ABSTRACT
Oneoftheclassicargumentsfortheclausalnature
appearstobeaCOMPpositionor,inmoreneutralterms,
theyhave+WHcounterparts.WeknowthatSshaveaCOMP
partsfollowsautomatically.Itisquiteevident,how-
ever,thatthisargumentisfarfrombeingconclusive.By
areNPs(orviceversa),fortheybothcantakePPsas modifiers.IfweassumethatNPsareVPsthisdistribu-
reasoningwedon'twanttomakelargescaleuseof.
