//
© 2026 Zoe. All rights reserved.
Zoe is a legal information platform. Always consult the Federal Register of Legislation for authoritative text.
[2020] HCA 12
appellant. Special leave granted; appeal allowed; Court of Appeal orders set aside; appeal to Court of Appeal allowed; convictions quashed and judgments of acquittal entered.
The High Court unanimously held that Cardinal Pell's convictions could not stand because, even if the choirboy's evidence was accepted as truthful, other reliable unchallenged evidence about how Mass ended made the alleged assaults so improbable that any rational jury would have harboured reasonable doubt. The Court stressed that an appeal court does not ask whether the offending was impossible or whether the complainant might still be right; it asks whether the whole body of evidence left open a reasonable possibility that the accused was innocent. Because the evidence of the applicant's movements, constant accompaniment while robed, and the busy sacristy after Mass could not be reconciled with the complainant's story, the convictions were unreasonable and acquittals were entered.
AI-generated legal information, not legal advice. Zoe can make mistakes. Check the cited source, and for advice about your situation consult a qualified Australian lawyer.
2,270 words · generated 24/04/2026
4 cited instruments linked from this judgment.
The applicant, George Pell, was installed as Archbishop of Melbourne on 16 August 1996. The Cathedral itself was closed for renovations until late November. The first Sunday solemn Mass he celebrated there was on 15 December 1996, with a further Mass on 22 December 1996. A third relevant occasion was 23 February 1997 when he presided but did not celebrate. Two 13-year-old choirboys, A and B, were members of the choir. A alleged that on one of the December dates, immediately after the external procession, he and B slipped away from the choir procession near the metal gate leading to the toilet corridor, re-entered the Cathedral through the south transept double doors, made their way along the sacristy corridor into the unlocked priests' sacristy, and began drinking altar wine. The applicant, still in full ceremonial vestments and allegedly unaccompanied, appeared in the doorway, challenged them, exposed his penis, orally raped B (charge one), orally raped A (charge two), indecently assaulted A by touching his penis and testicles (charge three) and masturbated himself (charge four). The entire episode was said to have lasted five to six minutes. A further month later, during an internal procession after Mass on 23 February 1997, the applicant allegedly pushed A against the wall of the sacristy corridor and squeezed his genitals painfully (charge five).
A first complained in June 2015. By then B had died; in 2001 B had told his mother he had never been interfered with in the choir. The prosecution case at the second trial (the first jury having been unable to agree) rested entirely on acceptance of A's evidence. The applicant did not give evidence but had denied the allegations in a 2016 police interview, asserting that rudimentary inquiry would show they were "fundamentally improbable". The prosecution called 23 witnesses, many of whom were "opportunity witnesses" describing post-Mass routines. After conviction on all counts, Pell appealed to the Victorian Court of Appeal on the ground that the verdicts were unreasonable. The majority (Ferguson CJ and Maxwell P) viewed the video recordings of A and several other witnesses, found A compellingly credible, and held that none of the "solid obstacles" to conviction required the jury to doubt his account. Weinberg JA dissented, concluding that the unchallenged opportunity evidence meant a reasonable jury ought to have entertained doubt. Special leave was granted and the High Court, in a single joint judgment, allowed the appeal, quashed all convictions and directed entry of judgments of acquittal.
The High Court began from the authoritative statement in M v The Queen that an appellate court must ask whether, on the whole of the evidence, it was open to the jury to be satisfied beyond reasonable doubt. The Court emphasised that this inquiry proceeds on the assumption that the jury found the complainant credible and reliable. The task is not to substitute the court's own view of demeanour but to examine whether, notwithstanding that assumed finding, the evidence as a whole left the jury with a rational doubt.
Central to the reasoning was the "compounding improbabilities" arising from three bodies of unchallenged or substantially unchallenged evidence. First, the applicant's "meet and greet" practice: Portelli gave unchallenged evidence that on both 15 and 22 December 1996 the applicant stood on the Cathedral steps greeting congregants for at least ten minutes, after which Portelli accompanied him back to the priests' sacristy. Potter, McGlone, Connor, Finnigan and several choristers gave mutually supporting evidence of the same practice. The respondent conceded that if the applicant had been on the steps for ten minutes the first four offences could not have occurred. The Court held that the evidence could not rationally be dismissed merely because memories of repeated rituals might blur after twenty years; the very repetition and the defining importance of liturgical regularity gave the evidence heightened probative force.
Second, the requirement that an archbishop never be unaccompanied while robed: Portelli, Potter, McGlone, Connor and the Dearings all gave evidence, again largely unchallenged, that the applicant was never left alone in robes inside the Cathedral. This reflected centuries-old church law and the specific duties of the master of ceremonies. The handful of occasions on which choirboys recalled seeing the applicant robed without identifying an accompanying person were held insufficient to displace Portelli's specific recollection of the December Masses.
Third, the "hive of activity" in the priests' sacristy: after the procession the altar servers entered, bowed to the crucifix, and then spent ten to fifteen minutes ferrying sacred vessels from the sanctuary under Potter's supervision. The sacristy was therefore not left unlocked and unattended. The Court rejected the Court of Appeal majority's hypothesis that the assaults could have occurred in a five-to-six-minute window of private prayer time, noting that the timing did not align, that A and B would likely have encountered the altar servers, and that concelebrant priests would also have been present.
The joint judgment repeatedly observed that the prosecution had never sought to prove the offending was impossible; rather, the defence had never been required to prove impossibility. By focusing on whether each obstacle left A's account "realistically possible", the Court of Appeal majority had impermissibly reversed the burden. Once the compounding effect of the three categories of evidence was recognised, the only rational conclusion was that the jury, acting rationally, must have entertained reasonable doubt. The same reasoning applied to charge five. Full weight was given to the jury's advantage, yet the improbabilities were "solid obstacles" that could not be overcome. The forensic disadvantage occasioned by the twenty-year delay was also expressly taken into account.
Prior to this judgment the governing test for unreasonable verdicts had been settled by M v The Queen (1994) 181 CLR 487 and elucidated in Libke v The Queen. Appellate courts were required to make an independent assessment of the record while respecting the jury's constitutional role. SKA v The Queen had already made clear that the mere existence of video recordings did not oblige an appellate court to watch them. What Pell clarified was the application of those principles to a case in which the prosecution case rested on a single complainant whose account was met by a powerful body of unchallenged evidence of invariable institutional routines. The Court rejected any suggestion that the accused must prove "impossibility" to raise a doubt. It also elevated the probative value of evidence of habit and religious ritual, noting that "adherence to ritual and compliance with established liturgical practice is a defining feature of religious observance".
After Pell, Australian courts must, when confronted with delayed complaints against public figures or within highly regulated institutions, carefully weigh whether the whole of the evidence, including evidence of routine, excludes reasonable doubt. The judgment reinforces that appellate review is not a mere rubber stamp of a jury's acceptance of a complainant. It underscores that a complainant's apparent credibility cannot, without more, overcome a body of evidence that, taken as a whole, leaves open a reasonable possibility of innocence. The forensic-disadvantage direction required by Victorian statute was integrated into the appellate analysis rather than treated as a mere jury instruction. The decision has therefore tightened the practical operation of the unreasonable-verdict ground in single-complainant historic cases.
"[T]here is 'a significant possibility that an innocent person has been convicted because the evidence did not establish guilt to the requisite standard of proof'." This is the Court's ultimate conclusion. In plain English it means that, no matter how believable A seemed, the other evidence left too great a chance that Pell was innocent.
"The function of the court of criminal appeal ... proceeds upon the assumption that the evidence of the complainant was assessed by the jury to be credible and reliable. The court examines the record to see whether, notwithstanding that assessment ... the jury, acting rationally, ought nonetheless to have entertained a reasonable doubt as to proof of guilt." Translation: even if we accept the jury thought the victim was telling the truth, we still have to look at everything else and ask whether any reasonable group of jurors would have been left with doubt. If yes, the conviction cannot stand.
"Evidence of a person's habit or practice of acting in a particular way to establish that the person acted in that way on a specific occasion may have considerable probative value. ... The evidence of religious ritual and practice in this case had particular probative value." In plain language: when people follow the same religious routine every week for years, evidence that they did so on the day in question is especially believable. Twenty years later it is still powerful because the ritual itself is so fixed.
"Portelli's evidence was unchallenged. ... Their Honours were required to take into account the forensic disadvantage experienced by the applicant arising from the delay of some 20 years." Translation: the prosecution never suggested Portelli was lying or mistaken on the critical points. The appeal court could not simply discount his evidence because of the passage of time; the delay itself made it harder for the defence to test the allegations and that disadvantage had to be weighed in the applicant's favour.
This authority is engaged whenever a conviction for sexual offending rests substantially or wholly on the uncorroborated evidence of a single complainant and is answered by unchallenged or substantially unchallenged evidence from multiple witnesses describing invariable institutional routines or practices that render the offending highly improbable. Typical triggers include: (1) delayed complaint measured in decades, giving rise to a forensic-disadvantage direction under statutes such as the Jury Directions Act 2015 (Vic) s 39; (2) an institutional setting (cathedral, school, boarding house, religious order) with well-documented liturgical, security or supervisory protocols; (3) "opportunity witnesses" who describe the accused's movements, accompaniment requirements or the constant presence of others in the alleged location; (4) evidence that the accused was never left alone while robed or in ceremonial dress; (5) a "hive of activity" or continuous traffic inconsistent with a private five-to-six-minute assault; and (6) mutually reinforcing testimony from witnesses whose honesty is not impugned.
The precedent is not limited to clergy; it applies to any case in which "compounding improbabilities" arise from evidence of habit. It is especially relevant where the prosecution opens on the basis that the routines left open a "realistic possibility" but then fails to challenge the detailed evidence of those routines. Conversely, if the complainant’s account is independently corroborated or the opportunity evidence is successfully contradicted in cross-examination, the Pell analysis will not compel acquittal.
Although the judgment itself does not refer to subsequent decisions, it carefully positions itself as applying without departure the test in M v The Queen and the approach to video evidence in SKA v The Queen. The joint judgment repeatedly returns to the language of M ("must as distinct from might have entertained a doubt") and treats Libke as merely an elucidation, not a modification. It therefore presents itself as orthodox rather than revolutionary. The Court expressly disapproves of the Court of Appeal majority's method of analysing each obstacle separately and asking only whether A's account remained possible. That approach is characterised as reversing the onus. Later benches are thus directed to examine the cumulative force of the improbabilities rather than to compartmentalise them.
The judgment also integrates the statutory forensic-disadvantage direction into the appellate reasonableness assessment, treating the delay not merely as a jury instruction but as a factor that must inform the independent appellate evaluation of the evidence. Courts are reminded that evidence of religious or institutional ritual carries "particular probative value" because of the defining importance of regularity. The decision therefore stands as a High Court restatement that appellate courts must not allow subjective belief in a complainant's demeanour to override a body of objective, unchallenged evidence that leaves open a reasonable possibility of innocence.
The judgment leaves several matters for future resolution. First, the precise boundary between permissible appellate viewing of video recordings and impermissible usurpation of the jury's demeanour-based advantage remains open. The Court states that such viewing should be "exceptional" and driven by a "real forensic purpose" identified by the parties, yet does not prescribe a bright-line test for when that purpose exists. Second, the weight to be given to evidence of habit in non-religious institutional settings (hospitals, schools, military units) is not exhaustively addressed, although the underlying Wigmore-derived principle is endorsed. Third, the interaction between the Pell compounding-improbabilities analysis and the Jury Directions Act provisions concerning "impossibility" arguments (ss 4A, 39) is expressly left undecided; the Court noted it was unnecessary to decide whether the respondent's concession about charge five contravened those sections.
Fourth, the Court does not decide whether A's account of the second, briefer incident was so inherently improbable as to require doubt even standing alone; it preferred to rest on the same opportunity evidence. Future cases will need to determine how far a complainant's evidence can be tested for inherent implausibility when opportunity evidence is weaker. Finally, the precise forensic weight to be accorded to a complainant's detailed knowledge of a location (the priests' sacristy) is left in an ambiguous state: the Court holds it is not corroboration yet acknowledges it may enhance credibility. How trial judges should direct juries on that distinction remains for later elaboration. These open questions ensure that Pell v The Queen will continue to generate litigation at the intersection of complainant credibility, institutional routines and the appellate unreasonable-verdict function.
(b) the appellant's convictions be quashed and judgments of acquittal be entered in their place.
B W Walker SC with R B Shann for the applicant (instructed by Galbally & O'Bryan)
K E Judd QC with M J Gibson QC and A S Ellis for the respondent (instructed by Office of Public Prosecutions Victoria)
Notice: This copy of the Court's Reasons for Judgment is subject to formal revision prior to publication in the Commonwealth Law Reports.
Criminal law - Sexual offences against children - Appeal against conviction by jury on ground that verdict unreasonable or cannot be supported having regard to whole of evidence - Where prosecution case wholly dependent upon acceptance of truthfulness and reliability of complainant's account - Where jury assessed complainant's evidence as credible and reliable - Where witnesses gave unchallenged evidence of specific recollections, practices and routines inconsistent with acceptance of complainant's account ("unchallenged inconsistent evidence") - Where Court of Appeal required to take into account forensic disadvantage experienced by applicant - Whether prosecution negatived reasonable possibility that applicant did not commit offences - Whether Court of Appeal required applicant to establish offending impossible to raise reasonable doubt - Whether unchallenged inconsistent evidence required jury, acting rationally, to have entertained doubt as to applicant's guilt.
Criminal practice - Appeal - Video evidence - Where evidence of complainant and other witnesses recorded - Where Court of Appeal viewed recorded witness testimony - Whether proper discharge of appellate court's function necessitated review of recorded witness testimony.
Words and phrases - "beyond reasonable doubt", "compounding improbabilities", "credibility and reliability", "function of the appellate court", "function of the jury", "impossibility", "improbability of events", "invariable practice", "jury's advantage in seeing and hearing the witnesses", "negatived the reasonable possibility", "opportunity witnesses", "realistic opportunity for the offending to have occurred", "religious ritual", "routines and practices", "significant forensic disadvantage", "significant possibility that an innocent person has been convicted", "solid obstacles to conviction", "standard and burden of proof", "unchallenged evidence", "uncorroborated", "video-recordings of the witnesses at trial".
Crimes Act 1958 (Vic), ss 45(1), 47(1). Criminal Procedure Act 2009 (Vic), ss 276(1)(a), 378, 379(b)(i). Judiciary Act 1903 (Cth), s 37. Jury Directions Act 2015 (Vic), ss 4A, 39.
KIEFEL CJ, BELL, GAGELER, KEANE, NETTLE, GORDON AND EDELMAN JJ. On 11 December 2018, the applicant was convicted following a trial before the County Court of Victoria (Chief Judge Kidd and a jury) of one charge of sexual penetration of a child under 16 years and four charges of committing an act of indecency with or in the presence of a child under the age of 16 years. The offences charged in the first four charges were alleged to have been committed on a date between 1 July and 31 December 1996. The fifth charge was alleged to have been committed between 1 July 1996 and 28 February 1997. All the offences were alleged to have been committed in St Patrick's Cathedral, East Melbourne ("the Cathedral"), following the celebration of Sunday solemn Mass and within months of the applicant's installation as Archbishop of Melbourne. The victims of the alleged offending were two Cathedral choirboys, "A" and "B". Procedural history A made his first complaint about the alleged assaults in June 2015. The prosecution case was wholly dependent upon acceptance of the truthfulness and the reliability of A's evidence. By the time A made his complaint, B had died in accidental circumstances. In 2001, B had been asked by his mother whether he had ever been "interfered with or touched up" while in the Cathedral choir. He said that he had not. This was the second trial of these charges, the jury at the first trial having been unable to agree on its verdicts. The applicant sought leave to appeal against his convictions to the Court of Appeal of the Supreme Court of Victoria (Ferguson CJ, Maxwell P and Weinberg JA). He was granted leave on a single ground (ground 1), which contended that the verdicts were unreasonable and could not be supported by the evidence. The members of the Court of Appeal viewed the recording of A's evidence, and that of a number of other prosecution witnesses. The majority, Ferguson CJ and Maxwell P, assessed A as a compellingly credible witness. There was evidence, adduced in the prosecution case from witnesses described as "the opportunity witnesses", with respect to the applicant's and others' movements following the conclusion of Sunday solemn Mass, which was inconsistent with acceptance of A's account. Their Honours concluded that no witness could say with certainty that the routines and practices described by the opportunity witnesses were never departed from. Their Honours reviewed a number of "solid obstacles" to conviction and in each case concluded that the jury had not been compelled to entertain a doubt as to the applicant's guilt. Weinberg JA, in dissent, considered that, in light of the unchallenged evidence of the opportunity witnesses, "the odds against [A's] account of how the abuse had occurred, would have to be substantial". His Honour concluded that the jury, acting reasonably on the whole of the evidence, ought to have had a reasonable doubt as to the applicant's guilt. The applicant applied for special leave to appeal from the judgment of the Court of Appeal on two grounds. The first proposed ground contends that the Court of Appeal majority erred by finding that their belief in A required the applicant to establish that the offending was impossible in order to raise and leave a doubt. The second proposed ground contends that the Court of Appeal majority erred in their conclusion that the verdicts were not unreasonable as, in light of findings made by their Honours, there remained a reasonable doubt as to the existence of any opportunity for the offending to have occurred. On 13 November 2019, Gordon and Edelman JJ referred the application for special leave to appeal to a Full Court for argument as on an appeal. The application was heard on 11 and 12 March 2020. Disposition For the reasons to be given, it is evident that there is "a significant possibility that an innocent person has been convicted because the evidence did not establish guilt to the requisite standard of proof". Special leave to appeal should be granted on both grounds and the appeal allowed. The respondent submitted that, in the event special leave were granted and the appeal allowed, the matter should be remitted to the Court of Appeal or relisted before this Court so that the whole of the evidence might be placed before it. The respondent's submission with respect to the consequential order is rejected. The submission that the Court does not have before it the material to enable it to determine whether the verdicts are unreasonable or cannot be supported by the evidence is specious. Each party placed before the Court all the evidence that it considered relevant to the determination of the applicant's proposed second ground of appeal and each party addressed written and oral submissions as to the inferences to be drawn from it. This Court is empowered to give, and should give, such judgment as ought to have been given by the Court of Appeal. As will appear, the Court of Appeal majority's findings ought to have led to the appeal being allowed. It follows that the order of the Court of Appeal must be set aside and in its place the appeal to that Court allowed, the applicant's convictions quashed and verdicts of acquittal entered. The layout of the Cathedral and the conduct of processions Before outlining A's allegations, there should be reference to aspects of the layout of the Cathedral and its surrounds, and to the manner in which the applicant, his attendants and the choir ordinarily processed from the Cathedral at the conclusion of Sunday solemn Mass at the time of the alleged offending. Adjoining the Cathedral at the rear is a modern building called the "Knox Centre". A metal gate at the eastern end of the Cathedral on its southern side opens into a short corridor between the Cathedral and the Knox Centre. The corridor gives access to several toilets and was referred to as the "toilet corridor". At the end of the toilet corridor, a glass door opens onto a small vestibule. To the right, as one passes through the glass door from the toilet corridor, there is a fire door which gives access to the Knox Centre. The choir room, in which the choir robed, was located in the Knox Centre. To the left, as one passes through the glass door from the toilet corridor, there is a door which gives access to another corridor ("the sacristy corridor"). The sacristy corridor is within the Cathedral but is not open to the public. On the southern side of the sacristy corridor, closest to the vestibule, is the archbishop's sacristy. The priests' sacristy is next to the archbishop's sacristy. Opposite both is a room which at the time was known as the "utility room" and was used by the altar servers as a robing room. Access to the sacristy corridor can also be gained through a set of double doors at the eastern end of the south transept. After the dismissal at the end of Sunday solemn Mass, the choir, attendants and clergy formally processed down the Cathedral's centre aisle to the great west door. The procession was led by the altar servers; next came the choir comprising around 50 boys, ranging from grade 3 to grade 12, and around 12 adult male singers; they were followed by the choirmaster, and concelebrant and other priests. The applicant, with his master of ceremonies, Monsignor Portelli, came last. They were accompanied by two further altar servers. The choir processed in file two-by-two in strict order with the sopranos first, followed by the altos, the tenors, the basses and the adult choristers. The choir marshal ensured that the procession maintained discipline. On fine days, the procession proceeded out through the west door. It was the applicant's practice to leave the procession at this point and remain on the steps of the Cathedral, with Portelli, to greet congregants as they were leaving. This "meet and greet" was estimated to have taken between ten minutes and half an hour on an ordinary day on which the applicant did not have an engagement in the afternoon. Meanwhile, the procession turned to its left and processed around the southern side of the Cathedral to the metal gate and into the toilet corridor. The choristers returned to the choir room in the Knox Centre and the altar servers and priests entered the Cathedral through the sacristy corridor. The altar servers entered the priests' sacristy and in formation bowed to the crucifix, thereby marking the formal end of the proceeding. If the weather was inclement, the procession processed down the centre aisle to the west door and processed back along the southern aisle, through the double doors in the south transept and into the sacristy corridor. The choristers walked along the sacristy corridor through the door to the vestibule and into the choir room in the Knox Centre, while the altar servers entered the priests' sacristy and bowed to the crucifix. A's evidence A and B were aged 13 years at the time of these events. A was a soprano. It was his evidence that, following Sunday solemn Mass, he and B had broken away from the procession at a point when it was approaching the metal gate to the toilet corridor. The two of them had slipped away and gone back into the Cathedral through the door to the south transept. The double doors from the south transept to the sacristy corridor were unlocked and they made their way down the corridor to the priests' sacristy, which was unlocked. They went inside and were "poking around". In a cupboard in an alcove they found a bottle of red altar wine. They had barely taken a couple of swigs from the bottle when the applicant appeared in the doorway. He was standing alone in his robes. He challenged them, saying, "[w]hat are you doing in here?" or "[y]ou're in trouble". A and B froze. The applicant undid his trousers and belt and started "moving ... underneath his robes". The applicant pulled B aside, took his penis out and lowered B's head towards it. A saw the applicant's hands around the back of B's head. B was crouched before the applicant and his head was down near the applicant's genitals (charge one). B said "[c]an you let us go? We didn't do anything." This assault took place for "barely a minute or two". Next, the applicant turned to A, pushing him down into a crouching position. The applicant was standing and his penis was erect. He pushed his penis into A's mouth. This assault took place over a short period of time that "wouldn't have been any more than 2 minutes" (charge two). The applicant then instructed A to undo A's pants and to take them off. A dropped his pants and underwear and the applicant started touching A's penis and testicles (charge three). As he was doing this, the applicant used his other hand to touch his own penis (charge four). The applicant was crouched almost on one knee. These further acts of indecency occupied "a minute or two". A and B made some objections but did not quite yell out. They were sobbing and whimpering. The applicant told them to be quiet, in an attempt to stop them crying. After the applicant stopped, A gathered himself and his clothing. He and B re-joined some of the choir, who were mingling around in the choir room and finishing up for the day. A and B then left the Cathedral precinct. A recalled that they were picked up by his parents or B's parents. He did not complain to anyone, including his parents, about the incident. Nor did he ever discuss the offending with B. At least a month after the first incident, again following Sunday solemn Mass at the Cathedral, A was processing with the choir back along the sacristy corridor towards the Knox Centre (the procession on this occasion was evidently an internal one). After A passed the doors to the priests' sacristy, but before reaching the door to the archbishop's sacristy, the applicant appeared and pushed A against the wall and squeezed his testicles and penis painfully. The applicant was "in his full regalia". The assault was fleeting. A did not say anything nor did he tell B about this second incident (charge five). A was uncertain of the date of each incident. He believed that both had occurred following a Sunday solemn Mass celebrated by the applicant in the second half of 1996, before Christmas. He maintained that the two incidents were separated by at least one month. The celebration of Sunday solemn Mass following the applicant's installation as Archbishop of Melbourne The applicant was installed as Archbishop of Melbourne on 16 August 1996 at a ceremony held in the Exhibition Building. The Cathedral was closed from Easter until the last week of November 1996 while renovations were being completed. The archbishop's sacristy was not available for the applicant's use throughout the period of the alleged offending. The applicant used the priests' sacristy to put on and remove his vestments in this period. Portelli and any other priests also used the priests' sacristy for robing. The first occasion on which the applicant celebrated Mass at the Cathedral was the vigil of Christ the King on the evening of Saturday, 23 November 1996. The first time the applicant celebrated Sunday solemn Mass in the Cathedral was on 15 December 1996. The only other occasion on which the applicant celebrated Sunday solemn Mass in the Cathedral in 1996 was on 22 December. The next occasion on which the applicant was present in the Cathedral for the celebration of Sunday solemn Mass was on 23 February 1997. The occasion was unusual in that the celebrant was Father Brendan Egan and not the applicant. The applicant presided at the Mass, a role which did not require him to speak. When presiding at solemn Mass the applicant wore his "choir robes": a purple cassock, which was worn under a white garment called a "rotchet" that extended down to the knees, and over which the applicant also wore a short purple cape. When celebrating solemn Mass, the applicant wore an alb, which is a white, ankle-length tunic, tied at the waist with a cincture, a rope knotted several times to keep it in place; a stole; a cross around the neck; a green and gold cord worn down the back; a chasuble; a purple skull cap; and a mitre. In procession, the applicant carried a crosier. In the way the prosecution case was left to the jury, it was alleged that the first incident occurred on either 15 or 22 December 1996 and that the second incident occurred on 23 February 1997. The applicant's denials The applicant did not give evidence at the trial. In October 2016, he voluntarily participated in a video-recorded interview with the police, which was in evidence. In the course of the interview the applicant emphatically denied A's allegations, stating that "[t]he most rudimentary interview of staff and those who were choirboys" at the time would confirm not only that the allegations were "fundamentally improbable" but also that they were "most certainly false". The applicant told the investigating police that he and his master of ceremonies were at the front of the Cathedral after Mass "as I always did", while the sacristan and his assistant would be in the sacristy cleaning up and bringing out the vessels and other items from the Mass. The prosecution's pre-trial application Consistently with its obligation to call all witnesses whose evidence was necessary to give a complete account of material events, the prosecution proposed to call 23 witnesses who were involved in the conduct of solemn Mass at the Cathedral or who were members of the choir in 1996 and/or 1997. The prosecution was presented with the difficulty that a number of its witnesses were expected to give evidence of practices that existed at the time of the alleged offending which, if followed, were inconsistent with the offending having occurred. The prosecution anticipated that a number of its witnesses would give evidence that these practices were followed with such strictness that there was no realistic possibility of the offending having occurred. In an attempt to confront this difficulty, the prosecutor applied to the trial judge for an advance ruling granting leave to cross-examine a number of his witnesses with respect to evidence that was expected to be unfavourable to the prosecution case and, in the case of the sacristan, Max Potter, with respect to a prior inconsistent statement. The trial judge held that evidence adduced by the prosecution that was inconsistent with, or likely to contradict, A's account of events, was relevantly "unfavourable". His Honour granted leave to the prosecutor to cross-examine a number of witnesses (and foreshadowed the grant of leave in relation to other witnesses) with respect to six such topics. These topics were: (i) whether the applicant was always in the company of another, including Portelli or Potter, when robed; (ii) whether the applicant always greeted congregants on the steps of the Cathedral following Sunday solemn Mass; (iii) whether the applicant's vestments could be moved to the side or parted so as to allow exposure of his penis; (iv) whether the doors from the south transept giving access to the sacristy corridor and the doors to the priests' sacristy were always locked in the period following Sunday solemn Mass; (v) whether the sacramental wine was always locked away and could not have been accessible; and (vi) whether it was possible for two choirboys to separate from the procession without being noticed. The leave granted, at least with respect to topics (i) and (ii), reflected the trial judge's satisfaction that the anticipated evidence, if accepted, excluded the realistic possibility of the offending having occurred as A described it. The prosecutor's opening In the event, the prosecutor pursued very limited cross-examination of his witnesses pursuant to the grants of leave. In opening his case to the jury, the prosecutor acknowledged that there were a number of seemingly irreconcilable differences between A's account and the evidence to be given by other prosecution witnesses. As the Court of Appeal majority encapsulated it, the prosecution case was that the evidence of the witnesses apart from A left open a realistic possibility that the offending that he described had occurred. The applicant's forensic disadvantage The trial judge was satisfied that the applicant had experienced a significant forensic disadvantage in being confronted with allegations of criminal offending more than 20 years after the events were said to have occurred. His Honour informed the jury of the nature of the disadvantage and directed them to take it into account when considering the evidence. His Honour's instruction as to the nature of the disadvantage covered the following considerations: (i) the delay meant that the applicant had lost the opportunity of making inquiries and exploring the alleged circumstances close to the time of the alleged events, which may have uncovered additional evidence throwing doubt on A's allegations or supporting the applicant's denials; (ii) most of the opportunity witnesses could only give evidence of practice or routine whereas, had the trial been held on a date closer to 1996, more might have had specific recall of the subject events; (iii) the effluxion of 20 years or so meant that some witnesses no longer presented the lucid and coherent evidence of younger men; (iv) the Dean of the Cathedral in 1996, whose evidence would have been material on the issue of the applicant's movements following Mass, was in a nursing home and incapable of giving reliable evidence; (v) the passage of time diminished the capacity for the defence to fully test A's evidence; and (vi) B would have been a material witness. The Court of Appeal views the recorded evidence The audio-visual recording of A's evidence at the first trial was admitted at the second trial as if its contents were A's direct testimony under s 379(b)(i) of the Criminal Procedure Act 2009 (Vic). Section 379 makes provision for the admission of the recording of the evidence of a complainant in any appeal from a criminal proceeding that relates to a charge for a sexual offence. Prior to the hearing in the Court of Appeal, the parties were informed that the Court proposed to watch video-recordings of the trial evidence of four witnesses (A, Portelli, Potter and an altar server, Daniel McGlone), and to attend a view of the Cathedral. In response to this information, the applicant submitted that there was no need for members of the Court of Appeal to watch any video-recordings of the witnesses at trial because his case on appeal did not depend upon an assessment of the credibility of any witness. The applicant agreed that the members of the Court should have the benefit of a view of the Cathedral. The applicant also submitted that, if the Court were nevertheless disposed to watch the video-recordings of some witnesses, the Court should also watch the recordings of a number of other named witnesses in order to avoid the risk of "imbalance" or "undue focus". The risk of "imbalance" was adverted to by French CJ, Gummow and Kiefel JJ in SKA v The Queen. The respondent agreed, both with the course proposed by the Court and with the further suggestions by the applicant. The respondent submitted that watching the video-recordings of the witnesses was "desirable given the existence of the relevant recordings". In this Court, the applicant maintained the position that it was unnecessary and undesirable for the members of the Court of Appeal to have watched the recordings of any of the witnesses. Nevertheless, the applicant was not disposed to contend that the course taken by the Court of Appeal was itself an appealable error. The respondent maintained the position that the existence of the recordings was enough to make it "appropriate" for them to be watched by the Court of Appeal. The position maintained by the respondent is not one that should generally be adopted by courts of criminal appeal. In SKA, French CJ, Gummow and Kiefel JJ rejected the suggestion that the mere availability of a video-recording of a witness' evidence at trial meant that the proper discharge of the function of the appellate court, to make its independent assessment of the evidence, necessitated a viewing of the recording. There may be cases where there is something particular in the video-recording that is apt to affect an appellate court's assessment of the evidence, which can only be discerned visually or by sound. In such cases, there will be a real forensic purpose to the appellate court's examination of the video-recording. But such cases will be exceptional, and ordinarily it would be expected that the forensic purpose that justifies such a course will be adopted by the parties, rather than upon independent scrutiny by the members of the court. Secondly, the assessment of the credibility of a witness by the jury on the basis of what it has seen and heard of a witness in the context of the trial is within the province of the jury as representative of the community. Just as the performance by a court of criminal appeal of its functions does not involve the substitution of trial by an appeal court for trial by a jury, so, generally speaking, the appeal court should not seek to duplicate the function of the jury in its assessment of the credibility of the witnesses where that assessment is dependent upon the evaluation of the witnesses in the witness-box. The jury performs its function on the basis that its decisions are made unanimously, and after the benefit of sharing the jurors' subjective assessments of the witnesses. Judges of courts of criminal appeal do not perform the same function in the same way as the jury, or with the same advantages that the jury brings to the discharge of its function. It should be understood that when the joint reasons in M v The Queen spoke of the jury's "advantage in seeing and hearing the witnesses" as being "capable of resolving a doubt experienced by a court of criminal appeal" as to the guilt of the accused, their Honours were not implying that it was only because there were, at that time, no practical means of enabling a court of criminal appeal to see and hear the evidence of the witnesses at trial that the jury's assessment of the credibility of the witnesses was of such potentially critical importance. The assessment of the weight to be accorded to a witness' evidence by reference to the manner in which it was given by the witness has always been, and remains, the province of the jury. Rather, their Honours in M were remarking upon the functional or "constitutional" demarcation between the province of the jury and the province of the appellate court. That demarcation has not been superseded by the improvements in technology that have made the video-recording of witnesses possible. The function of the court of criminal appeal in determining a ground that contends that the verdict of the jury is unreasonable or cannot be supported having regard to the evidence, in a case such as the present, proceeds upon the assumption that the evidence of the complainant was assessed by the jury to be credible and reliable. The court examines the record to see whether, notwithstanding that assessment - either by reason of inconsistencies, discrepancies, or other inadequacy; or in light of other evidence - the court is satisfied that the jury, acting rationally, ought nonetheless to have entertained a reasonable doubt as to proof of guilt. The Court of Appeal's analysis In their joint reasons, the Court of Appeal majority explained that they had approached the determination of the appeal "by trying to put ourselves in the closest possible position to that of the jury" by watching the recordings of A's evidence and that of a number of other witnesses, and by reading the transcript and attending a view of the Cathedral. Their Honours concluded, after viewing A's evidence, both before and after the hearing of the appeal, that he was a compelling witness because of the clarity and cogency of his answers and because of the absence of any indication of contrivance in the emotion he conveyed in giving his answers. A impressed their Honours as a witness "who was telling the truth" and whose answers appeared to be "entirely authentic". Their Honours proceeded to consider, in turn, whether each of a number of "solid obstacles" to acceptance of A's account was such as to compel the jury to have had a doubt. The Court of Appeal majority noted that defence counsel made the submission in his closing address that A's account could not be accepted because "it's impossible basically", there having been no opportunity for the events to have occurred in the way that A described. In their Honours' view, the submission made it incumbent upon the prosecution to seek to negative that the offending was impossible by demonstrating that there was a realistic opportunity for the offending to have occurred. While their Honours acknowledged that there was no onus upon the applicant to prove impossibility, their Honours' analysis proceeded by asking, in relation to each piece of evidence that was inconsistent with A's account, whether it was nonetheless realistically possible that that account was true. As Weinberg JA noted, defence counsel's choice to employ the language of impossibility in his closing address risked setting a forensic hurdle that the defence did not need to overcome. Regardless of counsel's rhetorical flourish, the issue was whether the prosecution had excluded the reasonable possibility that the applicant did not commit the offence/s. At the commencement of their reasons the Court of Appeal majority correctly noted that the approach that an appellate court must take when addressing "the unreasonableness ground" was authoritatively stated in the joint reasons of Mason CJ, Deane, Dawson and Toohey JJ in M. The court must ask itself: "whether it thinks that upon the whole of the evidence it was open to the jury to be satisfied beyond reasonable doubt that the accused was guilty". The Court of Appeal majority went on to note that in Libke v The Queen, Hayne J (with whom Gleeson CJ and Heydon J agreed) elucidated the M test in these terms: "But the question for an appellate court is whether it was open to the jury to be satisfied of guilt beyond reasonable doubt, which is to say whether the jury must as distinct from might, have entertained a doubt about the appellant's guilt." (footnote omitted; emphasis in original) As their Honours observed, to say that a jury "must have had a doubt" is another way of saying that it was "not reasonably open" to the jury to be satisfied beyond reasonable doubt of the commission of the offence. Libke did not depart from M. When it came to applying the M test, their Honours' subjective assessment, that A was a compellingly truthful witness, drove their analysis of the consistency and cogency of his evidence and the capacity of the evidence of the opportunity witnesses to engender a reasonable doubt as to his allegations. Their Honours reasoned, with respect to largely unchallenged evidence that was inconsistent with those allegations (the "solid obstacles" to conviction), that notwithstanding each obstacle it remained possible that A's account was correct. The analysis failed to engage with whether, against this body of evidence, it was reasonably possible that A's account was not correct, such that there was a reasonable doubt as to the applicant's guilt. At the trial and in the Court of Appeal, the applicant relied not only on the evidence of the opportunity witnesses, but also on the content of A's evidence, as giving rise to a doubt as to the truth and reliability of his allegations. It was submitted that A had adapted his evidence in material respects to address matters that had been raised with him for the first time at the committal hearing. These included whether A had changed his account of how the applicant had exposed his penis because of the suggested impossibility of pulling his vestments aside in the way A had first stated. They also included whether A had changed his account of how he and B had re-joined the choir after the assaults. The Court of Appeal majority did not consider that, in any of the respects in which A's evidence at trial varied from his earlier accounts, the variation was such as to have required the jury to entertain a doubt as to the credibility and reliability of his account of the offences. Weinberg JA, in dissent, considered that there was ample material upon which A's account could be subject to legitimate criticism: there were inconsistencies and discrepancies, and a number of his answers "simply made no sense". While his Honour accepted that A appeared to have embellished his account at times, he did not find that, had A's evidence stood alone, his allegations in respect of the first incident were fabricated. His Honour was not prepared to make the same assessment with respect to A's evidence of the second incident. Weinberg JA did not assess A to be such a compelling, credible and reliable witness as to necessarily accept his account beyond reasonable doubt. The division in the Court of Appeal in the assessment of A's credibility may be thought to underscore the highly subjective nature of demeanour-based judgments. A's evidence unsupported? Despite the fact that the prosecution case was left to the jury as being wholly dependent upon A's evidence, the Court of Appeal majority questioned that A's evidence was uncorroborated. Their Honours suggested that, to an extent, A's evidence was supported by reference to knowledge which he could not have come by unless he was telling the truth. The reference was to A's knowledge of the interior layout of the priests' sacristy, which their Honours found considerably enhanced the credibility of his account. "More striking still", their Honours said, was the fact that A identified the priests' sacristy as the setting of the assaults given that, at all other times, the applicant would have used the archbishop's sacristy. Their Honours said that the jury was entitled to discount the possibility that a tour of the Cathedral, which A may have taken at the time he joined the choir, would explain his detailed knowledge and recollection of the interior of the priests' sacristy 20 years later. So much may be accepted. It does not, however, provide support in the sense of corroboration of A's account. Satisfaction that A had been inside the priests' sacristy did not afford any independent basis for finding that, on such an occasion, he had been sexually assaulted by the applicant. Nor did the circumstance that A identified the priests' sacristy as the setting of the first incident afford independent support for acceptance of his account. A said that the applicant "sort of planted himself in the doorway [of the priests' sacristy]" and challenged the two boys before sexually assaulting them. On any view of the matter, acceptance of A's account involves that the applicant was not acting in accord with his regular practice and that he was an opportunistic sexual predator. A's account would be neither more nor less inherently credible if the archbishop's sacristy had been available for the applicant's use at the time. Defence counsel at the trial relied on a counter-argument with respect to the second incident. A placed this incident as having occurred at a point beyond the doors to the priests' sacristy, but before the door to the archbishop's sacristy. Counsel's submission was, in substance, "why would the applicant have walked beyond the priests' sacristy towards the archbishop's sacristy when it was not in use at the time?" Just as A's evidence that the assaults took place in the priests' sacristy does not enhance the credibility of his account, it might be thought that his evidence that the second incident took place past the entry to the priests' sacristy does not detract from it. There is no requirement that a complainant's evidence be corroborated before a jury may return a verdict of guilty upon it. Nonetheless, it was not correct to assess the capacity of A's evidence to support the verdicts on a view that there was independent support for its acceptance. And it was, with respect, beside the point to find that it was open to the jury to view A's knowledge of the priests' sacristy as independent confirmation of him having been inside it. The applicant's submissions The applicant submitted that, notwithstanding that the Court of Appeal majority correctly stated the standard and burden of proof, their Honours reversed it by asking whether there existed the reasonable possibility that A's account was correct, rather than whether the prosecution had negatived the reasonable possibility that it was not. On the Court of Appeal majority's findings, the applicant submitted, it was evident that the jury, acting rationally, ought to have entertained a reasonable doubt as to his guilt. The prosecution conceded that the offences alleged in the first incident could not have been committed if, following Mass, the applicant had stood on the Cathedral steps greeting congregants for ten minutes. Their Honours' conclusion that it was reasonably possible that the applicant had not adhered to his practice on the date of the first incident necessarily carried with it acceptance that it was reasonably possible that he had. This conclusion sufficed, in the applicant's submission, to require that his appeal be allowed, and his convictions quashed, in respect of the first four charges. The same logic applied to the offence charged in the second incident: if it was reasonably possible that the applicant was greeting congregants following solemn Mass for not less than ten minutes on 23 February 1997, he could not have been in the corridor outside the sacristies as the choir processed back through the sacristy corridor to the Knox Centre. The applicant's challenge in this Court was not developed by sole reliance on the evidence of his practice of greeting congregants on the Cathedral steps. The focus of his submissions was on the compounding effect of the improbability of events having occurred as A described them in light of unchallenged direct evidence and evidence of practice. The applicant adopted Weinberg JA's analysis of his submission below with respect to the "compounding improbabilities". His Honour distilled the applicant's case to ten claimed compounding improbabilities. In this Court, the respondent correctly noted that a number of the claimed improbabilities raise the same point. It remains that acceptance of A's account of the first incident requires finding that: (i) contrary to the applicant's practice, he did not stand on the steps of the Cathedral greeting congregants for ten minutes or longer; (ii) contrary to long-standing church practice, the applicant returned unaccompanied to the priests' sacristy in his ceremonial vestments; (iii) from the time A and B re-entered the Cathedral, to the conclusion of the assaults, an interval of some five to six minutes, no other person entered the priests' sacristy; and (iv) no persons observed, and took action to stop, two robed choristers leaving the procession and going back into the Cathedral. It suffices to refer to the evidence concerning (i), (ii) and (iii) to demonstrate that, notwithstanding that the jury found A to be a credible and reliable witness, the evidence as a whole was not capable of excluding a reasonable doubt as to the applicant's guilt. (i) The applicant's movements after the Mass Portelli had served the applicant's predecessor, Archbishop Little, as master of ceremonies and he continued in this role following the applicant's installation as Archbishop of Melbourne. Portelli's duties included meeting the applicant when he arrived at the Cathedral for Sunday solemn Mass and escorting him to the priests' sacristy, where he assisted him to put on his vestments. In evidence-in-chief, Portelli explained that, at the conclusion of Sunday solemn Mass, he was beside the applicant as they processed down the centre aisle to the great west door. The applicant always left the procession at the west door and stood on the steps to greet congregants as they were leaving. He handed his mitre to one of the two altar servers who accompanied them, and his crosier to the other. Portelli remained with him. The "meet and greet" could vary from "as little as ten minutes, say up to 15 or nearly 20. It would depend on what else we had to do that afternoon." Portelli disputed that, even on occasions when there was an engagement in the afternoon, the length of the "meet and greet" might be shorter, saying "it wouldn't be much shorter. It wouldn't make sense to stop for any less time than at least - at least six or seven minutes." He was asked: "Q. Sure, but was there an occasion or were there occasions, as best you can recall, where the Archbishop might depart from that practice and speak for a short period of time before returning to the sacristy? A. He may have done so on occasion, yes. Q. When I say short period of time, I'm speaking of just a couple of minutes? A. Yes, I suppose that's possible but I don't really recall it, but it's possible." In cross-examination, Portelli agreed that the two occasions in December when the applicant celebrated Sunday solemn Mass were memorable; there were a large number of congregants who wished to meet the applicant. The applicant remained on the steps of the Cathedral greeting people for at least ten minutes on each occasion. Portelli recalled that, at the conclusion of the "meet and greet" on each occasion, he accompanied the applicant to the priests' sacristy and assisted him to remove his vestments. In examination-in-chief, the prosecutor had obtained Portelli's acknowledgement that he did not remember whether there was an internal or external procession on 15 and 22 December 1996. In re-examination, Portelli further acknowledged that he did not remember where he and the applicant went after leaving the Cathedral. Portelli's evidence of the fact and the length of the "meet and greet" on 15 and 22 December 1996, and of accompanying the applicant to the priests' sacristy thereafter on each occasion, was unchallenged. Potter served as the Cathedral's sacristan for 38 years. He was aged 84 years at the date of the trial and he appears to have been suffering from some mental infirmity. At times, his recollection of events was apparently flawed. Potter gave evidence that the west door of the Cathedral was closed during the Mass. It was Potter's responsibility to open it for the procession or to have "one of [his] men" do so. Potter recalled the applicant's practice of greeting congregants on the steps of the Cathedral following Sunday solemn Mass. He estimated that this might take 20 minutes or half an hour. Potter maintained that he had an actual recollection of the applicant standing on the front steps of the Cathedral in 1996 at the time he first started saying Mass as Archbishop of Melbourne. When asked if it was possible that the applicant had stayed on the steps for a shorter period of time, Potter responded, "[i]t depends what function he was attending afterwards". The evidence suggested that the applicant did not have functions to attend on the afternoons of 15 and 22 December 1996. The prosecutor pressed Potter as to whether it was possible that the applicant had remained on the front steps speaking with congregants "for a very short period of time", to which Potter responded, "not the first time when he was the archbishop, it took him a while to adjust, and [he] stayed in there welcoming people for a couple of months in the cathedral". Potter agreed that it was possible that on occasions the applicant greeted congregants for a period of ten or 15 minutes rather than the 20 to 30 minutes that he had initially stated. He could not recall the applicant spending "just a short time" in this activity unless the weather was inclement. Potter disputed that on any occasion the applicant had returned to the sacristy unaccompanied; "[i]f Father Portelli wasn't there, he would let me know. I would go down and greet the Archbishop to bring him back in." Daniel McGlone was an altar server at the Cathedral in 1996. He was able to recall only one occasion when he served at a Sunday solemn Mass celebrated by the applicant. The occasion stuck in his memory because it was the first time that the applicant celebrated Sunday solemn Mass at the Cathedral, and his mother had made a rare visit to the Cathedral that day so that they could lunch together afterwards. At the conclusion of the Mass, McGlone walked with his mother to the west door, where the applicant was "doing the meet and greet". McGlone introduced his mother to the applicant, who said, "[y]ou must be very proud of your son". Mrs McGlone responded, "I don't know about that". McGlone recalled the occasion as taking place between October and December 1996. He believed that it was the first time that the applicant had celebrated Sunday solemn Mass in the Cathedral, although he allowed that it might not have been. It was McGlone's impression that the applicant was drawing a deliberate contrast between his administration and that of his predecessor, Archbishop Little, by adopting the practice of greeting congregants after Mass. Jeffrey Connor was an altar server in 1996. He ceased these duties in November 1997. Connor's personal diary entries recorded that he served at the solemn Masses on 15 and 22 December 1996. He did not have a specific recall of the services on those dates but said it was the applicant's "invariable" practice to greet congregants on the steps of the Cathedral after Mass. He recalled that the applicant would take off his mitre and hand it to one altar server and hand his crosier to the other. The altar servers would take them and join the procession at its rear. Connor had served on occasions as the applicant's mitre or crosier bearer. He said the applicant would return to the sacristy more than ten minutes after the procession. Peter Finnigan, the Cathedral choir marshal in 1996, recalled both Sunday solemn Masses celebrated by the applicant in December of that year. In his role as choir marshal, he was near the back of the procession as it left the Cathedral. Once it rounded the side of the Cathedral he moved up until he reached the front of it by the time the choristers were entering the toilet corridor. Finnigan was asked what the applicant was doing as the procession moved along. It was his understanding that the applicant would usually stand on the steps of the west door and greet parishioners for "something like" ten minutes. A number of men who were choirboys in 1996 gave evidence of the conduct of external processions following Sunday solemn Mass. Two of them recalled that on occasions the applicant processed back to the Cathedral with the choir. Anthony Nathan was asked if he ever saw the applicant "pause at the steps at the front entrance and not process out with you" and he said: "I've got memories of both. I think there may have been times where he um, stayed at the front of the steps and spoke to the congregation, and there's also times that I remember walking all the way around." Nathan was asked where the applicant would be at the time of the procession and he explained "so if he was in front of us, it was - by the time we go to that circular pool and then I wouldn't see him after that. Wouldn't really pay attention to where he was after that." The reference to the circular pool, it appears, was to a pool in the Cathedral's garden, which the procession passed by as it made its way back to the metal gate at the rear of the Cathedral. Luciano Parissi was a member of the choir between 1991 and 2001. Parissi was not able to say where the applicant was during the external procession following Sunday solemn Mass because "[h]e'd always be behind me. I would never really be looking back." Parissi recalled that the applicant remained with the procession and that usually the choir would stop and wait for him to enter the back of the Cathedral first. Parissi did not have any specific recollection of Sunday solemn Masses in the second half of 1996. Parissi's membership of the choir spanned the administration of three archbishops. In cross-examination he was asked if it was possible that it was not the applicant for whom the choir stopped at the end of the external procession. He said that to the best of his recollection there "would be times when that would happen with [the applicant] ... I can't recall definitively because I was there for a while, and sometimes those do blur into different priests and archbishops, yes." Andrew La Greca was aged 13 years in 1996. He had commenced singing with the choir in 1993. He had no recall of Masses or processions in December 1996. His recollection was that it was more common for the processions to proceed internally. He understood that whether the procession was external or internal depended upon the identity of the celebrant. Archbishop Little had a preference for external processions. External processions were also frequent when the applicant was the celebrant. La Greca recalled that as the external procession rounded the corner of the Cathedral sometimes the applicant "would just wait and speak to the congregation" and "[o]ther times he might have just kept on walking with us. I can't recall exactly." In cross-examination, La Greca agreed that the procession did not make a tight left turn around the Cathedral, it moved in an arc, and it was possible to look back and see congregants coming out of the Cathedral. He agreed he had seen the applicant greeting the congregants. Rodney Dearing was an adult member of the choir in 1996. It was his evidence that, after Mass, the applicant and Portelli would generally stay at the west door. He did not purport to have a specific recall of the solemn Mass on 15 or 22 December 1996. His evidence was of the applicant's general practice. Dearing was not aware of any occasion when the applicant had departed from the practice, although he acknowledged that, once the procession left the Cathedral, he had not had the applicant "under observation". He recalled occasions when, after returning to the choir room and removing his chorister's robes, he had gone back around to the west door and said hello to the applicant. This had happened reasonably often. Rodney Dearing's son, David, was a member of the choir in 1996. He was aged about 13 at that time. He recalled seeing the applicant stopping on the steps of the Cathedral after solemn Mass. He also recalled, on occasion, coming back through the Cathedral after he had changed out of his choir robes and seeing the applicant still on the main steps. He estimated that this would have been ten or 15 minutes after the end of Mass. (ii) The applicant was always accompanied within the Cathedral Portelli explained that the master of ceremonies is a church office with a long history. The duties of the master of ceremonies are set out in learned works which themselves date back some centuries. The teaching in these texts requires that an archbishop not be unaccompanied from the moment the archbishop enters a church. This evidence of Catholic church practice was unchallenged. Portelli's duties included accompanying the applicant back to the sacristy following Sunday solemn Mass and assisting him to remove his vestments. Portelli acknowledged that it was possible that there was an occasion when he did not return to the sacristy with the applicant although he had no recall of this happening and in such a case he would have made sure that the applicant was accompanied by Potter or a priest. Portelli also pointed out that he, too, needed to change out of his robes following the Mass. Portelli was able to recall the two occasions on which he had not acted as master of ceremonies for the applicant at Sunday solemn Mass in the Cathedral: in June 1997 he was overseas, and in October 2000 he underwent surgery. Apart from these two instances, Portelli had no recall of any occasion when he did not accompany the applicant to the sacristy to disrobe. He acknowledged the possibility of an occasion or occasions when, after escorting the applicant to the sacristy, he may have left him while he, Portelli, went back to the sanctuary to make sure that everything was in place if there was another service that afternoon. In that event, he would have been absent for around two minutes. He was able to say that he would not have left the applicant to check that everything was in place in the sanctuary on either 15 or 22 December 1996 because there were no other events fixed for those afternoons. Potter confirmed that the applicant would never return to the sacristy unaccompanied. It was Potter's responsibility to assist the applicant with the removal of his vestments and to make sure that the vestments were hung. Potter maintained there was always a priest to assist the applicant or "one of us", a reference it would seem to either Portelli or himself being present in the sacristy when the applicant removed his vestments. Potter agreed that, on the first two occasions on which the applicant said Sunday solemn Mass in the Cathedral, the applicant was assisted to disrobe by Portelli. McGlone's understanding was that an archbishop could never be left alone in the Cathedral, at least not during the course of ceremonies, and that the ceremony continued until the archbishop removed his vestments. McGlone explained that the vestments themselves are sacred, and that particular prayers are said when donning and removing them. Connor had no recall of ever seeing the applicant alone while he was robed. He agreed that such an occasion would have been memorable. David Dearing was asked if he had ever seen the applicant in robes without Portelli accompanying him. He replied, "I wouldn't have thought so, no. My recollection is that they were always together." He was asked to describe the distance between the two when they were walking together. His response was to say, "I described him as his bodyguard". His father, Rodney Dearing, agreed that whenever he saw the applicant robed, he was with Portelli. (iii) The timing of the assaults and the "hive of activity" Potter was responsible for the security of the sacristies and he had the key to the priests' sacristy, which was locked during Mass. Potter's evidence about when he unlocked it was unclear. Initially he said that he did so as the procession was making its way down the centre aisle while the recessional hymn was being played. Potter went on to explain that, at the conclusion of the Mass, some congregants would walk up to the sanctuary area and kneel to pray. He allowed them some five or six minutes of "private time" for prayer before he commenced removing the sacred vessels and other items from the sanctuary. Later in the course of examination-in-chief, Potter said that he did not unlock the priests' sacristy doors until after the five or six minutes of private prayer time. The altar servers, Connor and McGlone, gave evidence that, at the conclusion of the procession, they went into the priests' sacristy and each bowed to the crucifix. McGlone explained that the formal procession following the celebration of solemn Mass is only complete for the altar servers when they enter the priests' sacristy in formation and bow to the crucifix in turn. Connor recalled that usually Potter was waiting for them and he unlocked the doors. McGlone recalled that sometimes the door from the vestibule opening into the sacristy corridor was locked and "usually [Potter] would appear out of nowhere and unlock it. Most times though it was unlocked when we were processing there." He recalled that the doors to the priests' sacristy were unlocked. The altar servers assisted Potter in clearing the sanctuary. This took around ten minutes to a quarter of an hour. The chalices, ciboriums, cruet sets, tabernacle key and missals were all returned to the priests' sacristy. The candles (there were seven of them when the applicant celebrated Mass) and the thurible were returned to the utility room. The vessels and other items were carried one at a time. During this exercise people were continually coming into and going out of the priests' sacristy. Neither Connor nor McGlone could recall any occasion on which the sacristy had been left unlocked and unattended. In Connor's experience, that never happened. Dr Cox, the assistant organist, described the priests' sacristy as a "hive of activity" after Mass. The choir marshal, Finnigan, recalled that after Mass there were "people everywhere" in the sacristy corridor, with people "coming in and going out", including altar servers bringing implements into the priests' sacristy. He said that there were almost always a number of other priests acting as concelebrants who would vest and de-vest in the priests' sacristy. They would come in after Mass had finished and remain for the next ten to 15 minutes or so. McGlone likened it to the green room in an opera house, explaining that it was where the sacred vessels were taken, and that the servers "are moving back and forth into that room". Consideration - (i) the applicant's movements after Mass and (ii) the applicant always accompanied The Court of Appeal majority dealt with the evidence of the applicant greeting congregants on the Cathedral steps, observing: "But, on the critical issue of whether [the applicant] stood on the steps of the Cathedral on the day of the first or second Mass, and if so for how long, the recollection of the opportunity witnesses must necessarily be affected by their recollection of the ritual that developed thereafter." Notwithstanding that Portelli's evidence of having an actual recall of being present beside the applicant on the steps of the Cathedral as the applicant greeted congregants on 15 and 22 December 1996 was unchallenged, the Court of Appeal majority said it was open to the jury to have reservations about the reliability of his affirmative answers given in cross-examination. The Court of Appeal majority also considered that it was open to have reservations about the reliability of this evidence given the improbability of Portelli having a specific recollection of particular Masses in the absence of "some significant and unusual event" having occurred at one or other of them. Their Honours observed that, while Portelli may have had a general recollection of the first time the applicant said Sunday solemn Mass at the Cathedral, he had demonstrated a lack of detailed recall of the events of that day. The suggestion that witnesses' memories may have been affected by the ritual that developed thereafter has echoes of the prosecutor's closing submission, which was that the applicant's practice of greeting congregants may not have developed before 1997. It is a contention that finds no support in the evidence and was not pursued by the respondent on appeal to this Court. It will be recalled that Finnigan's understanding was that the applicant stood on the steps greeting congregants for "something like" ten minutes as the procession, with him towards the rear, made its way around the side of the Cathedral. His understanding in this respect was of Masses celebrated by the applicant in 1996, since Finnigan last acted as choir marshal on Christmas day of that year. The occasion when McGlone's mother was introduced to the applicant on the Cathedral steps after Sunday solemn Mass was in December 1996, as McGlone did not believe that he continued as an altar server after the end of 1996. The Court of Appeal majority observed that the encounter between McGlone's mother and the applicant was not in doubt but that there was some uncertainty about the date of its occurrence. McGlone was confident that this was the first time the applicant had said Mass in the Cathedral, but their Honours observed that McGlone had been mistaken in his belief that he had not attended the evening Mass celebrated by the applicant on 23 November 1996. Moreover, their Honours said that, accepting the encounter occurred on either 15 or 22 December 1996, it did not make the occurrence of the first incident impossible. It simply ruled out one of those two Sundays as the date of its occurrence. The Court of Appeal majority's treatment of what their Honours rightly identified as the critical issue in the case was wrong for two reasons. First, Portelli's evidence was unchallenged. Secondly, their Honours were required to reason in a manner that is consistent with the way in which a jury would be directed in accordance with the Jury Directions Act 2015 (Vic). Their Honours were required to take into account the forensic disadvantage experienced by the applicant arising from the delay of some 20 years in being confronted by these allegations. Their Honours, however, reasoned to satisfaction of the applicant's guilt by discounting a body of evidence that raised lively doubts as to the commission of the offences because they considered the likelihood that the memories of honest witnesses might have been affected by delay. The Court of Appeal majority acknowledged that there was general consistency and "substantial mutual support", in the account of the opportunity witnesses, as to what occurred after Sunday solemn Mass in the period when the applicant was archbishop. And, as their Honours also acknowledged, a defining feature of religious observance is adherence to ritual and compliance with established practice. However, their Honours again discounted this body of evidence, saying: "[A]ttempting to recall particular events is all the more difficult when the events being described are - as they were here - of a kind which was repeated week after week, year after year, and involved the same participants, in the same setting, performing the same rituals and following the same routines." Evidence of a person's habit or practice of acting in a particular way to establish that the person acted in that way on a specific occasion may have considerable probative value. As Professor Wigmore explained, "[e]very day's experience and reasoning make it clear enough". The evidence of religious ritual and practice in this case had particular probative value for the reason that their Honours first identified: adherence to ritual and compliance with established liturgical practice is a defining feature of religious observance. Contrary to the Court of Appeal majority's analysis, the absence of any "significant and unusual event" associated with solemn Mass on 15 and 22 December 1996 tells against the likelihood of Portelli having departed from his duties as master of ceremonies. The Court of Appeal majority took into account the evidence of four witnesses in concluding not only that it was possible that the applicant was alone and robed in contravention of centuries-old church law, but that the evidence of witnesses to the contrary did not raise a reasonable doubt as to the applicant's guilt. The evidence to which their Honours referred was that of three choirboys - Robert Bonomy, David Mayes and Nathan - and that of the organist and choirmaster, John Mallinson. Bonomy said that he had seen the applicant robed in the sacristy corridor and sometimes the applicant was with others and sometimes he was on his own. Bonomy had been lined up with the choir in the sacristy corridor waiting to process into Mass when he made these observations. Their Honours noted that Nathan and Mayes recalled the applicant coming into the choir room after Mass and that "[v]ery infrequently, Nathan said, [the applicant] would be robed". Nathan had a recollection of the applicant popping into the choir room to congratulate the choir on a good performance or a great Mass. He could not remember whether the applicant was alone or with someone else, nor whether he was robed. The occasion does not appear to have been further identified. Mayes' evidence, to which their Honours referred, was his memory of the applicant coming into the choir room "in the first five minutes while everybody was still there". When asked if the applicant was robed, he replied that it was "very rare to see him unrobed. Yeah, he would have been robed." Mayes could not say whether the applicant was accompanied or not on this occasion. Mayes agreed that there were infrequent special functions for the choir to which parents were invited at which the applicant would be introduced to the parents. He was unable to say whether the applicant was robed on these occasions. Mayes recalled leaving the Cathedral after Sunday solemn Mass and seeing the applicant on the Cathedral steps shaking hands or talking to congregants. He agreed that this would have been 15 minutes or more after Mass had finished. The Court of Appeal majority noted Mallinson's evidence of probably having seen the applicant in the sacristy corridor many times. Their Honours extracted the following exchange concerning that evidence: "Q. And again, was he on his own or with anyone? A. Sometimes he was with somebody and sometimes he would be on his own. Q. Would he be robed or unrobed? A. I've seen him both ways. For instance, after he'd gone to the sacristy and disrobed and he'd be in his normal clerical garb." It is by no means evident that Mallinson was departing from his evidence that, on the occasions when Mallinson saw the applicant in his robes, Portelli was always with him. It may be observed that Mallinson acknowledged that the applicant was a stickler for protocol and conservative in terms of church liturgy and tradition. The honesty of the opportunity witnesses was not in question. Portelli and Potter each gave evidence that Portelli accompanied the applicant to the priests' sacristy after solemn Mass on 15 and 22 December 1996. There appears to have been agreement that, in light of Potter's apparent infirmity, notwithstanding the grant of leave to cross-examine him, the prosecutor was not required to comply with the rule in Browne v Dunn. This understanding did not apply to Portelli. Portelli's evidence in this respect was unchallenged. So, too, was the evidence that Catholic church teaching requires an archbishop to be accompanied while in a church, at least while the archbishop is robed, unchallenged. And the evidence that it was Portelli's role as the applicant's master of ceremonies to ensure that this requirement was complied with was unchallenged. Whatever is made of Nathan's and Mayes' evidence of the applicant coming into the choir room in the Knox Centre, it was not evidence of the applicant being unaccompanied while robed in the Cathedral. Bonomy's evidence is a slim foundation for finding that the practice of ensuring that the applicant was accompanied while he was in the Cathedral was not adhered to. It provides no foundation for excluding the reasonable possibility that Portelli's actual recall of accompanying the applicant to the priests' sacristy after solemn Mass on 15 and 22 December 1996 was accurate. There was a powerful body of evidence of the applicant's practice of greeting congregants on the Cathedral steps following Sunday solemn Mass and that, while the length of this "meet and greet" varied, it occupied at least ten minutes. The applicant's practice in this respect contrasted with that of his predecessor, Archbishop Little. Portelli served as master of ceremonies for both and it might be thought unremarkable that he should recall that on the first and second occasions on which the applicant, as the new Archbishop of Melbourne, celebrated Sunday solemn Mass in the Cathedral, he had greeted congregants as they left after the service. The respondent's reliance in this Court on the two choirboys' evidence, that sometimes the applicant processed back to the Cathedral with the choir, is no answer to Portelli's evidence concerning the solemn Masses on 15 and 22 December 1996. Moreover, their evidence hardly calls into question the evidence of the opportunity witnesses of the applicant's practice of greeting congregants after Mass. Nathan's recollection was that, on the occasions that the applicant processed out of and around the side of the Cathedral, the applicant was in front of him. There does not appear to have been any question in the evidence of the other witnesses that when the applicant took part in the procession, as it entered the Cathedral or as it made its way down the centre aisle at the conclusion of the Mass, as the most senior of the participants, he was at its end. Parissi accepted that his memory of standing back to allow the applicant to re-enter the Cathedral complex might be wrong, as his memory of archbishops and priests tended to blur. Parissi was a chorister when Archbishop Little celebrated Mass and, as noted, Archbishop Little did not leave the procession to greet congregants. (iii) Consideration - the timing of the assaults and the "hive of activity" As the Court of Appeal majority observed, the effect of the altar servers' evidence was that the unlocking of the priests' sacristy doors, and the bowing to the crucifix, occurred soon after the procession finished. By the time the procession returned, and the altar servers reached the door giving access to the eastern end of the sacristy corridor, the doors to the priests' sacristy were unlocked. It will be recalled that it was A's account that he and B broke away from the procession at a point at which the choristers were congregated outside the metal gate which gave access to the toilet corridor. A and B made their way back into the Cathedral through the south transept door and from there through the double doors which opened from the south transept into the western end of the sacristy corridor. The Court of Appeal majority concluded that it was "quite possible" for the priests' sacristy to have been unlocked and that A and B might have entered the priests' sacristy after the altar servers had bowed to the crucifix. Their Honours further concluded that it was open to the jury to find that the assaults took place in the five to six minutes of private prayer time, before the "hive of activity" in the priests' sacristy, including the clearing of the sanctuary by the altar servers, commenced. The possibility for which their Honours allowed is not without difficulty. A, a soprano, was close to the front of the procession. If A and B broke away from it and re-entered the Cathedral through the door of the south transept and went through the double doors into the western end of the sacristy corridor, it might reasonably be expected that they would have encountered the altar servers. The altar servers were at the front of the procession. There were at least six of them and there may have been as many as 12. Those in the front of the procession waited for the two servers bookending it at the rear and then they bowed in order to the crucifix. A further oddity is that A and B did not encounter any concelebrant priests in the sacristy corridor or the priests' sacristy, notwithstanding that concelebrant priests would be expected to have gone into the priests' sacristy to disrobe after the procession broke up. It was Finnigan's evidence that there were other priests concelebrating solemn Mass on 15 and 22 December 1996. The principal difficulty with the Court of Appeal majority's analysis is that it elides Potter's estimate of five to six minutes of private prayer time with the estimate of five to six minutes during which A and B re-entered the Cathedral, made their way into the priests' sacristy and were assaulted. The two periods are distinct. The private prayer time commenced shortly after the conclusion of the Mass. Mallinson, the organist and choirmaster, referred to it as an "interval" of "decorum". He was asked when Potter would commence clearing the sanctuary after Mass, and he replied: "Well, it's difficult to define. Perhaps the clergy having left the sanctuary half a minute, a minute, perhaps a minute and a half, two minutes. It's difficult to say. I mean it depends on the circumstances, how many people are in the cathedral, but fairly soon after the clergy have left the sanctuary." The procession, of which A and B formed a part, was making its way down the central aisle of the Cathedral during the private prayer time. The procession processed with a degree of formality because it was a religious procession and its members were on display to the public. Assuming that private prayer time occupied five or six minutes, and not the lesser time that Mallinson recalled, it remains that, by the time the altar servers entered the sacristy corridor at the conclusion of the external procession, the private prayer time had been running for some minutes. The Court of Appeal majority's conclusion that it was possible that the assaults occurred after the altar servers had bowed to the crucifix in the priests' sacristy and before they commenced to clear the sanctuary invites the question "where were the altar servers during the five- to six-minute hiatus that their Honours hypothesised?" Although the timing of these events cannot be fixed with any precision, it was, as noted, plainly not the case that the private prayer time given to congregants, before items from the sanctuary were cleared to the priests' sacristy, did not commence until the front of the procession was close to the metal gate. It was not in issue that the altar servers entered the priests' sacristy and bowed to the crucifix at the conclusion of the procession or that they assisted Potter to clear the sanctuary. In closing submissions, the prosecutor invited the jury to find that, after bowing to the crucifix, the altar servers went to the "workers' sacristy" and waited for Potter to give them "the green light" to start clearing up. There was no evidentiary support for that submission and, following objection, the prosecutor withdrew it. In this Court, the respondent maintained that the assaults occurred after the altar servers had entered the priests' sacristy and bowed to the crucifix and before the "hive of activity" in the sacristy commenced. The respondent, relying on Mallinson's evidence, sought to lengthen the private prayer time, submitting that "[p]recisely when this interval would end would, of course, depend on the circumstances including how many people were in the Cathedral". The submission overlooked that, on Mallinson's account, circumstances such as how many people were in the Cathedral would only account for the private prayer time allowed prior to the clearing of the sanctuary varying from 30 seconds to two minutes. The respondent also submitted that "[t]he altar servers would have then left the Sacristy - either for the workers' room, where they disrobed, or for the sanctuary to assist Potter". The submission comes close to repeating the submission which the prosecutor withdrew at the trial. There was no evidence that the altar servers went to their room to disrobe prior to returning to the sanctuary in order to assist in clearing away the sacred vessels and other objects. Nor is there an evidentiary foundation for the conclusion that there was a hiatus between the time when the altar servers completed their bows to the crucifix and the clearing of the sanctuary. Conclusion It may be accepted that the Court of Appeal majority did not err in holding that A's evidence of the first incident did not contain discrepancies, or display inadequacies, of such a character as to require the jury to have entertained a doubt as to guilt. The likelihood of two choirboys in their gowns being able to slip away from the procession without detection; of finding altar wine in an unlocked cupboard; and of the applicant being able to manoeuvre his vestments to expose his penis are considerations that may be put to one side. It remains that the evidence of witnesses, whose honesty was not in question, (i) placed the applicant on the steps of the Cathedral for at least ten minutes after Mass on 15 and 22 December 1996; (ii) placed him in the company of Portelli when he returned to the priests' sacristy to remove his vestments; and (iii) described continuous traffic into and out of the priests' sacristy for ten to 15 minutes after the altar servers completed their bows to the crucifix. Upon the assumption that the jury assessed A's evidence as thoroughly credible and reliable, the issue for the Court of Appeal was whether the compounding improbabilities caused by the unchallenged evidence summarised in (i), (ii) and (iii) above nonetheless required the jury, acting rationally, to have entertained a doubt as to the applicant's guilt. Plainly they did. Making full allowance for the advantages enjoyed by the jury, there is a significant possibility in relation to charges one to four that an innocent person has been convicted. The second incident It will be recalled that the second incident is alleged to have occurred after Sunday solemn Mass on an occasion on which there was an internal procession through the sacristy corridor. A agreed that he, as one of the younger boys, would have been towards the front of the procession as it made its way through the sacristy corridor, with the older choristers, including some adults, behind him. They were all rushing to get back to the choir's robing room when the second incident occurred. The applicant appeared and shoved A against the wall and squeezed his genitals, causing pain, although he did not know if he had called out. The defence contended at trial that the notion that the applicant - a tall, imposing figure in his archbishop's robes - might assault a young choirboy in the presence of a number of choristers, including several adults, bordered on the fanciful. The Court of Appeal majority accepted that the sight of the applicant at close quarters with a choirboy might well have attracted attention. However, their Honours reasoned that the others in the corridor were intent on completing the procession and removing their robes as soon as possible. In this state of affairs, their Honours assessed that it was quite possible that the brief encounter went unnoticed. At all events, their Honours said, "the evidence once again falls well short of establishing impossibility". Weinberg JA considered that, had the second incident occurred in the way A described it, it was highly unlikely that none of the many persons present would have seen what was happening or reported it in some way. His Honour concluded that it was not open to the jury to be satisfied beyond reasonable doubt of the applicant's guilt of the offence charged in the second incident. The assumption that a group of choristers, including adults, might have been so preoccupied with making their way to the robing room as to fail to notice the extraordinary sight of the Archbishop of Melbourne dressed "in his full regalia" advancing through the procession and pinning a 13 year old boy to the wall, is a large one. The failure to make any formal report of such an incident, had it occurred, may be another matter. It is unnecessary to decide whether A's description of the second incident so strains credulity as to necessitate that the jury, who saw and heard him give the evidence, ought to have entertained a reasonable doubt as to its occurrence. The capacity of the evidence to support the verdict on this charge suffers from the same deficiency as the evidence of the assaults involved in the first incident. Portelli gave unchallenged evidence of his recall of being with the applicant at solemn Mass on 23 February 1997. Portelli recalled that this was an unusual occasion because Father Egan was the celebrant. The protocol remained that the applicant as the most senior person was last as the procession processed down the centre aisle of the Cathedral. The unchallenged evidence of the applicant's invariable practice of greeting congregants after Sunday solemn Mass, and the unchallenged evidence of the requirement under Catholic church practice that the applicant always be accompanied when in the Cathedral, were inconsistent with acceptance of A's evidence of the second incident. It was evidence which ought to have caused the jury, acting rationally, to entertain a doubt as to the applicant's guilt of the offence charged in the second incident. In relation to charge five, again making full allowance for the jury's advantage, there is a significant possibility that an innocent person has been convicted. This conclusion makes it unnecessary to consider whether the respondent's concession, that if the verdicts in relation to the offences charged in the first incident are unreasonable or cannot be supported by the evidence then it follows that the same conclusion should be reached in relation to the verdict concerning the offence charged in the second incident, amounts to a mode of reasoning that contravenes ss 44F and 44G of the Jury Directions Act. Orders For these reasons, there should be the following orders:
Pell
The Queen
Special leave granted; appeal allowed; Court of Appeal orders set aside; appeal to Court of Appeal allowed; convictions quashed and judgments of acquittal entered.