Two major factors have shaped writing research from the cognitive processing viewpoint. The first factor contains the strategic decisions about the course of such research, suggesting what is interesting and how best to proceed. The decisions made in regard to writing as a cognitive process have been to focus on the act of writing, to seek a process model of writing, to model individual writers, to work holistically (a "top down" approach), and to divide the writing task into parts for easier analysis. The second research-shaping factor is the necessity to attend to certain salient facts about the writing processes. These facts that nature has imposed on the writing researcher, and that result in part from the strategic decisions the researcher makes, include the following: Writing is goal directed. Writing processes are hierarchically organized. Some writing processes interrupt other processes. Writing processes can be organized recursively. Writing goals can be modified as writing proceeds. As yet, the model developed from the strategic decisions attends to all the salient facts, with the exception that no mechanism accounts for the modification of goals; so far, insufficient data have been collected to specify how such a mechanism works. (RL)