OUPCORRECTEDPROOF–FINAL,21/11/2015,SPi RETHINKING COMPARATIVE SYNTAX PUBLISHED TheSyntaxofYesandNo AndersHolmberg IN PREPARATION ParameterHierarchiesandUniversalGrammar IanRoberts CaseandAgreementAlignmentinRomanceandBeyond MichelleSheehan OUPCORRECTEDPROOF–FINAL,21/11/2015,SPi OUPCORRECTEDPROOF–FINAL,21/11/2015,SPi The Syntax of Yes and No ANDERS HOLMBERG 1 OUPCORRECTEDPROOF–FINAL,21/11/2015,SPi 3 GreatClarendonStreet,Oxford,OXDP, UnitedKingdom OxfordUniversityPressisadepartmentoftheUniversityofOxford. ItfurtherstheUniversity’sobjectiveofexcellenceinresearch,scholarship, andeducationbypublishingworldwide.Oxfordisaregisteredtrademarkof OxfordUniversityPressintheUKandincertainothercountries ©AndersHolmberg Themoralrightsoftheauthorhavebeenasserted FirstEditionpublishedin Impression: Allrightsreserved.Nopartofthispublicationmaybereproduced,storedin aretrievalsystem,ortransmitted,inanyformorbyanymeans,withoutthe priorpermissioninwritingofOxfordUniversityPress,orasexpresslypermitted bylaw,bylicenceorundertermsagreedwiththeappropriatereprographics rightsorganization.Enquiriesconcerningreproductionoutsidethescopeofthe aboveshouldbesenttotheRightsDepartment,OxfordUniversityPress,atthe addressabove Youmustnotcirculatethisworkinanyotherform andyoumustimposethissameconditiononanyacquirer PublishedintheUnitedStatesofAmericabyOxfordUniversityPress MadisonAvenue,NewYork,NY,UnitedStatesofAmerica BritishLibraryCataloguinginPublicationData Dataavailable LibraryofCongressControlNumber: ISBN –––– Printedandboundby ClaysLtd,Stlvesplc LinkstothirdpartywebsitesareprovidedbyOxfordingoodfaithand forinformationonly.Oxforddisclaimsanyresponsibilityforthematerials containedinanythirdpartywebsitereferencedinthiswork. Contents Acknowledgements viii Liosfat b breviations xi 1 Introduction 1 1.A1n swaerrdese ribvye de llipsis 1 1.A2n swbeypr a rtoirvc elreb 2 13. Howt oa nswneerg aqtuievset ions 4 1.S4o mtee rminodleotgaiiclasl 7 1.4Q.u1e stainosnwsre,er jso,i nders 7 1.4O.nu2 n grammaantsiwcearls 9 1.5 Ont hdea ta 10 2 Thsey notfqa ux estions 13 2.T1h mee anoifqn uge stions 13 2.A2l ternatiavnedd i qsujeusntcitoinosn 17 23. Wh-questions 21 2-C4h indeissej uynecst-qinuvoee s tions 23 25. Yes-qnuoe stiinFo innsn ish 27 2.O6n fioncq uuse stions 32 2.Y7e s-qnuoe stiinEo nngsl ish 34 2.N8e gaqtuievset ions 38 2.Y9e s-qnuoe stiinTo hnasi 42 2.9I.n1t rodFuicnqtauileo snpt:ai rotni cles 42 2.9Q.u2e stwiiotTnhys p1 ep articles 43 2.39 Q.uestwiiotTnhys p2 e particles 48 2.1C0o nclusions 50 3 Thsey notfaa nxs wers 52 3.1I ntrodAuncstwiaeorrfnesu: s l eln tences 52 3.2I denatniedtl yl ipsis 55 3.3 Ond iscontfihrneme ignagat litveer pnraotpiovsei tion ofan egaqtuievset ion 59 3-4A lterntaoat nisvwpeeasrr ticles 61 3.5V erb-aencshwoGe ernse:r al 62 3.6V erb-aencshwoae mrosnt ghl ea nguoaftg hweeos r ld 64 3.7 The syntax of verb-echo answers: The significance of inflection 72 3.8 Pro-dop and VP-ellipsis or big ellipsis? 73 3.9 Testing for pro-drop: The indefinite subject test 79 3.10 Another parameter: One verb or a string of verbs? 91 3.11 A case study: Welsh 92 3.12 The structure of Finnish answers 100 3.12.1 Some basic facts. The syntax of clauses in Finnish 100 3.12.2 The structure of answers to yes–no questions in Finnish 104 3.12.3 Deriving verb-echo answers in Finnish without remnant movement 109 3.12.4 Affirmative particles in Finnish 116 3.13 The structure of answers in Thai 117 3.13.1 General properties of questions and answers 117 3.13.2 The derivation of answers in Thai 121 3.13.3 Questions and answers with Type 2 particles 128 3.14 Answering questions with coordination 130 3.15 Conclusions 135 4.How to answer negative questions 138 4.1 The two systems for answering negative questions 138 4.2 The global distribution of the two systems 142 4.3 The English answering system 152 4.4 Negation in the polarity-based system 162 4.5 Swedish: A language without low negation 165 4.6 Finnish: Another language without low negation 168 4.7 Thai and the (in-)significance of low negation 178 4.8 Answering questions with high negation 181 4.8.1 Positive and negative bias 181 4.8.2 Tag questions 183 4.8.3 Positive-bias negative questions 187 4.9 Chinese negative questions and their answers 191 4.10 Japanese positive-bias negative questions 197 4.11 Another type of biased question 200 4.12 Other accounts of positive-bias questions 201 4.13 Conclusions and some typological implications 203 5.Some further issues 210 5.1 Yes and no as rejoinders expressing agreement or disagreement with statements210 5.2 Answering yes–no questions with narrow focus 216 5.2.1 Some cross-linguistic observations 216 5.2.2 The derivation of narrow-focus questions and their answers in Finnish 220 5.2.3 Negative answers to narrow-focus questions 224 6.Conclusions 227 References 233 Descriptive grammars 243 Index of Subjects and Authors 245 Index of Languages 249 OUPCORRECTEDPROOF–FINAL,21/11/2015,SPi Acknowledgements The research for this book was funded first by a Major Fellowship from the Lever- hulme Trust in –, for a project with the same title as this book, based at NewcastleUniversity,andthenbytheEuropeanResearchCouncilAdvancedGrant No.  ‘Rethinking Comparative Syntax’ (ReCoS) based at the University of Cambridge,whichIhavebeenpartofsince. Myfirststepstowardsatheoryofthesyntaxofanswerstoyes–noquestionswere takenasearlyas–,though,whenIspentthebestpartofayearasavisiting scholaratMITworkingonapaperonverb-echoanswersinFinnish.Ifiguredthatif the generative theory of syntax was any good, it would provide an elegant formal accountoftheseellipticalexpressionsmeetingtherequirementsofobservationaland descriptive adequacy, and it did. The paper was published in Studia Linguistica in , and still ranks as one of the achievements of mine that I’m most proud of, becausetherewasverylittlepreviousworkontheconstructioninanylanguage,and none at all for Finnish, and I made a host of new observations about Finnish in particular and the construction type in general, reported in that paper—although, embarrassingly, I wasn’t aware at the time of Ana Maria Martins’s work on Portu- gueseverb-echoanswerswhichpredatedmyworkbyatleastsevenyears,norofBob Morris Jones’s detailed work on Welsh, which predated my work by at least two years.Anyway,myworkonthisconstructionmadeitcleartomethatthisisarich source of data and generalizations which go far beyond the particular construction type. I also noticed that there are interesting syntactic generalizations that hold for verb-echoanswersandyes–noanswers.Thisisnotsurprising,astheyhavethesame meaning and use, but since superficially they do look very different it was an interesting challenge to make these generalizations more precise and explain them intheoreticalterms,whichiswhatIhavetriedtodointhisbook. FirstofallIwanttoextendmywarmestthankstoallthepeoplewhohavehelped mecollectdataforthisproject,byfillingoutquestionnaires,oransweringmyemails, or talking to me, patiently answering my questions. Second, I want to thank my colleaguesandstudentsattheSchoolofEnglishLiterature,Language,andLinguistics (SELLL)andtheCentreforResearchinLinguisticsandLanguageScience(CRiLLS) atNewcastleUniversityforshowinggreatinterestandsupportingmyworkinevery kindofway.IwanttoespeciallymentionWimvanderWurff,whohasdoneresearch himselfonyesandno,formanyinterestingdiscussions.Andabigthankyoutomy colleaguesatCambridgeintheReCoSprojectteam:IanRoberts,TheresaBiberauer, Michelle Sheehan, Jenneke van der Wal, András Bárány, Tim Bazalgette, Alison OUPCORRECTEDPROOF–FINAL,21/11/2015,SPi Acknowledgements ix Biggs,andGeorgHöhnforinspiration,critique,andhelp.Ian,Theresa,andMichelle havebeenparticularlyimportanttomyresearchalmosteversinceIcametotheUK in. Thanks to Alex Ho-Cheong Leung, Laura Bailey, and Mais Sulaiman all of who servedasresearchassistantsonmyLeverhulmeproject,inperiods.Specialthanksto SomphobYaisomanangwhowroteadissertationonquestionsandanswersinThai undermysupervision,submittedin,whichplaysanimportantpartinthisbook, andtoHofaMengJungWuwhoispresentlyworkingonadissertationonquestions and answers in Chinese. I should also mention the whole class in the Topics in EnglishSyntaxmoduleofatSELLL,which wasmainly spentonthesyntaxof answers in English, and where the syntax of the English answering system first became clear to me, also helped by the fact that I read Ruth Kramer and Kyle Rawlins’s paper around the same time. When I started this project I had no idea that English had such an intricate and interesting system of yes–no questions andanswers. Just about the first thing I did in my Leverhulme Fellowship period was go to LeipzigandspendaweekattheLinguisticsDepartmentoftheMaxPlanckInstitute of Evolutionary Anthropology, getting advice from colleagues there on how to proceed with the empirical side of my work and mining their wonderful library. ThanksespeciallytoMartinHaspelmathforinterestingdiscussionandusefuladvice. I’msogladIhadtheopportunitytovisitthisbeautifulandinspiringestablishment, especiallyastheLinguisticsdepartmentthereisnow,sadly,abouttoclose. AnothertripIdidinmyfellowshipperiodwastoHyderabad,India,whereIspent twomonthsattheEnglishandForeignLanguagesUniversitytogetherwithmywife GitteEdelmaninearly,collectingdata,givingtalks,discussingandfindingout about questions and answers. Many thanks to Amrit and Jay (R. Amritavalli and K.A.Jayaseelan)andtheircolleaguesandstudentsfortheirhospitalityandhelpand interestinmywork. In  weorganized a workshop at Newcastle University, funded by the Lever- hulme Trust, with Donka Farkas, Ana Maria Martins, Ruth Kramer, and Kyle Rawlinsasspeciallyinvitedspeakers.Igainedalotofnewknowledgeandinspiration fromthatoccasion. A reader of this book will soon notice that the Syntax of the World’s Languages (SSWL) database has been important in my project. With much help from Hilda Koopman, as the new, energetic director of the database, and in collaboration with Craig Sailor I posted a set of questions on the SSWL about yes–no questions and answers in , and since then more and more data from languages all over the world are gradually gathering on the database, to the benefit of myself and other peopleinterestedinthesyntaxofquestionsandanswers. Ihavepresentedpapersonthistopicatawiderangeofconferencesovertheyears (some people must be tired of hearing me talk about it by now). I could specially OUPCORRECTEDPROOF–FINAL,21/11/2015,SPi x Acknowledgements mention GLOW in Asia , at Mie University, Japan (organized by Seiki Ayanoandhiscolleaguesandstudents),andtheGLOWColloquiuminLund (organized by Halldór Sigurdsson with colleagues and students). At both of them Iwasaninvitedspeaker,andatbothofthemIgotalotoffeedbackandnewfactsand ideastothinkabout.IcouldalsomentiontwoinvitedtalksthatIgaveintheUSA,in thefallof,oneatMITandoneatNYU,whereIhadmanyinterestingcomments and generally received much encouragement to go on with my work. Thanks especiallytoAyakaSugawarafortellingmeaboutJapanese. Andfinally,veryspecialthankstomywife,BirgittaEdelman,thesourceoflotsof cleverobservationsaboutyesandnoandotherlinguisticissues,andgrammaticality judgements, and encouragement, as well as practical help finding my things and gettingaround,andgenerallyforcreatinganatmosphereathomewhichmakesita betterplacethananywhereelseforworkaswellasotherthingsthatmatter.