When I was a student struggling to become sophisticated in sociology, one of the continual I had in reading theory was that I agreed with practically everybody I read. I read Tocqueville and found him persuasive; I read Edmund Burke, found him persuasive; read Marx, found him persuasive too. Comte, Spencer, Durkheim, Weber-all full of illuminations for me. But gradually, one became aware of schools and traditions, of hidden implications and strategic omissions. Burke and Tocqueville were Conservatives; Marx was a revolutionary. Durkheim was a functionalist, Weber a nominalist. One simply couldn't agree with all of them. The graduate student subculture fixed that. One of the things a good graduate student subculture does is give students vested interests in taking positions and defending them; by making these names household words, it sensitizes students to crucial differences between thinkers, and therefore enables them eventually to confront a text with some semblance of critical perspective (background, it's more often called), so that each writerthinker doesn't have to be taken naively, in his own terms. Reading Mel Kohn, I find myself agreeing with everything he says. Maybe it's my lack of background in problems; but surely such agreement is not permissible, particularly under conditions which constrain me to perform in the role of Discussant. Still, it's difficult not to agree with his appraisal of progress in the field of research, and his consideration of missed opportunities and future possibilities. I think he's right that the past 25 years have seen a dramatic enlargement of research perspectives, and I think his illustration of these changes in research on sex roles, race relations, poverty, and crime are fairly drawn. I think too that the conceptual convergences he points to across the different areas of research (from descriptive to analytic designs, from concern with individuals to concern with institutional contexts and other structures, a sense of the interconnectedness of problems) and his ambitious hopes for future directions in research promise an increase in the rigor of research sufficient to bring such research up to methodological par for the rest of the sociology course. Well and good. It seems likely that we are more sophisticated, more knowledgeable today than 25 years ago about the conditions and consequences of sex role behavior, race relations, poverty, crime, and other features of the landscape. But (here come the buts ) I kept asking myself while reading Kohn's paper: What has all this got to do with the very conception of social problems? Like any other field of scientific enterprise, the more thinking and the more research we do, the more dialogue and mutual attention to and criticism of each others' work, the greater the illumination that's likely to be cast on the questions about crime, race, poverty, and sex that puzzle us. But what have more or better research on these matters got to do with the field of social problems? Crime, race, sex, and class are all more or less well established realms of sociological research, and, by invoking the traditional distinction between facts and values, it is possible for us to go on becoming wiser and more learned about them without ever talking about social problems or conceiving ourselves as experts in problems. To concern oneself explicitly with social problems (as distinguished from race, crime, or poverty) is to concern oneself with their solution or amelioration (and therefore with the relation to policy or reform). It is also to raise the question of how it is (to use Mel Kohn's brief definition) that some social phenomena that have a seriously negative impact on the lives of a sizable segment of the population are commonly regarded as social problems, while others (for example loneliness and Sunday afternoon football widowhood) are not. I