The Rule of Law

The entire enterprise of law can be regarded as one of governing human beings by the translation of moral values and prescriptions into justiciable rules,Footnote ¹ that is to say by rules which can be understood and applied both by those subject to them and by those applying them. It is an exercise in particularisation. Broad overarching principles and values expressed in terms of high abstraction are formulated in terms of lesser and lesser degrees of generality. Such rules describe and define their field of application, and in so doing may need to rely upon further rules and sub-rules to define the terminology and concepts they employ. The structure thus created is one of a vast set of interlocking definitions. It is no accident that formal legal rules are almost invariably expressed in indicative rather than in normative terminology.

At any one time this set of rules has to be regarded as comprehensive in the sense of being capable of application to any situation which may arise. In our, and most other, systems it must also be applied definitively to those situations, in the sense that the judge or court must determine a result by the application of those rules.Footnote ² In an ideal system all situations would further be understood unambiguously by those subject to the rules, and by those applying them. Unfortunately not all situations can be foreseen clearly, and even if they could, ordinary language is not so subtle or precise to allow the rules to remain both unambiguous and clear in their effects to those subject to them.Footnote ³ An important role of the courts is to apply the corpus of rules in such situations so as to arrive at a result in terms of prescription of what is to be done. In so doing, those applying the rules, especially in common law systems, contribute to the development and refinement of the rules. So most decisions apply, centrally the existing rules, and at the penumbra the determinations of the judiciary according to its perception of the policies underlying and motivating the formulation of the rules, and of the system as a whole. To maintain the rule of law to its greatest extent, the input of the judiciary should be kept as close as possible to the minimum required to allow the system to function. In this sense it can aspire to be seen as government by law, and not by men; by articulated rules, and not by some form of palm tree justice.

Application to Evidence

It might be argued that the law of evidence can, and should, be free to depart from the rule of law so understood. It might be thought since it is an aspect of the rules of procedure, and especially of procedure at trial, that nothing would be lost by simply permitting the parties to adduce whatever evidence they wish, and then to leave its use for determination by trial judges who inevitably acquire appropriate expertise to determine such matters, and on account of such expertise, and with the benefit of personal experience of seeing the witnesses and hearing them being examined and cross-examined, ought to be insulated from reversal by appellate courts who have no such advantage.

The largest flaw in such arguments lies in their isolation of the trial from consideration of the pre-trial effects of such procedure at trial, and of practice at trial from the effects of post-trial denial of appeal in such respects. It would mean that no party could be certain in advance of being able to adduce any particular piece of evidence, so adding to the uncertainty of being able to prove his case. Such an increase in uncertainty might well cause unfairness in inhibiting some less affluent potential claimants from bringing their cases to trial, and perhaps also encourage some less scrupulous claimants to bring speculative, or even fraudulent, cases to trial. The absence of clear rules for the judges to apply will tend to multiply and prolong arguments in favour of, or against, the use of particular pieces of evidence. Similarly the insulation of such rulings from appeal will exacerbate determination to succeed at trial, and add further incentive to multiply the evidence and prolong the argument. An additional consequence of abdication of appellate control of such decisions would necessarily be to inhibit the development of clearer, and so more justiciable, rules by traditional common law means of incremental clarification by the decisions of courts, and appellate tribunals.

A further problem is that if the rules are stated at too high a level of abstraction it may become difficult to direct juries with any precision as recently noted by the Court of Appeal in relation to evidence of bad character in R v. Lowe:Footnote ⁴

It may also be remarked that the corpus of pieces of evidence of potential relevance in a modern society, with its stress on the proliferation and preservation of information, is now very great indeed. It is also more complex, and sometimes more difficult to secure access to it and to assess it accurately, now that transactions so often taken place across jurisdictional boundaries, and in manipulable forms. This makes specific guidance all the more necessary.

Rules and Discretion

This basic distinction is between mandatory rules which, upon their antecedents being found to exist, exclusively require a conclusion; and discretions, which upon their antecedents being found to exist, may also inclusively permit that conclusion, but do not then require it. In relation to definitions the distinction is between those which are ungarnished, and those explicitly expressed to be inclusionary.

Discretion can, in principle, either broaden or narrow the application of rules. In the law of evidence it can take an inclusionary form, allowing the admission of pieces of evidence not explicitly permitted by the formal rules, or even explicitly excluded by them; or, more traditionally, an exclusionary form, denying the admission of pieces of pieces of evidence otherwise permitted by the formal rules.

Explicit invocation of discretion in the law of evidence is a modern phenomenon.Footnote ⁶ In criminal proceedings it started as judicial advice to counsel, developed into an exclusionary discretion at common law, attracted statutory endorsement; and finally found its twin in explicitly inclusionary discretion, replete with principles upon which it was to be exercised. Civil proceedings were even slower to accept discretionary exclusion by the judge, but eventually explicit recognition was accorded, notably in the form of rule 32.1 of the Civil Procedure Rules.

This sketch is somewhat misleading, since the expression of the antecedent parts of rules of evidence, both mandatory and discretionary, in terms of high generality and uncertain meaning tends to confuse. Thus in an attempt to preserve the distinction between mandatory rules and discretion courts have been driven to deny that the application of such terms, which must necessarily be applied by the judges, is tantamount to investing them with a discretion.Footnote ⁷ The resulting confusion between judgment and discretion can suggest that a provision apparently conferring a discretion, may operate as a rule.Footnote ⁸ The use of such general terms also tends to elide the distinction between inclusionary and exclusionary discretion,Footnote ⁹ to which the verbiage of countless cases invoking s.78 of the Police and Criminal Evidence Act 1984 bears depressing testimony. It may also have repercussions on the grounds for appellate review.Footnote ¹⁰

It may finally be noted that there is nowadays a tendency, especially in statutory formulations,Footnote ¹¹ to guide the operation of discretion by reference to considerations which should govern its exercise.Footnote ¹² While this probably seeks to reduce the open-ended scope of discretionary exclusion, and so to move toward greater compliance with the rule of law, its effects may actually move in the other direction,Footnote ¹³ since given the reluctance of appellate tribunals in relation to control of discretion to go much beyond ascertainment that the trial court took the proper considerations into account, and eschewed improper considerations, the provision of a list of such proper considerations effectively gives trial judges an insurance policy against successful appeal, so stultifying further development of the law. It also creates difficulty in relation to the burden of proof, since sometimes the operation of a discretion involves a finding of fact, which might raise issues of such a burden, but sometimes raises issues only of classification of agreed facts. Even in cases where such operation does depend upon a finding of fact the issue of the appropriate standard of proof may be complicated by the fact that invocation of the discretion, whether inclusionary or exclusionary, may be made either by defence or prosecution with different standards applying accordingly.

Rules and Principles

As noted above the use of concepts of high generality can blur the distinction between rules and discretion. Its more practical vice is that it is unable to provide clear guidance to those subject to, or seeking to apply, such rules.Footnote ¹⁴ The most egregious example in the law of evidence in this respect is to be seen in the deliberate decision of the Supreme Court of Canada to abandon a “categorical” approach to the hearsay rule, and to substitute an approach dubbed “principled”.Footnote ¹⁵ The genesis of this change is interesting. The court referred to Wigmore and found that he had rationalised exceptions to the hearsay rule in terms of reliability and necessity. While it may be true that such considerations played their part in moulding the rule and its exceptions, it is quite another matter to abandon the lower level rules so moulded in favour of reversion to these concepts themselves.Footnote ¹⁶ It has indeed thrown the whole of Canadian law in this area into the melting pot. The terms are vague and require elaboration in the context of each set of different facts.Footnote ¹⁷ In most cases resolution can come only by holding a voir dire,Footnote ¹⁸ and it has often been found necessary to interpret these terms in very odd ways.Footnote ¹⁹ Predictably enough the confusion caused by reversion to such “flexible”Footnote ²⁰ concepts has generated a huge increase in litigation, including many complex excursions to the Supreme Court,Footnote ²¹ and even larger numbers to the Courts of Appeal of the various provinces.

A further problem in the use of very high-level concepts is that they are liable to overlap, and frequently to point in different directions when applied to any given situation. This is prone to generate some form of compromise, frequently categorised as “balancing”, despite the patent absence of any common unit to which the concepts can be reduced. One of the most flagrant examples in the law of evidence is the balancing of prejudicial effect and probative force, despite one existing in the realm of emotion, and the other in that of logic, and despite the fact that in this context the prejudicial effect of a piece of evidence is understood to connote its being given a weight greater than its true value.Footnote ²²

Nor should it be forgotten that a major part of the law of evidence is used to determine whether or not a piece of evidence can be used at trial, that is whether or not the evidence is admissible. This will normallyFootnote ²³ be determined in advance of the evidence being adduced. It then becomes very difficult to apply backward looking concepts which may well be appropriate at the appellate stage, but which are of little use to trial judges.Footnote ²⁴ A striking example is reference to the fairness of the proceedings. It is very hard to see how such a concept can be applied to the admissibility of a piece of evidence which ex hypothesi has to be adduced before the proceedings have terminated, and so in advance of any definitive categorisation of them as fair or unfair. Nor can matters easily be rescued by reference instead to the likelihood of the proceedings becoming unfair, since that determination must be predicated upon some other factors which would themselves be more appropriate antecedent parts of the rule.

In order to mitigate this problem the judge has an obligation to keep any ruling on the admissibility of bad character under review, and it is not uncommon for it to be adjusted in the light of the course which the proceedings have taken.Footnote ²⁵

Overall it is hard to escape the conclusion that principles are resorted to just because they exist at such a high level of generality that they can easily be agreed in advance, but at the expense of potential disagreement at the point of their application to particular situations.

Discretion

So far as bad character evidence is concerned the scheme of the Act endows the court with an exclusionary discretion in relation to the otherwise admissible bad character of a non-defendant,Footnote ²⁶ but less clearly in relation to that of a defendant.Footnote ²⁷ The position is complicated because the terminology of s.101(3) is virtually identical to that of s.78 of the Police and Criminal Evidence Act 1984,Footnote ²⁸ except in being cast in mandatory rather than discretionary form, but applies to only two of the seven gateways provided by s.101(1). It seems that it was so cast because the nature of the conditions for the operation of s.78 effectively converted it into a rule, once they had been determined to be satisfied.Footnote ²⁹ But if this was the case, it is hard to reconcile the general retention of the operation of s.78 with the specification of only two of the gateways for the purposes of the operation of its modified form in s.101(3). To retain its operation seems to render nugatory the specification of those two gateways. What makes this even more surprising is that this view appears to have been taken by the government during the passage of the bill.Footnote ³⁰ Retention might be justified either on the basis that discretionary exclusion of evidence of the accused's bad character existed before the passage of the 2003 Act, and no explicit alteration has been made to that position; or on the basis that the operation of the discretion depends upon factors other than bad character. Such arguments have however received a cautious response.Footnote ³¹ Functionally and pragmatically the Court of Appeal has recommended a working practice of assuming s.78 to continue to apply,Footnote ³² since it will help to insulate the working of the 2003 provisions from successful challenge under article 6 of the European Convention of Human Rights.Footnote ³³ The court's determination to exercise discretionary control was perhaps most obviously displayed in R. v. Musone where in the absence of any discretion under the 2003 Act, or under s.78, and where none had existed at common law to restrain a co-defendant, it still contrivedFootnote ³⁴ to distil such a discretion from the Criminal Procedure Rules.Footnote ³⁵

In relation to hearsay the role of discretion is clearer. All previously existing discretionary powers to exclude hearsay were expressly retained,Footnote ³⁶ but the more troubling provision is s.114(1)(d)Footnote ³⁷ bestowing a new inclusionary discretion for cases in which the court considers it in the interests of justice for the hearsay statement to be admissible.Footnote ³⁸ It may be noted that in relation to this provisionFootnote ³⁹ “admissible” appears to mean “admitted”, given that non-hearsay reasons for exclusion are kept separate.Footnote ⁴⁰ Section 114(2) lists various factors to be taken into account in relation to the operation of the inclusionary discretion, not all of which will be applicable in every case.Footnote ⁴¹ The admissibility of a statement has even been upheld despite none of the factors having been considered at all.Footnote ⁴² It should also be noted that the Act contains further inclusionary discretions, in s.116(2)(e) for witnesses who are in fear,Footnote ⁴³ and in s.121(1)(c) for multiple hearsay,Footnote ⁴⁴ both predicated upon a more stringent basis than that specified in s.114(1)(d). What is the point of crafting an elaborate exception to admit hearsay, if failure to meet its requirements can be remedied by recourse to s.114(1)(d)? Indeed given the terms of s.114(1)(d), why was there any need to specify exceptions with more rigorous conditions and requirements?

The danger of this situation is that the highly flexible inclusionary discretion under s.114(1)(d) is capable of subsuming not only the more rigorous inclusionary discretions in ss.116 and 121,Footnote ⁴⁵ but also the boundaries of the specific exceptions to the general exclusion of hearsay contained throughout this part of the Act.Footnote ⁴⁶ Thus in R. v. O'Hare the court deploredFootnote ⁴⁷ an attempt by the defence to use s.114(1)(d) to by-pass the requirements of s.116, but was able to avoid succumbing only because the evidence had also to satisfy the conditions for the admission of fresh evidence on appeal. It is still more disquieting to find in R. v. Pulley Footnote ⁴⁸ that the Court of Appeal regarded the factors in s.114(2) explicitlyFootnote ⁴⁹ assigned to the interpretation of s.114(1)(d) as having been wrongly ignored in the determination of the admissibility of hearsay statements of deceased persons, and governed not by s.114, but under the different conditions defined in s.116.

In R. v. Xhabri the Lord Chief Justice first analysed the application of s.120, rejecting two of the routes to admissibility, but accepted a third, yet then, having done so, asserted that the evidence was clearly admissible under the inclusionary discretion in s.114(1)(d). In much the same way he found the higher inclusionary discretion of s.121(1)(c)Footnote ⁵⁰ satisfied, which again seems a pointless exercise if the statement could have been received under s.114(1)(d) without needing to consider the more stringent double-barrelled requirements of s.121.

While the reason for this overlap appears to be a change of attitude towards the scope of operation of the broad inclusionary discretion which was originally intended to be limited,Footnote ⁵¹ it has always been widely-phrased, and so capable of broad application.Footnote ⁵² This was recognised in McEwan v. DPP,Footnote ⁵³ but the court nevertheless took the view that the discretion should not be used as a simple way to disregard a long catalogue of errors and delays by the prosecution in securing admission of a hearsay statement under the provisions of s.116. The case is remarkable also in being a rare example of an appellate court being prepared to reverse the discretion applied by the magistrates below, who, as the court recognised, had meticulously considered all of the matters listed in s.114(2). The broad scope of s.114 was nevertheless re-emphasisedFootnote ⁵⁴ in Sak v. Crown Prosecution Service, where McEwan was distinguished on the basis that the prosecution conduct in Sak had been considerably less heinous. Similarly in R. v. Y the Court of Appeal was in no doubt that s.114(1)(d) was not confined to evidence adduced by the defence, and that it had effectively superseded the common law rule that a confession is admissible only against its maker,Footnote ⁵⁵ notwithstanding the explicit preservation of the common law rules relating to confessions.Footnote ⁵⁶

Vagueness

It would be hard to find a much vaguer or less technical provision than the definition of evidence of bad character in s.98(a) of the Criminal Justice Act 2003 as excluding evidence which “has to do with the alleged facts of the offence with which the defendant is charged.”Footnote ⁵⁷ Yet this definition underpins the whole panoply of statutory provisions relating to bad character in the Act,Footnote ⁵⁸ both that of the accused and that of third parties.Footnote ⁵⁹ It might be thought that relevance to their proof is one way in which evidence of bad character has to do with the facts of an offence.Footnote ⁶⁰ So broad a construction would however subvert all of the detailed provisions for admissibility and inadmissibility of evidence of bad character, since irrelevant evidence is automatically inadmissible.Footnote ⁶¹ So the phrase must be construed more narrowly,Footnote ⁶² but the formal wording provides no explicit indication of the basis upon, or extent to, which such an exercise should be conducted.

The Court of Appeal rapidly noticed the width with which s.98(a) was drawn, and was quite prepared initially to find bad character evidence manifested in a count which had been severed, although in the light of the explanation advanced by the accused, and accepted by the Crown, it found it ultimately to be irrelevant.Footnote ⁶³ It even went so far as to assert that if evidence of bad character fell within the exception of s.98(a) it was admissible “without more ado”. This seemed to ignore any question of relevance, and to be inconsistent with the common law discretion to exclude evidence the prejudicial effect of which exceeded its probative value. However this was subsequently ameliorated in R. v. Watson to the proposition that it might then be admissible,Footnote ⁶⁴ thus preserving the possibility of exclusion on the basis of prejudicial effect, or under s.78 of Police and Criminal Evidence Act 1984.Footnote ⁶⁵

The issue of relevance re-emerged in relation to the operation of s.100 in R v. Machado. The trial judge appears to have taken the view that despite the irrelevance of the evidence to the issues in the case, the tendered evidence of drug taking did amount to evidence of bad character, and was governed by s.100. The Court of Appeal took the view that since the drug-taking was contemporaneous with the facts of the robbery in issue, it fell within the plain meaning of s.98(a), and hence was not evidence of bad character for the purposes of the Criminal Justice Act 2003. It also indicated that it could not be excluded under s.78 of the Police and Criminal Evidence Act 1984, presumably because it was not the prosecution which sought to rely upon it. It thus came to the conclusion that the evidence had been wrongly so excluded.Footnote ⁶⁶

The issue of relevance may depend upon the sort of use to which the evidence of bad character is to be put. It was held in R. v. Malone that if the evidenceFootnote ⁶⁷ were used merely circumstantially, it might be admissible as being relevant on that basis; but if it were used to show the accused's propensity, it would have to be admitted under s.101.Footnote ⁶⁸ It may also be necessary to distinguish the aim of admitting the evidence from the means of achieving that aim. In R. v. Benguit the aim was to show that the accused was in possession of a knife at the time of the relevant stabbing, but the means of doing so was to show his propensity, as a drug dealer, to possess such a weapon. It was held that the need to show propensity entailed the use of s.101, and that the propensity to carry a knife was not ‘to do with the facts of the offence’ so as to take the evidence out of the scope of s.101.Footnote ⁶⁹ Similarly in R. v. Saleem the level of relevance for the purpose of s.98 and “having to do with” the facts of the offence seemed to be cast very high,Footnote ⁷⁰ in excluding evidence of bad character partly because it did not provide a motive or reason for committing the offence.

A further difficulty is that while the evidence might be adduced for one reason, the judge might then, as he did in R. v. Malone, direct the jury that it could use the evidence as propensity evidence.Footnote ⁷¹ It seems however that where evidence relating to different counts in the indictment is concerned, the application of s.112(2) requires that the counts be considered independently, and so if their circumstantial relevance relates to them taken together, this is not then possible, and if they are to be admitted at all, it must be under the provisions of s.101.Footnote ⁷²

It was further claimed in R. v. Malone that since the evidence had not been adduced under s.101, specifically 101(1)(d), the defence had been prejudiced by not being able to rely on discretionary exclusion under s.101(3), although it was conceded that exactly the same considerations would apply under s.78 of the Police and Criminal Evidence Act 1984.Footnote ⁷³

In a joint trial there is the further complication that evidence of the bad character of one of the co-accused may have to with the facts of the allegations against that one, but not against the other. In R v. R such evidence having been admitted against one, was regarded as admissible against the other under s.101, either as explanatory evidence under s.101(1)(c), or under 101(1)(d).Footnote ⁷⁴ Having been heard against the co-accused, it then mitigated any prejudice to the co-accused against whom it was to be used under s.101.

Despite the width of the words in s.98 it was determined in R. v. Tirnaveanu that it was limited by considerations of contiguity in time,Footnote ⁷⁵ although extending both to time beforeFootnote ⁷⁶ and afterFootnote ⁷⁷ the events in issue. In R. v. McNeill this was amplified by reference to the context of the rest of the Act, and particularly s.101(1)(c) and s.101(1)(d), on the apparent basis that there should be no overlap.Footnote ⁷⁸ It was further considered that a separate incident occurring some two days afterFootnote ⁷⁹ the events in issue was sufficiently contemporaneous. Nor, it was said in R. v. McKintosh, does evidence of bad character fall outside the provisions of s.98(a) when its relationship to the facts is by the indirect route of showing that the complainant, as a witness to the facts, is likely to be telling the truth, rather than lying.Footnote ⁸⁰ It is however hard to square this view with that espoused in R. v. Campbell Footnote ⁸¹ and R. v. T Footnote ⁸² that evidence of propensity to lie is not normally of sufficient importance to satisfy the requirements of s.101 or s.100. If by application of s.98(a) such evidence were to evade the need to satisfy those gateways it would appear that the only obstacle to its admission would either be discretionary, or by a determination of irrelevance.

As noted above the higher the level of abstraction, the vaguer and more ambiguous the proposition. Constitutional provisionsFootnote ⁸³ are prime examples of this, as for example illustrated by the treatment of slavery under the constitution of the United States of America. Similar considerations apply to quasi-constitutional provisions such as the European Convention on Human Rights, exacerbated in that case by its varying application in the domestic law of the jurisdictions to which it applies, and by the varying judicial attitudes and practices of the judges seeking to expound and apply it.

The law of evidence has certainly not escaped its uncertain application, either before or after the passage of the Human Rights Act 1998. The problems for the rule of law were well illustrated by the saga of R. v. A (No 2).Footnote ⁸⁴ In crude terms the issue was whether s.41 of the Youth Justice and Criminal Evidence Act 1999 which had essentially sought to eliminate explicit discretionary control from the determination of cross-examination of a complaint of sexual abuse about her previous sexual history, and to substitute a set of more mandatory rules, was, through failing to include in such rules entitlement to cross-examine about a prior sexual relationship between the complainant and the defendant, in breach of article 6 of the Convention by failing to provide for a fair trial. The majority of the House of Lords took the view that it was, and effectively wrote in to s.41 the very discretionary control that the legislature had set out to eliminate.Footnote ⁸⁵ It would have been open to them to have made a declaration of incompatibility between the legislation and the Convention under the provisions of s.4 of the Human Rights Act. That possibility highlights the problem. This section appears to have been intended to trigger re-examination of incompatible legislation, no doubt on the basis that the legislature would not knowingly wish to infringe the provisions of the Convention. If seen simply in terms of the legislature not wishing to infringe rights to a fair trial, that view might appear well-justified. The problem is however that the level of abstraction is too high, since the elaboration of the notion of a fair trial in less abstract terms, such as the point of division between allowing and disallowing cross-examination of complainants about their sexual history, and the ability of judicial discretion to establish the boundary between admission and exclusion, are much more controversial. In other words while the terms of the Convention might be universally acceptable, their interpretation and application by the European Court of Human Rights might well not be.

It is submitted that the rule of law is better promoted by detailed re-consideration of the problem, and the amendment, or, if necessary, enactment of remedial statutory provisions, than by simply allowing them to be set aside and replaced on such a vague basis by the judiciary.Footnote ⁸⁶

Judicial Hegemony

Appellate control of judicial discretion has traditionally been extremely limited, despite the fact that in many cases the facts of the situation are undisputed, and the only real issue is the categorisation of those facts by reference to the frequently vague terminology in which the conditions for the application of the discretion are cast. It might well be thought that in such a situation appellate control is both appropriate and useful.Footnote ⁸⁷ However in the earliest detailed guidance to the application of the bad character provisions of the Criminal Justice Act 2003 in R. v. Hanson Rose L.J. affirmed the traditional position:

Judge L.J. forcefully re-emphasised this view in R. v. Renda by assimilating fact specific judgments to discretion, and assigning their determination to the trial judge, by way of his “feel” for the case, deploring the creation of “authority” from such rulings, and devolving at least primary responsibility to trial courts.Footnote ⁸⁹ It is difficult to reconcile such an approach with the obligation to provide reasons for any ruling on issues of admissibility and reasons for exclusion imposed by s.110 of the Criminal Justice Act 2003,Footnote ⁹⁰ with the general rule laid down in Renda itself on the effect of a concession in cross-examination on whether a false impression had been given,Footnote ⁹¹ and more generally with the later proposition “that the Court of Appeal Criminal Division is the appropriate court in which the correctness of the judge's decision should be questioned.”Footnote ⁹²

By way of elaboration in R. v. Tirnaveanu the Court of Appeal regardedFootnote ⁹³ the interpretation and application of s.98(a) as “a fact specific exerciseFootnote ⁹⁴ involving the interpretation of ordinary words”. While in R. v. Al Badi the Court of Appeal assimilated s.101(3) to s.78 of the Police and Criminal Evidence Act 1984, describing both as discretionary, and regarded the appropriate approach as no different from that applying to the discretion whether or not to sever an indictment, and upheld the trial judge's exercise because:

So light an appellate rein is worth further exploration, distinguishing situations in which the application of discretion by the trial judge has been reversed; those in which the appellate court has applied its own view in the absence of any exercise of discretion by the trial judge; and those in which it has eschewed the language of review, and apparently exercised its own discretion, but then arrived at a conclusion agreeing with the result achieved by the trial judge.

Reversal

The first case to overturn a trial judge's ruling on such an issue appears to have been R. v. Murphy,Footnote ⁹⁶ where the Court of Appeal, while recognising the limitations on its powers in these respects expressed in previous authorities,Footnote ⁹⁷ nevertheless held the trial judge's determination of the relevance of an old conviction to the issues at trial to have been “plainly wrong”,Footnote ⁹⁸ and allowed the appeal. Similarly in McEwan v. DPP in relation to the admission of hearsay under the inclusionary discretion in s.114(1)(d) the Divisional Court, while recognising that the magistrates had conscientiously trawled through all of the relevant factors listed in s.114(2), still heldFootnote ⁹⁹ their exercise of that discretion to have been flawed, and here too allowed the appeal. Sometimes the exercise of discretion by the trial judge is condemned more in relation to procedural than substantive reasoning. This is traditionally described as taking into account factors which should not be so taken, or in failing to take into account factors which should. This reasoning was adopted in R. v. Pulley where the trial judge regarded his discretion as fortified by inconsistency between the hearsay statement and other evidence in the case which the trial judge felt assisted the accused who could then expose such inconsistency. The Court of Appeal found this reasoning flawed,Footnote ¹⁰⁰ since it effectively made an apparently less reliable statement better qualified for admission than an apparently more reliable one, and treated this as justification for substituting its own discretion, which it then proceeded to exercise contrarily to that of the trial judge, and held the evidence inadmissible.

It may be worth mentioning that although in R. v. Y the Court of Appeal reversed the trial judge who had held s.114(1)(d) inapplicable, this was on a question of law, not on the merits, and was in effect against the non-exercise of discretion rather than against its exercise.Footnote ¹⁰¹ So too R. v. McDonald is not strictly relevant in this context since there the evidence had been held admissible under s.103(1)(a), but not under s.101(1)(b), and the conviction was overturned not on the issue of admissibility, but simply because of misdirection as to permissible use.Footnote ¹⁰²

In effect in the earlier cases the Court of Appeal simply substituted its own “feel” for the case, and exercised its own discretion or judgement on a matter of fact for that of the trial judge. Such an exercise clearly reveals the uncertainty and unpredictability of reliance upon discretion and judgment unmediated by justiciable rules. At the close of such an exercise the position for the future is rendered still less transparent than it was before.

Exercise

Notwithstanding the justification for limited control on the basis of lack of appellate opportunity for direct observation of the trial, such lack has on occasion failed to deter the court from exercising its own discretionary and judgemental control when the trial court has failed even to attempt to exercise any at all. Thus in Maher v. Director of Public Prosecutions the magistrates were held by the Divisional Court to have been wrong to admit hearsay under s.117 since its conditions had not been satisfied,Footnote ¹⁰³ but the court upheldFootnote ¹⁰⁴ their decision on the basis that the result would inevitably have been the same had they considered their inclusionary discretions. This view was derived from the magistrates' abstention from the exercise of their discretionFootnote ¹⁰⁵ to exclude the evidence on the basis of its unreliability. It is far from clear that they considered this discretion either, and since reliability is mentioned in relation to only two of the nine factors specified in s.114(2), not all of which need be considered, and that the level of reliability required by s.121(1)(c) is specially enhanced, this argument is remarkably unconvincing. In the result the carefully considered limitations on the admissibility of business records were side-tracked, and the oral evidence of the observer of the incident apparently left untested by cross-examination in the relevant respect. A similar approach was takenFootnote ¹⁰⁶ by the Court of Appeal in R. v. T to the clearly fact-specific, but unarticulated, judgment of the lower court as to relevance.Footnote ¹⁰⁷ In R. v. Gyima the Court of Appeal was also quite specific in asserting: first, that the relevant condition was “a question of fact for the trial judge”;Footnote ¹⁰⁸ second, that the trial judge made “no express finding of fact”;Footnote ¹⁰⁹ and third that it “found no difficulty in making such a finding itself”.Footnote ¹¹⁰ Finally in R. v. Lamaletie and Royce Footnote ¹¹¹ the Court of Appeal took the view that although the application of the s.101(1)(g) had been argued at trial to be inapplicable as a matter of rule, and not as one of discretion under s.101(3), it was nevertheless able to agree with what it predicted the trial judge's ruling would have been, had the defence instead been rested on that basis.

Agreement

In a number of cases the Court of Appeal has agreed with the trial judge about the admission of the evidence, by the exercise of discretion or some fact-specific judgement, but in upholding his decision has failed to advert to the limited basis for its review,Footnote ¹¹² and instead appeared merely to record its own similar view of the relevant factors, which is somewhat confusing, and quite unnecessary, if it really is exercising only a limited power of review. In R. v. Gyima the position is worse confused by the court's characterisation of the exclusionary discretions under s.126 of the Criminal Justice Act 2003 and s.78 of the Police and Criminal Evidence Act 1984, as inclusionary discretions with the exercise of which it agreed.Footnote ¹¹³ Exactly the same approach has also been adopted by the Court of Appeal in relation to unfairness for the purposes of s.101(3) of the Criminal Justice Act 2003Footnote ¹¹⁴ and s.78 of the Police and Criminal Evidence Act 1984.Footnote ¹¹⁵ In R. v. Watson the court went out of its way to assert that the trial judge's decision in relation to the common lawFootnote ¹¹⁶ had to be made “by the exercise of his judgment in the light of all the information he had about the trial”, but then went on to record that “from all the information we have, we are of the view that the trial judge was quite right in reaching the conclusion he did.”Footnote ¹¹⁷ That certainly sounds more like positive agreement on appeal than reluctance to overturn on review.

It is rare for failure to exercise an inclusionary discretion to be considered on appeal, but in R. v. Reid and Rowe Footnote ¹¹⁸ such a discretion was discerned in relation to admitting evidence of bad character without giving the proper notice required by the Rules, and the Court of Appeal found no difficulty in expressing its unqualified agreement with the exercise by the trial judge of this discretion.

In many cases the determination of the trial judge has been upheld by the appellate court on the basis that any mistake has not rendered the conviction unsafe.Footnote ¹¹⁹ While the vagueness of this condition has caused considerable difficulty,Footnote ¹²⁰ its significance in this context is indirect. It does not itself render the rules more vague, but rather reduces any pressure to sharpen their accuracy.

A similar approach has also been made in a number of cases to the assessment of credibility of a witness, which might well be thought even more clearly to be the prerogative of the trial judge. It is disturbing that it has so often resulted in the exclusion of evidence for the defence. Thus in R. v. O'Hare the Court of Appeal refused to permit the adduction of defence hearsay as fresh evidence on appeal because of its own, admittedly convincing, view in his absence that its maker was unworthy of belief.Footnote ¹²¹ In R. v. Musone the Court of Appeal was unable to accept the principal reasons for rejecting the defence hearsay advanced by the judge, but upheld his fall-back position finding a new discretion to reject such evidence in the Criminal Procedure Rules, and then went beyond review to express its agreement with the substance of the trial judge's decision to exclude.Footnote ¹²² It is somewhat paradoxical that while elsewhere matters of law seem to be assigned to the exclusive control of the trial court, these matters of fact should be determined against the defence by an appellate tribunal without seeing the witnesses at all.

It has been accepted ever since the bad character provisions of the Criminal Justice Act 2003 came into force that the purpose for which the evidence has been admitted, and its use pursuant to that purpose, should be explained to the juryFootnote ¹²³ by reference to the facts of the case, and the live issues between the parties. In R. v. Edwards Rose LJ amplified his more general remarks in R. v. Hanson,Footnote ¹²⁴ by stressing:

Although it was held in R v. Highton that the gateways determine admissibility but not the extent of use of bad character evidence, the general guidance mentioned above was applauded.Footnote ¹²⁶ To the extent that the jury will require direction on such mattersFootnote ¹²⁷ Rose L.J. seemed to favour the use of model directions.Footnote ¹²⁸ It is true that in R. v. Campbell the Lord Chief Justice was somewhat critical of the use of specimen directions, and indeed assertedFootnote ¹²⁹ that it was “seldom helpful to rely upon previous decisions on particular facts as if they were legal precedents.” However this criticism seems to have been addressed to directions at a high level of generality, especially when the particular issues mentioned did not arise in the particular circumstances. At the other end of the scale the criticism may be understood to relate to over-particularisation, when the direction can provide no sufficient basis for the decision to become the foundation for the construction of a more general rule. The increased stress on directions is itself the consequence of the erosion of rules of admissibility into criteria for admission of evidence. It is contended here that the lawyer's art is to steer between these two extremes: to promote the development of justiciable rules by eschewing excessive generality which will fail to promote rules which are justiciable, and by eschewing unreasoned decisions on particular facts which will fail to promote the creation of rules of any sort, but propagate instead no more than a “wilderness of single instances.”Footnote ¹³⁰