Pollitt v The Queen

If, on the other hand, the Berrys gave evidence of statements made on distinct occasions, Mrs. Berry's statement was admissible and Mr. Berry's was not because it was not related in any way to the telephone conversation and infringed the hearsay rule. Consequently, Mr. Berry's evidence of the statement should not have been placed before the jury. The reception of the statement in evidence gives rise to a problem in that the jury may have regarded it as confirming the evidence given by Mrs. Berry. It was, of course, evidence which the applicant could not rebut or qualify otherwise than by reliance on his general defence that he was not the killer.

The question then is whether the trial judge's directions were sufficient to prevent any miscarriage of justice. He gave clear and positive directions that the evidence of both Mr. and Mrs. Berry could not be used against the applicant and that it could only be used to establish Allen's state of mind and that he was a party to an agreement to kill Williams. Having regard to the strength and clarity of the directions in relation to use of the evidence against the applicant, I do not consider that there was a miscarriage of justice arising from the reception of Mr. Berry's evidence.

Unfortunately, the trial judge's directions to the jury were deficient on two counts. First, he failed to instruct the jury as to the use they could make of Allen's statement in the presence of Mrs. Berry, if they accepted her evidence, for the purpose of identifying the applicant as the other party to the telephone conversation. Secondly, the judge erroneously informed the jury that they could use the Berrys' testimony as evidence of Allen's intention.

The first error entailed no disadvantage or prejudice to the applicant. On the other hand, the erroneous direction that the Berrys' testimony could be used as evidence of Allen's intention cannot be disposed of quite so easily. That is because it is difficult to know what the jury would have made of the direction when Allen's intention was not a relevant fact, nor was it relevant to a fact in issue. As the trial judge related the matter of Allen's intention to his being a party to an agreement to kill Williams, a fact which was an element in the Crown case, it seems clear that, in conformity with the instruction, the jury would have regarded the evidence as going to that issue and not treated it as evidence against the applicant. I acknowledge that the Berrys' evidence of Allen's statement that the applicant had made a mistake in doing a job for Allen and that he wanted to be paid was inadmissible and was very prejudicial to the applicant but, in the circumstances, as I have outlined them, I do not consider that the jury would have used that evidence for an improper purpose. There was, in any event, strong evidence identifying the applicant with the mistaken killing of Simpson, that evidence being independent of the Berrys' testimony. And there was evidence from the witness Jones that the applicant said that he was doing the killing for Allen and later that he had shot the wrong person, as well as other evidence linking Allen and the applicant in a discussion about disposing of Williams, the intended victim. In the result, it does not seem to me that there was any significant possibility that the jury would have used the evidence for an impermissible purpose when they were clearly and positively instructed not to do so.

I would grant special leave to appeal and dismiss the appeal.

The prosecution's case against Roy Anthony Pollitt, the appellant, was that he had shot and killed Lindsay Simpson by mistake, in intended performance of a contract he had made with one Dennis Allen that he, Pollitt, would murder Alan David Williams. There was no doubt about the fact that Simpson had been murdered. He had been shot by an assassin. On the appellant's trial before the Supreme Court of Victoria on a count of murder of Simpson, the issue was: was Pollitt the murderer? Proof that Pollitt had made a contract with Allen for the murder of Williams would tend to establish that Pollitt had both a motive to commit murder and an intention to do so, and evidence that Pollitt had a motive to murder Williams, coupled with evidence that the murderer had mistaken Simpson for Williams, was admissible as circumstantial evidence tending to prove that Pollitt was the murderer of Simpson [12] . The evidence tendered to prove that Pollitt had made a contract with Allen for the murder of Williams had to be admissible, of course, against Pollitt. The chief evidence of a contract between Pollitt and Allen was given by an accomplice, Jones, whose evidence if accepted established that Pollitt had agreed with Allen to murder Williams, that Pollitt expected a large sum of money in payment, that Pollitt obtained a gun and instructions during a visit to Allen's premises, that Pollitt was aware on the morning after the shooting that he had shot the wrong man and that Pollitt then rang Allen.

1. Plomp v. The Queen (1963), 110 C.L.R. 234, at pp. 247-250.

Allen died before the trial commenced. The prosecution, to assist in proving the alleged contract, tendered evidence given by Mr. and Mrs. Berry of conversations one or other of them had had with Allen on the days following Simpson's murder. Mrs. Berry gave evidence that, on the night of Simpson's murder, Allen had told the Berrys to watch the news. She did so and, when she saw Allen next day, he mentioned the "mistaken identity" item. She said that Allen was furious. She said that, at about 11 o'clock in the morning of that day, the phone rang and Allen answered it. Mrs. Berry's evidence continued:

Tell us what you can remember? - He said, "You get the rest of the money when you do the job properly." After the phone call, which lasted "maybe a couple of minutes", Allen was furious, "ranting and raving", and said to Mrs. Berry: I have already given him 5,000, he expects me to give him another five for something he hasn't done. Later she gave this evidence: going back to that telephone call, the telephone call where you say Dennis Allen became angry, now when he got off the phone, did he say who he had been speaking to? - Yes. Who did he say it was? - He said it was Ray [t]hat he had been speaking to. Mr. Berry gave evidence that, on the day following Simpson's murder, Allen was "jumpy and in a bad mood" and, when Allen was asked what was wrong, "he said that someone buggered up something they were supposed to do". Some days later, Berry was present at a conversation where Allen said that "someone wanted $5,000 and didn't do the job right". Later, according to Berry, Allen - said Roy got the wrong one and he wants to be paid for it - He wants his money, the rest of his money. Mr. Berry's evidence continued:

Did he say anything in that conversation about his attitude to the request that he be paid the extra money? - Yes, he said he wouldn't get it.

These pieces of evidence were admitted over a defence objection that the evidence was hearsay and in acceptance of a prosecution submission that Mrs. Berry's evidence established "Allen's state of mind at the time he made the statements [to the Berrys] and demonstrates that he was a party to the agreement to kill Williams, a fact which will be very much an issue in the trial". The learned trial judge thought that Mrs. Berry's evidence was admissible "to demonstrate the relationship which existed between the accused and Allen at the time these events took place". Mr. Berry's evidence was admitted "as demonstrating Allen's then state of mind and that he was party to the agreement to kill Williams".

The first condition of admissibility of evidence is relevance: apart from questions relating to the credit of a witness, a fact which evidence is tendered to prove (a "fact to be proved") must be a fact in issue or a fact relevant to a fact in issue. Where a fact to be proved is a fact in issue, admissibility of evidence tendered to prove it depends solely on the manner in which that evidence tends to establish the fact to be proved. Where a fact to be proved is a fact relevant to a fact in issue, admissibility depends first on the manner in which that evidence tends to establish the fact to be proved and, secondly, on the relevance of the fact to be proved to a fact in issue. I respectfully agree with Lord Oliver of Aylmerton [13] that "relevant" cannot be better defined than by the definition in Art. 1 of Stephen's Digest of the Law of Evidence [14] :

any two facts to which it is applied are so related to each other that according to the common course of events one either taken by itself or in connexion with other facts proves or renders probable the past, present, or future existence or non-existence of the other. Evidence of an out-of-court statement may tend to establish a fact to be proved either because that fact is asserted by the maker of the statement and the maker of the statement is to be believed to have made the statement truthfully - in which case the evidence is hearsay - or because the making of the statement is itself the fact to be proved or, when taken together with other facts established by evidence or common experience, tends to establish the fact to be proved - in which case the evidence is original evidence. The cogency of hearsay evidence depends on the credit to be attributed to the person making the statement in the circumstances in which it was made and on the credit of the witness who relates it in court. The cogency of original evidence depends solely on the credit of the witness who relates it in court. The distinction between hearsay and original evidence was stated by Lord Wilberforce in Ratten v. The Queen [15] in these terms: The mere fact that evidence of a witness includes evidence as to words spoken by another person who is not called, is no objection to its admissibility. Words spoken are facts just as much as any other action by a human being. If the speaking of the words is a relevant fact, a witness may give evidence that they were spoken. A question of hearsay only arises when the words spoken are relied on "testimonially," i.e., as establishing some fact narrated by the words. Authority is hardly needed for this proposition, but their Lordships will restate what was said in the judgment of the Board in Subramaniam v. Public Prosecutor [16] : "Evidence of a statement made to a witness by a person who is not himself called as a witness may or may not be hearsay. It is hearsay and inadmissible when the object of the evidence is to establish the truth of what is contained in the statement. It is not hearsay and is admissible when it is proposed to establish by the evidence, not the truth of the statement, but the fact that it was made." 1. In Reg. v. Kearley, [1992] 2 A.C. 228, at pp. 263-264. 2. 12th ed. (1936), p. 4. 3. [1972] A.C. 378, at p. 387. 4. [1956] 1 W.L.R. 965, at p. 970.

The classification of evidence according to this distinction, though clear enough in principle, is frequently obscure in practice. Yet the classification of an out-of-court statement either as hearsay or as original evidence is necessary to determine whether the general rule against the admission of hearsay evidence [17] (or some specific exception to that rule) applies or whether the general rule in favour of the admission of original evidence (or some exception to that rule) applies. To make the required distinction, it is necessary to ascertain the manner by which the evidence tends to prove the fact to be proved. The distinction may be extremely fine between a case where the existence of a fact to be proved is to be implied in (though not expressly asserted by) what the maker of the statement said and a case where the existence of a fact to be proved is to be inferred from the fact that the statement was made in the circumstances in which it was made. A distinction so fine that it turns on the difference between an implication from what the maker of a statement has said and an inference of fact drawn from the making of a statement in the surrounding circumstances is likely to produce uncertainty in classification. The distinction calls for a precise identification of the probative relationship between the statement and the fact to be proved, that is, the manner in which the tribunal of fact is invited to reason from the statement to the fact to be proved.

1. Myers v. Director of Public Prosecutions, [1965] A.C. 1001, at p. 1020.

As a fact to be proved may be either a fact in issue or a fact relevant to a fact in issue, and as the evidence tendered to prove a fact to be proved may be either original evidence or hearsay evidence, there are four categories of evidence for consideration: original evidence of a fact in issue, original evidence of a fact relevant to a fact in issue, hearsay evidence of a fact in issue and hearsay evidence of a fact relevant to a fact in issue. Unless there is some ground for excluding original evidence in a particular case (e.g., undue prejudice to an accused), original evidence is admissible; unless hearsay evidence falls into an exception (e.g., as part of the res gestae), it is excluded. The reasons why the law chose to exclude hearsay evidence are uncertain [18] , but at the heart of the objection to hearsay is its dependence on the credibility of a statement that is "uttered not under cross-examination" [19] . As the Privy Council pointed out in Lejzor Teper v. The Queen [20] :

The truthfulness and accuracy of the person whose words are spoken to by another witness cannot be tested by cross-examination, and the light which his demeanour would throw on his testimony is lost. The justification, if justification be needed, of the rule against the admission of hearsay is that a party - especially an accused in a criminal case - is not to be faced with evidence of an assertion made in circumstances in which the credibility of the assertion cannot be adequately tested. Whatever its origin and whatever its justification, the rule against hearsay is "fundamental" [20] , though it is qualified by numerous exceptions. 1. [1965] A.C., at pp. 1020-1021. 2. Wigmore's phrase: Wigmore on Evidence, vol. 6 (Chadbourn rev. 1976), par. 1788. 3. [1952] A.C. 480, at p. 486. 4. [1952] A.C. 480, at p. 486.

It would be possible, no doubt, to redesign the law of evidence with the assistance of psychologists and philosophers versed in epistemology but it cannot be done piecemeal or by extensions which are not linked to an underlying principle. To attempt such alterations, however pragmatically desirable in one case, will throw the next case into confusion. The law of evidence, though adjectival, is the working tool which a trial judge must keep constantly at hand and the principles of the law of evidence are, so to speak, the ground on which the dynamics of a trial, especially a criminal trial, are played out. There is an attraction in the notion that the admissibility of hearsay should be governed by a judicial assessment of its reliability, but there are countervailing arguments. Thayer [21] pointed to some of the unsatisfactory implications of that approach:

No doubt, in point of reason, hearsay statements often derive much credit from the circumstances under which they are made; and it would in reason have been quite possible to shape our law in the form that hearsay was admissible, as secondary evidence, whenever the circumstances of the case alone were enough to entitle it to credit, irrespective of any credit reposed in the speaker. This point of view is forever suggesting itself in that part of the subject relating to declarations which are a part of some admissible fact, - of the res gesta, as the phrase is. These are often spoken of as parts of a mass of circumstantial facts described as res gestae, all evidential, supporting and supported by each other in their tendency to prove some principal fact; instead of being regarded, as they should be, as parts of that fact itself, pars rei gestae, lying under the curse of hearsay, but received, by way of exception, on account of this special intimacy of connexion with the admissible fact. This part of the subject presents an instructive spectacle of confusion, resulting from the desire, on the one hand, to hold to the just historical theory of our cases; and, on the other, to resort to first principles, without being aware of the size and complexity of the task which is thus unconsciously entered upon. Accepting, for want of a practical alternative, the general principle that hearsay evidence is inadmissible, mention should be made of the two categories of original evidence identified above. 1. A Preliminary Treatise on Evidence at the Common Law (1898), p. 523.

The clearest instance of an out-of-court statement which is admissible as original evidence is a case where the making of the statement is itself a fact in issue. There is no need to examine either the probative relationship between the evidence tendered to prove the making of the statement and the fact to be proved (because the making of the statement is itself the fact to be proved) or the relevance of the fact to be proved to the fact in issue (because it is itself the fact in issue). Evidence of the making of the statement is not tendered to prove the truth of the facts asserted by the maker. For example, evidence of the uttering of seditious words by an accused charged with sedition must be admitted in proof of an element of the offence and evidence of the making of defamatory statements alleged to have been made by a defendant in a defamation action must be admitted in proof of the publication of defamatory matter.

The second category consists in the cases where the making of the statement is a ground for inferring the fact to be proved. For example, in Subramaniam where a person, accused in Malaya of unlawfully being in possession of ammunition, pleaded duress as an excuse, evidence of threats allegedly made to the accused by terrorists was held admissible to found an inference as to the accused's state of mind. The inference was to be drawn from the uttering of the threats and the circumstances in which they were uttered. Again, admission of evidence of this class does not offend the rule against hearsay, for the evidence is not tendered to prove a fact asserted by the maker of the statement. It is admissible on the footing that the fact that the particular statement was made, coupled with any relevant circumstances and considered in the light of common experience, warrants an inference that the fact to be proved exists.

In considering evidence of this class, the relevance of the making of the statement to the fact to be proved and the relevance of the fact to be proved to the fact in issue are of critical importance. Although relevance for the purpose of determining admissibility falls for decision by the trial judge, it is ascertained according to common experience [22] . Judicial minds have sometimes differed as to the capacity of evidence of the making of a statement to found an inference of a particular fact and have sometimes differed as to the relevance of the fact to be proved to a fact in issue. Statements made by intending customers of a person charged with the carrying on of an unlawful activity provide a good example. In Australia and New Zealand, on a charge of keeping a common gaming house, evidence that a number of people attend or telephone the premises seeking to place bets has been held to be admissible to prove that a betting business is carried on on the premises [23] . In Reg. v. Firman [24] , a Full Court of the Supreme Court of South Australia held that evidence that a number of people telephoned premises seeking to buy drugs from a named person who occupied those premises was admissible on a charge against that person of possessing heroin for sale. The Court regarded the making of the phone calls as relevant to the existence and nature of the business or activity carried on on the premises and the identity of the person carrying it on [25] . Although the callers in these cases might be thought to have known that a person at the premises was taking bets or selling heroin, as the case may be, evidence of what the callers said was not treated as implying an assertion of that fact. Had evidence of the callers' statements been understood merely as assertions of facts known to the callers, the evidence would have been classified as hearsay and held to be inadmissible in accordance with the observation of Dixon J. in Martin v. Osborne [26] :

The application of any general statement about relevancy is subject to the well-known specific rules of exclusion. For instance, the rule against hearsay and the warning implied in the descriptive phrase res inter alios acta lead to the exclusion of evidence not only of what a stranger to the cause has said but also of what he has done, if it is offered to prove his knowledge of some fact and thus the existence of that fact, notwithstanding that the fact itself be relevant and its proof by another medium be receivable. (Emphasis added.) In the cases mentioned, the evidence of the call was not treated as proof of the knowledge of the callers that bets were taken or heroin was sold at the premises, but as original evidence from which, in totality, the existence of a business of taking bets or selling heroin, as the case may be, could be inferred. The existence of a business was either a fact in issue or was relevant to a fact in issue. 1. Martin v. Osborne (1936), 55 C.L.R. 367, at p. 375, per Dixon J. 2. Davidson v. Quirke, [1923] N.Z.L.R. 552; Lenthall v. Mitchell, [1933] S.A.S.R. 231; McGregor v. Stokes, [1952] V.L.R. 347; Marshall v. Watt, Struthers, and County, [1953] Tas. S.R. 1; Gorman v. Newton; Ex parte Newton, [1958] Qd R. 169; Police v. Machirus, [1977] 1 N.Z.L.R. 288. 3. (1989) 52 S.A.S.R. 391. See also Morgan v. Texas (1980), Tex. Civ. App. 596 S.W. 2d 220. 4. (1989) 52 S.A.S.R., at pp. 394, 395, 397-398. 5. (1936) 55 C.L.R., at pp. 375-376.

A contrary view of the evidence of calls was formed by a majority of the House of Lords in Reg. v. Kearley. In that case, callers had offered to buy drugs from the appellant, identifying him by his nickname "Chippie". The calls were made in a period of several hours after the police had conducted a search for drugs in the appellant's house. The reasoning of the majority (Lord Bridge, Lord Ackner and Lord Oliver) contained the following steps: (1) the mere act of calling was meaningless and irrelevant apart from the significance attaching to the words used by the caller; (2) the words used by a caller contained or might be thought to have contained an implied assertion that the appellant was a supplier of drugs; (3) implied assertions in out-of-court statements offend the rule against hearsay; (4) though the words used by a caller showed the caller's state of mind, that state of mind was irrelevant to any fact in issue; (5) the inadmissibility of evidence of a single call is unaffected by proof that there were multiple calls to the same effect. Lord Griffiths, in dissent, regarded proof that the customers believed that they could obtain drugs from the appellant as founding an inference "that the appellant had established a market as a drug dealer by supplying or offering to supply drugs and was thus attracting customers" [27] . The fact that there were other possible explanations meant that it was for the jury to decide which inference they chose to draw. Lord Browne-Wilkinson, also in dissent, regarded evidence of the calls as tending to prove that there was a market for drugs that the appellant might satisfy and that that fact was relevant to the occupant's purpose in having possession of the drugs [28] . The points of departure between the majority in Kearley and the minority in that case and between that majority and the Australian and New Zealand cases earlier cited are to be found in the differing opinions as to the relevance of the fact that callers had sought to buy drugs or place bets to the question of the nature of the activity being carried on by the person charged - unless the statements were treated as hearsay assertions by the callers of the nature of that activity and, on that account, inadmissible. Although the question of admissibility in each case was a question of law, the answer to the question turned on the court's appreciation of the probative effect of the making of the calls, not of the truth of their content. Subject to one reservation, there was no divergence of opinion as to the inadmissibility of the contents of those statements to prove the truth of what the callers had impliedly asserted. The reservation relates to the belief of a caller that a particular activity was being carried on and the use of that belief - as distinct from the making of the statement which conveyed the belief - to found the relevant inference.

1. [1992] 2 A.C., at pp. 237-238. 2. [1992] 2 A.C., at pp. 278-279.

If it were permissible to infer the truth of a statement from the fact of the maker's belief in what he stated, the making of the statement could be used as evidence of the belief, and from that belief the truth of what was stated could be inferred. But, as Dixon J. said in Martin v. Osborne [29] , it is not permissible to use knowledge of a fact to prove the truth of what is known: to use the making of the statement in that way is to use evidence of knowledge as hearsay proof of what is known.

It follows that in those cases where evidence of an out-of-court statement can be used to prove the truth of what is stated - for example, evidence admitted as part of the res gestae - the evidence is not admitted as original evidence but under an exception to the hearsay rule. Subject to those exceptions, evidence of an out-of-court statement tendered to prove the truth of some fact asserted by the maker of the statement, expressly or impliedly, is prima facie inadmissible. But where the making of a statement in particular terms is capable, in conjunction with evidence of other facts and considered in the light of common experience, of founding an inference that a relevant fact exists, evidence of the making of the statement is admissible as original evidence of the relevant fact.

The most obvious exception to the hearsay rule is an admission by a party against the party's interests (in criminal cases, a confession by the person charged). That exception has no relevance to this case. Another possible exception to the hearsay rule - it is often stated as an exception, though it may be no more than an inference from the words uttered - is the proposition that an out-of-court statement is admissible to prove the contemporaneous state of mind of the maker of the statement. Such evidence is admissible, but only when that state of mind is itself a fact in issue or a fact relevant to a fact in issue [30] . Thus, in Reg. v. Blastland [31] , a person other than the accused was heard to make a statement indicating his knowledge at that time of the murder with which the accused was charged. The accused sought to tender evidence of a witness who heard the statement to prove that the maker of the statement, to have known of the murder at that time, must have been implicated in the murder. In the absence of evidence showing how that person came by knowledge of the murder, the fact of his knowledge was irrelevant to the issues. The statement was held inadmissible. If that person's statement had been treated as implying an admission by him of implication in murder, it would have been classified as hearsay and evidence of the statement would have been rejected on that ground.

1. Walton v. The Queen (1989), 166 C.L.R. 283, at pp. 288, 302, 307. 2. [1986] A.C. 41; cf. Reg. v. Szach (1980), 23 S.A.S.R. 504.

However, where a statement is tendered to prove the maker's knowledge of a fact and that knowledge is used to found an inference of fact other than the truth of the fact known, the statement is not classified as hearsay but as original evidence. Dixon J.'s restriction on the use of an out-of-court statement revealing the maker's knowledge of a fact has no application except where the statement is tendered to prove the fact known. The distinction between use of a statement revealing the maker's knowledge of a fact in order to found an inference that the fact known is true (a testimonial use of the statement) and use of such a statement in order to found an inference that some other fact exists (a circumstantial use) is discussed by Wigmore [32] . And the joint judgment of Wilson, Dawson and Toohey JJ. in Walton v. The Queen [33] , though recognizing the inadmissibility of evidence of a statement of knowledge which is tendered solely to prove the truth of the fact known, foresaw that a statement revealing knowledge of the maker of the statement might be used circumstantially and admitted in evidence for that purpose. Their Honours said:

It may be true in some cases to say that statements made by a person indicating his state of mind involve no element of hearsay. But in other cases a person's statements about his state of mind will only have probative value if they are truthful and accurate and to rely upon them is to rely to some extent upon the truth of any assertion or implied assertion contained in them. To that extent an element of hearsay may be said to be present. But the element of hearsay need not necessarily preclude evidence of that kind being treated as conduct from which an inference can be drawn rather than as an assertion which is put forward to prove the truth of the facts asserted. The distinction between the two approaches is one which can be fine, but it is one which in principle ought to be drawn. The distinction depends, as I have said, on the identification of the probative relationship between the particular statement and the fact to be proved. The distinction depends on the difference between proof by reliance on the truth of an assertion, whether express or implied, and proof by inference from all material facts which include the making of the statement. 1. op. cit., par. 1790. 2. (1989) 166 C.L.R., at pp. 302-303.

In Reg. v. Benz [34] Dawson J. held evidence of a statement that contained an assertion of fact to be admissible as original evidence tending to prove the fact when considered in conjunction with the circumstances in which the statement was made. The charge was murder. A man had been stabbed and his body, not then dead, had recently been pushed over the side of a bridge by two women in the early hours of a morning when a witness came upon the scene and asked if everything was allright. One of the women replied that it was O.K., "her mother was just feeling sick". The wife of the deceased man and her daughter were charged with his murder and the question was whether the statement of the woman on the bridge was admissible in proof of the identity of the accused as the murderers. Dawson J. held the statement to be original evidence of the fact that "the younger woman was acting in accordance with the relationship known to exist between the respondents" [35] and, together with other evidence, founded an inference that the two women on the bridge were the accused. The hearsay rule, in his Honour's opinion, had no application.

1. (1989) 168 C.L.R. 110, at pp. 134, 135. 2. ibid., at p. 134.

A further exception to the hearsay rule admits evidence of certain statements made in the course of, or approximately contemporaneously with, a transaction that is the subject of the court's inquiry: the res gestae exception. The res gestae exception has been considered in this Court in a number of cases in which the conditions of admissibility have been examined. In Brown v. The King [36] , where the question was the admissibility of a statement made by a witness to a mortally injured victim of a shooting as the victim was leaving the scene, Isaacs and Powers JJ. said this:

The incident offered in evidence was unconnected in causality with the shooting: if it had been so connected - as by flight to escape its continuance - the slight lapse of time and the mere fact of 25 yards distance would not have been sufficient in themselves to have destroyed the natural nexus.

But when there is no natural connexion by continuance - which may have liberal connotation - and there is a distinct and appreciable break of time and place, it would in our opinion be going beyond the limit of authority to admit evidence, which is in substance and reality a mere narration respecting a concluded event, a narration not naturally or spontaneously emanating from or growing out of the main transaction, but arising as an independent and additional transaction.

Their Honours' insistence on a natural connexion between the statement and the event was followed by Dixon J. in Adelaide Chemical & Fertilizer Co. Ltd. v. Carlyle [37] . His Honour noted the view then current among English lawyers that res gestae evidence was receivable only as original evidence and, on that account, was admissible "only as one of the parts or details of a transaction not complete when the statements were uttered and as supplying no proof of antecedent facts" [38] . The American view, on the other hand, treated the evidence as an exception to the hearsay rule "on the ground that a guarantee of their truth is to be found in their spontaneity, in the lack of "time to devise or contrive" and in the instinctive character of utterances made under the influence of excitement" [38] . The American view has now prevailed. In Ratten , Lord Wilberforce recognized the admissibility of hearsay in statements admitted under the res gestae exception when he rejected as a test of admissibility "the uncertain one whether the making of the statement was in some sense part of the event or transaction" [39] . After review of authorities (including the Australian cases), his Lordship said [40] : These authorities show that there is ample support for the principle that hearsay evidence may be admitted if the statement providing it is made in such conditions (always being those of approximate but not exact contemporaneity) of involvement or pressure as to exclude the possibility of concoction or distortion to the advantage of the maker or the disadvantage of the accused. Where the statement was made after the event (presumably his Lordship meant at a time approximately but not exactly contemporaneous with the event) [41] : it must be for the judge, by preliminary ruling, to satisfy himself that the statement was so clearly made in circumstances of spontaneity or involvement in the event that the possibility of concoction can be disregarded. In Vocisano v. Vocisano [42] Barwick C.J., speaking with the concurrence of the other members of the Court, reserved final consideration of the effect to be given to Ratten . His Honour added: A reason for the doctrine that statements made as part of the res are admissible as evidence is that, because of their contemporaneity and the circumstances of their making, they were unlikely to be concocted and therefore might well be reliable: but that does not mean that statements made on an occasion when they are unlikely to be concocted are for that reason admissible. It is the contemporaneous involvement of the speaker at the time the statement is made with the occurrence which is identified as the res which founds admissibility. 1. (1913) 17 C.L.R. 570, at p. 597. 2. (1940) 64 C.L.R. 514, at p. 533. 3. ibid., at p. 531. 4. ibid., at p. 531. 5. [1972] A.C., at p. 389. 6. ibid., at p. 391. 7. [1972] A.C., at p. 389. 8. (1974) 130 C.L.R. 267, at p. 273.

Once it is accepted that the res gestae principle represents an exception to the hearsay rule and admits statements which may be used to prove the truth of the facts asserted therein, it is understandable that admissibility should be made to depend, inter alia, on the judge's satisfaction that the conditions in which the statement was made were such as "to exclude possibility of concoction or distortion". But, as Barwick C.J. pointed out, non constat that any hearsay statement is admissible if the judge is so satisfied. The statement must be made in conditions "of approximate [if] not exact contemporaneity" and the impossibility of concoction or distortion must arise from the "spontaneity or involvement in the event" by the maker of the statement. In Walton [43] these propositions were stated in the joint judgment in these terms:

An assertion may be admitted to prove the facts asserted if it is part of the res gestae, but it is then an exception to the rule against hearsay: see Adelaide Chemical & Fertilizer Co. Ltd. v. Carlyle. The justification for that exception is now said to lie in the spontaneity or contemporaneity of assertions forming part of the res gestae which tends to exclude the possibility of concoction or distortion: Ratten [44] ; Reg. v. Andrews [45] ; see also Adelaide Chemical & Fertilizer Co. Ltd. v. Carlyle [46] . Of course, the discussion in Ratten and Andrews was in the context of the res gestae rule. The unlikelihood of concoction or distortion is not sufficient of itself to render a hearsay statement admissible: see Vocisano v. Vocisano [47] . In Walton evidence that, in answer to a telephone call, a child had said "Hello, daddy" was held inadmissible to prove the identity of the caller [48] . The statement was probative of the caller's identity only if credit were given to the child's implied assertion that the caller was, in fact, "daddy". The Chief Justice's dissenting view that the statement was admissible though it contained hearsay was based on his Honour's appreciation of "the extreme unlikelihood of concoction on the part of the child" [49] . His Honour's approach was not expressed to be limited by the condition of approximate if not exact contemporaneity with the res gestae, but emphasis was placed on spontaneity as a guarantee of non-concoction. In my respectful opinion, it is not desirable to subsume the res gestae principle under a general principle which would admit hearsay evidence when a trial judge believes that concoction was extremely unlikely because the hearsay statement was uttered with apparent spontaneity. The conditions stated in Ratten and in the joint judgment in Walton and substantially confirmed in Vocisano should continue to govern the admissibility of hearsay evidence under the res gestae exception. Deane J. in Walton also favoured a relaxation of the common law rule "where the circumstances are such that its inflexible application would confound justice or common sense or produce the consequence that the law was unattuned to the circumstances of the society which it exists to serve" [50] . His Honour thought but did not decide that, at least in some circumstances, statements made in the course of a telephone conversation by one party might be admissible to establish the identity of the other. The introduction of a telephone call exception to the hearsay rule has not been pursued thus far. For the purposes of this case, it is sufficient to observe that Walton stands as a recent decision of this Court affirming, at least by a majority, the conditions of admissibility of res gestae evidence stated by Lord Wilberforce in Ratten , namely, approximate if not exact contemporaneity with the res gestae and an assurance of non-concoction arising from the spontaneity of the statement or the involvement of the maker in the events of the res gestae. 1. (1989) 166 C.L.R., at p. 304. 2. [1972] A.C., at pp. 389-390. 3. [1987] A.C. 281, at pp. 300-301. 4. (1940) 64 C.L.R., at p. 531. 5. (1974) 130 C.L.R., at p. 273. 6. (1989) 166 C.L.R., at pp. 306, 309. 7. ibid., at p. 293. 8. ibid., at p. 308.

In Benz a majority of the Court regarded the statement of the woman on the bridge as admissible to prove the fact asserted by her, namely, that the other woman was her mother. Mason C.J. [51] held the statement to be admissible on the footing of res gestae. His Honour also expressed the view [52] that there was a "strong foundation" for receiving utterances of belief in a close familial relationship as evidence of that relationship. Deane J. [53] declined to grant special leave to appeal to determine the res gestae question and Dawson J. [54] , though finding that the statement would have qualified under the res gestae principle if it had been hearsay, held the statement to be admissible as original evidence. Gaudron and McHugh JJ. held [55] that the statement could have been used by the jury to find the truth of that assertion once they were satisfied that the two women on the bridge were the murderers engaged in disposing of the body when the witness saw them. Subject to that finding, the statement was admissible as part of the res gestae. The decision in Benz thus leaves in place, if it does no more, the decision in Walton as to the conditions of admissibility of hearsay evidence under the res gestae principle.

1. (1989) 168 C.L.R., at pp. 117-118. 2. ibid., at p. 117. 3. ibid., at pp. 127-128. 4. ibid., at p. 135. 5. ibid., at p. 144.

The principles above discussed govern the resolution of the present case. These principles must be applied to three distinct pieces of evidence: the telephone conversation by Allen on the morning after the murder overheard by Mrs. Berry; Allen's statement to Mrs. Berry after the telephone conversation, including his assertion that he had been talking to Ray; and Allen's statements to Mr. Berry during the days following the murder. None of these statements was made at a time substantially contemporaneous with the shooting of Simpson. None of them was made spontaneously. The res gestae here was the shooting of Simpson [56] , not the making, non-performance and subsistence of an assassination contract. The contract was relevant to prove the guilt of the party to the contract who had agreed to perform the assassination, but the dealings between the contracting parties were not part of the res gestae. None of the statements now in question qualifies as part of the res gestae. The proposition applicable to each piece of evidence is simply: was it admissible as original evidence of a fact relevant to a fact in issue? If yes, evidence of the statement was admissible; if no, that evidence was inadmissible. To apply any other or wider test would be to import a rule which has not hitherto been accepted as part of the law of evidence. It is necessary to consider each of the three categories of evidence separately.

The words spoken by Allen in the telephone call on the morning after the murder might be seen, in the circumstances, as a statement by Allen implying that the other party to the conversation had made a contract with him, that the other party had failed to perform the contract properly and that final payment had been deferred until the contract was fully performed. If that were the only way in which evidence of the conversation could be used, it would have been inadmissible as hearsay. There is no evidence that Pollitt, if he were the caller, admitted the truth of the grounds of Allen's remonstrance. But Allen's statement to the telephone caller that the caller would "get the rest of the money when you do the job properly" was itself a promise by Allen to pay the caller. From the making of that promise and from the fact that Allen had remonstrated with the caller, an inference could be drawn that the caller had undertaken to perform some service for Allen and had not then performed it. The evidence was therefore admissible as original evidence if the fact that the caller had undertaken to perform some service for Allen and had not then performed it (the fact to be proved) was relevant to a fact in issue. When the fact of the telephone call containing a remonstrance and a promise is coupled with the evidence of Jones that Pollitt rang Allen that morning after he, Pollitt, found out that he had shot the wrong victim, there is a foundation for an inference that the conversation overheard by Mrs. Berry was a conversation with Pollitt. If that inference were drawn, Allen's remonstrating with Pollitt about his mistaken attempt to perform the contract and Allen's promise to pay on proper performance were relevant to found a further inference that a contract for the murder of Williams existed and had not been performed, though the conversation could not be used (for want of any admission on Pollitt's part) to prove the shooting of Simpson by mistake for Williams. Evidence of the telephone conversation was therefore admissible as original evidence tending, in conjunction with other evidence, to prove the contract.

The statement by Allen to Mrs. Berry after the phone call that the caller had been given "5,000" and expected another 5,000 and that it was "Ray" that he had been speaking to is simply hearsay. Its probative force to prove the identity of the caller depends solely on the assertion of that fact by Allen. Clearly the statement proved Allen's knowledge of the identity of the caller but Allen's knowledge or state of mind was entirely irrelevant to any fact in issue. If Allen had been heard to say "I made a contract with Pollitt to murder Williams but he shot Simpson by mistake", the statement would have been inadmissible as hearsay. To infer the truth of the statement made by Allen after the phone call from Allen's state of mind or from his knowledge of the facts asserted in the statement is to treat the statement as hearsay. Evidence of that statement was inadmissible.

The statements made by Allen to Mr. Berry in the days following the murder were also simply hearsay. To the extent that they tended to prove any fact, the probative force depended on giving credit to Allen's assertions of the fact that "Roy" had had a contract to do a job that he had "buggered up" or, to put it another way, the probative force of the evidence depended on an inference of the truth of what Allen said drawn from Allen's apparent knowledge of the facts. Whichever way the probative force of that evidence be put, evidence of Allen's statements to Mr. Berry was inadmissible.

Thus, two inadmissible pieces of evidence apparently identifying Pollitt as the person who had "buggered up" the contract were laid before the jury. This identification came from the deceased person, Allen, who had allegedly made the contract that Williams should be murdered. Having regard to the source of those statements and their propinquity to the time of the murder, it is impossible to discount their prejudicial effect on the mind of the jury. Indeed, the very argument in favour of the admissibility of these statements assumes the credibility of Allen's assertions. A direction to the jury by the trial judge that Allen's statements were "not evidence against the accused man" was quite insufficient to remove the seriously inculpating evidence of Allen's express identification of Pollitt as the murderer. The admission of those statements in evidence was erroneous in law and the prosecution cannot show that the appellant did not lose a chance of acquittal that would have been open to him had the evidence been excluded. I would grant special leave to appeal, allow the appeal, quash the conviction and order a new trial.

Had it been necessary for me to address the other questions raised in the reasons for judgment of Dawson and Gaudron JJ., I would be in general agreement with their Honours, though I would state the duty of a trial judge to give a warning in the terms used in the joint judgment in Longman v. The Queen [57] : "the general law requires a warning to be given whenever a warning is necessary to avoid a perceptible risk of miscarriage of justice arising from the circumstances of the case."

I agree with Dawson and Gaudron JJ. that, for the reasons which they give, the learned trial judge was correct in not directing the jury that the witnesses Raymond John Denning ("Denning") and Gary Roger Jones could not corroborate each other. Two other issues remain for determination. The first relates to the adequacy of the learned trial judge's direction to the jury in relation to the evidence of Denning. The second relates to the admissibility of certain evidence given by Mr. and Mrs. Berry of statements made by Dennis Allen ("Allen"), in the absence of the appellant ("Pollitt"), to the overall effect that Pollitt had, on 18 September 1984, killed the deceased ("Simpson") as a result of a mistake of identity. The prosecution case was that Pollitt had entered into an arrangement with Allen (who died before the trial) to kill a man called Alan Williams but had killed Simpson after mistaking him for Williams.

The evidence of a witness about an oral confessional statement allegedly made to him by an accused while the witness and the accused were incarcerated together in a prison is liable to be unreliable for a number of reasons. One such reason is that such evidence is easily concocted. Another is that, where such evidence is concocted, an accused will ordinarily be denied the possibility of corroboration of his denial of it. Another reason is that it is likely that a "prison informer" will be of bad character. Another is the likelihood that a prison informer may be motivated to fabricate such evidence either by a perception that he will derive some benefit in terms of sentence, treatment or release on parole or by reason of any of a variety of pressures of a type which may easily arise in a prison environment and which may not be apparent to a jury.

All those reasons for the prima facie unreliability of the evidence of a prison informer about an alleged oral confessional statement made by another prisoner applied to the evidence of Denning in the present case. Denning's evidence was that the alleged confessional statement, which he claimed to remember "nearly word perfect", was made in the course of a 1988 conversation, initiated by Pollitt, in a prison yard at Goulburn gaol in New South Wales. It was not suggested that anyone else overheard any part of the conversation. Denning had an extraordinarily bad criminal record. At the time he commenced to provide information to the authorities, he was serving a life sentence in a Victorian gaol and, unless granted immunity (for which he had applied), expected to be charged with a serious crime in Queensland when released. He was at least partly motivated, in providing information and giving evidence, by a desire to receive favourable treatment from the authorities. He had been "absented" from prison "so as to assist the police with their inquiries" in relation to a large number of cases in which he alleged that he had been the recipient of confessional statements by fellow prisoners. Plainly, the ordinary rule that a warning should be given whenever it is necessary to avoid a real risk of miscarriage of justice by reason of the potential unreliability of evidence required that an appropriate warning be given by the trial judge in his directions to the jury in relation to Denning's evidence. As I followed the argument, so much was common ground. The real question as regards this issue in the case is whether the warning given by the learned trial judge was inadequate.

The primary warning given by the trial judge in relation to Denning's evidence was along the lines of a traditional corroboration warning. The jury was directed to "look to see if there [was] independent evidence" which corroborated the evidence of Denning and to look for "evidence from some other acceptable source which implicates the accused in the crime". In my view, such a corroboration warning is not really appropriate to deal with the potential unreliability of the evidence of a prison informer about an oral confessional statement allegedly made by a fellow prisoner. Where a prison informer supplies information after encouragement by the authorities, the likelihood is that there will have already existed some evidentiary material which was known to the authorities and which connected the accused with the alleged offence. Thus, in the present case, both Denning's evidence and Pollitt's unsworn statement indicate that Denning knew that the police had other incriminating evidence against Pollitt at the time when Denning first provided information to the police. In such circumstances, a warning to the effect that it is dangerous to rely on a prison informer's evidence of an alleged oral confessional statement unless that evidence is corroborated by other evidence connecting or tending to connect the accused with the offence charged really serves little useful purpose. Indeed, such a warning could conceivably have the unintended effect of conveying to the jury a suggestion that, once they find some corroboration, it is safe to rely upon such evidence. If a corroboration warning is to be of real use in such a context, it must be in terms which direct the need for corroboration towards the reliability of the particular evidence, that is to say, material which corroborates the actual making of the oral confessional statement. An example of that type of corroboration is evidence establishing both that disputed material in the alleged oral confessional statement is accurate and that that material would not have been known to the witness if the alleged confessional statement had not been made. The usefulness of even that kind of corroboration will, however, depend very much on the circumstances. If, for example, the prison informer has provided information to the police about alleged confessional statements made by a number of prisoners in relation to a number of different alleged offences, the possibility that he may have derived the corroborating information from police sources cannot be ignored.

It would be unwise to seek to formulate in the abstract the contents of an appropriate warning for every case where the prosecution places significant reliance upon the evidence of a prison informer. Such a warning is most likely to be understood by a jury and to be effective if it is moulded in the words of the trial judge to fit the circumstances of the particular case. The most that should be said is that there will inevitably be cases where the minimum necessary warning will be in unqualified terms which draw attention to the fact that the evidence is of its nature potentially unreliable, which stress the need for very careful scrutiny of it and which warn that it would be dangerous to base a conviction upon it. If the prosecution evidence had stood alone, the present would have been such a case. It is true that Denning's evidence contained details of matters relating to the killing which could well have been provided to him by Pollitt in the course of a confessional statement. It is also true that there was other evidence corroborating Denning's evidence in the sense that it connected Pollitt with the offences with which he was charged. Nonetheless, it appears to me that the overall prosecution evidence did not serve to reduce the inherent unreliability of Denning's testimony as that of a police informer seeking personal advantage by providing information to the authorities about an extraordinarily large number of confessional statements allegedly made by fellow prisoners. To the contrary, that inherent unreliability was underlined by Denning's insistence that he was innocent of the crime for which he had been serving a life sentence when he commenced to provide information to the police. It was also underlined by the contrast between Denning's evidence at the trial and his initial statement to the police which had, apparently, been tape-recorded. In his evidence at the trial, Denning provided an allegedly "nearly word perfect" account of a quite detailed confessional statement which he said had been made to him by Pollitt in Goulburn gaol in 1988. In the course of that statement, Pollitt allegedly expressly admitted that he had killed "the wrong bloke" while acting under an arrangement between himself and Allen that he would kill Williams. In contrast, Denning's initial statement to the police about the killing was as follows:

This would have been probably - oh, I'm not sure. Ah, early or middle '80's, I think 1985, something like that. It was a few years back, ah, ah, I don't really know the full story, but I know there was a bloke knocked and Roy'd done it and oh, arrested a few years ago, ah, I know a bit about the case and it's nothing to do with Russell Cox.

The prosecution case did not, however, stand alone. Pollitt's own unsworn statement at the trial provided corroboration of the accuracy and reliability of Denning's evidence in important respects: that, while Pollitt and Denning were fellow prisoners, they had had a conversation about the killing of Simpson; that Pollitt had said things about the killing being the result of a mistake in the identity of the victim. Indeed, in his unsworn statement, Pollitt denied the accuracy of Denning's evidence in only one respect. That respect was that, according to Pollitt, he was telling Denning about "what [he] thought the allegations were against [him]" and not about what actually had happened. While Pollitt's unsworn statement does not corroborate Denning's evidence on the critical issue between the two, namely, whether Pollitt had been speaking of police allegations or of what had actually happened, it goes a long way towards removing the ordinary possibility that the evidence of a police informer is a complete fabrication.

Even in the context of the support for important aspects of Denning's evidence which was provided by Pollitt's unsworn statement, it would, in my view, have been preferable for the learned trial judge to have directed the jury in unqualified terms that Denning's evidence was, for the reasons which I have indicated, potentially unreliable. That direction could, of course, have been accompanied by a comment drawing the jury's attention to the fact that Pollitt, in his unsworn statement, had expressly disputed the accuracy of Denning's evidence only to the limited extent that he claimed that he had told Denning only of allegations made against him. The adequacy of his Honour's directions must, however, be considered in the context of the fact that counsel for Pollitt, who would have been conscious of the conduct of the trial and the content of counsels' closing addresses, did not seek any relevant further directions. In that context, I have come to the conclusion that the combined effect of his Honour's remarks in summing up to the jury satisfied the requirement that an appropriate direction be given to warn the jury of the potential unreliability of the evidence. Those remarks included: (i) the descriptions of Denning as "a person with a long criminal history", "a person of bad character" and "a tainted person"; (ii) a warning that "common sense dictate[d]" that the jury "scrutinise [Denning's evidence] very carefully"; (iii) a direction that "experience has shown that persons of [Denning's] calibre" may "have motives for not telling the truth"; and (iv) an instruction to look for evidence "from some other acceptable source".

The relevant evidence of Mr. and Mrs. Berry fell into two categories. The first consisted of Mrs. Berry's evidence about what Allen had said in the course of, and immediately after, a telephone call which he received some time after Simpson had been killed. The second category consisted of the evidence of Mr. Berry about some subsequent statements made by Allen, independently of the telephone call, to the effect that someone wanted $5,000 and "didn't do the job right" and that "Roy" had got the wrong person and wanted to be paid for it. References to "Ray" or "Roy" were understood as references to Pollitt who was apparently known by both those names.

It can be said at once that the second category of evidence - i.e. the evidence of Mr. Berry - was not, in my view, admissible. That evidence was, of course, relevant to the central issues in the trial. It was common ground at the trial that Simpson had been killed as the result of an arrangement between Allen and the killer for the killing of Williams. Understood in context, the content of what Allen was alleged to have said established, if it was accepted as true, that there had been an arrangement between Allen and Pollitt that Pollitt would kill Williams and that, as a result of a mistake in identity, Pollitt had in fact killed Simpson. The evidence was not, however, admissible as proof of those matters. If tendered for the purpose of establishing the truth of them, it was hearsay.

Mr. Berry's evidence of what Allen had said also tended to establish what Allen believed at the time he made the relevant statements: his belief that there was an arrangement between himself and Pollitt; his belief that Pollitt had mistakenly killed Simpson. If tendered for the purpose of establishing that Allen believed those things, the evidence was not hearsay [58] . Nonetheless, it was inadmissible in that what Allen may or may not have believed about an arrangement between himself and Pollitt was not, of itself, a material fact in the trial of Pollitt for the alleged murder of Simpson. If, for example, Allen had been alive at the time and had been called as a witness, the Crown would not have been entitled to lead evidence that he believed that he had an arrangement with Pollitt and that he believed that Pollitt had killed Simpson. The reason why that is so is that Allen's beliefs were not, of themselves, admissible to prove anything against Pollitt on his trial for the murder of Simpson. If those beliefs had been based on things which Allen had seen Pollitt do or heard Pollitt say, he could have given evidence that he had seen or heard Pollitt do or say those things. He could not, however, have supplemented that evidence by evidence of the beliefs he had formed as a consequence of what he had seen and heard. On the other hand, if his beliefs were based not on anything that he had seen or heard Pollitt do or say but upon, for example, what a third person had told him that a fourth person had said, he could not give hearsay evidence of what the third person had said. Nor could he indirectly give such double hearsay evidence by giving evidence of what he believed as the result of what the third person had said.

1. See, generally, Walton v. The Queen (1989), 166 C.L.R. 283.

There are cases, apart from conspiracy cases, in which the evidence indicates that the accused and another person acted in preconcert in the commission of a specific crime. In such cases, evidence of what the other person did or said in furtherance of the commission of that crime may be admissible against the accused for the reason that the combination or preconcert to commit the crime is "considered as implying an authority to each to act or speak in furtherance of the common purpose" on behalf of the other [59] . It is at least arguable that the present case fell within that class of case with the result that things said and done by Allen, in furtherance of a criminal arrangement between Pollitt and himself of which evidence had already been adduced [59] , were also admissible against Pollitt. Even in such cases, however, "it can seldom happen that anything said by one which is no more than a narrative statement or account of some event that has already taken place, that is to say, some statement which would be receivable in evidence against the man who made it as an admission and not otherwise, can become admissible against his [companion] in the common enterprise" [59] . The reason why that is so is that such narrative statements are ordinarily not in furtherance of the common design and, therefore, are not within the scope of any implied authority. Thus, in the present case, the statements which Allen allegedly made, independently of the telephone conversation, to the effect that Pollitt had "got" the wrong person were not made in furtherance of any common design to kill Williams or do anything else. They were assertions of a purely narrative character. They would have been admissible against Allen, if he had been on trial, as admissions made by him. They were not admissible against Pollitt.

1. Tripodi v. The Queen (1961), 104 C.L.R. 1, at p. 7. 2. Tripodi v. The Queen (1961), 104 C.L.R. 1, at p. 7. 3. Tripodi v. The Queen (1961), 104 C.L.R. 1, at p. 7.

The position is not so clear as regards the evidence falling in the first category, that is to say, Mrs. Berry's evidence relating to what Allen had said during and after the telephone conversation. Her evidence was to the effect that, on the night after Simpson was shot, she had been present at Allen's home when he received a telephone call in the course of which he had said to the caller: "You get the rest of the money when you do the job properly" and that, immediately after that telephone call, Allen had "said it was Ray" (i.e. Pollitt) to whom he had been speaking. According to Mrs. Berry, Allen also made the comment: "I have already given him 5,000, he expects me to give him another five for something he hasn't done." No complaint is made on behalf of Pollitt about the evidence of what Allen said in the course of the telephone conversation. As has been said, it was apparently common ground at the trial that there had been an arrangement for the killing of Williams between Allen and the person who killed Simpson. The forensic dispute was about whether the killer was Pollitt or someone else. The complaint made on behalf of Pollitt is about the admissibility of Mrs. Berry's evidence of what Allen said, immediately after the telephone conversation, which identified Pollitt as the caller and, when understood in context, impliedly asserted that Pollitt had killed Simpson as the result of a mistake.

In the circumstances of the present case, the admissibility of what was said by Allen in the course of the telephone conversation ultimately depended upon whether the other party to the conversation was Pollitt. Evidence of comments made by Allen to Pollitt, in the course of a telephone call made to him by Pollitt, to the effect that Pollitt had already received part payment and that "the job" had not been done "properly" would be admissible for the reason that it could, in the absence of any indication of dissent on the part of Pollitt, base a finding that Pollitt had impliedly admitted those matters. Alternatively, evidence of the making of the telephone call by Pollitt and of what Allen said in the course of it would be admissible for the reason that the telephone call would be a step in the furtherance of the alleged common purpose to kill Williams which had miscarried and resulted in the killing of Simpson. On the other hand, if the person at the other end of the telephone was not Pollitt, Mrs. Berry's evidence of what Allen had said in the telephone conversation would be inadmissible against Pollitt. In a context where the evidence suggested that Allen had been involved in numerous other criminal dealings, a statement made to some person other than Pollitt that that other person would get "the rest of the money" when he had "do[ne] the job properly" could scarcely be seen as admissible on the ground that it would have been made in furtherance of the common criminal purpose which the Crown alleged existed between Allen and Pollitt. Putting to one side the furtherance of such a criminal purpose, evidence of what Allen said on the telephone to a person other than Pollitt would be inadmissible against Pollitt for the reasons given in relation to Mr. Berry's evidence about Allen's statements to him: except to the extent that it tended to establish Allen's beliefs, it would be hearsay; to the extent that it established Allen's beliefs, it would be inadmissible for the reason that what Allen believed was not a material fact in the trial of Pollitt.

It may be arguable that other evidence in the case about the existence of an arrangement between Allen and Pollitt could found an inference that the other party to the conversation which Allen had on the telephone was, in fact, Pollitt. If that were so, Mrs. Berry's evidence was direct evidence of part of a conversation between Allen and Pollitt, in the course of a telephone call made by Pollitt, and was, for the reasons given above, admissible against Pollitt. The question whether other evidence in the case founded such an inference was not, however, examined in argument. It is unnecessary that I pursue it since Mrs. Berry's evidence that Allen had, immediately after the conversation, identified Pollitt as the caller was admissible pursuant to what should, in my view, now be recognized and accepted by the courts of this country as a qualification of the hearsay rule.

Four of the principal reasons militating against the reception of hearsay evidence were succinctly identified by Lord Normand in Lejzor Teper v. The Queen [60] . First, it is not the best evidence. Second, it is not delivered on oath. Third, the truthfulness and accuracy of the maker of the relevant statement cannot be tested by cross-examination. Fourth, the light which the demeanour of the maker of the statement would throw on his or her testimony is lost. A fifth is that hearsay evidence of statements made by a third person is, especially where the third person is not available as a witness, easily fabricated and, even when fabricated, difficult to disprove. There are other considerations supporting the exclusion of hearsay evidence as a general rule. One is that the length and cost of legal proceedings could be intolerably increased if evidence of statements made outside the witness box by third persons were generally admissible as evidence of the truth of the matters alleged in them. Another is that, while the legitimacy of the Latin maxim res inter alios acta alteri nocere non debet as a basic tenet of the law of evidence is open to serious question [61] , ordinary considerations of justice lend some support for the "practical and convenient" test which treats "the presence or absence of the prisoner" as being prima facie decisive "of the admissibility of things said" by others as evidence in a criminal trial to prove that the accused did or said the things alleged against him [62] .

1. [1952] A.C. 480, at p. 486. 2. See, e.g., Cross on Evidence, 3rd Aust. ed. (1986), par. 1.120. 3. Tripodi v. The Queen (1961), 104 C.L.R., at p. 7.

On the other hand, the general rule against the admissibility of hearsay statements is a judge-made rule which is clearly subject to exceptions or qualifications. The question whether those exceptions or qualifications should be regarded as finally closed or whether room remains for judicial recognition of new exceptions or qualifications to accommodate the demands of justice and common sense in changing circumstances is not without difficulty. The arguments relevant to that question were well identified in a series of instructive speeches in the House of Lords in Myers v. Director of Public Prosecutions [63] and Reg. v. Kearley [64] . For my part, I respectfully agree with what was said by Lord Pearce in the following extract from his speech in Myers v. Director of Public Prosecutions [65] :

My Lords, the evidence whose admission is the ground of complaint was fair, clear, reliable and sensible. The question is whether the court was bound by a technical rule to exclude it. No one doubts that the general exclusion of hearsay evidence, subject to exceptions permitted where common sense and the pursuit of truth demand it, is an important and valuable principle. But it is a disservice to that general principle if the courts limit the necessary exceptions so rigidly that the general rule creates a frequent and unnecessary injustice. In my view, it is permissible and desirable that this Court, as the final appellate court of the nation, recognizes a new exception to or qualification of the hearsay rule when changing circumstances throw up an identified category of case in which it is plain that the inflexible application of the rule would confound justice or common sense or produce the consequence that the law was unattuned to the circumstances of the society which it exists to serve [66] . 1. [1965] A.C. 1001. 2. [1992] 2 A.C. 228. 3. [1965] A.C., at p. 1034. 4. See Walton v. The Queen (1989), 166 C.L.R., at p. 308.

In the times when the hearsay rule was in the course of being developed by the courts, almost all direct communications between persons were made either face to face, with the result that a witness could ordinarily see and hear both parties, or by writing, with the result that the written communication would itself commonly be preserved. These days, much of the private and commercial communications (including, presumably, criminal communications) between persons are by way of the telephone. The standards of our society preclude the legal use of interception devices in other than exceptional circumstances. Consequently, third party evidence of what is said in telephone conversations will ordinarily be subject to the restriction that a witness will have heard but one side of the conversation. In such a case, however, the identity of the other party to the conversation may be made manifest by what is said by the party whose side of the conversation can be heard. Strictly speaking, the comments of that party are hearsay to the extent to which they expressly or impliedly identify the other party. Even where evidence is given by a person who was actually a party to a telephone conversation, there will often necessarily be an element of hearsay in the identification of the other party. Thus, a person who receives a telephone call in relation to a commercial transaction from a person who identifies herself as AB may well be unable independently to identify AB's voice. It would, in my view, plainly be contrary to both common sense and the interests of justice if, in circumstances where there is nothing at all to suggest impersonation, the evidence of the caller's statement that it was AB on the telephone was inadmissible against AB or her principal as evidence prima facie establishing the identity of the caller for the reason that, until it was established that the caller was in fact AB, the statement to that effect was hearsay. Evidence of identification in such a telephone conversation is different from the case where a stranger comes to the door and asserts that she is AB. In the latter case, the direct observation of the other party to (or a witness of) the conversation enables subsequent physical identification to an extent that is unavailable in a case where all that has been heard is an unfamiliar voice on a telephone.

Once the view is taken that the hearsay rule should not be treated as being incapable of adaptation by the courts in circumstances where its inflexible application would confound justice or common sense, the conclusion seems to me to be inevitable that the rule must be adjusted to the extent necessary to permit the adoption of a sensible approach to an apparently spontaneous and reliable contemporaneous statement of one party to a telephone conversation identifying the other party to the conversation. In my view, the appropriate modification of the hearsay rule which should be recognized is that, where the circumstances do not on their face give rise to a significant possibility of fabrication or impersonation, a witness may give evidence that the party to a telephone conversation who was in his or her presence identified the person at the other end of the line by a statement made immediately before, in the course of, or immediately after, the conversation. Such evidence identifying the other party to the conversation will, of course, only be admissible if the identity of that person has the effect that evidence of the telephone conversation is admissible in the relevant proceedings against the party against whom it is tendered. So long as it is made immediately before, in the course of, or immediately after, the relevant telephone conversation, such a statement is likely to be spontaneous and reliable. Ordinarily, the other party to the conversation will be either the party to the litigation against whom such evidence is tendered or the agent of that party with the consequence that that party will be able to lead evidence disputing the evidence of identification if it is fabricated or the result of impersonation. A corresponding exception should apply in respect of the other category of case to which reference has been made, namely, where the witness was an actual party to the telephone conversation and the other party identified herself or himself expressly or by implication in circumstances which do not, on their face, give rise to a significant possibility of impersonation.

On the other hand, it does not appear to me that the appropriate qualification of the hearsay rule to allow evidence of the content of telephone conversations should extend to hearsay evidence of statements of one party to a telephone conversation about what was said by the other party in the course of the conversation. The function served by such statements, if they are tendered to prove the truth of their content, is that of narration. They can be affected and rendered inaccurate by faulty memory, interpretation or understanding of what was in fact said. Those are matters which cannot properly be investigated by cross-examination of a witness who has merely heard the statements as distinct from the person who made them. It is true that such statements, when made immediately after the conversation, may be spontaneous and less likely to be deliberately fabricated. Otherwise, however, the ordinary objections to hearsay evidence are applicable to them. Their exclusion, as hearsay, does not, in my view, offend either justice or common sense.

It follows that Mrs. Berry's evidence to the effect that Allen had had a telephone conversation with a person whom he identified as "Ray" (i.e. Pollitt) was admissible. It also follows that, once that evidence was led to the effect that the caller was Pollitt, Mrs. Berry's evidence that, in the course of the telephone conversation, Allen had said that Pollitt would get the rest of the money when he did the job properly was also admissible. On the other hand, Mrs. Berry's evidence that Allen had added the comment to the effect that he had already given Pollitt $5,000 and that Pollitt expected him to give him another $5,000 for something Pollitt had not done was inadmissible and should have been rejected. Mr. Berry's evidence to the effect that the shooting of Simpson had been the result of mistaken identity and that Pollitt had got the wrong person was also inadmissible and should have been rejected.

I have given careful consideration to the question whether any possibility of a resulting miscarriage of justice in the present case was precluded by the trial judge's directions to the jury about the very limited use which they were entitled to make of the relevant evidence of Mr. and Mrs. Berry. On balance, it appears to me that his Honour's directions did not have that effect. His Honour directed the jury that the statements made by Allen were "not evidence of the facts asserted by Allen" and were "not evidence against the accused man". The statements were, his Honour said, "admissible to establish Allen's state of mind at the time he made the statements and to demonstrate, if you accept the truth of them, that he was a party to the agreement to kill Williams". Those statements could, his Honour said, "only be used as establishing Allen's state of mind when he said them and to demonstrate, if you accept them, that he, Allen, was party to the agreement to kill Williams". It is difficult to be confident about how the jury would have interpreted those directions, particularly the direction that evidence which had been admitted on the trial of Pollitt was "not evidence against the accused man". It seems to me, however, to be likely that, in a context where it was common ground that Allen and the person who killed Simpson had been parties to an arrangement to kill Williams, a use of the evidence of Mr. and Mrs. Berry to establish what Allen believed (including, presumably, a belief that the other party to the arrangement with him was Pollitt) could have been just as damaging as the direct use of that evidence to establish that Pollitt was in fact that other party. It is true that Mrs. Berry's admissible evidence of what Allen said in the telephone conversation and of Allen's identification of the caller as Pollitt could well, in the context of all the evidence, have been accepted and interpreted by the jury as establishing that Allen and Pollitt had had a conversation based on a common understanding that Pollitt had killed Simpson by mistake. The inadmissible evidence of Mrs. and Mr. Berry about what Allen had said was, however, much clearer about Pollitt's involvement in the killing and, if accepted, left less room for doubt about Pollitt's guilt. In these circumstances, I am unable to conclude that there is no significant possibility that the wrongful reception of the inadmissible evidence affected the jury's verdict. That being so, I am not satisfied that the wrongful reception of that evidence did not relevantly deprive Pollitt of a fair trial according to law and thereby cause a miscarriage of justice.

I would grant special leave to appeal and allow the appeal. I would set aside the orders of the Court of Criminal Appeal and substitute therefor orders allowing the appeal to that court, quashing the conviction and directing that there be a new trial.

Roy Anthony Pollitt ("the applicant") was convicted of the murder of Lindsay Andrew Simpson. An appeal to the Full Court of the Supreme Court of Victoria was dismissed. The applicant now seeks special leave to appeal to this Court.

The application for special leave to appeal was refused at the conclusion of the hearing so far as it concerned two matters, namely, the admissibility of evidence identifying the applicant by his voice as the other party to a telephone conversation and directions given with respect to evidence which tended to identify a person other than the accused as the murderer. Two issues remain for consideration.

The prosecution case was that, on 18 September 1984, the applicant went to Lower Plenty with Gary Jones ("Jones") pursuant to a criminal arrangement with a third man, Dennis Allen ("Allen"): the arrangement was to murder Alan Williams ("Williams"), but the wrong man was killed. Williams was previously associated with Allen in illegal drug activities. On the day of the murder, the victim drove his wife and child to Williams' home. As he left the car, he was shot, protesting that the gunman had the wrong person.

The prosecution case relied heavily on the evidence of Jones and a prison informer, Raymond Denning ("Denning"), to whom the applicant allegedly confessed while both were in Goulburn gaol. The first issue is whether the trial judge adequately directed the jury with respect to the corroboration of the evidence of these persons.

On the prosecution case, Jones had conspired with the applicant to murder Williams. He pleaded guilty to a charge to that effect shortly before the applicant was brought to trial. Although not technically involved in the murder of Lindsay Simpson, Jones was quite properly treated as an accomplice [67] . The trial judge warned the jury that it was necessary for his evidence to be corroborated, although, strictly, all that is required in the case of accomplices is a warning that it is dangerous to convict on their uncorroborated evidence [68] . The trial judge also directed the jury that there was evidence which, if accepted, was capable of corroborating the account given by Jones, being the evidence of Jennifer Curillo, Mr. and Mrs. Berry, Detective Sergeant Brian Rix and, finally, the evidence of the prison informer, Denning.

1. McNee v. Kay, [1953] V.L.R. 520, at p. 530, per Sholl J.; Khan v. The Queen, [1971] W.A.R. 44, at p. 49, per Virtue S.P.J.; pp. 54-55, per Burt J. But cf. Davies v. Director of Public Prosecutions, [1954] A.C. 378, at p. 400 where Lord Simonds spoke of accomplices as persons who were "participes criminis in respect of the actual crime charged". 2. Davies v. Director of Public Prosecutions, [1954] A.C., at p. 399; Director of Public Prosecutions v. Hester, [1973] A.C. 296, at p. 325, per Lord Diplock; Reg. v. Medcraft and Medcraft (1981), 5 A. Crim. R. 186, at p. 192. For a history of corroboration, see Reg. v. Rosemeyer, [1985] V.R. 945, at pp. 960-966, per Ormiston J.

The evidence of prison informers has featured in a number of recent trials, particularly in New South Wales [69] . There is no rule of law or practice identifying evidence from a source of that kind as evidence which must be corroborated or as evidence upon which it is dangerous to convict without corroboration. However, it is always the duty of a trial judge to warn of the danger of convicting on evidence which is potentially unreliable [70] and it would only be in an exceptional case that the evidence of a prison informer would not fall into that category. Thus, in all but the exceptional case, it is necessary for a trial judge to warn of the danger of convicting on evidence of that kind unless corroborated by other evidence connecting or tending to connect the accused with the offence charged.

1. Reg. v. Drummond & Domican [No. 2] (1990), 46 A. Crim. R. 408; Reg. v. Anderson (1991), 53 A. Crim. R. 421. 2. Bromley v. The Queen (1986), 161 C.L.R. 315, at p. 319.

A corroboration warning was given in relation to Denning's evidence. But two complaints are made. First, it is argued that the direction "did not go far enough and did not cover the salient features of the problem that Denning raised". Then it is said that there should have been a direction to the effect that Denning and Jones could not corroborate each other.

It is convenient first to consider the argument that Denning and Jones could not corroborate each other. Leaving aside the special problems associated with the unsworn evidence of children [71] (which, of course, are not involved in this case) and the position with respect to the evidence of accomplices [72] , there is no rule of law or practice that evidence which attracts a corroboration warning cannot corroborate or be corroborated by other evidence attracting the same warning [73] .

1. See, e.g., Reg. v. Campbell, [1956] 2 Q.B. 432, at p. 438, per Lord Goddard C.J.; Director of Public Prosecutions v. Hester, [1973] A.C., at p. 314, per Lord Morris of Borth-y-Gest. 2. See, e.g., R. v. Baskerville, [1916] 2 K.B. 658, at p. 664, per Lord Reading C.J.; Director of Public Prosecutions v. Kilbourne, [1973] A.C. 729, at p. 751, per Lord Reid; Reg. v. Rigney (1975), 12 S.A.S.R. 30, at pp. 36-37, per Bray C.J.; pp. 51-52, per Hogarth J. 3. Director of Public Prosecutions v. Hester, [1973] A.C., at pp. 313-314, per Lord Morris; p. 326, per Lord Diplock; pp. 330-331, per Lord Cross of Chelsea; Reg. v. Rigney (1975), 12 S.A.S.R., at p. 40, per Bray C.J.; p. 57, per Jacobs J. See also Director of Public Prosecutions v. Kilbourne, [1973] A.C., at p. 747, per Lord Hailsham of St. Marylebone; pp. 750-751, per Lord Reid; p. 758, per Lord Simon of Glaisdale; p. 760, per Lord Cross.

The rule that one accomplice cannot corroborate another is based in large part on the common interest of accomplices in minimizing their involvement in the offence charged. That common interest raises the possibility that they may have conspired to give an identical but false account [74] . It is for this reason that a direction is given that accomplices cannot corroborate each other [75] . Thus, if there is no possibility of joint fabrication, as, for example, where the witness was an accomplice in other offences and is called to give similar fact evidence, a direction to this effect is not required [76] . Equally, there may be cases not involving accomplices where, because of the possibility of joint fabrication, it is necessary to direct a jury to look elsewhere for corroboration. But there was no basis for any suggestion of joint fabrication in this case and, thus, there was no need for a direction that Denning and Jones could not corroborate each other.

1. Reg. v. Button (1991), 54 A. Crim. R. 1, at p. 8, per Byrne J. See also Heydon, The Corroboration of Accomplices [1973] Criminal Law Review 264 . 2. Director of Public Prosecutions v. Hester, [1973] A.C., at p. 325, per Lord Diplock; Director of Public Prosecutions v. Kilbourne, [1973] A.C., at pp. 750-751, per Lord Reid. 3. See Director of Public Prosecutions v. Kilbourne, [1973] A.C., at p. 748, per Lord Hailsham.

Apart from directing the jury to "look to see if there [is] independent evidence which corroborates [Denning]", on several occasions the trial judge warned that Denning might "have motives for lying" and described him as "being of bad character" and "a tainted person". No direction was given as to the evidence capable of corroborating his account, but the trial judge instructed the jury to look for "evidence from some other acceptable source which implicates the accused in the crime".

Ordinarily, the issue with respect to the evidence of a prison informer is whether he or she has given an account of a conversation that simply did not occur. In cases of that kind, the ordinary warning to look for corroboration in evidence which implicates the accused in the crime charged will not direct attention to the real issue and, because there is usually some evidence of that kind, may lend a semblance of reliability to what is essentially unreliable evidence. But that is not the situation in this case. The applicant admitted that there was a conversation along the lines of that to which Denning deposed. The only issue was whether, in that conversation, he gave an account of events in which he had been involved or, as he claimed, of allegations that had been made against him.

The nature of corroborative evidence and the evidence which, if accepted, might constitute corroboration were fully explained by the trial judge in relation to the evidence of Jones. Having regard to the limited nature of the challenge to Denning's evidence, the directions given with respect to that evidence, in combination with the explanation of corroboration given in relation to Jones, was sufficient to alert the jury to the problems associated with it, to the danger of acting on it unless corroborated and to the manner in which they should look to see if it was, in fact, corroborated. There is, thus, no substance in the complaint that the directions "did not go far enough and did not cover the salient features of the problem that Denning raised".

The second issue in the application concerns the evidence of Mr. and Mrs. Berry who were regular visitors to Allen's home. As earlier mentioned, the prosecution case was that Allen arranged for the applicant to murder Williams, but, instead, it was Lindsay Simpson who was shot. Allen died before the applicant was brought to trial.

The evidence of Mrs. Berry was that on an occasion in 1984 and while she and her husband were visiting Allen's home, he told them to watch the late news. In the early hours of the next morning, Mrs. Berry saw a newsflash concerning a shooting. Later that same morning, she and her husband were again at Allen's home when Allen referred to the newsflash and described the shooting as one of "mistaken identity". Later that day or, perhaps, early the next morning, they were present when Allen received a telephone call during which he became angry and said to the caller: "You get the rest of the money when you do the job properly." A little later, Allen told Mr. and Mrs. Berry that he had been speaking to "Ray", that "he had given [Ray] a job to do and he had stuffed it up". According to Mrs. Berry, Allen also said: "I have already given him 5,000, he expects me to give him another five for something he hasn't done." Mrs. Berry knew the applicant as Ray. The evidence of Mr. Berry was to the same effect, save that, on his account, it was not until some few days after the telephone conversation that, in the course of a conversation about money, Allen indicated that it was the applicant to whom he had spoken on the telephone, saying: "Roy got the wrong one and he wants to be paid for it."

It was argued that the evidence of Mr. and Mrs. Berry, so far as it concerned the conversation in which Allen referred to "Ray" or "Roy", was hearsay and, thus, inadmissible. The rule against hearsay excludes evidence of an out-of-court statement if tendered to prove the truth of the matter stated [77] . But the making of a statement is a fact and, as such, evidence may be given of what was said if that itself is a relevant fact [78] .

1. Reg. v. Sharp, [1988] 1 W.L.R. 7, at p. 11; [1988] 1 All E.R. 65, at p. 68; Reg. v. Kearley, [1992] 2 A.C. 228, at pp. 254-255, per Lord Ackner. 2. Ratten v. The Queen, [1972] A.C. 378, at p. 387, per Lord Wilberforce. See also Walton v. The Queen (1989), 166 C.L.R. 283, at pp. 288-289, per Mason C.J.; pp. 305-306, per Wilson, Dawson and Toohey JJ.; p. 307, per Deane J.

There is no novelty in the proposition that evidence may be given of an out-of-court statement to prove the maker's contemporaneous state of mind [79] , including knowledge [80] , emotion [81] and intention [2] , if they or any of them be relevant. In criminal cases, the state of a person's mind is a matter which is commonly in issue and the cases concerning out-of-court statements are, in the main, concerned with that question. But, an out-of-court statement may be relevant to other matters including, for example, the relationship existing between particular individuals [3] , the presence of a person at some particular time or place [4] or the use of premises for particular purposes [5] .

1. Subramaniam v. Public Prosecutor, [1956] 1 W.L.R. 965; Concrete Constructions Pty. Ltd. v. Plumbers and Gasfitters Employees' Union of Australia [No. 2] (1987), 15 F.C.R. 64, at p. 82. 2. Thomas v. Connell (1838), 4 M. & W. 267 [150 E.R. 1429]; Wildman v. The Queen (1984), 12 D.L.R. (4th) 641. 3. Ratten v. The Queen. 4. Walton v. The Queen. 5. See, e.g., Reg. v. Benz (1989), 168 C.L.R. 110, at p. 116, per Mason C.J. 6. Reg. v. Hissey (1973), 6 S.A.S.R. 280. 7. See, as to the use of premises for betting, Davidson v. Quirke, [1923] N.Z.L.R. 552; Lenthall v. Mitchell, [1933] S.A.S.R. 231; McGregor v. Stokes, [1952] V.L.R. 347. See, as to the use of premises as a brothel, Woodhouse v. Hall (1980), 72 Cr. App. R. 39. But cf. Reg. v. Harry (1986), 86 Cr. App. R. 105 and Reg. v. Kearley where the issue was intention to supply, rather than the use of premises for the supply of, drugs.

When evidence of what was said is led to prove the fact that it was said and not to prove its truth, it is not hearsay, but primary and direct evidence. Its admissibility depends entirely on it being relevant, either because the statement itself is a fact in issue or because it tends to prove a fact in issue [6] . However, there may be cases where its relevance is so tenuous and, hence, its probative value so slight that, as a matter of discretion [7] , it should be excluded. There will also be cases where the evidence should be excluded as a matter of discretion because it contains an assertion which cannot be tested by cross-examination and there is a danger that a direction to treat the statement as a fact, and not as proving the truth of any assertion contained in it, may be ineffective. The absence of an opportunity for cross-examination is often advanced in explanation of the hearsay rule [8] , but the same problem can arise even though the evidence is led to prove what was said, rather than what was asserted, in the statement.

1. Walton v. The Queen (1989), 166 C.L.R., at pp. 288-289, per Mason C.J.; pp. 305-306, per Wilson, Dawson and Toohey JJ.; p. 307, per Deane J.; Reg. v. Kearley, [1992] 2 A.C., at pp. 242-243, per Lord Bridge of Harwich. 2. But cf. Reg. v. Blastland, [1986] A.C. 41 where it was held that unless the statement is "directly in issue or of direct and immediate relevance to an issue at the trial", it is inadmissible. 3. See, e.g., Lejzor Teper v. The Queen, [1952] A.C. 480, at p. 486 and Reg. v. Kearley, [1992] 2 A.C., at pp. 254-255, 258, per Lord Ackner; p. 259, per Lord Oliver of Aylmerton.

The foundation of the prosecution case was that Allen was the instigator of a criminal arrangement to murder Williams. That was a matter that had to be proved and, in the nature of such matters, proof depended very largely, if not entirely, on an inference being drawn from circumstances in respect of which direct evidence was available. Conduct by Allen which was contemporaneous, or nearly so, with the shooting of Lindsay Simpson was relevant to the inference that the jury was asked to draw: conduct consistent with it would strengthen the inference, whereas conduct inconsistent with it might destroy it entirely. And, of course, conduct includes what was said as well as what was done.

It is possible to view the evidence of Mr. and Mrs. Berry as evidence of a number of separate out-of-court statements. The first is Allen's instruction to watch the news; the second is his reference, the following day, to the shooting and the newsflash; the third consists of what he said during the telephone conversation; the fourth is his statement, either at the conclusion of that telephone conversation or some few days later, to the effect that "Ray" or "Roy" wanted to be paid even though he had "got the wrong one". No objection is or could be taken to the evidence concerning the first three of those statements, although, like the fourth statement, they too are out-of-court statements.

Although it is possible to treat the evidence of Mr. and Mrs. Berry as evidence of four separate out-of-court statements, the statements are more accurately described as interconnected events which, taken together, constitute a sequence or pattern of conduct on the part of Allen at or about the time that Lindsay Simpson was murdered. On that approach, the statement to the effect that Ray or Roy "got the wrong one" was no different from the earlier statements made by Allen and about which Mr. and Mrs. Berry gave evidence. All statements were circumstances which, taken with the other evidence in the case, supported the inference that Allen had arranged for Williams to be killed. As such, all were relevant and admissible, including the statement involved in this application.

There is one matter that distinguishes the statement in issue in this application from the other out-of-court statements about which Mr. and Mrs. Berry gave evidence. Unlike the other statements, the statement in issue contains an assertion which links the applicant with Allen's criminal arrangement and, hence, with the murder with which he was charged. But that merely describes the effect of the assertion: it does not detract from the admissibility of the statement as evidence of Allen's conduct, relevant to his involvement in a criminal arrangement to murder Williams.

There are particular problems when an out-of-court statement, itself a relevant fact, contains an assertion relevant to some other issue in the trial. It is in that situation that problems may arise if the truth of the assertion cannot be tested by the cross-examination of its maker. As earlier indicated, there may be occasions when, on this account, the proper exercise of discretion will result in the evidence being excluded. On the other hand, there may be occasions when it will be sufficient to direct the jury that the evidence should be used only for the purpose of drawing the inference which renders it admissible.

Clearly, evidence of an out-of-court statement which is relevant as a fact but which contains an assertion that is relevant to some issue must, if it is not excluded in the exercise of the trial judge's discretion, attract a direction that the making of the assertion does not go to prove the truth of its content. That was done in the present case, the jury being instructed that the evidence could "only be used as establishing Allen's state of mind and to demonstrate that he, Allen, was party to the arrangement to kill Williams" and that it was "not evidence against the accused man". It is appropriate to observe that, in the circumstances, that was adequate to deal with the problem associated with the evidence of Mr. and Mrs. Berry.

In the view which we take of the evidence of Mr. and Mrs. Berry, it is unnecessary to consider whether there should be a "telephone" exception to the rule against hearsay. That is a matter on which we would prefer to hear fuller argument before expressing any concluded view but, at present, we have some difficulty in extracting any principle upon which such an exception could be based.

Special leave to appeal should be granted, but the appeal should be dismissed.

The circumstances giving rise to this application for special leave to appeal are set out in the judgment of Dawson and Gaudron JJ. Their Honours' judgment also explains that only two questions remain for consideration by the Court.

It is true, as Dawson and Gaudron JJ. say of the prison informer, Denning:

There is no rule of law or practice identifying evidence from a source of that kind as evidence which must be corroborated or as evidence upon which it is dangerous to convict without corroboration. Despite the absence of any such rule of law or practice, there is an obligation on a trial judge to warn the jury of the danger of convicting on evidence which is "potentially unreliable" [9] , indeed "whenever a warning is necessary to avoid a perceptible risk of miscarriage of justice arising from the circumstances of the case" [10] . The evidence of prison informers is liable to be unreliable and so to give rise to a risk of miscarriage of justice. Deane J. has outlined the reasons why this is so. I am using the term "prison informer" for a prisoner, not connected with the offence in question, who purports to give evidence of statements of a confessional nature made by an accused while in prison. That was the situation here. The evidence of a "grass" or "supergrass" gives rise to other considerations [11] . 1. Bromley v. The Queen (1986), 161 C.L.R. 315, at p. 319. 2. Longman v. The Queen (1989), 168 C.L.R. 79, at p. 86. 3. See Greer, Supergrasses and the Legal System in Britain and Northern Ireland, Law Quarterly Review, vol. 102 (1986) 198.

Denning had a bad criminal record. He had previously told the police of confessional statements allegedly made to him by many other prisoners and had given evidence of such against one prisoner. The incentive for a prisoner in the position of Denning to give false evidence is so apparent as inevitably to attract a warning from the trial judge. But the nature and form of the warning call for comment. A warning, which states only that the evidence of a prison informer must be corroborated or that it is dangerous to convict upon such evidence without corroboration, tends to miss the point and, if anything, may militate against an accused. Almost by definition, there will be some "corroboration" of the prison informer's evidence, that is, of the matters to which the evidence relates. The accused will only be in prison because he or she has been charged with an offence and, in ordinary circumstances, that presupposes that some other evidence is available against the accused. Therefore, to tell the jury that it is dangerous to convict upon the evidence of a prison informer unless it is corroborated may, in many cases, almost carry an implied invitation to act upon that evidence. The situation of an accomplice is quite different. That person will have been involved in the offence. Certainly he or she may have something to gain by testifying against the accused but the evidence is generally capable of being tested by reference to the circumstances giving rise to the charge. There is virtually no way of meeting the evidence of a prison informer except by the accused's own denial.

It follows that the warning to the jury should be couched directly in terms that it is dangerous to convict upon the evidence of a person who is a prison informer, that such evidence should be scrutinized with great care and that the jury must be satisfied beyond reasonable doubt as to the guilt of the accused, having regard to the potential unreliability of the evidence of a prison informer and the risk of a miscarriage of justice by attaching too much importance to it. The direction should be properly tailored to the circumstances of the particular case. This will ordinarily require the trial judge to refer specifically to those facts which are likely to render the evidence of the witness unreliable.

Although the trial judge's direction, which relevantly appears in other judgments, did not contain a warning in the terms I have suggested, it was in the circumstances adequate to alert the jury to the dangers of acting on Denning's evidence. No miscarriage of justice was likely to have resulted from that aspect only of the direction.

The evidence of Wayne Desmond Berry and Sandra Margaret Berry

The evidence of Mr. and Mrs. Berry presents its own difficulties. The admissibility of that evidence must be assessed in the context of the prosecution case against the applicant. The presentment against him was originally on two counts, one of conspiring with Dennis Allen and Gary Roger Jones to murder Alan David Williams, and the other of murdering Lindsay Andrew Simpson. The trial of the applicant was on one charge only, that of murdering Simpson. Allen died before the applicant's trial. Jones had been charged with conspiring with the applicant to murder Williams; Jones pleaded guilty to that charge shortly before the applicant's trial and he was a witness for the prosecution at that trial. The case against the applicant was that Allen arranged for him to murder Williams but that, in error, the applicant killed Simpson. That is the context in which the admissibility of the evidence of Mr. and Mrs. Berry must be determined.

The evidence in question fell into several categories:

1. The evidence of Mrs. Berry that in 1984, while she and her husband were visiting Allen's home, Allen told them: "Watch the news, the late news." Mr. Berry said that there was a radio report that evening "that said somebody had been shot". In the early hours of the next morning Mrs. Berry saw a newsflash on television, apparently about a shooting.

2. The evidence of Mrs. Berry that the same day after the visit she and her husband were again at Allen's home when Allen said: "Did you watch the news? The mistaken identity." Mr. Berry said of Allen on that occasion: "No [in answer to a question whether the conversation of the previous night and the radio report were discussed], he was just a little bit agitated, and I just asked him what was wrong and he said that someone buggered up something they were supposed to do."

3. The evidence of Mrs. Berry that later the same day, or more likely early the next morning, she and her husband were present when Allen received a telephone call, in the course of which he said: "You get the rest of the money when you do the job properly." Mr. Berry did not depose to that conversation.

4. The evidence of Mr. and Mrs. Berry regarding what was said by Allen thereafter. In this regard there were differences in the evidence of the two witnesses. Mrs. Berry said that after the telephone conversation, and in the presence of herself and her husband, Allen was furious. He said he had been speaking to "Ray", that "he had given him a job to do and he had stuffed it up". He also said: "I have already given him 5,000, he expects me to give him another five for something he hasn't done." Mr. Berry did not depose to anything said by Allen after the telephone conversation had concluded. He did say that some days later Allen said that "someone wanted $5,000 and didn't do the job right". He then said that a couple of days later Allen commented: "Roy got the wrong one and he wants to be paid for it." In the light of all the evidence, it should be taken that the two witnesses were deposing to separate occasions rather than to the same occasion. That possible difference, while important, is not crucial to answering the question raised by the application.

That question is whether and to what extent anything said by Allen, not in the presence of the applicant, was evidence against the applicant. No real complaint was made of the evidence in the first three of the categories I have mentioned. It was the evidence in the fourth category that was the focus of attention, understandably so since that evidence provided a link between the applicant and Allen and could be regarded as an admission by the applicant that he had been involved in a killing arranged by Allen.

One thing is clear. Anything said by Allen to Mr. and Mrs. Berry could not be admissible evidence of the truth of what Allen said. That would be hearsay evidence in its clearest form. Equally, evidence of something said by Allen, not in the presence of the applicant, was admissible to prove Allen's state of mind if that was relevant and for some other purpose if any other purpose was relevant.

Dealing with Mr. Berry's evidence, the trial judge said to the jury that he wanted to give them a warning about the conversations with Allen "after the shooting". He said:

Those statements made by Allen are not evidence of the facts asserted by Allen. The only basis upon which they are admissible is to establish Allen's state of mind at the time he made the statements and to demonstrate, if you accept the truth of them, that he was a party to the agreement to kill Williams. But they do not establish the facts asserted by Allen and cannot establish the facts asserted because they were said in the absence of the accused man.

As to the evidence of Mrs. Berry as to what Allen said, his Honour observed:

The only basis upon which this evidence is admissible is to establish Allen's state of mind at the time and if you accept the truth of what she says, that he, Allen, was party to a plan to kill Williams. But it is no evidence against the accused man.

The direction of the trial judge immediately invites the question: What relevance did Allen's "state of mind" have to the case against the applicant? The Court of Criminal Appeal (Crockett and O'Bryan JJ., Cummins J. concurring on this aspect) put the matter this way [1] :

In our opinion the evidence was correctly admitted as tending to prove Allen's involvement with others in the killing of Simpson. The Crown did not rely upon the utterances of Allen as evidence of the truth of what he said but says that the evidence goes to establish Allen's state of mind and knowledge of an involvement in the killing. The essence of the Crown case was that Allen instigated a plan which involved the applicant to kill Williams. That Allen was a party to a plan to kill Williams and the wrong person was killed were facts very much in issue.

The evidence of the Berrys was admissible on the basis that the utterances of Allen tended to establish a fact relevant to a fact in issue namely, that he knew of the murder and was a party to it with another person or persons.

1. Reg. v. Pollitt (1990), 51 A. Crim. R. 227, at pp. 236-237.

When it is sought to adduce an out-of-court statement to prove the state of mind of the statement maker, the evidence is said to be original evidence. But the evidence is only admissible if it is of a fact in issue or because it tends to prove a fact in issue [2] . So the question must be: was Allen's state of mind in making the remarks in question a fact in issue or did it tend to prove a fact in issue? To that question the answer must be no. Any agreement between Allen and the applicant was a fact, to be proved in the ordinary way if relevant to the killing of Simpson. Statements made by Allen linking the applicant with Simpson's death did not go to Allen's state of mind when he spoke; rather, they went to some arrangement between Allen and the applicant. In any event, Allen's state of mind at the time he spoke was not relevantly a fact in issue nor did it tend to prove a fact in issue. Whether Allen had arranged for Williams to be killed was not of itself a fact in issue. Likewise, whether Allen had arranged for the applicant to kill Williams was not of itself a fact in issue. But it was relevant if it tended to prove that the applicant killed Simpson because he mistook Simpson for Williams. To say this is only to point up that what Allen said went to prove that the applicant killed Simpson in the course of carrying out an arrangement he had made with Allen. As soon as that is said, it is apparent that it is the truth of what Allen said that is relevant, thus branding the evidence as hearsay.

1. Walton v. The Queen (1989), 166 C.L.R. 283, at pp. 288-289, 307; Reg. v. Kearley, [1992] 2 A.C. 228, at p. 242-243.

So to conclude does not necessarily dispose of the matter. There are two further aspects that need to be examined; it is possible to view each as independent of the other though in truth they overlap. One aspect involves consideration of whether evidence which is technically hearsay may nevertheless be admitted if there are circumstances enhancing the reliability of what was said. The other aspect concerns the scope of res gestae, in particular the inclusion within its ambit of statements made so spontaneously or contemporaneously as to exclude the possibility of concoction. There are passages in the judgments in Walton that bear on these matters but it must be understood that that decision held statements to be admissible as constituting conduct on the part of the deceased victim from which her state of mind at the relevant time could be inferred. And in Walton , Wilson, Dawson and Toohey JJ. made it clear that "[t]he unlikelihood of concoction or distortion is not sufficient of itself to render a hearsay statement admissible" [3] . There must be something more. Generally, "[i]t is the contemporaneous involvement of the speaker at the time the statement is made with the occurrence which is identified as the res which founds admissibility" [4] .

1. (1989) 166 C.L.R., at p. 304. 2. Vocisano v. Vocisano (1974), 130 C.L.R. 267, at p. 273.

In its general operation the rule against hearsay is a valuable safeguard in the conduct of criminal trials. Hearsay evidence [5] "is not the best evidence and it is not delivered on oath. The truthfulness and accuracy of the person whose words are spoken to by another witness cannot be tested by cross-examination, and the light which his demeanour would throw on his testimony is lost."

1. Lejzor Teper v. The Queen, [1952] A.C. 480, at p. 486.

In Walton , Mason C.J. said that the hearsay rule "should not be applied inflexibly" [6] and with that statement there can be no quarrel. If applied rigidly, the rule can assume an unreality which gives little credit to the common sense of juries [7] . In Walton , Wilson, Dawson and Toohey JJ. commented [3] :

But if sometimes there is an element of hearsay in evidence which is led of statements made by a person other than a witness for the purpose of founding an inference concerning that person's state of mind, the justification for disregarding that element of hearsay may be thought to be of a similar kind. Such statements will rarely be purely assertive. Ordinarily they are reactive and are uttered in a context which makes their reliability the more probable. On the other hand, if a statement by a person about his state of mind is a bare assertion not amounting also to conduct from which a relevant inference can be drawn, then it ought to be excluded as hearsay. Of course those observations were made in a context in which a person's state of mind was relevant. For the reasons already given, that is not the present case. 1. (1989) 166 C.L.R., at p. 293; see also p. 308, per Deane J. 2. See the comments of Lord Griffiths in Reg. v. Kearley, [1992] 2 A.C., at pp. 236-237; and, more generally, Harriman v. The Queen (1989), 167 C.L.R. 590, at pp. 605-607. 3. (1989) 166 C.L.R., at p. 304.

What is it that lends such reliability to an out-of-court statement as to support its admission into evidence? What is it that allows a court to hold that "the possibility of concoction can be disregarded" [8] ? Spontaneity or contemporaneity provides some comfort. More accurately, perhaps, their absence makes it impossible to disregard the possibility of concoction and the rule against hearsay must prevail. Remarks made during the course of a telephone conversation are reactive rather than assertive and would ordinarily carry conviction as to the identity of the other party to the conversation. Common sense supports the admissibility of such remarks for the purpose of identification, if the remarks themselves are admissible, whether as original evidence or otherwise [1] . But it is a very considerable step to hold that evidence of remarks made by a confessed party to conspiracy to murder, at the conclusion of a telephone conversation, not only identifying the other party as "Ray" (Mr. Berry was told "Roy") but also purporting to repeat what the other person had said about his own criminal activities is admissible. Such remarks can hardly be described as reactive and they are much more open to the possibility of concoction. What Allen said to Mrs. Berry in those circumstances was hearsay and the rule against hearsay could not be relaxed to allow that evidence to be given. A fortiori, what Allen told Mr. Berry some days later was simply hearsay and, whatever flexibility of the rule is permissible, that evidence remains inadmissible.

1. Ratten v. The Queen, [1972] A.C. 378, at p. 389. 2. As to circumstances in which such remarks may be admissible, see Reg. v. Firman (1989), 52 S.A.S.R. 391, at pp. 394-396.

These deficiencies cannot be cured by calling in aid the concept of res gestae. The apparently simple Latin expression ("things done") cloaks considerable looseness of thought [2] . In Reg. v. Benz [3] Mason C.J., while acknowledging "the force of the criticisms made of the doctrine of res gestae", said that those criticisms did not "entail the conclusion that it has generated decisions that are incorrect". But the question still has to be asked: what were the "things done" of which the statements made by Allen to Mr. and Mrs. Berry might be said to form part? Clearly enough it was the killing of Simpson that was the thing done; with that, what was said by Allen some days after the shooting lacked contemporaneity and cannot be classified as part of the res gestae. It is argued that the killing of Simpson was the result (albeit the unintended result) of an arrangement made between Allen and the applicant for the killing of Williams. That arrangement was a relevant part of the prosecution case. But it does not make the arrangement part of the res gestae; the arrangement merely serves to explain why Simpson was killed.

1. See the criticisms collected in Cross on Evidence, 4th Aust. ed. (1991), vol. 1, p. 37,089. 2. (1989) 168 C.L.R. 110, at p. 117.

To the extent that statements made during or indeed immediately after a telephone conversation may constitute an exception to the hearsay rule, the scope of any such exception remains to be determined.

The implications for the applicant of the evidence of Mr. and Mrs. Berry as to what Allen said do not need to be dwelt on. Despite the warnings given by the trial judge, it would have been very hard indeed for the jury not to have been affected by what would have been understood as statements by Allen implicating the applicant in the killing of Simpson. To say to the jury that what Allen told Mr. Berry was "not evidence of the facts asserted by Allen" and that what Allen told Mrs. Berry was "no evidence at all against the accused man" could not be relied upon to dispel the incriminating nature of Allen's remarks. In those circumstances and despite other evidence incriminating the applicant, it is just not possible to discount the risk of a miscarriage of justice. There should be special leave to appeal, the appeal should be allowed, the applicant's conviction should be quashed and a new trial ordered.

This is an application for special leave to appeal against a conviction of murder. The applicant was convicted in the Supreme Court of Victoria of the murder of Lindsay Andrew Simpson. An appeal to the Court of Criminal Appeal of that Court was unanimously dismissed. Except to the extent necessary to explain my reasons, it is unnecessary to refer to the evidence in detail. The general nature of the case made against the applicant and the relevant evidence are discussed in other judgments.

The first question in the application is whether the trial judge erred in his directions to the jury concerning the need for corroboration of a confession allegedly made to Raymond John Denning, a prison informer. The second question is whether certain out-of-court statements were admissible in evidence on the trial of the applicant when they were made by a person who was neither a witness in, nor a party to, the proceedings. If one or more of the statements was inadmissible, a further question arises as to whether the admission of that evidence constituted a miscarriage of justice notwithstanding that the trial judge directed the jury that the statements were not evidence against the applicant.

The necessity for a warning in relation to the evidence of a prison informer

In the course of his summing up, the learned trial judge directed the jury as to the danger of acting upon the evidence of Denning. His Honour said:

Denning is a person with a long criminal history. He is a person one would describe as being of bad character or, as Mr. Ross put it, a tainted person or certainly up to comparatively recently. When considering the evidence of such a person commonsense dictates that you scrutinize it very carefully.

It does not automatically follow from the fact that Denning has a criminal record and has been a person of bad character that his evidence will be untrue. Nevertheless, experience has shown that persons of that calibre may - and I stress the word "may" - have motives for not telling the truth. So when considering what weight you will attach to his evidence it is wise to look for independent evidence which corroborates his account, that is, evidence from some other acceptable source which implicates the accused man in the crime.

Later on in his summing up, his Honour said: when you are dealing with such a person you must scrutinize his evidence very carefully when determining what weight you attach to it. You will also look to see if there is independent evidence you accept which corroborates the evidence of the witness in question. As I have said to you on a number of previous occasions such persons may have motives for lying. You must look at their evidence very carefully.

In his written submissions, counsel for the applicant contended that the learned trial judge had misdirected the jury because in "the case of oral and uncorroborated " gaol-yard " confessions, Juries ought to be cautioned to look for corroboration". The written submissions and parts of the oral argument seemed to assume that any independent evidence which implicated the accused in the crime was sufficient corroboration for this purpose. Indeed, for much of the argument, it seemed that the real complaint about the directions concerning Denning was that the learned judge had failed to direct the jury that the evidence of Jones could not be used to corroborate the evidence of Denning. But, during the course of the oral argument, counsel for the applicant said:

The sort of corroboration they should look for, independent corroboration, would be material which would make it more probable than not that in fact the appellant confessed to Denning; in other words, to support the confession, the fact that the confession was made. That is the sort of material with which the Court was really concerning itself in McKinney and Judge [4] . It is corroboration of the fact that the confession was made which was important as far as Denning is concerned, not just general supportive material as to the guilt of the accused. At the conclusion of his argument, counsel said that he was constrained to accept the force of the contention that what was required was "in fact not, properly speaking, a corroboration caution in the classical sense, but a warning such as in McKinney ". 1. McKinney v. The Queen; Judge v. The Queen (1991), 171 C.L.R. 468.

Prison informants do not fall into any of the categories of witness in relation to which a full warning concerning corroboration is required. If the evidence of a witness is potentially unreliable, however, the jury must be warned of the dangers of convicting on that evidence even though the witness does not fall into one of those categories of witnesses in respect of whom the trial judge must warn the jury that it is dangerous to convict upon their uncorroborated evidence [5] . But, independently of the circumstances concerning any particular prison informant, evidence from prison informers is potentially unreliable. The evidence of a prison informer may in fact be true. But, with the exception of some remand prisoners, the source is always tainted. All accounts of traditional prisons agree that beneath the veneer of law and order imposed by the rules of prison discipline lies a brutal world of fear and sudden, and often irrational, violence where conventional standards of conduct and values such as truth and respect for the rights of others have little relevance. It is not surprising, therefore, that, exposed to such an environment, some prisoners will become so indifferent to the rights and feelings of others that they will not hesitate to make false accusations of criminal conduct against other persons if acceptance of the accusations will advance their own interests. Many years of experience in hearing prisoners give evidence for and against accused persons has alerted the judiciary to the unreliability of the evidence of serving prisoners. But it is by no means certain that every juror fully appreciates that unreliability which arises not so much because the prisoner has been convicted of serious crime but because the character of that person has been altered for the worse by exposure to the values and culture of prison society.

1. Bromley v. The Queen (1986), 161 C.L.R. 315, at p. 319; Longman v. The Queen (1989), 168 C.L.R. 79, at pp. 86, 107.

Moreover, not only is the evidence of a prison informer potentially unreliable but there is a real danger that it will be accepted. The evidence which the prison informer frequently gives is that of a gaol-yard statement by the accused which amounts to a confession of guilt. Almost invariably, the statement is built around facts that the accused does not dispute or cannot disprove except by a bare denial. This circumstantial detail is apt to give the evidence the impression of inherent credibility. Furthermore, evidence given by prison informers is notoriously difficult to break down. The prison-informer witness is often a skilful liar, with a shrewd knowledge of the criminal courts and their procedures, a quick intelligence well suited to the giving of evidence and withstanding cross-examination and a good demeanour and friendly personality [6] . At the conclusion of the cross-examination, the jury is likely to be left with the impression of a personable witness whose evidence-in-chief was intrinsically credible and remained unshaken after a "searching" cross-examination. The danger from prison-informer evidence, therefore, lies in the plausibility as well as the character of those who usually give it.

1. cf. Brown, Prisoner informants: the new growth industry, Australian Society (June 1991) p. 11.

More than 300 years have passed since Sir Matthew Hale warned that "the truth" was that more mischief had come to "good men by false accusations of desperate villains, than benefit to the public by the discovery and convicting of real offenders" [7] . Whether "the truth" is the same today as it was in the time of Hale is in the nature of things unknowable. But certainly the dangers of acting on the evidence of "desperate villains" are no less today than when Hale wrote in 1650.

The use of prison informers seems to have increased significantly in recent years although statistical proof is not publicly available [8] . Various explanations for this phenomenon have been offered. None of them casts law enforcement or the administration of justice in a favourable light [9] . Whatever the explanation for the increase in the use of prison informers, their evidence is "suspect" evidence given by "suspect" witnesses [10] . A conviction made upon such evidence cannot be regarded as safe unless the jury were warned against the dangers of convicting on such evidence. But what form should the warning take?

1. See Brown and Duffy, Privatising Police Verbal: The Growth Industry in Prison Informants in Carrington et al. eds, Travesty! Miscarriages of Justice (1991), pp. 148-149; and cf. Reg. v. Drummond & Domican [No. 2] (1990), 46 A. Crim. R. 408; Reg. v. Anderson (1991), 53 A. Crim. R. 421. 2. See, e.g., Brown, op. cit., at pp. 11-12; Brown and Duffy, op. cit., pp. 159-162; Gibney and Woodyatt, Gaolhouse verbal, Legal Service Bulletin, vol. 16 (1991), 84. 3. cf. Reg. v. Hester, [1973] A.C. 296, at p. 324.

To direct a jury in terms of the conventional corroboration warning seems inappropriate. In many, perhaps the majority of, criminal cases where a prison informer gives evidence, the Crown case will contain corroboration in the traditional sense because it contains "independent testimony which affects the accused by connecting or tending to connect him with the crime" [11] . Very often, the accused person will have met the informer only because the accused is in gaol awaiting trial on the charge in question. But even in cases where the charge has only been instituted because of information provided to the authorities by the informer, it will often be the case that the accused is already suspected of being involved in the crime and that there is some independent evidence which tends to connect the accused with the crime. In these circumstances, the conventional corroboration warning gives the accused little, if any, protection against the dangers of acting on the evidence of prison informers. Indeed, a real possibility exists that such a direction might unwittingly induce the jury to believe that it is safe to act upon the evidence of a prison informer because his or her evidence is "corroborated".

1. R. v. Baskerville, [1916] 2 K.B. 658, at p. 667.

In McKinney [12] , this Court held that a trial judge was bound to direct the jury that:

they should give careful consideration as to the dangers involved in convicting an accused person in circumstances where the only (or substantially the only) basis for finding that guilt has been established beyond reasonable doubt is a confessional statement allegedly made whilst in police custody, the making of which is not reliably corroborated. Such a direction was required even though the Court made it plain that there was no suggestion [13] "that police evidence is inherently unreliable or that members of a police force should, as such, be put in some special category of unreliable witnesses". Because evidence by prison informers is inherently unreliable, it follows that a confession allegedly made to a prison informer must be the subject of a direction at least as stringent as that required when a disputed confession is alleged to have been made while the accused was in police custody. 1. (1991) 171 C.L.R., at p. 476. 2. ibid., at p. 478.

If accused persons are to be protected against the dangers arising from the evidence of prison informers and the fairness of criminal trials are to be ensured, juries must be directed as to the danger of acting upon such evidence. Ordinarily, that will mean that the jury must be directed that it is dangerous to act on that evidence where the account of the prison informer is not substantially confirmed by independent evidence. Thus, if the informer alleges that the accused has confessed to the crime, it will be dangerous to act on his evidence unless the making of the confession is substantially confirmed by independent evidence. If the informer alleges that he saw or heard the accused do or say something which incriminates the accused, it will be dangerous to act on the allegation unless the doing of the act by the accused or the making of the statement is substantially confirmed by independent evidence. Only in the most exceptional case, if at all, could the evidence of a fellow prisoner be regarded as independent evidence for this purpose. Furthermore, the jury should be warned not only as to the inherent or general danger of acting upon the evidence of a prison informer but, where appropriate, instructed to consider any specific matters which could reasonably be regarded as undermining the credibility of the witness. Finally, the trial judge should draw to the jury's attention the particular matters which could reasonably be regarded as confirming the evidence of the prison informer.

In this case, the learned trial judge did not direct the jury that it was dangerous to convict on the evidence of Denning unless they found that there was independent evidence which substantially confirmed that the applicant had made the confession in question. Nor did his Honour direct the jury as to any matters that could reasonably be regarded as confirming the making of the confession. Nevertheless, in the very special circumstances of this case, his Honour's directions cannot be said to have failed to bring home to the jury the danger of acting on the evidence of Denning. Significantly, counsel for the applicant at his trial did not object to his Honour's directions concerning Denning or ask for any redirection in relation to that evidence.

In his unsworn statement, the applicant admitted, expressly and by necessary implication, that he had had a conversation with Denning in Goulburn gaol about the murder of Simpson. In his statement, the applicant said that in 1987 he had heard that "there was a warrant out for my arrest" over the murder and that he had "tried to collect everything that I could on the subject to find out what they were alleging I had done". He said:

I wrote to my solicitor and asked him to make inquiries on a number of occasions. I asked quite a few people to dig up information for me relating to this matter. In the end, I had some news from the solicitor on what the case was about and some statements from witnesses. I thought at the time that the killer was supposed to have made the mistake because of the mix-up on cars, and that there was a photograph involved. Whether that came from the statements or the news, I can't remember.

Anyway, I told Ray Denning about what I was doing, what I thought the allegations were against me. I might have mentioned Kathleen Pettingill's name, but I cannot have said more about it than I knew, and that was no more than anybody knew. That is as much as I can tell you.

The unsworn statement of the applicant makes it clear that the conflict between the applicant and Denning concerning their conversation was a narrow one. The dispute was whether the applicant had told Denning what he had done or merely what he was alleged to have done. Once the applicant conceded that, well before he had been arrested for the murder of Simpson, he had spoken to Denning in Goulburn gaol concerning his knowledge of that crime, the case wore a very different complexion from the ordinary case of an allegation by a prison informer. It was not a case where he was denying that he had ever spoken to Denning about the matter. Apart from the dispute as to whether the applicant had merely told Denning what was alleged against the applicant, he conceded the accuracy of Denning's evidence. The applicant's statement also impliedly conceded the accuracy of Denning's evidence when Denning testified that it was the applicant who initiated the critical conversation and that the matter had been discussed on more than one occasion.

In determining the adequacy of the directions of the learned trial judge, this Court must keep firmly in mind that counsel for the applicant and his Honour had advantages which are denied to this Court. They understood the atmosphere and conduct of the trial, the tone and content of counsel's addresses to the jury, and the extent to which the Crown relied on the evidence of Denning having regard to the evidence of Jones, Jennifer Curillo and the Berrys. The effect of such matters can rarely be fully appreciated from a reading of the frozen record of the trial which comes before an appellate court, and this case is no exception. In the context of the applicant's concessions, the failure of counsel for the applicant to seek any further directions concerning the danger of acting on Denning's evidence has considerable significance.

The duty of the learned judge was to ensure that the jury fully understood the danger of acting on the evidence of Denning. Ordinarily, that duty would only be discharged by the trial judge informing the jury that it is dangerous to act on the evidence of a prison informer unless that evidence is substantially confirmed by independent evidence. That was not done in this case, but it was not quite the ordinary case. It is true that Denning was a prisoner who had been serving a life sentence since 1972 and that his evidence was subject to all the dangers which are inherent in the evidence of prison informers. But the conflict between the evidence of Denning and the statement of the applicant was a narrow one. Although it would have been better if the learned judge had given the ordinary warning, the applicant has not established that, in the special circumstances of this case, the jury would not have fully understood the danger of acting on the evidence of Denning. His Honour told them that Denning was a person of bad character who had committed many different criminal offences and that persons such as Denning "may have motives for lying". He also directed them that they must scrutinize his evidence very carefully and that they must "look to see if there is independent evidence you accept which corroborates the evidence of the witness in question". It is true that the jury would probably have understood the reference to corroboration as a reference to the applicant's involvement in the crime rather than as a reference to corroboration of the making of the confession. But in a case where the relevant issue was simply whether the applicant had told Denning what he had done or merely what he was alleged to have done and where at the trial counsel for the applicant found no fault with the direction, it cannot be said that the directions of the learned trial judge were inadequate to sheet home to the jury the danger of acting on Denning's evidence.

The Berrys gave evidence of four statements made by Dennis Allen during a period which was roughly contemporaneous with the murder of Lindsay Simpson. The trial judge admitted the statements on the basis that they showed Allen's state of mind and demonstrated that Allen was "a party to the agreement to kill Williams".

However, Allen's state of mind was not a fact in issue nor was it relevant to a fact in issue. On the trial of the applicant, it was not open to the Crown to prove an agreement between Allen and the applicant by reference to Allen's state of mind. Cases such as Walton v. The Queen [14] which permit proof of a person's intention or plan as a step in proving that the intention or plan was carried out have no application where the person's state of mind is sought to be used to prove the occurrence of a past fact or matter. If Allen had been alive, he could not have been called to give direct evidence of his state of mind in respect of those conversations which suggested that he had an agreement with the applicant. It follows that his state of mind could not be proved circumstantially by reference to his out-of-court statements. Consequently, the statements were not admissible on the ground that they showed Allen's state of mind and demonstrated that he was a party to an agreement to kill Williams. It is necessary, however, to determine whether the statements were admissible as relevant circumstances of the case or excluded by the rule against hearsay evidence.

A statement is "hearsay and inadmissible when the object of the evidence is to establish the truth of what is contained in the statement" [15] . The objection to hearsay evidence is that it is unreliable - the declarant is not subject to cross-examination and his or her truthfulness and powers of memory, recall, perception and narration cannot be tested. Because the reliability of an implied assertion is dependent upon the material expressly stated, the grounds for excluding express assertions are equally applicable to implied assertions. Consequently, for the purpose of the hearsay rule, implied as well as express assertions are regarded as "contained" in an out-of-court statement [16] .

1. Subramaniam v. Public Prosecutor, [1956] 1 W.L.R. 965, at p. 970. 2. Wright v. Doe d. Tatham (1837), 7 Ad. & E. 313 [112 E.R. 488]; Reg. v. Blastland, [1986] A.C. 41; Walton v. The Queen (1989), 166 C.L.R., at p. 292; Reg. v. Benz (1989), 168 C.L.R. 110, at p. 143; Reg. v. Kearley, [1992] 2 A.C. 228.

Allen's statements to the Berrys that they should watch the late news and that the shooting was one of "mistaken identity" tended to prove that Allen had knowledge that a murder was about to take place and that the wrong man had been murdered. But Allen's knowledge of those matters was not a fact in issue nor could it be used to prove any agreement between the applicant and Allen.

The statement "[y]ou get the rest of the money when you do the job properly" was also inadmissible. It could not be used testimonially to prove the facts expressly or impliedly asserted in that statement. Although the statement was circumstantial evidence of Allen's state of mind, viz. that he was not prepared to pay the caller for a job which had not been done, Allen's state of mind was not a fact in issue, nor, without infringing the rule against hearsay, could his state of mind as manifested by that statement be used to find that the job had not been done or that there was an agreement to do a job. The hearsay rule would be meaningless in practice if it prohibited a statement being used directly to prove a fact contained in the statement but allowed the statement to be used circumstantially to prove a state of mind from which could be inferred the existence of the very fact which could not be proved directly [17] .

1. cf. Reg. v. Blastland, [1986] A.C., at p. 54, per Lord Bridge of Harwich.

However, the statement "[y]ou get the rest of the money when you do the job properly" also constituted a promise by Allen that he would pay the caller for a service which remained to be performed. If the making of the promise at that time was a fact in issue or relevant to a fact in issue, the statement would have been admissible to prove that promise. But the statement was not admissible on either of those bases. The making of a promise by Allen, after the shooting of the deceased, to pay the applicant when the job was done was not a fact in issue. Moreover, it could only prove an earlier promise by Allen to pay the applicant to kill Williams by an inadmissible train of hearsay reasoning. Although the terms of the promise implied that there was an earlier agreement between the caller and Allen to carry out some "job", to use the statement for that purpose would be to use an out-of-court statement to prove the occurrence of a past fact or matter and would be an infringement of the rule against hearsay. The statement would not be offered as proof of a renewed promise to pay but as impliedly asserting that another agreement had been made in the past. Accordingly, the statement was not admissible on the ground that it constituted the making of a promise by Allen to pay the caller.

Furthermore, the statement and the contemporaneous identification of the caller as "Ray" were not admissible as exceptions to the hearsay rule. In the light of the widespread use of the telephone as a medium of communication and the inherent general trustworthiness of contemporaneous statements identifying telephone callers, the common law should now recognize, as an exception to the hearsay rule, the admissibility of evidence by a witness that, in his or her presence, a party to a telephone conversation identified the other party to the conversation by a statement made during or immediately before or after the conversation in those cases where the identity of the other party to the conversation is relevant. In Benz [18] , Gaudron J. and I pointed out that "notwithstanding the decision in Myers v. Director of Public Prosecutions [19] , a strong case can be made for developing and applying the common law rules of evidence by reference to the principle that hearsay evidence will be admitted when it appears to have a high degree of reliability". But the exception should be confined to categories of ordinary social and business conversations where there is no real ground for concluding that the identification was fabricated or fictitious. The identification of a party to an ordinary social or business telephone conversation by the other party, when it is made immediately before or after or during such a conversation, is ordinarily a spontaneous and trustworthy identification. Such evidence falls into a category which is sufficiently reliable to be admitted as an exception to the hearsay rule. However, identification made in the course of, or for the purpose of, a criminal venture does not seem to be sufficiently trustworthy to be admitted under this exception. Those engaged, or about to engage, in criminal ventures commonly use false names, particularly first names, to conceal the identity of those to whom they are speaking on the telephone. Moreover, it is not unknown for such persons to invoke the names of people in authority ("That was the Premier" or "That was the boss, he wants ") to give credence to their requests or importunities. Although in particular cases the application of the hearsay rule to telephone conversations made in the course of or for the purpose of a criminal venture may result in trustworthy evidence being excluded, the risk of injustice is sufficiently great to refuse to include such identifications in a new exception to the prohibition against receiving hearsay evidence. In many cases, the overheard part of a telephone conversation made in the course of or for the purpose of a criminal venture will be admissible as a statement made in furtherance of a common criminal purpose [20] . If it is not admissible on that basis, however, it should continue to be governed by the ordinary rules of evidence excluding the reception of hearsay evidence.

1. (1989) 168 C.L.R., at p. 144. 2. [1965] A.C. 1001. 3. See Tripodi v. The Queen (1961), 104 C.L.R. 1.

Furthermore, the overheard part of the conversation with the identified party should be admissible only in so far as it tends to prove, otherwise than as an admission or as hearsay, a fact in issue or a fact relevant to an issue. Because the response of the identified party has not been heard by the witness, it would not be appropriate to admit the overheard part of the conversation as an admission in cases where the identified party is also a party to the proceedings. And the mere fact that the conversation took place through the medium of a telephone and the other party to the conversation can be identified is not a reason for excluding the rule against hearsay evidence to the overheard part of the conversation.

Since it was the Crown case that the statements to which the Berrys testified were made for the purpose, if not in the course, of a criminal venture, neither the statements nor the identification of the applicant was admissible as an exception to the hearsay rule.

Finally, the statements that Allen "had given [Ray] a job to do and he had stuffed it up" and that "I have already given him $5,000, he expects me to give him another five for something he hasn't done" were pure hearsay statements and inadmissible.

Nevertheless, it is said that those statements of Allen which expressly or by implication refer to the applicant were admissible because they are to be classified as conduct or part of a pattern of conduct by Allen and not as hearsay statements. However, the rule against hearsay extends to at least some forms of conduct [21] . It would be surprising if it did not. The contrary view would mean that, on an issue of roadworthiness, the out-of-court statement by a motor mechanic that the brakes of a motor car have not been working properly would be inadmissible but his conduct in taking his children on a drive of the car immediately after making the statement would be admissible as evidence of his expert opinion that the car was roadworthy. It is unnecessary, however, to determine in this case to what extent the rule against hearsay applies to conduct.

1. Wright v. Doe d. Tatham; Chandrasekera alias Alisandiri v. The King, [1937] A.C. 220; and see Walton v. The Queen (1989), 166 C.L.R., at p. 292, per Mason C.J.

Here the "conduct" consists of nothing more than out-of-court statements by Allen which collectively assert, expressly or impliedly, that Allen had paid the applicant money to kill a person, that the wrong person had been shot, and that the applicant would get the rest of the money owed to him when the right person was killed. In so far as this "conduct" of Allen incriminated the applicant or Allen, it consisted of reports by Allen to Mr. and Mrs. Berry of what the applicant and Allen had agreed to do and of what the applicant had already done. That is to say, the evidence of the Berrys concerning this "conduct" of Allen was a hearsay account of transactions and events already concluded. In so far as the "conduct" evidenced Allen's state of mind, that was not a fact in issue or relevant to a fact in issue for the reasons I have already given. If, as is contended, the statements were simply some of the circumstances of the case which tended to prove that Allen had arranged for the applicant to kill Williams, nevertheless those circumstances consist of hearsay statements and statements showing Allen's irrelevant knowledge of the murder. If the Berrys could give admissible evidence of the statements in question, it would lead to the extraordinary result that it was also open to the Crown to call Allen, if he had been alive, to give evidence of the same statements as part of the proof that he had arranged for the applicant to kill Williams. The distinction between self-serving statements and original evidence and between hearsay statements and original evidence would be all but obliterated if a series of such statements as those in question could be admitted by classifying them as evidence of admissible conduct and not inadmissible hearsay.

Accordingly, the evidence of the Berrys concerning the statements by Allen was inadmissible.

In the course of his summing up, the learned trial judge strongly warned the jury on more than one occasion that statements made by Allen were not evidence against the applicant. Thus, his Honour said:

But they do not establish the facts asserted by Allen and cannot establish the facts asserted because they were said in the absence of the accused man. They can only be used by you as establishing Allen's state of mind when he said them and to demonstrate, if you accept them, that he, Allen, was party to the agreement to kill Williams. But they are not evidence against the accused man. Now is that clear? Very well.

Thus, the jury were clearly directed that the statements were not evidence against the applicant nor even evidence of the facts. Moreover, his Honour's question "Now is that clear?" and his comment "Very well" naturally lead to the conclusion that the members of the jury, by some form of words or conduct, acknowledged that they understood the use which they could make of the evidence. In that context, the question is whether, notwithstanding his Honour's warning and the jury's acknowledgment of their understanding of it, the admission of the evidence is likely to have prejudiced the fair trial of the charge. That question would have to be answered in the affirmative if there was any significant possibility that the jury either used the statements against the applicant or were so affected by them that they could not impartially evaluate the evidence which was admissible against the applicant. Having regard to the strength of his Honour's warnings, the first alternative can be discarded. However, the second alternative is not so easily discarded. The statements contained material which strongly incriminated the applicant. But the potential prejudicial effect of that material must be evaluated in the light of all the evidence in the case. The matters referred to in those statements were substantially in evidence independently of the evidence of the Berrys. Thus, Jones not only described the murder of Simpson by the applicant in detail but testified that the applicant had said that he was doing the killing for Allen and that, if Jones provided assistance, he would receive a large sum of money. Jones also testified that on the day after the killing of Simpson the applicant said that he had shot the wrong person and was worried about what Allen would say. Later that day Jones saw the applicant carrying a cardboard box containing a large sum of money. Jennifer Curillo testified that Allen and the applicant had talked about getting rid of Williams. She also testified that the applicant had telephoned and said: "Tell Dennis the job's done", that on the day following that call she had read a newspaper article concerning a mistaken-identity killing and that some days later she heard Allen tell the applicant in an angry voice that he had got the wrong man.

In this context, it should not be concluded that the admission of the evidence of the Berrys prejudiced the fair trial of the charge. This was a sordid case of a contract to kill a drug dealer which brought about the death of an unintended victim. Inevitably, the evidence took the jury into the brutal and ruthless world of drug dealing and contract killings. Furthermore, a strong body of evidence identified the applicant, expressly or by implication, with the callous killing of Simpson. It is difficult to believe that jurors who had heard the admissible evidence, much of which was horrifying, might have been so influenced by evidence of the statements made to or in the presence of the Berrys that they could not impartially evaluate the admissible evidence in accordance with the learned trial judge's directions. Although the evidence of the Berrys was wrongly admitted, this court should conclude that the jury followed his Honour's directions and did not use the statements as evidence of the facts or as evidence against the applicant and that the statements did not result in any miscarriage of justice.

Special leave should be granted in respect of the grounds dealt with in this judgment, but the appeal should be dismissed.