On Democracy's Epistemic Value

"On Friday, April 15, at 8:45 p.m., Great Britain sat down to dinner." So began an article I wrote in 1994 on the world's first deliberative poll. "The nation sat at tables covered in white linen in the dining room of the Piccadilly Hotel in Manchester. How does a whole nation fit in one room? With a bit of twentieth-century sleight of hand; with the science of polling and random sampling. Standing in for 55 million Britons were 300 who in their attitudes, incomes, educations, and manners perfectly mirrored the larger populace."¹

Fifteen years later, in the context of this special issue of The Good Society on democratic epistemology, it seems instructive to return to this and subsequent deliberative polls designed to show that deliberation could improve public opinion. The questions on the table now are these: What kind of value is there, if any, to a well-designed deliberative procedure? If it produces a better public opinion, in what sense is this opinion better? Are the outcomes better because they get to the normative "truth" of the matter or are the norms they articulate standards that are produced by the deliberation itself?

In this article I will concur with the "utopians" that think that a deliberative public might indeed be able to produce judgments that seem in their own right to be sound and productive. But I will take issue with the idea that these outcomes can be judged independently of the procedure itself or that it makes sense to say that these outcomes track or map onto any independent truth.² Apart from deliberations that are simply prudential procedures for ascertaining what are the best means to given and undisputed ends, the "truth" of the matter is produced in the procedure itself, in the discursive and pragmatic discussions of members of a political community as they try to decide what are the best ends for their community.

To make my case, I will consider three cases of deliberative polling, which are useful because they come as close as possible to reflecting how a large-scale public might deliberate, and in the course I will tease apart the differences between sorts of deliberations and the "ways of knowing" that are at work.³

The evening of April 15 at 8:45 was truly an extraordinary moment: a representative sample of an entire country sat down together to begin to deliberate on the issue of crime.⁴ The poll was designed to give a representative sample of a country an intensive experience of learning about and deliberating on an issue and hence to find out what the public at large would come to think about the issue if it were to have an opportunity to do the same. The results should have a "recommending force," as James Fishkin, the originator of the idea, often put it, in contrast to the largely unreflective public opinion captured in conventional opinion polls.

The poll came on the heels of an incident in Singapore when a Westerner, after committing a minor crime, was flogged in public. The British were experiencing their own crime wave, and their reaction to the Singapore flogging incident was adulation. "Flog the bastards," one could hear rumbling throughout the discussions in the room that very first night of their weekend of deliberations.

A majority of those polled before the deliberative poll were fed up with crime and ready to exact harsh measures against criminals, including doing away with the "right to silence" (similar to the U.S. right to freedom from self-incrimination), making prison a more wretched experience, and even bringing back the death penalty.

Over the course of the weekend these three hundred some people met in small groups of twenty, moderated by journalists from The Independent newspaper. They used issue guides that laid out the pros and cons of various courses of action to address the problem. (These issue guides had been vetted by all sides of the political spectrum so that all were satisfied that they were free of bias.) And participants had sessions in which they met with people expert in...