ABSTRACT
Ishallconcludethediscussionofanaphorawithabriefsurveyofthe caseswheremyanalysiseitherfailstocapturethefactsorcandoso onlywithsomeadhocmodifications.Theproblemspertainprimarily tothesyntacticaspectsoftheanalysis,i.e.tothec-commandrequire-
ment.Wewillseethattherearethreetypesofconstructionswherethe
c-commandconditiononboundanaphoraisviolated.Violationsmay showupinothercasesaswell,butoftentheyvarymorewithspeakers
andwithaparticularchoiceofexamplesandarenotfullydecisive.In
thecasesIsurvey,ontheotherhand,theconditionfailssystematically.
Sothesearerealproblemsthatshouldbesolved,butatthemomentI
donothaveanyimpressivesolution.Wemaynote,first,thatalthough
theproblemsappeartobespecifictothec-commandrequirement,they
havenoobvioussolutioninanyotherframework.Forexample,they
wouldposeadifficultyforasemanticallybasedtheoryaswell.Next,it
isinterestingtonotethateachoftheseproblemsshowsupequallyin
allthethreetypesofboundanaphoraweobservedinChapters5and7. Thismeansthatalthoughthepreciseconditionsonboundanaphoramay
notbefullyunderstoodyet,thetheoreticalclaimofChapter7isnot affectedbysuchproblems:reflexivisation,quantifiedNPanaphoraand sloppy-identityinterpretationstillobeypreciselythesameconditions.