History of Internal Deliberation

From "The Trouble With Elections: Everything We Thought We Knew About Democracy is Wrong," Chapter 11.8

Athenian voting disks

Most sortition advocates are deeply invested in give-and-take deliberation – what they call “deliberative democracy.” But it is my contention that they are only getting it half right. Active deliberation among a diverse and representative sample is essential for tapping collective intelligence, and crafting good proposals. But it is independent, internal deliberation that is essential to harness the wisdom of crowds, and to guard against the risks of groupthink that can result from active deliberation.

Internal deliberation embraces the independence necessary to elicit the sort of wisdom of crowds discussed by James Surowiecki. The value of separating the task of proposing, from the task of disposing (deciding) has been recognized for millennia. Daniela Cammack carefully analyzed the different ancient Greek words often translated into the same English word, “deliberate” in a paper entitled “Deliberation in Classical Athens: Not Talking, But Thinking (and Voting)” and convincingly argued that this internal-reflective concept of deliberation was used in the Athenian Assembly (as well as the Nomothetai), where citizens would listen to a small number of self-selected advisors (rhētores). While all Athenian citizens had a well-defined and fundamental right to speak (isegoria), in a body of 6,000 only a tiny fraction actually could, or possessed the gumption, to do so. The body as a whole primarily listened, evaluated, and voted.

The fourth century (BCE) restored democracy of Athens used a reformed system of lawmaking. A proposed new law would first be brought to the randomly selected Council of 500 (boule).  We don’t know for certain what sort of deliberation occurred at that stage. It was then sent to the mass Assembly (ecclesia). If they deemed such legislation worthy of serious consideration, they would refer it to a randomly selected jury (nomothetai) of perhaps 501 to 1,001 jurors. The Nomothetai listened to an equal number of pro and con presentations by designated speakers, and then, without debate voted by secret ballot on whether to adopt the new law. Each juror was provided with two bronze voting disks (one for “yes” and the other for “no”). One was solid, and the other had a hole in the middle (see photo at top). When held between the thumb and forefinger, nobody could see which sort of disk the member was depositing into the ballot box, and thus which way the member was voting, as they left the court. This blend of active deliberation followed by independent internal deliberation (private assessment) and voting by each juror may have effectively used collective intelligence and the wisdom of crowds sequentially by separate bodies.

James Harrington, in his 1656 landmark work describing a Utopian constitution, The Commonwealth of Oceana, also advocated separating proposing (with active deliberation) from disposing (with silent, independent reflection). In his design, a Senate of aristocrats (created partially through a random selection process) would debate and propose laws, and a popular House of Commoners (also formed partially through random selection) would approve or reject them – but without debate.

Because of the dangers of polarization and groupthink, a number of democratic theorists have proposed that final decision-making legislative bodies not engage in traditional debate — indeed, that they not speak at all. In the late eighteenth century, the important philosophers, Jeremy Bentham (the father of utilitarianism) in England, and the Marquis de Condorcet (a leader, and ultimately victim, of the French Revolution) in France, went one step further. They each argued that emotional influence and rhetorical skills could distort deliberation. Both favored requiring that all arguments be presented in written form. Condorcet wrote that although through discussion 

“one learns facts that one ignored and is made aware of objections one had not foreseen,… one is also seduced and worked up by the voice of an orator, led into error by clever sophistry without having the time to detect the trap. … [S]poken discussion harms the truth more than serving it, and …the preference of the majority would more often conform to the truth if one deduced it, without discussion…”

Bentham also advocated that only written arguments be used by legislators in order to avoid untoward psychological influences of skillful orators and focus attention on the logical merits of the arguments. It should be noted that brain research has highlighted the importance of emotional centers, such as the amygdala, in much of human decision-making. The goal can’t be to eliminate emotion from democratic decisions, but to avoid the manipulation of emotion by skillful orators, to the detriment of the listener, or the community. In order to avoid advantage for the side with the most eloquent presenters, it might be appropriate to reduce all presentations to written form, and have arguments from all sides presented by the same skilled speakers (or computer-generated voice).  

Although the US Constitution vests both the advocacy for legislation, and the judging of whether that legislation is worthy of adoption, in the same members of Congress, (unlike Harington’s notion that a Senate should propose and a House should dispose), James Madison’s Federalist Papers No. 10 actually presented an interesting argument against that lawmaking design.  

“No man is allowed to be a judge in his own cause, because his interest would certainly bias his judgment, and, not improbably, corrupt his integrity. With equal, nay with greater reason, a body of men are unfit to be both judges and parties at the same time; yet what are many of the most important acts of legislation, but so many judicial determinations, not indeed concerning the rights of single persons, but concerning the rights of large bodies of citizens?”

Keeping jurors in the dark about the leanings of their fellow jurors, combined with secret voting, promotes independent assessment and resists intellectual free-riding, groupthink, and polarization. But we mustn’t use wisdom of crowds independence exclusively. As Jon Elster notes in Securities Against Misrule: Juries, Assemblies, Elections,

“While it prevents bad interaction effects that arise through conformism and social pressure, it also excludes good interaction effects from information-pooling.”

Many experiments have shown that allowing many separate small deliberation groups within a large assembly resulted in better decisions than either full group deliberation or independent internal deliberation. This model of small break-out groups is typical of how citizens’ assemblies have functioned in recent decades. Since both procedures offer epistemological benefits and shortcomings in the decision-making process, the ideal sortition system would include both procedures, by separate groups of people. The active deliberation is most useful during the proposal development and amendment stage, and the independent internal “imaginative” deliberation is most appropriate when a final yes/no vote is ready to be made. 

Among the hundreds of citizens’ assemblies empaneled in recent decades, most have had between 24 and 150 members, though a few, such as the G1000 sortition implementation in Belgium had around 1,000. These bodies used random selection (often paired with stratified sampling) to promote diversity, rough representativeness, with a size that allowed for active deliberation. However, these samples are too small to be reliably representative of the population. To gain legitimacy for making a final public policy or legislative decision, as a body that is genuinely “like the community as a whole,” a final jury would need to number in the many hundreds or over a thousand members. 

However, since this jury is learning, listening, evaluating, and voting, without active give-and-take deliberation (other than asking clarifying questions), this jury has no reason to be physically gathered together (at substantial cost). Indeed, the sought after independence is enhanced through the use of remote Internet connections. Some online survey research has used a procedure that might be useful. After the participants have been presented with all of the information and arguments, they can be required to take a simple qualifying quiz to prove they paid attention to the material and actually did the necessary work, before voting.

This is a key element of my multi-body sortition design –  using both: diverse, active deliberation with collective intelligence, in smaller, longer duration bodies; and representative, independent assessment with the wisdom of crowds, in larger, shorter duration bodies, to gain the benefits of each method, while protecting against each method’s  shortcomings.