In a written discourse consisting of a string of "psychological paragraphs," there is in each such psychological paraqraph a "head" structure containing the topic which derives from the deep structure of the discourse. That "head" assertion differs from all other assertions in the psychological paragraph in that it carries new information. The topic expressed in the head assertion is maintained through the remainder of that psychological paragraph by the operation of "focus" which is achieved through appropriate choice among available syntactic alternatives. An attempt to provide empirical verification of these assumptions is described. It seems that native and non-native speakers differ in their ability to determine what presuppositions may in fact be shared between writer and reader in a given communication setting (a cultural problem). Secondly, native and non-native speakers differ in the strategies they can bring to bear on the maintenance of any given topic because they have available different sets of syntactic alternatives among which to choose. The non-native speaker uses available syntactic alternatives from his native language and arranges those alternatives in terms of the rhetorical preferences of it. Therefore it would appear that knowledge of the syntax of the second language is not in itself sufficient. There is a need for more elaborated studies of the intersentential syntax underlying focus in various languages. (Author/AMH)