Focusing on the problems of validity and representativeness of samples, a study examined the history of literacy in the United States since 1880 in order to set the contemporary debate on test scores, literacy and reading performance in the longer-range perspective of the last 100 years. The quality of the data and the arguments of literacy scholars concerning crude literacy, reading performance, and functional literacy were examined. Results indicated that although the problems of concept validity, representativeness of research samples, and noncomparability across time confuse the attempt to discern trends, three historical trends were identified. These trends are that (1) in the twentieth century, self-reported outright illiteracy almost disappeared as a percentage of the whole population; (2) the big story in twentieth century literacy is the rise in school attainment, not the relative effectiveness of schools to teach children at a particular grade level; and (3) the greatest difficulty was found in making a confident statement about the reading abilities of people at different points in time with the same amount of schooling. Findings suggest that present-day literacy policy should be argued on the basis of an assessment of the current condition and on the basis of shared educational goals and not on the basis of alleged declines or rises in literacy skills. (Five footnotes are included; 15 pages of references, 20 charts, and five appendixes of data are attached.) (RS)