The number and reason

The twelve number of Public reason offers a stimulating reflection on the “ collective wisdom ». This theme is of the great interest of registering for the confluence of political sociology and philosophy: the first classically questions the competence of the voter, and its social presuppositions, while the second, in a normative perspective , questions from Plato and Aristotle the ambiguous relationships between number and reason. American political sociology, including this issue of Public reason presents a certain number of contributions, focuses on the debate between the proponents of the paradigm of theHomo yeconomicus To those who dispute both the relevance of this approach to account for political behavior, and the incompetence of the greatest number that the former infer from the absence of immediate interest, “ self-centered », Likely to motivate political involvement and participation. Such a way of dealing with the advantages of democracy is of the interest of offering another look at this question than that of French political science. If democracy is justified, it is thus, as shown by a certain number of contributions, because it has an epistemic superiority: the large number, as Aristotle already claimed, is more likely to have reason that a limited group.

The reason for the number

The main argument made by Hélène Landemore, Professor in political science at the University of Yale, in favor of increased rationality of the decisions of a large number of people with regard to those of a small group of experts is the “ cognitive diversity “-that is to say the” Ability to see the world from different points of view “(P.10) – which, according to her, compensates for the lack of information affecting a major group compared to a reduced group. This argument, according to H. Landemore, is valid for the polls preceded by a deliberation as well as for the elections that no discussion of the voters prepares. H. Landemore is based on the acquired statistics and thus summons the paradox of the jury, which establishes only if, during a vote, the voters have more than one in two chance to choose the best of two options, that ‘They vote independently of each other, and in all sincerity, the probability that the majority of the best option is greater than that that an individual member choose it ; This probability, moreover, grows at the same time as the group’s workforce. THE “ Miracle of aggregation “, For its part, corresponds to the statistical phenomenon according to which the average of the responses given by the members of a group to a specific question relating to a quantitarian evaluation is statistically exact, the erroneous responses tending to cancel each other. The most famous example is that of Galton, which asked the population of a village to estimate the weight of a beef ; The average responses proved to be exact, at the nearest 500g. In addition to the objections to which these two statistical rules come up against, we can regret, taking into account the importance that H. Landemore attaches to this phenomenon of “ Miracle of aggregation », That she does not explain how it can apply to elections which consist of the aggregation of discreet choices, and not of quantified estimates (p. 35).

Political information and competence

The discussion that H. Landemore then engages with Bryan Caplan (professor of economics at the George Mason University) on the question of the competence of the voter echoes, to a certain extent, to the controversies experienced by French political science. H. Landemore argues that information is not synonymous with competence: empirical research which claims to establish the incompetence of the voter presuppose a certain conception of political competence, that of specialists. Bryan Caplan retorts that by refusing all the ordinary criteria which make it possible to measure political competence without proposing new ones, it becomes impossible to establish the epistemic superiority of democracy ; Indeed, it cannot be measured if one does not agree that political questions are subject to exact or erroneous answers. Note that almost all examples of Caplan are borrowed from the economy area, and relate to questions that admit only one exact answer ; However, Caplan also disputes, without developing this crucial point, that the absence of information does not biaise the decision if one limits ourselves to question the voters on the “ big choices “(According to the expression of sociologist Arthur Lupia, who maintains that truly political questions, relating to society choices, do not require any particular knowledge), even though they are a matter of moral choice rather than knowledge.

This tight and stimulating discussion on the link between information and competence continues between Bryan Caplan and Gerry Mackie, professor of political science at the University of California. This one attacks the economic approach of political behavior, in particular the concept of “ rational ignorance (Formulated by Anthony Downs, one of the pioneers of this approach): it is rational not to learn, given the infinitesimal impact of a vote on the result. Paradoxically, it is therefore because the voters are rational that they are incompetent. Given, however, electoral participation rates in Western democracies, we must take into account other motivations than those to influence the result directly to understand that individuals vote, such as the desire to express a preference. In other words, it is necessary to break with the theory of the voter only matured by his selfishness and engaging in a purely instrumental calculation when voting. In addition to this main thesis, G. Mackie also challenges the postulate of incompetence of the average voter, by mobilizing well -known arguments of contemporary political sociology: the voter, even if he would lack information, would make use of use Heuristic shortcuts – shortcuts -, such as the presentation of the candidates, their vocabulary, etc., to establish a summary but sufficient mapping, political positions.

Maximize collective wisdom

According to Adrian Vermeule (John Watson Junior Professor of Law to the Harvard Law School), arguments in favor of collective rationality must be used to promote the American Congress against the growing grip of the Supreme Court. A. VERMEUX in fact draws three normative consequences: first, the maximization of epistemic quality requires a broad reassignment of the legal authority of supreme courses to parliaments, due to their larger size and cognitive diversity. Then, in order to give its entire measure to the logic of the jury, it is necessary to increase the number of representatives at the American Congress: in addition to the gain of cognitive diversity that this increase would allow, vermale affirms, relying on the statistical calculations of List and Goodin, that if the average competence of a deputy is 0.51 and that there are two possible choices, raising the number of representatives from 101 to 601 members would allow collective competence to pass from 0.58 to 0.69 . Finally, the Supreme Court should include members who are not lawyers, in order to “ produce benefits in terms of both the group’s competence and diversity (P. 199).

Jon Elster, professor at the Collège de France, seeks to determine the best conditions for drawing up a constitution: a relatively large assembly, in addition to the epistemic advantage it involves, also makes it possible to minimize the role of interests, By avoiding giving too much weight to the haggling and the exchange of votes. J. Elster finds a concern specific to the proponents of the economic paradigm (with which, however, he distances himself) when he specifies that the assembly, however, should not be numerous to excess, so as to avoid only Representatives renounce to learn, given the low weight of each vote.

How to give strength to reason ?

Philippe Urfalino, director of studies at theEhessreverses to a certain extent the perspective of the file by taking an interest in, not in the wisdom of the crowds, but in the rationality of small groups. The question he poses is therefore how to ensure that the competence present in an assembly of wise men, a “ areopying », Gives all its measure, therefore, precisely, that it is not used for the number and rule of the majority. Based on the example of the medieval church, which weighed, during the elections of canons, bishops or the pope, majority and “ health », That is to say the number of votes and their alleged wisdom, Ph. Urfalino shows that the evaluation commissions of drugs in France and the United States reproduce the main features: the expression of reasons prevails on that of preferences in the two organizations ; The decision -making process is split, each proposal issued in the French commission being collegially discussed, each election of a Advisory Committee being ratified or rejected by the Food and Drug Administration ; The hierarchy is omnipresent in the assessment of the arguments or the final decision. The stimulating question posed by Philippe Urfalino is therefore to know why, when the Church had to abandon the concept of SANIOR PARS which threatened his unity, because of the always questionable nature of the wisdom invoked in favor of certain suffragents, these two organizations lead so effectively to a decision in each case which is submitted to them. In the case of commissiones Americans, whose members vote, number and wisdom do not contradict due to the wide convergence of experts ; In the French case, more intriguing since the commission decides to consensus, in addition to the particular influence of the president, the cognitive homogeneity of the members greatly facilitates the consensus. This last factor is particularly interesting: when cognitive diversity increases, as is the case in European drug assessment bodies, which bring together specialists from different European states, consensus is more difficult to reach. Although Philippe Urfalino does not himself develop this idea, it is possible to conclude that it is necessary to qualify the argument developed by other contributions from this file, according to which cognitive diversity tends to increase the probability that the Majority leads to the best option: it turns out that in small groups richly endowed with information, it induces blocking effects.

We will limit ourselves to conclude that a brief remark. One can wonder about the applicability of the paradox of the jury or the miracle of the aggregation to preferences, a question little discussed by the contributions. Politics indeed, far from being limited to an epistemic operation, is perhaps first of all the choice of values. However, if we admit that there is not, in the field of moral choices, of final criterion of good and just, the question of a possible epistemic superiority of democracy is in vain, and it is impossible to justify democracy other than by referring to the necessary autonomy of individuals. Politics is inseparable from the question of justice ; Except to consider, as Plato did from a cosmology which is no longer ours, that justice is a knowledge that some master better than others, the equal participation of all in the distribution of goods And charges is precisely the first indictment of purely procedural justice.