"[if] global rules exist as a unified phenomenon, ... we require a theory of global grammar. This is as much an innovation over TG as TG is over PS grammar.
Since global derivational constraints are constraints involving pairs (or perhaps some times triples) of syntactic phrase-markers, such a proposal would require that semantic representations be given in the same terms as syntactic phrase-markers. This is exactly what is claimed by the theory of generative semantics." (Lakoff 1970b: 638)
[Abraham, Werner: Terminologie zur neueren Linguistik. Tübingen: Niemeyer, 1988, Bd. 1, S. 249]