A Constructional Analysis for the Skeptical

This paper addresses the issue of phonologically null elements in HPSG by providing an analysis of the construction exemplified by NPs such as ?the rich?, ?the beautiful?, ?the unemployed?, which lack an overt noun. The properties of this construction are explored in detail, and a number of approaches described: in particular approaches which posit a phonologically empty noun, and constructional approaches. It is shown that a constructional approach is empirically superior. This is interesting, theoretically, because empirical differences between such approaches have proved elusive hitherto.