INSTITUTIONAL RESEARCH ON PUBLIC OPINION AND PARTICIPATIVE DEVICES From polls to public debate

Considering that the methods and techniques implemented in social research entail a certain definition of the concepts studied, it seems relevant to reflect on the possibilities of devices other than polls when studying “public opinion” as a sociological concept. Two arguments for such approach can be highlighted: the qualitative and discursive nature of public opinion, and polls’ difficulties for being collectively perceived as a reliable and interesting way to express points of view about public matters. The French CNDP (Commission Nationale du Débat Public) is conceived for encouraging national, regional or local debates about political decisions which involve an important governmental investment. Taking into account that deliberative polling has probably been the most systematized among numerous research devices which have tried to implement the premises of deliberative/participative democracy, we will explore the possibilities of a different and wider representation of “public opinion” from the analysis of the dynamics generated by the institutional device of

ces put in practice by public policy "are not tools with perfect axiological neutrality, equally available: on the contrary, they are bearers of values, fueled by one interpretations of the social".From this perspective, the symbolic efficiency of "official" or "institutional" practices is relevant to offering a "proven" vision of reality.We are therefore interested in reflecting upon the devices which have been institutionalised by the public policy in order to gather and share citizens' points of view and opinions, given their apparent role in the creation of their "official representation".The study of the different dynamics generated by these devices offers wide-ranging possibilities for reflecting on the concepts which "make them visible" from the institutions and, therefore, on their political implications.In this sense, the Centro de Investigaciones Sociológicas (CIS) in Spain and the Comission Nationale du Débat Public (CNDP) in France represent fairly specific cases: activities carried out by both institutions are based on legislation currently in force and of an official nature, as opposed to any other country in our environment.Despite their various aims and theoretical assumptions, one aspect is especially interesting to our analysis: both of them implement processes which imply the use of mechanisms to ensure citizens' opinions and points of view "come to light".Our attention will be focused on the systematic production of opinion polls at CIS, the exercising of methodological reflexivity which this institution implemented upon carrying out the study "Social perception of surveys", and the observation of emergent deliberative/participative dynamics in Public Debate sessions presided over by the CNDP, while understanding that these reflect a more complicated concept of public opinion and reality than that linked to polls.

the concePtion of "PuBlic" deriving froM PuBlic oPinion Polls
Possibilities for implementing participative methods in which perceptions, visions or opinions held by individuals are able to be effectively incorporated into the decision-making process (far from being just mass media published "data") definitely involve an "active" conception of "public".and if this is the group supposed to hold opinion, it is relevant to take a close look at the concept of "public", which has historically gone through different considerations, in most cases not coincident with that of "public" emerging from opinion polls.even if there are several factors that can explain this process, it is interesting to reflect on the way the "channels" used to give visibility and describe social concepts have a certain capacity to shape their social representation 2 .
When the bourgeois concept of "public opinion" appeared in the 18th Century, the Enlightened elites, in the belief they were the only group in the population able to display enough sense and judgment to decide public issues, considered themselves as the sole representatives of "real public opinion" (Mañas 2005: 90).The press and French salons in which debates took place became the privileged and "legitimate" places where public opinion, and therefore, criticism of the political regime, could be expressed.as the masses were completely excluded from this "public" and their elitist forms of expression, we can conclude there was a hierarchical consideration of opinions3 : the masses are supposed to hold just popular, vulgar and ordinary opinions (Farge 1992), very different from those which emerge from deliberative processes inherent to what was considered at that time "the real public opinion" represented by the enlightened bourgeoisie (Champagne 1990: 45-46).We see that, initially, public opinion was born as a deliberative process among those who are considered reasonable enough to control governmental excesses.
However, once the revolutionary bourgeoisie goals were achieved, the politically non-recognized popular opinion gradually found more direct ways of expression.During the 19 th century, the emergence of street demonstrations as a way of collective popular action, the establishment of universal (masculine) suffrage, and the emergence of a widely available popular national press are circumstances that would lead to bourgeoisie resistance, fearful of one group, "the masses" who are quantitatively stronger than the elites.From the beginning of the 20 th century, those demonstrations started being instrumentalised by political parties and trade unions: the revolutionary pressure would enlarge the old select group of "intellectuals" who were the only ones with the right to represent public opinion, welcoming, from that moment, a wider part of the population: "the citizens".Popular opinion was finally incorporated into public opinion by changing its source of legitimacy: from reason to quantity, to the number of individuals who subscribe to it (Blondiaux 1998: 46).When the masses -at least symbolically-started becoming part of "public" opinion, a previously vaguely-defined concept resulting from discussion and reflection became an operative and measurable concept.Quantity displaced quality in the explanation of public opinion.
although the Habermasian sense of public opinion was less influential in the United States than in europe (Zask 1999), the extension of suffrage and mass media increased the interest in understanding public opinion.The control of Governmental actions by enlightened critical discussion and reflection would not be a central goal any more, and the emergence of Social Psychology during the first decades of the 20 th century would contribute to the definition of the psychological mechanisms -"attitudes", "stereotypes", "instincts"-that seem to take part in the opinion formation process, always -and this is very important-from a new vision of "public" as a "mass of individuals" that are barely informed and easy to manipulate from exterior instances such as propaganda (Lippmann 1922).as a result of the historical coincidence between this process and that of the quantitative development of social research techniques, considered "scientific" and, therefore, "neutral" and "impartial" (Porter 1996), the conception resulting from this "public-opinion" binomial would fit perfectly with political and economic power objectives.That is how the opinion polls technique, based on a statistical survey methodology, became, first in the United States and later in europe, the new objectivisation device for public opinion, its new channel for making it visible.The success of the discourse that legitimizes surveys and polls will imply the general consideration of their results as the social and institutional "reality" of opinion.
What are the consequences for the conception of "public"?In america, in the context of World War II, Sociology seemed to be much more interested in measuring public opinion than in reflecting on its definition as a sociological or political concept, that tendency being coherent with a dominant conception of population as a group of individuals that can be studied by establishing frequencies (Mañas 2005: 108).So, the opinion embodied by a "wise" public turns into a public opinion conceived as the mere addition of individual opinions.By the spreading of statistical polls, public opinion will be understood as an objective entity of reality, as true data obtained by a scientific procedure, whose original source is the sum of individuals.Opinion polls can successfully work because they assume that individuals' usual behavior is a reflection of what the public thinks about the topics presented in the questionnaire, that being measurable by a pre-coded and standardized question approach.Individual or group social actors can hardly be "active" if their discourse is replaced by previously fixed stereotyped statements in a closed questionnaire.

an uncertain linK. citizen ParticiPation and oPinion Polls in social discourse
The success of opinion polls comes not only from the social representation of the "scientific method of statistics" as a form of true and legitimate knowledge of reality4 .as explained above, it arises in a context of an extension of suffrage in which the "opinion poll" product needs to be presented to society as a useful and functional tool for western democracies.The popularity gained by opinion polls can only be understood through their link with publication in mass media (almazán 2009: 86).Thanks to this, the "public" could see a reflection of their "opinion", which now boils down to a measure representing social consensus, namely, a numerical summary of the point of view of the "majority" with regard to public affairs.and it will remain in citizens' perceptions as an autonomous form of authority, "public opinion", which "speaks" and "is demonstrated".Therefore, a core element of the legitimizing discourse of opinion polls was to highlight their qualities as a "functional" technique for democratic dynamics and, accordingly, as a tool for providing visibility for the opinions of citizens, who would find in them a channel of expression and, therefore, participation.The opinion poll machinery had become big business and it was important to highlight one of the most highly valued dimensions associated with its representation in a context of an extension of suffrage amid a mass consumer society: its potential as a democratic instrument.
From an institutional and academic standpoint, it seems that the key arguments legitimizing opinion polls as a method for researching public opinion have survived the major critical intellectual process that developed both in Europe and the US, starting mainly in the second half of the twentieth century.Despite the methodological objections that "day-to-day" aspects of the professional work involving opinion polls raise, they continue to this day to be a core activity in the public and private institutions that specialize in performing them systematically, using formats and procedures that are not very different from those used over half a century ago.However, one problem has been detected that affects the possibility of the surveys themselves: it is becoming more and more difficult to find people prepared to respond to the pollsters' questions, either through an outright refusal to answer any of the questionnaire or through the trend in raising the percentage of "non-response" in certain parts of it.In other words, it is difficult to find individuals who decide openly to participate in an opinion poll.This lack of predisposition -does it indicate that citizens lack motivation for taking part in public affairs, or does it have something to do with a more or less explicit reservation towards the tool used?What seems clear is that the situation raises the need to take a second look at the discourse claiming a democratizing role for opinion polls.One way of ascertaining the continuity or current meaning of this argument consists of researching the perception that society has of statistical surveys or, put another way, understanding their collective representation and observing whether polls' connection with democratic participation continues to be important in that perception.
The Centro de Investigaciones Sociológicas, since its creation in 1963 as "Instituto de la Opinión Pública" to the present day, is the Spanish public institution responsible for carrying out social research mainly through surveys, most of which are opinion polls.On several occasions this institution has shown an evident interest in the perception that Spanish society has of opinion polls through specific surveys.The first two polls were carried out in the early years of the democratic transition in Spain and the justification for them centered on connecting opinion polls as a method of social research to the recent democratic development of the country and to the levels of academic modernity of other european countries and the United States.The first of the polls was published in 1976 under the title "Las encuestas a encuesta" ("Surveys being surveyed") (García Ferrando and García Llamas 1976), and the second, entitled "Las encuestas a encuesta, de nuevo" ("Surveys being surveyed again") was published in 1981 (Justel 1981).It already became evident in these surveys that there was an interest in finding the key to understanding the absence of responses from individuals when taking part in opinion polls.It is this motivation that can be found explicitly in the most recent survey carried out by the CIS on social perception of opinion polls.It is another quantitative study, performed 25 years later, which was published with a similar title to the previous studies (alvarez and Font 2007).recognizing the difficulties faced by technicians in obtaining responses from the population on public opinion matters, it is considered that the "image of the poll" might be a variable explaining "non-response".apart from the different formats (face-to-face, telephone and internet) in which the questionnaire was carried out, the particular feature of this poll was its complementary function with a previous qualitative study, which was based on ten discussion groups5 .This initiative has particular relevance: the institution responsible for carrying out "official" opinion polls not only intends to submit their tools to public "consideration", as occurred in the early studies of 1976 and 1981.rather, it also wants them to be carried out using a qualitative method; one step further, in other words, for reflexive analysis of surveys as techniques for researching opinions.We consider that the empirical content of this qualitative study sheds much light on how to understand social perception of polls, and offers information that not only covers the usefulness of helping to draw up a questionnaire.
although the discourse of the ten groups that form part of the study provides a fairly complete picture of the social perception of polls, we will not address all the conclusions of the study here, but rather only those dimensions regarding the role of polls to form a citizen participation mechanism within a democratic context.

A series of conclusions on the collective perception of the polls-participation relationship
The qualitative study discussion groups mentioned were held by the CIS in the autumn of 2006 in various Spanish towns, chosen for the size of the environment, a year before the economic crisis exploded and, therefore, in a context of growth and a reduced rate of unemployment.The majority of the groups included communities in relatively-stable working and economic conditions, and only one of them was formed by young adults in unstable working situations.These are circumstances which probably affect the discussion and, therefore, are important to the interpretation of citizens' perceptions of the surveys.We shall now make a brief reference to two dimensions of the perception of polls: usefulness in general and as a participation mechanism.
For respondents, a poll is "useful" if they perceive that, by completing it, it will lead to the achievement of advantages or protection of rights relating to issues close to their daily, personal lives.and these issues are closely linked to the participants' identity as consumers and users.Only in these cases, carrying out a survey seems justified, and the weight of its negative connotations with regard to "manipulation", "wasting time", reluctance to "provide information", etc. which arise insistently in all groups, eases.The poll "is useful", therefore, when it is considered to form part of an "after-sale service" where the subjects can recognize their rights as "customers".and, in the case of public services, where no sale occurs, the usefulness is in relation to the individuals' involvement as users of these services that they regard as important in their daily lives: "health" (doctors, nursing, hospitals, waiting lists, etc.) and "transport" (the subway, buses, etc.) are often referred to as topics to which participants would be happy to respond: "Now, the Social Security, for example, if you're in hospital and you've had an operation, they send you a form afterwards, a survey, to check how you've been treated, what things you've seen that you dislike...And that is being responded to... (…)" G8, 7 ( 6) "(…) Something you use every day, 'Mercadona', for example: we're going to carry out a survey on whether the products are arranged well…something like that.So, well, you might say: OK then, I'll answer that because it just happens that I shop at that supermarket and I want them to improve the service" G6, 17 (7)   This prioritization of "personal and individual use" enables us to reflect on its counterbalance: an appreciation of "public matters" as a power that goes beyond individuals and their immediate environment and could, in turn, claim the right to be a potential beneficiary of the possible uses of a poll, is absent from the discourse.The constant references to the first person singular indicate that individuals do not seem to include themselves in the category of what is "public" unless public services that affect them directly are put into question.The image of the poll, which remains tied up in a "strategic" function, for either businesses or political parties, also transfers the perception of its usefulness to highly specific and individualized strategies focused on satisfaction, where the subjects perceive their status as "citizens" more as "customers" or "users" than as members of a group with shared stances, aims or demands.
This is why it is difficult to find a discourse on polls making reference to their usefulness as a citizen participation mechanism from a more collective standpoint.The usefulness of polls to enable the free expression of opinions on matters that go beyond merely private issues, or their explicit association with democracy, only appear indirectly in the general discourse.The only group where the role of opinion polls as a tool for citizen participation is specifically addressed is that formed by young people in a precarious employment situation (G5).But what does participation mean in this sense?The participants' statements show that "participating" consists of something that goes beyond issuing a vote in an election.It is understood more as an attitude of responsible and informed commitment to issues that affect citizens as a group.The importance attributed to this need for information and analysis relating to public affairs appears when the possibility of crea-ting social debate through published polls' data is pointed out.It is therefore assumed that the published data are not static -as if they were a "snapshot" providing evidence of the "state" of reality-but rather are capable of generating dynamics subsequently when shared and analyzed collectively.However, that perception comes up against a much more pessimistic view of the real extent of the possibility that polls can be perceived as a useful formula for channeling collective participation.The following remarks summarize this stance: "Due to a lack of commitment and there being nobody who should give you that vision.To get you to join in and not suck the information out of you.Instead, you should be told that this poll will be useful for this and for that...you should be told a little what it's about and, maybe that way, it might occur to you to say, hey, look, this is important" G5, 27 (8)   It is very interesting to hear the graphic expression used by one of the participants that is repeated in other remarks about "sucking the information out of you", as if it were symbolically a sort of "extraction" of something private, personal and valuable, which is assumed, but does not involve the participant's will.It is difficult when the survey is represented with a meaning of this kind for the respondents to add their own motivation to it.When the poll is represented as an interaction in which the respondents' only role is to expose themselves with resignation to the "extraction" of the requested information, agreeing to respond to the questions means "to put up with" the interview rather than to take part in it.Most importantly, this imaginary representation of the situation totally decontextualizes the themes covered by the poll.It seems that the intended subject matter of the poll eventually loses importance and "drowns" in the very act of carrying out the survey, the format of which in the end impose themselves on the content.experiencing the "poll situation" in such a way, and perceiving its "usefulness" as an instrument of political manipulation -this is a constant dimension in all the groups-it is difficult for the respondents to appreciate the value of the subject matter of the research and to perceive the connection between the performance of the survey and the possible beneficial results that could arise from taking part in it.These young people assume that surveys "are not responded to well" because, for this to happen, there would have to be a motivation for taking part, which would include (provided there was a reward) the respondent's involvement in the subject matter under study.Moreover, the perception of remoteness and disconnection with the visible results of the survey with regard to collective interests eventually generates a sense of "frustrated participation".The tone is one of resignation when discussing the usefulness of polls as a participation mechanism: "you don't see the results", "you don't see any of this", "there's no point to it".This explains the low level of commitment to polls, even if people agree to respond to them.It is therefore perceived as a very weak channel for participation, and it is assumed that the results will have little influence on political decisions."Something else" would be needed in order for these opinions to be taken into account by political powers: "What I do believe is that polls are a source of knowledge for those who carry them out.What I don't believe in is the effectiveness of this knowledge.The government receives more pressure from the survey in the streets on housing than 20,000 surveys of whatever sort.What has coerced them or forced them to take measures is not the result of the survey" G6, 40.
"The streets" are becoming the symbolic place of "effectiveness" with regard to citizens' demands, one which is also "feared" by the political powers9 .The question that, for the analysis, this comment provokes is the difficulty of finding a way to mediate between opinion polls as a device that renders an idea of a passive public and "the streets" as the prime setting for citizen action.This gives evidence to the fact that there are no generalised discourses on channels through which citizens participate apart from the "elections" or an old link with opinion polls, the public image of which is widely damaged.Specifically, they do not come up in discourses because they are not known by the average Spanish citizen.This is where different understandings of "opinion" and "public" should be put forward, along with research devices capable of responding to a more discursive content of the former and more active and participative content of the latter.a new type of critique to the traditional idea of public opinion has been developed from the 1980s, which points to not so much the technical shortcomings of the device for measuring subjectivity or the very conditions of possibility of "public opinion" emanating from the polls (Bourdieu 2000), but rather to a more political dimension based on a "deliberative" approach which entails a different way of understanding the "public".
This perspective arises from worrying about giving greater social legitimacy to public policy due to greater participation by citizens in the taking of decisions.This is a participation which, according to the theoretical focus of "deliberative/participative democracy", would imply the implementation of mechanisms beyond those of classical elections with two aims: on one hand, to allow citizens to have a say in the decision-making process, and on the other, that this expression would come from opinions which are well-informed and discussed, after a process of collective discussion and deliberation aimed towards the interest of all.as such, the idea is to have the democratic ideal of participation (Bryce 1987) overlap the "enlightened" ideal of deliberation (Habermas 1994).The implementation of this approach has been carried out using various formats, each of them implying a certain "formula" which seeks to combine these two principal ingredients: participation and deliberation.But what they clearly share is the skepticism to traditional polls as a way to obtain "the real public opinion".These devices effectively represent major changes or innovations in the traditional poll procedure, while not discarding it in certain stages of the process, as occurs in the deliberative polls (Fishkin 1995) and the experimental surveys (Brugidou 2008(Brugidou , 2010;;Grunberg, Mayer and Sniderman 2002), or they differ from them, such as in the participative budgeting, citizen's juries, consensus conferences, other hybrid forums (Lascoumes, Callon and Barthe 1997) or the public debate (revel et al. 2007).
It is remarkable that the idea of the "public" as a unique and indivisible group, with a low level of involvement, and holder of "one" "majority opinion", is put into question.From a sociological approach which aims to analyze social research methods and the way they could influence or "shape" the concepts studied, the devices that try to combine the principles of participation and deliberation would involve a notion of the public similar to that of a "collective", where it is understood that this concept refers to existential contexts and shared social discourses that, in turn, might differ from one to another.While not all collectives occupy positions of similar social legitimization, this does not prevent them from having their own views and, therefore, their own discourse on day-to-day events.The consequences of this theoretical perspective on the concept of "public opinion" are significant since it starts to be considered more a social than a psychological process that takes place not so much inside individuals as "among them".This is, as stated by P. Beaud and L. Quere (1990: 4) an intersubjective process in which the collective subject shares a situation, a social position, which gives rise to the existence of a shared world of feeling.Opinion does not take root inside the psyche of the individuals, rather the individuals become "subjects" capable of interpreting and grasping the phenomenon that they will evaluate based on the perspective of a particular social position, which is in turn mediated by various types of interest networks, ideological conditioning, pressure groups, etc.

Deliberative Polling and public opinion research: implicit assumptions
as Sturgis et al. (2005: 30) point out, perhaps the most ambitious deliberative formula to try to resolve the weaknesses of classic opinion polls has been the deliberative poll.Its proponent, J. Fishkin (1995Fishkin ( , 2001Fishkin ( , 2005) ) highlights it as the only "wide-scale" mechanism capable of overcoming the dilemma between political equality and deliberation.In contrast to other techniques which the author considers more limited due to deficiencies or the absence of statistical representativeness (self-selection of participants in participative budgeting or in public debate, insufficient samples within citizens juries or in consensus conferences), the deliberative poll will allow a sample of citizens which are statistically representative of the entire population to deliberate thoughtfully through face-to-face discussion on certain matters of general interest (Fishkin 1995: 200).The process, described in detail in Cuesta et al. (2008), implies the administration of the same survey at the start and end of an experience in which the participants debate, both in small groups and full sessions, the arguments for and against a specific matter of general interest10 .The deliberative poll was conceived to offer a measure of citizens' opinions on a certain matter, as long as they had the time, resources and opportunity to think and debate on the matter.a measure is offered comparing the results obtained the first and second times the survey was given.
The methodology applied in this deliberative poll showed certain implicit assumptions.Firstly, the combination of participation and deliberation was resolved by focusing on the epistemic value of the device in terms of the importance of the second principle.In fact, in Fishkin's texts, there are not as many explicit references to "participation" as to "political equality", understood as fairly representative of every preference (Fishkin, 1995(Fishkin, :58, 2005:40):40), which would be guaranteed simply by the statistical representativeness of the chosen participants.Therefore, the participative ideal is understood here as theoretical equality of opportunities for all to participate through a sample of participants statistically representative of the entire population.The self-selection or reduced samples found in other participative formats would therefore act as obstacles to achieving political equality.In any case, this principle appears to depend on that of deliberation: on various occasions, Fishkin (1995:66, 93) refers to how promoting participation or political equality was insufficient if this is not accompanied by a deliberative process: deliberation allows citizens to think about the power that they are exercising.Without it, there is no guarantee that this power will be used for the good of all.This idea allows us to observe a second assumption: that the need for an ad hoc "deliberative treatment" implies a certain idea of incompetence of the masses inserted into their real, vital contexts, to make decisions on public matters.This is precisely the argument that Fishkin makes, not only to criticise the uninformed opinions which come from traditional opinion polls but also to justify the need for creating an artificial context which, in fact, involves an experimental logic.This would solve the non-expert public's informational and deliberative deficiencies on public matters.
By establishing deliberation as a fundamental value, this mechanism clearly sets the conditions of its development: it must be the product of informed discussion in which all arguments for and against the matter are deliberated equally and which the participants must weigh up simply for their quality as such (Fishkin, 2005:285).In this sense, Bernard Manin's thesis, according to which the quality of deliberation is not ensured from a generic plural discussion where participants spontaneously deliver their points of view about the topic in question (Manin 2005a(Manin , 2005b)), maintains that a mere discursive exchange of a group willing to voice their opinions on a given issue is in danger of ignoring any points of view whose direction and outlook differ from the majority opinion of the group11 .
Considering that the quality of deliberation in the end depends not so much on the opportunity for an equal say for all of the participants, but rather on the guarantee that all the participants are informed about all of the arguments for and against the decision to be adopted, Manin believes that "contradictory" debate proceedings (based on presenting opposing arguments) have greater epistemic value (Manin 2005a: 250).That is to say they are more able to bring forth the "correct" decision after the deliberation process.Such is the approach which would preside ideally the dynamics of deliberative polling.
However, these assumptions pose problems for sociological analysis of public opinion.Firstly, the evaluation of the device appears to have been focused on measuring the change of opinion (Fishkin, 1995:290), in such a way that obtaining differing results between the first and second polls guarantees the success of the procedure: the deliberative "treatment"12 would have been successful.In this sense, considerations should be given to not only the experimental approach which measures the efficiency of a technique through pre-and post-treatment questionnaires but also to why the quality of a deliberative process must necessarily imply a change of opinion.It doesn't seem to have been considered that participants may have a priori opinions which are well-formed or which are anchored in solid foundations.In fact, it seems to overestimate the capacity of mere information, however balanced and deliberated it may be, to imply a change that makes an opinion thoughtful, stable and, therefore, "desirable".as rosell explained when presenting another deliberative approach, ChoiceDialogues13 , "for controversial issues, a critical role is played by beliefs, values, emotions, and personal experience in addition to information" (Fishkin and rosell, 2004: 56).Information is only a small part of the opinion formation process, and it is hard to assume that participants approach it in a "sanitised" manner without the ideological or affective intervention which always, more or less consciously, goes through any social position from which information is transmitted or received.as Livet pointed out (2007: 339), the appearance of emotions does not necessarily spoil a debate.It can actually reveal participants' positions and their effective relationships.Actually, a deliberative process based on a "contradictory debate" implies a quantitative premise when considering the possible opinions that an issue can raise: how can we be sure that all points of view in favour of or against have been put forward, in order to guarantee a "balanced" stance?The fact that a great many conflicting arguments are presented does not mean that all of them carry the same weight or have the same symbolic effectiveness.Individuals form their perception through the social position they hold, so the mere exposure to opposing arguments will possibly not be enough to introduce these arguments into articulate and credible discourse, especially if the issue has strong ideological or social connotations.Therefore one needs to ask what guarantee there is, for topics that have particularly obvious connotations and that raise deep-rooted ideologies or positions, that the "balanced" exposure of opposing arguments can lead to an equally "balanced" opinion, since we understand the latter to be the outcome of a rapprochement of the points of discord.Similarly, in these cases, we could ask if that outcome is really sustainable in ordinary life, or if it is only virtual, simply a result of the device being put into practice.

The institutionalization of a participative device: the Public Debate
Despite the development and systematisation achieved by the deliberative poll, this activity has no legislative basis and, as such, has not been institutionalised as a mechanism regularly used to gather the opinions of citizens.The interest in focusing on the French experience of Public Debate is due to its special feature as an institutional practice of participative democracy: its activity comes from a public authority body with a nationwide reach, which was created, develops and works as stipulated by current legislation and, as a result, represents another in a line of government activities.This legislative basis constitutes an important difference with respect to other participative devices.Public Debate arose in the creation of the National Commission for Public Debate (CNDP) 14 , in the Barnier law of February 2, 1995.This institution was set up with the aim of encouraging debate on "major public renovation projects of national, state and local interest that entail a strong socioeconomic or environmental impact" (revel et. al. 2002: 9), becoming an independent and "neutral" authority (revel, et al. 2007: 12) charged with impartially mediating between the decision-making authority and the public.
From the theoretical approach of the "participative/deliberative democracy" and, therefore, from a consideration of the "public" based on these premises, the concept of public opinion that stems from the Public Debate device can be understood as a communicational process that arises as a result of public deliberation and discussion, where the participants ("ordinary" citizens, representatives, experts, technicians and public authorities), making use of cognitive, rhetorical and discursive abilities, find somewhere to pose questions, offer replies and pose the questions again.In contrast to traditional polls or even deliberative polls, this perspective would imply to talk of "publics" in plural that exist as forms of collective coordination around a matter or problem on which they project their interests and concerns; "publics" considered collectives that discuss and express their opinions from concrete social positions.The principles of the CNDP, the specific nature and magnitude of the issue to be addressed, the fact that the process lasts for months, the resources that are required, and the infrastructure necessary for carrying out dozens of debate sessions make this mechanism hard to put into practice without public investment.although Public Debate does not have "taking a decision" on the project being debated as its objective, and although its results are not linked to political authority (Delaunay 2008: 31), it is undeniable that "a possible decision" constitutes a horizon implicit in the discussion.For those reasons, the public debate is subject to permanent tension between its supposed use in elucidating a decision and the awareness that, in fact, nothing is decided here.It may be that this ambivalent relationship with the final decision contributes to the mechanism working in practice, even when it is not possible to guarantee a consensus as a final result15 .as Blondiaux pointed out (2009), public debate can be simultaneously considered as a place of expression of the conflict and a search for a consensus.However, in both cases, it is a place which, theoretically, allows for the expression of the opinions and points of view of all of those "members of the public" who wish to attend, without any form of pre-selection.The attendees' implication and interest in participation (whatever quality it may be) seems to be guaranteed a priori, in contrast to the deliberative poll.
Ideally, Public Debate is conceived for creating dynamics that re-distribute political positions, in order to get a symbolic equality of participants.The device attempts to guarantee equity among the agents who take part in it through interactive communication and feedback of the points of view, in clear contrast to the dynamic set in other formats for registering the respondents' position on public affairs, mainly the opinion poll and even discussion groups.By comparing the dynamics that are generated ideally in four devices, we can analyze, in a theoretical and schematic manner, the various forms of understanding two concepts that emerge from them: the "opinion" (whilst being aware that this generic name is not generally used in devices other than polls), and the group -the "public"-to which this refers (Figure 1).
In the first two cases, as a result of the asymmetry that arises in the respondent-pollster and group-moderator relationships, the opinion flows excessively in one direction, i.e. that which goes from the individual or group to the registry mechanism, whether a questionnaire or a recording of the conversation.Because of the symbolic "belonging" of these mechanisms to the pollster and moderator, their function consists merely of generating the conditions in order for the "collation" of what is expressed by the participant(s).
Clearly, the conditions of this unidirectionality are different if we compare the dynamics of polls and discussion groups: in the latter device there are possibilities for debate that do not exist in the former, but this occurs among equivalent identity positions with regard to the issues of the debate.The moderator, the only attendee belonging to the outside group of the participants, remains theoretically outside the debate and only intervenes in it to provide order.In deliberative polls, spaces are generated -the so-called "deliberative forums" (Cuesta 2008: 67), in which participants, selected at random, may debate the matter at hand by following a scheme known as "contradictory debate".However, the formula used to "register" the opinion at the end -a postforum survey with questions identical to those in the earlier survey-implies that more attention is given to measuring the effects of the dynamic than to the generation of an opinion formed through the exchange and feedback between different social positions with regard to the problem being debated.In contrast, the public debate is conceived in such a way that the "results" of the process are not produced until the various groups, which represent various stances with regard to the issues in question, have confronted their points of view on several occasions, therefore having the chance to revisit and amend them, where appropriate, in order to produce arguments from different positions, considered relevant for a potential decision, even if that decision does not finally arrive.Consequently, the opinion theoretically flows in several directions and is constructed and prepared time and again as the various collectives take part.This dynamic and discursive conception of how public opinion is formed is certainly shared by other devices -citizens' juries, consensus conferences, citizens' assemblies-which assume a concept of public opinion quite removed from that of the "snapshot" provided by traditional polling.But holding dozens of sessions over several months provides the public debate with a relatively prolonged period in which the opinion can "mature" and settle during the process.

Emerging concepts from four research devices
Source: Prepared by the author Moreover, the concept of "public" presents particular nuances in each of the four devices considered in figure 1.In contrast to what occurs in opinion polls (the implications of which were referred to in the previous section) it would be possible to find in the discussion groups and even in the public debate a contextualized "public", in other words, participants able to issue their opinions from the social position they occupy with regard to the issue addressed, shared with other participants with whom they identify, and that the device is capable in making visible.In the discussion groups, this shared identity in a "group work" situation would facilitate discursive opinions that eventually express collective, as opposed to individual, points of view.In the public debate, these sessions, which are always open, do not imply the formation of heterogeneous groups which debate separately one from another.However, it must not be forgotten that, although the a priori objective is to record all of the relevant arguments with respect to the matter at hand, they are still discourses which come from specific and various social positions with respect to the question addresed (the relevant arguments are not infinite and never come from the "social vacuum") and which the participants can identify during the dynamic.In a deliberative poll, opinions do not necessarily come from groups which are socially contextualised since they are formed randomly between participants: as the small deliberative groups are socially heterogeneous, it is difficult to obtain a "positional" discourse that clearly identifies the collective holding a certain view or opinion about the subject to be addressed or debated.By trying to function as a necessary process to solve the supposed lack of information from citizens in their daily lives on public matters, the implicit concept of "public" in the deliberative poll even finds parallels with Lippmann's work on the uninformed masses manipulated by the mass media.From this perspective, therefore, the public is only potentially deliberative: it needs to be stimulated -thanks to the dynamic implemented-in order to "escape" from its daily passivity and thoughtlessness.
On the other hand, the "passivity" referred to in figure 1 with regard to the discussion group does not mean the absence of conversational activity; clearly, the group's discursive production is its reason for being.We refer to the fact that the group attendees do not have in mind an expression for the action, i.e. they do not expect subsequently any type of consequence relating to their remarks.The group expresses itself to enable things to be "known" but not to "act".In contrast, the overriding idea behind the public debate includes the concept of an "active" public as an essential element of its philosophy.The fact that the mechanism is only implemented in very precise and identifiable themes, that the attendees may participate in further meetings where they expect other positions, different from those with which the participant identifies, and that the latent horizon of a decision is always present -even if it is not actually reachable-provides a huge help for the participants to feel involved with the device.A concept of an "active" public in which it is considered that attending meetings, expressing, exchanging views and revisiting one's own views in light of all the arguments contributes to shaping a decision to be adopted, is also shared by other participative devices.However, in most of them the number of attendees is rather low in comparison to what public debate allows, due to a kind of prior selection formula.entrance to the sessions of public debate is totally voluntary and free, which can give rise to dynamics of a large audience.anyone can enter or leave the meeting room at any moment, without there being any stipends to receive (in contrast to what happens, for example, in some citizens' juries and in the deliberative poll).This is not without importance because this lack of prior selection could generate a specific sort of implication for some of the attendees, constituting a differential feature of public debate: the role of the participant as an attentive observer who does not speak or express him/herself in the discussion.
However, in which sense are these considerations supported by practical reality?The empirical analysis of the dynamic of the sessions generated by the Public Debate, by putting a concrete example into practice, provides us clues to understand if the real format of the sessions is compatible with the premises of the device.

The development of a debate dynamic: the case of the project to extend the Paris metro
On September 30, 2010 a public debate process, promoted by the CNDP, was launched.It was set to last four months (until January 31, 2011), in which time 75 debating sessions were held in various points of the region known as Île-de-France.The theme to be debated was the extension to the metro line, which would affect both the city of Paris and the suburban areas of the region.The debating process, for which there is a budget of six million euros, does not start from "zero"; what is put into motion, in truth, is a double, parallel debating process in relation to the two projects that are quite well developed but are at odds: the Grand-Paris project, supported by the government, with a focus on pursuing the connection and territorial promotion of business and technology centers, and the Arc-Express supported by the STIF transport union, which is more focused on improving the precarious transport conditions suffered by the "Franciliens" in their daily commute, mainly those who live in the suburban areas (banlieues).The CNDP organized some joint sessions in which representatives of both projects took part, mainly at the beginning and end of the process, but most of them are devoted specifically to debate each of the projects.Therefore, the CNDP appointed partial commissions, each responsible for one of the two projects.
as well as the points raised in the debate itself and the final outcome of it 16 , it is now interesting to observe some of the characteristics of the dynamic of the sessions.The inaugural session took place in Paris in a large auditorium, the amphitheater of the Porte Maillot Congress Center, which has seating for 3,500 people and where it was really difficult to find a spare place.What was most striking was that the only information that the debate would take place was released on the same day (September 30) through advertisements in local newspapers17 .Despite this, the auditorium was practically full (entry was free to anyone who wished to attend) and, in the course of around three hours that the debate took place, the hall remained practically full, without any visible spare seats until 11 o'clock at night.
How can a participation dynamic be managed in a session where, as well as almost 3,500 attendees, "absent" participants were able to join in by videoconference?I could immediately perceive that one of the professed basic principles for guiding the public debate -the theoretical equivalence-gave way in practice to a highly formalized dynamic, which followed a strict protocol for granting and measuring speaking time for attendees who, moreover, occupied different physical (and symbolic) areas for expressing their views.as can be observed in the photograph below, the participants' seating location in the hall established a clear differentiation among them: representatives of the CNDP, experts, technicians and top representatives of each of the projects in question were spread around two large tables (each with members of each of the two projects) facing the public and located on the stage of the amphitheater, thereby occupying the preferential speaking positions, not only because of their central location but also in terms of the speaking time granted to them during the meeting.The other attendees, the vast majority, were seated, as can be observed, in the rows of seats of the amphitheater, in a position that, except for certain participants who turned out to be elected representatives of certain collectives with a major interest and involvement in the subject (and who were sitting in the front rows near the stage) was more that of attentive observers -as mentioned above, very few people left the venue before the end of the meeting-than of active participants in the debate.
The member of the Commission chairing the debate introduced the various stages comprising it, which were scheduled perfectly, and granted time to each speech.The meetings I attended presented a structured dynamic in a very similar way.In the opening minutes, the member representing the CNDP took the floor to legitimize its function, explaining the aims and philosophy overriding the Commission and the Public Debate, insisting that its sole mission was to moderate a debate from a standpoint of neutrality, quality, transparency and fairness, without taking part in the debate itself.Following that, the people who were sitting at the tables on the stage (occupying leading representative posts on the projects, together with technicians and experts connected with the projects) explained what the projects consisted of, their guidelines, route map, etc., supported by a video on a big screen.Only after making this presentation, could the attendees who made up the "public" take the floor.They were told that they could only speak for up to "three minutes", taking into consideration the high number of attendees who could choose to speak.The support staff provided a microphone to whomever, respecting their turn, raised their hand with the intention of speaking.The speeches that occurred in this phase of the session consisted mainly of brief comments on the perceived needs in the participants' close, day-to-day contexts or, essentially, of questions made to one or other of the participants who formed part of the group of experts sitting at the tables on the stage who, as alluded to, took their turn to reply.One point attracting attention was that these comments and questions were often made by the same people, who introduced themselves as representatives of interested collectives or associations.In fact, as shown in revel's study on the dynamics of six Public Debate processes, the institutionalisation of the device has generated progressive standardisation in practices carried out (revel 2007: 249) and, therefore, a certain "typical structuring" of the role that each one of the groups implied in the debate assumes.The representatives that defend the perspective of specific collectives are found to be especially active.Therefore, two clearly differentiated attitudes can be perceived among the "non-expert" public attending the debate: on resolution cellular phone.The CNDP website has published higher-quality photographs and videos of the sessions.
the one hand, the active participants who made comments or posed questions, most of them as representatives of associations, collectives, etc. and, on the other hand, the observers who wished to be present up until the last minute to be informed.Moreover, participation was not only possible by attending the meeting; the CNDP provided a space so that anyone who wished to take part via the internet could do so and their comments were noted and analyzed subsequently by the members of the commission.This point is of particular interest as it could be the key to extending participation to anyone unable to attend or unwilling to speak in public or through a "representative".another interesting aspect is the presence of information stands for each project in the corridors next to the hall in which the debate was being held.Before entering or on leaving the hall, attendees had the chance to gather a substantial amount of material to consult when not at the session, which consisted of "information" brochures on each of the projects being debated, the basic function of which was to advertise the projects.The volume of the published information was surprising.Videos and photographs of all the sessions, a transcription of the debates, reports, preliminary conclusions, final conclusions, etc. are all in the public domain and may be consulted by anyone who so requests, free of charge.
The information seems guaranteed but what about the participation?It is difficult for the dynamic described here, as well as the conditions in which it can develop (we should not forget that, in some cases, thousands of people gathered at these sessions), to have an obvious correlation with the premise of equity which presides the development of the debate.as revel points out (2007: 245), this principle does not neutralise the elements that, independently of the arguments expressed, produce asymmetry in the position of the participants: social position, competences and other potential sources of domination.Interestingly, the procedure that was actually applied during these sessions is close to what B. Manin considers as the most satisfactory proposition for ensuring effective deliberation.The privileged statement of the experts representing each of the metro extension projects (each of the representatives offered different informative formats about the advantages of each project and the reasons why one should support them, as well as allowing temporarily a limited voice to the rest of the attendees) closely resembles Manin's description of experiences which can, in his view, lead to satisfactory deliberation: "(…) in order for satisfactory deliberation to occur, it is relatively irrelevant whether these arguments are introduced without any dialogue between the deliberants.One can imagine that firstly orators, experts, personalities, present their points of view and deliberately opposing arguments in front of the deliberating assembly, and then that the assembly deliberates" (Manin 2005a: 254).What seems to emerge from this reasoning is a clear differentiation between deliberation and participation, giving priority to the former with regards to the epistemic value of any device which may incorporate both.and if this can be applied to the dynamics previously described, public debate, whilst offering interesting opportunities for deliberation, have not managed to consolidate ways in which this deliberation can acquire a fully participative nature.

conclusions
The representations associated with the research devices are influential when preparing the concepts under study and, therefore, when shaping the perception of the participants about them.The analysis of CIS qualitative study demonstrates that the meager expectations held by the participants with regard to the ability of polls to be a participation mechanism and their lack of predisposition to respond to them indicate that the device has become detached from the argument that, in its beginnings, attempted to link it to democratic development.Opinion in the form of a percentage is offered as such a unitary and consensus figure that it is easy not to see the position itself reflected in it.The public is shaped more like an "audience" than an active collective, although that does not mean, as is also demonstrated in the study, that the participants lack opinions and arguments about current affairs.Nevertheless, there is no perception that the individuals' participation as poll respondents will lead to any advantage or use for them, bearing in mind that polls have become identified more with commercial or party political strategies than with social research techniques, whether or not they are accredited by an institutional initiative.Such identification owes more to the over-exposure to advertising and sales promotions (telephone and internet companies, banking services, etc.) with which individuals are usually bombarded, than to the actual content and nature of the process which is carried out in order to obtain information.The group discourse of the afore-mentioned study shows that individuals do not take part in as many actual surveys as is supposed, and that their experience is a great deal more related to those sales strategies (which are confused with surveys), thus conditioning their perception of them.However, leaving to one side what are only sales promotions (whose appearance in the discourse about surveys is completely spontaneous and almost automatic, although they are really not related to surveys), when individuals try to consider the ability of polls to produce data which reflects "the reality of public opinion", the conclusions are rather discouraging.Whether they be market studies made by private research institutes, such as social and, especially, political polls, carried out by private or public institutions (INe, CIS) 19 , lurking in the background during the whole discourse there is a "strategy" that is adopted only to benefit a specific business corporation, political party or government.What influence does this perception have on the validity of the polls carried out by research centres that rigorously follow "scientific" procedures?In light of the analysis of the discourse, it can be observed that although individuals believe in the neutrality and impartiality of "science", which in polls or surveys manifests itself as the representative method of statistics, they are extremely dubious about the credibility of its results when they are able to disassociate the scientific procedure from the motivating intentions of the study, always perceived as having commercial and political strategies.So although the epistemic value of polls obviously depends on the respect for statistical representativity, there are other factors that, particularly in the study of "public opinion", pose difficulties.On the one hand, it would be illusory to think that by merely applying the statistical method correctly, one is providing an "objective" and "neutral" device: drawing up the questionnaire entails establishing the categories that "consensually" one has deemed worth investigating.The choice of these categories -always likely to be stated discursively-implies a subjective act, no matter how serious, detailed or meticulous its creation process may have been.
On the other hand, the evident lack of credibility that these surveys suffer from, combined with the feeling that participation in them has limited usefulness for ordinary citizens, equates to a lack of motivation and indifference in the responses.This not only has a bearing on the no-response percentages, but also suggests that it is highly possible many of the responses are not valid, since due to the desire to shorten the interview time, they are provided without the consideration they deserve.These and other problems (inhibitions in giving a genuine response when dealing with especially connoted subjects, possible misunderstanding of the statements, etc.) raise serious difficulties for viewing the results of the polls as a "true" reflection of public opinion, even when using statistically representative samples, a carefully designed questionnaire and a statistical treatment of the data; in short, citizens' perceptions of the research method have an impact on the ultimate validity of its results.
On the other hand, finding a sample of statistically representative citizens is not equivalent to ensuring equal opportunities for all to participate.We understand participation as a process which is more complex and where there are various degrees and shades difficult to make equal.Some degrees will depend, among other things, on the social group of the subject, on the opportunities for collective expression generated during the debate dynamics led by a moderator, and, no less important, of the subject's involvement with the object of opinion and even with the device itself.although "self-selection" of participants, something which is common is various participative devices, is considered an obstacle by the deliberative poll's assumptions, it is possible that it is that "will to be involved", quite apart from any statistical selection or economic compensation, which guarantees a true participative opinion, one which does not bow to pressure or indifference, independently of the form that this participation takes Despite the question marks raised by Public Debates about whether their practical implementation fully corresponds with the theoretical guidelines of participative or deliberative democracy, we consider that the device itself has a high level of symbolic effectiveness in involving participants, although the degrees of "activity" in the effective participation vary.Although Revel points out that "ordinary citizens" are the ones that are silenced and absent in the debate, we consider that not participating in the interaction does not mean that taking part is not possible: "attend and learn" is probably the attitude of many citizens who turn up to the debate.although they may not express an opinion, they attend voluntarily -without expecting any economic compensation-and stay in order to form a better-informed opinion.This is a relatively spontaneous attitude which may be interesting for the generation, a posteriori, of commitment and social debate.In contrast to other mechanisms which, like the deliberative poll, select their participants a priori, voluntary attendance and paying attention without expecting any compensation is interesting for the cultivation of a society in which social debate on matters of public interest is encouraged.
The specific nature and magnitude of the questions to be debated, the legitimacy attributed to the process through the involvement of a public institution that moderates and guarantees the conditions of the debate, the existence of a culture of citizen participation (particularly in the case of France) that is buoyed by the implementation of initiatives such as this one, holding numerous, actual sessions, on-going publishing of information on the process, etc. represent reasons that probably incite more interest in participants to become involved than the dynamic established in the rules governing the debate meeting.This is why we talk of symbolic effectiveness with real consequences.regardless of the final conclusions, the public debating process projects an image of public institutions concerned about the public's opinions on subjects that supposedly affect them, assuming that, in effect, the participants have something to say and to contribute on this matter.and this image, without a doubt, encourages participation, even if it means simply attending as an observer or in order to read published information.
If the limited credibility of the surveys can affect their validity as a research method, we can likewise imagine that, to involve participants in public affairs will take more than channels to facilitate this involvement -social legitimization is needed.Participants have to perceive the importance of their active role, not only as conveyors of a "state of opinion" but also as agents whose discursive and argued opinions play a role in decision-making.
To achieve this, it is essential that the representations associated with the devices designed to achieve participative opinions are related to the perception of their importance and usefulness from a collective standpoint.The role of institutions in this could be important, since although the provision of channels and format is always a way of constraining spontaneity, it might also be a new way of discovering citizens' interest towards their active role in public affairs, beyond "election routines" as the only "standardized" context in which citizens feel that voicing their opinion has consequences.
oPening public opinion to deliBerative and ParticiPative Practices.a reflection on deliBerative Polling and PuBlic deBate Figure 1.Emerging concepts from four research devices

Figure 2 .
Figure 2. Participants' seating location in a Public Debate's session Paris, 30 th September 2010