It is a widely accented assumption that the difference between restrictive and non-restrictive relative clause constructions should be expressed in their syntactic categorical structure as well as in their semantics. This view is held, e.g., by Rodman (1972). There, a syntactic function combines a common noun and a sentence to form a complex common noun in the case of restrictive relative clauses, and, in the case of non-restrictive relative clause construction, a syntactic function combines a term (noun phrase with determiner or a proper name) with a sentence to form a complex term. Rodman formulates this assumption in a Montague grammar in a way that rests on the transformational approaches to relative clause constructions. He thus has to formulate conditions that say under which circumstances which elements of a sentence can be relativized, namely, within a relative clause no other element can be relativized and in the case of non-restrictive constructions, a further restriction has to be imposed to the effect that coreference between the governing term and the relativized element has to be assumed.