Trials and Miscarriages: an evolutionary socio-historical analysis

Periodically miscarriages of justice become newsworthy and inform not only those who may have some responsibility for their occurrence and rectification, but the general public as well. At those times proposals for reform tend to ensue, and reforms occur. But such occasions are rarely considered historically or understood from an evolutionary perspective. This article undertakes to offer that missing feature. It attempts to inform the periodic highly charged discussion of miscarriages of justice with an understanding of their ingredients illustrated by both some recent and some much older history. The article presents the thesis that miscarriages of justice are a component of the workings of all criminal justice systems, part of their operations, rather than their malfunction. It shows how miscarriages of justice are the criminal justice system’s answer to a prior problem, the functional need to convict more persons than can be shown, with certainty, to have committed the crimes of which they have been charged. This thesis has the implication of inserting some modesty into proposals for reform, not to decry their attempts, but to inspire less naivety. The article focuses on the changing methods of the criminal trial, throughout the second millennium and up to the present day, as an expression of the underlying problématique that represents its thesis.


I INTRODUCTION
Since the 1990s, especially with advances in the use of DNA (deoxyribonucleic acid) as a forensic method within criminal justice, there has been considerable attention directed to miscarriages of justice. 2 Scholars publish articles and teach courses debating the definition of miscarriages, the scale of their occurrence, and the need for legal reforms to ensure their prevention and rectification. In the US, the debate on miscarriages has been intertwined with the issue of capital punishment. 'Hard' scientific evidence that the DNA of convicted murderers and rapists does not match that left by perpetrators on the murder weapon or at the crime scene has produced a list of persons who, but for this new evidence, would have been executed for crimes that others had committed. 3 These examples of 'exoneration' have stimulated the setting up of 'Innocence Projects' 4 dedicated to investigating the procedures, practices and outcomes of the criminal justice system leading to the conviction of persons claiming to be innocent of the crimes of which they have been convicted. The activities of these Projects (as well as other campaigning organisations and ad hoc individual campaigns) and how they might affect the work of the official channels for investigating claims of miscarriage of justice, has in turn become part of the debate. Their focus on factual innocence is regarded, by some, as likely to undermine support for the rights, constitutional and otherwise, afforded to defendants within the criminal justice system. 5 Meanwhile, within the legal system, the judiciary makes repeated attempts to re-define what constitutes a miscarriage of justice, both within the process of appeals, and in the context of claims for compensation brought by those whose convictions have been quashed. These definitions are inconsistent and unstable. 6 Footnote 4 continued national Commission of Jurists, apart from relevant government departments and quangos), and many ad hoc campaigns (see, for example, http://www.insidejusticeuk. com), journalistic initiatives and even television series (such as in the UK BBC's Rough Justice and Channel 4's Trial and Error). And significantly, in the UK, since 1997, the government financed Criminal Cases Review Commission -https://www. ccrc.gov.uk as set up by the Criminal Appeal Act 1995. In the UK the name Innocence Project has been used as an umbrella term for contemporary organisations associated mainly with university law departments, significantly at Bristol -see M. Naughton (ed), The Criminal Cases Review Commission: Hope for the Innocent? Requa 'The Supreme Court on Compensation for Miscarriages of Justice: Is it better that ten innocents are denied compensation than one guilty person receives it? ' (2012) 75 The Modern Law Review 387; and, as a short summary of recent jurisprudence including cases before the European Court of Human Rights, highlighting the proposition that 'The law on compensation for miscarriages of justice is currently in a very unsatisfactory state', A. Bailin and E.

TRIALS AND MISCARRIAGES
The activity stimulated by miscarriages goes beyond attempts to secure the exoneration of particular individuals. As already mentioned, exonerating those sentenced to suffer the death penalty operates as part of the campaign to remove this penalty by, in particular, offering a challenge to those who seek to defend it on a retributive basis. 7 At a more general level, miscarriages of justice provide a basis for assessing many of our current systems of criminal investigation, evidence, procedure, trial and appeal. If miscarriage of justice is a moral wrong, then avoidance of miscarriage is a moral good. This provides a perspective for assessing whether any part of the criminal justice system contributes to the wrongful conviction of the factually innocent or, though this is a more controversial question, whether conviction has been the result of procedures that are 'unfair', or involve an abuse of process, or in other ways fail to satisfy the requirements of due process and the rule of law. Some of this examination typically takes a prospective or hypothetical approachassessing aspects of the criminal justice process in terms of their likely contribution to future miscarriages of justice. But, and most notably in death penalty cases, actual examples of generally accepted miscarriages offer increased support to arguments for reform. Rather than the claim that a particular unreformed practice 'might' cause a miscarriage, one can point to the generally accepted cases, demonstrate the causal connection with the unreformed practice, and assert that this 'has' caused a miscarriage with the implication that, if it can be shown to have led to one miscarriage, it can be expected to lead to others. 8 In this article we subject miscarriages of justice to an evolutionary socio-historical analysis. By this we do not mean the application of sociological methods, quantitative or qualitative, to particular aspects of the criminal justice system, in order to generate 'social facts'. Such methods can produce information about the practices and has very little use or need for history. The empirical focus is on the present -what is happening now. Normative frameworks are usually similarly contemporary. If the focus is on the ability of the legal system to establish the truth of factual guilt or innocence, then the usual starting points of comparison are the most up to date bases for establishing empirical truths -current methods of forensic science, current knowledge about statistical proof, etc. If the normative framework is rights based, then typical sources for comparison are either 20th century human rights instruments, such as Article 6 of the European Convention on Human Rights, or current interpretations of rights based on constitutional protections that have existed since the end of the 18th century.
The socio-historical analysis we offer here does not seek to increase our empirical knowledge of contemporary criminal justice practice, but to subject those practices to an evolutionary approach. As such, we draw upon historical sources. In doing this, we attempt to go beyond a simple narrative of events, what Langbein has called a "story of how we came to live under a criminal procedure for which we have no adequate theory." 11 Using these sources, including some of Langbein's own works, we examine criminal conviction as a solution to a problem that needs to be solved (but never ultimately resolved) within changing societies, and the implications of legal systems' changing solutions to this problem: the need to convict more persons than we can know, with certainty, have committed the crimes of which they are accused. From this perspective, miscarriage of justice does not pose an isolated problem or even perhaps a complex of problems, which can be eliminated by specific reforms. Rather, miscarriages of justice are themselves the solution to a prior problem. The prior problem, and its various solutions, has a history, or rather, a describable evolution. Considering each stage in the evolution of trial procedures in light of this same problem, corresponding to what might be better described as its proble´matique, avoids our reading of contemporary arrangements as necessarily more fair or accurate than their predecessors. It also allows us to scrutinise the suggestion that our current arrangements have come into existence because they are 11 J. Langbein, The Origins of Adversary Criminal Trial (Oxford: Oxford University Press, 2003), 9. Langbein's approach and prognosis, which we find persuasive, is however disputed by others, for example the editors, A. Duff, L. Farmer, S. Marshall and V. Tadros of The Trial on Trial series of books (Volume 1: Truth and Due Process, Volume 2: Judgment and Calling to Account, and Volume 3: Towards a Normative Theory of the Criminal Trial (Oxford: Hart Publishing, 2004, 2006, 2007; see Vol 3,[19][20] more fair, accurate or rational than their predecessors. One substitutes an understanding of evolution in terms of the repeated solutions to a problem representing an underlying functional need, to its understanding in terms of progress towards a superior present. The article begins with a consideration of the background proble´matique of miscarriages of justice, and the solutions offered to it in the past. In constructing that probe´matique, and tracing the evolution of its solutions, we draw on the work of Niklas Lumann, particularly his earlier work A Sociological Theory of Law, and his later Law as a Social System. 12 Ours is a functionalist analysis. Following Luhmann, we are not claiming to attribute a single function to the criminal justice system, or law more generally. Nor does our analysis involve claims that society has inherent needs which the criminal justice system responds to automatically. Rather, our analysis proceeds on the basis that its criminal procedures, amongst its many and varied performances, offer a solution to a particular proble´matique: the need to identify those who have committed crimes in situations where one cannot be sure of the truth of what has occurred (what we have on other occasions described as its 'truth deficit'). 13 That proble´matique presented itself in the past, and continues to present itself in the present. Looking at the history of criminal processes (most notably the form of trial) as a solution to this proble´matique alters one's view of that past. In particular, it offers an antidote to understanding the past as a period in which there was less concern with factual innocence, or a blind faith in the ability of contemporary procedures to separate the factually guilty from the 12 For those who have some specialist understanding of Luhmann's work, we need to stress that we do not regard the autopoietic turn within Luhmann's writings as something which creates two separate bodies of work which cannot inform the same piece of research, one with an emphasis on functionalist analysis and the other on autopoiesis. We would argue that how a system's autopoiesis continues can be illustrated by focussing on the functionalist approach of Luhmann's work. The problems solved by autonomous autopoietic sub-systems, which we can understand as generated by functional need, are the restrictions on evolution presented by a system's need to continue itself autopoietically. (See N. Luhmann, Law as a Social System (Oxford: Oxford University Press, 2004, trans K. Ziegert), ch 3 'The Function of Law', esp 147-156; N Luhmann, A Sociological Theory of Law (2nd edn., Oxford: Routledge, 2014, trans. E. King-Utz and M. Albrow), ch 4 'Positive Law ', esp 167-174). 13 See, in particular, our chapter, 'Theorising the Criminal Trial and Criminal Appeal: Finality, Truth and Rights' in Duff et al, n 11 above, Vol 2, ch 14. This phrase is also used as a heading by Langbein in the final chapter 'The Emergence of Adversary Trial' of his 2003 book, n 11 above, at 331. factually innocent. It also operates to diffuse an understanding of our current procedures as the latest stage in a process of progressive improvement -that our system tolerates a lesser number of miscarriages of justice than its antecedents, through being more rational, scientific, or just. Instead we invite the reader to regard our current procedures as, from an evolutionary perspective, simply the current solution for the background proble´matique we have outlined. This history will, we believe, give us a different understanding both of our current arrangements, and their capacity for further change.

II THE EVOLUTION OF MISCARRIAGES AS PART OF THE EVOLUTION OF LAW
There is some agreement among common functionalist explanations that, at least in the West, the origins of law lie in the development of alternatives to blood feuds. 14 These feuds offer the prospect of neverending retaliation. One reason for the inability to end blood feuds is the lack of agreement as to what constitutes an adequate response to a wrong. Unless both sides to the feud share a common standard for retaliation there is a constant risk of what one side regards as an excessive retaliation, which is then a provocation to further retaliation. This problem can be solved by encouraging both sides to agree to common standards for compensation and punishment. That is why some of the earliest varieties of law take the form of stipulated compensations to be paid for different kinds of wrong. But the stipulation of penalties is only part of the problem. The instability of blood feuds is not only a result of disagreement over the responses appropriate to an act, it also involves disagreements over the act itself. Have cattle been stolen, and if so by whom? For in the blood feud, instability arises from different accounts of what wrongs have occurred, as well as how their perpetrators should be treated. Forcing the protagonists to submit their dispute to third parties for adjudication, and requiring them to accept authoritative pronouncements of what has occurred, is as important as requiring them to accept the penalties stipulated for the wrongs in question.
14 "It is commonly known that the early forms of legal procedure were grounded in vengeance. Modern writers have thought that the Roman law started from the blood feud, and all the authorities agree that the German law begun in that way." O. W. Holmes, Jr., The Common Law, (Boston: Little, Brown & Co., 1881), ch 1 'Early Forms of Liability', 2.
Just as the list of crimes and their punishments has varied between societies and over time, so too have the methods for deciding whether the crime (or other infraction) has occurred and who is responsible for it. But there are features which have remained constant. The problems to be solved have implications for the standard and kinds of proof required. Despite many rhetorical flourishes, absolute proof of guilt is too high a standard. If third party adjudication only resulted in the punishment of persons of whom there was absolutely no doubt as to their having committed the crimes in question then it would have had a much more limited role in resolving accusations of wrongfulness than it has consistently played. The underlying proble´matique has been, and continues to be, to provide forms of adjudication which are able to secure convictions in the presence of less than absolute evidence of guilt. 15 We can see this proble´matique, and various attempts at its resolution, in the history of Western forms of criminal procedure. This is a history in which each dominant form of procedure has, during its operation, suffered from known defects, especially in terms of likely miscarriages -allowing guilty persons to escape conviction and convicting the innocent. In each case, the reform of that procedure has not been a straightforward response to criticism, as the weaknesses of the procedure were more or less apparent throughout its use. The key to evolution has not just been criticism, but the generation of alternative solutions to the same proble´matique -solutions that appear to resolve that underlying proble´matique and its implications for miscarriages, which miscarriages however re-appear in a different form. What follows is an epigrammatic discussion of the history of criminal trial procedure (principally, as we approach the last two hundred years, by using only Anglo-American examples) in order to illustrate this universal proble´matique. We present this epigrammatic discussion in terms of the ability of each procedure to solve the need for conviction in the absence of absolute proof. It is a simplified history, and one in which one form of procedure can be seen as not simply replacing another at one time, but with different procedures overlapping, and with examples of 'earlier' procedures still operating, on some occasions, 15 It might be that 'the public' are either more aware of this, or more willing to recognise its implications than 'authorities'. Indeed, there is ample modern evidence that 'the public' recognise that miscarriages occur with some frequency and, probably, more frequency than that recognised by many official sources: see, for example, M. Zalman, M. Larson, and B. Smith, 'Citizen's Attitudes Toward Wrongful Conviction' (2012) 37 Criminal Justice Review 51. for example in some parts of Europe, hundreds of years after they had gone into decline within other particular jurisdictions. 16 III A BRIEF HISTORY OF THE CRIMINAL TRIAL It is convenient to start this brief history with trial by oath and combat, both of which can be traced back to Biblical, Roman Law and other references. Trial by oath involved the defendant denying the accusation by swearing to her/his innocence. In the case of trial by combat, those who swore oaths in their defence, effectively defaming their accusers as liars, could in turn be challenged to trial by combat over their slander. In procedural terms, in relation to most examples, oaths were not available to those who were caught in the act, while combat was equally not available in such circumstances or others, such as where there were witnesses or confessions. 17 They applied only in circumstances where guilt could be a matter of dispute. A single oath was considered insufficient, so the defence involved the use of conjugators -persons who would risk their own souls by swearing to the truth of defendants' innocence. In effect, and anticipating later features of the law of evidence, the conjugators were endorsing the character of the defendant. The major weakness of the oath, the willingness to commit perjury, was known to all involved (as was the conception that the survivor of a single 'judicial duel' was determinative of who could be proclaimed to be 'right'). In response, elaborate rules evolved to determine the number and status of the conjugators required for each kind of defendant, as well as the collateral addressee of the oath (God and a relic, God and my children, etc.). Each elaboration, in responding to the underlying risk of perjury, also drew attention to the 'miscarriages' likely to result. Whilst the primary risk of perjury might be diminished, these measures created their own bases for injustice -the greater ability of those able 16 For example, trial by battle/combat was only formally repealed in England for some very serious offences in 1819: Stat. 59 Geo. III. C. 46. That said, its operation had always been sparse in England, unlike within various German jurisdictions which, nevertheless, had ended its operation some two hundred years earlier. (See B. Thayer, 'The Older Modes of Trial' (1891) 5(2) Harvard Law Review 45, at 65-70).
17 "… even in the earliest times, this mode of proof was only an expedient resorted to in cases of doubt and on the necessity of its use the rachinborgs or judges probably decided." H. Lea, Superstition and Force: Essays on the Wager of Law -The Wager of Battle -The Ordeal -Torture (Pennsylvania, 1866), 40. https://archive.org/details/ superstitionforc00leahrich. to meet the required formalities to escape conviction, and the greater likelihood of convicting those who lacked the necessary social relations for conjugation. Confidence in the oath diminished further when the swearing required of conjugators was changed from an affirmation of the facts of the defence, to a mere affirmation of their belief that the defendant was telling the truth. 18 In its turn trial by ordeal, 19 for which there are many ancient, Biblical and Medieval references, operated to ameliorate the problems encountered with adjudication via the oath or combat. By contrast to the oath, both the ordeal and trial by combat provided an immediate physical deterrent for those tempted to deny wrongdoing and swear oaths in order to avoid the punishment and compensation sought by their accusers. In the case of the ordeal, denial would lead to an accused having to show their innocence by, for example, carrying or walking on a heated iron, or suffering near drowning by failing to float when submerged in water. As was the case with oaths, and later with torture, the purpose of the ordeal was to allow the punishment of persons for whom there could be some doubt as to their guilt. If a person was caught red-handed then the ordeal was not available. The thief caught in the act was not given the chance to prove her/his innocence via divine intervention. Nor could the ordeal be ordered for persons against whom there was no evidence as to their guilt. The ordeal was used to complete the inadequacies of evidential proof -it provided a means to resolve remaining doubts as to the guilt of the offender. If she/he failed the ordeal, or in the case of 18 See H. Black 'Antiquities of the Law of Evidence -Compurgation' (1893) 27 American Law Review 498. 19 For the editors, Duff et al., n 11 above, of the three volumes of The Trial on Trial, the ordeal (and, we surmise, similarly trial by oath and combat) "was not, strictly speaking, a trial in the modern sense, since it was not directly concerned with the discovery of past facts or the weighing of evidence or arguments before a tribunal acting judicially" but "a forum in which the person accused of having committed a crime was called to account for their actions before the community and before God." (Vol 3, 22) But the ordeal did concern itself with facts (whether accused persons had done what they were accused of), and it was applied judicially (within judicial proceedings); it simply did not require arguments or further evidence once the decision to use the ordeal had been taken. These editors (Vol 3, 28-29) similarly deny the accuracy of the description 'trial' (in a modern sense) to the early jury, which used its own knowledge of the parties and the facts to decide on guilt. Their particular approach, which seeks a normative theory for defending the general features of contemporary practice, involves a more limited kind of functionalist analysis than we adopt here: principally they are asking questions about the purposes and legitimacy of modern practices. trial by combat if her/his champion were defeated, then they were guilty despite the absence of irrefutable proof. If they passed the ordeal then the necessary extra element required for guilt was absent and, in these circumstances, it was God and not the authorities who had failed the complaining party. 20 The ordeal did not operate without criticism. Religious objections centered on the manner in which the practice forced God to act if the innocent were not to be punished and the guilty escape punishment, seen as an endeavour to 'tempt' the almighty. It was also unclear how the practice operated alongside that of confession and absolution: did the 'innocence' of a confessed penitent protect them against a finding of guilt through the ordeal? Aside from these religious issues, there was also less than complete faith in the accuracy of its outcomes. For example, the Assizes of Clarendon in 1166 provided that those acquitted by the ordeal who were still considered guilty by 'common report' should nevertheless be required to quit the kingdom on pain of outlawry. 21 And as part of his arguments against its use Peter Cantor, in the 12th century, recounted a case which has echoes of modern miscarriages of justice and how they have seeped into cultural understanding: a pilgrim accused, on his return from the Holy Land of having killed a companion pilgrim, was put to the ordeal, convicted and executed, only to have the unharmed pilgrim return home later. 22 The ordeal was effectively abolished in 1215 when Pope Innocent III issued an edict prohibiting the clergy from administering 20 "The directions of the codes … are generally precise … rendering the case one not to be solved by human means alone, … discretion of the tribunal to order an appeal to the judgment of God." Lea n, 17 above, 251. That said, it would appear that the various kinds of ordeal could offer a form of punishment in themselves: "The graduated scale of single and triple ordeals for offences of different magnitudes is so totally at variance with the theory of miraculous intervention to protect innocence and punish guilt, that we can only look upon it as a mode of inflicting graduated punishments in doubtful cases, thus holding up a certain penalty in terrorem over those who would otherwise hope to escape by the secrecy of their crime …" (255). 21 "…a malefactor who had established his innocence by hot water or iron obtained thereby only a commutation of punishment, and was forced to leave the kingdom in perpetual exile." (Lea, n 17 above, 263). 22 H. Lea, The Ordeal (Pennsylvania: University of Pennsylvania Press, trans A. Howland, 1973), 155. The Ge´rard Depardieu 1982 film 'The Return of Martin Guerre' is an example of such cultural reference. Public awareness of the potential for miscarriages of justice, and public involvement in petitions or other expressions of concern with miscarriages of justice, can be traced to the second half of the 19th century with the advent of popular detective fiction; Wilkie Collin's 1868 novel The Moonstone often being referred to as the first full length detective novel. it on behalf of secular authorities. 23 Its abolition posed an immediate problem throughout Western Europe: how were accusations against persons who were not caught in the very act of committing their crimes to be processed in the absence of ordeals?
On the Continent the remedy was the introduction of a system of trial by torture, the subject of John Langbein's classic work: Torture and the Law of Proof. 24 Langbein argues that the use of judicial torture was adopted because it provided a solution to the proble´matique we have outlined, and which Langbein expresses in the following statement: "No society will long tolerate a legal system in which there is no prospect of convicting unrepentant persons who commit clandestine crimes." (at 9) The need he refers to is not an abstract logically derived societal need, but one that would have made itself felt immediately within legal procedures, 25 with new accusations continuing to be brought to the secular authorities whilst previous accusations remained unresolved. As well as recognising the need for a process that could compensate for the lack of sure proof, Langbein identifies the reason why torture, rather than judgment of those appointed by the secular authorities, was preferred: "How could men be persuaded to accept the judgment of professional judges today, when only yesterday the decision was being remitted to God?" (at 8) More than a description Langbein, in this book, seeks to challenge what he calls the 'fairy tale' that the ending of judicial torture to obtain confessions was a response by enlightened European monarchs to its sustained criticism by liberal writers of the 18th century such as Voltaire and Beccaria. As he points out, the problems of torture as a means to obtain evidence were known throughout its use: that 23 Introduced at the Fourth Lutheran Council, re-enforced by the decree of Honorious III in 1222. See J. Langbein, Prosecuting Crime in the Renaissance: England, Germany and France (Cambridge, Mass: Harvard University Press, 1974), 134. The ordeal was not prohibited, only the participation of the clergy was. But in the absence of the clergy, the claimed connection to God lacked plausibility. Nevertheless the ordeal continued to be used in various jurisdictions, for various offences, despite formal attempts to abolish it throughout the 13th century (see Lea, n 17 above, 272-279). 24 J. Langbein, Torture and the Law of Proof: Europe and England in the Ancien Regime (Chicago: University of Chicago Press, 1977). 25 For a fascinating account of the hiatus that occurred within the English system of criminal prosecution when the ordeal was abolished, see R. Groot, 'The Early-Thirteenth-Century Criminal Jury' in J. Cockburn and T. Green (eds), Twelve Good Men and True: The Criminal Trial Jury in England, 1200-1800 (Princeton: Princeton University Press, 1988), ch 1. someone may know details of the crime without having committed it; that suggestive questioning may occur and not be detected; that torture tests endurance, not truth. And, in a statement that anticipates current debates on miscarriages of justice, that: "Long before Voltaire, French writers of the sixteenth and seventeenth centuries are pointing to cases in which an innocent person confesses and is executed, after which the real culprit is discovered." (at 9) On the Continent, where Roman Law had spread since the 12th century, the introduction of trial by torture had a very specific legal context. The formal system of proof required two eye witnesses to the crime for a conviction. This was full proof. Could the secular authorities restrict themselves to convictions only in the presence of full proof, or would they find ways to convict with less than this? The increased use of judicial torture in response to the 'abolition' of the ordeal was, as Langbein points out, a less dramatic change than it would appear today, as it shares so many of the characteristics of the procedure that it replaced -both trial by torture and trial by ordeal require those who would deny charges of wrongdoing to undergo physical pain. He also stresses that this resort to torture was not a general freedom to apply torture in order to force confessions. It was a judicial application of torture. It required the presence of a professional judiciary throughout the jurisdictions where it was applied. It could not operate with the same degree of plausibility if it was simply a mechanism applied by political rulers against those accused of crimes. And being a judicial device, it was administered via a complex system of rules. Torture was not supposed to be applied where there were two eye witnesses to the crime, so those caught 'red handed' would not be able to elude conviction through their ability to avoid confessing under torture. And at the other extreme, people could not simply be accused by others and then subjected to torture. A person could not be put to the torture in the absence of any other proof as to their guilt. Torture, administered to extract testimony that only the wrongdoer could know, was used to complete the guilt of those for whom proof of guilt was otherwise insufficient. The Romancanon system of proof listed all kinds of indications of guilt that were not evidence themselves, but which were given various numerical values when calculating whether there were sufficient grounds to put someone to the torture. Whilst this required judgment, these detailed rules also gave the Continental system of proof a formal and objective character. Conviction was something more than a political judgment by those in power, or by those that they appointed to judicial positions. To quote Langbein again: The system of statutory proofs was the answer. Its overwhelming emphasis is upon the elimination of judicial discretion, and that is why it forbids the judge the power to convict upon circumstantial evidence. Circumstantial evidence depends for its efficacy upon the subjective persuasion of the trier, the judge. … By contrast, the system of statutory proofs insists upon objective criteria of proof. (at 6) The difficulties caused by the effective abolition of the ordeal were solved in England in a different manner. Torture, though applied (most often through the Tudor period) was never a normal feature of investigation and prosecution. 26 In 1215, when the English faced the same crisis as the Continent, the procedure they adopted to replace the ordeal was jury trial. 27 The jury was already used as preliminary to the ordeal, since only those 'suspected' of crime by either a jury drawn from the hundreds, or the part of that jury selected from the immediate vicinity of the crime (the visne), were required to undergo the ordeal. The remedy for the loss of the ordeal was to offer those suspected of serious crimes, who were facing perpetual imprisonment in the absence of the ability to proceed to the ordeal, the alternative of having the jury decide the issue of guilt in their cases. Faced with death in prison awaiting trial, most chose this alternative. The right to refuse jury trial proved only a short term measure, with defendants soon being required not only to accept the trial, but to begin the trial by pleading guilty or innocent, with 'strong and hard imprisonment' (in practice, what we would now call torture) if they refused. 28 Whilst this is a very different system, one can see that it represents a 'functionally equivalent' solution to trial by torture. The involvement of the jury offered a local solution to what would have been a local accusation of wrongdoing. The jury was chosen from persons expected to know the accusers and the accused, and to use their knowledge of this and the circumstances of the crime to reach a verdict which was acceptable to the local community. It advanced a local truth, based on local assumptions of character as well as evidence going to the immediate circumstances of the crime. 29 Whilst this is viewed with hindsight as a more rational 30 way of finding the truth than the ordeal, its satisfactoriness appears to lie in its competence to adjudicate between accusers and those accused in a manner likely to prove acceptable to the local community.
By comparison with the Continental system of formal proofs, reliance on local juries seems to have offered a particularly subjective basis for conviction; nevertheless it shared the feature of providing a more plausible grounding than would have been possible if guilt were solely based on the judgment of those appointed to be judges. At the beginning of the Common Law, English Justices deferred to local communities in deciding both what constituted a 'wrong' and whether a particular person had committed such a wrong. The local character of substantive legal rules was quickly lost as the individual Justices, conferring amongst themselves, developed a Common Law for the realm. By contrast, the local character of fact finding underwent a more gradual process of erosion. Whilst the formal rules (the directions to the Sheriff to select juries in advance of an assize) continued to require the jury to include persons living close to the scene of the crime, the problems of ensuring sufficient numbers of 29 "… they might then be expected to know something not only of 'the truth and nature of the offense' but also of the 'quality of the [prisoner] … and happily the credit of the accuser and his witnesses'." (Zachary Babington, Advice to Grand Juries in Cases of Blood (1667) pp 3-4 quoted in J. Beattie, Crime and the Courts in England (1660-1800) (Oxford: Clarendon Press, 1986), 440. This quote is far more apt to understand the beginnings of jury trial, than was the case in the 17th century. The difficulties of finding local jurors led to the selection of men from a wider area -the hundred, and eventually the county. There also developed a practice of selecting men who had to visit the assizes for other reasons -coroners, witnesses, prosecutors and court officials, in order to make up the numbers (talesmen). Research which has looked at the location of those tried, and compared it to the homes of jurors, indicates that the ability of the jury to decide the issue of guilt substantially through their knowledge of the crime declined much earlier than was previously thought, perhaps as early as the 14th century. See J. Post 'Jury Lists and Juries in the Late Fourteenth Century' in Cockburn and Green, n 25 above, ch 3. 30 That said, however: "The ordeal is an appeal to the supernatural, was not the jury exactly of the same sort? … It gave a verdict, guilty or not guilty … It was nobody's business how that verdict was reached. Like the ordeals, the jury also was inscrutable." T. Plucknett, Edward I and Criminal Law (Cambridge: Cambridge University Press, 1960), 75. jurors to serve the assizes led to the use of juries staffed by persons who lived too far away from the crime to have local knowledge of its circumstances. The local jury that came to the assizes 'to speak, rather than to listen' was worn down, having exceeded the willingness and ability of Sherrifs to summon them, and their willingness to undertake this civic responsibility. But rather than stop trials, trials themselves altered, with the calling of jurors with less property, who lived further from the scene of the crime, with more jurors who sat repeatedly, and with juries who would otherwise be inquorate being made up of persons who had other reasons to be at the assizes (talesmen). This must have changed the nature of the jury's role, for the lack of actual knowledge of the facts would have necessitated a procedure that allowed much more scope for the presentation of evidence to the jury. What altered in turn was the role played by reputation. With a jury drawn from the locality of the crime actual knowledge of the parties would have played a central role in the decision: with those known to be respectable being less easily convicted than those with dubious reputations, and outsiders being more easily accused than insiders. 31 With jurors drawn from a wider area, character and reputation still formed part of the assessment, but in the form of a judgment of what might be expected from persons with the status and role of the accused and defendant, rather than a community's tangible experiences. Though lapses from the official standards for the selection of jurors varied between different circuits of the assize, over different periods and even between different towns on the same assize, they nevertheless gradually transformed the jury into a non-local, non-self-informed body of the 'middling sort', 32 31 "Changes since the eighteenth century … the inexorable move to urbanization … deprived the criminal justice system of reliable sources of local knowledge on which the early eighteenth century criminal process … had been able to rely in the prosecution of 'insiders', while also undermining the long-standing English practice of judging 'outsiders' in terms of their appearance." N. Lacey deciding guilt or innocence with a heavy reliance on the judge and a core of experienced jurors. 33 We have little evidence of how jury trials were actually conducted between the 12th and early 18th centuries. 34 The legal system had no need for records, since there was no provision for appeal against how a trial was conducted, or the verdict it reached. 35 From accounts that show the time taken to complete the assizes, and the number of prisoners dealt with, we know that jury trial was short: "Where criminal trials before a jury are today measured in days and weeks, if not months, a trial in 1700 would be measured in minutes, only 33 And linked to this, when it is no longer possible to suppose that the jury, being in sole command of the facts, would have been able to nullify the law by failing to find persons guilty of laws or the applications of laws with which they disagreed. See Green, n 31 above, ch 9, esp 356-366. 34 Some idea of the nature of trial during this period takes the form of speculations of what would be possible, on the basis of the little that can be known. For example, lists of jurors called, empanelled and actually serving can, where their home villages or occupations are given, provide evidence as to how local they were to the crimes on which they adjudicated, which in turn points to the need for evidence to be presented. And the usually large number of arraignments allocated to a single jury during the period of a single assize, gives some idea of the limited possibilities for evidence or deliberation thereon. See J. Baker, The Oxford History of the Laws of England, Vol VI: 1483-1588 (Oxford: Oxford University Press, 2003), 516-520, who gives a description of trial in the 16th century using a manual from 1550. 35 Contemporary statements about the nature of those trials are suspect. There is some tendency (still common!) to hold the trial up to a standard supposed to be generally accepted, and then point to some examples of a failure to meet those standards as evidence of a general problem. For example, the property qualifications were interpreted as a means to ensure juries of a high calibre, and evidence that Sheriffs were substituting jurors who did not meet those standards interpreted as evidence that actual juries were inferior. Lack of any research at the time into what actually occurred allowed three different contemporaneous observers on these practices to make contradictory claims. The practice of meeting the property qualification for jury selection varied considerably, and where it was lowered, actual selection still excluded those who, it was feared, might acquit against the weight of evidence ( occasionally in hours, never in days." 36 By the late 17th and early 18th century, alternative records (newspaper accounts and judges' diaries) became available. From these and other sources we know that defendants were not allowed access to lawyers to present their cases, and that prosecutors, though allowed lawyers, rarely employed them. Cross examination of witnesses, such as occurred, was undertaken by the judge. Defendants could not give sworn evidence, which is not to say that they did not participate in their trials. The swearing of oaths as a defence against accusations was, we must remember, a form of trial that had been largely displaced by the ordeal. Rather than risk their souls by swearing to an untruth in order to escape punishment, defendants were free to make unsworn statements. 37 Indeed, as the prosecutor's evidence was being presented, defendants would repeatedly be asked to answer it, with failure to do so being taken as acceptance of the truth of what was being said. The speed of these trials was made possible by the absence of lawyers, and by the experience of jurors, who not only heard numerous trials over the several days of the assizes, but were likely to be manned by a significant number of members who had been called for jury service before. 38 How should we understand the manner in which this form of jury trial separated the innocent from the guilty? We must resist the temptation of treating every difference from our current procedures as a shortcoming, evidence that our current procedures represent a 36 J. Beattie, Policing and Punishment in London 1660-1750: Urban Crime and the Limits of Terror (Oxford: Oxford University Press, 2001), 259-260. 37 The sentiment supporting such a rule was strong on both sides of the Atlantic, sustained by many statements about the risks involved in allowing sworn evidence from those who so obviously had a strong interest in the outcome of the trial. In England it was not until the Criminal Evidence Act 1898 that defendants were allowed to testify at their own trials (while in the US, such reforms occurred in various States from the 1860s). That said, defendants' rights to give evidence under oath, and that of the witnesses called to assist them, was not a concession simply intended to assist defendants, but also to re-introduce the threat of perjury (with its criminal and judgment day sanctions), and to deter false evidence which might prevent convictions. (See D. Bentley, English Criminal Justice in the Nineteenth Century (London: The Hambledon Press, 1998), especially chs 16-18). 38 In brief, see Beattie, n 36 above, 266-267; for a fuller account, see Hay, n 35 above; and, on the make-up of juries, see J. Beattie, 'London Juries in the 1690 s' ch. 8, and P. King, '"Illiterate Plebians, Easily Misled": Jury Composition, Experience, and Behavior in Essex, 1735-1815' ch 9, both in Cockburn and Green, n 25 above. greater commitment to separating factual innocence from guilt. 39 Such forms of comparison are particularly misleading if we simply compare jury trial in the 21st century with its operation in earlier centuries. For a start, only a relatively small proportion of current defendants in the UK and US 40 actually experience jury trial, even in cases involving serious crimes of dishonesty or violence. The over seventy percent of UK defendants facing serious charges who plead guilty are 'encouraged' by the knowledge that this may well result in a substantial reduction in the final sentence. 41 By contrast, in the 18th century almost all those charged with felonies in England experienced an actual trial before a jury. Trials, though short, were considered important. Guilty pleas were actively discouraged, even in cases where part of the evidence to be given was a prior confession before the magistrates. 42 This should not be interpreted as a hostility to the evidential value of confessions. It was more to do with the different role played by the jury and judge at this time. In a period when so many offences had a statutorily fixed penalty of death, the jury and the judge had the task of selecting who, amongst those facing the death penalty, might be spared. The trial offered an opportunity to reach these decisions, with juries undertaking, often at the direction of the judge, to find guilt on a lesser charge, in the face of the evi-39 For a classic example of such an approach, see G. Williams, The Proof of Guilt: A Study of the English Criminal Trial (3rd edn., London: Stevens and Sons, 1963), ch 1 'The Evolution of the English Criminal Trial'. 40 For example, the US early 21st century nationwide figures, which vary considerably across jurisdictions, suggest an average of above 70% of guilty pleas: see US Sentencing Commission, Sourcebook of Federal Sentencing Statistics, annual reports. The Bureau of Justice 2007 report: Statistics for State Court Sentencing of Convicted Felons, indicates that two-thirds of felony defendants were convicted and, of those convictions, 95% were made up of guilty pleas. At a similar time in the UK Crown Courts: "guilty pleas rose from 56% in 2001 to 70% in 2010." (Judicial and Court Statistics 2010, p 90) Statistics for Magistrates' Courts suggest that since 2000 an average of over 90% of defendants enter guilty pleas, which dispose of many potential trials that might, in previous times, have been heard by juries. However, those statistics are complicated by the particular time when guilty pleas are entered (if, after the beginning of a trial, then described as 'cracked trials'). 41 See our later discussion. This expectation has received some recent fine tuning in the UK, with the sentence reduction being adjusted to take account of the point in the process when the defendant pleads guilty: see Criminal Justice Act 2003, s. 144 (1). The guidelines for such an approach are currently being reviewed: http://www. sentencingcouncil.org.uk/wp-content/uploads/Reduction-in-sentence-for-a-guiltyplea-consultation-paper-web.pdf. 42 Langbein, n 11 above, 18-20. dence, in order to allow the alternative sentences of branding and transportation. When the assizes ended, the judge had a role in further reducing the pool of those facing death by identifying some for conditional pardons. 43 Without seeing defendants responding to the accusations made against them, and giving them the opportunity to present character witnesses in their defences, this sorting would have been made much more difficult. 44 What then of the attitude towards guilt versus innocence, rather than sentencing? In his assessment of the 18th century trial, John Langbein concludes (for example in the final chapter of his 2003 book) that the nature of the offences that were brought to trial meant that the preponderance of cases involved no significant contest over the issue of guilt. With no police force to carry out investigations the majority of prosecutions were brought by the victims themselves, without lawyers, and their accused were persons caught in the act of committing the crime, or caught in possession of the stolen goods with no credible explanation of how they came to have them. For these accused, the trial was mostly a sentencing procedure, with whatever evidence they could present regarding their overall character (employment history, persons who could speak to their good character) being less important as proof of innocence than as evidence that they were not so far down the path from honesty to depravity as to justify being made examples of through hanging. But the acquittal figures suggest that innocence was indeed taken seriously. Around a third of the trials at the criminal assizes ended in acquittal. 45 Relying on reports of trials at the Old Bailey, Langbein attributed acquittals to two kinds of cases. Juries were reluctant to convict on the basis of a single witness, and they tended to acquit whenever there was genuine doubt over identification.
We can see the importance of credibility within these trials. The prosecution case was usually uncomplicated, but it depended crucially on the jurors' assessment of the truthfulness of prosecutor and defendant. This kind of decision was familiar to them: "For juries made up of employers and masters, of men experienced in civic affairs as well as the ways of the court, making judgments about men and women who were not unlike their servants and employees was a natural and familiar activity." 46 Trial at this time was essentially still a confrontation between victim and accused. The victim had to present a cogent reason for believing that the accused had committed the crime, and the defendant would be asked by the judge to respond directly to the evidence as it was given. Confessions obtained by the accuser through promise of non-prosecution were routinely discounted for a number of reasons, including their obvious encouragement to the innocent wrongly to admit guilt. 47 Ideally, the victimprosecutor would have a case supported by at least one other reliable witness to the crucial facts. The defendant would have the opportunity to explain her/himself. Within this process, as with the kinds of judgements made of servants and employees that it duplicated, the ability to tell who was telling the truth was crucial. This made character important to the issue of guilt as well as sentencing. The character of the parties was crucial to the assessment: what kind of person was one dealing with -a known rogue, or an honest person. 48 Against that background, the jury observed the responses of the 46 Beattie, n 36 above, 273-274. 47 A judge "might disregard the impropriety if there were other strong evidence of the prisoner's guilt. But he might equally reject the confession entirely, particularly where it formed the core of the prosecution's case." (Beattie, n 36 above, 413.) Langbein traces the beginnings of a rule excluding confessions tainted in this way to the late 1730s (n 11 above, 220). One can assume that jury members were as aware as judges of the ordinary person's fear of the criminal law, and the unreliability of a confession encouraged by a promise of non-prosecution; so the discounting of such evidence would have preceded the rule-based practice of excluding it from consideration. 48 "The character of the prisoner (in the sense of his disposition and his reputation) was especially important information and was often crucial to the outcome of the trial. That consideration encouraged prisoners to bring character witnesses to speak on their behalf. One can see from the questions often put to such witnesses that what was important was their evidence about the prisoner's habits of life: did he work regularly; did he support his family; was he sober and honest in his dealings with others; did he, in other words, have an established place in a community, and was he known to his respectable neighbors as a man who could be trusted? Or, on the other hand, was he a troublemaker and an idler, a man without visible means of defendant to the evidence of the victim prosecutor, and decided what credence it could be given. For those administering such trials, the possible benefits of legal representation were outweighed by the loss of opportunity to observe those accused speak in person in direct response to whatever evidence justified their having to explain themselves. 49 These trials were not regarded as foolproof, as evidenced by the reasons given by judges in their subsequent recommendations for pardon.
Judges sought pardons in murder cases in the eighteenth century for men who they thought had been wrongly convicted by juries that had acted out of prejudice or that had been heavily influenced by local opinion or that had misunderstood the evidence and had insisted on convicting in the face of judicial advice that the facts could not sustain such a verdict in law. In other felonies judges can be found occasionally recommending a pardon for men convicted upon evidence that in their view was insufficient to sustain the charge. 50 This use of the pardon did not however indicate a problem with miscarriages of justice. Because the jury, at least since Bushel's Case, 51 could not be compelled to convict those that they regarded as innocent, the judge's power to recommend a pardon represented a second chance for those who were found guilty to escape punishment. As such, it was seen as a reason why appeals against wrongful conviction were unnecessary -evidence that the process worked rather than the converse. The fact that the pardon excused punishment without undoing the finding of guilt does not seem to have raised the concerns it does today. 52 There was some contemporary criticism of the system of jury trial during this period, even amongst the judiciary. A minority questioned the benefits of asking a defendant to reply to evidence without notice, Footnote 48 continued support who might be presumed to live dishonestly? A man who could produce no witnesses was likely to have a difficult time in court." Beattie, n 29 above, 440. 49 "When prisoners did speak, their testimony was given great weight, especially the way they asserted their innocence -or at least that was always said on behalf of the rule prohibiting defense counsel." Beattie, ibid., 349. 50 Beattie, ibid., 409.
and without the aid of counsel. 53 The absence of a right to counsel whilst facing a possible death penalty in criminal cases was contrasted with the right to counsel in civil cases dealing with even small property values. The absence of counsel was sometimes linked to suggestions that their employment would operate to prevent the innocent from being wrongly convicted. 54 But any claim that the presence of counsel was essential to ensure correct decisions on guilt and innocence was, of course, inhibited by the knowledge that few if any of those facing trials for felonies would have been able to employ counsel even if the right were granted. Looking at these trials with modern eyes, it seems incredible that decisions which could lead to a person's execution could be taken in such a short space of time, with little or no deliberation, with verdicts commonly being given at the same time for a number of prisoners after a similar number of trials. That such verdicts were possible is explained by the presence of leading jurors, often foremen, who not only served for numerous trials in one assize, but had done so previously. Through this practice, a jury could have a core of experienced jurors who understood the criminal law, the procedure in which they were involved, and what could be expected of them. They would also be familiar with the narratives and explanations which were regularly extolled by prosecutors and defendants and, if modern research on the effects of such familiarity are applicable to the past, be therefore more likely to convict. All this seems to point to a reckless disregard for miscarriages of justice. On the other hand, in the absence of a professional police force, forensics or the involvement of lawyers, the task which they undertook was less complicated than that faced by a modern jury. Cases requiring difficult evidentiary issues were not going to be presented by victims to the JPs whose role it was to record this evidence and bind victims and witnesses to appear at the assizes. Hence Langbein's view was that the issues which 53 However: "It remained the dominant view that since trials for felonies were essentially confrontations between the victim and the accused, prisoners were at no disadvantage in speaking for themselves", and a view often defended: "If the jurors could watch the accused respond to the evidence as he heard it for the first time and hear his defence as it naturally occurred to him … the truth of his innocence or guilt would be apparent." (William Hawkins, A Treatise of the Pleas of the Crown, ii, 400-402, quoted in Beattie, n 36 above, 264). 54 Beattie, n 29 above, 358 illustrates the strong contemporary criticisms, for example: "… keeping a man in prison before trial, and then denying him the help of a lawyer in court, he concluded was 'downright tying a man's hands behind him, and baiting him to death'." (Quoting Sir John Hawles, Remarks upon the Trials, 1689). needed to be decided by these juries were well within their competence. In addition, as evidenced by the large number of acquittals, juries were not rubber stamping accusations of serious crime. Thus, despite our modern concerns regarding the brevity of these trials for capital offences, contemporaneous views about the likelihood of miscarriages of justice which claimed a priority for wrongful acquittal over wrongful conviction, should not be regarded as mere cant. The classic statement here is that of Blackstone, who opined that: "… the law holds that it is better that ten guilty persons escape than that one innocent suffer." 55 Even discounting for hyperbole (a common feature of Blackstone's Commentaries) a claim that the English trial was organized on the basis of a willingness to sacrifice those known or believed to be innocent, is not made out. 56 The next significant evolution in jury trial was the increased involvement of lawyers. 57 This was preceded by changes in the nature of prosecutions in the late 17th and early 18th centuries. These changes were in many respects the result of a decreasing willingness of the authorities to rely on private prosecutions for the enforcement of the criminal law. 58 Certain crimes, important to the interests of the Government, had no immediate victims, for example 'coining' (the removal of small amounts of precious metals from coins) which threatened to undermine the value of the currency. 59 Another reason for change was the belief that crimes involving gangs of offenders could not be left to the victims to prosecute without assistance. 60 55 W. Blackstone, Commentaries on the Laws of England, Book IV (7th edn., Oxford: Clarendon Press, 1785), Ch 27, 358. 56 Needless to say, a lot has been written about Blackstone's ratio. For a strong modern critique that views it as an impratical guide to the actual operation of the criminal justice system, both historically, and today, see D. 58 "Anxiety about the unwillingness of victims to prosecute was expressed in a variety of ways in this period." Beattie, n 36 above, 317, fn 4. 59 This crime diminished when the mint reissued the coinage at the end of the 17th century. 60 This concern with gangs began in the 1690s: see T. Wales, 'Thief-takers and their clients in later Stuart London' in P. Griffith  Whereas a simple theft would involve a single complainant and a single defendant, with the accused person being removed from the community to await trial once the complaint had been made to the magistrates, crimes such as highway robbery were commonly committed by gangs, which left the victim exposed to retribution once a complaint had been made against an identified gang member. 61 In response to these developments, the authorities began to issue rewards for information leading to the successful prosecution of particular classes of offence. The first of these statutory rewards was for information leading to the prosecution of those committing robbery, for which a reward of £40 was offered. 62 This was at a time when the average wage of a skilled artisan was 10 s (namely half a pound) per week. 63 Periodically, at times of public disquiet, the Government added a further £100 reward for highway robbery. 64 Between 1690 and 1713 rewards were introduced for convictions of highwaymen, burglars, counterfeiters and coin clippers. 65 In addition, in those areas where highway robbery was considered a particular problem, local parishes might offer additional rewards. Alongside the system of statutory rewards, and with a particular focus on the problems represented by criminal gangs, the authorities also offered immunities. Those who had undertaken crimes in association with others could, if accused, escape capital punishment or even all punishment by offering evidence against their associates. In this system, rewards and indemnities or pardons intertwined. Rewards were intended not only to encourage victims to prosecute, but to encourage accomplices to turn King's or Queen's evidence. The system of rewards encouraged 61 A contributor to the Gentleman's Magazine in 1744 wrote of the infamous Black Boy Alley gang, that they "are so insolent that they go to the houses of peace officers, make them beg pardon for endeavouring to do their duty and promise not to molest them." Quoted in R. 63 Though the incentive was diluted by the practice of sharing the reward between all of those involved in the prosecution: informant, prosecutor, witnesses, etc. Paley estimates that the thief-taker might have earned as little as £3 to £5 per conviction (Paley, n 61 above, 322). 64 Paley, ibid., 324. These additional rewards were offered by proclamation, not statute, and were more targeted. 65 For the relevant statutes, intentions behind them and some description, see Beattie, n 36 above, 315-318. the growth of a new profession: thief-takers, 66 who relied on their knowledge of the criminal underworld to identify, for reward, those likely to have committed offences. They also acted as intermediaries between thieves and victims, securing, for a fee, commonly split with the thief, the return of stolen goods. 67 Inevitably, many of these thieftakers were ex-felons themselves. 68 Their activities also extended to using the threat of prosecution to obtain payment from those innocent of crimes, bringing charges for the sole purpose of harassing their enemies, 69 or to create conspiracies to entice vulnerable persons into committing crimes in order to obtain a reward. 70 The dangers presented by the presence of rewards and immunities were obvious to all. The logic which underlay the simple jury trial where a defendant confronted their accuser was that the victim could be expected to act in good faith. No person would ordinarily undertake the expense and inconvenience of reporting crime and appearing at the local assizes unless they had actually suffered from a crime and had good cause to believe they knew the identity of the perpetrator. 71 In addition, where a probing of the complainant's 66 See Beattie, ibid., ch 5 'Detection and Prosecution: Thief-Takers, 1690-1720'. 67 This system of rewards from the authorities operated alongside a system of private rewards offered by the victims of crimes. But these private rewards were primarily aimed not at the prosecution of offenders, but the return of stolen goods. The focus of victims on restitution rather than punishment was evidenced by the abundance of adverts placed with rewards for 'lost' items. See Wales, n 60 above, 69-71. 68 Notorious thief takers Anthony St Leger and Anthony Dunn were former felons (see Beattie, n 36 above, [233][234][235][236][237][238][244][245][246]. "… in general one is looking at men with a clear and unambiguous history of involvement in the world of professional crime." Paley, n 61 above, 304. 69 Paley, ibid., 312. 70 Paley argues that the manner in which rewards were distributed amongst parties, and illegal fees charged for their administration by corrupt officials, led thieftakers to incite crimes in circumstances where rewards were largest, and sharing least. (ibid., 323-324). 71 The significant costs and time involved in bringing a prosecution in the middle of the 18th century are described by D. Hay and F. Snyder in 'Using the Criminal Law: 1750-1850: Policing, Private Prosecution and the State'in D. Hay and F. Synder, n 61 above, ch 1, 25-26. In a later essay in the same volume, Hay argues that the threat of prosecution was used as an exercise of class power, as those with less social standing, and especially those viewed as members of the criminal classes, had little ability to defend themselves against such accusations at trial. So, for example, employee servants who sought to enforce their rights to unpaid wages might find themselves threatened with prosecution for theft in respect of items that they had evidence revealed that they had motives to dissemble, or did not have good grounds for their accusation, these were treated as reasons to acquit. But with statutory rewards and immunities things were very different. 72 Defendants 73 who faced the death penalty unless they could identify associates had every incentive to find such persons. Judges could not simply rule out such evidence without undermining the effectiveness of the statutory scheme, and challenging a practice that had been expressly approved by Parliament. Similar logic applies to the system of rewards. Whatever judges may have thought of evidence offered in the hope of attaining such lucrative rewards, they could not simply reject it. Convictions would thus have to occur in circumstances where there was good reason to suspect the bona fides Footnote 71 continued been allowed to keep in lieu of some part of those wages. Whilst such threats, to be credible, had to be capable of being carried through, the power imbalance Hay describes implies that such threats were rarely necessary. He contends that there is reason to believe that a 'significant' number of prosecutions were malicious, in the sense of being knowingly false, but notes the absence of any general recognition of this as a problem in the 18th century. 'Prosecution and Power: Malicious Prosecution in the English Courts, 1750-1850', ch 8. That said, the assumption that a prosecutor was disinterested would not apply to actions based on 'informations qui tam´where the successful prosecutor was entitled to a part of the fine (usually half). However, these were normally summary offences prosecuted before the magistrates. 72 Note that the claim here is not that prosecutions motivated by malice as well as loss did not occur, nor that wholly false prosecutions motivated entirely by malice were rare, but that the trial could ordinary proceed on the basis of good faith, and the judge and jury could decide whether malice, such as previous quarrels, went so far as to lead to false prosecutions. The offering of significant rewards, coupled with indemnities for those who informed, created a self-consciousness of the likelihood that convictions based on such evidence would be unsound, which had previously been absent. 73 One can make a similar argument in respect of the possibility identified by Hay, n 71 above, that private prosecutions would sometimes be brought by those who knew them to be false as a means of winning a dispute. Until an alternative (public prosecutions) became available, it was difficult to recognise this risk within the legal system without paralysing the system of prosecution. Hay points to the lengths that judges went to, not to recognise officially the possibility of malicious prosecution, by denying the certificates required by criminal defendants who wished to bring civil actions for malicious prosecution. The dependence of trial on private prosecutions makes it difficult for the constant possibility of false prosecution to be acknowledged. Allowing civil remedy in anything other than a case involving the most extreme and obvious evidence of perjury would deter prosecutions, and inviting jurors to consider the possibility of false allegations in the absence of any evidence of its presence would represent a challenge to the system of trial in general, rather than a common trial procedure. of accusers and their witnesses, for the alternative was to secure convictions only in circumstances which had proved sufficient for the prosecution of simple thefts. The introduction of rewards and immunities was premised on the belief that limiting conviction to such circumstances was indeed not sufficient.
The system of rewards produced a dramatic scandal in the middle of the 18th century. 74 The circumstances of a prosecution of persons accused of highway robbery attracted the suspicions of a constable. At the end of a subsequent trial, where the defendants faced the death penalty, the constable announced, after the guilty verdict, that the witnesses to the offence were a professional gang who had persuaded gullible youths to commit highway robbery against other members of the gang in order to prosecute them for rewards. By this point the gang involved had been responsible for the prosecution and execution of six persons. The story, and the dramatic circumstances of its telling, captured the interest of the new press, who interpreted it not as an isolated occurence, but as an indication of the unsatisfactory nature of a system of trial that relied on rewards and indemnities. 75 One can see in this incident an early example of a miscarriage of justice, taken up by the mass media, creating pressure for reform. 76 But one can also see it in terms of resistance to reform until the problem identified by this article -the unacceptability of restricting convictions to circumstances contemporaneously accepted as representing clear proof -had been solved by other means. The system of statutory rewards was maintained for some 50 years. Its removal was made possible by further changes in the organization of prosecution, in particular by the creation of a professional police force charged with the investigation and prosecution of crime. This overcame the need to rely on victims, or incentivize them into prosecuting crime. With regard to the system of indemnities, this has, as we know, never left the system, only varied in its formality. The 18th century reward statutes created a situation akin to a contract. 77 Whilst the certainty 74 As described by Paley throughout her chapter, n 61 above. 75 Paley (ibid., 323) suggests that "the major effect of the provision of £40 rewards was to provide an incentive not to the detection of crime but to the organization of thief-making conspiracies……thief takers were in the business not to detect crime but to commit it". 76 This becomes a pattern, which pattern is illustrated throughout our book: Nobles and Schiff, n 8 above, 1. 77 Following a failure by magistrates to honour one of these agreements, their status was considered by eleven judges in R v Rudd 1 Leach 115,168 Eng. Rep. 160 of the benefits from informing on others may have diminished from this high point, those involved in group crimes could still expect a reduction in their sentence, in addition to that earned by a guilty plea, if they offered evidence against their associates. And such evidence, whilst it can be expected to be probed for its reliability, is not excluded.
During the period when trials continued in reliance on evidence which was known to be unreliable, the system of trial itself adapted. Langbein attributes the increasing involvement of defence counsel in ordinary trials during the late 18th and early 19th centuries to the same concerns that had led to their earlier involvement in treason trials. Following the notoriously biased trials for treason that proceeded the Glorious Revolution of 1688, Parliament had passed the Treason Act of 1695. One of the major protections offered to accused persons by this statute was the right to be represented by counsel, so that treason trials would no longer take the form of the defence being required to answer, in person, the charges made against them. Langbein regards the concerns with the bona fides of prosecutors and their witnesses which resulted from the system of rewards and indemnities, as something that encouraged judges towards a similar development in the trials of ordinary felons. What must have begun as something allowed by the judges ripened into an accepted right, and gradually in turn developed into a general practice. With the increasing involvement of lawyers the trial itself changed into more of an adversarial contest between opposing counsel 78 with the judge, freed of the responsibility for interrogating the prosecution evidence, adopting a more passive role. The involvement of counsel also provided an incentive for more reporting of the exchanges between judges and counsel. Reporters attended who published what might be of interest to lawyers rather than just the general public, allowing a system of precedent to develop, and for the conduct of the trial to be administered by reference to ever more complex understandings of the respective rights of prosecution and defence. 79 Footnote 77 continued (KB, 1775). Their Lordships recognised that such promises had to be honoured or the ability to prosecute gangs of criminals would be severely undermined. The way they achieved this is discussed by Langbein, n 35 above, 91-96.
78 See Langbein, n 11 above, ch 5. 79 In the absence of lawyers there had been no law of evidence as such. Recording the reactions of judges to objections by counsel to the production of evidence led to its formation. For background on the practice of magistrates in response to judge's The procedural changes, which occurred from the end of the 18th century, are commonly expressed as a discourse of rights and a form of trial which is equated with English 'justice'. 80 These include the right to be represented by counsel, to introduce evidence including expert evidence, to maintain silence in the face of police questioning and to avoid giving evidence at trial, and more generally, the claim that English justice, in comparison with Continental forms of trial, is adversarial in nature. 81 The current self-understanding of the modern Anglo-American trial as adversarial in nature, committed to the identification of factual guilt through 'fair' procedures, 82 exists alongside knowledge that Footnote 79 continued rulings, see Beattie, n 29 above, 272-281). By the end of the 18th century the presence of counsel had "ushered into criminal procedure the divisions between examinationsin-chief and cross-examination and between evidence and argument, nourished the growth of the law of evidence, changed the nature of the judicial involvement in the trial, and supplemented the haphazard efforts of prisoners to defend themselves with professional advocacy". (D. Cairns, Advocacy and the Making of the Adversarial Criminal Trial 1800-1865 (Oxford: Clarendon Press, 1998), 3. Relying on Wigmore, Twining notes that "the number of rulings on evidence at Nisi Prius in the period 1790-1815 was more than in all the previous reports of the preceding two centuries." (W. Twining, Theories of Evidence: Bentham and Wigmore (London: Weidenfeld & Nicolson, 1985), 2. 80 See Williams, n 39 above, 10, who describes the pre-18th century English criminal trial as 'un-English'. Much legal scholarship is devoted to the presentation of the normative basis for these forms of procedure, and criticism of common law and statutory changes which are perceived as erosions of rights and/or deteriorations in the capacity of trial to separate the factually guilty from the factually innocent. 81 In this morphogenesis of trial procedure, one reform is worthy of particular attention. This is the right of defence counsel to address the jury, introduced by the Prisoners' Counsel Act 1836 in the face of resistance from the majority of judges and barristers. Those opposed to this reform were committed to the understanding which informed the 18th century trial: that jurors could understand the evidence presented to them without the assistance of lawyers. Addressing the jury would go from assistance to persuasion, and distort the truth finding function. Prior to this, the ability of lawyers to assist defendants had depended on the trial judge's discretion. "What is clear from Hansard and media records is that there was a very real fear that [with the passage of the Act] the felony trial could descend into a theatre for avaricious counsel to actively mislead the court for financial and reputational gain." C. Griffiths, 'The Prisoners' Counsel Act 1836: Doctrine, Advocacy and the Criminal Trial' (2014) Crime and History 28, at 36. 82 "The twin objects of a criminal trial are to accurately determine whether or not a person has committed a particular offence and to do so fairly." A. Ashworth and M. Redmayne, The Criminal Process (4th edn., Oxford: Oxford University Press, 2010), 23. the criminal justice system is hugely reliant on self-incrimination. The most formal expression of self-incrimination is the guilty plea. In the United States around 95% of those found guilty are so found following a guilty plea. In the UK, the figure for guilty pleas in serious cases rose from an average close to 15% less in the previous 50 years to just over 70% in 2011. 83 With this reliance on self-incrimination, any attempt to move from our current situation to one which offered a full trial to all defendants would have huge resource implications, and be politically contentious: "Why go through a long and expensive process. Trials for the obviously guilty [are] a waste of good money and time." 84 But the difficulties of moving from our current condition does not explain how we arrived at this juncture. If we glance back to the 18th century, we have a situation where it seems hard to see how plea bargaining could ever have begun. Defendants could expect little reward for giving up their short trials 85 and, as mentioned, pleading guilty robbed the judge and jury of the opportunity to decide, in light of defendants' character, what penalty was appropriate. There had been periods, particularly at the end of the 16th and beginning of the 17th century, where pressure on particular courts from increases in crime had led to an increase in the willingness to accept guilty pleas, and informal plea bargaining, but this had fallen into disuse, and had not become an institutionalised and dominant practice. 86  85 "[Although] the potential was ever present to deal with caseload pressure by encouraging guilty pleas, so long as jury trial remained rapid and lawyer-free, there was usually no disposition to do it." Langbein, n 11 above, 19 (cited in McConville and Mirsky, ibid., 333). 86 Cockburn finds virtually no use of guilty pleas from 1559 to 1574, but rising to above 14% between 1591 and 1617. Overall, in half of the sixty-six years of his study, between 10 and 30% of convictions transpired without trial. His analysis of the records where this occurred suggests an early form of plea bargaining, with the value of the goods stolen being reduced below that which made the offence capital, in cases where the offender had decided to plead guilty. Cockburn attributes the sudden increased reliance on guilty pleas to the pressures on courts of increased numbers of prisoners, at a time when it was difficult to increase the number of jurors available for trials. (J. Cockburn, Calendar of Assize Records, Home Circuit Indictments, Elizabeth I and James I: Introduction (London, HMSO, 1985), 65-70. to the evolution from jury trials for nearly all serious offences, to jury trials for only a few?
The conventional account of this evolution is internal to the legal system: treating the move to guilty pleas as a knock on effect of the evolution of trial, with the emphasis on complexity. As trials became more adversarial, with greater involvement of lawyers, and professional policing leading to more complex forms of evidence, the incentive on judges and lawyers to negotiate guilty pleas increased. 87 McConville and Mirsky's 2005 book has challenged this explanation, via a longitudinal study of prosecutions and trials in the US. They found that the period when guilty pleas became the norm did not coincide with changes to trial practice and an increased presence of lawyers. Lawyers were present and active from the beginning of the 19th century but the move towards extensive plea bargaining did not begin in the US before 1850. 88 Rather, it coincided with wider changes in social attitudes to crime. The successful prosecution of crime ceased to be a matter of principal concern to the victim, as it was in the 18th century. With the move from aristocratic to democratic forms of government, crime became a political issue, with pressure on government to show that it could successfully prosecute crime. Thus the detection of criminals, and their certification as criminals via conviction, provides not only an incentive to fund policing, but also a pressure on those prosecuting (in the US District Attorneys) to achieve convictions in a high number of cases.
McConville and Mirsky describe this as a move from a system of individual private justice, to one of 'aggregate justice'. 89 In the UK, the move to formal plea bargaining has been much more recent, and 87 See McConville and Misky, n 84 above, 1-9 and works cited therein. This internal evolution is justified normatively by the argument that the introduction of defence lawyers diminished the need for full trials, as these 'experts' could take a correct view of likely outcomes at trial, and advise defendants when conviction would most likely follow if the matter proceeded to trial. The Special Issue of the Law and Society Review in 1979 (Vol. 13, No. 2), with an introductory article by Feeley, 'Perspectives on Plea Bargaining', includes a number of significant studies which offer information relevant to understanding the history and current development of plea bargaining. 88 ibid., see particularly ch 14 'Understanding System Transformation'. 89 ibid., see especially ch 13 'Aggregate Justice and Social Control'. The growth of 'law and order' as a political issue, and the professionalisation of policing and prosecution that followed, coincided with the rise of a free press willing to publish articles both on sensational crimes, and the overall levels of crime and their 'clear up rate' as established via convictions following trial or guilty plea. See P. Schlesinger and H. Tumber, Reporting Crime: The Media Politics of Criminal Justice (Oxford: its introduction criticised. 90 But even in the absence of formal agreements or mandatory reductions in sentences in exchange for guilty pleas, the percentage of cases that proceed to trial has steadily declined. Reduced sentences for guilty pleas became part of the English criminal justice system, even in the absence of these formal agreements, justified (principally in terms of a plea of guilty showing contrition) until fairly recently by the appeal courts in terms that still echo the concerns of the 18th century trial. 91 Until 1985, the functions of US District Attorneys were, in England and Wales, mostly retained by the police themselves, who decided on the charge to bring, and against whom, 92 and were able to convey honestly to any unaware prisoner the likelihood that a confession, followed by a plea of guilty, would lead to a reduced sentence. Whilst the persons responsible for prosecution differed between the US and the UK, as did the formality of plea bargains, there is no reason to suppose that the factors which led to a decrease in the use of trial in the US were not also present in the UK. Professional policing does not simply alter the chances that crime will be detected. It also changes social attitudes towards a broad 'responsibility' for crime: it becomes a government responsibility, not least because the resources devoted to policing have to be justified. To the extent that this is a statistical matter, the major evidence of police efficiency is either the comparative level of crime committed (compared to earlier periods or other locations) 93 or the percentage of crimes 'cleared up' through a successful conviction or its equivalent (caution, agreement to be bound Footnote 89 continued Clarendon Press, 1994), for an introduction to understanding of the political context of contemporary reporting of the criminal trial and criminal justice process. 90 For a full account, see McConville and Marsh, n 83 above. 91 "Counsel must be completely free to do what is his duty, namely to give the accused the best advice … This will often include advice that a plea of Guilty, showing an element of remorse, is a mitigating factor which may well enable the court to give a lesser sentence than would otherwise be the case. Counsel, of course, will emphasise that the accused must not plead Guilty unless he has committed the acts constituting the offence charged." Lord Parker CJ in R v Turner (1970) 54 Cr. App. R. 352, at 360. Recently the justification has focussed more on the saving of costs. See McConville and Marsh, ibid., ch 3 'State-induced guilty pleas and legitimacy'. 92 Particularly important as a consideration where there is evidence to support the charging of other members of the accused's family or friends. 93 There is no agreement on what level of crime is acceptable in a society, nor an ability to claim that any particular level of crime is directly attributable to a given over, pleading guilty to one crime and requesting other previously unsolved crimes be 'taken into consideration', etc.). 94 But the pressure to secure convictions cannot be traced solely, or perhaps even mainly, to statistics. One needs to take account of the role played by the mass media. 95 Without attributing a single function to the policethey have many -a major part of their job, as that task is understood within the mass media, is 'catching criminals', which requires them to secure convictions using resources which include their ability to 'persuade' suspects to confess and plead guilty.
Alongside the growth of the guilty plea, we can see a continued reliance on confessions as evidence of guilt. Within the 18th century pre-trial procedure, the magistrate was expected to question the suspect and present incriminating answers as part of the document that he transmitted to the court. If the accused failed to speak at court, or tried to deny the charge, the record of her/his prior confession would be disclosed to the court. The introduction of legal representation for defendants was by its nature also a right for the defendant not to speak for her/himself, and formed the basis of what we now know as the doctrine against self-incrimination. 96 But pretrial investigations continue to focus on obtaining confessions as the Footnote 93 continued level of policing. So one always needs comparators to support claims that crime has been reduced in response to a given level of policing in particular locations. 94 These arguments inevitably lead to the mirky subject of 'police targets'. The CPS (Crown Prosecution Service) operates under government targets which require it to "narrow the justice gap defined as the difference between the number of crimes recorded by the police and the number of crimes for which an offender is brought to justice". (D. Jeremy, 'The prosecutor's rock and hard place', [2008] Criminal Law Review 925). 95 The separation between statistics and perception was noted in a study of the Press in Cleveland USA as far back as 1919. The press reported a 'crime wave' despite statistical evidence that crime had increased only marginally, resulting in pressure on public officials to sacrifice due process in order to avoid adverse pub- best evidence of a defendant's guilt. 97 Initially, judicial responses to confessions obtained during police custody were conditioned by the limited role played by constables within the pre-20th century jury trial: whereby they arrested accused persons solely in order to ensure that they could be brought to trial: Neither judge, magistrate nor juryman, can interrogate an accused person unless he tenders himself as a witness, or require him to answer questions tending to incriminate himself. Much less, then, ought a Constable to do so, whose duty as regards that person is simply to arrest and detain him in safe custody…. (Lord Brampton's address to the police, 1882). 98 Contrary police practice was ratified in 1912, when judges abandoned their periodic practice of excluding evidence obtained by the police from a defendant held in custody, in favour of a set of rules that set out the conditions required to be met before a confession obtained in custody could be presented as evidence at trial: the Judges' Rules. 99 The justification for accepting confession evidence was given two years prior in R v Booth and Jones: 'If this sort of investigation were not allowed very few crimes would ever be discovered.' 100 The underlying problems of a reliance on self-incrimination are obvious. Guilty pleas do not contribute to the identification of the factually guilty. Rather, they select, from amongst the pool of possibly guilty persons, those who are risk adverse or for whom the stigma of conviction is less important, and rewards them for the 97 See J. Baldwin and M. Mcconville, Confessions in Crown Court Trials (Royal Commission on Criminal Procedure Research Study No. 5, London: HMSO, 1980): their model study (as relevant today, despite many statutory and other changes, as then) found that confessions were crucial evidence in around 30 per cent of the cases that proceeded to jury trial. However, in the majority of cases where a confession is obtained, the defendant will subsequently plead guilty. The standard of 'best evidence' is not independent of the criminal justice system. For those charged with the investigation of crime the 'best evidence', in serious crimes, is that most likely to convince the jury of the defendant's guilt. Confessions have, and continue to have, a status at court that resist psychological understanding of reasons why people make false confessions, on which there is considerable research data. For the classic account, see G. Gudjonsson, The psychology of interrogation and confessions: A handbook (Chichester: Wiley, 2003). 98  provided that a failure to impart evidence on arrest that might assist in a suspect's defence could itself be treated as evidence of guilt. The English system of plea bargaining has also become more formal and certain, with sentence guidelines introduced that require a judge to give reasons for any failure to recommend a stipulated discount as a reward for pleading guilty. It is probably better to interpret McConville and Misky's US study not as a single explanation for the decline in the percentage of suspects that experience full trial, but evidence of a process of multiple causation and co-evolution that has occurred in both the US and 101 In the US there is evidence that prosecutors compensate for the weaknesses of their cases by increasing the number and seriousness of the charges brought against an accused, and then increasing the discounts offered in return for pleading guilty. In a weak case, the defendant faces the certainty of a small sentence in return for a guilty plea, or the small possibility of an extremely harsh sentence if she/he goes to trial and a jury convicts. For a discussion of this practice in the context of prosecutions for murder, by shaking of small infants, see D. Tuerkheimer, Flawed Convictions: 'Shaken Baby Syndrome' and the Inertia of Injustice (New York: Oxford University Press, 2014), ch 8. 102 The oppressiveness of police custody and the consequent involuntary nature of confessions obtained in such custody was acknowledged by the Royal Commission on Criminal Procedure in their 1981 Report, and formed part of their reasoning for recommending the abolition of the common law requirement that confessions be 'voluntary'. See their Report, Cmnd. 8092, 1981, paras 3.116, 4.68 and 4.74 UK. One moved from the 18th century situation of full trials and few guilty pleas to our present one, with its reliance on self-incrimination, through opportunities generated via changes in both trial and policing practices. Professional policing created the incentive for expedient convictions without trial, and the judiary could not but become aware of the importance to the police of confessions and guilty pleas in executing their crime solving role; and the increased complexity of the adversarial trial has led to a situation in which discounts for unretracted confessions and guilty pleas are not only seen to be necessary if the police are to carry out their functions, but also necessary if the courts are not to be overwhelmed by the burden of offering a full trial to every suspect. These developments, described by McConville and Misky as 'aggregate justice' are also captured in Packer's famous two models for the analysis of contempory criminal procedures -the crime control and due process models. 104 The first is managerial and administrative, and depends on a willingness to accept that the police, with sufficient accuracy to maintain the deterrent effect of the criminal law, can identify those who are factually guilty of crimes. The due process model requires a higher level of accuracy: the elimination of all avoidable errors (miscarriages of justice?). These models allow one to describe and assess different aspects of contemporary criminal procedure in terms of a willingness to give effect to police judgments that defendants are guilty, rather than submit those judgments to independent scrutiny. Since not all aspects of due process can be understood as contributions to the issue of factual guilt, the due process model can only account for criminal procedure in terms of a commitment beyond 'factual guilt' to other values as well, such as fairness, and the need to restrain illegal police actions. 105 But as Packer acknowledges, when introducing these models, the crime control and due process models are not independent of each other. There is no 'pure' crime control model, in which the test of a person's factual guilt is simply left to the police, who punish those that they decide are guilty of crimes. Packer's crime control model assumes the existence of criminal procedures -legal procedures which exist separately from policing. (Like tribal rulers, the police need something outside their own coercive powers to establish the 'truth' of the guilt of the accused). 106 The issue which Packer's models seek to identify, in their application to developments in doctrine, legislative provisions, media coverage, and public opinion, is how much process is 'due', at a minimum, for a police willingness to consider a person guilty to morph into an individual's conviction.
As Packer observed, the pressures placed upon our standards of due process depend crucially on what is made criminal, and how much criminal conduct we wish to prosecute. Packer's own concern was with the scope and complexity of the criminal law. The more conduct that is criminalized, and the more complex the proofs required to establish criminal conduct, the greater the resources that must, ceteris parabus, be devoted to the criminal justice system. 107 With respect to any particular defendant, the extent to which their experience approximates to either of these two models depends crucially on the role played by rights, and the agency of the defendant in the exercise of those rights: Finally, there is a complex of assumptions embraced within terms like 'the adversary system,' 'procedural due process,' 'notice and an opportunity to be heard,' 'day in court,' and the like. Common to them all is the notion that the alleged criminal is not merely an object to be acted upon, but an independent entity in the process who may, if he so desires, force the operators of the process to demonstrate to an independent authority (judge and jury) that he is guilty of the charges against him. It is a minimal assumption. It speaks in terms of 'may', not 'must'. It permits but does not require the accused, acting by himself or through his own agent, to play an active role in the process; by virtue of that fact, the process becomes or has the capacity to become a contest between, if not equals, at least independent actors. Now, as we shall see, much of the space between the two models is occupied by stronger or weaker notions of how this contest is to be arranged, how often it is to be played, and by what rules. 108 106 This view of the role played by due process challenges the common presentation of Packer's due process model as a restraint, or 'obstacle course' placed in the way of a system that instrumentally aims at crime control. Due process is necessary to give police investigations a routine and technical way of demonstrating that they have 'cleared up' crime. The question is how much due process is required, on what occasions, for the police to continue to communicate these achievements. 107 There is another dimension to this pressure. One can criminalize conduct whose commission can rarely be proved with any level of certainty -some examples from the difficulties with prosecutions for the offence of corporate manslaughter illustrate this, and continue to do so even since the enactment of the Corporate Manslaughter and Corporate Homicide Act 2007. 108 Packer, n 104 above, 8-9.
Some of this allocation takes the form of entitlements which are not matters of choice -so, for example, the restrictions on the 'right' to jury trial leave defendants with only a 'right' to summary trial. 109 This operates as a significant restriction on the resources that can be absorbed through criminal trials. But, as Packer stressed in the above passage, a large part of the rationing of resources within the Anglo-American system operates through affording defendants the right to make choices (or have them made on their behalves by lawyers): to plead guilty or not, to forgo jury trial and opt for trial before magistrates 110 ; to conduct their own defences with, if they choose, the assistance of counsel; to call their own witnesses, including expert witnesses. Rights construct the procedures that currently separate the guilty from the innocent whilst at the same time operating as switches that control the resources absorbed by criminal procedures. In the absence of the move to guilty pleas, a gradual increase in the complexity of trials and consequent delays in the processing of defendants might have led to other changes, such as increased resources (e.g. increased numbers of courts and sittings) or some attempt to decrease complexity. From this perspective, the enhanced use of the guilty plea is a functional equivalent of these other two possibilities. But it has a radical potential which the alternatives to it do not. It allows trial to become ever more complicated, without taking up more resources, provided that the percentage of guilty pleas increase and the number of persons taking up their full panoply of rights diminishes. The same logic applies to Packer's concern with the increased resources required, ceteris parabus, if more activities are criminalised. Increasing the guilty plea percentage lowers the cost of this. As already described, jury trial has altered from a process afforded to all persons accused of serious crimes, to a right that all persons accused of serious crimes may insist upon, within a system that could 109 F. Belloni and J. Hodgson, Criminal Injustice: An Evaluation of the Criminal Justice Process in Britain (Basingstoke: Macmillan, 2000, 6) argue that the routine miscarriages of justice that occur in the magistrates courts "attracts little media, and therefore public, attention because of the widespread perception that cases dealt with by magistrates involve only trivial offences." As Ashworth comments in introducing his 1998 book: '… discussion devoted to criminal trials, particularly trials by jury, far outweighs their numerical significance within the criminal process.' A. Ashworth, The Criminal Process: An Evaluative Study (2nd edn., Oxford: Oxford University Press, 1998), 3. 110 For choice of trial, the presence of the switch may be more important than its allocation to the defendant as a right; in Scotland it is the prosecutor who chooses whether trial by jury or magistrates is appropriate in cases triable either way. not continue in its present form, without a huge increase in resources, if this right was not forgone by a substantial majority of such accused persons. Decisions on this right, normally aided by counsel, take place in the shadow of jury trial. They require an assessment of the pros and cons of proceeding to trial and losing the advantages of a guilty plea which depends, alongside the defendant's attitude to risk, on an assessment of the likelihood of a guilty verdict. This means that the expectations of what a jury will do play a pivotal role in the current system. And, as we shall show shortly, the jury also plays a pivotal role post-conviction in terms of what the current system is capable of recognising as a miscarriage of justice. This increased dependence on the jury throws into question attempts to describe our current system as more rational, accurate, or fair than those which have gone before -and points us once again towards the suggestion that our current forms of criminal procedure are the present solution to a ongoing functional need.

IV THE MODERN JURY AND MISCARRIAGES
Let us consider what we ask of modern juries. As already pointed out, in the 21st century the jury can commonly be invited to consider a much more complex series of facts than its predecessor. The simplicity of the evidence put before an 18th century jury reflected the rudimentary nature of the investigations that preceded the prosecution of most suspects. 111 But professional policing brought increased possibilities for the investigation of crime and with it the need for juries to assess more complex combinations of different kinds of evidence, in order to decide on the innocence or guilt of the accused. Not only are there likely to be more witnesses, which involves more assessments of recall and veracity, but juries can be asked to form assessments of forensic or other expert evidence, either directly, or by forming a judgement on the weight to be given to the testimony of experts. This involves a trust in the fact-finding capacity of jury members (with fuller and more carefully crafted instructions from judges) which goes beyond anything corresponding to this in the 18th century. Trial has altered in response to the increased burdens placed upon juries, creating new combinations of rational and irrational elements. For example, the complexity of evidence justifies the need for the judge, when summing up, to take the jury through the evidence they have heard, pointing to ways in which they might reach an opinion in reliance on it, in light of the applicable law. As a result, within the system of appeals, the judge's summing up speech has become like a sacred text, examined on the basis that the jury would have heard and reacted to every word of it. These practices continue in the face of evidence of the limited ability of individuals to recall or learn from an oral delivery of complex instructions. 112 The increased reliance on various forms of 'scientific' evidence for the investigation of crime has created new challenges for the jury. The ability of experts to give their opinion serves to shield jurors from some of the complexities of understanding expert evidence, but even here, there is still the issue of what weight to give scientific evidence such as DNA, fingerprints, or some other more controversial types of expert evidence such as that which underpins 'shaken baby syndrome', 113 in the context of the rest of the evidence. Where experts are in dispute, jurors are either required to receive evidence on the methods and calculations by which the experts reached their conclusions, or to decide between experts on the patently irrational basis of whether their performance as witnesses inspires belief in their opinions. The question of how juries can be made to understand scientific evidence has also focussed attention on how juries assess other kinds of evidence. In particular, if scientific evidence is to be understood in terms of how it alters the probability that the prosecution case is true (as advocated by those who take a Bayesian approach to proof) then why shouldn't the jury be facilitated and required to assess every piece of evidence, as it is admitted in the course of the trial, using this methodology? In the UK the appeal courts have rejected this approach on the basis of the jury's inability to understand and apply such a methodology, which invites critics to find ways in which juries could apply Bayesian methods without needing to understand the underlying mathematics (applying the metaphor of the manner in which lay persons use calculators). 114 The idea that some 'scientific' method for understanding evidence should be extended to the whole trial assumes that it would be appropriate to give 'reasonable doubt' or 'being sure' a mathematical meaning -x% which, if it were to occur, would immediately prompt questions such as: what is the correct percentage, is the same percentage required from different juries/jurors in respect of the same crime, can different levels of probability be applied in respect of different crimes, or does what is acceptable differ from context to context and juror to juror? 115 Despite the difficulties of generalisation reflected in unsystematic piecemeal reform, it is probably fair to say that the recent evolution of the law of evidence as a whole can be understood as an ever increasing faith in the ability of juries to deal with the evidence presented to them, with exclusion rules repeatedly removed in favour of allowing juries to hear more evidence, and form an assessment, accompanied where necessary by warnings from the judge of possible dangers of unreliability. So, for example, the exclusion of involuntary confessions given in police custody was dropped in favour of allowing juries to rely on these so long as they 'could be reliable'. 116 The current rules acknowledge the possibility of false confessions made under these conditions, but assume that a jury is competent to decide which of the confessions made under conditions of oppression or hope of reward are genuine. This increased faith in juries also has implications for the numbers facing trial -as the admissibility of confessions, and the expectation that jurors will treat a confession as strong evidence of guilt, creates pressure on those who have confessed to then plead guilty at trial.
The progressive removal of rules of exclusion in favour of allowing evidence to be heard and assessed, subject to warnings and instructions where necessary, forms part of what has been called the 'rationalist tradition' in evidence scholarship. The assumption which underlies this approach is that a 'fact-finder' cannot be assisted in reaching a decision by the exclusion of any evidence that could be relevant to deciding what has occurred: the 'best-evidence principle'. 117 This approach challenges the exclusion of any evidence from the trial, unless one can argue that a jury is patently ill-equipped to give such evidence proper weight. But any claims regarding the limitations of a jury's competence have to operate within the context of a legal system which relies on, and celebrates, the ability of a jury to identify guilt in all manner of complex situations. Such reliance and celebration also has to be maintained in the face of knowledge that some crimes cannot be prosecuted and brought to trial, if this limitation on the jury's competence is acknowledged, and no alternative system of trial is provided. 118 This last difficulty, which goes to the heart of the functional need which jury verdicts are used to solve, can be illustrated by considering recent developments on the law surrounding historical child sexual abuse. There is no statute of limitations in the UK for these offences, which means that some allegations are made decades after the claimed offence by adults, against those in charge of their care when they were children -most typically parents, teachers, care workers, football coaches and others with some authority. The lapse of time commonly makes it impossible to establish the precise day and time of the claimed offences, making alibis impossible. To show its support for the victims of child sexual abuse, and to allow prosecutions that would not otherwise be possible, Parliament removed the require-ment that all allegations of sexual assault required the victim's testimony to be supported by some evidence which would corroborate it, thus implicitly authorising prosecution and conviction on the basis of victim testimony alone. 119 This has left English judges in a situation which has echoes of the 18th century experience of statutory rewards for evidence leading to convictions. Parliament has endorsed trials in circumstances which leaves prosecuting authorities and judges uneasy as to the ability of their procedures to produce a verdict that can convince.
This sense of unease was demonstrated most clearly in the case of R v B. In this case the alleging victim was a step-daughter of the accused, who claimed numerous sexual assaults had been committed on her between ages 7 and 11, almost 30 years prior to her formal complaint to the police. Except for claims of abuse made to her psychiatrists, she had taken no previous action, claiming that she did not feel able to cause distress to her mother, and had waited until after her mother's death to make formal allegations. The Court of Appeal had no criticism to make of the conduct of either the victim or the trial, including the trial judge's decision that a fair trial was possible in circumstances where it boiled down to which of two witnesses to believe. The Court of Appeal accepted that the jury had been correctly warned of the dangers of convicting on the basis of 119 Prior to the Criminal Justice and Public Order Act 1994, the Sexual Offences Act 1956 had provided that persons should not be convicted of various sexual offences on the evidence of one witness alone unless that witness testimony is corroborated. Prior to the 1956 Act, juries were instructed that they could convict without corroboration but that they should know that it was very dangerous for them to do so. It was a requirement of the common law that juries should have the corroboration warning. However, the 1994 Act changed the position. Section 33 provides: "(1) The following provisions of the Sexual Offences Act 1956 (which provide that a person shall not be convicted of the offence concerned on the evidence of one witness only unless the witness is corroborated) are hereby repealed …." There is now abundant cogent evidence that many serious child sexual abuse crimes have been unpunished, both in the distant and recent past, which evidence is universally accepted, but there is also some convincing evidence that juries are ill equipped to judge sexual abuse cases impartially (see, for example, N. Vidmar, 'Generic Prejudice and the Presumption of Guilt in Sex Abuse Trials' (1997) 21 Law & Human Behaviour 5. On the evidential complexities involved in prosecuting sexual abuse cases, see the CPS guidelines: http://www.cps.gov.uk/legal/a_to_c/ child_sexual_abuse/#content. These complexities are added to considerably in relation to historic child sexual abuse (some of which are noted at paras. 105-107). This example well demonstrates, in relation to the substantial 'public concerns' and desire for 'retribution', the apparent need to achieve convictions despite absence of the clearest proof. uncorroborated evidence. The Court nevertheless claimed that it had a 'residual' discretion to set aside a conviction on the basis that it was unsafe or unfair, even where the trial process itself could not be faulted. Lord Woolf's judgment sets out the difficulties faced by the court: One thing is clear: the jury saw the witnesses and we have not. Therefore they were in a better position to judge where the truth lay than this Court. Furthermore, the trial process depends upon our confidence in the jury system. We have to have confidence that they made the appropriate allowance here for delay, and we also have to have in mind the intervention of Parliament. Parliament made the decision as to where they considered the right balance between the prosecution and the defence should lie in regard to the question of corroboration. We must not seek to go behind the decision of Parliament. Therefore juries in cases of this sort must be left with the difficult task of determining where the truth lies. (our emphasis) Having asserted that the decision here -who to believe -must be left to the jury, he continues: However, there remains in this Court a residual discretion to set aside a conviction if we feel it is unsafe or unfair to allow it to stand. This is so even where the trial process itself cannot be faulted. It is a discretion which must be exercised in limited circumstances and with caution. 120 The Court could not declare that uncorroborated victim testimony is never sufficient for prosecution without disregarding a statutory provision. It could not announce a period after which such prosecutions should not be brought without creating a separate limitation period for sexual offences. It could not criticise the trial judge for failing to stay the proceedings as an abuse of process without articulating the basis of her/his error, and thereby changing the standard of what constituted a 'faultless' trial. The Court could not throw doubt on a jury's ability to decide which of two opposing witnesses to believe as this is a quite normal basis for conviction and the reason Parliament regarded the need for corroboration in sexual assault cases to be an anomaly. They could not proceed on the ground that the jury failed to heed the judge's warning about the difficulties of convicting on the basis of uncorroborated evidence, as the assumption that juries follow a judge's directions is central to their normal appeal process, which involves the scrutiny of these directions. Their eventual decision, to overturn the conviction, was justified in terms of 120 R v B [2003] EWCA Crim 319 at paras 26 and 27. the commitment of criminal justice to avoid wrongfully convicting the innocent, but without identifying the factors which would require another Court of Appeal to reach the same conclusion. 121 The dependence of the current system on the jury is at its most transparent when one considers the basis for appeals. 122 Showing deference to the jury allows the Court of Appeal to resist appeals based solely on the grounds that the jury could have reached a different verdict. 123 This is justified by judges in terms of constitutional requirements, the respect to be given to a judgment by 'peers' or laymen over that of professional judges, or even that jury trial represents an embodiment of justice 124 -all claims that appear quite fanciful in jurisdictions that operate comfortably without a 'right' to jury trial, 125 or even without the jury at all (as happened in Northern 121 Indeed their decision could only be presented as an exercise of discretion on the assumption that another Court of Appeal, facing an indistinguishable case, could fail to quash the conviction. Later Courts of Appeal have taken a different approach, focussing on the effects of delay on the ability of defendants to present their defences to specific allegations made against them. By being specific as to dates and locations, the victim increases the ability of the defence to claim that absences of evidence resulting from delay (such as missing witnesses and documents able to prove an alibi defence) means that the delay has removed opportunities to demonstrate innocence objectively. This creates a situation whereby the more specific the allegations of alleged victims, the more likely it is that delay will support the proceedings being halted as an abuse of process. But whilst this approach may weed out many prosecutions, it leaves a case on all fours with R v B (where no dates were stipulated and the abuse took place in the family home without witnesses) as one where the only defence left to the accused is to deny the allegations. They cannot here complain that the delay makes any difference to their ability to prove their innocence. If found guilty, their only hope is that another Court of Appeal will follow R v B and reject the jury's verdict. 122 Appeals are the operations which identify what the system regards as an error. Within the functionalist perspective we are adopting in this article, they also invite us to consider what the system is able to regard as an error without undoing the current solution to the problem of convicting those for whom one has no certainty as to their guilt. 123 The reader can, if they wish, follow our more extended analysis of this issue in R. Nobles and D. Schiff, 'The Right to Appeal and Workable Systems of Justice ' (2002)  Ireland for terrorist offences during 'the troubles' 126 ). There is often a circularity in these statements, as when the appeal court asserts that it is less well equipped to decide on the veracity of witnesses that it has not heard, whilst at the same time refusing to hear them and limiting itself to reviews of the evidence rather than rehearing. But one can also consider these statements in terms of the evolutionary alternatives. Within this system, announcing a willingness to redecide any case where a jury could have reached a different verdict on the evidence would produce a huge increase in the number of appeals. And were these appeals to result in a large number of quashed convictions, justified solely on the basis that the judges took a different view of the evidence than the juries, then we would be back to the fork in the road that occurred in 1215 -where the decision to convict in the absence of overwhelming evidence of guilt led not to trial by judges, but to trial by juries. In many respects, the situation is more difficult than in 1215. For then, one had a situation of crisis -without some innovation in procedure there could be no processing of the accusations of criminal acts. Whilst the judgment of a judiciary was not the selected solution, either in the UK or on the Continent, one can see that decisions by judges might well have been the chosen solution if the alternative was that no cases with less than absolute proof could proceed. But in the 21st century, opening all jury convictions to a full reappraisal by judges would not only represent the undoing of one solution to this proble´matique and the substitution of another. It would also create a situation in which every appealed conviction could generate an immediate comparison of the relative merits of the two solutions. One already has some sense (from examples of media reporting) of the hostility of the American or British media to a system of appeal from jury trial by way of review when one considers their reporting on appeals where persons found guilty by a jury have their convictions quashed on the basis of errors of procedure. 127 Now, imagine a newsworthy case (perhaps a particularly brutal murder, or rape) and then consider the likely reporting of an appeal by review if, following a conviction in which there had been no breach of procedure, three judges decide that a person whom the jury 126 See J. Jackson and S. Doran, Judge Without Jury: Diplock Trials in the Adversary System (Oxford: Clarendon Press, 1995). 127 Such hostile reporting will often ignore or simplify the appeal process which, in practice, requires the Court of Appeal to consider not merely the procedural error, but whether the procedural error might, or might not, be relevant to the appellant's factual guilt (or innocence). found guilty should be set free on the basis that they are not as satisfied, as the jury had been, that the appellant was guilty. In addition, another likely reason for the judiciary's reluctance to engage in a full review of jury verdicts is their awareness of the implications of this for their workload, or to put this in terms of justice, the delays caused to persons who appeal against their convictions in cases where there has been some substantial mishap or error in the conduct of their trial. 128 The judiciary might well expect large numbers of prisoners facing long prison sentences to seek a full reappraisal of their convictions. So, as well as a willingness to face strong media criticism, embarking down this route could be expected to require either a huge increase in the size of the criminal judiciary, or new measures to deter appeals, both of which would generate their own new criticisms. 129 The deference shown to jury verdicts does not result in a situation in which no jury verdicts are ever quashed on the basis that the judiciary reach a different view on the facts -as shown by the interpretation of the normal ground for an appeal, that the conviction is 'unsafe', in R v B. As 128 In the UK, the 19th century saw a large number of attempts to introduce Bills in Parliament which would require the judiciary to consider appeals against jury verdicts on the basis that the jury had reached a wrong conclusion on the evidence, all of which were robustly opposed by the judiciary. See R. Pattenden, English Criminal Appeals 1844-1994(Oxford: Clarendon Press, 1996, 6-27; Nobles and Schiff, n 8 above, 45-47. The judges claims that juries rarely made mistakes was taken then, as now, as evidence of their complacency towards miscarriages of justice. But their claim that they would do no better, and would commonly be understood to do worse, is less easily dismissed. Consider the following statement by Mister Justice Bingham in 1906: 'It [the verdict of the jury] gave to the people a sense of security; and it placed the judgment of the Court above criticism, because it based them on the assent of the public …. This new Bill proposed to violate it [the principle of jury verdict] by an appeal to a Court of three lawyers whose decision … was to have the finality which at present attached to the verdict of a jury only … It was, of course, conceivable that a jury might make a mistake; but such a possibility must not be allowed to paralyse justice or to delay the execution of its decrees. The faintest possibility of the improbable mistake of a jury being corrected by a court of two or three lawyers offered … no compensation for the mischief which the contemplated Court of Appeal might introduce. Would the public have confidence in the decisions of such a tribunal?' (Quoted in Nobles and Schiff, n 8 above, 47). 129 For example, by reinforcing the penalty used to curb applications for leave to appeal or appeals, where some or all of the time spent in prison waiting for the appeal to be decided does not to count towards the sentence served (loss of time directions -Criminal Procedure Rules 2011, The Consolidated Criminal Practice Direction, Part II.16). The criticism is that, ceteris parabus, this deters the innocent but risk averse individual more than the risk prone but guilty one.
with R v B, the assertion of this possibility (which is sometimes called the 'lurking doubt doctrine') 130 operates retrospectively, as a means by which the Court of Appeal can decide to quash a conviction, without operating prospectively, to provide a substantive ground for an appeal. To get to the Court, prisoners have to show either that their trial involved a procedural error, or that there is new evidence, and that one or both indicate that their convictions are not 'safe'. Here again, the need to preserve jury trial as the solution to the functional need (sometimes articulated in terms of the need for finality) leads to further restrictions on what can lead to a conviction being quashed. The jury provides no reasons for its decisions, which leads to appeals which focus on the decisions made by the judge at trial, especially her/his summing up to the jury. Having no way of knowing how the jury did reach its decision, leads to a system of appeals organised around how the jury could justifiably have reached its decision. This has the effect of severely restricting the relevance of procedural errors or new evidence. The jury are treated as a fact finding body, who are assumed to have followed the judge's instructions, and to have found as true all of the facts identified by that summing up as necessary for the conviction. Adopting this perspective, the appeal court judges effectively ask themselves if the new evidence, or the breach of procedure, could have made a difference to the verdict of this jury. Whilst the standard of 'possible difference' suggests a liberal approach to appeals, given its semantic distance from the trial standard of proof beyond reasonable doubt, adopting what normally amounts to the prosecution case as fact, leads to a situation that can reasonably be described as a 'reverse burden of proof.' 131 Further restrictions on appeals arise from the Court's approach to new evidence. 132 There will often be some evidence or arguments that a jury did not hear that, with hindsight, might, if heard, have led to a 130 A doctrine that many commentators and practitioners have questioned or, at least, questioned in terms of its continuing significance. The usual suggestion is that 'lurking doubt' implies other specific grounds of appeal, and it would be more acceptable, especially from the point of view of articulating clear precedents, to identify what those specific grounds are. See, for example, the practitioner text by P. Taylor, Taylor on Criminal Appeals (2nd edn., Oxford: Oxford University Press, 2012) 119-121. 131 Although this can be shown to apply more generally, we have previously specifically described how and why this may operate in relation to the reassessment of scientific evidence on appeal: Nobles and Schiff, n 8 above, 199-215. 132 Technically, this is now governed by section 23 of the Criminal Appeal Act 1968 (as amended by the Criminal Appeal Act 1995). For analysis, see Taylor on Appeals, n 130 above, 174-191. different decision. This is not least because both prosecution and defence commonly seek to offer alternative narratives, and choose not to introduce evidence that weakens the chosen narrative. Allowing appeals based on these excluded alternatives, even if it were to be examined on the basis that the prosecution case had been accepted as factually correct, would open the Court to an enormous increase in the number of appeals. These are primarily screened out by attributing agency to defendants. Their lawyers' choices are attributed to them. Having made these choices, they cannot now ask for a consideration of arguments and evidence that could, but were not, made at trial. Of course this focusses attention on the standard of the advice and representation enjoyed by defendants. If the agency attributed to the defendant were reliant on the very highest standards of professional competence and the resources that facilitate such competence, then the reduction in appeals achieved by the Court's interpretation of new evidence would be much less. Instead, we find a reluctance to hear arguments or evidence that were available but not made at trial unless the legal assistance given could be shown to be an example of 'flagrant incompetence': This is an unpromising ground of appeal, the court is extremely reluctant to accede, and in the interests of the appellant D it must be hoped that there are other more promising grounds of appeal. It is not enough that counsel was inept, he made a mistake, he was under a misapprehension, he made a tactical error, his judgment was flawed, he was unwise, another counsel might have conducted the case differently, or better. 133 The practice of restricting what constitutes new evidence, refusing to undertake a rehearing of trial evidence, and the commitment shown to the presumed construction placed upon the evidence by the particular jury have been subjected to sustained criticism in the media, most notably in the aftermath of the drawn out failure in the UK of the Court of Appeal to quash the convictions of high profile prisoners convicted of terrorist related offences in the early 1970s: the Birmingham 6, Guildford 4, Maguire 7 and Judith Ward cases. 134 The deference shown to the jury by appeal court judges was also noted and criticised in the Report of the Runciman Royal Commission on 133 A. Samuels, 'The incompetence of counsel as a ground of appeal' Criminal Lawyer (2013) 3.
134 See our analysis of this media reporting: Nobles and Schiff, n 8 above, chapter 4.
Criminal Justice set up after the successful third appeal of the Birmingham 6: In its approach to the consideration of appeals against conviction, the Court of Appeal seems to us to have been too heavily influenced by the role of the jury in Crown Court trials. Ever since 1907, commentators have detected a reluctance on the part of the Court of Appeal to consider whether a jury has reached a wrong decision … the court should be more willing to consider arguments that indicate that a jury might have made a mistake … [and] more prepared, where appropriate, to admit evidence that might favour the defendant's case even if it was, or could have been, available at the trial. 135 In presenting these observations on the level of deference shown to the jury, we are not asserting that appeal court judges never lower their standard of respect for jury verdicts. Indeed, the example of R v B shows the contrary. Rather, one has to consider the possibilities for adjusting deference in terms of the role played by the jury in establishing the truth of what cannot be known with certainty. There was no suggestion, at least from the Runciman Royal Commission, that the appeal courts should abandon deference towards jury verdicts, and open themselves to the possibility of having to re-examine every guilty verdict for possible error based on their own judgment of the facts. 136 Nor any suggestion that they should convert a system of appeal by review into one involving a full rehearing, which might give more plausibility to their decisions to quash jury verdicts. The assumption underlying the Royal Commission's criticism is that the jury is still the body entrusted with solving this problem (namely conviction without certainty) but that, nevertheless, on more frequent occasions than have occurred previously, the Court of Appeal must be willing to take on this task itself. But it is unclear how the Court 135 Royal Commission on Criminal Justice Report (1993) Cm 2263, ch. 10, para.3. For a discussion of the persistence of this deference in the face of changing statutory formulations of the appeal provisions, see D. Schiff and R. Nobles, 'Criminal Appeal Act 1995: The Semantics of Jurisdiction ' (1996) 59 The Modern Law Review 573, and on the necessity for deference, Nobles and Schiff, n 123 above. 136 In a case like R v B, where there is neither new evidence nor an error of procedure, the general position has to be that outlined by Lord Woolf -it has to be accepted, amongst all judges in the Court of Appeal, and all those who direct cases up to them (judges giving leave and counsel advising on appeals, and even the Criminal Cases Review Commission in making a post-appeal reference) that there is a considerable unwillingness to overturn the jury's verdict. That is why R v B is both an example of the fact that this could occur, and a general reminder to judges and counsel that this cannot be anything other than a quite exceptional reason for an appeal succeeding.
can do this whilst still treating the jury as a body which can establish the truth of whether a person is guilty. If the jury is regarded as having found the facts, by accepting the evidence necessary for there to be a conviction, this will always restrict the possibilities for a successful appeal. As such, the instruction to show 'less deference' is a call for the appeal judges to operate somewhere between these positions, without any clear sense of what this might entail, how it could be operated consistently, nor how the task is to be communicated between the various judges who staff the Court of Appeal, both now and over time, or the judges who award leave to appeal, or the Criminal Cases Review Commission in their decisions to refer cases in anticipation of what may amount to a successful appeal. 137 There can, of course, be adjustments of doctrine which will alter what can result in an appeal. The standard of professional incompetence can be changed, and new evidence can be defined with reference to the practical difficulties of obtaining evidence in time for trial, rather than whether it was available at time of trial and, in theory, could have been obtained and called. And such changes will lead to increased numbers of appeals. But if these appeals are not to undo the current solution to the functional need that we have outlined throughout this article, and begin an evolution towards a new one, then they must continue to be decided in terms of an acceptance that the jury are a fact finding body, whose finding of facts must form the basis of any reassessment of the conviction. And, with this approach, accusations of 'excessive' deference from the academic and media critics alluded to by the Royal Commission, are likely to continue.
Lastly, in this brief review of modern UK practice, we need to consider how the appeal process relates to one of the central features of the modern procedure for conviction -the guilty plea. This practice was justified in the Courts, even as late as the 1990s, on the basis that those who show contrition for their crimes, by admitting them, should be rewarded by receiving a lower sentence than those who do not. This rationale, however unconvincing, justifies the reliance on the guilty plea on the basis that it represents evidence (via 137 The Criminal Cases Review Commission, set up by the Criminal Appeal Act 1995, may refer cases to the Court of Appeal where there is a 'real possibility' of a successful appeal (s.13.1.a). We explore a number of the complications of the relationship between these different bodies involved with considering miscarriages of justice in the UK, in R. Nobles  self-expression) of guilt. The more recent changes to the English criminal justice system adopt a self-consciously managerial approach to the practice -justifying the reduction of sentences by reference to the resources being saved by those who do not insist on their right to jury trial and plead guilty early, with sentences normally adjusted upwards by reference to a tariff the later the guilty plea occurs within the process. 138 Rewarding defendants for forgoing jury trial quite obviously puts pressure on those who cannot afford trial in terms of the defence costs, or who prioritise sentence reduction over the benefits of acquittal. And acknowledging that the justification for this is not contrition, but simply resources saved, undermines any claim that a guilty plea, like a confession, represents evidence of guilt.
In light of the obvious dangers of incentivising defendants to plead guilty, how ready is the English legal system to allow appeals from prisoners who decide later to assert their innocence? As with the jury verdict, the evidence and advice available to the defendant could always have been different. Evidence which could have been disclosed may not have been, and evidence not available may subsequently become available. Counsel or solicitors may have taken greater care to interrogate the prosecution case, or have exercised better judgment. 139 But allowing these factors to undo a conviction has, within a system with its considerable reliance on the guilty plea, as much or even more capacity to undo the solution to the functional need described throughout this article than a greater openness to quashing 138 S.114 Criminal Justice Act 2003. The sentencing guidelines make it clear that the strength of a defendant's case (and the likelihood of being innocent) is not a relevant consideration for the court, when offering an incentive to plead guilty. The issue to be considered by the court is how much an early plea has saved in economic and emotional (for the victim and witnesses) costs. See Sentencing Council Guidelines 'Reduction in Sentence for a Guilty Plea: Definitive Guide, revised 2007, statement of purpose, para 2.2. McConville and Marsh (n 83 above, 225, n 32) claim that the courts commonly object to providing this discount to persons where the case against them is overwhelming. They do not however object to the likelihood that those who have a less then overwhelming case against them (they might be innocent) will be incentivised to plead guilty at an early stage. : both of these studies report negative assessments of the quality of advice and representation provided to their clients by criminal solicitors. Newman (at 165) found a rhetorical commitment to provide a good service and facilitate justice amongst the lawyers working within both 'radical' and 'sausage factory' legal practices, and a common failure to live up to this rhetoric within both. jury verdicts. Prior to the current UK reforms the illusion, that pleas could be relied upon provided that they were made without pressure or coercion, was maintained by the fiction that such pressure did not arise so long as the judge took no active part in articulating the reductions likely to follow a guilty plea. 140 But the new system of statutory tariff reductions creates a similar level of certainty, and pressure, to that of a judge spelling out the terms of the reduced sentence that will follow from a guilty plea. The standard is that the guilty plea must be a 'true acknowledgement of guilt'. This sounds liberal, until one understands that it operates on the basis that all guilty pleas entered under routine circumstances are to be treated as meeting this standard. One finds examples of successful appeals, but only in cases where there is evidence that, at the time of the plea, the defendant suffered from rather more than the 'normal' pressure to plead guilty. And as noted by McConville and Marsh, there is something approaching a wilful refusal by the judiciary to acknowledge the obvious probability that a considerable number of guilty pleas will be made by innocent defendants.
[O]nce he has admitted such facts by an unambiguous and deliberately intended plea of guilty, there cannot then be an appeal against his conviction, for the simple reason that there is nothing unsafe about a conviction based on the defendant's own voluntary confession in open court. A defendant will not normally be permitted in this court to say that he has changed his mind and now wishes to deny what he has previously thus admitted in the Crown Court. 141 Instead, the discussion typically proceeds on the basis that discounts operate only to persuade the guilty to give up the possibility of securing an acquittal. 142 The possibility that the innocent may plead 140 Classically articulated in R v Turner (1970) 54 Cr. App. R. 352. 141 R v Asiedu [2015] EWCA Crim 714, para 19. In this case it was held that irregularities at trial, such as non-disclosure of evidence that favoured the defence, would not allow a defendant to appeal successfully after a guilty plea unless those irregularities amounted to the legally defined and 'closely confined' notion of an abuse of process. 142 This, we would argue, is demonstrably the approach adopted throughout the latest significant official report setting out detailed, comprehensive proposals in relation to the UK criminal justice system: "streamlining all their processes, increasing their efficiency and strengthening the effectiveness of their relationships …". A review of the Criminal Courts of England and Wales by The Right Honourable Lord Justice Auld, September 2001: http://webarchive.nationalarchives.gov. uk/+/http://www.criminal-courts-review.org.uk See: for trenchant criticism especially as regards sentencing discounts, McConville and Marsh, n 83 above, ch 4 guilty is either ignored, or discounted on the basis that this will in some sense be the innocent defendant's own fault: There are no doubt many defendants who, although they know they are guilty of the offence alleged against them, nevertheless enter a plea of not guilty in the hope of being acquitted. The Bar Council's Code of Conduct makes clear that defence counsel should explain to the accused the advantages and disadvantages of a guilty plea. It goes on that he must make it clear that the client has complete freedom of choice and that the responsibility for the plea is that of the accused. It is common practice, endorsed by paragraph 12.5.1, to tell an accused that he should plead guilty only if he is guilty. 143 (our emphasis)

V ASSESSING THE PRESENT
In light of this brief history, how should we describe the current situation? Should we understand the rights granted to defendants as protections against wrongful convictions or, in addition, as compensations for the necessity to produce 'sufficient' convictions in circumstances in which we cannot be absolutely certain as to the guilt of accused persons? Is our present form of adversarial trial (as we have described it in the UK or, we would argue, other forms of trial in other jurisdictions), with its various rights and procedures, a current expression of the continuing need to overcome a deficit in our ability to be certain about the truth of accusations that particular individuals are guilty of particular crimes? And if it is, does this make miscarriages of justice not simply a problem in themselves (a moral wrong, with many highly reprehensible and unacceptable consequences), but the inevitable outcome of a solution to a prior proble´matique and, in this sense, themselves part of the solution to that proble´matique? And how does our understanding of our present situation alter if we accept this kind of analysis?
Let us recap on the historical evidence. Our earliest procedures can be understood in these terms. The oath, the ordeal, trial by combat and the continental system of trial by torture were expressly reserved for cases where the clearest proof of an offender's guilt was not available. These legal procedures compensated for the lack of evi-Footnote 142 continued 'Lowering the Bar'; for the fuller context, Ashworth and Redmayne, n 82 above, ch 10 'Plea'. 143  dence, and created expectations, in communities, that legal redress could be forthcoming in situations where accusers had less than the clearest evidence of guilt. It is not hard to redescribe these forms of trial in functionalist terms as mechanisms which solved the proble´matique of the need to punish more persons than could be demonstrated, clearly, to be guilty of crimes (assuming of course that those crimes had been committed and that, therefore, someone was 'guilty'). When we move on to consider the jury, the situation becomes more difficult. Because we continue to use the jury in Anglo-American criminal trials, we are less willing to describe it in these terms and, correspondingly, more willing to describe proven miscarriages as evidence of failure in themselves -whether of the fallibility of reliance on trial by lay persons, or of other deficiencies (rather than seeing miscarriages as the solution to a prior proble´matique). It is clear that the jury has its origin in the same functionalist needs that justified the ordeal, as its introduction was expressly to take over the role formerly accomplished by the ordeal. This was not the result of any faith that juries would do a better job of finding the truth than the ordeal. We can suppose that the local jury that took over the function of the ordeal at least had some local knowledge of the crime, and we can also suppose that the later selection of jurors who would have no such local knowledge would be more impartial in their response to evidence than the kind of jury they replaced. But neither of these two changes in the history of juries was a response to any perception that the prior system was failing to identify the factually guilty. Both changes were the result of the need to continue to process persons accused of crime -not a vague and general social need, but the immediate pressure of persons being accused and detained in ever increasing numbers should the assizes fail to reach decisions on their guilt or innocence. With the 18th century jury we see a system which failed to meet current standards for the evidence necessary to justify capital punishments, but which was generally regarded by those responsible for its operation as an appropriate mechanism for getting at the truth of an accusation. However, the system of rewards that undermined contemporary confidence in the bona fides of accusers was allowed to continue for 50 years after the starkest possible evidence of its corrupting tendencies. The fear that there would be insufficient prosecutions (and convictions) was given greater priority than the fear of miscarriages of justice. Until the creation of a professional police force, this concern with the possibilities of corrupt prosecutions could only be dealt with via an increased burden on the jury to separate the false but corrupt witness from the truthful but corrupt witness (aided by warnings from judges on the need for corroboration and the increased use of lawyers to challenge and question witnesses).
The solution to the predicament of the corruption produced by a system of bounty hunting -a professional police force capable of investigating crimes -has produced its own further difficulties. Aside from new forms of corruption (the pressure on police to justify their existence by 'solving' crime -as evidenced by successful prosecutions) developments in policing also operate to increase the complexity of the evidence presented to a jury. We may question the accuracy of jury verdicts reached in minutes by 18th century juries, but these juries had to consider very limited evidence. Today's jury has to consider the results of today's forms of police investigation: forensic evidence, CCTV, telecommunications interception, identification and confessions. In addition, the judicial reaction to dangers represented by professional informers saw the beginnings of a process that remains with us: the juridification of trial procedure. These procedures are organized in terms of the respective rights of the prosecution and defence: to question witnesses, have disclosure, notice of the other's 'case', challenge the jury's membership, etc. Our current situation is therefore one which cannot be described as a single-minded pursuit of the factual truth of a defendant's guilt. Alongside the pursuit of truth, we also have a strong, or system constructed, concern with rights. This prompts the continuing question, the proble´matique: whether this extra element is, like the oath, the ordeal, trial by combat, torture, the local jury, and the short 18th century trial with its reliance on the defendant's character and demeanour, a modern (and in many ways equivalent) means to enable the conviction of persons in the absence of certainties as to their guilt.
Our modern criminal processes undergo considerable changes, or at least what is suggested to be considerable changes, on a regular basis, and often in response to the symbolic importance of the unacceptability of miscarriages of justice. But, at this particular juncture, in the still early stages of the 21st century, with the development of new forms of forensic evidence posing different challenges, and the development of plea bargaining and the guilty plea as a modern equivalent of the need for conviction in the absence of absolute proof, we can expect miscarriages of justice to remain with us. Their background proble´matique is a part of the criminal justice system -an outcome of its operations. Of course, that does not mean that every effort should not be made to reduce their numbers, or guard against their occurrence. But, as they are an intrinsic side-effect of the system's operation, they need to be approached with a less naı¨ve assumption that they merely reflect human failures (whether those of individuals or the systems which they construct). They, in practice, reflect the operation of the system, as much as the punishment of those actually guilty of the crimes with which they have been charged and convicted.

OPEN ACCESS
This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/ licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.