Howzat?

Stephen Sedley

  • The Origins of Adversary Criminal Trial by John Langbein
    Oxford, 376 pp, £30.00, February 2003, ISBN 0 19 925888 0
  • Archbold: International Criminal Courts edited by Rodney Dixon, Richard May and Karim Khan
    Sweet and Maxwell, 1000 pp, £125.00, December 2002, ISBN 0 421 77270 0

Three hundred years ago an Englishman charged with, say, robbery could expect to be interrogated by a local magistrate, held in jail until the King’s justices next rode in on circuit, arraigned before a jury of local property-owners on an indictment he had never seen, and tried in less than an hour. He would not be allowed legal representation even if he could afford it. He had no right to give sworn evidence in answer to the witnesses who – as the saying was – were swearing away his life. He was allowed to question them as best he could and then tell the jury from the dock why he was innocent. Nobody would tell the jury that he was presumed innocent until proven guilty beyond reasonable doubt, because that was not the law. There was effectively no appeal. If convicted, he would probably be hanged before the judge rode on to the next assize town.

Most of the changes which gave us the modern criminal trial, and which are the subject of John Langbein’s book, came about in the last part of the 18th century and the first part of the 19th. Their legal and practical effect from 1822 to the present has been tracked in the 54 editions of the Criminal Pleading, Evidence and Practice of John Frederick Archbold, an otherwise obscure barrister who supplemented his earnings by knocking out legal textbooks, one of which has turned out to be his passport to immortality and has made his name a kitemark. Hence the delphic title of the second volume included here, which carries forward Archbold’s work from the Georgian courts at the centre of Langbein’s study into the post-Nuremberg world.

The growing use of barristers during the 18th century to conduct prosecutions – a high proportion of them brought by local prosecuting societies – highlighted the unfairness of denying the same right to defendants on felony charges. (Oddly, there was always a right to counsel for those charged with misdemeanours, some of which were quite serious: the acquittal of the Seven Bishops in 1688 was in large part due to their having been charged with seditious libel, which was classed as a misdemeanour, and thus having been very competently defended.) The widespread use of professional thief-takers and informers made perjury by prosecution witnesses a daily event, although there was little chance of exposing it in the absence of competent cross-examination. An entrepreneurial legal profession, well represented in Parliament and close to the judges, was becoming aware of a large new market for its services.

But to break into it required fundamental changes in both law and legal culture, starting with a recognition of every accused person’s right to speak through an advocate. Samuel Johnson gave this wings:

‘As it rarely happens,’ he said to Boswell, ‘that a man is fit to plead his own cause, lawyers are a class of the community, who, by study and experience, have acquired the art and power of arranging evidence, and of applying to the points at issue what the law has settled. A lawyer is to do for his client all that the client might fairly do for himself.’

And once the lawyers were involved, other things became axiomatic: that the prosecution must prove its case without the help of the accused, and that it must prove it beyond reasonable doubt. The gallows and the transports, the Black Act’s ministers, gave urgency to the sense that these were inalienable rights. The first inroad into the no-counsel rule for felony trials had been made by the post-Bill of Rights Parliament, which in 1696 passed an act giving a right to defence counsel to persons – unlikely to be common criminals – accused of treason. But there it stalled until the lawyers started pushing for change in the succeeding century.

There are many histories – economic, cultural, moral – to be written of the vast lurch into modernity of the criminal justice process. The American scholar John Langbein traces its narrative and procedural dimensions, basing himself extensively on the Old Bailey sessions papers, a rich and only recently mined resource, but also on less obvious sources such as the Post Office solicitors’ archives at Mount Pleasant in Clerkenwell, which record early prosecutions for interfering with the mails, and the Royal Mint’s archives, which record prosecutions for coining. The picture it confirms – a number of legal historians have already done detailed work in the field – is of a Hogarthian havoc of authoritarianism and anarchy, for which the state provided the forum but in which the prosecution of offences was commonly a private enterprise and often a pretty unscrupulous one.

It was into this unsavoury system of felony trials that the judges began during George III’s reign to admit defence lawyers. By the third quarter of the 18th century the Old Bailey hack had become a byword in the legal profession for disreputable advocacy. But in the world outside, the lawyers who stood between the accused and the gallows commanded admiration, and none more than William Garrow, the most successful and most celebrated defence counsel of the later Georgian years. To Langbein, however, Garrow is an object of contempt. He rates over ten times as many index entries as Thomas Erskine, a more important figure, because he is the exemplar for a thesis which rings through the book like the 13th chime of the clock – that where the old form of English trial was a truth-seeking process, albeit an inefficient one, the modern Anglo-American adversarial trial is dedicated to the suppression or evasion of the truth. ‘Adversary procedure,’ he says, ‘represented a material worsening of the truth deficit.’

There is a great deal to be said about, and in many respects against, the modern criminal judicial process both in the US and here. It is still caught up, for example, in the absurdities of the rule against hearsay evidence, which originated in the period Langbein discusses and which even lawyers have difficulty in understanding and applying. (Is it permissible to testify that when the accused ran off, someone shouted ‘Stop, thief!’ and so on.) But there is practically nothing to be said, even comparatively, in favour of the form of trial it replaced; and it is not obvious why Langbein should want to suggest otherwise. There is no intelligible sense in which the old form of trial was more apt than the modern criminal trial to elicit the truth. It was directed to establishing, by largely unjust means, the guilt of the accused as a prelude to brutal punishment or, occasionally, ostentatious clemency. It was the jury, albeit composed entirely of male property-owners, which stood between the accused and his accusers, and not infrequently it acquitted. The fact that the record of the defendant’s interrogation by a local magistrate was sent on to the trial court, and the fact that at trial he had to speak for himself or perish, hardly amounted to a collaborative search for the truth.

Langbein does not argue for a return to the old English system; he accepts that ‘two-sided partisanship may indeed have been better than one-sided partisanship’; but he advocates a resumption of the search for the truth – whatever that signifies – by means of a modern inquisitorial system.

The suggestion that modern inquisitorial procedures are superior to adversarial ones (Langbein is also a comparative lawyer) deserves serious attention. But discussion of it is not assisted by denouncing as ‘the trickster Garrow’ a forceful participant in the burgeoning adversarial system who, if his efforts inevitably shielded a good many guilty people from conviction, also saved many whose guilt was far from certain from hanging or transportation. Not a single verbatim instance of Garrow’s advocacy cited in Langbein’s book shows him to have been less than professional in his standards. A similar animus prompts Langbein to denounce Henry Brougham’s celebrated exposition of the duty of the defence counsel, delivered at the trial of Queen Caroline, as ‘self-serving prattle’ which became ‘window-dressing for a truth-be-damned standard of defensive representation’, and to describe as ‘antics’ John Lilburne’s courageous refusal in 1649 to incriminate himself on a treason charge. Seen through such spectacles, history starts to blur.

No modern system of criminal justice is either truly inquisitorial or truly accusatorial. All have elements, in the initial phase, of inquiry, intended to determine whether there is a case worth putting before a trial court. All have elements, at the point of trial, of open contest between the accused and the state. None, at least in the 44 states which have now signed up to the European Convention on Human Rights, allows the railroading of defendants which characterised what Langbein approvingly calls the ‘accused speaks’ mode of trial. All require the state to establish the accused’s guilt, even though part of the proof may well be what the accused himself has chosen to say. If as a result we have trials which go on for days where they used to be over in an hour or less (though one of Langbein’s sound criticisms of the early days of the adversary system is the indecent speed with which cases continued to be dispatched), and if as a result guilty people sometimes walk free, it is not necessarily an excessive price to pay for trying to ensure that people are not convicted unless a fair trial has made a jury sure of their guilt.

The great unresolved truth in any criminal trial, whatever the procedure, is not whether the state has satisfied the jury of the accused’s guilt – the verdict tells us that – but whether the accused is really guilty. It is ineluctable even in a fair system that guilty people will sometimes go free and – worse still – innocent people be convicted. Some are released on appeal; but we do not know, any more than other countries do, how many people are in prison for crimes they did not commit. The fact that no system is infallible means that there are some, and simple denials on the part of those convicted (the stock in trade of sex abusers in particular) is not going to tell us who they are. The recent use of DNA analysis in death-row cases in the US has dramatically illustrated the technique’s potential for eliminating suspects, though too frequently in cases where there ought not to have been a conviction anyway. But nothing either in the ancient English trial or in modern inquisitorial procedures seems capable of dispelling a worry which hangs like a cloud over even the best criminal justice systems.

The French system does not, any more than the English system, involve a roving investigation into the truth. Although the French code of criminal procedure mandates the judge to carry out all investigations that he considers ‘necessary for establishing the truth’, both systems are concerned, first of all, to see if there is reliable evidence to support a specific accusation and, second, to test the evidence in open court in a process which requires the state to prove its case. A truly inquisitorial system which required people to exculpate themselves from a bare accusation, or to answer any and every official inquiry, would be regarded by most people as belonging to a police state. The proper end of criminal process is regarded, at least in liberal polities, as the conviction of those whose presumptive innocence on a specified charge the state can conclusively negate: to answer the question ‘Howzat?’ That requires a spelled-out charge, a full chance to challenge the evidence which supports it, an equal chance to advance the accused’s own case before an independent court, and a presumption of innocence which can be defeated only by compelling proof. Truth as an independent entity does not get lost in such a process because truth in that sense has never been its object. The jury is sent in search of a single verity: has the accusation been proved? And this, pace Langbein, is true of every Continental legal system.

Where England and Wales have since 1985 had a Crown Prosecution Service, designed to take the prosecution function away from police forces which have too intense an interest in securing a conviction, France has examining magistrates who are sometimes criticised for becoming surrogate prosecutors. Most of the public concern about criminal justice in France in recent times, however, has concerned controversial discontinuances or acquittals. England and Wales in the 1970s and 1980s suffered a catastrophic series of erroneous convictions, mostly but not solely in terrorism cases, which finally led the courts to insist on the prosecution disclosing everything it had collected but was not proposing to use. In 1996 this process was interfered with by legislation which handed back to the police the crucial decision as ” to what materials should be classified as too sensitive to hand over. Crown prosecutors are expected to review the list but the defence never gets to see it, and research suggests that important material regularly gets left off the non-sensitive schedule. But none of this deserves the stigma of truth-suppression: it represents at the worst a flawed endeavour to get all the relevant cards on the table. Again, the modern system in England and Wales involves a quid pro quo process by which the Crown has to start by disclosing any material which might undermine its case, a judgment requiring a painful degree of candour. The defence then has to set out its case (a provision much honoured in the breach), and in return the Crown must disclose anything it has which might assist that case. It is arguable that where the Crown has material which might point towards the accused’s innocence, it should hand it over unconditionally. But the legislative aim of preventing the defence from setting an ambush at trial or, where it has no worthwhile case, cobbling together a case out of the prosecution’s materials, is intelligible and would no doubt have Langbein’s support. Although defenders of the ‘pure’ adversarial system continue to regard any obligation on the defence to show its hand as an invasion of the rights of the accused – a view strongly rooted in the period of Langbein’s study – there is little evidence that to do so works injustice and good reason to think that it helps in arriving at what the jury oath calls a ‘true verdict’.

The same is true of the caution which has replaced the old formula that you are not obliged to say anything but that anything you do say will be taken down (or, as the bitter joke went in the days before obligatory tape-recording of police interviews, anything you don’t say will be made up) and may be used in evidence against you. Today you are also told that if you fail to mention something which later forms part of your defence, adverse inferences may be drawn. This, too, is entirely reasonable, not least because it only recognises what juries have been saying to themselves for generations in spite of judicial exhortations not to. The big difference is that since the great reforming measure of the 1984 Police and Criminal Evidence Act (PACE), what a suspect says will ordinarily be excluded unless it has been taped in an interview. With the saving of the time – sometimes days – that used to be spent on challenging the veracity of conversations recorded in police notebooks in a prose encountered nowhere else in the English-speaking world, the trial can now at least get down to the real issues; and with the right to have a lawyer present (though not now in terrorism cases), the interview has become a reasonably safe place for putting explanations, or their absence, on the record. In France there are no such protections except for children. The garde à vue, sometimes lasting for days, takes place without access to a lawyer except in the first hour and then the 20th, and without being tape-recorded. Suspects are not entitled to be told that they have a right to remain silent or even why they are being held. There is nothing in inquisitorial justice which demands this – it doesn’t happen, for example, in Germany, where every suspect has to be told his rights, or in Italy, where the suspect’s lawyer can be present throughout the interrogation – but it is defended precisely on the ground that it makes it easier to get at the truth. Nor does an adversarial system necessarily carry such protections: in Scotland there is no right to have a lawyer present at a police interrogation. Insofar (and it really is not very far) as the examining magistrate in inquisitorial systems acts as a neutral fact-gatherer, the problem remains that, as a French judge recently wrote to me, ‘c’est un homme seul qui écrit l’histoire à sa façon.’

In the nature of things, no doubt, we have stumbled rather than made an orderly advance into modernity; but Langbein is wrong to accuse the modern Anglo-American system, as he does, of being atheoretical. Like most things outside the Napoleonic world it was not built on a single theory, but it is shot through with theoretical propositions about the right of silence, the privilege against self-incrimination and the right to counsel – all issues with which Langbein’s critique is concerned. It is unsettlingly unclear, in fact, how much of the old system Langbein is endorsing when he writes: ‘Because the accused conducted his own defence at trial, he necessarily made himself an informational resource for the court . . . The trial procedure put pressure on the accused to speak in person about the charges and the evidence adduced against him.’ If it is the denial of counsel to the accused which he favours, his view could generously be thought utopian: Thomas More’s polity did away with lawyers, ‘so shall there be lesse circumstaunce of wordes and the trwth shal soner cum to light; whiles the judge with a discrete judgement doth waye the wordes of hym whom no lawier hath instruct with deceit; and whiles he helpeth and beareth out simple wittes agaynst the false and malicious circumversions of craftie chyldren’. But in real life, as Langbein’s account shows again and again, dumbstruck prisoners got no help from judges, perjurers went unchallenged and wretches hung that jurymen might dine. If Langbein means to endorse the pressure to speak or be convicted, to a modest extent his advocacy of the use of the accused as an ‘informational resource’ has been heeded, at least in this country. But there is a meaningful gap between giving the suspect fair warning that a failure to explain things now may make it harder to do so later and compelling the suspect to contribute to the assembly of a case against himself – a process redolent less of inquiry than of inquisition. And when Langbein asserts that the modern right of silence operates in practice to deny the court the testimony of the person usually closest to the events, namely the accused (he has evidently overlooked Sam Weller senior’s advice to Mr Pickwick about the indispensability of a good alibi), he is mistaken. Since 1898, when the law first allowed defendants to give evidence, staying out of the witness box has been a high-risk tactic. It is sometimes done, and it sometimes works, but generally only where the defence advocate can show the jury that the prosecution case really doesn’t stack up; and even then the defendant’s silence may prove too eloquent for his own good.

The length of modern criminal trials reflects as much as anything the enhanced technology and procedures of criminal investigation, all of which the defence has to be able to probe. But the manner of cross-examination of witnesses remains a serious issue, especially when the witness is the victim of the crime. The Bar Council has rules of professional conduct which forbid, for instance, questions merely designed to wound or denigrate; but enforcing such rules isn’t easy, especially when the damage is generally done before the judge can intervene. It is alarmingly easy for an aggressive lawyer to cross-examine a perfectly honest witness into confusion or collapse, and to the extent that Langbein finds counsel’s ability to deflect or obstruct the court’s inquiry objectionable, he has a sound point. The Runciman Royal Commission’s call for judges to exercise greater control of advocates still needs heeding; but whether the defect goes to the heart, rather than simply the handling, of the adversarial process is another question. (There are occasional records of witnesses getting the better of counsel – not many, it’s true. But trial excerpts on a Massachusetts website include a dogged cross-examiner asking a pathologist whether he could be sure the patient was dead when he began the autopsy. The pathologist is sure. ‘How can you be so sure, doctor?’ ‘Because his brain was sitting in a jar on my desk.’ ‘But could the patient still have been alive nevertheless?’ ‘It’s possible,’ the pathologist says, ‘that he could have been alive and practising law somewhere.’)

Perhaps the sharpest conundrum about truth in the criminal justice process concerns the revelation of previous convictions. These are facts – truths if you like – but the ground for their exclusion is not just what Bentham called a foxhunter’s reason, designed to give the quarry an even break. You have to start by asking which is the relevant truth. Is it that the accused has a record which makes it likely that he has now done the same thing again? Or is it that the prosecution is relying on propensity and suspicion where it should be relying on evidence? France adopts the first view and allows its courts to take the accused’s record into account in determining guilt. England at present takes the second view and routinely – though with important exceptions – excludes it. Both are trying with blunt tools to reach a reliable result by fair means. French trial judges (who deliberate with their juries) tend increasingly to keep the accused’s record in the background, and many would like it held over until sentencing takes place. In conventional thinking on this side of the Channel, the exclusion of previous convictions discourages the police from simply rounding up the usual suspects and the prosecutor from substituting suspicion for fact. But the current pressure for the routine disclosure of previous convictions springs from the legitimate sense that propensity is a relevant fact. The common law, which Langbein shows to have arrived by halting stages at the principle of exclusion early in the 18th century, has long since forfeited any claim to consistency or logic in this department. It excludes evidence of bad character but, by letting in evidence of good character, drops loud hints to juries when the latter is not given. It excludes evidence which indicates a propensity to commit particular crimes but lets in previous convictions (and in extreme cases previous acquittals) which show method or habit. Spotting the difference can be taxing, and a lot – probably too much – now depends on the prosecutor’s judgment.

It’s precisely because any rigid rule will offend against some legitimate value that the conundrum has no principled solution. Langbein’s preference for what he calls the Continental method – letting it all in for what it is worth – is no more principled than any other solution. A judge sitting without a jury has no alternative but to look at the evidence she is offered, however irrelevant or prejudicial. But where you have a judge who can rule on admissibility in the absence of the jury – and countries like Norway and France (the latter reputedly on the instigation of Voltaire) have juries for serious cases patterned on the English system – an exclusionary rule can be adopted and applied. You then have either to let the judge grapple with it or to give up and let juries hear everything, admissible or not – as they did in the 18th-century system which Langbein excoriates.

One way and another the rebel flag of truth-seeking in which Langbein’s research is wrapped probably makes the Anglo-American adversarial system look better than it is. Even shorn of the grandiose idea that a criminal trial should be expected to do more than determine guilt on a specific charge, the adversarial criminal trial could do with an overhaul. Although the cost of civil litigation in this country remains shameful, in civil procedure the recent Woolf reforms have been largely successful in changing the culture from confrontation with blood on the walls to something nearer co-operation. The trail had been quietly pioneered by the family law courts, although there, too, property and maintenance disputes continue to make lawyers rich and ex-spouses poor. The Auld Report is the start of an endeavour to initiate similar reforms in criminal procedure, but its implementation is bedevilled by the fact that, unlike civil litigation, criminal justice means votes.

It is the role of the prosecutor which probably marks the sharpest division between accusatorial and inquisitorial systems. Sir Robin Auld came to the conclusion that early identification of the issues was the key to an efficient criminal justice system, and that this meant not only a competent defence but strong and independent prosecutors. One legacy of the common law’s history has been a measured distance between the judge and the prosecutor, where in many other systems the prosecutor has judicial status and a concomitantly elevated role in court. Common law practitioners, who set great store by the visible equality of prosecution and defence in the choreography of justice, would be shocked to see, in France, the prosecutor sitting on a level with the judge and jury while the defence sits in the well of the court. The French procureur in fact both directs the police inquiry and determines the scope of the juge d’instruction’s investigation. The Scottish procurator fiscal, whose office was probably derived from France in the 16th century, comes from this mould. It is a mould which may soon break: the European Court of Human Rights has recently condemned Romania for allowing the public prosecutor to remand people in custody. Crown prosecutors in England and Wales have similar responsibilities but no such powers or status: they are civil servants who bring charges, and since 2000 can act as prosecutors. But Auld has been proved startlingly right by a recent experiment authorised by the Attorney-General in which having crown prosecutors formulate charges instead of police officers led to a drop of over half in abortive trials and a climb of 15 per cent in the conviction rate.

The international criminal jurisdictions which have been established in the last decade have not been set at odds by this variety of national experiences: they have profited from it. The common law adversarial trial has proved a viable model for hearings of the sort that the Milosevic case has made familiar. It is not only that it confronts accused with accuser in a public forum; it seems to offer the best prospect of sifting reliable from unreliable evidence. Truth-seeking on a large scale may be one way of seeing the process; but the job of the International Criminal Tribunal for the Former Yugoslavia is not to write the recent political and military history of Greater Serbia. Deciding whether the indictments laid by the prosecutor against Milosevic and others have been proved is as much truth as a court of law is equipped to cope with.

There are now five international criminal courts. The big one, less than a year old and yet to try its first case, is the International Criminal Court set up by the 1998 Rome Treaty. Two, for the former Yugoslavia and Rwanda, were established in 1993 and 1994 by Security Council resolutions. Two others, covering East Timor and Sierra Leone, have been set up by the UN, the former under its transitional administration, the latter by agreement with the national government. Both of them have the power to deal with war crimes and crimes against humanity as well as crimes in domestic law, and both have finite lives. (With UN backing, Cambodia has just set up its own two-chamber tribunal to try members of the Khmer Rouge.) The Sierra Leone arrangement shades into the Lockerbie one, in which a Scottish prosecution was conducted on neutral territory – the Netherlands – in return for Libya’s surrender of the two suspects. All start from the difficult premise that a court of law without a parent state has only those coercive powers it can beg or borrow. They have been particularly dependent for this reason on external ground forces or internal divisions to catch their suspects.

These courts are inevitably feeling their way. They are staffed by a probably unique mix of common law and civil law judges and lawyers. Their constitutive documents offer the kinds of procedural guarantee that modern Anglo-American process takes more or less for granted. But they also permit, on the civil law pattern, a closer relationship between the prosecutor and the court. This can be dangerous, but it can also be healthy. It can stop a prosecutor from doing what investigators commonly do, getting carried away with a single theory. But it can also make it possible for a strong prosecutor to influence a weak court. The statutes of the ICC and the Yugoslavia and Rwanda tribunals say that the prosecutor is ‘a separate organ’ of the court or tribunal – a metaphor nicely eliding the separation familiar in common law systems with the symbiosis of civil law systems; or, if you take another view, a diplomatic oxymoron. But it has enabled prosecutors to produce more manageable and tenable indictments, and it has protected the courts from being presented with untriable cases. Even so, it needs to be watched: there is relatively little preliminary management that cannot be done openly in the presence of the accused.

Perhaps the most intractable problem, to which no legal system has a comprehensive answer, is the protection of frightened or vulnerable witnesses. What do you do when the source of key testimony is someone whose life will be put at risk if their identity is disclosed, but whom the accused cannot effectively challenge without knowing who they are? Early in its life, the Yugoslavia tribunal got its fingers burnt in the Tadic case, when a damning prosecution witness whose identity had, on seemingly good grounds, been kept from everyone including the accused, was accidentally identified by the defence lawyers and found to be a pathological fantasist. The Yugoslavia tribunal has never made another such order. Instead, it will grant anonymity from publicity and will disguise the witness’s image or use screens where it is satisfied that protection is required. An inquisitorial system which heard and evaluated such evidence out of the accused’s hearing would no doubt obviate the problem, but it would not be a system of justice.

One efficiency of many inquisitorial systems is that evidence assembled by the prosecutor does not need to be repeated orally at trial if it is not directly challenged. At its most brusque, this can mean that the entire prosecution dossier, starting with the investigator’s report, is placed before the trial court as evidence. The Yugoslavia tribunal has refused to let this be done. It has shut out the investigator’s report, on common law principles, as weightless hearsay. Although the tribunal has power to admit written testimony, in fairness to the accused it has declined to admit panoptic accounts of events unless the author is available for cross-examination. In a court largely concerned with military and paramilitary activities and their effects on populations, such a stance is time-consuming but indispensable.

The South African jurist Richard Goldstone, the tribunal’s second appointed prosecutor, who saved it from an early demise and made it a court to be reckoned with, remarks in his foreword to Archbold: ‘There were no precedents on which we could rely. We came from many legal backgrounds and we tended to rely on our own training and legal systems. Our task, however, was to construct a new international jurisprudence.’ It’s an encouraging fact, visible in these and other international courts and tribunals, that when faced with common problems, divergent legal systems and traditions tend to converge in devising workable solutions. It’s when they remain in their own territorial boxes that convergence is slower. It took Italy until 1988 to modify Mussolini’s criminal code and, by enacting a criminal procedure code, to move towards an adversarial trial system; and it was only on the eve of the millennium that the principle of contradittorio – the adversarial principle – finally entered the Italian Constitution.

There is really not much to make one nostalgic today for the criminal trial process with which England and Wales and the North American colonies entered the 18th century. The rolling process of change since then, both in common law and civil law systems, has been erratic and imperfect but not negative. One thing, however, that has not changed since the 18th century and probably should have done is the provision by the City of London of a formal daily lunch for the Old Bailey judges; though at least, with the abolition of 15-hour sittings at which the judges presided in relays, the burghers are spared the Ordinary of Newgate’s ordeal of dining with them at 3 p.m. and again at 5 p.m. It was some years ago at the modern lunch, still eaten in wigs and robes, that an Old Bailey judge with a reputation for snobbery let drop that he had spent the weekend with the Bedfords at Woburn. ‘How nice,’ said the acidulous Mr Justice Melford Stevenson, who was presiding. ‘What’s the admission charge these days?’