Coordination of predicates is not, as claimed by many researchers, pragmatically conditioned; neither can it be described or explained satisfactorily in terms of purely syntactic theories. The present study on conjoined predicates shows that this structure is used referentially and covers varying scopes and types of knowledge representations. Its use is examined in various linguistic systems, ranging from contact languages and a developing second language to native usage. While conjoining of predicates is realized in serial verbs in contact languages to expand the deficient lexicon, its function in initial stages of learner language is the creation of tight knowledge chunks. When the scope is widened with the expansion of the general linguistic system, it is extended to cover larger, novel contents. An accompanying investigation of native English shows a whole gamut of usages with the content varying from a narrow to a broad scope as well as to on-line chunking, which means that conjoined predicates symbolize information that is stored, perceived, or imagined and serves as a stylistic device in a large number of contexts. The development and the native use of this structure are described and explained in terms of Role and Reference Grammar, which takes a semantic and syntactic view on linguistic structures.