Schools should be instructing students in formal thought and expression—what we call “comprehending”—rather than in everyday or “home” thought and language—what we call “understanding.” In this essay we suggest general changes in the standard reading and writing curricula. Finally, we examine the language of writing instruction, in college-level individual writing conferences, to take a close look at issues involved in implementing the curricula for higher and lower achieving students.

sources over the past several years (Templeton, 1969) . Literacy is no longer a luxury in this country; it is a necessity. Every child deserves a chance to become a skilled, competent reader. Our knowledge of the nature of the reading process and the acquisition of reading has increased noticeably over the past ten to twenty years, largely as a result of government funding of basic research on reading (Levin & Williams, 1970; Kling, 1971; Kavanaugh & Mattingly, 1972). To be sure, no adequate review of this progress is currently available, and the impact of these research findings on classroom practice has been minimal. As recently as this year (1974), a review of the psychology of reading introduces the area of research on reading acquisition: Despite all the current emphasis on literacy, the wealth of 'programs' commercially available, the 'learning specialists' who have set up in shopping centers and the arguments over phonics or whole word methods, it is the beginning phase of learning to read that we seem to know least about. All the talk is of what the teacher does or should do and not of what happens or should happen in the child. This is a very peculiar situation. There is presumably a learning process going on, but it is a rare psychologist who studies it. (Gibson Zc Levin, 1975, p. 264) Large amounts of money continue to be poured into the development and evaluation of competing reading curricula, with outcomes that are disappointing to say the least (Bond & Dykstra, 1967; Corder, 1971) . With few exceptions, these evaluation projects have fallen far short of minimum standards of experimental research in the behavioral sciences (Corder, 1971) . There is little one can learn from bad data. It is not surprising to find, on reanalysis, that the major outcome of the large First Grade Cooperative Reading study was the discovery that children of high IQ were more successful in learning to read than children of low IQ (Lohnes & Gray, 1972) . There have been at least three recent major efforts to synthesize the research