Introduction
In the Criminal Justice Act 2003, Parliament attempted to simplify, clarify and rationalise the rule against hearsay, in so far as it applies in criminal proceedings. Overall, this has been applauded as a success: a “crafted code” for the fair admission of hearsay evidence.Footnote 1 But central to the statutory reforms was a redefinition and clarification of the concept of hearsay itself. Redefinition was needed. For years, the courts had struggled but failed to provide a clear, watertight definition of what was or was not hearsay at common law. Six years on, however, it seems that the new concept of hearsay is no more satisfactory, and no better understood, than the old. As with the new rules governing evidence of bad character, one set of complexities has merely been exchanged for another, and the new complexities include some that the courts have yet to master, as recently illustrated by the confusion of the Divisional Court in West Midlands Probation Board v. French, Footnote 2 and of the Court of Appeal in R. v. Leonard.Footnote 3
To make matters worse, some of the definitional changes brought about by the Act appear to be unfortunate and almost certainly unforeseen, whilst one of the principal targets for reform, namely the notorious decision of the House of Lords in R. v. Kearley,Footnote 4 has been struck only a glancing blow.
I. Hearsay at Common Law: Definitional Problems and their Consequences
At common law, the general rule was that hearsay was inadmissible in evidence. A recognised exception to this general rule had to be found and applied before any such evidence could be received. As those exceptions formed a largely shambolic and random collection, offering no guarantee that reliable evidence would be thrown a lifeline, the definition of hearsay, and the correct designation of cases on the periphery of hearsay, assumed huge importance.Footnote 5
The Basic Rule
The classic definition of the hearsay rule at common law was that formulated by Sir Rupert CrossFootnote 6 and subsequently endorsed by the House of Lords in R. v. Sharp:Footnote 7
An assertion other than one made by a person while testifying in the proceedings is generally inadmissible as evidence of any fact asserted.
This was in practice a rule against second-hand evidence. Witnesses were (and are still) ordinarily required to testify orally in court and to confine their testimony to matters within their own first-hand knowledge or experience. Such testimony is known as “original” or (commonly, but less accurately) as “direct oral” evidence.Footnote 8 Where first-hand evidence was required, the mere repetition of some previous out of court statement could not suffice. Cross was therefore right in stating that such statements were “generally inadmissible as proof of any fact asserted”, but the demand for evidence to be given at first-hand was the real essence of the rule, and is in our submission a better way of explaining it.Footnote 9
This is apparent from R. v. Rothwell,Footnote 10 in which a police officer (DC Groom) testified that he had seen R hand small “wraps” or packages to a number of “known heroin users” including a man called Law. On the face of it, applying Cross's definition, no hearsay issue arose from this. Groom made no reference to any previous out-of-court statement, but purported to describe facts that he himself had witnessed. The obvious inference was that the wraps contained heroin that R was supplying to these “known users”. But closer examination demonstrated that Groom was testifying only in part from personal knowledge. He had, in particular, no such knowledge of Law's heroin habit, but had relied upon information passed to him by others and upon the reputation that Law had acquired in his dealings with fellow police officers. In other words, Groom was merely repeating what he had been told, and the court was being invited to rely on it.
The Court of Appeal held that Groom “should not have been allowed to give that evidence”. His reference to Law's heroin usage did not cease to be hearsay merely because he initially avoided making reference to his sources. His evidence on that point was inadmissible, said the court, because he “had no direct personal knowledge” of it,Footnote 11 and could not properly swear that his evidence was true.Footnote 12
Hearsay and Reliability
The finding that Groom's evidence was hearsay is not, of course, the same as saying that Groom's sources were necessarily unreliable. The prima facie exclusion of hearsay at common law rested on the generalisation that such evidence is potentially afflicted by dangers of misunderstanding or distortion in transmission or receipt, by the fact that the original maker of the statement was almost certainly not under oath and by the difficulty of challenging the truth or accuracy of a hearsay assertion when the person repeating it to the court has no real knowledge of its truth. Where such evidence is relied upon by the prosecution, it may sometimes be impossible for the accused to challenge it effectively and thus receive the fair trial to which he is entitled under the European Convention on Human Rights, Art. 6,Footnote 13 but this depends upon the circumstances. In other cases, a conviction based on hearsay may be both fair and safe.Footnote 14 At common law, the hearsay rule frequently deprived the prosecution of cogent and reliable evidence, as in Myers v. D.P.P.,Footnote 15 and because the rule applied with equal ferocity to the defence, defendants could be deprived of credible evidence of innocence.Footnote 16
1. Definitional Problems at Common Law
The inflexibility of the hearsay rule and its exceptions at common law was only one part of the problem. There was also dissatisfaction as to the way in which the concept of hearsay had been interpreted by the courts. It had been held for example that an attempt to describe an offender or suspected offender in pictorial form was not a “statement” concerning the appearance of that offender and not therefore subject to the hearsay rule or the rule against narrative.Footnote 17 A sketch or composite image purporting to show (for example) that a robber had a moustache was held to be admissible as the evidential equivalent of a photograph, even in circumstances where a verbal statement (“the robber had a moustache”) would have been excluded as hearsay. A moment's thought reveals that to be nonsense, because the image is merely the product of the eye-witness's assertion, but such was the law, prior to commencement of the Act.
The most problematic definitional issue, however, concerned “implied assertions”. As we shall see, this is a problem that the Act has not entirely resolved.
2. The Problem of Implied Assertions
If, with a view to proving some fact (eg, that D committed a crime) a witness (W) testified that X had expressly told him of that fact, W's testimony was clearly hearsay at common law. By the same token it was hearsay if W testified that X had deliberately insinuated or hinted at this fact, or that X had accidentally betrayed it in the course of an unguarded remark. In either case, W would be relying on the truth of what X had told him and inviting the court to do the same. In R. v. Kearley,Footnote 18 for example, police officers heard unidentified callers asking for the appellant and requesting their “usual” supply of drugs. Such requests carried within them an implied assertion that they had been supplied by him before, and were inadmissible to prove that this was indeed the case.
Some proponents of the “implied assertion” doctrine went further, classifying as hearsay any report of X's behaviour from which the court might be invited to infer that he knew of the fact in question, even if he had neither asserted nor intended to assert any such thing. If, for example, W reported that X had turned up at D's house, trying to buy drugs, this was said to be an “implied assertion” by X that D was a drug dealer.Footnote 19 The debate ground on for many years, and was not resolved even by the House of Lords in Kearley, in which a majority of the Appellate Committee was widely perceived as supporting this broader interpretation of the implied assertion doctrine.
Dissatisfaction with Kearley was a major feature of debate in the period leading up to the hearsay reforms contained in the Act. The intended impact of the statutory redefinition was to retain the thrust of the common law's definition, as described by Cross and accepted in Sharp, whilst limiting it to deliberate statements (verbal, pictorial or otherwise) and leaving inadvertent assertions outside the rule. The courts have rightly rejected arguments that some vestige of the hearsay rule survives at common law, independently of the Act.Footnote 20 To determine what is or is not hearsay, one must accordingly look in all cases to the provisions of the Act itself; but this does not mean that the common law definition can now be forgotten. On the contrary, some provisions of the Act cannot easily be understood without it.
II. The Criminal Justice Act 2003: Importance of Proper Classification of Hearsay
In addition to redrawing the definitional boundaries of hearsay, the Act makes significant changes to the circumstances in which hearsay may be admissible in evidence. Section 114 creates broad new pathways to admissibility: a four-lane highway where reworked statutory and preserved common law exceptions sit alongside a completely novel route of hearsay admissible by agreement, and where all three are potentially by-passed by section 114(1)(d), which provides a structured inclusionary discretion for otherwise-inadmissible hearsay which it is in the interests of justice to admit. Together, these exceptions add significantly to the chances of a trial being influenced by second-hand evidence of which the witness in court has no first-hand knowledge. Evidence such as that rejected in Rothwell might now be received if the court could be persuaded to exercise the inclusionary discretion; an exercise that would include consideration of the reliability of Groom's sources, the importance of the evidence that the men were “known users” and the problems faced by R in challenging the evidence at second-hand.Footnote 21
Thus the threshold question of what is or is not hearsay may perhaps have lost some of its old importance – where it is going to be admitted, the practical significance will lie in ensuring that adequate notice is given of the intention to use it, and that the tribunal of fact is adequately aware of its shortcomings. In such cases a court or judge may thus be tempted to reason that “it doesn't really matter whether this is hearsay or not”. But there are still good reasons for seeking to define the exact boundaries of the concept:
• If evidence is hearsay, a statutory or preserved common law basis for admissibility must be identified before such evidence can be admitted; whereas none of these need ordinarily be considered if the evidence is original;
• If evidence is hearsay, but is admissible under one of the new provisions of the Act (rather than under an older, preserved exception to the rule) notice must be given in accordance with the Criminal Procedure Rules, part 34, and its associated Practice Direction, before any such evidence can be adduced;
• If there is multiple hearsay, the requirements of section 121 may additionally need to be satisfied, but these may be ignored if no more than one hearsay statement is involved;
• Similarly, the provisions of sections 122–124 (which deal with the production of hearsay documents as exhibits, and the competence, capacity and credibility of the makers of hearsay statements) can all safely be ignored if the evidence in question is not hearsay in the first place;
• Section 125, which empowers a court to stop a case based on unconvincing evidence adduced by the prosecution, and 126, which empowers a court to exclude seriously unsatisfactory and time-wasting evidence, are similarly inapplicable unless it is first determined that the evidence in question is hearsay. Prosecution evidence is subject to a judicial power of exclusion, whether it is hearsay or not,Footnote 22 but that has no application to defence evidence. This means that a court may have some discretion to exclude potential defence evidence, but only if it takes the form of hearsay.
• Section 133, which allows for the use of authenticated copies in place of original documents, applies only to documents containing statements, and must therefore (by section 115(1)) be restricted to copies of documents that contain at least one statement by a person within the meaning of section 115(2).Footnote 23 Such statements are hearsay whenever adduced as proof of any “matter stated” therein.Footnote 24
III. Problems with the Redrawn Concept of Hearsay
The redrawn concept of hearsay has given rise to a number of problems. Some of these are embedded in the Act, while others arise from interpretational errors in the courts.
1. Problem (1): Kearley, Relevance and Hearsay
In lawyer's shorthand, it might be said that a principal aim of the Act was to solve the Kearley problem,Footnote 25 but Kearley is a problematic decision at more than one level, and only one of these is addressed in the Act.
The facts are well-known: K's house was searched by the police and a modest quantity of amphetamine sulphate was found hidden in a rabbit-hutch. The main issue that arose (primarily for reasons connected with a confiscation order) was whether it could be proved that K was a drug dealer rather than a mere user or occasional sharer. The prosecution attempted to prove he was a dealer by adducing testimony from police officers to the effect that in a short period following K's arrest no fewer than 18 calls were made to his telephone or front door by unidentified persons who asked for him by name (“is Chippie there?”) and were evidently seeking to purchase drugs from him. The House of Lords held this evidence to be inadmissible.
The Court of Appeal in R. v. Singh Footnote 26 and more recently in R. v. Nation Footnote 27 assumed that the disputed evidence in Kearley was deemed inadmissible only because it took the form of implied assertions, and that evidence of such assertions would now be admissible because they are no longer classified as hearsay thanks to the Act.Footnote 28 As Rose L.J. put it in Singh: Footnote 29
What was said by the callers in R. v. Kearley would now be admissible as direct evidence of the fact that there was a ready market for the supply of drugs from the premises, from which could be inferred an intention by an occupier to supply drugs. The view of the majority in Kearley, in relation to hearsay, has been set aside by the Act.
With respect, this ignores the fact that Kearley was first and foremost a decision about relevance. All five members of the Appellate Committee in that case agreed that certain inferences could properly be drawn from the behaviour of the callers. It could be inferred that those callers were drug abusers who believed K might be able to supply them with the drugs they wanted. Their actions also proved that there was a ready market for any amphetamine that K might have been willing to sell. For the dissenting minority on the Appellate Committee, that was enough to ensure the admissibility of that evidence. But the majority disagreed. Lord Ackner said:Footnote 30
Each of those requests was, of course, evidence of the state of mind of the person making the request. He wished to be supplied with drugs and thought that the appellant would so supply him. It was not evidence of the fact that the appellant had supplied or could or would supply the person making the request. But the state of mind of the person making the request was not an issue at the trial; accordingly evidence of his request was irrelevant and therefore inadmissible. If the prosecution had sought to call any of the persons who had made such requests, merely to give evidence of the making of the request, in order to establish their appetite for drugs and their belief that such appetite would be satisfied by the appellant, such evidence could not have been properly admitted …. The application of the hearsay rule does not, on the facts so far recited, fall for consideration. The evidence is not admissible because it is irrelevant. It is as simple as that.
Lord Oliver and Lord Bridge agreed. Lord Bridge said:Footnote 31
The state of mind of the person making the request for drugs is of no relevance at all to the question whether the defendant is a supplier.
In other words, the prosecution could not prove K was a drug dealer merely by proving that other persons believed he was one, or (pace Rose LJ) that they provided a ready market for him to sell to. The “ready market” argument was expressly rejected by Lord Oliver, who said:Footnote 32
Whilst it is no doubt true that the mere supplying of drugs from premises necessarily implies the existence of customers anxious to acquire them, that is not a proposition which can legitimately be used to demonstrate the converse, i.e. that the existence of such customers implies the supply of drugs. That can be demonstrated only by establishing the truth of the customer's belief that they will be supplied. That belief cannot by itself be prayed in aid to substantiate its own veracity. Footnote 33
The hearsay rule would need to be considered, according to the majority, only where the caller's request carried with it an express or implied assertion that the appellant was indeed a supplier of drugs: as Lord Bridge put it, “the sole possible relevance of the words spoken is that by manifesting the speaker's belief that the defendant is a supplier they impliedly assert that fact”. In particular, two of the callers asked for their “usual” supply, thus impliedly asserting that K had supplied them with drugs before. To the extent that the words, thus construed, now had some obvious relevance to the issue, they became objectionable instead on grounds of hearsay. Lord Bridge posed this question:Footnote 34
If the speaker had expressly said to the police officer that the defendant had supplied him with drugs in the past, this would clearly have been inadmissible as hearsay. When the only relevance of the words spoken lies in their implied assertion that the defendant is a supplier of drugs, must this equally be excluded as hearsay?
His answer to that question was affirmative, and Lords Ackner and Oliver took the same view.Footnote 35 To this extent, but only to this extent, the 2003 reforms have an impact. The caller who refers to his “usual supply” doubtless supposes he is speaking to someone who already knows about his past dealings, because the reference would otherwise be unintelligible. The primary rule on implied assertions within the Criminal Justice Act limits hearsay to statements made with a purpose to “cause another to believe the matter”;Footnote 36 a purpose which would be lacking on the part of such a caller, so that the call becomes relevant non-hearsay evidence to the extent that it suggests an ongoing involvement in drug dealing.
But if (as in the majority of the calls logged in Kearley) the caller simply asks whether a particular person is around and has any drugs to sell, the “relevance” problem identified in Kearley remains untouched by the Act. Proof of the existence of a ready market for drugs does not of itself prove either past dealing by that person or that he has any intention to supply drugs in the future. In the absence of further evidence, we have no means of knowing why the callers in Kearley came to believe K was a dealer in the first place. If one of the callers had testified in person, but only to the extent of narrating that he had offered to buy drugs from K, then in the view of the majority his evidence would have been dismissed as irrelevant; and as Lord Oliver explained:Footnote 37
It cannot be right that that which would be irrelevant and objectionable if tendered directly nevertheless becomes relevant and admissible if tendered as hearsay.
We would support a purposive interpretation of the Act, and admit that there is much to be said for a more realistic approach to relevance, such as was advocated by the dissenters in Kearley itself.Footnote 38 However a purposive interpretation can only be taken so far, and in our submission the Act contains nothing that could support a judicial reappraisal of what is or is not relevant.Footnote 39
In some at least of the cases decided before the Act came into effect, notably R. v. O'Connell, Footnote 40 the Court of Appeal correctly discerned the basis of the decision in Kearley; but in Singh and Nation it appears to have forgotten this, and has therefore misunderstood the ambit of the Act.
2. Problem (2): Implied Assertions
Given that the Act sets out to replace the common law of hearsay, it is surprising to find that the term is not explicitly defined in it. Section 114(1) (“Admissibility of Hearsay Evidence”) provides that “a statement not made in oral evidence in the proceedings is admissible as evidence of any matter stated if, but only if …” and then details four circumstances in which evidence of such statements may be admissible.Footnote 41 This echoes Cross's famous definition of hearsay at common law, and in R. v. Horncastle the Court of Appeal clearly regarded the rule as continuing to provide for the careful regulation of second-hand evidence, based on the premise that it is “ordinarily essential that the truth of a matter be given in person by a witness who speaks from his own observation or knowledge”Footnote 42.
In affirming the status quo, Horncastle assumes that “evidence of any matter stated” imports the common-law distinction that hearsay applies only where a statement is adduced or received as evidence of the truth of a matter stated. This assumption is clearly correct, and is implicit in section 121(2), which provides that for the purposes of section 121 (multiple hearsay), “‘hearsay statement’ means a statement, not made in oral evidence, that is relied on as evidence of a matter stated in it”. As Roberts and Zuckerman observe,Footnote 43 Parliament cannot have intended to use the term “hearsay” in section 121 in a different sense from that used the rest of the Act.
The Act's failure clearly to articulate the concept of hearsayFootnote 44 is however unfortunate and is particularly dangerous because “matter stated” receives further definitional attention in section 115(3), which sets out a primary rule that a “matter stated” is one to which the hearsay Chapter applies if (and only if) the person making the statement has a particular purpose – in the most common case a purpose to cause another to believe the matter.Footnote 45 Here the focus is on the purpose of the person making the statement, rather than that of the party seeking acceptance of the truth of the evidence, and the object, as noted above, is to excise from the hearsay rule unintended implied assertions such as those of the two callers in Kearley who asked for their “usual”.
The trap that the Act thereby lays for the unwary is based on a “false converse” invited by section 115(3) in combination with the incomplete section 114(1). Where the maker's purpose was clearly to convince another person of some fact, it might seem that his statement must be hearsay whenever the facts he states are put in evidence, and irrespective of whether it is tendered to establish the truth of the fact in question.
Suppose for example that the defence wish to prove that a person, not called as a witness, told the accused that a bridge had collapsed in a storm. Suppose further that the reason for adducing this statement was to explain why the accused had then driven to his destination by a long and tortuous route, this being an essential ingredient of his alibi for a serious crime. Here there is a “matter stated” – “the bridge is down” – but the evidence is not hearsay because the purpose of the defence in tendering the statement is not the same as that of the maker, but is rather to explain the actions of the accused on receiving the relevant information. Such evidence may be relevant and admissible even though all parties now agree that the bridge was not down at all.
At common law, numerous examples could be found of assertive statements that were relevant for some reason independent of their truth, and did not therefore contravene the hearsay rule when received for that purpose. The classic example was provided by Subramaniam v. Public Prosecutor,Footnote 46 where S sought to prove that he had acted under duress by telling the court of threats that had allegedly been made against him. As the Privy Council rightly discerned, the “truth” of these threats was not in issue. The only issue was whether they had caused S to act under duress. In this essential sense, the acid test for hearsay has not changed. In terms of the interpretation of hearsay in the courts, however, the litmus paper is coming up the wrong colour on a worryingly regular basis.
R v. Leonard – A question of purpose
From the foregoing discussion one might predict that confusion is most likely to arise in cases where the purpose of the person making the statement is closely allied to, but significantly different from, that of the party adducing that statement in evidence; and so it proves. In R. v. Leonard,Footnote 47 L was convicted of possession of a Class A drug with intent to supply, but the Court of Appeal held that the trial judge had wrongly admitted evidence of two text messages sent to L which were relied on by the prosecution as evidence of recent dealing. In one, the unknown texter thanked L for “the gear” which he proclaimed to be “well sound”. In the other, the texter (again unknown) complained about the “£5 joey” for which he had paid his last £10. The Court of Appeal considered that these messages each constituted hearsay, and went on to find that no saving exception could be found,Footnote 48 which meant that the evidence had been wrongly admitted.Footnote 49
In so finding, the court appears to have overlooked the essential distinction between “the matter” which the texter in each case intended to assert, and “the matter” on which the prosecution had sought to rely. In each case, the texter intended to cause L to believe a matter – that the gear was good and satisfying, or that the joey contained short measure. There was, however, no intention to assert the one matter for which the prosecution sought to rely on the texts, i.e., that L had supplied each with drugs. The texters operated on the assumption that L already knew that, so did not seek to cause him to believe that matter. The mere fact that a statement is expressly assertive does not (as in “the bridge is down”) preclude reliance on the statement for another, non-hearsay purpose.
This distinction can be hard to draw when (as in Leonard) the party seeking to adduce the evidence is seeking to draw out an unspoken assumption behind the words. This seems to have been the stumbling block for the court. Having noted that the texters were asserting facts or opinions about the quality and quantity of the drugs,Footnote 50 the court rightly spotted that the prosecutor had a different purpose in tendering the texts in evidence, namely to invite the jury to infer that L had recently supplied drugs to the texters. The error creeps in here:
But in order that the jury might make that inference, the Crown first has to establish as facts the matters stated in the texts. The Crown's interpretations of those statements were that, in the first one, the sender of the text was happy with the drugs that had actually been supplied; and in the second he was not happy ….
With respect, the prosecution had no interest in establishing the level of satisfaction of the customer, but merely invited the jury to infer that these observations (true or false) must have been made by persons who had received drugs from L. The process is not dissimilar to that in Lydon, where the prosecution relied on the finding of a note bearing the words “Sean Rules OK” in circumstances suggesting it had been left by the offender, not to suggest that Sean ruled anything, but to suggest that the writer was likely to have been called Sean. The only argument that could be put against the reception of the evidence in Leonard is that the matter relied on – recent supply – is being asserted by the texter, if not expressly then impliedly, but this can be answered by saying that, as to that matter, the texter had no intention to cause the recipient to believe it, so it cannot be a “matter stated” under section 115(3).
Reference to the intention of the Law Commission in proposing the clause that became section 115(3) reinforces the conclusion that the court in Leonard got it wrong. According to the Law Commission, words or conduct that are not intended by the maker to be assertive of the fact in issue (whether they are intended to assert some other fact or are entirely non-assertive) ought not to be classed as hearsay.Footnote 51 Much of the rationale for exclusion – the risk that the maker may be lying – ordinarily disappears where the maker believes that the person he is addressing already knows the fact in issue: why for example would two of the callers in Kearley have asked for their “usual” supply unless they had dealt with him before? The callers might perhaps have been mistaken, but if the possibility of deliberate untruthfulness can be discounted, the Law Commission was persuaded that the risk of error should not render the statement hearsay. In the case of a statement tendered by the prosecution, it would be necessary to establish beyond reasonable doubt that the maker had no purpose to convey the relevant fact – a high hurdle, but a necessary one if the risk of mendacity was to be discounted.Footnote 52
In formulating this proposal, the Law Commission was not following the reasoning of the dissenting minority in Kearley, who drew from the conduct of the callers a circumstantial inference of a market or opportunity for dealing that was both relevant and independent of any inference as to the belief of those making the calls. Rather, the proposal was intended to echo the legislative framework propounded for Australia and that in the United States Federal Rules, as well as the writings of English academics of the stature of Cross and Tapper. Had the Australian precedent been directly incorporated it would have stated:
Evidence of a previous representation made by a person is not admissible to prove the existence of a fact that the person intended to assert by the representation.Footnote 53
The Law Commission, however, preferred its “cause to believe” formula to the more familiar language of “intention to assert” for two main reasons that may ultimately prove unsatisfactory. The first was to avoid confusion where an assertion could be said to be intentional, but was not intended to persuade anyone of anything (“I have come for my usual”). The second was to avoid suggesting that a statement had to be “implied” to benefit from declassification as hearsay. Thus the statement to an alleged dealer “that stuff you sold me last week was bad” would be hearsay as to the quality, but not as to the previous week's dealing, as, from the speaker's viewpoint, the dealer already knew about it, “and in that sense there is no intention to assert”.Footnote 54
In retrospect, and in the light of Leonard, it might have been safer for the Law Commission to have conveyed its intentions in a single proposition distinguishing, as the Australian Uniform Evidence Acts distinguishes, between the intention (if any) of the maker of the statement and the purpose served by putting the statement in evidence. This, together with an express articulation in the definition that the purpose of adducing the evidence should be to establish the truth of a matter stated would perhaps have prevented the confusion which has served to keep within the hearsay rule the very sort of case which the Law Commission intended to liberate from it whilst preserving the principle that the acid test is, and always has been, contingent on the purpose of the party by whom the statement is tendered in evidence – ie the use which the court is invited to make of it.Footnote 55
3. Problem (3) – Licences, Contracts and Operative Words
A similar criticism can be levelled at West Midlands Probation Board v. French,Footnote 56 but to understand what went wrong in that case it is necessary first to introduce what the Law Commission considered to be a necessary rider to its primary rule that hearsay requires the maker of the statement to intend to “cause another to believe it”. This came about because it was anticipated that the primary rule would, without elaboration, lead to the wrong conclusion where the purpose of the maker was not to persuade of the truth of his statement but to cause another to act on it as though it were true, whether the hearer believed it or not. The example given was of an expenses claim where the purpose of the maker is to secure payment and the maker may be indifferent as to whether the recipient believes him. Yet the nature of the representation is such that the possibility of concoction by the maker cannot be disregarded, therefore (reasoned the Law Commission) the formula requires elaboration to bring the example within the hearsay rule, together with the equivalent problem posed where the representation may be directed at a machine rather than a person – the maker is still seeking to convey the relevant impression and might be lying. Hence the rider (which became section 115(3)(b)): “or to cause another person to act, or a machine to operate, on the basis that the matter is as stated”.Footnote 57
Neither of these elaborations was prompted by hearsay problems that had arisen in practice – but in the parallel universe of the law of deception and fraud, both had been incorporated into legislation dealing with different forms of lying behaviour. Incorporating them within the already convoluted world of implied assertions creates the potential for enormous confusion. Does the unknown caller who asks for drugs not have as his purpose to “cause another to act on it [the request] on the basis that the matter is as stated”? This is not the kind of case that the rider was intended to address, but excluding it requires a sophisticated approach to what matters are stated. In the above example, it is relatively easy to respond that a simple request does not state “a matter” at all.Footnote 58 What then if the unknown caller asks for more of his “usual” supply? If this falls within the rider, however, the Law Commission has shot itself in the foot, as it was precisely this sort of implied assertion that section 115(3) was intended to liberate from the trappings of the rule.
Now for French: F was charged under the Criminal Justice Act 1991, s. 40A(4), with breaching the terms of a licence under which he had been released from an earlier sentence of imprisonment. Specifically, he was charged with failing to attend supervision appointments with his appointed probation officer.
F would not formally admit that he had been subject to any such licence or that he had been made aware of any requirement to meet with a probation officer, and it was accordingly necessary for the Board to prove those matters as part of its case. To do so, it adduced evidence in the form of a photocopy of the licence document or “notice of supervision” itself. This was addressed to F and signed both by him and by a governor of H.M.P. Blakenhurst, who signed it on behalf of the Secretary of State. It stated, inter alia:
You are being released on licence for a period of three months. You will be under the supervision of a probation officer or social worker … and must comply with the conditions of this licence …. Your supervision commences on 01/05/2007 ….
It then stated that F was required to report immediately to his probation officer at a specified address and maintain regular contact with him thereafter.
For some reason the Board assumed that the licence was a form of hearsay evidence and that notice of intention to adduced it was required under part 34 the Criminal Procedure Rules 2005. These assumptions prevailed throughout the trial, at the end of which the magistrates concluded that the licence was not only hearsay but inadmissible. On appeal, the Divisional Court also concluded that the licence was hearsay, but hearsay admissible either as an official document within the terms of Criminal Justice Act 2003, s. 117, or as a certified copy of an order within the terms of the Documentary Evidence Act 1868.Footnote 59
The error in the reasoning of the court in French was twofold. First, it failed to apply the “acid test” to distinguish between the purposes of the maker of the statement in the licence and the purpose of the Board in tendering it in evidence. It held that the matters in issue were to be proved by “statements” not made in oral evidence, and that the prosecution was relying on “matters stated” within the licence because the purpose of the Governor who was the official signatory was either to cause others (F, prison staff, the Home Office, etc.) to believe the matters or it was to “cause other persons to act on the basis of those matters” within section 115(3). Therefore, the court concluded, “the licence falls within the definition of section 114 of the Criminal Justice Act 2003”. That conclusion, however, was misconceived because the purpose of the Board in adducing the various statements was simply to show that they had been made to and signed by F. No issue arose as to whether they were or were not “true”. The purpose of the Governor in signing the licence did not determine its hearsay nature; it was the purpose of the party putting the statement in evidence that was crucial. In any event, the court also failed to recognise that the Governor's purpose was not to induce a belief in the truth of the conditions, or to cause another to act on the basis that they were as stated, but simply to set them out in the appropriate form for signing by F.
In our submission, therefore, the use of the licence to prove the facts in question did not involve any issue of hearsay. Indeed, production of the original licence, far from being hearsay, would have been the best form of original evidence open to the Board: “Here is the licence; here are the terms; and here is the respondent's signature of acceptance”. As with a written contract, production of the document itself is ordinarily superior to parol evidence of any kind.Footnote 60
The error in French was foreshadowed in R. v. Walker, a case dealt with under the “multiple hearsay” provisions of section 121, but which on proper analysis involved only one degree of hearsay. W was alleged to have shot and killed a man. His cousin, B, testified that shortly after the shooting he was given a gun (later identified as the murder weapon) by C, who said “Your cousin said hold this and he'll collect it in the morning”. The court took the view that C's purpose in speaking to B was to “cause B to act”, namely, “take the [gun] and temporarily look after it for [W]”. That, it was said, was the first degree of hearsay. As C was W's co-accused, he could not be called by the prosecution to testify as to what W had told him, and so the multiple hearsay provision was supposedly engaged. But (with respect) B's testimony involved only one degree of hearsay. He had only hearsay knowledge (derived from C) of what W supposedly wanted him to do, but he could testify at first-hand that C told him to hold the gun for W. Words do not become hearsay merely because the speaker wants others to act on them. Nor did any issue arise as to the truth (or otherwise) of anything that W had reportedly said to C. It did not matter that he failed to collect the gun from B in the morning. Had C testified, his evidence as to W's instructions concerning the disposal of the gun would also have been original evidence.
If French and Walker are followed, much evidence that has hitherto been regarded as original will be labelled as hearsay. Words of offer and acceptance, tendered in proof of the existence of a contract, do not contravene the rule as traditionally understood because their truth is not in issue. They are “operative words” that have an effect in law independently of the actual intentions of the parties. If we were to ask, “Who has the necessary first-hand knowledge to give evidence of these matters?” the answer would be, “Anyone who heard those words being spoken”.
In Woodhouse v. Hall, which was approved in Kearley, the issue was whether a massage parlour was operating as a brothel. By analogy with cases on contract, the Divisional Court held in that case that, because a brothel was a place where more than one woman “offers” herself for sexual services, it was not hearsay for the offers to be proved by the undercover police officers to whom they had been made, without calling the women concerned.
This analysis was not intended to be threatened by the Act. Approaching the matter as the court did in Walker, however, would yield a different answer. Does the woman offering the service have the purpose of causing another person to act? Of course she does: there is little point in making an offer otherwise. Similarly, applying the logic in French, does she seek to cause the offeree to believe, and to act, as though the offer were genuine? Again, the answer is that she does. The words are significant to the hearer precisely because they are regarded as indicative of the intention of the speaker to carry out the offer should he decide to accept. Even if the woman is secretly intending to take the punter's money and run, she still wants him to believe that the offer is genuine. But her purpose does not determine that the evidence is hearsay. What section 115(3) is intended to do is to exempt some statements (not intended to be assertive) from the hearsay rule. Misunderstandings in decisions like French and Walker, prompted by the over-elaborate inclusion of the “cause to act” formula, permit the provision a far wider sphere of operation that threatens to deflect attention from the essential hearsay question of the purpose for which the out of court statement is tendered as evidence.
IV. Hearsay, Narrative and a New Take on Confessions
1. Statements as Narrative and as Hearsay
Regulating the evidential use of previous statements is the function not only of the hearsay rule but also of the rule against narrative, by which a party is ordinarily precluded from reinforcing a witness's original testimony by reference to his previous consistent statements on the subject. It is in some respects convenient to conflate the hearsay and narrative rules, but there were (and are still) good arguments for keeping the two variants distinct, if only because the rule against narrative was (and remains) based on a different rationale and was (and remains) subject to a largely different set of exceptions.
Few of the concerns about hearsay apply to cases in which the court or jury is told of a witness's previous statement. Even though the statement in question may not have been made on oath, the witness will testify on oath and can be cross-examined as to its truth or accuracy. The rationale behind the narrative rule is that a previous consistent statement merely duplicates, at second hand, evidence already before the court. The only purpose such evidence could serve is to enhance the credibility of the witness's sworn testimony by emphasising his consistency. Some might argue that it should indeed be admissible for that very purpose, because consistency is the opposite of inconsistency, and inconsistency clearly damages credibility; but the common law recognised that logic only up to a point and only in certain exceptional circumstances. The narrative rule was based, rightly or wrongly, on the assumption that a witness's consistency ordinarily proves nothing. As Humphreys J. observed in R. v. Roberts,Footnote 61 it may be of no real help to the court or jury in assessing the credibility of an accused's defence to learn that he has previously told other persons what his defence would be. A good liar will lie consistently, but remains a liar.Footnote 62
Although the hearsay and narrative rules thus had different rationales and were each subject to a largely different set of exceptions, there was common ground between them wherever one witness was called to prove the consistent statement of another. If, for example, W, a rape complainant, was permitted to testify that she had promptly told her sister, X, about her ordeal, it was necessary to call X to confirm that W did indeed tell her this.Footnote 63 X would not however be able to do anything more than confirm that W had indeed made the complaint. A “recent complaint” of rape was a recognised common law exception to the narrative rule, but the hearsay rule (in its narrower sense) still precluded X from purporting to give evidence as to the truth of W's complaint. If for some reason W failed to give the oral evidence expected of her, X's evidence of the complaint would generally have been left with no legitimate purpose to serve and was then inadmissible unless some exception to the hearsay ruleFootnote 64 could be invoked in respect of it.Footnote 65
The same principles applied where evidence was admitted to rebut a suggestion or recent fabrication by the witness or to show that a witness to identification had previously identified the defendant under controlled conditions. In Sealey v. Trinidad and Tobago,Footnote 66 Lord Hutton explained the position by adopting this passage from Blackstone's Criminal Practice 2002:
In the ordinary course of events, where the identifying witness has testified adequately against the accused at trial, the pre-trial identification serves to prove his consistency and his ability to make an identification under fair and objective circumstances. It is admissible, in other words, by way of an exception to the rule against previous consistent statements …. If the police officer who supervised the identification parade is called to testify as to the identification, he can do so only in support of the identifying witness. His testimony cannot go to the issue of the accused's guilt, because he has no first-hand knowledge of it…. There is no special rule making hearsay admissible in identification cases.
Section 120 largely eliminates the distinction between admissible evidence of fact and evidence going only to credit; and section 119 performs a similar function in respect of a witness's previous inconsistent statement. By section 120(2) for example:
If a previous statement by the witness is admitted as evidence to rebut a suggestion that his oral evidence has been fabricated, that statement is admissible as evidence of any matter stated of which oral evidence by the witness would be admissible.
As David Clarke J. noted in R. v. T,Footnote 67 this is not itself a provision governing admissibility:
What the sub-section does is to regulate the use to which such evidence, once admitted, may be put. It is then admissible as evidence of the truth of its contents, not merely as evidence going to the issue of consistency.
According to the Court of Appeal in R. v. Athwal,Footnote 68 this has brought within the concept of hearsay evidence that previously fell outside it. In one sense this is correct. But one must be careful here not to get things back to front. The Act has in this respect relaxed the strictness of the old law. It has made admissible as hearsay evidence that previously would have been inadmissible as hearsay but admissible only as narrative.
In practice, that change is far more significant in respect of previous inconsistent statements under section 119. Where the witness testifies in court and is shown to have said the same thing before, the practical effect of that (whatever the law may say) is to establish his consistency. But if, as in R. v. Joyce,Footnote 69 the witness has changed his story or suffered a real or pretended loss of memory (“I don't remember/I saw nothing /I must have been mistaken”, etc.) the effect of a previous inconsistent statement under section 119 may be to put before the court evidence that would not otherwise be before the court at all; and by virtue of section 119 a court or jury may now convict on the basis of that previous statement where the prosecution case must otherwise have collapsed.
2. Unintended Consequences – Diaries
One surely unintended consequence of section 115(3) is that some deliberate out of court statements may no longer be regarded as hearsay, even where the court is then asked to rely upon them as true. In R. v. Knight, Footnote 70 C complained of sexual offences allegedly perpetrated against her when she was under 16, and was supported in her account by her aunt, A, who had read C's secret diary in which her alleged relationship with K was chronicled in detail. On appeal, it was contended that A's recollection of the contents of the diary (which C burned when she discovered the intrusion) was inadmissible. Prior to commencement of the Act this would have been a safe bet – A's testimony as to the diary would have breached both the rule against narrative and the rule against hearsay. But under the Act, the diary escaped the operation of the hearsay rule, in consequence of the definition of “matter stated”. C was writing only to herself, and thus had no intention to cause another to believe any matter. The rule against narrative was not mentioned in the judgment – presumably for reasons subsumed in the court's conclusion that A was giving “direct evidence … parallel to observing someone fleeing, or lashing out at someone, or blushing at a significant moment”.Footnote 71
The description of a secret diary as direct evidence derives from the earlier case of R. v. N(K), Footnote 72 which was followed in Knight, but involved significantly different facts. In N(K), entries in the secret diary of L were relied upon by the defence in support of N's complete denial of sexual behaviour towards her, with the object of showing that she was a fantasist. In consequence, four entries were before the jury including one in which L wrote that she had “slept with [N]”, in exchange for money: this was said to undermine her evidence that the intercourse she alleged was accompanied by force on his part and resistance on hers.
On the question whether section 119 applied so as to make L's inconsistent statement, once admitted, evidence of the truth of any fact stated, the Court of Appeal concluded that the private nature of the diary meant that L's purpose in making the statements was not to “cause another to believe the matter” under section 115(3). As the definition of “matter stated” also extends to section 119 (which makes a witness's previous inconsistent statement admissible as evidence of any matter stated) that section could not be relied upon to render this particular variety of inconsistent statement admissible. The court went on to avoid the absurdity of ruling that an inconsistent statement in a diary was of less value than a statement shouted from the rooftops by holding that, as the diary was not hearsay, its contents were freely admissible at common law as a “fact” from which an inference could be drawn that intercourse had taken place – “in exactly the same way as if she had been observed by other people kissing him”.
This is (with respect) a dangerous road to tread. Statements do not cease to be statements merely because they are not evidence of matters stated. Section 115 quite properly separates the concepts: a statement, for the purposes of hearsay-related provisions, including section 119, is “any representation of fact or opinion, made by a person by whatever means”.Footnote 73 The account in L's diary was thus still a statement: a “representation” of her thoughts that could be true or false. Although section 115(3) appears to exempt a statement made to oneself from the operation of the hearsay rule, common sense tells us that it may also be the product of mendacity in the form of fantasy, as teenage diaries form a plausible exception to Falkenor's rule that an individual “does not lie to himself”.Footnote 74 Furthermore, it is precisely because of certain functions that it may perform as a statement that such evidence is traditionally rendered admissible, in a case like N(K), to be set alongside the testimony of L for the purpose of highlighting inconsistencies. If the statement then functions as evidence of its truth, contrary to the position before the Act, this is only because the hearsay rule at common law no longer exists to prevent this happening.
If the statement is miscast as a fact, however, then in a case such as Knight the consequence is the disapplication of the rule against narrative as well as the rule against hearsay. Seen as a statement, the diary entry in Knight would probably have been admissible for the prosecution by way of exception to the narrative rule only if of value as a “complaint”. It might also, of course, go to rebut recent fabrication, but this would be contingent on such an attack being launched by the defence.
Could such an entry qualify as a complaint? The common law authorities were confused. In one case, a note written by the complainant describing abuse, and given to a friend in mistake for another note, was somewhat dubiously regarded as a complaint, despite the lack of any apparent intention to communicate the contents to another person,Footnote 75 but in another, a letter written by a complainant to herself as a form of therapy, describing abuse suffered at the hands of the accused, was held to have been inadmissible and its wrongful admission to support the complainant's testimony was sufficient to render the conviction unsafe where the case depended entirely on the credibility of the two principal witnesses.Footnote 76 Thus, the best that a secret diary entry could have hoped to achieve at common law (assuming no defence of recent fabrication was advanced) was to secure a very dubious toehold as a complaint in respect of an alleged sexual offence.
The Act retains the concept of the rule against narrative as a distinct exclusionary rule, subject to exceptions. In section 120(4) and (7) it provides a vehicle whereby complaints, once admissible as evidence of consistency, may now also be received as evidence of their truth. This would however be of no assistance unless an entry in a secret diary can properly classed as a “complaint” in the first place, and despite the broadening of the common law doctrine in section 120(7),Footnote 77 that seems very doubtful. By the simple trick of denying its status as a statement, however, Knight contrives to make such an entry admissible as circumstantial evidence and, presumably, capable of fulfilling either function.Footnote 78
3. Confessions in secret diaries?
What then would happen in a “secret diary” case where the diary is that of the accused, and contains an admission to the offence charged? Is the law of confessions thereby engaged, and if so, with what result?
The answer, it is submitted, is that the law of confessions is indeed engaged, because the diary entry falls within the definition of a confession in PACE 1984, s. 82(1), by which “confession” includes “any statement wholly or partly adverse to the person who made it, whether made to a person in authority or not and whether made in words or otherwise”. The entry would therefore be admissible against the accused, unless (somewhat improbably) it falls foul of section 76(2) of that Act.
This involves treating the entry as a “statement” for the purposes of section 82(1) when according to N(K) and Knight it is not a statement at all; but the law of confessions has never partaken directly of the statutory components of the hearsay definition. In PACE itself, the draftsman declined to import into section 82(1) the definition of a “statement” provided for the documentary hearsay exceptions originally provided in section 68 of that Act, so this dual approach is defensible. It also avoids the question of whether there needs to be a “matter stated” within the meaning of Criminal Justice Act 2003, s. 115(3), but again the argument could be that the “confessions” provisions of the 2003 Act, unlike those dealing with previous consistent and inconsistent statements, do not import that concept. Thus section 118, in preserving aspects of the common law of confessions and admissions, refers to “any rule of law relating to” confessions etc, rather than any rule by which a confession is evidence of a matter stated. And section 128(1), which inserts section 76A into PACE to govern the use of confessions by or on behalf of a co-accused, adopts the same form of words as section 76 in specifying that a confession may be given in evidence “in so far as it is relevant to any matter in issue”. Of course, the purpose of proving any confession is to establish the truth of its contents,Footnote 79 but this does not require it to become embroiled in the convoluted constructs of section 115(3).
A potential snag with this reading arises if one accepts that the law of confessions operates only as an exception to the hearsay rule – forming what Tapper calls “the single largest exception” to that rule. As a general observation this is of course true – most confessions are hearsay, being tendered to prove the truth of a matter stated, and as such attract the relevant provisions of the Act. Thus, a confession made to a person whose evidence is received in his absence under section 116 ought to attract the provisions of section 121 (multiple hearsay).Footnote 80 But if the true rule is that only a confession which passes the acid test for hearsay can progress to be considered under section 76, the incriminating diary fails at the first hurdle – under N(K) and Knight the entry is admissible as a “relevant fact”, without reference to whether it satisfies the conditions in section 76(2). It therefore falls to be grouped with other aspects of incriminating conduct which are generally regarded as non-confession evidence and subject to exclusion, if at all, by operation of PACE 1984, s. 78.
The reason that most commentators are happy to see these other types of evidence excluded from the clutches of section 76 is either that it is hard to see the accused as “making” an adverse statement (where the conduct consists, e.g., of bursting into tears or running away) or that, to the extent that he has made a statement, it is not properly speaking “adverse” when made (as in the telling of lies, or the more difficult case of the ex facie exculpatory statement which ends up counting against the accused once other evidence comes to light).Footnote 81 With the diary entry, common sense suggests that an adverse statement has been made, the value of which will depend upon the extent to which it is the product of the accused's free will.
Even if it is accepted that N(K) and Knight have wrongly mis-cast statements as facts, it remains the case that a diary confession is not “a matter stated” to which the hearsay rules now apply, so that it is inconvenient to regard the confessions rules as being applicable only to a true hearsay statement. A solution was advanced by Mirfield long before the present controversy arose, in the course of an argument that the law on confessions needed to be insulated from the uncertainty at common law as to the hearsay status of implied assertions. He proposed that the “principle or principles lying behind the core examples of confessions have developed a force of their own in that context and should be allowed also to govern penumbral cases”.Footnote 82 Although the case of the diary entry is now “penumbral” not so much because of uncertainly but because of the unforeseen consequence of section 115(3), the argument that it should not lead to the disapplication of the rules intended to prevent reliance on incriminating admissions obtained in breach of the proper processes of interrogation seems a strong one.
This would appear to have been the assumption on which evidence was admitted by the experienced trial judge in R. v. Henton,Footnote 83 where H was covertly recorded “confessing” to his pet cats that he had murdered his partner, and it does not appear to have been suggested that anything other than a routine application of PACE, s. 76, was called for in that case.
It is unlikely that there will be many cases in which self-incriminating words that were not intended to be seen or heard by anyone else could, at the same time, be susceptible to challenge under section 76(2), but such cases are not impossible to envisage. D may for example write something incriminating in his private diary after being plied with hallucinogenic drugs. Suppose instead that D's alleged victim covertly tapes a conversation between them while using a concealed recording device, but uses oppression to elicit an admission. Is section 76 applicable if D then blurts out something he assumes E already knows (e.g., “OK, so I raped you, but we were both drunk”)? This cannot be hearsay under section 115(3), but arguably it is still a confession.Footnote 84
Does any of this really matter, given that unfairly obtained prosecution evidence can in any event be excluded under PACE, s. 78? According to Lord Steyn in Hasan,Footnote 85 sections 76 and 78 between them provide a “coherent and comprehensive way for the just disposal of all decisions about statements made by accused persons to the police.” There is no need, on that view, to strain the concept of a confession beyond its natural meaning in order to protect defendants who are bullied or tricked into incriminating themselves in other ways.
This is true up to a point; but if such utterances are not confessions, then the same must be said where it is a co-accused who wishes to adduce the evidence: a situation beyond the reach of section 78, but which would have been governed by section 76A in the case of a confession. To that extent, the distinction clearly does still matter.
4. Self-serving entries in a secret diary?
The problems caused by section 115(3) and Knight do not stop there. If Knight is correct, the accused must be entitled to adduce entirely self-serving entries from his (or from a co-accused's) secret diary, provided that those entries do not appear to have been written with any such use in mind. If, following Knight, such an entry is deemed not to be a statement, but a fact, neither the rule against hearsay nor the rule against narrative can be invoked to exclude it.Footnote 86
V. Conclusion
Whatever other virtues it may possess, the new statutory regime is undermined by its failure to provide a clear and precise definition of the basic concept of hearsay, and by its failure properly to integrate that concept with the existing concepts of narrative and confessions. Those failings have been exacerbated by some unfortunate case law, in which the judges appear to have been led astray by the confusing signposts erected by the Act itself.
Some at least of the erroneous precedents may in time be overruled or isolated as the courts become more familiar with the legislation. But fundamental weaknesses in the Act may be harder to put right. In particular, it seems that the spectre of Kearley has not, after all, been exorcised. In attempting to exorcise it, section 115(3) appears to have succeeded only in creating a new and different spectre – that of a “statement” that contains no “matter stated” for the purposes of the Act, even though a court or jury may be invited to rely on it as true. It is not surprising that the courts have struggled with that concept. As things stand, one can only predict more trouble ahead.
