The semantic differential scale developed by Charles Osgood, George Suci, and Percy Tannenbaum can be modified to create a method of evaluating student writing that is clear, efficient, flexible, and specific. Based upon the premise that most terms have several dimensions of meaning, the theory creates scales by drawing "lines" through the "semantic space" of a term. In a writing class, discussion, stimulated by teacher prompting and guidance, can be used to develop appropriate criteria and evaluation scales for judging assignments. For instance, "evident" and "obscure" might serve as the evaluation scales for two criteria of "purpose." After receiving copies of the evaluation scales they have evolved, students engage in peer review. The teacher then checks off his or her own evaluation according to the scales and comments on the major strengths and weaknesses of the paper and its writing technique. Students may always use the evaluation to guide revision of any paper. All in all, this system has five advantages over conventional forms of evaluation: (1) the grading process is no longer mysterious, (2) student responsibility is increased, (3) evaluation is integrated into the writing-rewriting process, (4) arbitrariness is decreased, and (5) flexibility is increased. (Includes a sample evaluation scale.) (JL)