges them as true.this truth judgment is a veridicality judgment, and the speaker judges the past and the present progressivetobeveridical. Theveridicalassertionisthenaddedtothecommonground, andthediscourseproceedsbyonlylookingatworldswherenicholasbroughtdessertoris washingthedishes. Considernowthesentencesmentionedearlierwiththesubjunctiveandtheoptative:

7 107 (16)WhatshouldNicholasbringfordinner? a Na(subj.) feri kanenaglyko. Itwillbeagoodidea tobring a dessert. b As(optative) ferikanenaglyko. Lethimbringadessert. Bothsentencesareinvitationstobringadessert(p),butdonotentailthatpistrue,or that it will be true. In both cases, there is also nonassertive illocutionary force; but neitherforcenorthedesirethatpentailpnoworinafuturetime.withthesesentences thespeakerexpressesadesirethattherebeaneventdescribedbyp,atsomepointinthe future, but has no commitment to the event happening; a desire does not commit the speaker to p, or add p to the common ground. (The same, by the way, is true of imperatives.).thesesentencesarenonveridical. So, when we talk about truth judgments, we talk about them by appealing to speakers commitmenttothetruthofasentence.prospectiveorientedparticlesdonot conveycommitmenttothetruth,butwhenaspeakerassertsapastorpresentsentence, they are referring to facts, and by asserting them they are committed to them. In this context,veridicality beingajudgmentaboutthetruthofasentence itappliesintwo ways. First,objectively: if an expression entailsthe truth of its argument, it is veridical. ThiswastheoriginalinceptionofZwarts(1995),andGiannakidou(1994,1997): (17)Def1:Objectiveveridicality ApropositionalfunctionFisveridicalifFentails;otherwise,Fisnonveridical. Wecanthinkofthisas objective veridicality.itisveryclosetofactuality asentence underfistrueifitreferstoafact.inthissense,veridicalityismoreorlessequivalentto thetraditionalrealis,andanysentencethatdoesnotrefertoafactisnonveridical.pastor presentsentencesareveridicalinthisobjectiveway,hencegiannakidouandzwarts claim thatpastisdeterministic.theprospectivedomainsaren tveridical.inabranchingtime model,futurebranchesshareahistory,apastupthetimewheretheybranchout,and thisdividesthemodelintoadeterministic(past/now)spacewhichisobjectivelyveridical, and nondeterministic, prospective space, which branches and allows for a number of possibilities.icomebacktothisinsection3. AsIhaveshowninearlierwork(Giannakidou1994,1998,1999,2009),thisnotionof objectiveveridicalitycanaffordsubstantialresultsinnaturallanguage,butitneedstobe enriched with an epistemic component when we consider mood choice, NPIs, and modality.truthjudgmentisdonebyindividuals(inmainclausesthespeaker),andwhen

8 108 the speaker assesses truth, she does so based on her beliefs and knowledge. This is subjectiveveridicality.subjectiveveridicalityisshown,inrecentwork,tobeimportantin extracting truth assessment from texts (de Marneffe et al. 2012), and in Trnavac and Taboada (2012) and Giannakidou (2013, to appear), a correlation is posited between nonveridicality and evaluation that further supports the subjective nature of the veridicalityjudgment. Tocapturetheideathattruthisassessedrelativetoanindividual,Idefined models ofevaluation(giannakidou1998,1998,1999).thesemodelsaresetsofworlds,relativeto anindividuali,theepistemicagent,correspondingtowhattheagentbelievesorknows. Wecanthinkofthemasinformationstates,ormodalbases,associatedwithindividuals: (18) Epistemicmodelofanindividuali (Giannakidou1999:(45)) AnepistemicmodelM(i)Misasetofworldsassociatedwithanindividuali representingworldscompatiblewithwhatibelievesorknows. (19) Truthinanepistemicmodel ApropositionpistrueinanepistemicmodelM(i)iffM(i)p: w[wm (i)ww'.p(w')] Inmainpastsentences,themodelrepresentstheepistemicspaceofthespeaker,andit includesworldscompatiblewithwhatshebelieves.apropositionpofamainassertion willbeevaluatedwithrespecttotheepistemicspaceofthespeaker: (20) a Johnwontherace. b [[Johnwontherace]]=1iffw:wM(speaker)ww'. Johnwontheraceinw' If the speaker truthfully asserts the sentence John won the race, she must believe that Johnwontherace.Believingthepropositionmeansthatallworldsinherepistemicspace arejohnwontheraceworlds.hence:m(s)p.thisexplainsmoore sparadox,i.e.why #John won the race, but I don t believe he did is odd. In the main unmodalized past assertionthespeakerisfullycommitted,withinherbelief/knowledgespace,tothetruth ofthepropositionsheexpresseswiththesentencesheutters.veridicalityisthisstateof fullcommitment:

9 109 (21) Def.2.Veridicality,nonveridicality,antiveridicalitywrtindividuals i. AfunctionFisveridicaliffFconveysanepistemicstateM(i),relativeto anindividuali,suchthat:allworldsinm(i)areworlds,i.e.m(i). ii. Ifi.doesn thold,fisnonveridical. iii. IfallworldsinM(i)arenonworlds,thenFantiveridical. Underthisdefinitiontoo,optativeandsubjunctiveparticlesarenonveridicalbecauseofii. Inbothdefinitions,nonveridicalityistheabsenceofveridicality.Negation,ontheother hand, is the opposite of full commitment, antiveridicality: the speaker is fully not committedtop.anantiveridicaloperatorisalsononveridical,sincei.isvalidforittoo. (22) AriadnedidnotkissNicholas. (23) AriadneisnotkissingNicholas. (24) [NOT(AriadnekissNicholas)]istrueiff:allworldsinM(s)areworldswhere Ariadnedid/doesnotkissNicholas So,bothversionsofveridicality,objectiveandsubjective(Defs1andDef.2)capturethe sensitivityofnpistobothnegationandthe,seeminglyunexpectedfromtheperspective of negation and DE, prospective operators. Possibility modals like might, may, regardlessoforientation arealsononveridicalandlicensenpisaswesaw. (25) a Nicholasmightbringadessert. b Nicholasmaybringadessert. (26) a Nicholasmighthavebroughtadessert. b Nicholasmayhavebroughtadessert. Pastpossibilitiescontrastwiththesimplepastassertionsinthatthespeakerisnolonger committed to the truth of p; p is regarded merely as possibility. Adding the possibility modalweakensthecommitmentandcreatesnonveridicality.asdemarneffeetal.(2012: 102) put it, declaratives like Ariadne left conveys firm speaker commitment, whereas qualified variants with modal verbs or embedded sentences imbue the sentence with uncertainty. Giannakidou (2013) calls assertions with possibility modals inquisitive assertions,andonecanthinkofthesubjunctive,optative,andpossibilitymodalsasaclass ofcommitmentweakeningoperatorswhosefunctionistocreateanonveridicalspace:

10 110 (27) Commitmentweakeningandpossibilityoperators i. Commitmentweakeningisthecreationofanonveridicalepistemicspace. ii. Possibilityoperators(subjunctive,optative,weakmodals)havethefunction ofweakeninganindividual scommitmenttoaproposition. The epistemic effect of commitment weakening can be viewed now as conveying a partitionedspacecontainingbothpandnonpworlds. (28) Def.3:Veridicality,nonveridicality,antiveridicalityandcommitment Anepistemicspace(asetofworlds)M(i)relativetoanindividuali: i. IsveridicalwithrespecttoapropositionpjustincaseallworldsinM(i)are pworlds.(fullcommitment) ii. IfthereisatleastoneworldwinM(i)thatisanotpworld,M(i)is nonveridical. (Weakenedcommitment) iii. IfallworldsinM(i)arenotpworlds,M(i)isantiveridical. (Countercommitment) Thenonveridicalepistemicspaceinthisdefinitioncontainsatleastonenotpworld.As previously,intheextremecasewhereallworldsarenotp,wehaveanantiveridicalspace, that I call here countercommitment. Countercommitment as well as weakened commitmentarenoncommitmenttop.onlyinveridicalspacesisicommitted(i.e.fully committed)top. Noticethatcommitmentweakeningisirrespectiveofthepastprospectivecontrast, since modals, as well as na and as, can be used with past tenses and weaken commitment: (29)a Isosna ixe gripi. PerhapsSUBJ had.2sgflu Maybehehadtheflu. b. Malon *na/thaixegripi. probablysubj/futhad.2sgflu Probablyhehadtheflu. (30) As toumilouse. OPThe.gen.talked.3sg Ifonlyhehadtalkedtohim.

11 Thesubjunctivenaisusedtogetherwiththepossibilityadverb isos maybe,butnotwith thestrongeradverbmalon probably.thelatteriscompatiblewiththefuture,suggesting thatthefutureitselfhasstrongerforce(seediscussioninsection3).thecooccurrenceof nawithmaybecanbeviewedasacaseofmodalconcord,anditsuggeststhesubjunctive being akin to a possibility modal (see also Giannakidou 2014). The optative has counterfactualforce,afactobservedforoptativesingeneral(grosz2011),soitbehaves antiveridicallywiththepast thoughnonveridicallywiththenonpast,aswesawearlier. So,tosumup:possibilityexpressions,includingthesubjunctiveandtheoptative,are commitmentweakeners:theycreateanonveridicalepistemicspace.withsomeofthem, thespeakerisinwhatgiannakidou2013calls nonveridicalequilibrium. (31) Nonveridicalequilibrium(Giannakidou2013:14) AnepistemicspaceM(i)isinnonveridicalequilibriumiff: M(i)ispartitionedintopandnotp,andthereisnobiastowardspornotp. The equilibrium means that the speaker considers p and not p as equally good possibilities,sosentenceswithequilibriumastypicallyneutral.butasiarguedinthe2013 paper, the equilibrium is easily disturbed. This happens with the optative and other counterfactualoperatorsforinstance,whichinthepastcreatebiastowardnotp.when weusefutureexpressionsandnecessityepistemicmodals,asweseenext,positivebiasis created (Giannakidou and Mari 2013, 2014), and it brings with it the projected truth mentioned by Giannakidou and Zwarts. Projected truth is responsible for the stronger flavorthatuniversalmodalsandthefuturehave. 3.Thefuture:nonveridicality,partialcommitment,andprojectedtruth ThestartingpointofdiscussionsoffuturesentencesisoftenAristotle sveryfamoussea battleexample(,deinterpretatione9). (32) a. Therewillbeaseabattletomorrow. b. Therewillnotbeaseabattletomorrow. A major goal of Aristotle in is to discuss the thesis that, of every contradiction,onemembermustbetrueandtheotherfalse(the LawoftheExcluded Middle ).Regardingthefuturesentences,Aristotleacknowledgesthatthetruthorfalsity ofeachsentencewill,intime,befullydeterminedbyhowthingswillturnout:therewill eitherbeornotbeaseabattle.aristotlealsoacknowledgesthat,atpresent(i.e.,atthe 111

12 112 speechtime),itisnotknown,inthesensethatapastsentencecanbeknown,thatthere willbeaseabattletomorrow.so,thefuturesentencesatnareobjectivelynonveridical, i.e.thefutureisnondeterministic. Thefutureisalsosubjectivelynonveridical,asitiscompatiblewithanepistemicstate thatincludes,forafuturetime,bothp(therewillbeaseabattle)andnotp(therewillnot be a sea battle). This is the position Alda Mari and I defended in recent publications (Giannakidou2012,GiannakidouandMari2013,2014),andIwilltrytospellitoutinthe rest of this section.in FUT p, the epistemic state M(s) of the speaker at present is nonveridical and allows, for afuture time, both p and not p; but unlike with possibility modals, there is a subspace within M(s) that fully supports p. This creates partial commitmentofthespeakertop,andmakesthestatementwiththefuturestrongerthan themerepossibilitystatements. 3.1.Thefuture,thepresentandthenonpast From the perspective of nonveridicality, future statements are pretty much like statementswithnecessitymodalslikemust: (33)a. ForallIknow,theremustbeaseabattletomorrow. b. (Inorderforthisconflicttoend),theremustbeaseabattletomorrow. Modal verbs, in epistemic and mostly deontic uses, come with the same kind of indeterminacy about the prejacent proposition p, and are therefore also nonveridical. BroekhuisandVerkuyl(2013)pointouttheparallelismofthefutureauxiliarywithother modalverbsindutch,e.g.,intheirexample(9),givenherebelow: (34)a. Elsazaldanwandelen. a'. Elsamoetdanwandelen. Elsawillthenwalk Elsamustthenwalk Elsawillwalkthen. Elsamustwalkthen. b. Ikzaljebellen. b'. Ikgajebellen. Iwillyoucall Igoyoucall Iwillcallyou Iamgoingtocallyou. c. Wezullenmorgenthuiszijn. c'. Wekunnenmorgenthuiszijn. Wewilltomorrowhomebe Wemaytomorrowhomebe Wewillbehometomorrow. Wemaybehometomorrow

13 113 Futureorientationiscommontozullenandnonzullenmodalities,asweseeintheprime examples. Broekhuis and Verkuyl claim that, despite the future orientation, it doesn t makesensetosaythatthemodalverbsmoeten must,kunnen may,andthereforealso zullenarefuturetenses.futureorientationisgoingtobeattributedtothepresenttense on the auxiliaries, which in their account carries prospective meaning. The modal verb itself(i.e.,minusthepresent)isthusanindicatorofpuremodality. LetusnowreturntotheobservationthatprospectiveorientationinGreekcomes withnonpast. (35)a. Asfiji ojanis. asleave.pnp.3sgthejohn LetJohngo. b. Nafiji ojanis. naleave.pnp.3sgthejohn LetJohngo. c. Thafiji ojanis. tha leave.pnp.3sg thejohn Johnwillleave. TheimperfectivenonpastistheactualsemanticpresentinGreek(Giannakidou2009,in press).thepnpisadefectiveperfectivenonpastthatcan tfunctionasapresent,because ofitsaspectuallimitationofperfectivity.so,greekactuallypossessesasemanticpresent (morphologically: imperfective nonpast) and a semantic nonpast. In the occurrences of PNPplusparticlewehaveprospectiveorientation,sofuturereference,justlikeinDutch, is not a privilege of the future particle tha. Giannakidou (2009) argues that the Greek perfective nonpast denotes a prospective interval but unlike the prospective present interval whose left boundary is n, the left boundary of the nonpast is undefined. It containsadependentvariablet. (36) [[perfectivenonpast]]=ptp(t,) Adependentvariablecannotremainfree,butmustbevaluedbysomehighervalue.This ideaisinspiredbyabusch's(2004)analysisofwollasasubstitutionoperator.according to Abusch, "In the substitution operator, t is a bound variable that corresponds to the tense argument of will [which is n, coming from an implied higher PRES; clarification

14 114 mine].foratopleveloccurrenceofwill,theeffectistosubstitute(n,)forn"(abusch 2004:39). The Greek perfective nonpast then is a WOLL, but unlike will where n is triggered by default (Abusch 2004: 48) the Greek perfective nonpast does not trigger PRES; so it becomes necessary to have an overt exponent of n in the structure, otherwise the structureisillicit: (37) *TP:te[write(j,e)e(t,) grapsiojanis Johnwrite.PNP T 0 :nonpast AspectP:te[write(j,e)et] PtP(t,) Asp 0 :PFT= Pte[P(e)et] VP:twrite(j,t) t v Theinterval(t,)isillformed,becausetisunvalued.Theparticlessavethestructureby providingn.ifweadd,forexample,thefuturetha,tcannowbeidentifiedwithn: (38)[[tha]]=n (39)[[tha]](TP(19))=te[write(j,e)e(t,)](n)=e[write(j,e)e(n,)] The event of John s writing will now be located at the interval that starts at n and stretchesthroughinfinity.thisexplainsthepossibilityoffutureforthepnp.theanalysis saysthattheadditionalstructureprovidedbytheparticlesgivesalocusfortriggeringof thespeechtime. 3.2.EpistemicfutureandMUST Itisacommonobservationthatfuturemorphemesexhibitpurelyepistemicreadings,and ingiannakidouandmariwecallthis epistemicfuture.herearesomewellknown examples(seealsoenç(1996),andtsangalidis(1998)): (40)a. TheFrench llbeonholidaythisweek. (Palmer1987) b. Nodoubt,you llrememberjohn. c.edwilllayinbedalldayreadingtrashynovels. (Huddleston1995) d. Oilwillfloatonthewater. (Haegeman1993)

15 115 AccordingtoPalmer,willhereexpressesconclusionofreasoningthatthespeakerdoes, (see also Broekhuis and Verkuyl (2013)). What kind of reasoning? Obviously, reasoning that relieson what the speakerknows or has evidence for. In concluding with will, the prejacentissupported.however,andwecomebacktothis,thespeaker sconfidenceis notashighasitwouldhavebeenhadshechosenanonmodalizedform,e.g.,thefrench are on holiday this week. The nonmodalized sentence is veridical, and therefore expressesfullcommitment. In Greek, as we see below, the future particle in the nonpredictive use can be followedbyapresentorapastform.pasttenseiscompatiblewithpastadverbials: (41)a IAriadne tha pezi tora. (nonpredictive) theariadnefutplay.imp.nonpast.3sgnow Ariadnemustbeplayingnow. (42)a. IAriadne thakimithikeprin2ores. (nonpredictive) theariadnefutsleep.pp.3sg beforetwohours Ariadnemusthaveslepttwohoursago. b. IAriadnethamilise xthes. (nonpredictive) theariadnefuttalk.pp.3sgyesterday Ariadnemusthavespokenyesterday. Noneofthesecasesis future inthesenseofmakingreferencetoaneventthatfollowsn. Rather,ascanbeseeninthetranslations,weusemust.GiannakidouandMaricallthisthe epistemic future. Epistemic future and MUST convey strong support for p, but no full commitment,noveridicality.thisisshownbythetestbelow(giannakidou&mari2013): (43)a. IAriadneitanarosti #aladhenimekeendelossigouri. Ariadnewassick #butiamnotentirelysure. b. IAriadnethaitanarosti aladhenimekeendelossigouri. Ariadnewill/musthavebeensick butiamnotentirelysure. Tha,FUT,andmodalizationingeneralconveynonveridicalmodalspaces,andstillallow for notp worlds. Hence, all modalized sentences, even with necessity modals, are weaker than nonmodalized assertions (pace von Fintel & Gilles 2010; for critical discussionseealsolassiter2013).futurethaintheepistemicuseisthusakintomust,i.e. auniversalquantifier(recallitscompatibilitywithmalon probably ),andthemodalbase (f(w))isepistemic.

16 116 (34) Foranyworldw,andconversationalbackgroundsf,g: (Giannakidou2012) [[prepi/tha/must]]w,f,g=q<st>.w Best g(w) (f(w)):q(w )=1; wherebestg(w)(x)selectsthemostidealworldsfromx,giventheorderinggiven byg(w) Crucially, only in the Best (see Portner 2009 and discussion therein) worlds is p true, therefore the universal modal is nonveridical, since the modal base still contains not p worlds.intermsoftruthconditions,then,epistemicfutandmustareequivalent,andin Greektheycancooccurwithmodalconcord:e.g.Thaprepinaexigripi Shemusthave theflu.thaprepiisequivalenttoprepi must (Giannakidou2012). So,FUTandmustcontainnonveridicalspaces,butatthesametime,theycontaina privileged(best)innerdomainofcommitment,thedomaintheyquantifierover.forthis reason,universal epistemic modals are stronger than the mere possibility statements which express no commitment, as we discussed earlier. The effect of bias will be seen morestronglyinthepredictiveusethaticonsidernext.beforeimoveontoshowthis,i wanted to note a similarity, in exactly this respect of combining nonveridicality with partialcommitment,betweenfut/mustandthesocalled modal discourseparticlesin Germanicsuchaswohl(Zimmermann2011): (44) a. MaxistwohlaufSee. (examplefromzimmermann2011) Maxisprtatsea Maxmustbeatsea. b. (ForallIknow),Maxwillbeatsea. Zimmermann claims that with wohl, the epistemic commitment of the speaker is weakened compared to the plain sentence, while also conveying a confidence that the propositionislikelytohold.still,though,thisisonlypartialcommitment:incaseiknowp, Icannotusetheparticle. (45) A:Ican tseehein. EristwohlaufSee. Hemaybeatsea. (46) A:Iknowforsure: #HeinistwohlaufSee. Heinisprtatsea

17 117 Likewise,inGreek,ifIknowforsure,Ican tutterthainestithalasa Hemustbeatsea. Inusingtheepistemicfutureandwohl,Iaminanonveridicalstateofknowledgethatstill allowsnotp,andifthecontextforcesfullknowledge(asisthecasee.g.ofdirectevidence) theresultisodd.theseparticles,andmust,giannakidouandmari2014bargue,depend onpartialknowledge.crucially,dutchzullenseemstofunctionsimilarly,asweseebelow. TheDutchcounterpartofwohl,welcanalsocombinewithzullen(andwecanthinkofthis asmodalconcordtoo,onaparwithcooccurrencesofthaandprepimentionedearlier): (47)A: Ican tseehein. Heinzal(wel)opzeezijn. Hemaybeatsea. (48)A: Iknowforsure: #Heinzalopzeezijn. Hemaybeatsea. (49)A: Heissogrumpy! Hijzalwelslechtgeslapenhebben! Hemusthavesleptreallybad! Zullenappearstoepistemicallyweakenstatementsaboutthepasttoo,justliketha,wohl. So,FUT,wohl,wel,andMUSTseemtoformanaturalclassofnonveridical,positivebias operators: (50)Positivebiasoperators,partialsupportofpropositions AnonveridicallinguisticexpressionFcreatespositivebiasif: (a) FconveysapartitionedepistemicspaceM(i)intopandnotp;(nonveridicality), and (b) FpartiallysupportspviauniversalquantificationoverasubsetofworldsinM(i), selectedbytheorderingsource. Becausetheorderingsourceisthe ideal (i.e.thegoldenstandardinthecontext),partial supportofpinthebestworldsmakesthatsetprivilegedandcreatesbias.inthecaseof partial support, we talk about projected truth within the set supporting p. All universal modal operators thus have this dual nature of allowing both p, not p, while privileging optionp.andforthisreasontheygivetheimpressionthattheyare strong. Imoveonnow,finally,toshowhowtheseideasexplainthemodalityofprediction, basedontheanalysisofgiannakidou&mari(2013b,2014a).

18 Thefuture:projectedtruthandbiasinreasonableworlds Predictionhasanepistemicbasis(seealsoBroekhuisandVerkuyl2013).InpredictingFUT patthespeechtimen,thespeakerisconfidentthatthepropositionwillbemadetrueat somepointfollowingn.thisconfidencereliesonknowledgeatthepresent,butitisnot itselfknowledgeofthefutureevent,sincesucheventshavenotyetmaterialized. The future can be thought of in terms of branching times (Thomason 1984, Kaufmann et al. 2005), and Giannakidou and Mari argue that the speaker uses her knowledgeasadomainrestriction,i.e.,touniversallyquantifyoveronlyasubsetofthese metaphysicalalternatives.speakersprojecttheirknowledgetocleanupthemetaphysical branches, to carve them out into reasonable and unreasonable ones (Mari 2013). Reasonable futures are those where everything proceeds as expected and nothing peculiar happens (see Mari 2013 for extended discussion of this). The knowledge a speakerhasatnallowshertoknowwhichbranchesarereasonable,andwhichnot.inthe figure below, the dotted line to w 3 indicates unreasonable future (Giannakidou & Mari 2014:58): NowconsiderthesentenceJohnwillbehereat5.

19 119 (51) Theeventunfoldsintheactualworldw 0,whichhasareasonabledevelopmentw 1.The sentence says that if the course of events remains reasonable, the speaker is highly confidentthatjohnwillbehereat5.however,itisstillpossiblethatanaccidenthappens. Inthiscase,theactualworldtocomebecomesanunreasonableone: When we predict, we normally don t consider as relevant such possibilities (only if we knowthatjohnispronetoaccidents,dotheybecomerelevant,inwhichcasetheforceof our prediction John will be here at 5 becomes weaker). As a universal quantifier over reasonableworldsthen,futexpressessupportforpwithinthereasonableworlds,thus projectedtruth ontothatset:

20 120 (52) TruthconditionsforpredictiveFUT (GiannakidouandMari2014) Atspeechtimenandwithrespecttothespeaker sepistemicstatem(s)atn: i.[[fut(p)]]is1iffw ReasFut(n):t (n,)&p(w,t ),i.e.pistrueinall reasonablefutures,givenpresentknowledgeinm(s). ii.notallworldsinm(s)arereasfutatn.(nonveridicality) Thesetruthconditionsrenderpredictivefutureapositivebiasnonveridicaloperator;and allowustoseewhy,depsitenonveridicality,wegetmoore sparadox(thankstoitamar Francezforraisingthisquestion): (53) #Johnwillbehereatfive,butIdon tbelieveit. Thereasonwhythissentenceisbadisnotbecausethespeakeriscommittedtopwiththe futureconjunct,butbecausethespeakerprojectstruthontothesetofreasonableworlds. FUT p is supported in the reasonable worlds,i.e. all reasonable worlds w are p worlds. Though the epistemic state M(s), which included the modal base, allows not p, the continuationtargetsthereasonableworlds,andidon tbelieveitcontradictstheuniversal quantification,byassertingthatinthereasonableworldstherearealsononpworlds.in otherwords,withpositivebiasnonveridicaloperatorsareboundtogetthemooreeffect becausethesecondclausetargetsthesupportingworlds,hencetheyareweakerthanthe veridicaloperatorswherethemooreeffecttargetsthewholem(s)(seegiannakidouand Mari2014formoredetails).Whythesupportingworldsaretargetedisprobablyduetoa pragmaticfact:theyare,bytheuseoffut,themostsalientones. To sum up, the predictive future statement has this dual nature: nonveridical and thus weaker than the unmodalized past assertion and allowing for NPIs while also strongerthanmereexistentialstatements,andexpressingpartialcommitmenttopand projectedtruthontothepsupportingset. 4.Conclusion MygoalofthispaperwastoshowFranswhatanextraordinaryjourneyourinitialideasin GiannakidouandZwarts1999haveinitiated.Giventhebreadthofdatadiscussedhere andinmyworksincethe1990s,franswillbehappytoseethatthenotionsofprojected truth(nowunderstoodaspartial,biascreatingcommitment)andnonveridicality,thatwe so humbly envisioned as relevant fifteen years ago, have farreaching consequences consequences that, in fact, go beyond the mere licensing of polarity items. In the end,

21 121 veridicalityjudgmenthasbeenshowntobedecisiveformoodtriggering,i.e.inorderto capture commitment weakening with subjunctive and optative particles. And it is also necessary for the analysis of all modalities while discriminating between weaker modalities (possibility ones, with nonveridical equilibrium) and stronger ones, i.e. universalmodalitiesofthefutureandmustwithpartialsupportandpositivebiasforp. Thenotionsofactualandprojectedtruththataccompanytheveridicalityjudgmenthave alsobeenshowntobeusefultoolsformodelinglinguisticsemanticfacts.ultimately,then, the(non)veridicalityjudgmentappearstobecomplex,andthisisagoodthing:itmirrors the complexity of (non)veridicality phenomena in human language. The way epistemic agentsjudgetruthisnotasimplematter,andfrankly,lifewouldbequiteboringifitwere. Acknowledgements. In honoring Frans, I also wanted to acknowledge the unfailing intellectualsupportofjackhoeksemathroughtheyears.victorsanchezvalenciahasalso beeninstrumentalinmythinkingabouttimeandnonveridicality,anditissuchapitythat heleftussoearly.partsofthispaperinparticularwerepresentedataclcgcolloquiumat the University of Groningen (April 2013), and I am grateful to the audience for their comments and suggestions, most notably John Nerbonne and JanWouter Zwart. Many thanks,finally,tomyclosecollaboratoraldamari.shehasdeservedlytakenfrans place inmywanderingsinthelandofthefutureandnonveridicality,asamostinsightfuland enthusiasticcompanion. 5.References Abusch,Dorit.2004.Onthetemporalcompositionofinfinitives.In:TheSyntaxofTime, ed.jaquelineguéron,andjaquelinelecarme,1 34.Cambridge:MITPress. Broekhuis,Hans,andHenkVerkuyl.2013.Binarytenseandmodality.NaturalLanguage andlinguistictheory.doi /s Copley,Bridget.2002.Thesemanticsofthefuture,PhDdiss.,MIT,Cambridge. Enc,Mürvet.1996.TenseandModality.InTheHandbookofContemporarySemantic Theory,ed.ShalomLappin, Oxford:Blackwell. Fernald,TheodoreB.&EllavinaPerkins2006.NegativepolarityitemsinNavajo.In:A. Berez,S.Gessner&S.Tuttle(eds.).ProceedingsoftheDene(Athabaskan) LanguagesConference(=ALC)2006(AlaskaNativeLanguageCenterWorking Papers7).Fairbanks,AK:ANLC,Publications,19 48.

22 122 vonfintel,kai,anthonys.gillies.2010.must...stay...strong!naturallanguagesemantics 18: Giannakidou,Anastasia.1994.ThelicensingofNPIsandtheMoernGreeksubjunctive.In LanguageandCognition3,UniverisityofGroningen. Giannakidou,Anastasia.1997.TheLandscapeofPolarityItems,PhDdiss.,Universityof Groningen. Giannakidou,Anastasia.1998.PolaritySensitivityas(Non)veridicalDependency. Amsterdam:JohnBenjamins. Giannakidou,Anastasia.1999.Affectivedependencies.LinguisticsandPhilosophy22: Giannakidou,Anastasia.2009.Thedependencyofthesubjunctiverevisited:temporal semanticsandpolarity.lingua120: Giannakidou,Anastasia.2011.Negativepolarityandpositivepolarity:licensing,variation, andcompositionality.inthehandbookofnaturallanguagemeaning(2edition), ed.klausvonheusinger,claudiamaienborn,andpaulportner, MoutondeGruyter,Berlin. Giannakidou,Anastasia.2012.Greekfutureandmodalconcord.InProceedingsofthe InternationalConferenceonGreekLinguistics(ICGL)10. Giannakidou,A.2013a.Inquisitiveassertionsandnonveridicality.InThedynamic, inquisitive,andvisionarylifeof,?,andpossibly AfestschriftforJeroen Groenendijk,MartinStokhofandFrankVeltman,ed.byMariaAloni,Michael Franke,F.Roelofsen: Giannakidou,A.2013b.(Non)veridicality,epistemicweakening,andeventactualization: evaluativesubjunctiveinrelativeclauses,intrnavacandtaboada,eds. NonveridicalityandEvaluation:theoretical,computational,andcorpusapproaches. Emerald,StudiesinPragmatics. Giannakidou,A.2014.Thesubjunctiveasevaluation.ToappearinvolumewithUniversity ofchicagopress. Giannakidou,Anastasia&AldaMari.2013.Atwodimensionalanalysisofthefuture: modaladverbsandspeaker sbias.proceedingsoftheamsterdamcolloquium, Giannakidou,Anastasia&AldaMari.2014a.ThefutureinGreekandItalian:truth conditionalandevaluativedimensions.submitted. Giannakidou,AnastasiaandAldaMari.2014b.Futureanduniversalepistemicmodals: reasoningwithnonveridicalityandpartialknowledge.toappearinblaszczaketal., UniversityofChicagoPress. Giannakidou,Anastasia,andFransZwarts.1999a.Aspectualpropertiesoftemporal

23 123 connectives.ingreeklinguistics:proceedingsofthe3rdinternationalconference ongreeklinguistics,ed.amaliamozer, athens:ellinikagrammata. Giannakidou,Anastasia,andFransZwarts.1999b.Temporal,aspectualoperatorsand (non)veridicality.ms.universityofgroningen. Grosz,Patrick.2011.OntheGrammarofOptativeConstructions.MITPhDthesis. Haegeman,LilianeM.1983.TheSemanticsofwillinPresentDayEnglish:AUnified Account.Brussels:RoyalAcademyofBelgium. Hoeksema,Jacob2010.DutchENIG:Fromnonveridicalitytodownwardentailment. NaturalLanguageandLinguisticTheory28, Huddleston,Rodney.1995.TheCaseagainstaFutureTenseinEnglish.Studiesin Language9:2, Kaufmann,Stefan,CleoCondoravdi,andValentinaHarizanov.2006.Formalapproachesto modality.intheexpressionofmodality,ed.williamfrawley,71106.berlin: MoutondeGruyter. Kissine,M.(2008).Whywillisnotamodal.NaturalLanguageSemantics16: Lassiter,Daniel.2013.Mustisweak.Presentationslides,StanfordUniversity. Lee,Chungmin1999.TypesofNPIsandnonveridicalityinKoreanandotherlanguages.In: G.Storto(ed.).SyntaxatSunset2(UCLAWorkingPapersinLinguistics3).Los Angeles,CA: Lin,JoW.1996.PolarityLicensingandWhPhraseQuanticationinChinese.Ph.D.thesis. UMass,Amherst. Mari,Alda.2013.Eachother,asymmetryandreasonablefutures.JournalofSemantics. DOI: /jos/fft003 DeMarneffe,M.,C.Manning,andC.Potts.2012.Didithappen?Thepragmaticcomplexity oftheveridicalityjudgement.computationallinguistics38: Matthewson,Lisa1998.DeterminerSystemsandQuanticationalStrategies.Evidence fromsalish.thehague:hollandacademicgraphics. Palmer,FrankR.1987.MoodandModality.Cambridge:CambridgeUniversityPress. Portner,Paul.2009.Modality.OxfordUniversityPress. Thomason,R.1984.CombinationofTenseandModality,inD.Gabbay;F.Guenthner (eds.),handbookofphilosophicallogic:extensionsofclassicallogic,dordrecht: Reidel, Trnavac,Radoslava,andMaiteTaboada.2012.Thecontributionofnonveridicalrhetorical relationstoevaluationindiscourse.inlanguagesciences34: Tsangalidis,Anastasios.1998.Willandtha:aComparativeStudyoftheCategoryFuture. UniversityStudioPress.

24 124 Xherija,Orest,2013.NothingandanythinginAlbanian.Ms.UniversityofChicago. Zimmermann,Malte.2011.Discourseparticles.InSemantics(HSK33.2),eds.Klausvon Heusinger,ClaudiaMaienborn,&PaulPortner, Berlin:MoutondeGruyter. Zwarts1986.CategorialeGrammaticaenAlgebraïscheSemantiek.Eenonderzoeknaar NegatieenPolariteitinhetNederlands.PhDthesis.UniversityofGroningen. Zwarts,Frans.1995.Nonveridicalcontexts.LinguisticAnalysis25,34: Zwarts1996.ThreeTypesofPolarity.In:F.Hamm&E.Hinrichs(eds.).Plural Quantication.Dordrecht:Kluwer,1998,