Throughout the early years of the twentieth century, literacy education was based on the solid understanding of grammar. Yet as early as 1923, empirical data indicated that the link between knowledge of grammar and correct use of English was tenuous at best. Despite formidable evidence, some educators still advocate the use of grammar as a principal form of English language instruction. Noam Chomsky challenged supporters of traditional grammar instruction by advancing an alternative explanation of language development. Better working concepts of sentence length, clause length, and T-units were developed. Sophisticated studies confirmed the legitimacy of sentence combining as a means of developing composition skills. There were further developments in composition research and theory: (1) student composition should be a natural outgrowth of classroom discussions and the reading of literature; (2) students use 2 different approaches to writing, "extensive" and "reflexive"; (3) "free writing" has utility as a composition tool; (4) journal writing can increase the fluency and self awareness of student writers; and (5) theorists and researchers studied the process of writing as well as the product. Yet an expected improvement in writing proficiency scores in the 1980s failed to materialize as composition teachers continued to emphasize grammar and structure at the expense of developing writing skills. Classroom teachers must be made aware that an extraordinary body of well-documented research has value and can contribute to a better way of teaching writing. There is only one way that students will learn to write and that is by writing. (Eighty references are attached.) (RS)