Abstract
Sociological debates about public opinion have to weigh up theoretical dilemmas. The first of these dilemmas concerns whether public opinion appeared only with the advent of modern society, or whether it was a phenomenon in other, earlier types of society.
The second dilemma is associated with the opposites of ((realistic}} and ((nominalistic}} approaches. The third dilemma is the contradiction between ((monistic}} and ((pluralistic}} positions.
However, if we go from the theory of public opinion to its survey, we find a surprising thing: all the mentioned theoretical dilemmas are irrelevant from the survey point of view. Researchers (pollsters) are basically satisfied with a concept of public opinion which can be described as ((operational}}.
According to this, public opinion is what is measured by polls of public opinion. The basic pitfall of the operational conception is that surveys which should capture and analyze public opinion can easily fit the purpose of their specific producer, as the method of surveying public opinion can be transformed into an instrument for shaping it.