Domains of discourse and the semantics of ambiguous utterances : a reply to Gauker