Having seen off God, Professor Dawkins turns his attention to another object of veneration, the jury, in the The New Statesman. He thinks that the requirement, as he sees it, of proof ‘beyond reasonable doubt’ for guilty verdicts, is at odds with what juries actually do when deliberating. The high standard of proof should leave no room for discussion, or for the forceful intervention of unusually persuasive jury members. Cases should speak for themselves. He says
“beyond reasonable doubt” is a hollow and empty phrase. If you defend the single-jury system as delivering a verdict “beyond reasonable doubt”, you are committed to the strong view, whether you like it or not, that two juries would always produce the same verdict. And when you put it like that, will anybody stand up and bet on 100 per cent concordance?
If you place such a bet, you are as good as saying that you wouldn’t bother to stay in court to hear the verdict, because the verdict should be obvious to anybody who had sat through the trial, including the judge and the lawyers on both sides. No suspense. No tenterhooks. There may be no practical alternative, but let’s not pretend. Our courtroom procedures make a mockery of “beyond reasonable doubt”.
He says he has been on three juries. Unless his experiences was from decades ago, he would have heard the Judge tell him that he could find the defendant guilty only if he was ‘sure’ of guilt. Judges have not used ‘beyond reasonable doubt’ to describe the standard of proof for a very long time – exactly because it invites pointless speculation about the percentage required. The requirement of being ‘sure’ is much easier to use as the standard of proof. We know when we are sure about something, without having to put a percentage on it; or more importantly, we definitely know when we are not sure. (In the bad old days there was a rural judge who gave the ‘fridge direction’: he would tell the jury that their decision was as important as other big events in their lives – such as whether to buy a fridge. Perhaps Professor Dawkins was in his court.) That was manipulative was patronising and unnecessary. Judges now do not qualify what is meant by sure. The official Crown Court ‘Bench Book’ (2010), which gives guidance for virtually every legal direction that a Judge has to give the jury, states that
The prosecution proves its case if the jury, having considered all the evidence relevant to the charge they are considering, are sure that the defendant is guilty. Further explanation is unwise.
If the jury are not sure they must find the defendant not guilty.
The point is that even people less smart than the Professor can be trusted to know whether they are sure of a defendant’s guilt, without a disquisition on what ‘sure’ means.
A jury trial is not a mathematical or scientific process. When deciding if some one has told the truth (the invariable key issue) we weigh up everything they have said for consistency and for reasonableness and for its fit with known facts about the case. All humans and their stories tend to be unreliable, even when they don’t deliberately set out to mislead. The criterion of judgment that makes the jury so valuable is human experience. A jury is not a computer which mechanically process information. A courageous jury will on (rare) occasions refuse to convict a technically guilty person if they think there is a gross unfairness about the prosecution: remember Clive Ponting’s alleged breach of the Official Secrets Act, and take heart from the jury that refused to convict William Penn, even when the Judge imprisoned them for failing to obey his directions to convict him. We should not eliminate the human dimension and the conscience verdict, for no appreciable gain. We have a Court of Appeal to deal with cases that have gone seriously wrong.
When entrusted with so great a decision, it is right that the jury should take all the time they need and scrutinise the evidence and arguments they have heard. 12 is a good number to even out differences between slower and faster individuals, men and women young and older, ethnically and socially diverse. In a long trial with masses of evidence you see them changing from 12 baffled-looking individuals at the start, into a powerful coalition of minds working together to answer the questions the trial gives them.
The process disgusts those who prize efficiency above all other virtues because it is slow, cumbersome and gives unpredictable results. But it has legitimacy. A group of citizens, not state appointees, decide whether other citizens have committed crimes. For all its imperfections and occasional failures (human, all too human), the jury system works well and has earned its place at the heart of our criminal justice system. It has always had powerful enemies and it is a shame to see the Great Debunker joining them.
Nothing like The Sun 
I think it’s worth mentioning that in France many criminal cases are settled by a single investigative judge, who I suppose qualifies as a “state appointee”.
The great debunker isn’t joining them at all, perhaps you didn’t read his article in the statesman, but I am sure that I read that he said that although he didn’t like the current method, it was still the best around, and far from trying to get rid of jury trials, it seems to me as though he would rather have two juries at each trial. I think the point of the article was about the insincerity of the phrase ‘reasonable doubt’
Out of curiosity could a jury be found in contempt of court for delivering a perverse aquittal? For example in the case of Clive Ponting the jury could legitimately have said that he wasn’t guilty of that particular offence; but what if the jury hadn’t agreed with the concept of a ‘secrets act’ from the outset? Would it have been an offence to not bring that up at jury selection?
Informative and interesting post though.
Interestingly, a friend of mine who served on a jury last year said that her jury interpreted the judge’s instruction that they should go find a verdict, or a majority verdict, as an instruction that some of them should compromise their position, because that is what the judge said they should do.
The meaning of “beyond reasonable doubt” was referred back to the judge by the jury of which I was a member. He acknowledged the difficulty of the concept and I believe we were encouraged to adopt the concept of certainty. I personally found this difficult too because there are very few things we can be certain of (apart from death and taxation).
As for the quality of decision making on the jury, it was appalling. Fortunately, “not guilty” verdicts were returned for both trials and so perhaps not too much harm was done.
With regard to the French inquisitorial style, I understand there is no love lost between the prosecution service and the judge, thus providing a safeguard for the accused. I doubt the outcomes of trials differ much between the two jurisdictions.
Yes, there’s a ‘Mr. Logic’ manner about Dawkins’ pronouncements on so many things and it sounds like he’s done it again here (I should say, I’ve not read his article). But, Francis, don’t many jurors still imagine the criterion is ‘beyond reasonable doubt’, because of books, telly, films and proverbial parlance? I was a juror in my first trial recently and I’m fairly ‘sure’ (though I couldn’t swear!) that that phrase was bandied about in the jury room quite a bit, although I can’t at all remember what language the judge used to us. So perhaps judges need to remind jurors that it is not in fact the standard of proof? By the way, I’m in favour of juries, precisely because law is not a science. And science sort of uses juries too – what else is peer reviewing?
Thanks for this comment – I agree with you. True, ‘BRD’ has entered the language, but judges are discouraged from using it – probably because people like Mr Logic get hung up on it and stop seeing wood for trees in the case in hand. ‘Peer reviewing’ is exactly it – on the basis that the jury and the accused are peers as fellow citizens and fellow humans,. Your judge probably told you that you had to bring your various experiences of real life to bear when judging the case, which may include scientific expertise and rare powers of analysis, but may not.
Love your blog BTW – haven’t seen it before.
In the cases on which I was a juror I would have voted for conviction had the standard of proof been based on the balance of probabilities. The criterion of certainty seems to me to be implausibly high because it is almost impossible to be sure of something. Call it pedantic, or call me Mr Logic, if you will, but so often common-sense arrives at incorrect conclusions. Inconsistent statements and circumstantial evidence may well point to someone’s guilt but is it proof? Bearing in mind, the often (in my view unjustly harsh) penalties that can be imposed there is a strong motive to return a not guilty verdict even if one believes or suspects otherwise. There have been too many instances of miscarriages of justice, some irrecoverable, as when convicts were hanged for me to place much faith in the decisions of juries.
I think I probably fall into the same camp as Professor Dawkins in this matter.
I was under the impression that Lord Denning held that beyond reasonable doubt means that the Jury ( or Judge or magistrate ) must be sure the defendant is guilty of the charges brought such that no reasonable person can be in doubt that the defendant is guilty. It is a common sense standard. The tribunal of fact does not need to reach the level of proof a sceptical philosopher would require about truth. It is a higher standard than in civil Law as personal Liberty is in issue rather than property. Jurors I think can understand this distinction. Even if Philosophers or professors of Biology are confused. May be the latter also get confused by Jury nullification which is a political issue rather than a matter of proof.
Excellent post Mr. Fitzgibbons. In fairness to Dawkins, whose logic is quite correct, he has strayed into an area in which he is not expert and has simply misunderstood the point. A related point: Truth, in the scientific sense is not really what juries are aiming to achieve. Juries are defendants’ peers, they are there to judge them and decide if they should convict someone for society’s sake. That they should be sure the Crown have proved the case is a pre-requisite for a conviction but it is certainly not the only thing running through their minds… they’re only human after all. This is perhaps best demonstrated by the many juries (almost) centuries ago who would give special verdicts to avoid sending people on a long drop on a short rope!
Lewis K.
I think you are being a little unfair on Prof D.
First, because his original article discussed both the US and the UK, with specific references to the death penalty and the OJ Simpson trial. So references to “beyond reasonable doubt” are an attempt to develop and argument that would straddle the Atlantic.
Second, because I wonder how much he would have wanted to change his argument if he had been required to replace “beyond reasonable doubt” with “sure”. Dawkins is arguing that juries do not, in practice, deliver verdicts based on the high standard of proof required. You seem to be challenging him on a different point that he didn’t actually make.
So far as I can tell, the only way the except you quoted would not be applicable with “sure” used in place of “beyond reasonable doubt” is if one understands “sure” to be a lower standard of proof than BRD. (And, if one does understand it to be a lower standard, that raises a very interesting set of questions, albeit different ones from Dawkins’ article.)
On reflection, my opening sentence was wrong. I think you are being very unfair.
Apologies for the several typos. I’m sure you realise that penultimate para should say “excerpt” not “except”.
Simon – thanks for your comment. It’s always a pleasure to find some one who reads and comments on the blog, especially the old stuff. The piece was meant to be polemical, but I hope I didn’t misrepresent what Dawkins said: I thought his complaints were a case of wanting the unattainable best, while unintentionally being an enemy of the good. There are lots of ways in which criminal trials are imperfect, but my view is that the jury remains an invaluable counterweight to the power of the state. Or, in old money, better that n (choose how many) guilty people go free than one is sent to jail (or put to death, as in the USA).
I hadn’t noticed that this was an oldie. It was linked to on my twitter timeline this morning from (via David Allen Green). I’ll try to keep up!
Most interesting. I came across this article when writing up my experiences of jury service as a response to the recent Serial podcast.
I have to say that my interpretation of being “sure” sets a higher bar in terms of probability than “beyond reasonable doubt”. In terms of specifics I was on the jury in a case where the evidence was the most simple possible. Just two people gave evidence, accuser and accused. The accuser claimed the defendant had threatened force order to extort money from him. There were no other witnesses to this or other material facts.
There is no doubt that the accuser made a better case. Both he and the defendant were from the same ethnic community but otherwise they were quite different. The accuser was smart both in appearance and manner, he spoke English fluently and was articulate and engaging. The defendant appeared sullen, had to speak through an interpreter and seemed to have difficulty forming his answers.
When I heard the judge’s direction to be “sure” I was in a quandary. Had this been the X factor it is clear who I would vote for. But in terms of proving whether or not the defendant was guilty I was much less “sure”. The “facts” in terms of testimony were contradictory. Effectively I was being asked who I believed, but then was I making a judgement on the appearance of the two men. I liked the accuser more for “sure”. But then I was concerned that I may just be prejudiced against the defendant and judging him on the basis that he was a bit stupid (and perhaps a bit scared /confused given his situation)
So I voted “not guilty” because I wasn’t “sure”. I was not alone but in a clear minority 9-3. The majority were firm in their belief and after an hour or so they persuaded me and the other two that we were just being too liberal in our definition of “sure”. So we ended up with a unanimous guilty verdict. Fortunately my qualms of a possible miscarriage were averted when I found out that the defendant had already been found guilty of a string of other crimes.
Still I was left with clear impression that “sure” means different things to different people and that ultimately the decision is as much subjective as objective.
This impression was reinforced by my experience as a juror on another trial. This was a quite different case, involving two defendants, separately represented, facing an extremely serious crime and took five weeks before we were asked to consider our verdicts. When we did one juror made it quite clear that he regarded all the evidence as pretty irrelevant. He was “sure” and prepared to find the defendants guilty and send them to prison for a very long time just because they appeared very “shifty” in how they looked/behaved while in the dock.
We spent three days trying to reach a decision and in the end failed. With hindsight we probably made a big mistake at the start of deliberation when we all had an open vote and as a result split into two distinct camps. After that it was largely a case of “us vs them” and only one person changed their view. Still we all agreed the only time we felt we made some useful progress was when we used a flip chart to mark 2 extremes of 100% guilty and 100% not guilty and we each marked for each defendant our judgement on this axis.
This showed two clear things:
– In terms of percentages we were actually not in groups at all, our answers were distributed along the axis. The key difference was what we felt “sure” meant. For some it was 90% positive, others just 50%.
– The defendants were different. For one it half put a mark on the “not guilty” side of the chart. For the other all 12 votes were on “guilty”. Again the defining factor was what “sure” meant.
In the end I feel we came up with the right answer given the evidence. But like a sausage it was probably best that no-one saw how we managed this.
Overall my personal view is that the system could be improved if juries were given more direction in terms of process. Specifically they should be advised
– To make votes blind so that people can feel less pressure both in reaching a decision and changing their minds. It would be a simple matter to include equipment to facilitate this e.g. a bag containing 12×2 balls of different colours.
– If split put the question of guilt aside for a time and just focus on a narrative that includes the aspects all can agree on such as time-line, witnesses who every one agrees are to be trusted.
Just a PS, sort of linked to the above.
The biggest surprise to me on my jury service was that apparently the justice system seems to find it quite normal and acceptable that witnesses routinely fail to tell the truth and often outright lie.
In the second trial we heard from a succession of people for both the prosecution and defence. They flatly outright contradicted each other. It was a simple matter of fact they could not all be telling the truth. Specifically:
– The main witness for the prosecution was involved in drugs and had a history of violence. Everyone agreed this to be the case, including the prosecution counsel and many of the prosecution witnesses. Nonetheless in his evidence this witness denied this to be the case, stating on oath that he had not been involved in drugs or threatened others.
– The wife of one of the accused gave him a cast iron alibi. She stated under repeated questioning that he had been with her for the entire period of time when the crime in question was committed.
This was crucial in our deliberations. Those who believed the defendants guilty said that the obvious lies of the main prosecution witness were understandable given he could not incriminate himself and should not affect our judgement of his other evidence which was the largest single factor in deciding the case. They also said it was obvious that the wife was going to lie to protect her husband so her evidence could be ignored.
Those who argued not guilty said the reverse. That they would prefer to take the word of a woman who on the face of it had never broken the law over an admitted criminal who was obviously lying on other matters.
Our task was not made any easier as it seems witnesses can by and large lie with impunity.
My simple belief was that if witnesses committed perjury they would, as a matter of course be charged if this was proven to be the case. This seems not, in fact, to be so.
The main prosecution witness clearly lied under oath. However as far as I know he was never charged.
Had we reached a guilty verdict would the wife, as a matter of course be charged? My guess is no, it would only be under exceptional circumstances.
I can see good, practical reasons why this should be the case. Based on my experience if everyone who lied in the witness box was charged hen the judicial system would consist of nothing but perjury cases.
But still it does make being “sure” or otherwise a lot harder if witnesses cannot be relied on to be telling the truth and/or are not under any real sanction for failing to do so.