On the occasion of the twentieth-anniversary issue of the Public Opinion Quarterly in 1957, an issue devoted to twenty years of public opinion research, Herbert Hyman reflected on the existing state of theory about public opinion (Hyman, 1957; pp. 54-60). The state of theory was bleak, he commented, if one looked for a grand theory that integrated the vast empirical findings of past and contemporary studies of public opinion. Things looked brighter, however, if one looked for theoretical orientations of more modest scope, for example, theories of the middle range, a phrase suggested by Robert K. Merton (Merton 1949; pp. 4-5).