[Ebooks PDF] download The oxford handbook of experimental syntax jon sprouse full chapters

Page 1


The Oxford Handbook of Experimental Syntax Jon Sprouse

Visit to download the full and correct content document: https://ebookmass.com/product/the-oxford-handbook-of-experimental-syntax-jon-spro use/

More products digital (pdf, epub, mobi) instant download maybe you interests ...

The Cambridge Handbook of Experimental Syntax (Cambridge Handbooks in Language and Linguistics) Grant Goodall

https://ebookmass.com/product/the-cambridge-handbook-ofexperimental-syntax-cambridge-handbooks-in-language-andlinguistics-grant-goodall/

The Oxford Handbook of the History Phenomenology (Oxford Handbooks)

https://ebookmass.com/product/the-oxford-handbook-of-the-historyphenomenology-oxford-handbooks/

The Oxford Handbook of Daniel Defoe (Oxford Handbooks) Seager

https://ebookmass.com/product/the-oxford-handbook-of-danieldefoe-oxford-handbooks-seager/

The Oxford Handbook of Religious Space Jeanne Halgren Kilde

https://ebookmass.com/product/the-oxford-handbook-of-religiousspace-jeanne-halgren-kilde-2/

The Oxford Latin Syntax, Volume II: the Complex Sentence and Discourse Harm Pinkster

https://ebookmass.com/product/the-oxford-latin-syntax-volume-iithe-complex-sentence-and-discourse-harm-pinkster/

The Oxford Handbook of Polling and Survey Methods (Oxford Handbooks)

https://ebookmass.com/product/the-oxford-handbook-of-polling-andsurvey-methods-oxford-handbooks/

The Oxford Handbook of Decadence Jane Desmarais

https://ebookmass.com/product/the-oxford-handbook-of-decadencejane-desmarais/

The Oxford Handbook of Pierre Bourdieu Unknown

https://ebookmass.com/product/the-oxford-handbook-of-pierrebourdieu-unknown/

The Oxford Handbook of Psychotherapy Ethics Trachsel

https://ebookmass.com/product/the-oxford-handbook-ofpsychotherapy-ethics-trachsel/

EXPERIMENTAL SYNTAX

OXFORDHANDBOOKSINLINGUISTICS

RECENTLYPUBLISHED

THEOXFORDHANDBOOKOFREFERENCE

EditedbyJeanetteGundelandBarbaraAbbott THEOXFORDHANDBOOKOFEXPERIMENTALSEMANTICS ANDPRAGMATICS

EditedbyChrisCumminsandNapoleonKatsos THEOXFORDHANDBOOKOFEVENTSTRUCTURE

EditedbyRobertTruswell

THEOXFORDHANDBOOKOFLANGUAGEATTRITION

EditedbyMonikaS.SchmidandBarbaraKöpke THEOXFORDHANDBOOKOFNEUROLINGUISTICS

EditedbyGreigI.deZubicarayandNielsO.Schiller

THEOXFORDHANDBOOKOFENGLISHGRAMMAR

EditedbyBasAarts,JillBowie,andGerganaPopova THEOXFORDHANDBOOKOFAFRICANLANGUAGES

EditedbyRainerVossenandGerritJ.Dimmendaal THEOXFORDHANDBOOKOFNEGATION

EditedbyVivianeDéprezandM.TeresaEspinal THEOXFORDHANDBOOKOFLANGUAGECONTACT

EditedbyAnthonyP.Grant

THEOXFORDHANDBOOKOFLANGUAGEANDRACE

EditedbyH.SamyAlim,AngelaReyes,andPaulV.Kroskrity THEOXFORDHANDBOOKOFLANGUAGEPROSODY

EditedbyCarlosGussenhovenandAojuChen THEOXFORDHANDBOOKOFLANGUAGESOFTHE CAUCASUS

EditedbyMariaPolinsky

THEOXFORDHANDBOOKOFGRAMMATICALNUMBER

EditedbyPatriciaCabredoHofherrandJennyDoetjes THEOXFORDHANDBOOKOFCOMPUTATIONAL LINGUISTICS

Secondedition

EditedbyRuslanMitkov

THEOXFORDHANDBOOKOFTHEMENTALLEXICON

EditedbyAnnaPapafragou,JohnC.Trueswell,andLilaR.Gleitman

THEOXFORDHANDBOOKOFETHIOPIANLANGUAGES

EditedbyRonnyMeyer,BediluWakjira,andZelealemLeyew THEOXFORDHANDBOOKOFEXPERIMENTALSYNTAX

EditedbyJonSprouse

ForacompletelistofOxfordHandbooksinLinguisticspleaseseepp675–78.

theoxfordhandbookof EXPERIMENTAL SYNTAX

GreatClarendonStreet,OxfordOX26DP, UnitedKingdom

OxfordUniversityPressisadepartmentoftheUniversityofOxford. ItfurtherstheUniversity’sobjectiveofexcellenceinresearch,scholarship, andeducationbypublishingworldwide.Oxfordisaregisteredtrademarkof OxfordUniversityPressintheUKandincertainothercountries ©editorialmatterandorganizationJonSprouse2023 ©thechapterstheirseveralauthors2023

Themoralrightsoftheauthorshavebeenasserted FirstEditionpublishedin2023

Allrightsreserved.Nopartofthispublicationmaybereproduced,storedin aretrievalsystem,ortransmitted,inanyformorbyanymeans,withoutthe priorpermissioninwritingofOxfordUniversityPress,orasexpresslypermitted bylaw,bylicenceorundertermsagreedwiththeappropriatereprographics rightsorganization.Enquiriesconcerningreproductionoutsidethescopeofthe aboveshouldbesenttotheRightsDepartment,OxfordUniversityPress,atthe addressabove

Youmustnotcirculatethisworkinanyotherform andyoumustimposethissameconditiononanyacquirer PublishedintheUnitedStatesofAmericabyOxfordUniversityPress 198MadisonAvenue,NewYork,NY10016,UnitedStatesofAmerica

BritishLibraryCataloguinginPublicationData Dataavailable

LibraryofCongressControlNumber:2022935323

ISBN978–0–19–879772–2

DOI:10.1093/oxfordhb/9780198797722.001.0001

Printedandboundby CPIGroup(UK)Ltd,Croydon,CR04YY LinkstothirdpartywebsitesareprovidedbyOxfordingoodfaithand forinformationonly.Oxforddisclaimsanyresponsibilityforthematerials containedinanythirdpartywebsitereferencedinthiswork.

Preface

Listoffiguresandtables

PARTIJUDGMENTMETHODSIN

1.Acceptabilityjudgments 3 JonSprouse

2.Acceptabilityjudgmentsofbindingandcoreference: Methodologicalconsiderations 29 ElsiKaiserandJeffreyRunner

3.(Quantifier)scopejudgments

KrisztaEszterSzendrői

4.Experimentalsyntaxandlinguisticfieldwork

PARTIIACQUISITIONMETHODSIN

5.Behavioralacquisitionmethodswithinfants 137 LaurelPerkinsandJeffreyLidz

6.Behavioralacquisitionmethodswithpreschool-agechildren 171 KristenSyrett

7.Modelingsyntacticacquisition 209

LisaS.Pearl

8.Artificiallanguagelearning 271

JenniferCulbertson AnnotatedbibliographyforPartII 301

PARTIIIPSYCHOLINGUISTICMETHODS INSYNTACTICTHEORY

9.Self-pacedreading 313

MasayaYoshida

10.Eye-trackingandexperimentalsyntax 333

DaveKushandBrianDillon

11.Speed–accuracytrade-offmodelinganditsinterfacewith experimentalsyntax 363

StephaniForaker,IanCunnings,andAndreaE.Martin

12.Formalmethodsinexperimentalsyntax 393

TimHunter

13.Investigatingsyntacticstructureandprocessinginthe auditorymodality 453

MaraBreenandKatyCarlson

14.Language-processingexperimentsinthefield 491

MatthewWagersandSandraChung AnnotatedbibliographyforPartIII 513

PARTIVNEUROLINGUISTICMETHODSIN

15.Electrophysiologicalmethods 533

JonSprouseandDiogoAlmeida

16.Hemodynamicmethods 559

JonathanR.Brennan

Preface

Thefieldofsyntaxhasalwaysbeeninterdisciplinary.Partofthisissimplythenatureof cognitivescience—theimmensityoftheproblemposedbyhumancognitionrequires aconcertedeffortfrommultipledisciplines.Andpartofthisisthenatureofsyntactic theory:Itmediatesbetweensoundandmeaning,itisatheoryoftherepresentations constructedduringsentence-processing,anditisatheoryoftheendstateforlanguage acquisition.Astechnologyhasadvanced,sotoohavethemethodsthatsyntacticians havebroughttobearonthecentralquestionsofthefield.Thepasttwodecadesinparticularhaveseenanexplosionintheuseofvariousexperimentalmethodsforprobing thesyntaxofhumanlanguages.ThisHandbookisanattempttobringthesestrands ofresearchtogetherintoasinglevolume.IhavethreegoalsforthisHandbook:(i)to providehigh-levelreviewsoftheexperimentalworkthathasbeendrivingthefieldof experimentalsyntax,(ii)toinspirenewresearchthatwillpushtheboundariesofthe theoryofsyntax,and(iii)toprovidehigh-levelmethodologicalguidanceforresearchers whowishtoincorporateexperimentalmethodsintotheirownresearch.Ihopereaders willagreethatthecontributorstothisvolumehavecreatedchaptersthatsucceedinall threegoals.

Forthishandbook,Ihaveintentionallydefined experimentalsyntax inthebroadest possibleterms—astheuseofany(andall)experimentalmethodsinserviceofsyntactic theory.Iamawarethattheterm experimentalsyntax issometimesusedinanarrower sensethatismoreorlesssynonymouswith formalacceptabilityjudgmentexperiments (Ihaveuseditthatwaymyselfinmyownwork),butIbelievethissynonymyismerely asymptomoftheimportantrolethatacceptabilityjudgmentsplayinsyntactictheory, andnotameaningfuldelimiterofthetypesofmethodsthatsyntacticianscanprofitably employintheirresearch.Thespaceofpossiblemethodsislarge—toolargeforanysingle volume.InassemblingthisHandbook,Ihavechosentofocusonmethodsthatare(i) relativelywell-understood,(ii)relativelypracticalintermsoftheequipmentrequired, and(iii)relativelylikelytoyieldinformationthatisrelevanttosyntactictheory.Allof thesechoicesaresubjective.Idonotintendtheexclusionofanygivenmethodtomean thatitdoesnot,orcouldnot,fallunderthebroaddefinitionofexperimentalsyntax. Infact,whatIhopethisHandbookshowsisthatthisbroaderdefinitionofexperimentalsyntaxisstillinitsinfancy.Wehavenotyetexploredallofthemethodsthatcould potentiallycontributetotheoriesofsyntax,norhaveweseenthefullpotentialofthe methodsthatwehaveexplored.Assuch,thishandbookisaspirational—itissimultaneouslyasnapshotoftheknowledgewehavecollectedtodateandapointertothekind ofworkthatwillbepossibleinthefuture.

xpreface

Onecriticalcomponentofallexperimentalworkisalinkinghypothesis—ahypothesisthatlinkstheobserveddatatotheunobservedtheoreticalconstructsthatareunder investigation.Inexperimentalsyntax,weneedhypothesesthatlinkeachofthemethods discussedinthishandbookbacktosyntactictheory.Becausethereislikelynomethod thatprovidesadirectlinktosyntactictheory(atleastatourcurrentleveloftechnology), foreachandeverymethodinthishandbook,creatingalinkinghypothesisbetween thedataandsyntactictheoryentailscreating(orinvestigating)alinkinghypothesis betweensyntactictheoryandanothercomponentofthetheoryoflanguage,suchas thetheoryofsentence-processingorthetheoryoflanguageacquisition.Manyofthe chaptersinthisHandbookdiscussthisissueindetail,soIwillnotbelaborthepoint here.Thepracticalconsequenceofthisisthatthereisatheoreticalthemethroughoutthishandbook—thelinkingofsyntactictheorytoothercomponentsofacomplete theoryoflanguage.IhaveorganizedtheHandbookaroundthistheme.Therearefour sections,eachcorrespondingtothelinkinghypothesesnecessarytoleveragethemethodsineachsectioninserviceofsyntactictheory:(i)judgmentmethods,whichrequire alinkbetweenthetheoryofofflinejudgmentsandsyntax,(ii)acquisitionmethods, whichrequirealinkbetweenthetheoryofthelanguage-acquisitionprocessandsyntax, (iii)psycholinguisticmethods,whichrequirealinkbetweenthetheoryofsentenceprocessingandsyntax,and(iv)neurolinguisticsmethods,whichrequirealinkbetween neurobiology,sentence-processing,andsyntax.

AfewnotesontheorganizationoftheHandbookareinorder.First,IhaveaskedOxfordUniversityPresstokeepthereferencesforeachchapterwiththatchapter(andnot inaglobalreferencelistattheendoftheHandbook).Myhopeisthatthiswillallowthe chaptersinthisHandbooktotrulyserveasaguideforexploringthepotentialtheoreticalcontributionsthateachmethodcanmaketosyntax,withthereferencelistsserving asafirstreadinglist.Second,becausethesemethodsarenotusefuliftheycannotbe learnedbynewresearchers,Ihaveaskedthecontributorsofeachchaptertocreatean annotatedlistofresourcesforlearningthemethodintheirchapter.Ihavecollatedthe listsbysection,compilingthemintofourstand-alonechaptersthatoccurattheendof eachsection.Finally,Ihaveaskedeachcontributortowriteamini-essayaboutwhat theyseeasthefutureofexperimentalsyntax.Myhopeisthatthesemini-essayswill provideinspirationtoreaderswhoareconsideringadoptingexperimentalsyntaxinto theirownresearchprograms,andalsoserveasasortoftimecapsulebywhichwecan measuretheprogressofthefieldinfutureyears.Ihavecollectedthesemini-essaysinto asinglechapterattheendofthehandbook.

Thisvolumecouldnotexistwithoutthetalent,energy,andeffortofinnumerable colleagues.First,IwouldliketothankthecommissioningeditoratOxfordUniversity Press,JuliaSteer,forlayingthefoundationofthisvolumebyencouragingmetoexplore abroaddefinitionofexperimentalsyntaxratherthananarrowreviewofworktodate. Second,Iwouldliketothankeachofthecontributorsforsharingboththeirvisions andtheirexpertise.Ifthisvolumesucceedsinanyofitsthreegoals,itwillbebecause

prefacexi

oftheirhardworkanddedication,bothinwritingtheirchaptersandindoingthekind ofresearchthatpushestheboundariesofthefield.Finally,Iwouldliketothankeveryonewhohassupportedmethroughoutmycareer—advisors,collaborators,colleagues, students,family,andfriends.Scienceisacommunityeffort.AndIamgratefulbeyond wordsforthecommunitythatIhavesomehowbeengiveninthislife.

Listoffiguresandtables

Figures

1.1Thetwopredictionsofthe2×2designforwhether-islands(leftpanel andcenterpanel),andtheobservedresultsofanactualexperiment (rightpanel) 17

1.2Threedemonstrationsofthecontinuousnatureofacceptability judgments 21

2.1SceneverificationdisplayfromtheexperimentbyKaiseretal.(2009) 43

2.2PictureselectiondisplayfromtheexperimentbyKaiseretal.(2009) 44

3.1OutcomeofexampleteststoryfromConroyetal.’s(2009)TVJTtask 87

7.1ModeloftheacquisitionprocessadaptedfromLidzandGagliardi(2015) 213

11.1SATfunctionforonecondition,illustratingthethreephasesof processing 369

11.2IdealizeddifferencesinthethreephasesoftheSATfunctionsfortwo conditions 370

11.3Idealizeddifferencesinthefinishingtimedistributionscorrespondingto theSATdifferencesshowninFig.11.2 370

12.1Wecanusesurprisaltoformulatealinkinghypothesiswhich,taken togetherwithaprobabilitydistributionoversentences,produces empiricalpredictionsaboutsentencecomprehensiondifficulty 401

12.2Sincesurprisalcanactasatestofprobabilitydistributionsand probabilitydistributionscanbeseenasconsequencesofhypothesized grammars,surprisalcanactasatestofhypothesizedgrammars 402

12.3Graphicalillustrationoflc-predictandlc-connect 431

13.1ExampleitemfromBreenetal.(2010)designedtoelicitnaturalistic productions 459

14.1AsampleitemsetfromSussmanandSedivy(2003) 494

14.2SomeculturallyspecificillustrationscreatedfortheChamorro PsycholinguisticsnaProject 497

16.1TheBOLDsignal 562

xivlistoffiguresandtables

16.2Schematicrepresentationofsyntax-relatedbrainregionsoftheleft hemisphere 566

16.3Linkinghypothesesconnectpropertiesofthegrammarwithneural signals 582

17.1AcceptabilityjudgmentdatareproducedfromLinebargeretal.(1983) 611

17.2Functionalneuroanatomyoflanguageandworkingmemory(WM)as relevanttoourproposal 618

17.3Aschematicofhealthyandagrammaticsentence/phraseproduction withrespecttothedorsalandventralpathwaystoarticulation 622

Tables

3.1Percentageofsurfacescoperesponseforcomprehensionquestion 64

3.2Summaryofexperimentalfindingsofthelanguageacquisitionstudies reviewedinthispaper 74

4.1Experimentalparadigmforstudyingsubjectpreference, morphologicallyergativelanguages 108

7.1ThequalitativefitYangdiscoveredbetweentheunambiguousdata advantage(Adv)perceivedbyaVarLearnerinitsacquisitionalintake andtheobservedageofacquisition(AoA)inchildrenforsixparameter valuesacrossdifferentlanguages 236

7.2Optionalinfinitiveexamplesinchild-producedspeechindifferent languages,andtheirintendedmeaning 236

8.1Summaryofkeyartificiallanguagelearningmethods 276

12.1Afirstillustrationofbottom-upparsing 417

12.2Theeffectofcenter-embeddingonbottom-upparsing 420

12.3Theeffectofleft-embeddingonbottom-upparsing 421

12.4Theeffectofright-embeddingonbottom-upparsing 423

12.5Afirstillustrationoftop-downparsing 424

12.6Theeffectofcenter-embeddingontop-downparsing 426

12.7Theeffectofleft-embeddingontop-downparsing 427

12.8Theeffectofright-embeddingontop-downparsing 428

12.9Afirstillustrationofleft-cornerparsing 430

12.10Theeffectofcenter-embeddingonleft-cornerparsing 433

12.11Theeffectofleft-embeddingonleft-cornerparsing 434

12.12Theeffectofright-embeddingonleft-cornerparsing 435

16.1Syntacticrepresentationsstandinamany-to-manyrelationshipwith sentence-processingoperations 564

16.2Summaryofbrainregionsrelatedtosyntax 565

16.3Examplesofphrasesandsentencesmadeofofrealwordsornonsense pseudo-words 572

16.4Hypotheticalcountsforutteranceswithintransitiveverbs 580

17.1Examplesofstimulifromeachcondition 612

Thecontributors

DiogoAlmeida isanAssociateProfessorofPsychologyatNewYorkUniversityAbu Dhabi.Hisresearchcapitalizesonbehavioralandelectrophysiologicaldata(EEGand MEG)toinvestigatequestionsaboutlinguisticrepresentationsandprocessesatmultiplelevels(phonology,morphology,andsyntax).HeholdsanMAincognitivescience fromtheÉcoledesHautesÉtudesenSciencesSociales(2003)andaPhDinlinguistics fromtheUniversityofMaryland(2009),andcompletedhispost-doctoraltrainingat theUniversityofCalifornia,Irvine.

MaraBreen isanAssociateProfessorintheDepartmentofPsychologyandEducation atMountHolyokeCollege.Herresearchexplorestheroleofprosodyinspeechperceptionandproduction.Usingbehavioraltechniques,eye-tracking,andevent-related potentials,sheinvestigateshowspeakersuseprosodiccuestoprovidemeaning,how listenersuseprosodytocomprehend,andhowimaginedprosodyduringreadingcan affectunderstanding.Inaddition,sheexploreshowsoundcuesareprocessedsimilarly acrossmusicandlanguage.Herworkappearsinjournalssuchas Cognition,Journal ofMemoryandLanguage,JournalofExperimentalPsychology:General,and Language, Cognition,andNeuroscience.

JonathanR.Brennan isanAssociateProfessorofLinguisticsandPsychologyatthe UniversityofMichigan,wherehedirectstheComputationalNeurolinguisticsLaboratory.HereceivedaPhDinlinguisticsfromNewYorkUniversityin2010andcompleted post-doctoraltrainingattheChildren’sHospitalofPhiladelphia.

KatyCarlson isaProfessorofEnglishintheDepartmentofEnglishatMoreheadState University.Herresearchconcentratesprimarilyonhowprosodycanaffectsentenceprocessing,withspecialinterestsinfocuseffectsinellipsissentencesandprosodic influencesonattachment.Shestudiesbothpitchaccentsandprosodicboundaries,and haspublishedinjournalssuchasLanguageandSpeech,Glossa,andLanguage,Cognition andNeuroscience.

SandraChung isDistinguishedProfessor(emerita)ofLinguisticsattheUniversity ofCalifornia,SantaCruz.Herresearchinvestigatestheoreticalissuesinsyntaxand otherareasthroughfieldworkonChamorroandotherAustronesianlanguages.Shehas collaboratedonresearchinsemanticswithWilliamA.Ladusaw,andonresearchinpsycholinguisticswithMatthewWagers.Since2009shehasbeeninvolvedinacommunitybasedeffortintheNorthernMarianaIslandstoupgradethedocumentationofthe Chamorrolanguage.

xviiithecontributors

JenniferCulbertson isaProfessorintheDepartmentofLinguisticsandEnglishLanguageattheUniversityofEdinburgh,andafoundingmemberoftheCentrefor LanguageEvolution.Sheusesexperimentalandcomputationaltoolstoinvestigate howthehumancognitivesystemshapeslinguistictypology.ShereceivedtheRobert J.GushkoPrizeforOutstandingDoctoralDissertationsinCognitiveSciencein2012, waselectedtotheYoungAcademyofEuropein2019,andcurrentlyholdsaEuropean ResearchCouncilStartingGrant.

IanCunnings isanAssociateProfessorofPsycholinguisticsintheSchoolofPsychologyandClinicalLanguageSciencesattheUniversityofReading,UK.Hismainresearch interestsareinsentenceanddiscourse-processingindifferentpopulationsofspeakers. Hisworkhasexaminedthememory-encoding,storage,andretrievalmechanismsthat subservetheresolutionofdifferenttypesoflinguisticdependenciesduringlanguage comprehension.Hismostrecentresearchexamineshowthesedifferentmemoryoperationscaninformourunderstandingofthefactorsthatinfluencesuccessfulsentence comprehensioninnativeandnon-nativespeakerpopulations.

BrianDillon isanAssociateProfessorintheDepartmentofLinguisticsattheUniversityofMassachusetts,Amherst.Hisresearchfocusesonadultsentencecomprehension, aimingtounderstandhowlinguisticconstraintsaredeployedinreal-timetoconstrain sentenceprocessing.Inhiswork,heintegratesinsightsfromlinguistictheorywith process-levelcognitivemodels,withaparticularinterestintheprocessingofagreement andpronominalreference.

StephaniForaker isanAssociateProfessorintheDepartmentofPsychologyatState UniversityofNewYorkCollegeatBuffalo,specializingincognitionandpsycholinguistics.Hermainresearchinterestsareinsentenceanddiscourseprocessing,focusingon theroleofmemoryandfocalattention.Shehasusedthespeed–accuracytradeoff(SAT) proceduretoinvestigatelong-distancedependenciesandpronounresolution,training underBrianMcElree,whopioneeredtheapplicationofSATtopsycholinguisticissues. Hercurrentresearchexaminesthecontributionofhandgesturesaspartofencoding, storage,andretrievaloperations,particularlyinanaphoraresolution.

TimHunter isanAssociateProfessorintheDepartmentofLinguisticsattheUniversityofCalifornia,LosAngeles.Thebulkofhisresearchusescomputationalperspectives toinvestigatetheformalpropertiesofnaturallanguagegrammar,withonemaingoal beingtoclarifytheconsequencesoftakinglinguistictheoriestobetestablecognitivehypothesis.Thislineofworkincludesstudiesconnectingminimalistsyntaxto experimentalworkinlanguage-processing,andstudiesoftherelationshipbetweendeterminers’truth-conditionsandverificationprocedures.Hehasalsoworkedonthe argument/adjunctdistinctionandthesyntaxofellipsis.

ElsiKaiser isaProfessorintheDepartmentofLinguisticsattheUniversityofSouthern California.ShereceivedherPhDinlinguisticsfromtheUniversityofPennsylvania,afteraBAinGermaniclanguagesandliteraturesfromPrincetonUniversityandanMAin

thecontributorsxix

psychologyfromtheUniversityofPennsylvania.Herresearchfocusesontheprocesses andrepresentationsinvolvedincomprehensionandproduction,especiallyindomains involvingmultipleaspectsoflinguisticrepresentation(syntax,semantics,pragmatics), suchasreferenceresolution.Shehasinvestigatedmultiplelanguages(e.g.Finnish,Estonian,French,German,andDutch,includingcollaborativeworkonBangla/Bengali, Hindi,Italian,Korean,Chinese,andVietnamese).

DaveKush isanAssistantProfessorofLinguisticsatUniversityofToronto.Heis interestedinsentence-processing,syntactictheory,andcross-linguisticvariation.

JeffreyLidz isDistinguishedScholar-TeacherandProfessorofLinguisticsattheUniversityofMaryland.Hisresearchexploreslanguageacquisitionfromtheperspective ofcomparativesyntaxandsemantics,focusingontherelativecontributionsofexperience,extralinguisticcognition,anddomain-specificknowledgeinlearners’discovery oflinguisticstructureandlinguisticmeaning.

AndreaE.Martin isaLiseMeitnerResearchGroupLeaderattheMaxPlanckInstitute forPsycholinguistics,andaPrincipalInvestigatorattheDondersCentreforCognitiveNeuroimagingatRadboudUniversityinNijmegen,theNetherlands.Herwork hasspannedstructuralandsemanticaspectsofsentenceprocessing.Shehasusedthe speed–accuracytrade-offprocedureandcognitiveneuroimagingtostudytheroleof memoryinsentenceprocessingviaellipsis,alineofresearchbegunwithBrianMcElree, whopioneeredapplicationofSATtopsycholinguisticissues.Thecurrentfocusofher lab,LanguageandComputationinNeuralSystems,isondevelopingtheoriesandmodelsoflanguagerepresentationandprocessingwhichharnessthecomputationalpower ofneuraloscillations,suchthatformalproperties(viz.,constituency,compositionality) canberealizedinbiologicalandartificialneuralnetworks.

WilliamMatchin isanAssistantProfessorofCommunicationSciencesandDisorders intheArnoldSchoolofPublicHealthattheUniversityofSouthCarolina.Aspartof theCenterfortheStudyofAphasiaRecovery,hedirectstheNeuroSyntaxlab,using functionalneuroimagingandlesion–symptommappingandincorporatinginsightsof linguistictheorytounderstandthearchitectureoflanguageinthebrain.Heiscurrently investigatingthenatureofgrammaticaldeficitsinaphasia,includingparagrammatism andagrammatism.

LisaS.Pearl isaProfessorintheDepartmentofLanguageScienceattheUniversity ofCalifornia,Irvine.Herresearchliesattheinterfaceoflanguagedevelopment,computation,andinformationextraction,includingbothcognitivelyorientedresearchand appliedlinguisticresearchthatcombinestheoreticalandcomputationalmethods.Her cognitivelyorientedresearchfocusesonchildlanguageacquisition,withaparticular focusontheoryevaluationviaacquisition-modeling,andhowchildren’sinputaffects theirlinguisticdevelopment.

LaurelPerkins isanAssistantProfessorintheDepartmentofLinguisticsattheUniversityofCalifornia,LosAngeles.SheearnedherPhDinlinguisticsfromtheUniversity

xxthecontributors

ofMarylandandheldapostdoctoralfellowshipintheLaboratoiredeSciencesCognitivesetPsycholinguistiqueattheÉcoleNormaleSupérieure.Herresearchstudiesthe earlieststagesofsyntaxacquisitionininfancy,drawingfromformallinguistics,developmentalpsychology,andcomputationalcognitivemodelling.Sheisarecipientofa Post-DoctoralStudyGrantfromtheFyssenFoundation,aDoctoralDissertationImprovementGrantfromtheNationalScienceFoundation,andaGlushkoDissertation PrizefromtheCognitiveScienceSociety.

MariaPolinsky isProfessorofLinguistics,AssociateDirectoroftheLanguageScience Center,andDirectorofResearchFieldStationsattheUniversityofMaryland.Shehas conductedextensiveprimaryworkonseverallanguagesoftheCaucasus,Austronesian languages,andChukchi.Sheisalsoengagedinacomprehensiveresearchprogramon heritagelanguages.Herworkemphasizestheimportanceoflesser-studiedlanguages fortheoreticallinguistics.RecentpublicationsincludeDeconstructingErgativity(2016), HeritageLanguagesandTheirSpeakers (2018),and TheOxfordHandbookofLanguages oftheCaucasus (2021).

CorianneRogalsky isanAssociateProfessorofSpeechandHearingScienceinthe CollegeofHealthSolutionsatArizonaStateUniversity(ASU).AsDirectorofASU’s CommunicationNeuroscienceandNeuroimagingLab,Rogalskyusesbehavioraland neuroimagingtechniquestobetterunderstandtheneuralandcognitiveresourcesthat supporteffectivecommunicationineverydaylifeforindividualswhohaveexperiencedabraininjurysuchasastroke.Rogalsky’scurrentfocusisinvestigatinghow executivefunctionssuchasselectiveattentionandworkingmemorysupportspeech comprehensioninneurotypicaladults,andhowthatsupportmaychangeafterastroke.

JeffreyRunner isaProfessorofLinguisticsandBrain&CognitiveSciences,Deanof theCollege,andViceProvostandUniversityDeanforUndergraduateEducationat theUniversityofRochester.HeearnedaBAinlinguisticsattheUniversityofCalifornia,SantaCruz,in1989andaPhDinlinguisticsattheUniversityofMassachusettsat Amherstin1995.HejoinedthedepartmentofLinguisticsattheUniversityofRochester in1994.Hisresearchusesexperimentalmethodologiestoinvestigatenaturallanguage syntax.In2017,hebecamedeanoftheCollegeinArts,SciencesandEngineering, andisresponsibleforthecurricular,co-curricular,andextra-curricularundergraduate experience.

JonSprouse isaProfessorofPsychologyatNewYorkUniversityAbuDhabi.HereceivedanABinlinguisticsfromPrincetonUniversity(2003)andaPhDinlinguistics fromtheUniversityofMaryland(2007).Hisresearchfocusesontheuseofexperimentalsyntaxtechniques,includingacceptabilityjudgments,EEG,andcomputational modeling,toexplorefundamentalquestionsinsyntax.Hehasauthoredoverfortyjournalarticlesandbookchaptersonexperimentalsyntax.Hisworkhasbeenrecognized bytheBestPaperin Language award,theEarlyCareeraward,andtheC.L.Baker mid-careerawardfromtheLinguisticSocietyofAmerica.

thecontributorsxxi

KristenSyrett isanAssociateProfessorintheDepartmentofLinguisticsatRutgers,the StateUniversityofNewJersey–NewBrunswick,withaco-appointmentattheCenter forCognitiveScience(RuCCS).SheistheDirectoroftheLaboratoryforDevelopmental LanguageStudies.Herresearchfocusesonsemanticsanditsinterfacewithpragmatics andsyntaxinlanguageacquisitionanddevelopment,andonexperimentalsemantics andpragmaticsinadultpsycholinguistics.

KrisztaEszterSzendrői isaProfessorofTheoreticalandExperimentalLinguisticsat theUniversityofVienna.Sheworksoninformationstructure,includingitssyntaxand prosody,usingboththeoreticalandexperimentalmeans,workingwithbothadults andchildren.Shehasalsoworkedonthesyntaxofscope,especiallyontheacquisitionofscope.Sheisalsointerestedintheinteractionsbetweeninformationstructure andscope.Onadifferentnote,forthepastfewyearsshehasbeenleadingaresearch projectstudyingthegrammarofContemporaryHasidicYiddish.

MatthewWagers isProfessorofLinguisticsattheUniversityofCalifornia,Santa Cruz,wherehehastaughtsince2009.Thefocusofhisresearchishowsyntacticinformationisrepresentedinmemoryandhowmorphologicalcuesguideincremental interpretation.Thesearecross-cutbyaninterestinbroadeningthecontributionofpsycholinguisticallyunder-investigatedlanguagestotheorydevelopment.HeholdsaPhD inlinguisticsfromtheUniversityofMaryland(2008)andanABinMolecularBiology fromPrincetonUniversity(2003).

MasayaYoshida isanAssociateProfessorintheDepartmentofLinguisticsatNorthwesternUniversity.Researchinterestsincludeonlinesentenceprocessingandsyntax. Hehasworkedonthesyntaxandprocessingofellipsisconstructions,long-distance dependencies,andislands.Someofhisrecentstudieshaveexploredthestructureassociatedwiththeellipsissiteinclausalellipsisconstructions,andhowstructureinthe ellipsissiteisbuiltduringonlinesentence-processing.

JUDGMENT METHODSIN

SYNTACTIC THEORY

acceptability judgments jonsprouse

1.1Introduction

Thegoalofexperimentalsyntax,atleasttomymind,isstraightforward:touseexperimentalmethodstocollectdatathatisrelevantfortheconstructionandevaluationof syntactictheories.Fordatatypesthatcanonlybecollectedusingaformalexperiment, suchasreactiontimesorEEG,theworkofexperimentalsyntaxissimplytheworkof leveragingthesemethodsforquestionsintheoreticalsyntax.However,thingsappearto beabitmorecomplicatedwhenthedatatypeinquestionisacceptabilityjudgments,as acceptabilityjudgmentscanbecollectedbothrelativelyinformally,asistypicalinmuch ofthesyntaxliterature,orrelativelyformally,asistypicalintheexperimentalsyntax literature.Itakethecoexistenceofthesetwomethodsofjudgmentcollectiontoimply thatthegoalofexperimentalsyntaxwithrespecttoacceptabilityjudgmentsisnotsimplytocollectacceptabilityjudgments,becausethatiswhatisdoneinallsyntacticwork, butrathertoexplorewaysinwhichtheformalcollectionofjudgmentscanaddnewinsightsoverandabovethosethatderivefrominformalmethods.Therefore,mygoalin thischapteristoidentifyfourareasinwhichformaljudgmentexperimentshavemade substantialcontributions—twothatleantowardmethodologicalissues,andtwothat leantowardtheoreticalissues—andtoreviewthecurrentstateoftheevidencethatwe haveforeachofthoseareas.Tobeclear,thischapterisnotintendedasanexhaustivereviewofallpossibleareasinwhichformaljudgmentexperimentscouldpotentiallymake acontribution;rather,itisintendedasastartingpointforthinkingaboutthekindsof questionsintheoreticalsyntaxthatmightbenefitfromformalacceptabilityjudgment experiments.Myhopeisthatthesequestionswillhelptoinspirenewquestions,and newwork,inthegrowingfieldofexperimentalsyntax.

4jonsprouse

Beforedelvingintotheprimarycontentofthischapter,Iwouldliketobrieflymentionafewassumptions(and/ordecisions)thatIammaking.ThefirstisthatIassume, followingmanyworkingsyntacticians,thatacceptabilityjudgmentsareinprinciple valuablefortheconstructionandevaluationofsyntactictheories.Iwill,therefore,not attempttomotivatetheuseofacceptabilityjudgmentsingeneral(seeSchütze1996 foracomprehensivediscussionofthis).ThesecondisthatIwillassumearelatively minimallinkinghypothesisbetweenacceptabilityjudgmentsandthecognitivepropertiesofsentenceprocessing.Underthislinkinghypothesis,anacceptabilityjudgment isarelativelyautomaticbehavioralresponsethatariseswhenaspeakercomprehends asentence,andthatthisbehavioralresponseisimpactedbyalargenumberofcognitivefactors,suchasthegrammaticalityofthesentence,theprocessingdynamicsof thesentence,thesentence-processingresourcesrequiredbythesentence,themeaning ofthesentence,theplausibilityofthesentencerelativetotherealworld,andeventhe propertiesofthespecifictaskthatisgiventothespeaker.Ibelievewholeheartedlythat amorepreciselinkinghypothesiswouldbehelpfulforusingjudgmentsasevidencein syntax;however,Ialsobelievethattheminimallinkinghypothesisaboveismorethan sufficienttobegintoexplorethevalueofformalacceptabilityjudgmentexperiments insyntax.Mythirdassumptionisthatthereisnosubstantivequalitativedifference between“informal”and“formal”judgmentexperiments.Bothareexperimentsinthe sensethattheyinvolvethemanipulationofonevariable(syntacticstructure)torevealacausalrelationshipwithanothervariable(acceptability).Thereforebothinvolve allofthecomponentsthattypifypsychologyexperiments:asetofconditions,asetof itemsineachcondition,asetofparticipants,ataskfortheparticipantstocomplete usingtheitems,andaprocessforanalyzingtheresultsofthetask.Thedifferenceappearstometobeprimarilyquantitative,inthat“formal”experimentstendtoinvolve moreconditions,moreitemspercondition,moreparticipants,andmorecomplexanalysisprocesses.Tomymind,thelabels“informal”and“formal”simplypointtoward differentendsofthisquantitativespectrum.Inpractice,whenIsaythatformalexperimentsarevaluableinsomeway,whatImeanisthatincreasingthenumberof conditions,items,orparticipants,and/orincreasingthecomplexityoftheanalysis, canyieldinsightsthatfewerconditions,items,participants,and/orlesscomplexanalysescannot.Thelabels“informal”and“formal”areamoreconcisewaytoexpress thisidea.MyfourthassumptionisthatSchütze1996alreadyprovidesacomprehensivereviewofexperimentalsyntaxworkthatwaspublishedbefore1996.Therefore, inordertoprovidesomethingnewforthefield,Iwillfocushereonworkpublished after1996.

Finally,thischapterisnotahow-toforconstructingformaljudgmentexperiments. Thegoalisforthistobethechapteronereads,eitherbeforeorafterreadingahowto,forinspirationaboutthetypesofquestionsonecanaskwiththemethod.Iwill providesomereferencesforlearningacceptabilityjudgmentmethodsintheannotated bibliographyforPartIofthisHandbook.

1.2Thevalidityandreliabilityof acceptabilityjudgments

Perhapsthemostfrequentlyaskedquestionintheexperimentalsyntaxliteratureisto whatextenttheinformallycollectedjudgmentsthathavebeenpublishedintheliteraturecanbetrustedtoformtheempiricalbasisofsyntactictheory.Thisquestionhas arisensincetheearliestdaysofgenerativegrammar(Hill1961;Spencer1973);itplayed acentralroleinthetwobooksthatusheredinthemostrecentwaveofinterestinexperimentalsyntax(Schütze1996;Cowart1997);andithasgivenrisetoanumberof high-leveldebatesintheexperimentalsyntaxliteratureoverthepast15yearsorso (seeEdelmanandChristiansen2003;Ferreira2005;WasowandArnold2005;Featherston2007;GibsonandFedorenko2013forsomeconcernsaboutinformalmethods;see Marantz2005andPhillips2009forsomerebuttals,andMyers2009foraproposalthat attemptstosplitthedifferencebetweeninformallycollectedjudgmentsandfull-scale formalexperiments).Theexistenceofthisquestionisunderstandable.First,informally collectedjudgmentsformthevastmajorityofthedatapointspublishedinthe(generative)syntaxliterature.Second,thepropertiesofinformalcollectionmethodsarenot identicaltothepropertiesoftheformalexperimentalmethodsthatareoftenusedin otherdomainsofcognitivescience:Informalmethodsofteninvolveasmallernumberofparticipants,thoseparticipantsareoftenprofessionallinguistsinsteadofnaïve participants,theparticipantsareoftenpresentedasmallernumberofitems,andtheresultsareoftenonlyanalyzeddescriptively(withoutinferentialstatistics).Ifonebelieves thatthepropertiesofformalexperimentsarewhattheyaretoensurethequalityofthe data,thenitislogicallypossiblethatthedifferencesbetweeninformalmethodsandformalexperimentscouldleadtolower-qualitydata.Theconsequencesofthiscannotbe understated.Iftherearesystemicproblemswithinformallycollectedjudgments,then therearelikelytobesystemicproblemswith(generative)syntactictheories.

Thisquestiontouchesuponanumberofissuesinpsychometricsandthebroader philosophyofmeasurement.Thefirstquestionis:Whatdowemeanwhenwesaythat datacanbe“trusted”toformthebasisforatheory?Psychometrictheorieshaveidentifiedanumberofpropertiesthatgoodmeasurementmethodsshouldhave.HereIwill mentiontwo(andonlyinacoarse-grainedway,settingasidesubtypesoftheseproperties):validityandreliability.Ameasurementmethodis valid ifitmeasurestheproperty itispurportedtomeasure.Ameasurementmethodis reliable ifityieldsconsistentresultsunderrepeatedmeasurements(withunchangedconditions).Theconcernsabout informalmethodsthathavefiguredmostprominentlyintheliteratureappeartobea combinationofconcernsaboutvalidityandreliability,suchastheconcernthatsmall samplesizeswillleadtoanundueinfluenceofrandomvariation,theconcernthata smallnumberofexperimentalitemswillleadtoanundueinfluenceoflexicalproperties,andtheconcernthattheparticipationofprofessionallinguistswillleadtotheoreticalbias.Ineachcase,theconcernseemstobethatinformallycollectedjudgments

6jonsprouse

willnotreflectthetrueacceptabilityofthesentence(validity),andfurthermorethatthe judgmentsthemselveswillbeinconsistentoverrepeatedmeasurements(reliability).

Thisleadstoasecondquestion:Howdoesoneestablishvalidityforthemeasurement ofacognitivepropertylikeacceptability?Thedirectmethodforestablishingvalidityis tocomparetheresultsofthemeasurementmethodwithasecond,previouslyvalidated, measurementmethod.Thisisobviouslyunavailableformostcognitiveproperties—if cognitivescientistshadamethodtodirectlymeasurethecognitivepropertyofinterest, wewouldnotbotherwiththeunvalidatedmeasurementmethod.Thatleavesonlyindirectmethodsofvalidation.Oneindirectmethodistoaskwhetherthetheorythatresults fromthedatahasthepropertiesofagoodscientifictheory.This,ofcourse,interacts withbroaderissuesinthephilosophyofscienceaboutwhatpropertiesagoodtheory wouldhave,soIwillnotattempttoprovideanexhaustivelist.Buttwopossiblecriteria are:(i)makingpotentiallyfalsifiablepredictions,and(ii)explainingmultiplephenomenawitharelativelysmallnumberoftheoreticalconstructs.Inthecaseofacceptability judgments,Iwouldarguethattheresultingtheoryofsyntaxdoes,indeed,havethese properties.Anotherindirectmethodistoaskwhetherotherdatatypesprovidecorroboratingevidence,modulothelinkingtheoriesbetweenthedatatypesandtheunderlying theory.Inthecaseofacceptabilityjudgments,wecanaskwhethertheresultingsyntactictheorycanbelinkedtoasentence-processingtheoryinawaythatmakespotentially falsifiablepredictionsaboutotherpsycholinguisticmeasures,suchasreadingtimes,eye movements,orEEG,andultimatelywhetherthesemeasurescorroboratethesyntactic theory.Iwouldarguethatthecurrentresultsintheliteratureconnectingsyntactictheoriesandsentence-processingtheoriesarepromising.Thatsaid,indirectmethodscannot guaranteevalidity.Itislogicallypossiblethatacceptabilityjudgmentscouldgiveriseto atheorythathasallofthehallmarksofagoodtheory,butthatdoesnotultimately explainhumansyntax(perhapstheresultingtheoryisactuallyaboutprobability,or plausibility,orevenprescriptivegrammaticalrules).

Thisleadstothefinalquestion:Howdoesoneestablishreliability?Inprinciple,establishingreliabilityisrelativelystraightforward,asitsimplyentailsreplicatingthemeasurement.Theexactreplicationcanvarybasedonthetypeofreliabilityoneisinterested in:Between-participant(orinter-rater)reliabilityaskswhetherthesamejudgmentsare obtainedwithdifferentsetsofparticipants;within-participant(ortest–retest)reliability askswhetheronesetofparticipantswillgivethesamejudgmentsattwodifferenttimes; between-taskreliabilityaskswhetherdifferentjudgmenttaskswillyieldthesamejudgments(eitherbetween-participantorwithin-participant).Inpractice,establishingthe reliabilityofinformalmethodsiscomplicatedbytheirinformality.Bydefinition,informalmethodscontrolthevariouspropertiesofthejudgmentcollectionprocessless strictlythanformalmethods,makingastrictreplicationdifficultifnotimpossible.One waytocircumventthisproblemistocomparetheresultsofinformalmethods,perhaps asreportedinthesyntacticliterature,withtheresultsofformalexperiments.Thiswould beatypeofbetween-taskreliabilityforinformalandformalmethods,andtotheextent thatthetwosetsofresultsconverge,itwouldestablishakindofreliabilityforinformal methods.Manyoftheresultsreportedbelowtestpreciselythiskindofreliability.But

acceptabilityjudgments7

itisimportanttonotethatwhileconvergencebetweenthetwomethodscanbeinterpretedasestablishingatypeofreliabilityforboth,divergencebetweenthetwomethods canbeinterpretedinthreeways:Itcouldbethecasethatinformalmethodsareunreliable,oritcouldbethecasethatformalmethodsareunreliable,orboth.Itistempting toassumethatformalmethodsenjoysomesortofepistemologicalpriorityinadivergence(i.e.thatformalmethodsrevealthegroundtruth),butasmanylinguistshave pointedout,itiseasytoimagineexperimentalmaterialsthatleadtounreliablejudgmentsfromnon-linguistparticipants,butnotfromlinguistparticipants(suchasgarden pathsentenceslike Thehorseracedpastthebarnfell).Resolvingthesourceofthedivergencebetweentwomethodsrequiresfollow-upexperimentsthatmanipulatespecific hypothesesforthedivergence.Tomyknowledge,thoughtherehavebeenmanystudies oftheconvergence/divergencebetweeninformalandformalmethods,therehavebeen nosystematicstudiesofthesourceofthedivergencesthatdoarise(presumablybecause, aswewillseepresently,therearerelativelyfewdivergencesbetweenthemethods).

Inreviewingtheevidencecollectedsofarontheconvergencebetweeninformaland formalmethodsforjudgmentcollection,itisusefultomakeadistinctionbetweenstudiesthatsampledthedatapointstoretestwithbias,andstudiesthatsampledthedata pointstoretestrandomly.Biasedsamplingmeansthatthedatapointswerechosenbecauseofsomepropertythattheyhave;inthesestudies,thisistypicallythebeliefthatthe specificdatapointsareinvalidorunreliable.Typicallythisbeliefcomesfromdebates intheliteratureaboutthestatusofthedatapoint,ortheresearchers’own(informally collected)judgments.Biasedsamplingstudiescanbeusedtoestablishthattheconvergencebetweeninformalmethodsandformalmethodsisnotperfectbyshowingthat thedatapointsinquestiondonotreplicateusingformalmethods.Butbiasedsampling cannotbeusedtoestimateaspecificconvergencerate.Abiasedsamplecouldeither overestimateorunderestimatetheactualconvergenceratebyvirtueofthebiasedselectionprocedure:Aprocedurethatfocusesonselectingknowninvalidorunreliabledata pointswillalmostcertainlyunderestimatethetrueconvergencerate;similarly,aprocedurethatfocusesonselectinglikelyuncontroversialdatapoints(e.g.ajudgmentfor Thisisapen)islikelytooverestimatetheconvergencerate.Thereareonlytwooptions fordeterminingthetrueconvergencerate:Anexhaustivecomparisonofalldatapoints, whichwouldestablishtheconvergenceratewithcertainty,orarandomsamplingprocedure,whichwouldestimatetheconvergenceratewithinamarginoferrordetermined bythesizeoftherandomsamplerelativetothesizeofthepopulationinquestion.

Biasedsamplingstudiesdominatedmuchofthedebateaboutthevalidityandreliabilityofinformallycollectedjudgmentsuntilrelativelyrecently,presumablybecause ofthetimeandfinancialcostassociatedwithtestinglargenumbersofdatapointsprior tothecreationofcrowdsourcingplatformslikeAmazonMechanicalTurk.HereIwill brieflyreviewsomeofthemoreprominentbiasedsamplingstudies.WasowandArnold (2005)testedaclaimfromChomsky(1957)thattheorderingpreferencebetweenNPs andparticlesinverb–particleconstructionsisbasedonthecomplexityoftheNP,not thelength.TheyfoundthatthejudgmentsfollowChomsky’sreportedjudgmentswhen averagedovertheentiresampleofparticipants,butthatsomeindividualparticipants

Turn static files into dynamic content formats.

Create a flipbook
Issuu converts static files into: digital portfolios, online yearbooks, online catalogs, digital photo albums and more. Sign up and create your flipbook.