This paper continues the discussion of the restriction operator Kaplan and Wedekind 1993 and whether it can provide a linguistically adequate solution to the problem posed by syntactic complex predicate formation. The solution introduced here has been implemented as part of an on-going project aimed at the development of a computational grammar for Urdu and can be shown to model the linguistic facts of syntactic complex predicate formation as described by Alsina 1996 and Butt 1995. This also allows for a straightfoward extension to related phenomena in other languages such as German, Japanese, Norwegian, and French.