Evolutionary biologist and former Professor of the Public Understanding of Science at Oxford University Richard Dawkins has a bone to pick with the U.S. judicial system. Specifically, the phrase “beyond reasonable doubt”. He argues that it doesn’t mean what it says it means — that two different juries would come to two different verdicts (or perhaps the same verdict but in a different way). Therefore, the phrase loses its power. It’s semantics, sure, but when people’s lives hang in the balance, Richard argues that we should perhaps take a second look at the phrase. Richard Dawkins’ new book is Science in the Soul: Selected Writings of a Passionate Rationalist.

Richard Dawkins: In Science in the Soul I have a chapter on reasonable doubt and it’s about, of course, the phrase. “Reasonable doubt” comes up in courts of law where juries are told that they must convict somebody, say a murder, only if it’s beyond reasonable doubt that they are guilty. 

And that sounds all very good, it should be beyond reasonable doubt, but when you think about the fact that—I think about courtroom dramas, which are so popular on television, for example, and I suspect that this accurately portrays something like what goes on real courtrooms, and I’ve certainly been on three juries myself, there is a note of suspense in the court when the jury comes back, which way will it go? Will it be guilty or not guilty? And then if they say “not guilty,” certain people heave a great sigh of relief. If they say guilty other people do. 

So there is a lot of doubt in the courtroom among people who have sat through the entire trial, the judge for example, the lawyers, the audience sat through the entire trial, as the jury has. 

So if the jury comes in and brings in a verdict that is beyond reasonable doubt, everybody in the court should know that. If it’s beyond reasonable doubt there can be no doubt at which way the jury will jump. And yet when the jury do give their verdict, how can that be if it’s beyond reasonable doubt?

Imagine the following experiment: suppose that you had two juries listening to the same evidence and the two juries are not allowed to talk to each other, they're sent off onto separate jury rooms and they come up with their own separate verdicts, who would bet on the juries coming back with the same verdict every single time? Virtually nobody would. 

If you think about the O.J. Simpson trial, for example, would anybody bet on another jury coming up with the same verdict? 

And yet unless you can bet, unless you can say “yes, they would come up with the same verdict,” you cannot really take the phrase beyond reasonable doubt seriously. 

Now I'm not suggesting that we should have two juries in every trial, I'm just pointing out that the phrase beyond reasonable doubt doesn't actually mean what it says.