HML v The Queen

HIGH COURT OF AUSTRALIA

GLEESON CJ,
GUMMOW, KIRBY, HAYNE, HEYDON, CRENNAN AND KIEFEL JJ

Matter No A23/2007

HML  APPELLANT

AND

THE QUEEN  RESPONDENT

Matter No A19/2007

SB  APPELLANT

AND

THE QUEEN  RESPONDENT

Matter No A28/2007

OAE  APPLICANT

AND

THE QUEEN  RESPONDENT

HML v The Queen
SB v The Queen
OAE v The Queen
[2008] HCA 16
24 April 2008
A23/2007, A19/2007, A28/2007

ORDER

Matter No A23/2007

Appeal dismissed.

Matter No A19/2007

Appeal dismissed.

Matter No A28/2007

1.        Special leave to appeal granted.

2.Appeal treated as instituted, heard instanter and dismissed.

On appeal from the Supreme Court of South Australia

Representation

Matter No A23/2007

T A Game SC with C S L Abbott for the appellant (instructed by Herman Bersee)

C J Kourakis QC, Solicitor-General for the State of South Australia with S A McDonald for the respondent (instructed by Director of Public Prosecutions (SA))

Matter No A19/2007

A L Tokley with C S Gallagher for the appellant (instructed by Gallagher & Co)

A P Kimber with K G Handshin for the respondent (instructed by Director of Public Prosecutions (SA))

Matter No A28/2007

N M Vadasz for the applicant (instructed by Kyrimis Lawyers)

M G Hinton QC with K G Handshin for the respondent (instructed by Director of Public Prosecutions (SA))

Notice:  This copy of the Court's Reasons for Judgment is subject to formal revision prior to publication in the Commonwealth Law Reports.

CATCHWORDS

HML v The Queen
SB v The Queen
OAE v The Queen

Criminal law – Evidence – Similar facts – Sexual crimes – "Uncharged acts" – Relevance – Admissibility – Applicability of test in Pfennig v The Queen (1995) 182 CLR 461.

Criminal law – Evidence – Similar facts – Standard of proof – Whether "uncharged acts" must be proved beyond reasonable doubt – Directions to jury.

Practice and procedure – Application to amend notice of appeal – Whether leave should be granted to amend notice of appeal to raise issue of admissibility of evidence, to which no objection was taken at trial.

Words and phrases – "context", "guilty passion", "propensity evidence", "relationship evidence", "relevance", "similar fact evidence", "uncharged acts".

Evidence Act 1929 (SA), ss 34CA, 34I.

1. GLEESON CJ.   These matters raise issues concerning the admissibility at a criminal trial of a certain kind of similar fact evidence, and the proper directions to be given to a jury in the event that such evidence is admitted.  In each matter, the evidence was that of a complainant who, in addition to giving an account of specific acts the subject of the charge or charges in an indictment, testified that other such acts had taken place between the accused and the complainant.  This was described in argument as evidence of uncharged acts.  I am content, for the purpose of stating my reasons, to adopt the description used in argument, although I do not suggest that it would always, or even usually, be a helpful phrase in a trial judge's directions to a jury.  Of course, evidence of uncharged acts might come from a source other than the complainant; and uncharged acts of the same kind as the charged acts are themselves a particular example of evidence that reveals criminal or discreditable conduct of an accused other than the conduct with which he or she is charged.  There are wider issues involved.

2. In cases of alleged child sexual abuse, it is not uncommon for a complainant to assert that the incidents the subject of charges against the accused were part of a pattern of behaviour that extended over a period of time, perhaps many years.  There is nothing new about this kind of evidence, although in recent years the increase in reporting of, and prosecution for, child sexual abuse has drawn wider attention to some of the problems involved.  In KRM v The Queen[1], McHugh J pointed out that, in cases of sexual offences, evidence of uncharged acts between the accused and the complainant has long been admitted[2].  He said that such evidence tended to explain the relationship of the parties or made it more probable that the charged acts occurred.  In a footnote, he referred to a number of authorities, the first of which was R v Ball[3], a decision of the House of Lords in 1910.  In that case, which concerned incest, the Lord Chancellor referred to the law "which is daily applied in the Divorce Court ... to establish ... the existence of a sexual passion"[4].  His Lordship was referring to evidence of "guilty relations between the parties" in aid of proof of what was then the matrimonial offence of adultery.  In R v Hartley[5], the English Court of Criminal Appeal said, of a complainant in a case of a sexual offence, that "where a person alleges that an offence such as that with which we are concerned here has been committed against him and that the occasion was not an isolated one, he is entitled to give evidence that the offence was indulged in habitually."  The reasons why, and the circumstances in which, that is so must be examined in order to decide the present matters.

   [1](2001) 206 CLR 221; [2001] HCA 11.

   [2](2001) 206 CLR 221 at 230 [24].

   [3][1911] AC 47.

   [4][1911] AC 47 at 71.

   [5][1941] 1 KB 5 at 6-7.

3. In some Australian jurisdictions, there are statutory provisions governing these questions.  The matters before the Court (two appeals and an application for special leave to appeal) come, however, from South Australia, where it is the common law that must be applied[6].  Since there is a question of the admissibility of evidence, the logical starting point is relevance.

   [6]See, for example, R v Nieterink (1999) 76 SASR 56.

   Relevance and proof

4. Evidence is information which, according to certain governing general principles and more detailed rules, will be received by a court for the purpose of deciding issues of fact that arise for its decision.  The issues in civil cases are defined by the pleadings or other corresponding procedure.  They are determined by the principles of substantive law that apply to the dispute, and by choices made by the parties within the boundaries set by those principles.  In a criminal trial of an indictable offence, the indictment identifies the alleged offence.  The prosecution sets out to prove the elements of the offence, that is to say, the specific offence alleged to have been committed by the accused.  The jury will be directed, as a matter of law, that for a verdict of guilty it is necessary to be satisfied beyond reasonable doubt of those elements[7].  The elements of the offence, to the extent to which they are disputed, identify the facts in issue, which may be refined by particulars[8].  Depending upon the way in which the prosecution seeks to prove its case, or the way in which the defence is conducted, it may appear, as a matter of fact, that an element of the offence charged will not be established beyond reasonable doubt unless some subsidiary fact, relevant to a fact in issue, is proved to that standard.  However, the legal requirement as to onus and standard of proof is related to the elements of the offence charged.  In some cases, there may be only one available path to a conclusion of guilt, but often that is not so.  Jurors are commonly instructed that they may be selective in their approach to the evidence, and even in their approach to different parts of the evidence of the one witness.

   [7]Shepherd v The Queen (1990) 170 CLR 573 at 579-580; [1990] HCA 56. The elements, or what Dawson J described as the "essential ingredients" of the elements, of an offence are identified by statute and/or common law, and by the terms of the indictment.

   [8]It is unnecessary, for present purposes, to go into the question of the circumstances in which the prosecution will be limited by particulars of a charge or by the conduct of its case.

5. The basic principle of admissibility of evidence is that, unless there is some good reason for not receiving it, evidence that is relevant is admissible[9].  Evidence that is not relevant is inadmissible; there is then no occasion to consider any more particular rule of exclusion.  Reasons for not receiving relevant evidence may relate to its content, or to the form or circumstances in which it is tendered.  Evidence is relevant if it could rationally affect, directly or indirectly, the assessment of the probability of the existence of a fact in issue in the proceedings[10].  That directs attention, in a criminal case, to the elements of the offence charged, the particulars of those elements, and any circumstances which bear upon the assessment of probability.  The prosecution may set out to establish that an accused had a motive to commit an offence charged.  Motive may rationally affect the assessment of the probability of the existence of one or more of the elements of an offence.  Evidence that tends to establish motive, therefore, may rationally affect such assessment.  If so, it is relevant.  When the prosecution sets out to establish motive, that is often a step in the prosecution case that is not indispensable.  If it is established, motive may support (sometimes powerfully) the prosecution case, but juries are often told that failure to establish motive does not mean the case must fail.  The legal necessity is to establish beyond reasonable doubt the elements of the offence.  What that entails as a matter of fact may depend upon the circumstances of the particular case.  Some of the statements made in Chamberlain v The Queen [No 2][11] could have been interpreted as abrogating the fundamental legal principle, but what was there said was subsequently clarified in Shepherd v The Queen[12].

   [9]This principle is reflected in s 56 of the Evidence Act 1995 (Cth).

   [10]Washer v Western Australia (2007) 82 ALJR 33 at 35-36 [5]; 239 ALR 610 at 612; [2007] HCA 48; Goldsmith v Sandilands (2002) 76 ALJR 1024 at 1025 [2]; 190 ALR 370 at 371; [2002] HCA 31; cf Evidence Act 1995 (Cth), s 55.

   [11](1984) 153 CLR 521; [1984] HCA 7.

   [12](1990) 170 CLR 573.

6. Information may be relevant, and therefore potentially admissible as evidence, where it bears upon assessment of the probability of the existence of a fact in issue by assisting in the evaluation of other evidence.  It may explain a statement or an event that would otherwise appear curious or unlikely.  It may cut down, or reinforce, the plausibility of something that a witness has said.  It may provide a context helpful, or even necessary, for an understanding of a narrative.  An example is some evidence given in R v Wickham[13].  A female complainant in a child sex abuse case gave an account, directly relevant to a charge, of a sexual encounter she had with her father when she was 14 years old.  She said that her father entered her bed, and had sexual intercourse with her.  After some brief conversation, they both went to sleep.  The father denied that any such event occurred.  There was other evidence to show a history of similar sexual activity before the occasion in question.  In the absence of that evidence, the complainant's account of what otherwise would have been presented as a single, and apparently isolated, act might have been regarded by the jury as difficult to believe.  The complainant expressed no surprise when her father came to her bed.  She made no protest.  She behaved as though this was a common occurrence.  She said that, in fact, it was a common occurrence.  If she had not been permitted to say that, her evidence could have appeared hard to believe.  To have put her evidence forward as though she were describing an isolated incident would have been misleading, and, it might be added, unfair.  Jurors are told that, in evaluating evidence, they should use their common sense and their experience of life.  Whether or not expressly invited to do so, jurors are likely to assess competing versions of events or conduct by reference to their ideas of normal or predictable behaviour.  In R v Boardman[14] in a passage later cited with approval in this Court, Lord Cross of Chelsea said that there are cases in which to exclude evidence of the kind presently in question would be an affront to common sense.  The law must apply a more definite test, but common sense and relevance are closely related.  A jury's assessment of some kinds of evidence is likely to be based more upon common sense than upon scientific method.

   [13]Unreported, New South Wales Court of Criminal Appeal, 17 December 1991.

   [14][1975] AC 421 at 456.

7. Evidence of uncharged acts in child sexual abuse cases may also be relevant because of a matter mentioned above, that is, motive.  As both Deane J[15] and McHugh J[16] have said, evidence which tends to show that a father has treated a daughter as an object of sexual gratification may tend to show a motive for committing the offence charged.  If it appears that a parent has a sexual desire for a child, then that may make more credible the child's allegation that a particular alleged sexual incident occurred.

   [15]B v The Queen (1992) 175 CLR 599 at 610; [1992] HCA 68.

   [16]KRM v The Queen (2001) 206 CLR 221 at 230 [24].

8. There may be little difficulty in establishing the relevance of uncharged acts, although that is by no means the end of the question of admissibility.  Specifying the nature of the relevance may bear both upon admissibility and upon the appropriate directions to a jury.  Words such as "relationship" and "propensity" may cover both aspects of potential relevance already mentioned, but they may cover more, and may require closer definition before their application to the circumstances of a given case.  Evidence of a sexual interest of a father in a child is evidence of a certain kind of propensity, a kind of propensity that jurors may regard as bearing upon the probability that the testimony of the child as to a particular act is true.

9. As to the potential use of uncharged acts to evaluate a complainant's evidence by furnishing an explanation for apparent lack of surprise, or protest, Gaudron J said, in Gipp v The Queen[17], that evidence of general sexual abuse is relevant and admissible on that basis, but only if the conduct of the defence case raises such considerations.  I regret that I am unable to agree.  Questions of admissibility of a complainant's evidence of uncharged acts usually arise for decision either before the trial or during the evidence-in-chief of the complainant.  There may be no relevant conduct of the defence case by reference to which a decision can be made.  Furthermore, the conduct of the defence case may not be a fixed point of reference.  It is important not to overlook the legitimate opportunism that may be involved in the conduct of a defence under an accusatorial system of trial.  It is one thing to require a prosecutor to give particulars.  It is another thing to bind defence counsel to a certain line of argument.  It should also be remembered that jurors, in assessing probabilities, are not bound by the conduct of defence counsel.  When jurors evaluate the evidence of a complainant they are not limited to considering arguments advanced by the lawyers.  If the complainant's evidence concerning a charge were given as though it were an account of an isolated event, then regardless of the line taken by the defence it might create a false impression, and that impression could colour the jury's assessment of the evidence.  In some cases, the possibility is too obvious to be ignored, regardless of the line adopted in defence.  An example is provided by the evidence, in the first of these three matters, concerning the method of persuasion that the complainant was required to use in order to obtain permission to go shopping.  If she had described the conduct involved in that transaction as if it were an isolated incident it might have sounded like fantasy.  Jurors bring their ideas of normal behaviour to the assessment of probabilities.  Trial judges and advocates cannot ignore that fact, and the law of evidence must take account of it.

   [17](1998) 194 CLR 106 at 113 [12]; [1998] HCA 21.

10. It is the tendency of evidence that determines its relevance.  The trial judge decides whether evidence could rationally affect the jury's assessment of the probability of the existence of a fact in issue.  The ultimate effect of the evidence is a question of fact to be decided by the jury.

11. The kind of similar fact evidence in question, that is, a complainant's evidence of uncharged acts, even when received and used as evidence of motive, is unlikely to compel, as a matter of logic, a conclusion that the charged offence or offences occurred.  To prove that a person did something many times does not compel a conclusion that he did it again.  However, it might make it more likely that sworn testimony that he did it again is true.  People do not act in accordance with all their inclinations at every opportunity, but proof of a person's inclinations may provide strong support for direct testimony as to that person's conduct.  Decisions as to the relevance of evidence are made by asking how, if accepted, it bears on the assessment of the probability of a fact in issue.  Assessments of probability are rarely the subject of syllogistic reasoning.

    Exclusion

12. Whatever the purpose for which similar fact evidence is adduced, it has an effect which the law regards as capable of providing a good reason for excluding it:  if accepted, it shows a disposition or tendency to engage in crime or other discreditable conduct.  If that is all it shows, and the prosecution adduces the evidence for the purpose of leading to the conclusion that the accused is a person likely from his criminal conduct or character to have committed the offence for which he is being tried, then the law excludes such evidence as a matter of fundamental principle[18].  The purpose just described is often referred to as propensity, although that itself is ambiguous.  The reason for the exclusion is not the irrelevance of propensity, but its prejudicial effect.  In this context, prejudice means the danger of improper use of the evidence.  It does not mean its legitimate tendency to inculpate.  If it did, probative value would be part of prejudicial effect.  It is the risk that evidence of propensity will be taken by a jury to prove too much that the law seeks to guard against.

    [18]Makin v Attorney-General for New South Wales [1894] AC 57 at 65.

13. In addition to the possibility of prejudice just mentioned, which is common to most similar fact evidence, there is a further prejudicial effect of the kind of similar fact evidence with which we are presently concerned, that is, a complainant's evidence of uncharged acts.  Typically, as in the present matters, the uncharged acts will be disputed, and sometimes the only evidence of them will be that of the complainant.  The form in which the evidence emerges may create a serious risk of unfairness.  It may range from a general assertion that conduct similar to that the subject of the charges had occurred on other occasions, perhaps over many years, to a detailed account of other specific acts.  The accused is on trial for the charged offences.  He may seek to deal with the charges by obtaining particulars, and testing the complainant's evidence by all available forensic methods.  His capacity similarly to test the evidence of the uncharged acts may be limited.  The adversarial process by which charges are laid, particularised, and contested may be ill-adapted to an investigation of these other allegations.  This problem is not limited to a complainant's evidence of uncharged acts in sexual abuse cases.  It may arise in other forms of similar fact evidence where the alleged facts are disputed.  Questions of form, as well as content, need to be taken into account. 

14. The common law excludes evidence if its probative value is outweighed by its prejudicial effect.  Examples of prejudicial effect are given above.  The concept of probative value involves relevance and weight.  The probative value of evidence must be considered by reference to the purpose or purposes for which it is used.  In Pfennig v The Queen[19], McHugh J pointed out that prejudicial effect and probative value are incommensurables.  So, it might be said, are many other forms of competing considerations that judges routinely "weigh".  A great deal of judicial and other decision-making involves forming a judgment about where the balance is to be struck between competing considerations that are not amenable to any fixed standard of comparison. 

    [19](1995) 182 CLR 461 at 528; [1995] HCA 7.

15. To require a judgment as to what is just by taking into account probative value and prejudicial effect is the way in which the common law in England, Canada and New Zealand still deals with propensity evidence.  The authorities before 1995 were discussed by this Court in Pfennig v The Queen[20].  In Pfennig, the High Court accepted the same general principle, but refined its application to similar fact evidence in an attempt to ensure that what is to be applied is a rule of law, not a discretion, and that the rule of law provides an adequate response to the danger of unfair prejudice.  In Pfennig the issue was identity.  The truth, as distinct from the admissibility, of the similar fact or propensity evidence was not in dispute.  It was a murder case.  The accused was charged with abducting and murdering a young boy.  The evidence in question showed that he had admitted abducting and indecently interfering with another young boy on a separate, subsequent occasion.  The evidence also established that the accused met the murder victim shortly before the victim disappeared.  When the pattern of similarity, underlying unity or "signature" common to both incidents was taken into account, the later incident was cogent, circumstantial evidence pointing to the accused's guilt of murder of the first boy[21].  The propensity revealed by the second incident was used as circumstantial evidence in relation to the first incident.

    [20](1995) 182 CLR 461 at 476-480.

    [21](1995) 182 CLR 461 at 488-489.

16. The plurality judgment, of Mason CJ, Deane and Dawson JJ, accepted that the underlying necessity was to make a judgment about probative value and prejudicial effect.  They quoted what was said by Lord Cross of Chelsea in Boardman[22]:

    "The question must always be whether the similar fact evidence taken together with the other evidence would do no more than raise or strengthen a suspicion that the accused committed the offence with which he is charged or would point so strongly to his guilt that only an ultra-cautious jury, if they accepted it as true, would acquit in face of it."

    [22][1975] AC 421 at 457.

17. They also quoted Lord Mackay of Clashfern LC who said, in Director of Public Prosecutions v P[23]:

    "[T]he essential feature of evidence which is to be admitted is that its probative force in support of the allegation that an accused person committed a crime is sufficiently great to make it just to admit the evidence, notwithstanding that it is prejudicial to the accused in tending to show that he was guilty of another crime."

    [23][1991] 2 AC 447 at 460.

18. However, accepting that underlying principle, the plurality judgment went on to formulate a more specific test, which had its origin in the use of circumstantial evidence to convict.  It should be remembered that the case in Pfennig was entirely circumstantial, and the (undisputed) evidence of propensity formed part of the circumstances.  It revealed a propensity to abduct young boys for sexual purposes, a propensity which, when added to the other circumstances, was held to be conclusive of guilt of murder.  Without the circumstance of propensity, the other circumstances were inconclusive.  As noted above, other evidence in the case showed that the accused met the victim at or about the time of his disappearance.  The propensity evidence showed that the accused was a child molester.  It was thought to be very unlikely that there were two child molesters in the particular area at the time, and that the other one also had met the victim.  This, it may be noted, involves certain societal assumptions, not syllogistic reasoning.

19. The refinement of the general principle advanced in the plurality judgment in Pfennig was encapsulated in the following passage[24]:

    "Because propensity evidence is a special class of circumstantial evidence, its probative force is to be gauged in the light of its character as such.  But because it has a prejudicial capacity of a high order, the trial judge must apply the same test as a jury must apply in dealing with circumstantial evidence and ask whether there is a rational view of the evidence that is consistent with the innocence of the accused."

    [24](1995) 182 CLR 461 at 482-483 (reference omitted).

20. Since they had earlier accepted Lord Cross of Chelsea's identification of the question as one concerning the value of the similar fact evidence taken together with the other evidence, their Honours must have been speaking of "the evidence" as the similar fact evidence taken together with the other evidence[25].  That, indeed, is the way their reasoning in relation to the case before them proceeded.  If there were any uncertainty as to what their Honours meant, the surest guide to their meaning is to be found in the way they applied it to the facts. 

    [25]Phillips v The Queen (2006) 225 CLR 303; [2006] HCA 4.

21. An earlier passage in the plurality judgment stated[26]:

    "In other words, for propensity or similar fact evidence to be admissible, the objective improbability of its having some innocent explanation is such that there is no reasonable view of it other than as supporting an inference that the accused is guilty of the offence charged."

    The reference to "its having some innocent explanation" was elliptical.  The question was whether, when the propensity evidence was taken into account, there was no reasonable view of the totality of the evidence other than as supporting, with the degree of strength described in other passages, an inference that the accused was guilty of murder.

    [26](1995) 182 CLR 461 at 481-482 (reference omitted).

22. McHugh J criticised the reasoning in the plurality judgment, saying that the test propounded was impossible to relate to many well-known cases, including similar fact evidence in sexual offences.  He distinguished between cases where the use of the evidence was for a reason other than the accused's propensity and cases where the prosecution relied on propensity reasoning.  Cases in the first category, he said, such as cases where evidence of relationship simply explains other evidence that directly implicates the accused, could not be subject to the "no rational explanation" test.  The correctness of that observation seems to have been assumed in Gipp v The Queen[27], as McHugh J pointed out in KRM v The Queen[28].  In any event, there is no logical answer to this point.  Pfennig was not a case about evidence that happened to reveal propensity; it was a case about the use of the fact of propensity as circumstantial evidence in proof of the offence charged.  The use of propensity as circumstantial evidence was the key to the formulation of the refined test.  What was said in Pfennig must be understood in its context.

    [27](1998) 194 CLR 106.

    [28](2001) 206 CLR 221 at 228-233 [20]-[31].

23. There are commonplace examples of admissible evidence that reveals a criminal tendency, or discreditable behaviour, but that is not tested by reference to what might be described as the Pfennig refinement of the general principle concerning probative value and prejudicial effect.  The most obvious example is evidence of bad character that is received to contradict evidence of good character.  There are also examples of admissible evidence of motive which reveals criminal acts but has nothing to do with propensity reasoning.  Suppose D is charged with the murder of X.  Suppose the prosecution sets out to prove motive, the alleged motive being that X was blackmailing D because X had became aware that D had engaged in criminal or other discreditable conduct.  Evidence that D, to the knowledge of X, had engaged in such conduct would be relevant, as supporting the alleged motive, but the propensity revealed by such conduct may be completely irrelevant.

24. The Pfennig refinement upon the general principle as stated, for example, in Boardman does not supplant the general principle in all cases of evidence which reveals the commission of criminal offences other than the charged offences.  Where evidence of uncharged acts is introduced for the common, and acceptable, purpose of explaining that a complainant, in giving an account of conduct the subject of a charge, is not purporting to describe an isolated event, so that the account of the event may properly be evaluated by the jury, the test to be applied in determining admissibility is whether the probative value of the evidence outweighs its prejudicial effect.  Evidence may have probative value in the assistance it gives in assessing other evidence.  What is sometimes called "relationship evidence" may have value in this way.  So also may evidence of what are sometimes called res gestae.  The evidence that was held to be admissible in O'Leary v The King[29], of similar acts prior to and after the events charged, helped to explain or make intelligible the course of conduct pursued[30]. 

    [29](1946) 73 CLR 566; [1946] HCA 44.

    [30]See also Martin v Osborne (1936) 55 CLR 367 at 375; [1936] HCA 23.

25. In a sexual abuse case, a complainant's evidence of uncharged acts, admitted only for the purpose of explaining or making intelligible her account of the charged acts, or to show that she was not purporting to describe an isolated event where otherwise her account may appear implausible, need not offend rules against investigation of collateral matters or impermissible attempts to bolster a witness's credit.  It is, however, subject to the general principle concerning probative value and prejudicial effect, and the possible potential unfairness resulting from both form and content earlier discussed may affect its admissibility.

26. Evidence of uncharged acts has another potential use as evidence of motive.  The form of particular propensity involved in a sexual interest of a parent in a child could be regarded as providing a motive for conduct of the kind alleged in the charge.  This form of propensity reasoning might not be relied on, in which case it may be necessary for a trial judge to warn a jury against employing it.  Where, however, it is pursued, then the Pfennig reasoning, that is, reasoning about propensity as a circumstantial fact making more likely the offence charged, is in point. 

27. Pfennig was a case about the legitimate use of propensity reasoning, and the probative value, in such a context, of the evidence of propensity.  It expressed a test for deciding whether the evidence of propensity reached a certain level or standard of probative value.  The concept of probative value is about assessment of probabilities, which includes the reasonableness of inferences.  In deciding admissibility, the trial judge assesses the probative value of the evidence in question upon the assumption that it is accepted[31], and in the context of the other evidence.  It is a test of admissibility of evidence, not a test of the reasonableness of a jury verdict.  In the present matters, unlike Pfennig, there was direct testimony that the accused had engaged in the acts alleged in the charges.  In each case, if the evidence of the complainant about the uncharged acts were accepted, when added to the other evidence, including the direct testimony, it would have eliminated any reasonable doubt that might be left by the other evidence.  The observations of Hodgson JA in WRC[32] are in point.  The nature of the issues in each case was not such as to require a different conclusion.  There may be cases in which the nature of the dispute about the complainant's testimony, considered as a whole, is such that acceptance of the evidence of the uncharged acts is inconclusive.  These cases are not of that kind.  In each case, the probative value of the evidence of uncharged acts would have satisfied the Pfennig standard.  However, as will appear, the evidence was not left to the jury as evidence of motive, and warnings were given against propensity reasoning.  In those circumstances, while the Pfennig refinement did not apply, it was still necessary to consider whether probative value was outweighed by prejudicial effect.

    [31]Phillips v The Queen (2006) 225 CLR 303.

    [32](2002) 130 A Crim R 89 at 101-102 [26]-[29].

28. One further observation should be made about prejudicial effect.  The forms of prejudice earlier discussed are in some cases amenable to management by limiting the use to which evidence may be put, controlling the form in which it may be adduced, and giving suitable directions and warnings to juries.  If a trial judge concludes that the risk of prejudice is such as to put it beyond reasonably effective management, then the evidence should be excluded.  There may be cases in which fairness is best served by confining the evidence of uncharged acts to brief and general evidence that the occasion the subject of an alleged offence was not an isolated instance.  In Gipp v The Queen[33], McHugh and Hayne JJ referred to the possibility of a defence preference for evidence of sexual history that was given shortly and without detail.  The circumstances of particular cases will vary, and the appropriate judicial response to the requirements of fairness cannot be anticipated by a general rule save that, as already mentioned, both form and content will require consideration.

    [33](1998) 194 CLR 106 at 132 [75].

    Standard of proof

29. It is the elements of the offence charged that, as a matter of law, must be proved beyond reasonable doubt.  (I leave aside presently irrelevant cases where insanity or some other defence is raised.)  If evidence of a fact relevant to a fact in issue is the only evidence of the fact in issue, or is an indispensable link in a chain of evidence necessary to prove guilt, then it will be necessary for a trial judge to direct a jury that the prosecution must establish the fact beyond reasonable doubt; generally, however, the law as to standard of proof applies to the elements of the offence, not particular facts.  The decisions of this Court concerning corroboration in Doney v The Queen[34], and proof of lies as evidence of consciousness of guilt in Edwards v The Queen[35], illustrate the point.  Trial judges commonly, and appropriately, direct juries in terms of their possible satisfaction of particular matters relied upon by the prosecution, without referring to a standard of proof in relation to each such matter.  To do otherwise would risk error.

    [34](1990) 171 CLR 207 at 211; [1990] HCA 51.

    [35](1993) 178 CLR 193 at 210; [1993] HCA 63.

30. Where evidence is adduced for the purpose of explaining a context or similarly assisting the evaluation of the evidence of a witness, no separate question of the standard of proof of such evidence arises.  Thus, if a complainant, giving direct evidence of the facts which constitute the elements of the offence charged, says that it was not an isolated incident but part of a wider pattern of behaviour, and does so either generally or with specificity, no separate question of a standard of proof in relation to the latter evidence ordinarily would arise.

31. There is no general principle that whenever, at a criminal trial, the prosecution sets out to prove, as a fact relevant to a fact in issue, that some criminal conduct occurred, that fact must be established beyond reasonable doubt.  In the example earlier given, where certain behaviour by D is relied upon in support of an alleged motive on the part of D to murder X who was said to be blackmailing D, it would make no difference in principle whether the behaviour was criminal or whether it was otherwise discreditable.  Unless it was indispensable in the sense earlier mentioned, it would not have to be proved beyond reasonable doubt.  In the recent case of Washer v Western Australia[36], evidence was admitted to show that an accused, who was charged in connection with a certain drug importation, was in the business of drug dealing.  This was circumstantial evidence relevant to the alleged intent to supply the drugs involved in the importation.  It was not an indispensable fact; it was part of a web of circumstances.  It did not have to be established beyond reasonable doubt, or at all.

    [36](2007) 82 ALJR 33; 239 ALR 610.

32. Where a complainant's evidence of uncharged acts is relied upon by the prosecution as evidence of motive in order to support the complainant's evidence of the charged acts, two considerations may arise.  First, if that evidence is an indispensable step in reasoning towards guilt then it may be necessary and appropriate to give a direction about the standard of proof in respect of such evidence.  Secondly, it may be unrealistic, in cases such as the present, to contemplate that any reasonable jury would differentiate between the reliability of the complainant's evidence as to the uncharged acts and the complainant's evidence as to the charged acts.  That will not always be so.  There may be cases where some parts of a complainant's evidence are corroborated and others are not, or where an accused's response to part of the evidence is different from the response to other parts.  Generally speaking, however, the indispensable link case apart, it is ordinarily neither necessary nor appropriate for a trial judge to give separate directions about the standard of proof of uncharged acts.

33. The views expressed by Doyle CJ in R v Nieterink[37], which were acted upon by the trial judges in these three matters, are consistent with what is said in the preceding paragraph.

    [37](1999) 76 SASR 56.

    The present matters

34. The facts and the issues in each matter are set out in the reasons of Hayne J and Heydon J.

35. In HML v The Queen, the trial judge left the evidence of the uncharged acts to the jury, not as evidence of motive, but only as evidence of the context in which the complainant's evidence of the charged acts was to be evaluated.  I have already referred to her evidence as to asking the accused for permission to go shopping and his response.  The trial judge referred to the father's confidence to offend, and the complainant's lack of surprise or complaint.  The judge gave warnings against the use of propensity reasoning.  The evidence of the uncharged acts was admissible, and the directions were adequate.  I agree, for the reasons given by Hayne J, that the action or inaction of the Victorian authorities in relation to the uncharged acts in Victoria was irrelevant[38].

    [38]cf Washer v Western Australia (2007) 82 ALJR 33; 239 ALR 610.

36. In SB v The Queen, the evidence in question was not the subject of objection at trial.  Leave to amend the notice of appeal to raise the question of admissibility should be refused in accordance with the principles referred to in Crampton v The Queen[39].  The trial judge told the jury that the evidence was potentially helpful in evaluating the complainant's evidence of the charged acts which "may otherwise appear to be unreal or not fully comprehensible."  He directed the jury not to use propensity reasoning.  The evidence was not received or used as evidence of motive.  The directions involved no error or unfairness.

    [39](2000) 206 CLR 161; [2000] HCA 60.

37. In OAE v The Queen, the prosecution, as sometimes happens, charged the accused with the first and the last of a series of happenings.  Presumably this is done because a complainant may have a clearer recollection of the first and the last such acts, unless there is something particularly memorable about the intervening occasions.  Here again, the trial judge did not admit the evidence as evidence of motive, and warned the jury against propensity reasoning.  The directions to the jury referred to the permissible use of the evidence only as establishing a context for the evidence of charged acts.  The evidence was admissible, and the jury directions were sufficient.  In particular, for the reasons given earlier, there was no occasion to tell the jury that they could not rely on the evidence in question unless they found it established beyond reasonable doubt.

    Conclusion

38. In each of HML v The Queen and SB v The Queen the appeal should be dismissed.  In OAE v The Queen, special leave to appeal should be granted but the appeal should be dismissed.

39. GUMMOW J.   The appeals by HML and by SB were heard together and with the special leave application by OAE.  Special leave should be granted and the appeal by OAE treated as having been heard instanter.

40. All these appeals are brought from the South Australian Full Court sitting as the Court of Criminal Appeal. The issues of the law of evidence which have been argued in this Court turn upon the common law, with one qualification. This is the belated attempt, which should not succeed, made in oral argument on the appeal by HML to rely upon s 34I of the Evidence Act 1929 (SA).

41. I agree with what is written by Hayne J respecting matters of general principle.  In particular, I agree with what appears in his Honour's reasons under the heading "Pfennig v The Queen".

42. I agree with the reasons given by Hayne J for the disposition of the appeals by SB and by OAE.  With respect to the appeal by HML, my agreement has the reservation respecting the treatment of the laying of charges in Victoria which is developed by Kirby J in his reasons.  I agree with what Kirby J has written on that aspect of the appeal by HML, including the application of the proviso.  The upshot is that in this, as in the other appeals, I agree with the orders proposed by Hayne J.

43. KIRBY J.Three proceedings are before this Court.  Two are appeals, by special leave already granted, from orders of the Court of Criminal Appeal of South Australia[40].  The other is an application for special leave to appeal from orders of the same court[41].  The application was directed to be heard with the appeals because of the similarity of some of the issues raised[42].  I agree that special leave should be granted in the third matter.

    [40]In R v H, ML [2006] SASC 240 and R v S, B [2006] SASC 319.

    [41]See R v O, AE (2007) 172 A Crim R 100.

    [42]OAE v The Queen [2007] HCATrans 473.

44. The appeals have been considered together because of uncertainties that have arisen in trial and intermediate courts in respect of evidence in criminal trials involving accusations of sexual offences committed against under-aged children, commonly by family members.  In particular, the appeals present controversies relating to:

    (1)The rulings to be made in such trials in respect of the relevance and admissibility of evidence of discreditable sexual conduct involving the accused, apart from that alleged in the specific charges brought by the prosecution; and

    (2)The directions or warnings that should be given to a jury by a trial judge in such a trial, where such evidence is ruled admissible.  Such directions or warnings might relate to:  (a) the potential uses of such evidence; (b) the standard of proof to be applied by the jury in deciding whether or not they accept such evidence and whether they should use it in reasoning to their conclusion about the guilt of the accused of the offence(s) charged; and (c) the dangers of propensity reasoning based upon such evidence.

45. So much has been written about the foregoing questions in earlier decisions of this Court[43], and now in these proceedings, that I hesitate to add to the elaboration lest what I write ends up contributing to the uncertainties.  Rulings on evidence of this type must often be made by trial judges on the run, in the course of the criminal trial.  Of its nature, such a trial will often be fraught and emotional.  In addition, trial judges face great burdens in framing their directions and warnings to juries in cases of the present kind.  Such directions or warnings must be framed so as to be understood by a jury of ordinary Australian citizens who do not have the luxury of hours (still less months) of cogitation.  Therefore, this is a case where, if at all possible, this Court should make a particular effort to speak with a clear voice.

    [43]See in particular Gipp v The Queen (1998) 194 CLR 106; [1998] HCA 21; KRM v The Queen (2001) 206 CLR 221; [2001] HCA 11; Tully v The Queen (2006) 230 CLR 234; [2006] HCA 56.

46. In so far as there are differences between the opinions expressed in the reasons of Gleeson CJ, Hayne J, Heydon J, Crennan J and Kiefel J, I prefer and endorse (as Gummow J does) the principles stated by Hayne J.  I do so because I agree with Hayne J, for reasons that I will detail, about:

    (1)the purposes for which, in trials of this character, evidence of "uncharged acts" may be admitted[44];

    (2)the applicability to the admissibility of such evidence[45] of the holding of this Court in Pfennig v The Queen[46]; and

    (3)the necessity, where such evidence is admitted, for the trial judge to instruct the jury that they must be satisfied beyond reasonable doubt about the truth of such evidence if they are to use it to reason towards guilt[47].

    [44]Reasons of Hayne J at [103]-[111]; cf reasons of Gleeson CJ at [5]-[11]; reasons of Heydon J at [274]-[336], [364], [387], [390]-[394]; reasons of Crennan J at [423]-[433]; reasons of Kiefel J at [491]-[501].

    [45]Reasons of Hayne J at [106], [112]-[118]; cf reasons of Gleeson CJ at [24]-[27]; reasons of Heydon J at [289], [364], [387]; reasons of Crennan J at [455]-[467]; reasons of Kiefel J at [502]-[511].

    [46](1995) 182 CLR 461.

    [47]Reasons of Hayne J at [132], [244]; cf reasons of Gleeson CJ at [29]-[32]; reasons of Heydon J at [339], [376], [395]; reasons of Crennan J at [477]; reasons of Kiefel J at [512]-[513].

47. In particular, I agree in what I take to be Hayne J's insistence upon conformity with what was said by this Court in Pfennig (observed, for example, in the approach of the Court of Appeal of Victoria in R v Vonarx[48]) in preference (where it is different) to the approach adopted by the Court of Criminal Appeal of South Australia in R v Nieterink[49].  It was the reasoning in Nieterink that influenced the Court of Criminal Appeal of South Australia in deciding, in the ways that it did, the three appeals that are now before this Court.

    [48][1999] 3 VR 618.

    [49](1999) 76 SASR 56 at 66 [48]-[49].

48. With one exception, I also agree with Hayne J about the proper application of the relevant principles to the present appeals.  The exception relates to the exclusion, on grounds of relevance, of evidence that the appellant HML sought to tender concerning the then current state of criminal proceedings against him in Victoria.  However, this error does not affect the outcome of that appeal.  The "proviso"[50] is applicable.  HML's appeal should be dismissed.

    [50]Criminal Law Consolidation Act 1935 (SA), s 353(1); cf Liberato v The Queen (1985) 159 CLR 507 at 518; [1985] HCA 66; Gillard v The Queen (2003) 219 CLR 1 at 15 [29], 32-33 [94]-[97], 41-42 [133]-[134]; [2003] HCA 64.

49. I will also offer some additional comments about the serious inadequacies in the directions given to the jury in the appeal of OAE, both as to the use that the jury in that case might make of "uncharged acts" as part of the "context" and as to the want, there, of a sufficiently clear indication that evidence of the uncharged acts had to be proved beyond reasonable doubt.  I agree with Hayne J that the defect is not one to which the "proviso" applies.  Alike with his Honour, I would allow OAE's appeal.

50. In relation to the appeal of SB, I have nothing to add to what Hayne J has written.  I agree with the reasons and conclusions of Hayne J (including on the application for leave to enlarge the grounds of appeal).  It follows that that appeal should be dismissed.

51. These reasons will therefore explain why:

    (1)      I agree with Hayne J as to the applicable general principles;

    (2)I differ, in one respect, from Hayne J as to their application in the appeal of HML, but without dispositive consequences; and

    (3)I agree with Hayne J as to the disposition of the appeal of OAE.

52. The analysis in these reasons adopts the assumption, inherent in much appellate examination of jury decision-making, that members of a jury reach their conclusions by a process of deliberation from evidence to verdict by way of an accurate application of judicial directions on the law[51].  Such empirical evidence as there is casts serious doubts upon such assumptions[52].  Indeed, psychological research applied to judicial or other decision-making, including investigations based on the cognitive reflection test, suggests the very large role played by intuition in such decisions.  In such matters, the human brain has a tendency to make automatic, snap judgments[53].  However, in default of contrary argument, these reasons will continue to make the law's assumptions, however dubious they may be in scientific terms.

    [51]cf reasons of Heydon J at [353], reasons of Kiefel J at [488].

    [52]See Zoneff v The Queen (2000) 200 CLR 234 at 260-261 [65]-[67]; [2000] HCA 28. See also Cush and Goodman-Delahunty, "The Influence of Limiting Instructions on Processing and Judgments of Emotionally Evocative Evidence", (2006) 13 Psychiatry, Psychology and Law 110 at 113.

    [53]Guthrie, Rachlinski and Wistrich, "Blinking on the Bench:  How Judges Decide Cases", (2007) 93 Cornell Law Review 1 at 19.

    The facts and legislation

53. The facts:  The facts of each appeal are set out in considerable detail in other reasons.  Those reasons disclose the relevant objections to, and rulings on, the evidence at trial, the grounds of appeal and dispositions in the Court of Criminal Appeal in each case, and the arguments advanced in this Court.  It will be necessary for me to add a little more detail concerning the error in the trial of HML just mentioned.  However, otherwise, I am content to rely on the detailed expositions by my colleagues.

54. The legislation:  As Heydon J explains in a note to his reasons, this Court's expression of the law in these appeals is substantially confined to those jurisdictions of Australia in which the common law rule stated in Pfennig survives[54].  Other than in South Australia, the Northern Territory, and to some extent Queensland, the rule in Pfennig has been amended, either by the adoption of the Uniform Evidence Acts[55] or by the enactment of particular State legislation[56].  Subject to any constitutionally protected principles of due process, it is competent for the Parliaments of Australia to regulate the substantive and evidentiary law that is in issue in these proceedings.  No constitutional argument has been raised by any party.

    [54]Reasons of Heydon J at [288], fn 227.

    [55]See ss 97 and 98 of the Uniform Acts.  These Acts are applicable in federal courts and in the Australian Capital Territory, New South Wales, Tasmania and Norfolk Island.  See reasons of Heydon J at [288], fn 227.

    [56]Evidence Act 1977 (Q), s 132A; Crimes Act 1958 (Vic), s 398A; Evidence Act 1906 (WA), s 31A. See Washer v Western Australia (2007) 82 ALJR 33 at 46 [58]; 239 ALR 610 at 625-626; [2007] HCA 48.

55. In several jurisdictions, including South Australia, an attempt has been made to address the issues arising in these proceedings by the creation of so-called "relationship crimes"[57] and by the enactment of special evidentiary rules for cases involving sexual offences[58].  As explained in other reasons, none of these special legislative provisions is determinative of the present appeals.

    [57]See eg Criminal Law Consolidation Act 1935 (SA), s 74. See reasons of Heydon J at [259]. A similar provision was considered in KBT v The Queen (1997) 191 CLR 417; [1997] HCA 54.

    [58]See eg Evidence Act 1929 (SA), s 34I. See reasons of Hayne J at [185]-[187]; reasons of Heydon J at [337]. See also s 34CA of that Act; cf reasons of Heydon J at [310].

    Admissibility of relationship evidence

56. Factors favouring admission:  I accept that, as a matter of legal principle or policy, several considerations tend to support the reception of evidence by complainants of alleged acts of sexual abuse different from, and additional to, those identified in the charges preferred against the accused by the prosecution:

    (1)Although criminal trials address specific charges alleged in an information or indictment, the experience of the courts shows that sexual abuse of young persons is often, or typically, manifested in multiple and repeated incidents over a period of time.  It is commonly impracticable, or even impossible, to include them all among the formal charges.  The repeated character of the events may render them individually unmemorable either to the complainant or to the accused.  A court process directed to eliciting a truthful description of what has happened to a complainant will take account of such practical considerations;

    (2)Where sexual assault cases are not prosecuted under the new provisions establishing "relationship crimes", a practice is often observed by prosecutors of charging the first, or earliest, alleged incident of a sexual offence remembered by the complainant and also the most recent incident that can be described[59].  Others may be included because of special features in the facts or surrounding circumstances which are said to trigger the memory of the complainant and to permit particularity.  However, almost inevitably, and whatever the wishes and precautions of lawyers, evidence may emerge of other incidents not made the subject of charges.  This may be due to factual links between such incidents and the matters charged[60], or because such incidents are allegedly remembered whilst the complainant's evidence is being adduced.  Alternatively, the complainant, unaware of (or impatient with) the conventions of the criminal trial, may assert that many other similar instances occurred, leaving it to the trial judge to deal with the admissibility of such evidence and with the directions that should then be given.  Attempts to quarantine the charged acts may, in practice, be both artificial and futile;

    (3)From the point of view of the complainant, and respecting his or her entitlement to provide a truthful version of what is recalled, it is important for legal procedure to facilitate, so far as basic principle permits, the giving of a "fair and coherent account"[61] of what has allegedly occurred resulting in the criminal prosecution[62];

    (4)The law has an important obligation to protect truthful complainants about sexual abuse.  It is an appreciation of the significance of this consideration that led Lord Hope of Craighead to observe in R v A (No 2)[63] that "the balance between the rights of the defendant and those of the complainant is in need of adjustment if [complainants] are to be given the protection under the law to which they are entitled against conduct which the law says is criminal conduct".  This observation has particular force where the abuse has allegedly been suffered by children as a result of the conduct of family members who owe the child special duties of trust and protection;

    (5)Self-evidently, sexual assault against children is a very serious crime both in terms of its incidence in our society and in its impact on the victim, the victim's family and the community.  There is compelling evidence of historical "under-enforcement" in this area[64].  The increase in prosecutions for offences of the present kind observed by the courts in recent years is, in part, a reflection of changing community, police and prosecutorial attitudes.  These developments ought not to be permitted to be frustrated by unjustifiably restrictive court procedures; and

    (6)The retention of jury trial for most contested allegations of such offences in Australia suggests a continuing acceptance of the need to entrust decision-making in such cases to "the ordinary experiences of ordinary people"[65].  Juries resolve disputed issues and distinguish false or unproved accusations from those which they consider to have been proved to the requisite standard by applying their collective experience of life and of their fellow human beings[66].  In recent years, the House of Lords, in Director of Public Prosecutions v P[67] and R v H[68], has demonstrated a greater willingness to trust juries with sensitive evidence than, for example, was apparent in the earlier case of R v Boardman[69].  Thus, Lord Griffiths, in the case of H[70], suggested that a "less restrictive form" of the rules excluding relevant evidence was appropriate given today's "better educated and more literate juries".  So far as the common law of Australia is concerned, the result may also be a greater willingness in this country to permit jury access to relevant but sensitive, and potentially prejudicial, evidence[71].  The fact that potential prejudice may be susceptible of limitation through careful directions and warnings is an additional factor that tends to favour reposing greater trust in juries in cases such as the present.

    [59]Reasons of Gleeson CJ at [37].

    [60]As was the case in the trial of HML; see reasons of Heydon J at [318].

    [61]White v The Queen [1999] 1 AC 210 at 217 cited reasons of Heydon J at [299].

    [62]cf reasons of Crennan J at [474]-[475].

    [63][2002] 1 AC 45 at 71 [55]. See also DS v Her Majesty's Advocate [2007] UKPC D1 at [5].

    [64]See Hamer, "Similar Fact Reasoning in Phillips:  Artificial, Disjointed and Pernicious", (2007) 30 University of New South Wales Law Journal 609 at 634-635.  Hamer cites Australian Bureau of Statistics figures published in 2005 suggesting that 80% of women victims of sexual assault do not report the assault, and remarks that "[d]ata is unavailable, but under-reporting is likely to be higher still for sexual offences against children". 

    [65]Doney v The Queen (1990) 171 CLR 207 at 214; [1990] HCA 51.

    [66]cf R v Best [1998] 4 VR 603 at 611 per Callaway JA.

    [67][1991] 2 AC 447.

    [68][1995] 2 AC 596.

    [69][1975] AC 421. See reasons of Crennan J at [443].

    [70][1995] 2 AC 596 at 613.

    [71]cf reasons of Crennan J at [473].

57. Factors favouring exclusion:  As against the foregoing considerations, a number of others need to be kept in mind:

    (1)In general, criminal trials of serious offences in Australia observe an accusatorial form[72].  As a matter of law, the accused is ordinarily entitled to put the prosecution to proof of its allegations.  In the usual case, it is essential that an accused person should be informed in advance of the trial not just of the "legal nature of the offence with which he is charged but also of the particular act, manner or thing alleged as the foundation of the charge"[73].  In Australia, this has led to rules of law and practice requiring a high degree of specificity of accusations and of criminal charges[74].  To the extent that uncharged accusations or generalised "relationship evidence" intrude upon such a trial, they have a tendency to impair the right of the accused to know in advance, and to prepare to test and to meet, the particular charges alleged.  This, in turn, has the tendency to endanger a fundamental feature of the criminal trial;

    (2)From the viewpoint of the accused, the foregoing elements of the criminal trial afford important protections.  They permit the accused to prepare for the trial; to test the accusations; to assemble a defence; and (if so decided) to gather rebutting, alibi and other evidence.  They also permit the accused to object to evidence as it is tendered where it is not relevant to the issues for trial, as those issues are defined by the information or the indictment, supplemented perhaps by particulars.  To the extent that a complainant introduces other accusations and allegations that are not contained in the charges or particulars, serious prejudice may sometimes arise which it is difficult, or impossible, to cure on the run in the course of the trial;

    (3)Although the foregoing features of the accusatorial trial are particularly important in common law countries, it is arguable that a clear delimitation of criminal accusations before the beginning of any trial is a universal requirement of international human rights law.  Thus, Art 14 of the International Covenant on Civil and Political Rights states a number of basic rights by reference to the "determination of any criminal charge" against a person.  The determination of a "criminal charge" apparently postulates a degree of particularity and notice to the person accused of the exact allegation that is made;

    (4)Whilst proper attention must be addressed to the protection of complainants, so that they may place relevant testimony before the trial without artificial or irrational impediments, it is the accused, and not the complainant, who is on trial.  Ordinarily, in cases involving allegations of repeated child sexual assault, the accused faces, if convicted, serious (commonly custodial) punishment.  It is therefore the duty of courts, and of prosecutors, to ensure the fairness of the trial, especially so because accusations of criminal offences against children are specially likely to arouse feelings of prejudice and revulsion in the community which will normally be shared by jurors[75];

    (5)Uncontested evidence sometimes later proves that accusations earlier made to police about a sexual assault are false, resulting in the initial conviction of an innocent person[76].  Cases also arise where such accusations are withdrawn and disclaimed after the accused has been convicted[77].  It is not correct to assume that all such accusations are accurate and reliable; and

    (6)Although criminal appeals are necessarily conducted on the assumption that the jury understand and observe directions given to them about the law[78], there are risks, once certain evidence becomes known to the jury, that they may treat that evidence as disclosing a general disposition on the part of the accused to act as alleged in the charges.  To the extent that the common law retreats from rules withholding particular evidence from the jury, and to the extent that the law permits the jury to receive and consider such evidence although not the subject of any charge, there may be a commensurate need to enlarge the judicial obligation to direct and warn the jury about the dangers of pure propensity reasoning.

    [72]See RPS v The Queen (2000) 199 CLR 620 at 632-633 [27]-[29], 653-654 [101]; [2000] HCA 3. See also Thompson (2002) 130 A Crim R 24.

    [73]Johnson v Miller (1937) 59 CLR 467 at 489 per Dixon J; [1937] HCA 77.

    [74]Walsh v Tattersall (1996) 188 CLR 77; [1996] HCA 26. See KBT (1997) 191 CLR 417 at 429.

    [75]cf De Jesus v The Queen (1986) 61 ALJR 1 at 3; 68 ALR 1 at 4-5; [1986] HCA 65.

    [76]See R v Button [2001] QCA 133; Edwards, "Ten things about DNA contamination that lawyers should know", (2005) 29 Criminal Law Journal 71 at 73.

    [77]See eg W (1989) 44 A Crim R 363.

    [78]Gilbert v The Queen (2000) 201 CLR 414 at 420 [13]; [2000] HCA 15.

58. Conclusion on admissibility:  When all of the foregoing considerations of legal principle and policy are given their due weight, I am prepared to retreat from opinions that I earlier expressed in KBT v The Queen[79], Gipp v The Queen[80] and other cases as to the admissibility of propensity evidence, including "relationship evidence" and evidence of "uncharged acts". 

    [79](1997) 191 CLR 417.

    [80](1998) 194 CLR 106.

59. I defer to what Hayne J has written on these subjects[81].  I do so because only Hayne J's approach in these appeals gives appropriate significance, in my view, to all of the considerations of principle and policy mentioned above.  Thus, I agree that, in cases such as the present where sexual offences have been charged, "relationship evidence", including evidence about "uncharged acts", may be received as relevant to the charges against the accused provided such evidence meets the requirements of the test stated by this Court in Pfennig[82].  Where that evidence relates to other offences, different from those that are the subject of the charges concerning the same accused and complainant, the Pfennig test will ordinarily apply to such a case. 

    [81]Reasons of Hayne J at [102]-[133].

    [82](1995) 182 CLR 461.

60. The wider foundations propounded for the admission of such evidence (such as to bolster the credibility of the complainant, or to provide evidence of the general "context") would not ordinarily meet the Pfennig standard.  In my view, such evidence is not admissible simply to provide "background".  If such a vague criterion were adopted, virtually any evidence of discreditable conduct, uncharged in the information or indictment, would arguably be relevant and admissible in such a trial, because every alleged crime has a "context".  Such a rule would be destructive of the particularity of the accusatorial trial.  It would potentially be most unfair to the accused.  It would undermine the proper discipline required of prosecutors in framing accusations.  It would be damaging to the jury's central function, namely to return verdicts on the specific charges presented rather than to condemn the accused as a "nasty" or "disreputable" person.

61. Once the linchpin for admissibility of such evidence is accepted as being that stated in the test expressed in Pfennig, the foundation for the reception of "relationship evidence" and evidence of "uncharged acts" becomes clearer.  Such evidence may only be admitted if relevant to a permitted step in reasoning towards the accused's guilt of the charges framed in the information or indictment.  Once this is clear, the requirement for directions or warnings to the jury to apply the criminal standard of proof becomes plain.  That course is justified whether one invokes a metaphor and classifies the "relationship evidence" or evidence of "uncharged acts" as "links in a chain" of reasoning to guilt of the charges brought[83], or whether one views such evidence as "so intertwined with the charged acts" as to necessitate satisfaction to that standard[84].  In any such case, "the trial judge must direct the jury that they must be satisfied that the uncharged acts have been proved beyond reasonable doubt"[85].

    [83]cf Shepherd v The Queen (1990) 170 CLR 573 at 579; [1990] HCA 56.

    [84]O, AE (2007) 172 A Crim R 100 at 108 [38].

    [85]O, AE (2007) 172 A Crim R 100 at 108 [38].

62. Standing back from the mass of decisional authority mentioned in other reasons, much of it difficult to reconcile, the approach endorsed by Hayne J achieves, in my view, an appropriate adjustment of the competing considerations of legal principle and policy that I have identified.  It departs, to some extent, from the strict particularity favoured by the accusatorial tradition.  However, it acknowledges the need, where relevant, for a clear direction as to the standard of proof to be applied to uncharged acts in cases of this kind.  As well, there will often be a need for a clear warning from the judge about the dangers of pure propensity reasoning, that is, reasoning from a conclusion that the accused is a bad type of person to the conclusion that he or she is guilty of the particular offences charged.

63. In approaching the issues raised by these three appeals in jurisdictions where the common law applies unaffected by statutory modification, Australian judges should apply the principles expressed in the reasons of Hayne J.  Specifically, for the reasons Hayne J has given, a trial judge should instruct a jury "that they must only find that the accused has a sexual interest in the complainant if it is proved beyond reasonable doubt"[86].

    [86]See reasons of Hayne J at [247].

    Application of principles in HML v The Queen:  a question of relevance

64. A particular but relevant issue:  Questions of relevance can sometimes arise in cases involving accusations against the same accused of multiple sexual offences.  An instance is Phillips v The Queen[87].  There, this Court said[88]:

    "It is essential at the outset to identify the issues at the trial on which the similar fact evidence is tendered, for this is central to the identification of relevance, and to the assessment of probative force on which the admissibility of similar fact evidence depends."

    [87](2006) 225 CLR 303; [2006] HCA 4.

    [88](2006) 225 CLR 303 at 311 [26] (footnote omitted).

65. The decision in Phillips, and the way in which the issue of relevance arose in that case, have been criticised[89].  It is neither necessary nor appropriate in these appeals to address the criticism.  The circumstances in which the issues of relevance arose here were quite different.  They were argued at trial in the case of HML.  They were pressed on appeal, including in this Court.  The particular point in issue is a small and discrete one.  Because it is one upon which I depart from the conclusion of Hayne J, I will explain how it arises; why I disagree; and why the consequence is not ultimately determinative of the disposition.

    [89]Hamer, "Similar Fact Reasoning in Phillips:  Artificial, Disjointed and Pernicious", (2007) 30 University of New South Wales Law Journal 609.

66. A point reserved at trial:  The prosecution case against HML was that, whilst the two charges contained in the information concerned sexual offences against his natural daughter that occurred in Adelaide in September/October 1999, other and different sexual misconduct had begun years earlier (and continued afterwards) during visits made by the complainant to her father, then living in Victoria[90]. 

    [90]See reasons of Hayne J at [136]; reasons of Heydon J at [251].

67. The prosecutor expressly opened to the jury with the fact that HML had been interviewed by Victoria Police "in relation to the allegations of what had been taking place in Victoria".  The jury were told that they would "have an opportunity to see that interview played on a video player later in the trial".  So indeed the jury did.  On the prosecution case, the reference to the Victorian events was justified on the basis that they showed that the alleged offences in Adelaide:

    "didn't just happen out of the blue; there had already been inappropriate behaviour toward her and indeed sexual offending continued afterward.  Without knowing that, it might seem odd that the accused would suddenly commit the offences in a hotel in Adelaide.  It puts the Adelaide offending into context.  …  [I]t demonstrates that the accused was someone who actually had a sexual interest in [the complainant]; he was sexually attracted to her.  The evidence of the ongoing sexual conduct might explain the reasons for this offending.  He offended against her because he found her sexually gratifying and that sexual interest in her continued over a number of years."

68. Quite detailed evidence was then given in HML's trial, including by the complainant, about the sexual offences that allegedly occurred in Victoria.  Those offences were said to have happened both before and after the charged (Adelaide) events.  The only offences that were the subject of the trial in South Australia were those alleged to have happened in Adelaide.

69. On an initial voir dire, counsel for HML had indicated that he wished to question Detective G J Beanland of Victoria Police, to be called in the prosecution case, as to "whether or not charges [had] been laid in Victoria".   The prosecutor opposed this course on the basis that the answer would not be relevant to a fact in issue.  In his submissions to the trial judge, counsel for HML explained:

    "I would not be asking him as to why the charges didn't proceed.  But if, as the prosecution's issues [suggest], the jury are going to be hearing about uncharged acts, then it should be very plain that that's exactly what they are, otherwise it would be unfair to the accused."

    In response, the prosecutor submitted that the admission of such evidence would encourage speculation, and open "a can of worms".  The trial judge indicated that he was inclined to agree with the prosecutor, stating that "[t]here shouldn't be any questions to elicit the fact that nothing occurred in Victoria".  However, the trial judge expressly left it open to counsel for HML to make further submissions on the issue.

70. Thus it was that counsel for HML renewed his application to adduce the contested evidence when Detective Beanland was called to give evidence.  However, following short argument, the trial judge refused the application.  He stated what was, in effect, his conclusion on this point as follows:

    "The fact that [HML] wasn't charged in Victoria is not probative of the fact that he was charged here or probative as to what the outcome of this proceeding might be.  That's propensity reasoning at its worst.  …  I will be telling the jury that they are not to speculate and … I am entitled to assume that they will do as I tell them".

71. Thus, although Detective Beanland was permitted to give evidence that he had questioned HML in August 2003 at the Mount Gambier Police Station in South Australia, he was not allowed to tell the jury that, to the date of the trial in March 2006, no charges based on the alleged Victorian offending had been laid by police.

72. Suggested irrelevance of evidence:  Other members of this Court have concluded that the trial judge's ruling was correct and that the evidence that trial counsel sought to adduce was rightly excluded as irrelevant[91].  An identical conclusion was reached by the Court of Criminal Appeal[92].  With all respect to those of that view, I disagree.

    [91]Reasons of Gleeson CJ at [35]; reasons of Hayne J at [190]; reasons of Heydon J at [353]; reasons of Crennan J at [478]; reasons of Kiefel J at [515].

    [92]H, ML [2006] SASC 240 at [12]-[13]. See reasons of Heydon J at [351].

73. Reasons for relevance:  Evidence is relevant to an issue if the acceptance of it could bear on the demonstration of a matter in contention at the trial.  It is not uncommon for courts to disagree over questions of relevance[93].  Judges must commonly reach and express their conclusions on contested questions of relevance quickly and intuitively.  On this issue I certainly acknowledge the respect that is owed to the opinion of the trial judge, affirmed on appeal.  However, for several reasons, I regard that conclusion as erroneous:

    (1)Statements in the trial of HML about the Victorian allegations (and the serious criminal offences that those allegations suggested) were made in the prosecutor's opening to the jury.  Evidence about those allegations was given in the complainant's testimony.  The issue was revived in the prosecutor's closing address and in the judge's summing up.  The allegations therefore constituted an important and repeated theme in the trial.  They were deliberately introduced into the trial by the prosecution, allegedly to provide "context".  Yet although (as this Court holds) evidence of them was receivable for that purpose, the ruling of the trial judge denied HML the opportunity that he sought to attempt to neutralise the Victorian allegations as best he could;

    (2)The Victorian allegations related to alleged incidents both before and after the Adelaide visit.  According to the complainant, HML, in Victoria, would place one or two fingers in her vagina in the morning, doing so "regularly", and would also kiss her goodnight, trying to insert his tongue into her mouth in an inappropriate and suggestive fashion.  Allegedly, on at least one occasion after the Adelaide visit, HML penetrated his daughter's vagina with his penis and, separately, performed an act of cunnilingus upon her[94].  Having regard to the time when these offences were alleged to have occurred in Victoria, the report about them to Victoria Police, the investigation of the complaints by those Police, the interview of HML by Detective Beanland at Mount Gambier (conducted in conjunction with South Australian Police), and the subsequent lapse of time, a jury would arguably have been entitled to assume that (in the ordinary course of events) a decision would have been made, one way or the other, on whether or not to prosecute the offences, or at least the most significant of them.  From silence, the jury might conclude that HML had been charged, and perhaps was awaiting trial or had even been convicted upon them;

    (3)The relevance of the alleged Victorian offences was clearly regarded as established.  But if they were relevant, it was strongly arguable that the failure in the available time to prosecute such offences was also relevant.  Fairness suggests that HML should have been afforded the chance to attempt (so far as he could) to deal with such potentially prejudicial, and effectively unanswerable, evidence and statements.  The only means available to him to do so rested on the fact (undisputed in the absence of the jury) that no charges had been brought in Victoria.  There was a distinct element of inequality in permitting the Victorian evidence to be led by the prosecution but precluding HML from establishing the current status of the accusations;

    (4)The trial judge's direction to the jury that they should not "speculate" about the outcome of the Victorian allegations (which he incorrectly described in his charge as "offences") did not, in my view, neutralise the Victorian evidence.  On the contrary, such a direction was almost bound to attract the jury's curiosity about the outcome, in consequence of the specific mention of it.  Whilst it is true that a decision on the prosecution of HML in respect of the Victorian allegations depended on decisions by officials absent from the trial in South Australia, it would not have been difficult to frame a factual explanation to the jury to the effect that no Victorian charges had been brought; but that this did not prevent them being brought in the future; and that the jury should focus their attention strictly on the alleged Adelaide offences which were the only charges upon which the jury's verdicts were to be returned.  When the "uncharged" Victorian acts were given such attention in the trial, they were clearly treated as relevant to the issues in some way.  Basic fairness should then have led to acceptance of HML's submission and to permission to procure evidence on the issue from Detective Beanland.  It is difficult to deny that HML's attempted response was relevant without accepting that the entire evidence of the Victorian allegations was irrelevant and should have been excluded on that basis.  The one was an attempted qualification, albeit partial, of the other.  Rejection of HML's application was, in my view, erroneous.  The resulting error was only compounded by the direction that the judge then gave.

    [93]As this Court did in Smith v The Queen (2001) 206 CLR 650 at 656 [12]; cf at 657-659 [19]-[24]; [2001] HCA 50.

    [94]Reasons of Heydon J at [256].

74. Application of proviso:  It follows that I differ in my conclusion on this issue.  The exclusion of the evidence which HML sought to tender on this issue amounted to a "miscarriage of justice".  Prima facie it enlivens a right to have the jury's verdicts quashed and a retrial ordered. 

75. Nevertheless, under the "proviso" in South Australia it is necessary for a court, reaching such a conclusion, to proceed to consider for itself whether "no substantial miscarriage of justice" has "actually occurred".  This familiar language[95] requires this Court either to express its own conclusion on the point or to remit the question to the Court of Criminal Appeal for its decision on the issue[96].  Given the extensive consideration of the evidence by this Court, I consider that the decision on the application of the proviso can and should be made immediately.

    [95]See Liberato (1985) 159 CLR 507 at 520 citing Mraz v The Queen (1955) 93 CLR 493 at 514; [1955] HCA 59; Weiss v The Queen (2005) 224 CLR 300; [2005] HCA 81.

    [96]cf Mahmood v Western Australia (2008) 82 ALJR 372 at 379 [31]; 241 ALR 606 at 614; [2008] HCA 1.

76. The submissions of HML at trial on this point could not have resulted in the complete exclusion of the evidence of the Victorian allegations from consideration by the jury.  For the reasons already given, the jury properly had access to that evidence, being evidence of other offences admissible on the Pfennig test to demonstrate HML's sexual interest in the complainant. 

77. All that HML therefore lost by the incorrect ruling on the question his counsel propounded was the chance to present to the jury a factual integer indicating that no prosecutions had "yet" been brought in respect of any Victorian allegations.  Had such evidence been adduced from Detective Beanland, it would have been necessary for the trial judge to qualify it by explaining to the jury that no one (including Detective Beanland) knew if, or whether, any such prosecution would, or would not, be brought by the Victorian prosecution authorities, not themselves members of the police force.  The most that would have been added was a factual ingredient that would have made the instruction to the jury not to "speculate" appear more rational and understandable. 

78. Given the nature of the matters in issue in HML's trial, the absence of that integer is not a cause of a substantial miscarriage.  Nor am I convinced that, in consequence of the omission, an actual miscarriage of justice has occurred.  This issue could, and should, have been handled better.  But in the context of the ultimate focus of the trial on the Adelaide offences, it is not necessary, on this ground, to set aside the convictions based on the jury's verdicts.  Those convictions should stand.

    Application of principles in OAE v The Queen:  direction on standard of proof

79. Standard of proof:  general principles:  In his reasons, Hayne J concludes (as I also would) that whether or not evidence of "uncharged" acts is admissible is not to be determined[97]:

    "by asking whether the evidence in question will put evidence about the charges being tried 'in context', or by asking whether it describes or proves the 'relationship' between complainant and accused".

    [97]Reasons of Hayne J at [106].

80. The mistake involved in this approach is that, if it were endorsed, it would effectively allow any relevant discreditable facts to be tendered against an accused simply because such evidence threw some light on the "context" of the offences.  The risks of unfairness inherent in such an approach are obvious.  The purpose of adopting the more stringent approach set out in Pfennig is to obviate, or at least minimise, such risks in cases of the present kind. 

81. It is because the Pfennig approach is correct that, in relation to any "uncharged" acts (at least in cases of sexual offending), the jury must be charged that they have to be satisfied beyond reasonable doubt that the prosecution has proved the "intermediate facts" propounded as constituting an indispensable step on the way to an inference of guilt of the offences charged[98].

    [98]See reasons of Hayne J at [196].

494. More difficulty attends the question as to the relevance of relationship evidence where it is used in a more general way, to provide a setting or context for the offences charged[440].  The perceived need for the evidence is that the charged acts may otherwise seem unreal and not very intelligible[441].  Professor Birch distinguishes "background evidence" from that tendered as similar facts.  She emphasises the assistance it gives to the jury by putting them in the general picture.  If it involves references to prior offences, it does so because an account would otherwise be incomplete or incoherent.  It is not so much that it would be an affront to common sense to exclude the evidence, rather that it is helpful to have it and difficult for the jury to do their job if events are viewed in total isolation from their history, in her view[442]. 

     [440]See B v The Queen (1992) 175 CLR 599 at 610 per Deane J.

     [441]Gipp v The Queen (1998) 194 CLR 106 at 130 [72] per McHugh and Hayne JJ; KRM v The Queen (2001) 206 CLR 221 at 264 [134] per Hayne J; [2001] HCA 11; Tully v The Queen (2006) 230 CLR 234 at 278 [145] per Callinan J; [2006] HCA 56; R v Etherington (1982) 32 SASR 230 at 235 per Walters J; R v Wickham unreported, New South Wales Court of Criminal Appeal, 17 December 1991 per Gleeson CJ; R v Nieterink (1999) 76 SASR 56 at 65 [43] per Doyle CJ.

     [442]Commentary to case note on R v Stevens [1995] Crim LR 649 at 651.

495. A description of the work relationship evidence might do does not, however, complete the identification of its legal relevance.  I do not understand the cases dealing with relationship evidence of the kind here in question to suggest that such evidence could be admitted as part of the res gestae.  It could not be said to be relevant on account of being contemporaneous with the offences charged[443].  The inclusionary aspect of the res gestae doctrine, extending to bad disposition[444], suggests that no wide view should be taken of evidence properly falling within it.  As McHugh J reminded in Harriman v The Queen[445], if evidence is characterised as part of the transaction charged, it is not subject to any further condition of admissibility. 

     [443]Cross on Evidence, 7th Aust ed (2004) at [37001].

     [444]Cross on Evidence, 7th Aust ed (2004) at [37001].

     [445](1989) 167 CLR 590 at 633; [1989] HCA 50.

496. Some cases have admitted evidence of events extending beyond those proximate to the act charged on the basis that they make the act understandable and because the events are regarded as being bound up with the offence.  On closer examination it may be thought that the evidence in question qualified as part of the res gestae.

497. In O'Leary v The King[446] evidence was admitted of the accused's drunken and violent conduct towards fellow employees on the day leading up to the murder of one such employee.  Dixon J said that if the events of the day were not presented in evidence the murder would seem "an unreal and not very intelligible event".  But his Honour also described them as a series of connected events which could be considered as one transaction[447].  Latham CJ said that the evidence made it possible to obtain a real appreciation of the events of the day and night, but also said that the events formed "constituent parts or ingredients of the transaction itself"[448].  McHugh J in Harriman[449] appears to have viewed O'Leary as concerned with res gestae evidence.  R v Bond[450] is another case where the evidence was described generally as "necessarily admissible" and unable to be excluded "without the evidence being thereby rendered unintelligible"[451].  However, it was also described as involving acts "so closely and inextricably mixed up with the history of the guilty act itself as to form part of one chain of relevant circumstances"[452], which may suggest that it was in truth viewed as part of the res gestae.

     [446](1946) 73 CLR 566; [1946] HCA 44.

     [447](1946) 73 CLR 566 at 577.

     [448](1946) 73 CLR 566 at 575.

     [449](1989) 167 CLR 590 at 628-629; and see also R v Etherington (1982) 32 SASR 230 at 235 per Walters J.

     [450][1906] 2 KB 389.

     [451][1906] 2 KB 389 at 400 per Kennedy J.

     [452][1906] 2 KB 389 at 400 per Kennedy J.

498. It has been said that evidence may be admissible, in a general way, to show the true relationship between the parties[453].  More particularly, in cases of the kind in question, it may show that a sexual relationship exists[454].  Wilson v The Queen[455] was a case where evidence was relevant to explain a relationship between husband and wife, on the trial of the husband for the murder of his wife.  The evidence admitted showed the acrimony in their relationship, disclosed in particular by statements made by the wife as to her belief that her husband wished to kill her.  Menzies J said that the evidence was admitted because the jury would want to know what the relationship was between the husband and wife, for otherwise they would be deciding the matter "in a vacuum"[456].  But the evidence was not relevant in some general way.  It permitted an inference to be drawn, as to the husband's state of mind, such as might negative the prospect of an accidental shooting, which he had raised by way of defence.  In cases of the kind presently under consideration evidence tendered to show the relationship between the parties would provide more than context.  It would focus attention upon its sexual nature and the accused's role in its creation and continuance.  It would invite inferences to be drawn about the accused's state of mind towards the complainant and therefore his tendency.  This does not provide support for a more general role for the evidence. 

     [453]KRM v The Queen (2001) 206 CLR 221 at 230 [24] per McHugh J.

     [454]R v Vonarx [1999] 3 VR 618 at 623 [15], 625 [21]-[22]; Gipp v The Queen (1998) 194 CLR 106 at 156 [141] per Kirby J.

     [455](1970) 123 CLR 334; [1970] HCA 17.

     [456](1970) 123 CLR 334 at 344 per Menzies J, McTiernan and Walsh JJ agreeing.

499. It may be accepted that relationship evidence has long been admitted for the explanation it provides to the jury[457].  In KRM v The Queen McHugh J said that it might explain why, on the occasions charged, the complainant did not rebuff the accused or showed no distress or resentment[458].  Cases involving sexual offences against children have identified other questions which are likely to occur to a jury and require explanation.  They include:  whether the offences are isolated incidents; why the accused felt confident enough to demand the acts in question; why the child was compliant; and why he or she did not make a complaint to another person[459].  The present case of HML provides another example.  The jury would naturally be concerned to know whether any conduct of a sexual nature had preceded the alleged demands by the accused for fellatio and anal intercourse.  They would wonder at, if not be concerned by, the complainant's apparent detachment in the way she described the circumstances of the offences.  The evidence she gave, as to a statement made by the accused, makes the point.  He encountered some difficulty with penetration when having anal intercourse with the complainant and commented that it had not happened before.  The complainant was able to explain that he had had intercourse of this kind with her on previous occasions. 

     [457]KRM v The Queen (2001) 206 CLR 221 at 230 [24] per McHugh J.

     [458](2001) 206 CLR 221 at 230 [24].

     [459]Harriman (1989) 167 CLR 590 at 631 per McHugh J; Gipp v The Queen (1998) 194 CLR 106 at 113 [12] per Gaudron J, 131 [73] per McHugh and Hayne JJ; R v Etherington (1982) 32 SASR 230 at 235 per Walters J; R v Wickham unreported, New South Wales Court of Criminal Appeal, 17 December 1991 per Gleeson CJ; R v Josifoski [1997] 2 VR 68 at 77 per Southwell AJA; R v Wackerow [1998] 1 Qd R 197 at 209 per Byrne J; R v Nieterink (1999) 76 SASR 56 at 72 [76] per Doyle CJ.

500. In my view relationship evidence is relevant, but not in a general way and not by way of background or contextual evidence.  It is relevant to answer questions which, in cases of the kind under consideration, may fairly be expected to arise in the minds of the jury were they limited to a consideration of evidence of the offences charged.  So understood the basis for its admission is not to bolster the complainant's credit.  It is relevant to answer questions and thereby rebut or negative an inference which might otherwise be drawn by the jury.  In Gipp v The Queen, McHugh and Hayne JJ accepted that general relationship evidence might be admitted for a limited purpose, one which did not rely upon the accused having a sexual interest in the complainant[460].

     [460](1998) 194 CLR 106 at 131-132 [75]-[76].

501. Relationship evidence tendered for this limited purpose does not depend, for relevance, upon a question being raised by the defence.  Gaudron J in Gipp[461] accepted that issues may arise as to the complainant's lack of surprise or failure to complain, but considered that they could only be raised by the defence.  I must respectfully disagree.  Gibbs ACJ in Markby v The Queen[462] did not consider that the admissibility of evidence, relevant otherwise than as to tendency or propensity, depended upon the line taken by the defence at trial, that is, whether the accused had raised or disclaimed a particular defence[463].  The position of the defence may not be clearly exposed on cross-examination of the complainant.  It may not be until addresses that reliance is placed upon gaps in the complainant's account.  Even if the defence eschewed reliance upon what might be drawn from the absence of particular evidence from the complainant, it would not always be sufficient to settle a concern held by the jury.  In any event, if it be accepted that the evidence is relevant to meet questions which may be fairly anticipated to occur to a jury, it cannot be seen as dependent upon the course taken by the defence. 

     [461](1998) 194 CLR 106 at 113 [12].

     [462](1978) 140 CLR 108; [1978] HCA 29.

     [463](1978) 140 CLR 108 at 116-117, referring to Harris v Director of Public Prosecutions [1952] AC 694 at 710.

     Propensity evidence and admissibility

502. Relationship evidence tendered for the purpose of providing answers to the jury, in the way explained, discloses the other misconduct.  It does not, however, involve the use of any tendency of the accused, in the reasoning of the jury, so long as the jury are properly instructed.  It will be necessary, where it is relied upon for this limited purpose, for the trial judge to carefully direct the jury as to the use they can make of the evidence[464].  In BRS v The Queen McHugh J acknowledged that a direction may be effective to overcome the potential for prejudice[465].  To achieve that it will be necessary that the jury be told that they must use the evidence only to answer the questions, identified at an early point by the prosecution and accepted as relevant by the trial judge, which are considered likely to occur to them; but that they are not to use it to reason that the accused is likely to have committed the offences.  In some cases a trial judge might fairly observe that the reference to other acts, which are likely to be of the same kind as those charged, does not logically prove the prosecution case or enhance the complainant's credit.

     [464]BRS v The Queen (1997) 191 CLR 275 at 305-306 per McHugh J; [1997] HCA 47; Gipp (1998) 194 CLR 106 at 156 [141] per Kirby J.

     [465](1997) 191 CLR 275 at 310.

503. It must be accepted that relationship evidence may have dual purposes.  Where it is tendered for both, the more stringent test for admissibility must necessarily be applied.  If it is not tendered for the purpose of showing the accused's tendency it does not follow that it is inadmissible for the more limited purpose of informing the jury, according to general principles of the law of evidence.  In Bull v The Queen, McHugh, Gummow and Hayne JJ said that the fact that evidence which is relevant and legally admissible on one issue, may be logically but not legally relevant to another issue, does not make the evidence inadmissible on the first issue[466].  When an exclusionary rule of evidence otherwise applies, the trial judge will need to warn that the evidence can be used for the admissible purpose and no other[467].

     [466](2000) 201 CLR 443 at 463 [68]; [2000] HCA 24.

     [467]Bull v The Queen (2000) 201 CLR 443 at 463 [69], referring to B v The Queen (1992) 175 CLR 599 at 619.

504. This rule of admissibility does not overcome the rule of exclusion, where it is shown that the risk of prejudice to the accused is far greater than the probative value of the relationship evidence.  A risk of prejudice is most likely to arise, in cases of this kind, where specific, detailed incidents of other sexual misconduct are recounted by the complainant.  The possibility must be accepted that there may be cases where the trial judge may consider that the risk cannot be negatived or limited to an acceptable degree by a direction.  It is not possible to envisage the circumstances of every case.  But it may be expected that in most cases the evidence necessary to answer the particular questions, identified as likely to occur to the jury, will be of a general nature.  Evidence beyond that necessary to answer the question would exceed the bounds of relevance.  However, in some cases the evidence may be so general as to be objectionable on that account[468].

     [468]Gipp (1998) 194 CLR 106 at 132 [75] per McHugh and Hayne JJ.

505. In Gipp, McHugh and Hayne JJ expressed the view that relationship evidence tendered for the limited purpose did not offend the policy of the law upon which the rule of exclusion is based[469].  I respectfully agree.  The direction to the jury prohibits its use as evidence of tendency and it is therefore to be distinguished from other similar fact evidence[470].  In Pfennig it was said in the joint judgment that relationship evidence is a type of propensity evidence[471].  I take their Honours to mean that this is so when it is used as propensity evidence.  In that situation it belongs to a special class of circumstantial evidence which may attract the test there propounded[472].  The test applied in Pfennig can have no application to the limited purpose here discussed.  It has been assumed that the test did not apply[473].  In Conway v The Queen relationship evidence was considered to fall outside the special rules in Pfennig, because Pfennig dealt with the more difficult and dangerous category of similar fact evidence[474].

     [469](1998) 194 CLR 106 at 134 [84] per McHugh and Hayne JJ.

     [470]Gipp (1998) 194 CLR 106 at 112-113 [11] per Gaudron J.

     [471](1995) 182 CLR 461 at 464-465 per Mason CJ, Deane and Dawson JJ.

     [472](1995) 182 CLR 461 at 482-483.

     [473]KRM v The Queen (2001) 206 CLR 221 at 233 [31] per McHugh J; R v Nieterink (1999) 76 SASR 56 at 66 [49] per Doyle CJ; R v Vonarx [1999] 3 VR 618 at 622 [13].

     [474](2000) 98 FCR 204 at 233 [95].

506. The admission of relationship evidence to show the accused's sexual interest in the complainant clearly involves use of the accused's tendency to engage in acts with the complainant such as those charged.  Where the accused has already offended that propensity or tendency may be taken as showing a preparedness on the part of the accused to act upon it and to continue to act upon it.  It is to be recalled that in cases such as this there is usually no independent evidence to prove the acts relied upon as relationship evidence.  It is for the jury to determine whether all, or some, of the evidence is acceptable for the purpose suggested by the prosecution, assuming for present purposes that they do not accept the direct evidence of the offences given by the complainant as itself sufficient.  A finding of propensity on circumstantial evidence is one as to an intermediate fact.  In the ordinary course a jury would be instructed by the trial judge that they must only find that the accused has a sexual interest in the complainant if it is proved beyond reasonable doubt[475].  From that point they may consider that it is more probable that the accused committed the offences.

     [475]Shepherd v The Queen (1990) 170 CLR 573; [1990] HCA 56; and see Gipp (1998) 194 CLR 106 at 132 [76] per McHugh and Hayne JJ.

507. The more difficult question, when relationship evidence is tendered for this purpose, is whether the test for admissibility in Pfennig must be applied to it before it is put before the jury.  As the passage from the joint judgment in Pfennig earlier referred to[476] shows, the rationale for the test is that propensity is applied directly to a step in proof of the prosecution case.  There it was to be used to remove any question about the identification of the accused as the person who committed the murder and thereby to conclude guilt.  In cases such as the present matters its use, and the process of reasoning in which the jury are involved, is different.  The identifiable step is proof of propensity itself which, by reason of the nature of the cases, may leave little to add to the evidence of the commission of the offences themselves to conclude the question of guilt. 

     [476]See above at [489].

508. The use of the accused's tendency or propensity in Pfennig was relatively straightforward, although uncommon.  The evidence in question showed that the accused had abducted a boy on another occasion, for sexual purposes.  The tendency may have been in operation at the time of the murder of the boy in question.  Absent this evidence the prosecution could prove only that the boy had been abducted and that he had met and talked with the accused near the scene of his disappearance.  The fact of the boy's murder was to be inferred.  The critical issue before the jury was the identity of the murderer, as the joint judgment acknowledged[477].  The abnormal propensity of the accused was, clearly enough, regarded as admissible as an indicium of identity.  This is similar to the use made of the accused's propensity in Thompson v The King[478].  The accused's propensity in Pfennig excluded the possibility of another person being the murderer, although this result is perhaps not clearly stated.  It may be inferred that that was the only innocent explanation thought necessary to be excluded.  That left no rational view of the evidence as a whole consistent with the accused's innocence.  The evidence was therefore admissible.

     [477](1995) 182 CLR 461 at 475 per Mason CJ, Deane and Dawson JJ.

     [478][1918] AC 221.

509. The joint judgment in Pfennig spoke, at various points and in a general way, of the application of the test to cases involving similar facts.  Cases involving the use of relationship evidence as disclosing propensity were not, however, the subject of discussion in the reasons.  The submissions for each of the accused in the present matters did not suggest that the Pfennig test should not be applied because that decision was wrongly decided.  They assumed that the test had no application to cases of this kind.  How that was so was not fully investigated.  Nevertheless it seems plain enough that the circumstances of these cases are very different from those pertaining in Pfennig, as is the use to which the evidence is to be put and the reasoning in which the jury would be engaged.  These considerations raise the question whether the test is necessary and whether it could have any real practical operation.

510. The starting point in the application of the Pfennig test, in cases involving relationship evidence, is the assumption, on the part of the trial judge, that there is a reasonable doubt arising from the prosecution case absent the relationship evidence.  The enquiry undertaken by the trial judge, for admissibility, is whether there remains any explanation consistent with innocence when the evidence of the accused's propensity is applied.  The second assumption necessary to be made, in cases such as these, is that the jury will accept the relationship evidence in full, which of course they may or may not do, depending on the view taken of the complainant's credit and the plausibility of her account.  At this point the relationship evidence may be applied in order to determine whether there is an explanation which might be consistent with the accused's innocence, that is to say whether the reasonable doubt remains.  When applied to cases involving relationship evidence which shows the accused's propensity in relation to the complainant there will rarely be a case where an innocent explanation is left.  This result will be brought about because evidence of the offences themselves will largely be indistinguishable from the acts the subject of the relationship evidence.  The reason that the relationship evidence is highly probative is because it is of the same type and it is specifically directed towards the complainant.  It is used to establish the accused's propensity and then to reason as to the likelihood of the commission of the offences.  It may be contrasted with the situation in Pfennig, where the evidence was of a general propensity, towards boys, and it was used to identify the accused as the murderer.

511. If the rule were to be applied, I respectfully agree with the view, stated by each of Gleeson CJ[479], Hayne J[480] and Heydon J[481], that the evidence in each of the three matters would pass the test.  In my view, however, that largely follows because the test is somewhat artificial, and therefore not very useful, in its application to cases of this kind.

     [479]Reasons of Gleeson CJ at [27].

     [480]Reasons of Hayne J at [171], [216] and [234].

     [481]Reasons of Heydon J at [287], [364] and [387].

     Summary as to admissibility

512. In my view relationship evidence is admissible for two purposes.  It is admissible to show the sexual interest the accused had in the complainant at the time of the commission of the offences.  That tendency, if proved, will in most cases make it likely that the accused committed the offences.  The jury will require the usual directions with respect to the use of circumstantial evidence and clear directions as to proof of the sexual interest as an intermediate fact.  The need for and practical effect of the test for admissibility referred to in Pfennig does not suggest its application in cases such as these as obvious.  If applied, the evidence in these cases would qualify for admissibility in accordance with that test.

513. Relationship evidence is also admissible for the more limited purpose of providing answers to questions which might naturally arise in the minds of the jury, such as questions about the complainant's reaction, or lack of it, to the offences charged, or questions about whether the offences charged were isolated events.  These examples are not exhaustive.  It follows that no more evidence than is necessary to answer the enquiry could be considered relevant.  Admissibility for this purpose is conditioned by the requirement of a direction to the jury as to the limits on the use to which the evidence can be put.  Where the direction is not considered sufficient to overcome the potential for misuse of the evidence, perhaps because of the nature of the evidence, it should not be admitted on this ground.

     Determination of the cases

514. Details of the summings up in each of the three cases are provided in the reasons of Hayne J, Heydon J and Crennan J.

     HML v The Queen

515. The directions provided by the trial judge were consistent with the use of the relationship evidence for the limited purpose discussed.  His Honour the trial judge identified it as providing background and an explanation as to why the accused was confident enough to demand oral sex and have intercourse with the complainant, and why she acquiesced.  Insofar as the existence of a relationship was said to be disclosed by the evidence, its relevance was attached to the question about what it might explain about the circumstances of the offences.  The fact of the accused having a sexual interest in the complainant was, in any event, governed by his Honour's direction requiring proof to the criminal standard.  His Honour gave warnings about the use of propensity reasoning.  The fact that no charges had been laid in Victoria, with respect to some of the acts forming part of the relationship evidence, was irrelevant and nothing more was required to be said by his Honour about that.  Leave to amend the notice of appeal should be granted and the appeal dismissed.

     SB v The Queen

516. Leave to amend the notice of appeal should be refused because it raises matters with respect to evidence which was not the subject of objection at trial.  The trial judge explained to the jury that the evidence was relevant only to assist them in understanding how a particular incident came about and gave a strong warning against the use of it as evidence of the accused's propensity.  The directions were appropriate and sufficient.  This appeal should be dismissed.

     OAE v The Queen

517. This matter requires special leave to appeal.  That leave should be granted.  The trial judge directed the jury that the evidence showed the nature of the relationship between the accused and the complainant.  More specifically, his Honour explained that its purpose was to show that the event the subject of the second and third offences charged, which took place some four years after the first, would have otherwise appeared to have "happened out of the blue, so to speak".  No reliance was placed upon the evidence as disclosing the accused's sexual interest.  His Honour in any event directed the jury, in clear terms, not to use the evidence to reason to the guilt of the accused by reference to what it disclosed about the accused.  The directions were sufficient.  The appeal should be dismissed.