"The opinion public not exist. " Pierre Bourdieu This is a classic text Pierre Bourdieu, a brilliant and provocative, indispensable for anyone who wants to understand how to "manufacture" public opinion. Bourdieu's assertion that public opinion "there" is far from aporística: Highlights it is above all a fiction, a construct of the media and pollsters, which is built on false assumptions. Reduce "the opinion of a country" to a figure, say for example that a certain percentage of Colombians "think something" is a spurious claim.
The argument is developed from three observations on the opinion polls that reveal its inherently limited, usually handled as follows:
1. Polls mean that everyone should have an opinion about something, which is not true.
2. Also assume that all opinions are of equal worth. That is, part of the false premise that all opinions have the same real power.
3. Simply ask the same question everyone is an arbitrary definition (manipulated) the field which is the subject of inquiry, ie it assumes that there is a consensus on key issues. So is the way it asks the question, which often involve bias.
Good reading.
Public opinion is not
statement made Noroit (Arras) in January 1972 and appeared in Modern Times , 318, January 1973, pp. 1292-1309. Incorporated in Issues Sociology , Paris, Les Editions de Minuit, 1984, pp. 222-235.
I would like to clarify first that my purpose is not to mechanically terminate and easy to opinion polls, but to conduct a rigorous analysis of their operation and their duties. This implies that we put in question the three assumptions incurred by implication. Any opinion survey assumes that everyone can have an opinion , or, in other words, the production of an opinion is within the reach of all. Collision leaves one feeling naively democratic, I would argue with this first assumption. Second assumption: we assume that all opinions are equal. I think we can demonstrate that it is not so and that to accumulate opinions that have not even the real force tends to produce artifacts meaningless. Third assumption implicit in the mere asking the same question everyone is involved the assumption that there is a consensus on issues, that there is an agreement on issues that deserve be asked. These three assumptions imply , I believe it, a whole series of distortions that occur even when all conditions of rigor methodology are met in the recollection and analysis of data.
is very often to opinion polls blame technology. For example, it calls into question the representativeness of samples. I think in the current methods used by offices for the production of surveys, the objection is not valid. They were also accused of asking leading questions, or rather to bias issues in their formulation: this is already true and it happens often that we induce response through the asking question. Thus, for example, violating the basic tenet of the construction of a questionnaire that requires "let their opportunity" to all possible answers, it frequently fails in the questions or the answers provided one of the options , or it has several times the same option in different formulations. There are all sorts of biases of this type and it would be interesting to consider the social conditions of emergence of these biases. Most of the time they take the conditions in which working people who produce the questionnaires. But they are mainly due to the fact that the problems that manufacture institutes opinion polls are subject to a request from a particular type. So, having undertaken an analysis of a large national survey of French opinion on the education system, we found in the archives of a number of offices, all matters concerning the teaching. This has shown us that over two hundred questions on the education system have been laid since May 1968, against less than twenty between 1960 and 1968. This means that issues which apply to such organizations are deeply related to economic conditions and dominated by a certain type of request social. The issue of teaching by example can be raised by an institute of public opinion only when it becomes a political problem. We see immediately the difference between these institutions of research centers that generate their problems, if not in a clear sky, at least with a much greater distance with respect to the social demand as it directly and immediately.
A statistical summary of the questions made us see that the vast majority of them were directly related to the political concerns of "Personal politics". If we had fun tonight playing at small papers and if I told you to write five questions that you feel are most important in education, surely we would get a list very different from what we obtain by raising questions that were actually asked by opinion surveys. The question: "Should we introduce the policy in the schools?" (Or variants) has been asked very often, whereas the question: "Should we change the software? ; " or "Should we change the mode of transmission of content?" has rarely been asked. Also: "Should we retrain teachers?". All these questions are very important, at least in a different perspective.
The issues that are proposed by the opinion polls are subordinated to political interests, and it controls very strongly to both the meaning and significance of the responses is given publication of results. The opinion poll in the state, an instrument of political action, his most important function is perhaps to impose the illusion that there is a public opinion as purely additive summation of opinions individual, to impose the idea that there would be something like the average of the average opinion or opinions. The "public opinion" which is manifested in the front pages of newspapers in the form of percentages (60% of French people support ...), this is a public artifact outright whose function is to conceal the state of opinion at some point of time is a system of forces, tensions, and it is nothing more inadequate to represent the state of the opinion that a percentage.
We know that any exercise of force is accompanied by a discourse to legitimize the power of those exercises, we can even say that the own any power relationship, is to have its full force only to the extent where it hides itself. In short, to put it simply, the politician is one who says: "God is with us." The equivalent of "God is with us" is now "public opinion is with us." This is the fundamental effect of the opinion survey: the idea that there be a unanimous opinion, therefore legitimize and strengthen political power relations which are based or make it possible.
Having initially said that I meant to finish, I'll try to tell very quickly what are the operations by which it produces this effect consensus . The first operation, which has its starting point the premise that everyone should have an opinion, is to ignore the non-response. For example you ask people: "Do you support the government Pompidou?" You save 30% of non-responses, 20% yes, 50% non. You can say, from people is negative than positive from people and then there is the residue of 30%. You can also recalculate the percentages for and against excluding non-responses. This simple choice is a theoretical operation of a great importance on which I would like to reflect with you.
Eliminate non-response is to do what we do in an electoral process where there are blank or invalid, it is imposing on survey of opinion philosophy implicit electoral survey. Looking more closely, we observe that the non-response rate is higher in general among women than among men, the gap between women and men is even higher than the problems are of a more strictly political. Another observation: the more an issue relates to matters of knowledge, understanding, the gap is wide between the rates of non-responses from more educated and less educated. Conversely, when the questions relate to ethical issues, changes in non-response by level of education is low (example: "Must be tough on children"). Another observation: the more a question raises conflicting issues, concerns a knot of contradictions (an issue on the situation in Czechoslovakia for the people who vote Communist), plus one question generates voltages for a given category, the more non-responses are common in this category . Consequently, simple statistical analysis of non-responses provides information on what is meant by the question and also on the relevant category, the latter being defined as much by the probability associated with it have opinion that the conditional probability of having a favorable or unfavorable opinion.
The scientific analysis of opinion polls shows that there is practically no problem bus, no matter who is reinterpreted according to the interests people to whom it is asked, the first imperative is to wonder what question the various categories of respondents believed respond. One of the most pernicious effects of the opinion survey is precisely to put people on notice to answer questions they have not asked. Consider for example the issues that revolve around issues of morality, whether questions about the severity of parents, relations between teachers and students, pedagogy directive or nondirective, etc.., Issues which are of the more perceived ethical problems we descend further into the social hierarchy, but which may be political problems for the upper classes: One effect of the survey responses is to transform ethical policy responses by simple effect of taxation issues.
In fact, there are several principles from which we can generate a response. First there is what one might call the political competence by reference to a definition that is both arbitrary and legitimate, that is to say dominant and concealed as such policy. This political competence is not universally popular. It varies roughly as the level of education. In other words, the probability of having an opinion on any issues involving political knowledge is quite comparable to the probability of going to the museum. We observe great differences: where such a student engaged in a leftist movement receives fifteen divisions to the left of the PSU, for a middle nothing. In the political scale (far left, left, center left, center, center-right, right, far right, etc..) That investigations of "political science" use for granted, certain categories Social intensive use a small corner of the far left, others only use the center, others use the entire scale. Finally an election is the aggregation of very different spaces; we add people who measured in centimeters with people who measure in kilometers, or rather, people who note from 0 to 20 and people with reported between 9 and 11. Competence is measured among other things the degree of fineness of perception (the same thing in aesthetics, which can distinguish some five or six ways successive one painter).
This comparison can be pushed further. In terms of aesthetic perception, there is first a permissive condition: it requires that people think the artwork as a work of art; Then, having seen as a work of art, they must have categories for collection for the building, structure, etc.. Suppose a question asked: "Are you for an education directive or nondirective education?" To some it may be constituted as political representation of parent-child integrated into a vision Systematic society, for others it is a pure question of morality. Thus the questionnaire that we developed and in which we ask people if, for them, it's politics or not to strikes, to have long hair, to participate in a pop festival, etc.., shows great variations between social classes. The first condition to adequately address a political issue is to be able to establish as the policy and the second, having established a policy, is to be able to apply specifically political categories that may be more or less adequate, more or less refined, etc.. These are the specific conditions of production of opinions, those opinion survey assumes universally and uniformly filled with the first assumption which anyone can produce an opinion.
Second principle from which people can produce an opinion, what I call the " class ethos" (for do not say "ethics class), that is to say a value system implied that people have internalized from childhood and from which they generate answers to problems that are extremely different. The opinions that people can share the output of a football match between Roubaix Valenciennes owe much of their coherence, their logic, the class ethos. A variety of responses which are considered policy responses, are actually produced from the ethos of the classroom and at the same time can take a meaning quite different when they are interpreted in the political arena. Here I must refer to a sociological tradition, especially popular among some sociologists of U.S. policy, which very commonly speak of conservatism and authoritarianism of the working classes. These arguments are based on international comparison surveys or elections which tend to show that whenever we question the popular classes in any country whatsoever, on issues concerning the relations of authority, personal freedom, freedom of the press, etc.., they respond more "authoritarian" than other classes, and overall we conclude that there is a conflict between democratic values \u200b\u200b(in the author which I think Lipset, this American democratic values) and values \u200b\u200binternalized What the popular classes, values \u200b\u200bof authoritarian and repressive. From there, it takes a kind of eschatological vision: raise living standards, raise the level of education and, since the propensity for repression, authoritarianism, etc.., is linked to low income, low education levels, etc.., and we will produce good citizens of American democracy. I think what is at issue is the meaning of responses to certain questions. Suppose a set of questions like the following: Are you in favor of equality between the sexes? Are you in favor of sexual freedom of the spouses? Are you in favor of a non-repressive education? Are you favorable to the new company? etc.. Suppose another set of questions like: Do teachers have to strike when their position is threatened? Teachers have to be in solidarity with other staff in times of social conflict? etc.. Both sets of questions provide answers strictly inverse structure in respect to social class: the first set of questions concerning a certain type of innovation in social relations in the symbolic form of social relations, raises Answers more favorable than it rises in social hierarchy and the hierarchy by level of education and, conversely, the issues that relate to real transformations of power relations between classes elicit responses more negative with height in the social hierarchy.
short, the proposition "The lower classes are repressive" is neither true nor false. It is true insofar as, to a range of problems such as those related to domestic morality, relationships between generations or between sexes, classes tend to be far more legalistic than other social classes. Instead, issues of political structure, which involve the preservation or transformation of social order, and not just conservation or transformation of modes of relationship between individuals, classes are much more favorable novation, that is to say to a transformation of social structures. You see how some of the problems in May 1968, and often ill-posed, in the conflict between the Communist Party and leftists, are related very directly to the central problem that I tried to ask this evening, that of the nature of responses, that is to say, the principle from which they are produced. The opposition that I made between these two groups of questions boils down because of opposition between two principles of production for the views: a purely political principle and an ethical principle, the problem of conservatism of the working classes being the product ignorance of this distinction.
The effect of taxation issues, effect exercised by any opinion poll and any interrogation policy (starting with the election) results because the questions asked in an opinion poll are not really issues that arise in all respondents and that responses are not interpreted in terms of the problem against which the various categories of respondents actually responded. Thus the problem dominant , whose list of questions for two years by pollsters provides an image that is to say that the problem mainly concerns the people who hold power and who intend be informed on how to organize their political action is unevenly controlled by different social classes. And, importantly, they are more or less able to produce a problem-cons. About the televised debate between Servan-Schreiber and Giscard d'Estaing, an institute of public opinion polls had asked questions like: "Is that academic success is dependent on donations, intelligence, Labour merit? "Responses to actually deliver information (unknown to those who produced them) on the degree to which different social classes are aware of the laws of inheritance cultural capital: membership of the myth of the gift and ascension through the school, school of justice, fairness in the distribution of posts according to titles, etc.. is very strong in the classroom popular. The cons-problem may exist for some intellectuals, but it has no social force although it was echoed by a number of parties, groups. The scientific truth is subject to the same laws of distribution than ideology. A scientific proposition is like a bull of Pope on birth control, it preaches only to the converted.
It combines the idea of \u200b\u200bobjectivity in an opinion poll in asking this question in more neutral terms to give every chance to have all the answers. In fact, the opinion survey would probably be closer to what happens in reality though, completely violating the rules of "objectivity," we gave people the means to locate as they actually are in actual practice, that is to say with respect to the views already expressed, if, instead of saying eg "There are people friendly to birth control, others are negative, do you? ... ", it outlined a series of explicit policy positions mandated groups to form opinions and disseminate them so that people can be located relative to the responses already made. We commonly speak of "positions" there are positions that are already planned and they takes. But we do not take the chance. We take the position that one is predisposed to take based on the position one occupies in a certain field. A rigorous analysis aims to explain the relationship between the structure of positions taken and the structure of the field occupied positions objectively.
If the opinion polls very poorly capture the virtual state of public opinion and more accurately the movement of opinion is, among other reasons, the situation in which they apprehend the opinions is quite artificial. In situations where there is public opinion, especially crisis situations, people are made to the opinions, views supported by groups so that choosing between opinions, it is very obviously choose between groups. This is the principle effect of politicization that produced the crisis: we must choose between groups who define themselves politically and to define more and more positions according to principles explicitly political. In fact, what seems important is that the opinion survey treats the public as a simple sum of individual opinions, collected in a situation that is at the bottom of the compartment where the individual will stealthily express an opinion in isolation alone. In real situations, opinions are forces and reports opinions are conflicts of power between groups.
Another law emerges from this analysis: we have even more opinions about an issue that we are more interested in this problem, that is to say that one has more interest in this problem. For example the education system, the response rate is very closely related to the degree of proximity to the educational system, and the probability of having an opinion depends on the probability of having power over this about what they opine. The opinion asserts itself spontaneously, it is the opinion of people whose opinion carries weight, as they say. If a Minister of Education acting according to a poll (or at least from a superficial reading of the survey), it would not do what he does when he really acts like a man politics is to say, from phone calls he receives the visit of such a union leader, like Dean, etc.. In fact, it acts against these forces of opinion actually declared that outcrop on its perception that since they have strength and where they have strength because they are mobilized.
regards to predict what will happen to the University in the next ten years, I think the mobilized opinion is the best basis. However, the fact, attested by the non-response, that the provisions of certain categories do not access the status of opinion, that is to say consisting of speeches claiming consistency, claiming to be heard, s impose, etc.., should not conclude that in situations of crisis, people who had no opinion will choose at random: if the problem is politically constituted for them (problems of wages, work rate for workers), they will choose in terms of political competence, whether it is a problem which is not created for them politically (repressiveness in the relations within the company) or is being formed, they will be guided by the provisions deeply unconscious system that guides their choices in the areas most different from the aesthetic or sport to the economic preferences. The investigation of traditional opinion ignores both the groups and the provisions virtual that can not be expressed in terms of explicit discourse. That is why it is unable to generate any reasonable prediction about what would happen in a crisis.
Suppose a problem like the education system. We may ask, "What do you think of the policy of Edgar Faure?" This is a very close election of an investigation, in that it is the night where all cows are black: everyone agrees roughly without knowing what, we knows what was meant by the unanimous vote of the law Faure to the National Assembly. Then asked: "Do you support the introduction of politics into the schools?" Here, there is a clear distinction. It's the same when asked: "Teachers can they strike?" In this case, members of the working classes, by transferring their political competence particular, know what to say. One can still ask: "Should we change the curriculum? Do you support continuous testing? Do you support the introduction of parents in the councils of teachers? Do you support the abolition of the aggregation? etc.. "Under the question" Are you in favor of Edgar Faure? "there were all these questions and people took a stand at once on a set of problems that a good questionnaire could arise only by means of at least sixty issues where there is variation in every direction. In one case the views are positively related to the position in the social hierarchy, in the other, negatively, in some cases markedly, in other cases weakly or not at all. Just think of that electoral process represents the limit of a question like "Are you in favor of Edgar Faure?" To understand that specialists in political sociology can be noted that the relationship usually occurs in almost all areas of social practice, between social class and practices or opinions, is very weak when it comes to electoral phenomena, so much so that some do not hesitate to conclude that it does There is no relationship between class and the social act of voting for right or left. If you have in mind an electoral process poses one question syncretic what could reasonably take that two hundred questions, each measured in centimeters, the other in kilometers, the strategy of candidates is wrong to ask questions and to play up divisions on concealment to win floating votes, and many other effects, you find that may need to ask the traditional question of the relationship between voting and social class and wonder how it is that there is still a relationship, albeit weak; and examine the function of the electoral system, an instrument which by its very logic, tends to mitigate conflicts and cleavages. What is certain is that by studying the functioning of the opinion poll, you can get an idea of \u200b\u200bhow does this particular type of opinion survey that is the election and the effect it produces.
short, I really wanted to say that public opinion does not exist, as in any case shall give those who have an interest in asserting its existence. I said that there was one hand opinions formed, mobilized, pressure groups mobilized around a system of interest explicitly formulated and on the other hand, provisions which, by definition, do not mind if we meant, as I have done throughout this analysis, something that can be formulated in discourse with some claim to coherence. This definition of opinion is not my opinion on public opinion. It is simply the clarification of the definition that implement public opinion polls asking people to comment on opinions expressed in producing and, by simple agrégation statistique d'opinions ainsi produit, cet artefact Qu'est l'opinion publique. Je dis simplement que l'opinion dans l'acception implicitement post par ceux qui font perm Sondages d'opinion ou des ceux qui utilisent les résultats in, je dis simplement that opinion-là cette n'existe pas.
The original link to the text they find it by clicking here
