A drunken brawl, a stabbing, a man falls dead, pierced through the stomach as he attempts to prevent a violent attack. The assailant is arrested, charged and brought to trial. The jury is required to reach a simple verdict. Has enough evidence been produced in court to convict the accused, or not?

So far so familiar. But the date is May 1728, and the whole course of judicial history is about to be rewritten. The dead man is the Earl of Strathmore, a Scottish grandee, his assailant one James Carnegie, and the defence is that there had been no intent to kill – the Earl was simply in the wrong place at the wrong time. “I had the misfortune to be mortally drunk, for which I beg God’s pardon,” Carnegie told the court.

Until that date, under Scots law, the verdicts “guilty” or “not guilty” had been unknown. Instead, a jury determined whether the case was “proven” or “not proven” – resting its decision entirely on evidence heard in court. But in the Strathmore case, a flamboyant lawyer, Robert Dundas, urged the jury to clear his client completely. To everyone’s surprise, it pronounced Carnegie “not guilty,” thus introducing a third verdict into the legal system. As the not guilty verdict became more widely used, “proven” was replaced by “guilty” – but the third option was retained. To this day, anyone in Scotland accused of a crime has the privilege of being found guilty, not guilty, or, if there remains some uncertainty about the evidence, “not proven” – effectively acquitted.

The “not proven” verdict – unknown in English law – has been controversial ever since. Sir Walter Scott, writer and lawyer, described it as “that bastard verdict” after a woman he had been prosecuting for poisoning her servant was let off by the jury, who found the case not proven. “All I can say,” snarled Scott, “is that if that woman was my wife, I should take good care to be my own cook.”

Now there are renewed attempts to scrap it altogether, with the Scottish Nationalist government considering strong pleas from victims’ groups that the bastard verdict should be abolished because it offers juries a “get-out” clause and allows guilty offenders to escape the retribution of the law.

In fact, the case for a “not proven” verdict has grown, not shrunk, in recent years. Not only should it be retained in Scotland, there are strong grounds for the English legal system to embrace it too. If justice means defending the innocent from wrongful conviction as well as representing the victim of crime, then allowing a jury to decide that there remains an element of doubt about the prosecution case would be of enormous benefit.

The case against Barry George, convicted of the murder of Jill Dando on doubtful scientific evidence, could have been set aside by a jury deciding that neither side had definitively proved its case.

Stefan Kiszko, who served 16 years for a murder he never committed, and who died a year after his release, might have been spared his ordeal. Other cases of wrongful conviction stretch back over the years – the Birmingham Six, the Guildford Four, the Carl Bridgewater case, Judith Ward, accused of a series of bombings in the 1970s, Angela Canning, convicted on the faulty evidence of a pathologist, Andrew Adams, wrongly convicted of the murder of a retired teacher, and whose sentence was finally quashed in January. In each case, a jury, offered an alternative to the unequivocal extremes of guilt or innocence, might have chosen a third way and prevented a gross miscarriage of justice.

Juries in Scotland use the third verdict infrequently, but effectively. In 2005-2006, the last year for which records are available, they brought in 128,400 guilty verdicts, 14,400 acquittals, and 800 not proven decisions. Out of 138 homicide cases, just two not proven verdicts were returned. An accused whose case is not proven walks free from the court and cannot be retried for the same offence. With no relaxation of the double jeopardy rule, a murderer, for instance, may escape justice simply because police have failed to mount a strong enough scientific case. The arguments against its use come mainly from victims’ families, who claim that it allows offenders who should have been convicted to escape punishment. There is also the risk that anyone saddled with a “not proven” verdict remains a permanent suspect.

But the real argument for its wider use outweighs all of these, and it lies in the crime of rape. Here the not proven verdict would act in the interest of the victim rather than the assailant.

Recent controversy over the low conviction rates for rape has focused on the reluctance of juries to convict when there is an absence of corroborative evidence. With only two conflicting accounts to go on, it is hard to decide whether sex was forced or not.

Women, who already face the ordeal of reliving their experiences in court, are confronted with the possibility that the man accused of assaulting them may walk free, while they in turn are cast in the role of false accuser or fantasist.

The not proven verdict would spare women that unwarranted slur. A jury that was genuinely undecided could inform the judge that the prosecution, in its view, had failed to prove its case – but without implying that the female victim was lying. It would mean, of course, that the accused left court with a shadow hanging over his reputation. But given that the balance is so heavily tilted at present against women, that would be a small price to pay.

Against a background of steadily diminishing civil liberties, the introduction of a third option for juries would strike a blow for justice. Not proven may have begun life as a bastard verdict, but now is the time to legitimise it.

Times Recommends

• Bank charges hit us for being human
• Sorry, Father, thou still shalt not steal
• The shed where Santa’s failures go to die