While I was otherwise engaged, the media and blogosphere exploded yesterday with fallout from the jury discharge in the Vicky Pryce case. The judge's remarks that the jury had displayed "absolutely fundamental deficits in understanding" predictably took centre stage. The implication was that it was the jury's fault. Predictably, protagonists drew battle lines as to whether trial by jury is a busted flush or not.

I think that is drawing the wrong line. A much more interesting set of questions is raised about how judges direct juries and what we know about how juries take decisions. These points lead me to agree very strongly with Lord Woolf when he says that there is a good deal of sense in more research being conducted into juries to help judges better advise jurors.

I was prompted to write the blog by some comments on Twitter, including those from two QCs, but let me first set out some of the background. The judge's comments follow on the jury asking several questions at the end of several days deliberation.

Q1. You have defined the defence of marital coercion on page 5 of the jury bundle and also explained what does not fall within the definition by way of examples. Please expand on the definition, provide examples of what may fall within the defence, specifically 'will was overborne' and does the defence require violence or physical threat?

Answer: "The pressure applied by the husband need not involve violence or physical threats. The law requires that a husband was present and coercion was to such an extent that she was impelled to commit an offence because she truly believed she had no real choice but to do so."

Q2. In the scenario that the defendant may be guilty but there may not be enough evidence provided by the prosecution at the material time when she signed the notice of intent to prosecute to feel sure beyond reasonable doubt, what should the verdict be, not guilty or unable or not safe to bring a verdict?

Answer: "Turning to page three of my written directions, the direction is combining the burden and standard of proof with the need for a majority verdict. If, having carefully considered all of the evidence, at least 10 of you feel sure of the guilt of the defendant then it would be your duty to return a verdict of guilty. On the other hand, if after careful consideration at least 10 of you were feeling less than sure of guilt, then it would be your duty to return a verdict of not guilty. And so it follows that if at least 10 of you are not sure, the appropriate verdict is one of not guilty."

Q3. If there is debatable evidence supporting the prosecution case can inferences be drawn to arrive at a verdict? If so can inferences/speculation be drawn on the full evidence or only where you have directed us to do so?

Answer: "The drawing of an inference is a permissible process. Speculation is not. In this case the evidence on which the prosecution relies is largely undisputed, and where you are willing to draw inferences from that is entirely a matter for you."

Q4. Can you define what is reasonable doubt?

Answer: "The prosecution must make you feel sure beyond reasonable doubt. A reasonable doubt is a doubt that is reasonable. These are ordinary English words that the law does not allow me to help you with, beyond the written directions [he had already given them]".

Q5. Can a juror come to a verdict based on a reason that was not presented in court and has no facts or evidence to support it?

Answer: "The answer to that question is a firm no. That is because it would be completely contrary to the directions I have given you."

Q6. Can we infer anything from the fact that the defence didn't bring witnesses from the time of the offence, such as the au pair or neighbours?

Answer: "You must not, as I have now emphasised many times, speculate on what witnesses who have not been called might have said or draw inferences from their absence. Her evidence is that no one else, other than Mr Huhne, was present when she signed the form."

Q7. Does the defendant have an obligation to present a defence?

Answer: "There is no burden on the defendant to prove her innocence and there is no burden on her to prove anything at all. The defendant does not have an obligation to present a defence, in this case the defendant has given evidence and it is for you to judge the evidence from her in the same way you would any other witness."

Q8. Can we speculate about the events at the time Miss Pryce sent the form or what was in her mind when she sent the form?

Answer: "The answer to that is an equally firm no. The position in a criminal is that no one must speculate. There is a difference between speculation, which is not permitted, and inference, which is the drawing of common-sense conclusions from the facts of which you are also sure. Speculation is guesswork. That is not the same as inference at all."

Q9. The jury is considering the facts provided but is continuing to ask the questions raised by the police. Given that the case has come to court without answers to these questions please advise on which facts in the bundle the jury should count on to determine a not guilty or guilty verdict.

Answer: "You must decide the case on the evidence [put before the court]. It is for you to decide which you consider to be important, truthful and reliable then decide what common-sense conclusions you can safely draw. It is not for me to tell you which piece or pieces of evidence are important and which are not. That is a matter for you to decide."

Q10. "Would religious conviction be a good enough reason for a wife feeling she had no choice i.e. she promised to obey her husband in her wedding vows, he ordered her to do something and she felt she had to obey?

Answer: "This is not, with respect, a question about this case at all. Vicky Pryce does not say that any such reason formed any part of her decision to do what she did. Answering this question will not help you in any way whatsoever to reach a true verdict in this case. I must direct you firmly to focus on the real issues in this case."

Mr Justice Sweeney went on: "I want to repeat the absolutely vital importance of your following my directions of law to the letter and the fact that it is an equally important part of each of your individual duties to ensure that all of you do follow my directions of law to the letter.

"Without doing so, you are simply not in a position to reach a true verdict according to the evidence one way or the other.

"It is essential that each of you ensure that my directions of law are faithfully applied by all of you. If, for whatever reason, any one or more of you feel you do not understand my directions, then it would be wholly wrong to reach a verdict one way or the other.

"Equally, the duty of all of you is to ensure that no one of your number does reach a conclusion one way or another unless they are confident they are able to understand and follow to the letter my directions. It does come in some cases that jurors are simply not able to agree in sufficient numbers on the verdict.

"If, after further consideration, you find yourselves in a position where you are simply not able to agree, then you must of course have the courage to say so. I hope all that is clear."

It is worth noting in passing the language used. Readability and intelligibility depend on short words in short sentences. But I do not want to dwell on that. Nor do I want to strain too hard against the judge's criticism of the jury; though voicing it publicly in such a controversial trial may not be a step which makes juries better in the future. Some of the questions look odd, but there may be reasonable explanations which do not suggest a jury having lost a fundamental grip. One possibility is that the more eyebrow-raising questions reflected points being made within the jury room by one, or a small number of, members of the jury that the rest were seeking to deal with. I think it is worth saying though that some of the questions, which have nonetheless been ridiculed, are perfectly reasonable ones. These questions and the way the judge answers them may raise more modest, but important, questions.

Take for example the question about "beyond reasonable doubt". The judge answer is, understandably perhaps, tautologous and he declines to help further. One suggestion might be that the judge may have been wrong to say he was prohibited from providing a further explanation. Archbold 4-447 (f) suggests that where asked a judge should offer an explanation: "a reasonable doubt is the sort of doubt that might affect the mind of the of a person in dealing with matters of importance in his own affairs" (Walters v R). Though if the judge were wrong, frankly, I struggle to see how this further definition would help.

Another interesting question is whether the judge should have introduced the "beyond reasonable doubt" test. The weight of opinion that I have canvassed is that the normal direction to give is that the jury must be sure that that the defendant is guilty. The Crown Court Benchbook gives a specimen direction based on this test, and says that being sure is the same as entertaining no reasonable doubt. It also counsels against providing further guidance on the "sure" test.

In this case the judge directed using both versions:

"The Prosecution has chosen to bring this case, and it is for them to prove it, on all the evidence before you, if they can. There is no burden on the Defendant to prove her innocence. On the contrary, there is no burden on the Defendant to prove anything at all.

"The standard of proof that the Prosecution must achieve before you could convict is simply this – the prosecution must make you feel sure of guilt (that is the same as, but no more than, the proof of guilt beyond reasonable doubt)."

I'd tend to the view that there is nothing wrong with saying this, unless you think that the jury might be confused by being offered two tests rather than one. Importantly, though, they might well be confused by twin tests of such abstraction. We simply do not know whether that helps or hinders them. The more important issue in some ways is whether the jury can understand either test. That they could not has been seen as evidence of their inadequacy. I am not at all sure that it is. Let me turn to the twitter conversation from last night. James Turner QC (an intelligent – which you'd expect of course – and engaging presence on Twitter) said this that caught my eye:

"Well, I've never understood what "reasonable doubt" means – reasonable to whom, for example; deep philosophical stuff…"

"…And the concepts of 'dishonesty' and 'recklessness' are even worse – they mean what you want."

On the sure test he said this:

"How do I know if sure? What does that mean? Have I got to exclude all doubts to be sure, or only 'unreasonable' ones?"

And continues with…

"And that's only the first direction as to law – then add the evidence to the mix – not so simples."

Sean Jones QC (another great presence on Twitter) made several comments pointing out similar linguistic uncertainties in other fields of law and then made this important point:

"[I] suspect complete precision of meaning is both impossible and undesirable."

I have a lot of sympathy with both positions. And my answer is that we should do more research on these questions. My colleague, Cheryl Thomas, has conducted a study for the Ministry of Justice which found this (amongst other things):

"Most jurors at Blackfriars (69%) and Winchester (68%) felt they were able to understand the directions, while most jurors at Nottingham (51%) felt the directions were difficult to understand.

"Jurors' actual comprehension of the judge's legal directions was also examined. While over half of the jurors perceived the judge's directions as easy to understand, only a minority (31%) actually understood the directions fully in the legal terms used by the judge. Younger jurors were better able than older jurors to comprehend the legal instructions, with comprehension of directions on the law declining as the age of the juror increased."

More fundamentally, I understand there has been no research in this country on how juries (real or experimental) approach the issue of burden of proof. It is interesting to note that the shift from the 'reasonable doubt' to the 'are you sure?' test was done without such research. It is worth exploring whether, when we apply a linguistic test to a matter of uncertainty, we all understand the test and apply it in a similar way.

That kind of research can be done. One study which particularly caught my eye was Sorting Guilty Minds. That study tested US law through an experimental survey based design. It looked at how notions related to intent ("purposeful, knowing, reckless, or negligent") were assessed by a random sample of respondents. The law's assumption that people naturally sort mental states into these four categories with a high degree of accuracy was partially justified. Respondents, " were able to regularly and accurately distinguish among purposeful, negligent, and blameless conduct." The respondents struggled however with the difference between knowing and reckless conduct. This kind of research is worth doing so that the system can be sure that when it communicates responsibilities to juries and innocence or guilt to defendants and society it does so in a way that is reliable.

These questions are raised before we get into the more tricky territory of the marital coercion defence, on which the judge directed:

"A wife's will would not have been overborne (in the sense that I have just described) if, for example, she was persuaded by force of argument to choose (albeit reluctantly) to commit the offence rather than to take another course, or if she was persuaded (albeit reluctantly) to commit the offence out of love for, or loyalty to, her husband or family, or to avoid inconvenience (whether to herself or others). Her will must have been overborne in the sense that she was impelled to commit the offence because she truly believed that she had no real choice but to do so.

"It is not, however, for the defendant to prove that Mr Huhne coerced her – rather it is for the prosecution to prove that he did not do so. The Prosecution may do that (as they seek to in this case) either by making you feel sure that Mr Huhne was not present when Ms Pryce committed the offence, or by making you feel sure that her will was not overborne (i.e. that she was not impelled to commit the offence because she truly believed that she had no real choice but to do so)."

Again, I want to raise some questions about the nature of the text here. Long, complex sentences with unnecessary parenthesis and obscure language bedevil lawyers (overborne?). We really should dwell at some point on how lawyers and judges communicate; particularly, as here, where the legitimacy of an institution and the liberty of an individual is at stake.

Of course it is not always easy and, I say with the strongest emphasis, the job of judging is one of the most challenging and important jobs in our society.

But it is not, I hope an unfair point to make that the communication of the constituents of the offence could be clearer. And it is not at all surprising that the jury struggled and asked for examples. These questions are one of the few ways in which juries can speak. As James Turner QC put it last night:

"It's all very well for judges and press to criticise a jury, but isn't it unfair that jury not then permitted to explain?"

Richard Moorhead is Director of the Centre for Ethics and Law at the Faculty of Laws, University College London. Click here to visit his blog, Lawyer Watch, and click here to follow Richard on Twitter.