TEXTURE, PATTERN, AND COHESION IN WRITTEN TEXTS: A STUDY WITH A GRAPHIC PERSPECTIVE (PRONOUNS, DEFINITE ARTICLES, COMPUTER ANALYSIS)

That some entity called \u27text\u27 exists has rarely been questioned, but because of its complexity, the synergistic nature of such texts is little understood. This synergism is the dynamic product of reader processing which is more than, but dependent on, the physical text produced by a writer. One aspect of this synergism is \u27texture\u27 which is caused by an overlaying, as it were, of numerous text patterns, including cohesion. This study provides an insight into the rudiments of texture, and hence synergism, by analyzing the patterns of three types of cohesion in thirty-five published texts representing five genres. The patterns of cohesion for these types--definite articles, pronouns, and agent displacements--were computer-analyzed for frequency of occurrence. The statistical results demonstrate that these types occur with some degree of regularity in all texts in all genres, but they also show that some linguistic signals for cohesion are ambiguous or unfulfilled. Furthermore, when abstracted locationally and drawn graphically on the computer as \u27maps\u27 of cohesive networks, these patterns uncover aspects of cohesion that the numerical data cannot. That is, as the number of cohesive elements per cohesive node varies and as the lengths of the cohesive ties vary, so will the relative cohesiveness of the texts vary. An index of cohesion developed to show this variability reveals that the personal essay genre is the most cohesive while the nonfiction genre is the least cohesive. Because the mind of the reader must simultaneously manage all types of cohesion (as well as other text components), the processing of texts is extremely complex. In order to describe this complexity for patterns of cohesion, a set of maps showing the networks of each of the three types of cohesion was transferred to colored transparencies. These maps, when overlaid, suggest not only the complexity of the processing required to read a text, but also the multidimensionality of textual synergism