The Ash Twins Blog

Dawkins offers an enjoyable devastation of the jury adjudication process:

[W]hy are twelve jurors preferred to a single judge? Not because they are wiser, more knowledgeable or more practised in the arts of reasoning. Certainly not, and with a vengeance. Think of the astronomical damages awarded by juries in footling libel cases. Think how juries bring out the worst in histrionic, gallery-playing lawyers. Twelve jurors are preferred to one judge only because they are more numerous. Letting a single judge decide a verdict would be like letting a single chick speak for the whole herring gull species. Twelve heads are better than one, because they represent twelve assessments of the evidence.

But for this argument to be valid, the twelve assessments really have to be independent. And of course they are not. Twelve men and women locked in a jury room are like our clutch of twelve gull chicks. Whether they actually imitate each other like chicks, they might. That is enough to invalidate the principle by which a jury might be preferred over a single judge.

In practice, as is well documented and as I remember from the three juries that it has been my misfortune to serve on, juries are massively swayed by one or two vocal individuals. There is also strong pressure to conform to a unanimous verdict, which further undermines the principle of independent data. Increasing the number of jurors doesn’t help, or not much (and not at all in strict principle). What you have to increase is the number of independent verdict-reaching units.

So, what should we do?

I’ll call it the Two Verdicts Concordance Test. It is based on the principle that, if a decision is valid, two independent shots at making it should yield the same result. Just for purposes of the test, we run to the expense of having two juries, listening to the same case and forbidden to talk to members of the other jury. At the end, we lock the two juries in two separate jury rooms and see if they reach the same verdict. If they don’t, nothing can be proved beyond reasonable doubt, and this would cast reasonable doubt on the jury system itself.

I remember hearing about the robot controlled by rat neurons last month, but the details are extraordinary. The machine can learn:

Researchers say the robot, which they nicknamed Gordon, can learn behavior to a certain extent. When it hits a wall, for example, it gets an electrical stimulation from the robot’s sensors. As it confronts similar situations, it learns by habit. To help this process along, the researchers also use different chemicals to reinforce or inhibit the neural pathways that light up during particular actions. Lead researcher Kevin Warwick says that Gordon currently swerves to avoid a wall eight out of 10 times: “The signals and the pathways are strengthening as each action gets repeated.”

The NYTimes has a nice article about the public sphere’s digestion of the first presidential debate and the campaigns’ efforts to manipulate that process.

The positioning was in keeping with what is now a quadrennial rite in which the campaigns go full bore to convince the news media, and ultimately the public, that their candidate won — or more than that, to argue that the debate spotlighted some sort of character or issue defect in their opponent. This often involves highlighting some supposedly fatal mistake by their opponent — the sighs of Al Gore during a 2000 debate; the first President George Bush’s peek at his wristwatch while debating Bill Clinton in 1992.