Gleeson CJ, Gummow J, Hayne J, Callinan J, Heydon J & Crennan J
The applicants were presented for trial in the Supreme Court of Victoria on a single presentment alleging two counts against each: a count alleging the murder of Steven John Borg and a count of intentionally causing serious injury to Paula Michelle Rodwell. Each was convicted on both counts. Each sought leave to appeal to the Court of Appeal of the Supreme Court of Victoria against the convictions. The applications for leave to appeal against conviction on the second count (of intentionally causing serious injury) were allowed and a new trial ordered on that count: Hartwick, Clayton and Hartwick (2005) 159 A Crim R 1. Each application for leave to appeal against the conviction for murder was dismissed.
The applicants sought special leave to appeal to this Court on a number of grounds. In so far as each applicant sought special leave to appeal on grounds which, in effect, invited the Court to reconsider its decisions in McAuliffe v The Queen (1995) 183 CLR 108 and Gillard v The Queen (2003) 219 CLR 1, the applications for special leave were referred for argument, as on an appeal, before the whole Court. In so far as the applications sought special leave to appeal on other grounds, they were dismissed.
At the conclusion of oral argument on the grounds inviting reconsideration of McAuliffe and Gillard, the Court announced that the invitation to reconsider those cases was declined and ordered that each application for special leave be dismissed. It follows that intermediate and trial courts must continue to apply the principles established by those decisions. It also follows that it is neither necessary nor desirable to attempt to elaborate or explain those principles in any way. Nothing that is said in these reasons should be understood as doing so. Rather, what follows are our reasons for joining in declining to reconsider those cases and for making the orders that were made.
It is necessary to say something about the facts of the case. Although the trial of the applicants lasted 46 days, it is possible to describe the essential facts of the matter quite shortly.
The deceased, Mr Borg, had a relationship with Ms Rodwell, the alleged victim of the second count charged against the applicants. Mr Borg did not live with Ms Rodwell but was visiting her house when he met his death. The applicant Lisa Hartwick lived in the same street as Ms Rodwell. The applicants John and Lisa Hartwick had been married to one another but at the relevant time were divorced. Mr Hartwick often visited Ms Hartwick and stayed at her house. The other applicant, Ms Clayton, was a friend of Lisa Hartwick. The Hartwicks both knew Mr Borg and Ms Rodwell, but Ms Clayton had met Ms Rodwell only twice before the date of the offences and had never met Mr Borg.
On the day Mr Borg died, Lisa Hartwick accosted Ms Rodwell and assaulted her saying, among other things, that she owed Mr Hartwick money and had accused Mr Hartwick of being a "dog", a police informer. Celia Clayton was present during these events. Lisa Hartwick, Celia Clayton and Ms Rodwell then went together to Ms Hartwick's home to clear these matters up with Mr Hartwick. It was agreed that there had been a mistake, and the two women apologised to Ms Rodwell.
Ms Rodwell went back to her house and told Mr Borg what had happened. He had consumed a quantity of drugs and he became very angry at the way she had been treated. He drove to Lisa Hartwick's house in a stolen car, accelerated into her driveway and smashed his car into the rear of Ms Clayton's rented car. That car was propelled into another that was parked directly in front of it. Both cars were seriously damaged, and some damage was done to the house. Mr Borg then returned to Ms Rodwell's house and, with her, consumed a cap of heroin.
When the applicants discovered what had happened they became very angry and very agitated. There was evidence at trial that the applicants uttered various threats of violence against the person or persons responsible for the damage that had been done, and that these threats were uttered both before and upon arriving at Ms Rodwell's house. Who said what was a matter of dispute at trial.
There was evidence at the trial that each of them armed himself or herself in some way before leaving Ms Hartwick's house. The weapons were variously described as including metal poles, wooden poles, and a large carving knife. Who had what weapon or weapons, and who knew what weapon or weapons were being carried (if any), were matters of dispute at trial.
Ms Rodwell gave evidence of the applicants coming to the door of her house, assaulting her and one of them then detaining her at knifepoint while there was a prolonged assault on Mr Borg. The applicants disputed her account of what happened. But the injuries done to Mr Borg were consistent with her account of an attack lasting 30 to 40 minutes in which he was severely beaten with poles and stabbed a number of times. One of the stab wounds caused fatal injuries.
The prosecution's case at trial was that, although it could not identify which of the applicants inflicted the fatal stab wound, each was guilty of murder. This contention was put in three ways. First, it was said that the killing occurred in the course of the applicants' implementation of a plan to cause really serious injury to the deceased. That is, the prosecution alleged that each applicant had participated in a joint enterprise. Alternatively, the prosecution argued that each applicant was guilty of murder because each had agreed to assault the deceased using weapons, and reasonably foresaw the possibility that death or really serious injury might be intentionally inflicted on the victim by one of them in the course of their carrying out the agreed assault. It is this second way of putting the case ("extended common purpose") that engaged the principles described in McAuliffe and Gillard. Finally, the prosecution argued that the two applicants who did not inflict the fatal wound had aided and abetted the person who did, by intentionally helping, encouraging or conveying their assent to that person in his or her commission of the murder. This third contention depended upon principles of accessorial liability.
The prosecution accepted that if murder was not established, it would be necessary for the jury to consider manslaughter, but the principal contention advanced on behalf of the prosecution was that each applicant was guilty of murder on one or other of the three bases described.
Questions of provocation were mentioned at trial but, in the end, no party asked for provocation to be left to the jury and it was not. Self-defence was in issue because there was evidence that the deceased was armed with a knife. It was necessary, therefore, at trial, to consider the possible application of a number of intersecting legal principles. The prosecution contended that three separate legal paths led to the conclusion that each applicant was guilty of murder. Each applicant contended that the prosecution had to disprove that he or she had acted in self-defence. And these legal principles were to be applied where there was much dispute about who had done what in the events culminating in Mr Borg's death.
Questions of extended common purpose had to be considered at trial. The principles authoritatively established in McAuliffe and Gillard were to be applied both at trial and by the Court of Appeal and there was, and could be, no argument to the contrary. The applicants sought special leave to appeal to contend that those principles, which had to be applied in the courts below, should now be abandoned or at least substantially modified, presumably on the basis that "on any ground there was a miscarriage of justice": Crimes Act 1958 (Vic), s 568(1).
There are several reasons for the Court not to reconsider what was said in McAuliffe or Gillard about extended common purpose. First, contrary to the applicants' central submission, it is not demonstrated that the application of the principles stated in those cases has led to any miscarriage of justice in this case or, more generally, has occasioned injustice in the application of the law of homicide. The applicants pointed to no decided case said to reveal the alleged injustice. Rather, for the most part, the argument was advanced in a wholly abstract form.
The applicants' contentions about "unjust" results, or results said to demonstrate disconformity between legal and moral responsibility, proceeded, in the main, from an unstated premise that the crime of murder should be confined to cases in which the accused intended the death of the victim. The allegation of injustice or disconformity, though variously expressed, fastened upon the fact that applying principles of extended common purpose could result in a person being found guilty of murder where that person did not agree or intend that death should result, but foresaw only the possibility that an assault with intent to kill or cause really serious injury might be made in the course of the joint enterprise. The applicants sought to compare this outcome with the case of a person assaulting another, knowing of the possibility, but not intending, that death or really serious injury might result. Such a person, the applicants submitted, would be guilty only of manslaughter.
A person who does not intend the death of the victim, but does intend to do really serious injury to the victim, will be guilty of murder if the victim dies. If a party to a joint criminal enterprise foresees the possibility that another might be assaulted with intention to kill or cause really serious injury to that person, and, despite that foresight, continues to participate in the venture, the criminal culpability lies in the continued participation in the joint enterprise with the necessary foresight - McAuliffe v The Queen (1995) 183 CLR 108 at 118; Gillard v The Queen (2003) 219 CLR 1 at 36 . That the participant does not wish or intend that the victim be killed is of no greater significance than the observation that the person committing the assault need not wish or intend that result, yet be guilty of the crime of murder.
Secondly, the applicants could point to no other court of final appeal accepting the proposition that the applicants put at the forefront of their submission, namely, that the doctrine of extended common purpose should be abolished or modified by replacing foresight of the possibility of a murderous assault with foresight of the probability of such an assault. Other common law countries continue to apply generally similar principles to those stated in McAuliffe and Gillard - see, for example, Chan Wing-Siu v The Queen  AC 168; Hui Chi-ming v The Queen  1 AC 34; R v Powell  1 AC 1.
Thirdly, if there are to be changes in this area of the law (and we are not to be taken as suggesting that there should be) there could be no change undertaken to the law of extended common purpose without examining whether what was being either sought or achieved was in truth some alteration to the law of homicide depending upon distinguishing between cases in which the accused acts with an intention to kill and cases in which the accused intends to do really serious injury or is reckless as to the possibility of death or really serious injury - R v Crabbe (1985) 156 CLR 464. That is a task for legislatures and law reform commissions. It is not a step that can or should be taken in the development of the common law. The nature and extent of the difficulties that are encountered in adopting some other principle are revealed in the report of the United Kingdom Law Commission Inchoate Liability for Assisting and Encouraging Crime, Law Com No 300, (July 2006) Cm 6878. The Law Commission rejected a proposal to abolish what it described as secondary liability for a collateral offence committed in the course of a joint venture - Ibid at 19 [2.24]-[2.25].
Further, no change could be undertaken to the law of extended common purpose without examining the whole of the law with respect to secondary liability for crime. The history of the distinction between joint enterprise liability and secondary liability as an aider, abettor, counsellor or procurer of an offence has recently been traced by Professor Simester in "The Mental Element in Complicity", (2006) 122 Law Quarterly Review 578 at 596-598. As that author demonstrates (ibid at 598-599), liability as an aider and abettor is grounded in the secondary party's contribution to another's crime. By contrast, in joint enterprise cases, the wrong lies in the mutual embarkation on a crime, and the participants are liable for what they foresee as the possible results of that venture. In some cases, the accused may be guilty both as an aider and abettor, and as participant in a joint criminal enterprise. That factual intersection of the two different sets of principles does not deny their separate utility.
Finally, it was submitted that the doctrine of extended common purpose should be discarded because it renders the trial of homicide too complex. The trial of these applicants was said to provide a good example of the complexities that arise. But upon examination, this trial does not support this particular criticism.
In the present case, the trial judge gave the jury written directions upon which he elaborated by extensive oral directions. Unsurprisingly, the trial judge's written directions followed a pattern formed by the way in which the prosecution put its case. The jury were thus instructed in a way which assumed that each of the three arguments advanced by the prosecution should be considered separately. There was, therefore, a separate statement of each of the elements said to be necessary to establish the crime of murder according to whether there had been a joint enterprise, whether the applicants had been party to a common purpose such as engaged doctrines of extended common purpose, or whether the applicants who had not struck the fatal blow had nonetheless aided and abetted the person who had. These directions had to be overlaid with the need for the jury to deal with the questions about self-defence that were presented by the evidence.
It may greatly be doubted that it was essential to identify the issues which the jury had to consider according to a pattern determined only by the legal principles upon which the prosecution relied. The written directions took that shape, but the oral directions focused more immediately upon the factual questions that arose.
The real issues in the case (Alford v Magee (1952) 85 CLR 437 at 466) which the jury had to decide were issues of fact. It was for the trial judge to determine what those real issues were and to instruct the jury about only so much of the law as must guide them to a decision on those issues. It may have been possible to instruct the jury in a way that avoided repetition of what, in the end, were relatively few issues for their consideration.
The case against each applicant had to be considered separately. The injuries suffered by the deceased were consistent only with a prolonged assault upon him. There seemed little doubt that one of the applicants had inflicted the fatal wound. Because the prosecution did not contend that the evidence revealed who had struck the fatal blow, the principal issues in each case centred upon:
what did the applicant agree was to happen when they went to Ms Rodwell's house?
what did that applicant foresee was possible? and
what did that applicant do at the house, if anything, to aid and abet whoever it was who had fatally assaulted the deceased?
If, as the prosecution contended was the case in respect of each applicant, the particular applicant under consideration was shown, beyond reasonable doubt, to have agreed with one or both of the other applicants to cause really serious injury to the deceased, a verdict of guilty of murder had to be returned. If the prosecution demonstrated beyond reasonable doubt that the applicant under consideration was party to an agreement with one or other of the applicants to assault the deceased to some lesser degree, and foresaw the possibility that death or really serious injury might intentionally be inflicted on the deceased in the course of that assault (otherwise than in self-defence), again, a verdict of murder had to be returned. In this latter respect, if persuaded beyond reasonable doubt that the applicant concerned went to the premises armed, or knowing that others were going armed, it would be open to the jury to infer that that applicant foresaw the possibility of assault with the requisite intent, but such an inference was not inevitable.
Finally, if the jury were persuaded beyond reasonable doubt that the applicant under consideration detained Ms Rodwell, knowing that Mr Borg was being assaulted with intent to kill or cause really serious injury, and that the applicant in question detained her to help or encourage the making of that assault, a verdict of murder had to be returned.
There was a great deal of evidence that bore on these issues. Several different accounts had been given of what had happened before and during the fatal assault on Mr Borg, by the applicants when interviewed by police, by witnesses to what was said and done before the applicants arrived at Ms Rodwell's house, and by Ms Rodwell herself. And it was necessary for the judge to tell the jury what evidence was admissible against each applicant. But the issues (as distinct from the evidence) were relatively simple. What did the applicant agree was to happen; what did that applicant foresee might happen; what did that applicant do at the house?
Applying the principles of extended common purpose did not require the over-elaboration or over-complication of the issues in this case. And the applicants offered no example of a case where it would.
It is for these reasons that we joined in the orders that were made at the conclusion of oral argument.
Three applicants asked this Court to grant special leave to permit the simplification, rationalisation and re-expression of the Australian common law principles governing accessorial liability on the basis of an extended common purpose on the part of secondary offenders. The challenge has been anticipated for a time because of the injustice, asymmetry and complexity of the present law. It is a field "uncertain and controversial" and "notoriously difficult" that suffers from "incongruent principles" with consequent injustice - Simester, "The Mental Element in Complicity", (2006) 122 Law Quarterly Review 578 at 578, 579, 600.
In my view, the applications were entitled to succeed. The applicants should have been afforded relief. In his dissenting opinion in the Supreme Court of the United States in Bowers v Hardwick 478 US 186 at 202-203 (1986), since overruled by Lawrence v Texas 539 US 558 at 578 (2003), Blackmun J, in another legal context, stated:
The Court's cramped reading of the issue before it makes for a short opinion, but it does little to make for a persuasive one.
Respectfully, that is my conclusion in the present case. Upon a full analysis of the detailed arguments of the parties, the applicants have made out their contentions. Special leave to appeal should have been granted. The appeals should have been allowed.
THE DECISIONAL BACKGROUND AND HISTORY
CONVICTIONS AND APPEALS
Celia Clayton, John Hartwick and Lisa Hartwick (together "the applicants") applied to this Court seeking to challenge orders of the Court of Appeal of the Supreme Court of Victoria - R v Hartwick (2005) 159 A Crim R 1. Those orders followed applications for leave to appeal to that Court against the convictions of each of the applicants for the murder of Steven John Borg ("the deceased") at Karingal, near Melbourne, on 23 May 2001.
The Court of Appeal, constituted by Charles, Chernov and Nettle JJA, upheld the applicants' challenge to their concurrent convictions of intentionally causing serious injury to the deceased's friend, Ms Paula Michelle Rodwell. It set aside such convictions, ordered a retrial on those counts ((2005) 159 A Crim R 1 at 50-53 -) and varied the sentences accordingly to a period of eighteen years imprisonment, with fourteen years non-parole period, in respect of the count of murder alone ((2005) 159 A Crim R 1 at 64 ). However, the Court of Appeal confirmed each of the convictions for murder. Those convictions had followed jury verdicts of guilty on the counts charging the applicants with the murder of the deceased. The guilty verdicts, in turn, followed directions given to the jury by the trial judge (Smith J). Such directions were in accordance with the authority of this Court in Johns v The Queen (1980) 143 CLR 108, McAuliffe v The Queen (1995) 183 CLR 108 and Gillard v The Queen (2003) 219 CLR 1.
APPLICATIONS TO THIS COURT
Against the orders of the Court of Appeal disposing of the applications in respect of the murder convictions, the applicants made their applications to this Court. Those applications, mounted separately, raised several grounds. One issue, common to each of the three applications, was a request by the applicants that this Court should reconsider and re-express the law on extended common purpose liability for murder as stated in McAuliffe and as applied most recently in Gillard. It was that issue that the Special Leave Panel (Hayne and Heydon JJ - see  HCATrans 331) reserved to the Full Court.
In my reasons in Gillard at 16-17 -, I noticed a number of doctrinal difficulties presented by the law of extended common purpose liability as stated in McAuliffe. I did so because the ultimate issue in Gillard was whether the trial judge had erred in failing to direct the jury as to the availability of a conviction of the appellant in that case of manslaughter. Despite arguably strong evidence exculpating Mr Gillard from the consequences of the intention of his co-offender to kill the deceased in that case - (2003) 219 CLR 1 at 22-23 , if the principles in McAuliffe were applied in their full rigour, they appeared to leave little, if any, space for the conviction of Mr Gillard of manslaughter. Nevertheless, in the end in Gillard at 27-31 -, together with the other members of this Court, I found a sliver of room for manslaughter. I therefore joined in the orders for the retrial of Mr Gillard, made in that case.
CHALLENGES TO McAULIFFE AND GILLARD
In joining in the orders in Gillard at 22 , I expressed my unease that an opportunity had been missed in that appeal to re-examine McAuliffe and "to clarify and simplify the task of trial judges and the juries they instruct". The other members of this Court in Gillard applied the principles in McAuliffe without questioning them. In his reasons, Hayne J in (2003) 219 CLR 1at 36 , expressed the view that "[r]econsideration of McAuliffe [was] neither sought nor required". Gummow J, ibid at 15 , expressed agreement with those remarks.
In these applications the issue presented tangentially in Gillard was squarely raised. The applicants, facing sentences of extended terms of imprisonment for murder, asked for leave to afford an opportunity to re-examine the law stated in McAuliffe and Gillard. It is that question that the Special Leave Panel referred to a Full Court for decision. The three applications were consolidated and returned for hearing before the entire Court, sitting in Adelaide. Whilst advancing certain individual arguments appropriate to the case of each applicant, counsel substantially shared the common attack on what they suggested were the defects of justice, logic and rationality in the legal rule expressed in both McAuliffe and Gillard.
At the conclusion of the oral argument, the Court adjourned briefly and then announced its refusal to reconsider the authority stated in the two cases. An order was made dismissing the three applications. The Court's reasons for adopting this course were reserved.
At the conclusion of the oral argument, I preferred to avail myself of the opportunity to reflect on the applications and the arguments advanced for and against them. My own preference in this respect could not delay the orders of the Court, my colleagues being firm in their conclusions and proposed orders. I have taken time for study and reflection before reaching my conclusions and stating my reasons and orders. Upon further consideration, it is my view that the applicants were entitled to succeed. The time has come to re-express the Australian law of extended common purpose liability so as, at least in homicide cases where those rules are most important, to restore greater concurrence between moral culpability and criminal responsibility; to reduce doctrinal anomalies and asymmetries; and to reduce the risk of miscarriages of justice in the applicants' particular cases arising from the application of the law as hitherto expressed.
NEED TO EXPLAIN THE CHALLENGES
Without explaining the factual background of the applications in greater detail, their merits may not be fully understood. Without recounting the arguments that the applicants deployed, the disposition that I favour will not be appreciated. Without illustrating the serious doctrinal weaknesses of the comparatively recent expressions of the law on extended common purpose, the reasons for favouring re-expression will not be perceived. Without recounting the past failures of others to repair the defects in this branch of the law, the necessity for this Court to accept its own responsibility for re-expressing the common law, shown to be defective, will not be accepted.
Although, therefore, my reasons constitute a minority opinion, out of fairness to the arguments of the applicants and to those who in the future, in different times, may return to this issue, I will explain, with no more detail than necessary, my differing opinion and conclusion. When a like question was before the House of Lords in 1997, in R v Powell  1 AC 1, two of their Lordships (at 11 per Lord Mustill, 14-15 per Lord Steyn - see also at 25 per Lord Hutton) specifically acknowledged the anomalies in the law of homicide now complained of in these applications; the need to address such anomalies; and the hope that the legislature would do so. While acknowledging the capacity of the legislature to rectify such defects in the law, on reflection I have concluded that no further delay should be tolerated. This part of the common law is in a mess. It is difficult to understand. It is very hard to explain to juries. It involves a portion of the law made by judges. What the judges have expressed with imperfect results, they can re-express with greater justice and rationality in the light of experience and the submissions in this case - cf Chamberlains v Lai  NZSC 70 at  per Elias CJ, Gault and Keith JJ. Ultimately, in expressing and applying the common law in Australia, that is the responsibility of this Court - Lange v Australian Broadcasting Corporation (1997) 189 CLR 520 at 565-566. It is a responsibility that we should be ready to shoulder in these proceedings.
Some of the background facts leading to the death of the deceased are set out in the reasons of the other members of the Court ("the joint reasons" - at -). As to the main facts, there was no dispute. However, as to particular facts, relevant to the respective parts played by each of the applicants in the assaults upon, and ultimate fatal stabbing of, the deceased, there were differences in the versions given by the respective applicants in their records of interview with the police.
None of the applicants gave oral evidence before the jury. The jury therefore had to rely, substantially, on the applicants' several video-tape recorded interviews with the police. In this Court, the parties relied on the factual background as summarised by the Court of Appeal - Hartwick (2005) 159 A Crim R 1 at 5-8 -. That Court, in turn, drew upon the police interviews and the other evidence adduced at the trial. The applicants also sought to supplement the summary of the Court of Appeal with the summary of the findings of fact given by Smith J in his reasons for sentence - R v Hartwick, Hartwick and Clayton  VSC 63 at -.
One of the complaints advanced by the applicants against the law of extended common purpose, as stated in McAuliffe and Gillard, was that it excludes the jury from the requirement, and responsibility, of differentiating the moral culpability of each criminal accused, so as to determine their separate, respective responsibilities for the offence in issue. Typically, in homicide cases, any suggested disparity in the moral culpability of accused, alleged to have acted in common, arises from what is claimed to be an unpredicted, unexpected and unwished-for introduction by one of the accused of a weapon, later used to kill the deceased. Essentially, the applicants complained that this Court's doctrine of extended common purpose unjustly encourages juries to lump all accused together, finding them all equally guilty but without addressing in a more precise way, as is usual for criminal punishment, the individual responsibility of each for the acts proved against them severally. According to the applicants, the present approach has led to lazy and unprincipled determinations. Such an approach, particularly in homicide cases where the doctrine is specially important, works injustice and departs from the basic principles of criminal responsibility now accepted in our law.
In the applicants' joint trial, the prosecution did not prove which of the applicants had actually administered the fatal knife wound to the deceased - Hartwick (2005) 159 A Crim R 1 at 7-8 . In the way the trial was conducted, it was virtually certain that it was one of the applicants. However, because of the ambit of the present extended common purpose doctrine, the applicants were each liable to be found guilty of the murder of the deceased. This was so, although the jury might, if they had been required to do so, have decided that one, two or all three of the applicants had not actually intended the deceased's death or had not regarded it as a virtually certain or a "probable" outcome of participating with the others in the shared purpose they actually had when they engaged in their activity in common.
Given that this complaint goes to the heart of the present applications, it is appropriate to say a little more about the factual details and the part that each applicant took in the events leading to the deceased's death, as disclosed by the evidence.
The deceased had crashed a stolen car into a car rented by Ms Clayton which had been parked outside Mrs Hartwick's home in Karingal. The collision pushed Ms Clayton's car into another car, which in turn collided with the front wall of Mrs Hartwick's house, causing much damage - Hartwick at 5-6 ; R v Hartwick, Hartwick and Clayton  VSC 63 at . None of the applicants actually saw the deceased driving the offending vehicle -  VSC 63 at . Nevertheless, they assumed that the incident was connected with an earlier altercation between Mrs Hartwick and Ms Rodwell (the deceased's domestic partner). That altercation had initially arisen out of a belief by Mrs Hartwick that Ms Rodwell had accused Mr Hartwick (Mrs Hartwick's ex-husband) of being a "dog", i.e. a police informer - (2005) 159 A Crim R 1 at 5-6 ;  VSC 63 at -.
Following the initial altercation, Mr Hartwick informed Mrs Hartwick that her belief in this respect had been mistaken and that it was the deceased, not Ms Rodwell, who had made the objectionable allegation - (2005) 159 A Crim R 1 at 5-6 ;  VSC 63 at -. This news led to the apparent conclusion of the dispute with Ms Rodwell. But when, shortly afterwards, the car crash occurred, the inference that the deceased had caused it, in retaliation for the earlier confrontation with Ms Rodwell, was inescapable. The three applicants became "very angry and agitated" - ibid at 6 . A neighbour said that she saw both Mr Hartwick and Ms Clayton leaving Mrs Hartwick's house with a knife. Another described the two women applicants as angry, with Mr Hartwick telling them to "calm down" and promising that he would "sort it out" - ibid at 6 .
For the purposes of sentencing the applicants, following the jury verdicts and their convictions of murder, the trial judge made a number of findings upon which the Court of Appeal was also prepared to act -  VSC 63 at -; (2005) 159 A Crim R 1 at 5-7 -. These included that, within thirty minutes or so of the car crash, the applicants drove to Ms Rodwell's house with the intention of assaulting the deceased. They had with them at least one knife and some poles (variously described as a vacuum cleaner pipe, a shopping trolley handle and a wooden cricket stump). They demanded entry to the house where Ms Rodwell and the deceased had just consumed a cap of heroin - (2005) 159 A Crim R 1 at 5-6 . On Ms Rodwell's urgings, the deceased temporarily left the premises by the back door. When Ms Rodwell opened the front door of the house, Ms Clayton struck her on the head with one of the poles. This incident was the subject of the second count of the presentment upon which each of the applicants was also found guilty by the jury and convicted but in respect of which the convictions were quashed and a new trial ordered.
Inside the house, Ms Rodwell was held at knife-point throughout most of what ensued either by Ms Clayton or Mrs Hartwick. At various times, Mrs Hartwick and Mr Hartwick attempted to stem the bleeding from the wound received by Ms Rodwell when she was struck on the head.
Mr and Mrs Hartwick went through the house looking for the deceased and searching for items of value, inferentially to compensate them for the damaged vehicles and premises. It was when they were so engaged that the deceased suddenly re-appeared. He was armed with a knife. A struggle ensued in which the deceased was severely beaten with the poles. Five of his teeth were knocked out. His left ear was almost severed. According to Ms Rodwell, the deceased and Mr Hartwick were observed attempting to "dag" (stab) one another. Mr Hartwick called for assistance from his fellow intruders. It was at some stage during the ensuing mêlée that the deceased was fatally stabbed on the left side of his chest. His heart and lung were punctured by the knife. Mr Hartwick was also stabbed, possibly by the deceased but, according to Ms Rodwell, by Mrs Hartwick, as a result of an accident. The applicants took items from the deceased's pockets including money and cannabis. Mrs Hartwick made some effort to remove identifiers of the presence of the applicants at the scene of the crime. The three applicants then decamped leaving Ms Rodwell alone with the deceased's body.
Although the broad outline of what happened, thus described, was clear enough, the respective records of interview introduced differential assertions of culpability on the part of each of the applicants.
In her interview, Ms Clayton said that, whilst she had accompanied the others to Ms Rodwell's home, her role was largely that of a look-out whilst the others were engaged in ransacking the premises. Ms Clayton said that throughout the event, she was holding a knife at Ms Rodwell's throat whilst the latter was seated on a couch beside her. She said that she knew that the deceased was receiving a severe beating. However, she claimed that she did not fully appreciate what the others were doing to the deceased, that resulted in his death.
Armed with the present statement of the law on extended common purpose, the prosecution argued before the jury that Ms Clayton was guilty of murder on the basis of her own statements to the police. This was so, notwithstanding her claim that she had taken no part in the fatal stabbing; had no intention to participate in violence of that degree; no actual participation in the physical attacks on the deceased; and no involvement in the ultimately fatal act of stabbing the deceased.
Mr Hartwick's record of interview was more guarded. He also denied that he was guilty of murder. Initially, he said that he believed that the police had got the wrong people. However, although he never admitted to being in the premises where the deceased was killed, his defence at the trial was conducted on the basis that he was present when the deceased was attacked. His essential case was that he did not stab the deceased; did not have any weapons; and had no intention to kill or inflict really serious injury on the deceased. To the contrary, he was the "peace maker who wanted to sort things out".
Mr Hartwick attended the Frankston Hospital for treatment for his own stab wounds. It was argued that, if he had had anything to hide, he would not have done this. He claimed that he did not at first realise that he had been stabbed. Additionally, Mr Hartwick relied on Ms Rodwell's evidence of the "dagging" fight. On the basis of this evidence, he claimed that, if it was he who had stabbed the deceased, he only did so in self-defence which the prosecution had failed to exclude as the cause of the fatal stabbing.
Mrs Hartwick's case was also substantially contained in her record of interview with police. She claimed that she had accompanied Mr Hartwick and Ms Clayton after the car crash because, otherwise, she would have been bashed by Mr Hartwick. She claimed not to have known that Ms Clayton had a knife with her until after she had seen it when Ms Clayton struck Ms Rodwell as she opened the door of her home. Mrs Hartwick said that she had stayed near the door throughout most of the ensuing events, although at one point she had obtained a towel for Ms Rodwell because she was bleeding from the head. She described how the deceased had suddenly "come from nowhere" and stabbed Mr Hartwick, leading to the scuffle between the two men. She described how Ms Clayton became involved in that scuffle and denied that she had assaulted or touched the deceased in any way. She admitted that she and Ms Clayton had driven to her home together after the fatal stabbing and that Mr Hartwick had returned there separately. He and Ms Clayton had asked her to clean two knives, which she did. One of these was a knife from her own house which, by then, she understood to have been in Ms Clayton's possession. The other knife she understood to have been in the deceased's possession. According to Ms Rodwell, when the applicants were leaving the scene of the crime, Mrs Hartwick had said to her "You know the rules - no dogging" - (2005) 159 A Crim R 1 at 7 .
THE PROSECUTION CASE
In respect of the first count of the presentment (murder) the prosecution accepted that it could not prove which of the applicants had inflicted the fatal stab wound on the deceased. However, according to the prosecution, as a matter of law, that incapacity did not matter. Each of the applicants was guilty of murder equally with the others. This submission was advanced upon the basis of the three legal principles on which the prosecution relied to inculpate in the killing of the deceased whichever of the two applicants who were not involved in the fatal blow, equally with whichever applicant actually stabbed the deceased.
The prosecution advanced three bases to establish liability for murder as equal possibilities - see (2005) 159 A Crim R 1 at 7-8 :
The first basis was that the applicants were each acting in concert. In particular, the prosecution submitted that each of the applicants had gone to Ms Rodwell's home with the agreement and intention to inflict really serious injury on the deceased and that, pursuant to this agreement, the deceased was stabbed with the intention of causing such really serious injury. The prosecution did not submit that the three applicants left Mrs Hartwick's home actually intending to kill the deceased. Having regard to the evidence adduced at the trial, the prosecution accepted that such a conclusion would be unrealistic -  VSC 63 at . However, to obtain convictions of murder for acting in concert, it was enough if the jury concluded that the applicants had gone to confront the deceased and to inflict on him really serious injury. That, it was argued, the prosecution had proved to the requisite standard;
The second basis relied on extended common purpose liability. Amongst other ways, this ground of accessorial liability was advanced on the footing that the applicants had agreed to assault the deceased with a weapon or weapons and that each of them "foresaw as a possibility in the carrying out of the agreed understanding or arrangement that death or really serious injury would occur by a conscious, voluntary and deliberate act of one of them not done in self defence" - written directions provided by Smith J to the jury, DPP Exhibit 22 (emphasis added). This argument affords the suggested foundation for the jury's verdicts that the applicants challenged in these applications; and
The third basis was aiding and abetting. Amongst other ways in which this basis of liability was put, the prosecution claimed that each applicant was present when the deceased was stabbed and each was "aware that the Deceased was being consciously, voluntarily and deliberately assaulted with intent to kill or cause really serious injury and not in self-defence" and that each applicant "intentionally helped the stabber to commit the crime"; or "intentionally encouraged" him or her by words, presence or behaviour; or "intentionally conveyed" to him or her "assent to or concurrence in the commission of the crime" - DPP Exhibit 22.
THE EMERGING ISSUE OF RESPONSIBILITY
From the foregoing description of the three ways in which the prosecution separately urged the jury to find each of the applicants (including those who had not inflicted the fatal stab wound on the deceased) guilty of murder, it will be noticed that there is an important distinction between the second and the other two propounded bases of legal liability.
To secure a conviction of murder on the basis either of the jury's conclusion that the applicants had acted in concert, or were guilty of aiding and abetting whichever one of them actually inflicted the fatal wound, an ingredient of legal liability in each case was the jury's satisfaction, beyond reasonable doubt, of the requisite "intention" of each accused. By way of contrast, to secure a conviction on the basis of the extended common purpose of the three accused, the prosecution was not obliged to establish a relevant specific intention. It was sufficient that the prosecution should establish something short of intention on the part of each applicant for a verdict of guilty of murder to be returned. It was enough if the jury concluded that each applicant foresaw the possibility that death or really serious injury would occur from the deliberate act of one of them.
This feature of extended common purpose liability (in this case for the crime of murder) involves an important distinction, both in law and on the facts. It highlights the contested ways in which the prosecution case against the applicants was presented on the basis of extended common purpose liability, on the one hand, and alleged acting in concert or aiding and abetting the principal offender (whoever that might have been), on the other. The form of the case brought in respect of the two latter charges is more consonant with the usual principles of criminal liability recognised by our law.
Because the verdicts of the jury in the trial of the applicants did not differentiate the foundation upon which the jury convicted each of the applicants, it is impossible for this Court to exclude the available inference, on which each of the applicants relied, that the jury could have proceeded to determine the liability of each applicant by the easiest, and most readily proved, of the bases which the prosecution had advanced - cf Domican v The Queen (1992) 173 CLR 555 at 566-567. Because that inference could not be excluded, it was necessary, in assessing the applicants' common legal challenge before this Court, to assume the jury might have proceeded to decide the guilt of the applicants of the crime of murder by applying the stated law on extended common purpose liability. The applicants complained that this illustrated the injustice involved in the disharmony between the establishment of criminal responsibility on the basis of extended common purpose and doing so on the prosecution's other proposed bases of accessorial liability. This was particularly unjust where, to secure conviction, it was enough for the prosecution to prove that each applicant merely foresaw what happened as a possibility in carrying out the common understanding. Anything is possible, the applicants submitted. Why, they asked, should they be convicted of murder not on the basis of their proved individual intentions but on the basis of estimations of possible foresight of eventualities that had certainly transpired?
THE UNEDIFYING CIRCUMSTANCES
The facts that I have earlier described, leading to the violent death of the deceased, present an unedifying story of anger and violence. None of the applicants emerges from the story with much credit. It is understandable that their appeals to legal principle and to the correct legal doctrine might occasion a degree of impatience and an intuitive feeling that the applicants deserve little sympathy. However, that is not the way the criminal law functions in Australia. Often, important principles are established in cases involving unattractive individuals and confronting facts. It is when the law deals with such cases and such individuals that it is tested for its adherence to basic principle - cf Adelaide Company of Jehovah's Witnesses Inc v The Commonwealth (1943) 67 CLR 116 at 124.
Having been found guilty together, the applicants were each convicted of murder and sentenced to very lengthy terms of imprisonment. Their counsel have argued the point of legal principle which they advanced in this Court. They have asked this Court to re-express the applicable rules as to accessorial liability in their case and, hence, for the many others which (as the facts of Gillard at 22-23  illustrate) may occasion an even greater sense of individual injustice over the outcome that follows from the present expressions of the law on extended common purpose liability.
For England, the House of Lords, in Powell, addressed many of the issues argued in these applications in a hearing that lasted three days. The appeal was relisted for orders five months later. Their Lordships delayed the publication of their speeches for a further three months. Their reasons reveal a full awareness of the legal anomalies and asymmetries that were at stake -  1 AC 1 at 1-10. For the justice and integrity of the common law of Australia, this Court should be no less attentive to the detailed arguments that the applicants have urged upon us.
Three issues were presented in these applications. They are:
Reconsideration of extended common purpose
Should this Court re-express the Australian common law with respect to the doctrine of extended common purpose liability in the criminal law? Is such re-expression required or appropriate having regard to the comparatively recent statements of this Court on the subject in McAuliffe and Gillard? If anomalies, inconsistencies and asymmetries of legal doctrine are revealed, is the present expression of the principle sustainable as a practical rule of accessorial liability, apt to circumstances such as the instant case? If convinced that re-expression of the law is required, should this Court leave any such re-expression to the legislature, assisted by law reform or like bodies? Or should it perform the re-expression in discharge of its own functions as the final national court of appeal entrusted with stating, and where necessary re-expressing, the basic principles of the common law applicable throughout Australia?
Re-expression of the doctrine
If it is concluded that the defects in the extended common purpose doctrine are such as to justify, and require, a re-expression by this Court of the applicable rule of the common law, should the re-expression state that to secure convictions of co-accused on the basis of an alleged common purpose, the prosecution must prove that the secondary offender, which it seeks to render liable equally with the principal offender, desired the latter to act with the intention that he or she did, or knew that it was "not merely a 'real possibility' [that the primary offender might do so] but virtually certain that he would do so" - Smith, "Criminal Liability of Accessories: Law and Law Reform", (1997) 113 Law Quarterly Review 453 at 465 (original emphasis)? Or should any such statement render the secondary offender equally liable with the primary offender on the ground of extended common purpose with the primary offender, only if the prosecution proves that the secondary offender foresaw as a probability, in the carrying out of the agreed understanding or arrangement, that death or really serious injury would occur by the act of one of them - cf R v Crabbe (1985) 156 CLR 464?
Disposition of the appeals
If it is concluded that the liability of co-accused on the basis of extended common purpose at common law should be re-expressed, what dispositions should follow in these proceedings:
The setting aside of the applicants' convictions and the ordering of a retrial on the counts of murder (as the applicants each sought); or
The setting aside of the orders of the Court of Appeal dismissing their appeals and the remitter of the proceedings to the Court of Appeal for disposal by that Court in the light of the re-expression by this Court of the common law; the consideration of the application of such law in the applicants' respective cases; and also the consideration of whether the applicants' convictions ought nonetheless to be confirmed by the application of the "proviso" (Crimes Act 1958 (Vic), s 568(1)), having regard to the principles explained by this Court in that respect in Weiss v The Queen (2005) 80 ALJR 444 at 452-456 -; 223 ALR 662 at 671-675?
JUSTIFICATION AND MAINTENANCE OF EXTENDED COMMON PURPOSE LIABILITY
JUSTIFICATION OF THE DOCTRINE
Before this Court, the prosecution rejected the applicants' criticisms of the doctrine of extended common purpose. It defended that doctrine as a long-standing feature of the common law of Australia. However, the oldest authority in this Court cited by the prosecution in support of that proposition was decided as recently as 1980.
In fact, as Sir Robin Cooke pointed out, giving the advice of the Privy Council in Chan Wing-Siu v The Queen  AC 168 at 176, a Hong Kong appeal, the need to express the test for such liability more elaborately in terms that a jury would have to apply came about "in association with the modern emphasis on subjective tests of criminal guilt". The traditional expression of the test for such liability had originally been voiced in objective terms. It is in those terms that the test found its way into the Queensland Criminal Code drafted by Sir Samuel Griffith - See Criminal Code (Q), ss 7, 8 and 9; Darkan (2006) 80 ALJR 1250 at 1261-1262 -, 1274 ; 228 ALR 334 at 346-347, 363. The current law on the subject is thus scarcely a law of great antiquity.
Nevertheless, the prosecution emphasised that the Australian law on extended common purpose liability had been stated (in Johns (1980) 143 CLR 108), reaffirmed (in McAuliffe (1995) 183 CLR 108) and recently reapplied (in Gillard (2003) 219 CLR 1) by this Court in unanimous decisions. The starting point for the applications was thus from a position where the applicable authority was abundantly clear. The prosecution submitted that, to abolish, or modify, the doctrine of extended common purpose, so stated, would, to the extent that this course was attempted, unsettle the present law in a way that could not be justified.
Viewed from its interests, the prosecution's defence of the present law was vigorous and understandable. As expressed in McAuliffe and later cases the impugned law is highly favourable to the prosecution. In instances involving multiple offenders acting to some extent in concert, it eases the path to securing the conviction of all participants, without irksome differentiation between them or troublesome attention to what the prosecution has actually proved about the acts and intentions of each of the individuals accused. The prosecution justified this approach on the basis that it tendered a realistic criterion for criminal liability of those who become involved in common criminal enterprises of the kind illustrated by the facts of this case.
Secondly, whilst accepting that extended common purpose liability did not require proof of actual common intention (or even actual foresight of the virtual certainty or probability of what in fact occurred) the prosecution emphasised that the requisite foresight nonetheless imposed a type of subjective criterion, albeit one that needed only to be proved to the level of a "possibility".
In order to obtain an accessory's conviction, it still remains for the prosecution to establish beyond reasonable doubt that the secondary offender agreed to take part in a criminal enterprise; that he or she participated or continued to participate despite being aware that, in carrying out the agreed enterprise, it was possible that another participant might commit the more serious crime; and that to be found guilty as an accessory to murder, it was necessary for all elements of the crime to have been established in the course of carrying out the criminal enterprise. This meant proving that the principal offender must have acted with the necessary intent, in a case such as the present, to kill the victim or cause really serious injury to the victim - Gillard at 11-12 -, 13-14 -, 28-29 -. See also Powell at 12 per Lord Steyn. The prosecution submitted that these remained rigorous requirements which could be safely left to the good sense of jury determination. It suggested that some of the criticisms voiced by the applicants were more theoretical than real - cf Powell at 12 per Lord Mustill, 25 per Lord Hutton.
Thirdly, the prosecution contested that there were anomalies in the doctrine of extended common purpose which necessitated a re-expression of the law to make its operation more principled, simpler, and fairer for all concerned. Thus, the prosecution laid emphasis on the requirement that, in every case, it must prove beyond reasonable doubt that there was an agreement to participate in a criminal enterprise. It submitted that this was such a fundamental element of liability that comparison with other forms of criminal complicity, absent that factor, was illogical and unhelpful.
Whilst this is a fair deployment of rhetoric by the prosecution, its force breaks down somewhat when the content of the agreement is explored. Thus, it might have been an agreement to attack another person. Typically in cases of homicide, the problem for joint liability is presented when one participant to the common purpose goes much further than the others say they desired, anticipated or expected. The question then for consideration is whether, in such circumstances, by a kind of legal fiction, the law should attribute to the secondary participant equal liability for the outcome of the criminal enterprise, although that outcome was never in fact desired, anticipated or expected by the secondary offender but only by the principal.
Fourthly, the prosecution supported the present law of extended common purpose liability by reference to policy arguments addressed to the desirability of discouraging gangs and other persons acting in common participating in criminal enterprises which have a nasty tendency to escalate and go wrong - Powell at 25 per Lord Hutton with whom the other Law Lords agreed: Lord Goff of Chieveley at 10, Lord Jauncey of Tullichettle at 10, Lord Mustill at 12, and Lord Steyn at 15. Certainly, this was an important consideration in dissuading the House of Lords from re-expressing the English law in this respect, although their Lordships acknowledged anomalies, lack of symmetry and other defects in the resulting law - Powell at 12 per Lord Mustill, 13 per Lord Steyn. Thus, in Powell at 25, Lord Hutton said:
I recognise that as a matter of logic there is force in the argument advanced on behalf of the appellants, and that on one view it is anomalous that if foreseeability of death or really serious harm is not sufficient to constitute mens rea for murder in the party who actually carries out the killing, it is sufficient to constitute mens rea in a secondary party. But the rules of the common law are not based solely on logic but relate to practical concerns and, in relation to crimes committed in the course of joint enterprises, to the need to give effective protection to the public against criminals operating in gangs. [His Lordship referred to R v Majewski  AC 443 at 482 per Lord Salmon.]
A similar explanation is given by A P Simester and G R Sullivan in their text Criminal Law Theory and Doctrine, 2nd ed (2003) at 221. Responding to the anxieties voiced by some of their Lordships in Powell, the authors express doubt that it is possible to "distinguish so readily between matters of intelligence and those of practicality". They offer their own justification for treating common criminal liability for the outcomes of a joint enterprise as "a special case of complicitous participation and not merely a sub-species of assistance and encouragement": Ibid at 224. They support the stance of the law expressed in Powell (affirmed for Australian purposes by the then recent decision of this Court in McAuliffe cited in 1997 in Powell at 21, 27 per Lord Hutton) in these terms - Simester and Sullivan, supra at 226. See also Simester, "The Mental Element in Complicity", (2006) 122 Law Quarterly Review 578 at 592-601:
The law has a particular hostility to criminal groups: conspiracy to defraud, for example, is an offence even where individual fraud is not. As with the inchoate crime of conspiracy, the rationale is partly one of dangerousness: 'experience has shown that joint criminal enterprises only too readily escalate into the commission of greater offences.' [Powell  1 AC 1 at 14.] Criminal associations are dangerous. They present a threat to public safety that ordinary criminal prohibitions, addressed to individual actors, do not entirely address. Moreover, the danger is not just of an immediately physical nature. A group is a form of society, and a group constituted by a joint unlawful enterprise is a form of society that has set itself against the law and order of society at large. Individuals offending alone do not do this. Thus concerted wrongdoing imports additional and special reasons why the law must intervene.
This argument represents the strongest policy justification for the exceptional ambit of the present Australian common law rule of extended common purpose liability. It constitutes a justification for a rigorous principle of accessorial liability that goes beyond that imposing liability for acting in concert and for aiding and abetting (upon which the prosecution alternatively relied in its proceedings against each of the applicants). The question, however, remains whether the resulting formulation is too drastic a departure from the now ordinary requirements that the prosecution must prove that the intention of the accused went with his or her conduct. And whether, in homicide cases like the present, the formulation now adopted leaves adequate room for the offence of manslaughter, as an alternative to the crime of murder, in terms that are realistic so that judges can explain the distinction and so that jurors can understand it - Hartwick at 19-24 -. In fact, at the trial of the applicants, manslaughter was left to the jury, although not in terms entirely to the applicants' liking.
Although in this case, as in Gillard, manslaughter, by law, remained an available verdict in respect of each applicant, the difficulty of sustaining it, as a matter of practicality, was correctly noted by the Court of Appeal - (2005) 159 A Crim R 1 at 24 . How, for example, knowing that the applicants were armed with weapons of the kind described (a knife, a vacuum cleaner pipe or a cricket stump), the jury could properly and reasonably reach a conclusion of an intention to inflict a trivial injury on the deceased is difficult to see. There may be space for manslaughter, as Gillard holds. But in all truth it is a tiny and elusive one.
SUGGESTED MAINTENANCE OF THE DOCTRINE
In addition to its defence of extended common purpose liability, the prosecution argued that, regardless, this Court should refrain from re-expressing the law on the subject. It should stay its hand, as the House of Lords did in Powell, even if it were to conclude that the present law was defective. These arguments are by no means negligible. It is necessary to notice them.
First, the prosecution pointed out that the current expression of Australian law on extended common purpose liability was generally compatible with statements of the common law in England (Powell at 29-30); Northern Ireland (Gamble at 283-284); New Zealand (R v Tomkins  2 NZLR 253 at 255-256) and in the Privy Council in respect of Hong Kong (Chan Wing-Siu  AC 168 at 176-179). In these circumstances, demonstration of a need to strike forth on a new Australian approach was said to be unproved. On the other hand, both in the New Zealand and Privy Council decisions (unsurprising because of the common author, Sir Robin Cooke) a series of formulae, expressed otherwise than in terms of "possibility", are advanced, including "substantial risk", "real risk", and "a risk that something might well happen" (Tomkins at 255-256; Chan Wing-Siu  AC 168 at 179), revealing a degree of ambivalence in the foregoing authority about the test to be applied. In Powell, the anomalies and difficulties of the present expressions of the law on extended common purpose liability were acknowledged but put aside for suggested reasons of practicality rather than legal principle.
Secondly, the prosecution argued that, if any change to the expression of the Australian common law on this subject were to be attempted, it should be left to the legislature, acting with the assistance of a law reform or other advisory body. The Court was reminded that, in Victoria, Parliament had reformed the law with respect to felony murder - Crimes Act 1958 (Vic), s 3A. Within Australia, law reform agencies have addressed aspects of the law of homicide. An offence of "wilful murder" was introduced in (and also removed from) the Griffith Queensland Criminal Code. More recent investigations into the law of homicide have considered the introduction of degrees of murder as recognised in some jurisdictions of the United States - New South Wales Law Reform Commission, Partial Defences to Murder: Diminished Responsibility, Report No 82, (1997); see R v Lavender (2005) 222 CLR 67 at 109-110 -.
Against the background of law reform investigations (and reports of the Criminal Law Officers Committee of the Standing Committee of Attorneys-General in Australia - Discussion Draft: Model Criminal Code, Chapter Two, General Principles of Criminal Responsibility, (1992) at 79), the prosecution argued that this Court should leave any reform that might be considered appropriate to Parliament. Obviously this submission has force. I understand the weight accorded to it in the joint reasons at . Not least must the argument be given weight because of a recent report of the Law Commission of England and Wales (see Inchoate Liability for Assisting and Encouraging Crime, Law Com No 300, (2006) Cm 6878 at 19 [2.24]-[2.25]) rejecting the proposition that legislation should be recommended to the United Kingdom Parliament to overturn the effect of the decision in Powell.
CONCLUSION: NEED FOR CHANGE
I therefore acknowledge the merits of the arguments deployed against the applicants' submissions. They have persuaded the majority of this Court - joint reasons at - - who have decided that the law on extended common purpose, as expressed in McAuliffe and Gillard, should not be reconsidered. I have reached the opposite conclusion for reasons that I will now express.
THE NEED FOR CHANGE IN EXTENDED COMMON PURPOSE LIABILITY
DEFECTS IN LIABILITY FOR POSSIBILITIES
There are many defects in the statement of the applicable law of extended common purpose liability for the criminal acts of others. No doubt they arise from time to time in other offences, but they are most visible where the offence involved is one of homicide. This is the area of legal discourse to which this Court, and other courts, keep returning. They do so, because murder is ordinarily a crime of specific intent. Obviously, murder is one of the most serious crimes provided by law. In Australia, upon conviction, it often carries the mandatory imposition of a sentence of life imprisonment. In our society, a conviction of murder also carries a special opprobrium because of the particular respect accorded to human life and the special revulsion felt towards those who unlawfully terminate it - cf Charlie v The Queen (1999) 199 CLR 387 at 399 .
In Powell, in the context of a law requiring a mandatory life sentence upon conviction for murder, Lord Steyn observed at 15:
There is an argument that, given the unpredictability whether a serious injury will result in death, an offender who intended to cause serious bodily injury cannot complain of a conviction of murder in the event of a death. But this argument is outweighed by the practical consideration that immediately below murder there is the crime of manslaughter for which the court may impose a discretionary life sentence or a very long period of imprisonment .... [T]he problem is one of classification. The present definition of the mental element of murder results in defendants being classified as murderers who are not in truth murderers. It happens both in cases where only one offender is involved and in cases resulting from joint criminal enterprises. It results in the imposition of mandatory life sentences when neither justice nor the needs of society require the classification of the case as murder and the imposition of a mandatory life sentence.
The same reasoning applies where, as under the law of Victoria, persons, like the applicants, convicted of murder on the basis of extended common purpose liability, must each, consistent with that conviction, receive very lengthy terms of imprisonment that reflect their common legal liability of this most heinous crime.
Secondly, in all criminal offences (but particularly in the offence of murder) it is highly desirable that legal responsibility should generally accord with community notions of moral culpability - Johns, McAuliffe, Gillard, Powell and Chan Wing-Siu were all cases involving extended common purpose liability for homicide.
The principle of adhering to "a close correlation between moral culpability and legal responsibility" is not one expressed by philosophers, ethicists, academics and social scientists alone. The words were used by Mason CJ, Toohey, Gaudron and McHugh JJ in their joint reasons in Wilson v The Queen (1992) 174 CLR 313 at 334. Their Honours were there considering whether "battery manslaughter" remained part of the common law in Australia. They observed at 332 that "[t]he notion of manslaughter by the intentional infliction of some harm carries with it the consequence that a person may be convicted of manslaughter for an act which was neither intended nor likely to cause death". Their Honours therefore concluded at 334 that a conviction for manslaughter in circumstances that had previously been recognised as battery manslaughter "does not reflect the principle that there should be a close correlation between moral culpability and legal responsibility, and is therefore inappropriate". This Court therefore decided that battery manslaughter was not part of the Australian common law of homicide. Manslaughter by an unlawful and dangerous act was retained, although in an amended form.
In effect, in these proceedings, the applicants asked this Court to undertake a similar analysis, and to apply the same criterion, as adopted in Wilson. They submitted that proof by the prosecution of no more than the possibility that a principal offender might intentionally cause grievous bodily harm to a victim did not, of itself, establish conduct sufficiently culpable to warrant conviction of the offence of murder. They suggested that to punish persons who have "neither mentally nor physically committed an offense to the same extent as .... those who have" (Mueller, "The Mens Rea of Accomplice Liability", (1988) 61 Southern California Law Review 2169 at 2173) was an unjustifiable departure from the fundamental tenet of the Australian criminal justice system, as it has now evolved. They argued that the accessory's level of foresight and intention, posited by the present Australian legal test, was insufficient to warrant conviction of murder.
Such evolution has seen developments in the features of accessorial liability long recognised by the common law. As the joint reasons in this Court pointed out in Darkan v The Queen (2006) 80 ALJR 1250 at 1266 ; 228 ALR 334 at 353, different mechanisms have been used, over time, to limit a secondary offender's liability for acts done by a principal offender. Originally, the limit was expressed objectively in terms of the secondary offender's foresight of the probable consequences of the common enterprise. This view, frozen in time, is still reflected in Code provisions, as in Queensland. However, in the meantime, the common law has moved toward a general harmony with the recognition of the requirement for a subjective element (mens rea) to render particular conduct (actus reus) criminal in character.
The test adopted by the common law to constitute what is, in effect, the subjective element in crimes established by extended common purpose liability, falls short of obliging proof of actual intent. All that is required is that the relevant outcome must be foreseen by the accessory as a possibility. The applicants argued that this step forward did not go far enough. It exposes a secondary offender to liability of conviction of murder upon proof by the prosecution of nothing more than foresight of the possibility of homicide. In my view, there is force in the applicants' submission that many juries are likely to conclude that the fact that a murder has occurred shows that it was possible that it would. And if it was possible in fact, it is but a small step to conclude that the secondary offender foresaw, as a possibility, at least that in effecting the common purpose, the victim might suffer really serious harm with intent from the act of the principal offender.
It follows that this form of secondary liability is disproportionately broad. It tilts the scales too heavily in favour of the prosecution.
It is also the experience of the criminal law that subordinate offenders, who become involved in common criminal enterprises, are sometimes weak, impressionable, vulnerable individuals whose will is insufficient to resist the unexpected, violent acts perpetrated by a ring-leader. Such was certainly the case of Mr Gillard (Gillard (2003) 219 CLR 1 at 22-23 ), who looked up to the principal offender as the "ringmaster" of their joint enterprise. Such was also the case of Mr English, who succeeded before the House of Lords where the other appellant, Mr Powell, failed - Powell  1 AC 1 at 30.
Foresight of what might possibly happen is ordinarily no more than evidence from which a jury can infer the presence of a requisite intention. Its adoption as a test for the presence of the mental element necessary to be guilty of murder, amounts to a seriously unprincipled departure from the basic rule that is now generally reflected in Australian criminal law that liability does not attach to criminal conduct of itself, unless that conduct is accompanied by a relevant criminal intention - See e.g. R v O'Connor (1980) 146 CLR 64 at 96-97 per Stephen J; He Kaw Teh v The Queen (1985) 157 CLR 523 at 530-531, 565.
On this ground, the many criticisms that have been voiced by scholars and others about the over-inclusive concept stated in McAuliffe, and confirmed in Gillard and like cases, should be accepted by this Court. The present law of extended common purpose liability is unjust, overbroad and anomalous. It should be re-expressed to narrow its ambit.
This re-expression should take the advantage of reducing the legal anomalies and asymmetries that were identified and acknowledged by the House of Lords in Powell. There are many such unprincipled disparities in the current law. I will mention only some of them.
If a principal offender were to kill the victim, foreseeing only the possibility (rather than the probability) that his or her actions would cause death or grievous bodily harm, that person would not be guilty of murder - cf Crabbe (1985) 156 CLR 464 at 469-470. Yet a secondary offender with a common purpose could, on the current law, be found guilty of murder of the same victim on the basis of extended common purpose liability if the jury were convinced that he or she had foreseen the possibility that one of the group of offenders might, with intent, cause grievous bodily harm and if, in the result, one of the group does indeed kill the victim with the intention to cause such grievous bodily harm - McAuliffe at 118.
On the face of things, the secondary offender's moral blameworthiness in such a case is significantly less than that of the principal offender. Yet (particularly in separate trials - cf Osland v The Queen (1998) 197 CLR 316 at 368-370 , ) it is quite possible, on current legal doctrine, that the secondary offender might be convicted of murder whilst the principal offender is acquitted, or convicted of a lesser offence.
There is a further anomaly and lack of symmetry upon which the applicants relied. In Giorgianni v The Queen (1985) 156 CLR 473, this Court expressed the mens rea required for other forms of complicity at common law, in the case of an accused charged as an aider, abetter, counsellor or procurer of the offence in question, in terms firmly anchored in a requirement of proof of:
intentional participation .... by lending assistance or encouragement .... The necessary intent is absent if the person alleged to be a secondary participant does not know or believe that what he is assisting or encouraging is something which goes to make up the facts which constitute the commission of the relevant criminal offence.
Adherence to the present requirements of liability for an extended common purpose is difficult, or impossible, to reconcile with this approach to criminal liability.
It is not necessary here to enter upon the controversies that have arisen as to the exact knowledge necessary to establish the accessory's guilt of the principal's crime. It is sufficient to identify the serious disparity that has arisen in the requisite subjective element for aiding, abetting, counselling or procuring an offence (murder in particular) and that required for the establishment of guilt as an accessory on the basis of extended common purpose.
The applicants ask why, in terms of justice and legal principle, there should be such a difference in the legal elements to be proved to establish, as in their case, guilt of the crime of murder. Why, especially (without the authority of statute), should the prosecution be able to rely on significantly divergent tests for convincing a jury that the same accused was guilty of the one offence? Why, in point of legal principle, should murder in consequence of acting in concert require proof by the prosecution of a specific intention on the part of the secondary offender when no specific intention at all was required for proof of murder in the course of carrying out a purpose held in common that did not include murder?
NEED TO RE-EXPRESS THE LAW
By these and like arguments, the applicants have demonstrated, in my view, the serious anomalies, disparities, inconsistencies and lack of symmetry that have been introduced into this area of secondary liability for acts done by others. There is no justification of legal principle for persisting with these defects when they are called to the attention of a final national court with the power and function to address them at the suit of those directly affected.
The justification presented by Simester and Sullivan - see above at - - is ultimately unpersuasive. The law may indeed dislike group anti-social activities, particularly where they result in death. But a rational and just legal system will dislike such activities equally, whether the conduct charged is prosecuted as an offence of acting in concert or of aiding and abetting others in carrying out the group activity. The law will not withdraw from one means only of establishing the offence (by reliance upon extended common purpose liability) the normal requirement of the modern criminal law that the prosecution prove a requisite intention on the part of the secondary offender.
To hold an accused liable for murder merely on the foresight of a possibility is fundamentally unjust. It may not be truly a fictitious or "constructive liability" - Powell  1 AC 1 at 13. But it countenances what is "undoubtedly a lesser form of mens rea" - ibid at 14. It is a form that is an exception to the normal requirements of criminal liability - Gillard at 18-19 -, 28-29 , 30 ; Powell at 11 per Lord Mustill. And it introduces a serious disharmony in the law, particularly as that law affects the liability of secondary offenders to conviction for murder upon this basis.
By providing a legal footing upon which a jury might find a secondary offender guilty upon proof of mere foresight of the possibility that the victim will suffer really serious harm as a result of the common purpose of the accused, the present doctrine expands the liability of secondary offenders, in the case where a murder is charged, so far that, realistically, there will ordinarily be very little, if any, room left for manslaughter.
It is true that in Gillard at 29-30 - this Court upheld the need to ameliorate this potential outcome of extended common purpose liability by reserving the availability of a conviction of manslaughter. However, the room left for a verdict of not guilty of murder but guilty of manslaughter is confined in such cases almost to disappearing point - Gillard at 26 , 29-30 . In Gillard (as in this case) that point lay somewhere between contemplation of the possibility that the principal offender would do really serious harm to the victim, and the rejection of that possibility by the contemplation that the harm that would possibly ensue would be trivial, despite the weapons taken to the homicide scene. Little wonder that, in the present case, the Court of Appeal concluded that - (2005) 159 A Crim R 1 at 24 :
Given the way in which the case was conducted and the evidence which was before the jury, the suggestion that the jury would have taken the trial judge to be referring to 'merely technical assault' is completely hypothetical.
I agree with that estimate. It demonstrates what is, in any case, self-evident. As the ambit of liability for murder is expanded by reference to a possibility that the principal offender might inflict really serious harm on the victim, the availability of an alternative verdict of guilty of manslaughter from a jury, acting rationally and honestly, is virtually nil.
This is not a theoretical problem. To the extent that the availability of a realistic verdict of manslaughter permits a court to reflect notions of culpability in accordance with estimates of moral responsibility for the crime assigned by the jury to individual offenders, it serves a most useful social function - cf Powell at 15 per Lord Steyn. See also Gillard at 25-26 -, 32 . Especially so because, in several Australian jurisdictions, conviction of murder requires the imposition of a sentence of life imprisonment. Conviction of manslaughter, on the other hand, affords a judge a very wide sentencing discretion. It permits a more sensitive reflection of established individual culpability. Generally, this is desirable in the law, but especially so in cases of homicide. Either of the reformulations of extended common purpose liability suggested by the applicants would permit directions to the jury that would be more harmonious with the judge's directions about the other two ways in which the prosecution propounded the applicants' guilt of murder. It would also be more protective of the jury's right to a real choice, consistent with the finding of guilt of homicide but short of murder.
The need for re-expression of the law on extended common purpose liability is also demonstrated by the undue complexity that has been introduced by the separate and disharmonious principles to be applied in respect of each of the three ways in which the prosecution sought to justify the applicants' guilt of murder at their trial. For two of these ways (acting in concert and aiding and abetting) intention on the part of the accused at least to cause really serious injury has to be proved. But for the third (extended common purpose liability) proof of intention as such is unnecessary. This distinction introduces a needless disparity and complexity that must be extremely confusing to juries, as well as difficult for trial judges who have the responsibility of explaining secondary criminal liability to a group of lay citizens performing jury service. What jurors must make of the disparity, and the nuances of difference between the distinct modes of possible reasoning to their conclusion, is best not thought about.
The unreasonable expectation placed upon Australian trial judges (affirmed by appellate courts) to explain the idiosyncrasies of differential notions of secondary liability to a jury is something that should concern this Court. Especially so in the case of major points of difference in the governing legal principles (such as the absence of reference to specific intention in the explanation of extended common purpose liability). In my view it behoves this Court to try harder to find a unifying principle for secondary criminal liability. After all, the object is to explain to a jury, on the basis of common facts, how they may reason to a single conclusion, namely guilty, or not guilty, of murder. The law should not be as unjust, obscure, disparate and asymmetrical as it is. Its present shape can only cause uncertainty for trial judges and confusion to juries. Where, as in these applications, a specific application was made to this Court to rationalise and unify the applicable law, we should not rebuff the request so peremptorily and uncritically. On the contrary, this is precisely the kind of case in which a court such as this fulfils its role as expositor of the general principles of the common law for this country.
The joint reasons at  suggest that the issues for the jury's verdicts need not be over-elaborate or over-complicated in trials of the present kind. I wish that I could agree. The experienced trial judge who presided at the trial of the present applicants that lasted forty-six sitting days, charged the jury over three days. In the hope, no doubt, of avoiding accidental error in his oral directions, he provided the jury with written instructions that reflected the substantial complexities of the elements of each of the three ways in which the prosecution put its arguments for verdicts of guilty of murder. In accordance with McAuliffe, the written direction asked the jury, relevantly, to decide whether:
The Accused you are considering foresaw as a possibility in the carrying out of the agreed understanding or arrangement that death or really serious injury would occur by a conscious, voluntary and deliberate act of one of them not done in self defence.
For issues which the majority say are relatively simple, a great deal of effort was consumed to explain the different principles to the jury, and correctly so. Much of counsel's addresses at the trial were devoted to the same points. The prosecution was entitled to present the cases against the three applicants in every way lawfully available to it. However, the ensuing disparities and inconsistencies in the applicable principles of secondary liability introduce undue complexity to the applicable legal rules upon which the jury are told they must act.
Such complexity is also inconsistent with the basic function of jury trial. In these proceedings this Court cannot solve all of the problems presented by the complexity. However, in my view, the Court should endeavour, when the opportunity is presented, to remove or reduce at least the most obvious inconsistencies of which the applicants complain. If ever there was a part of the law where consistency and symmetry should be at a premium, it is where murder is charged and where the trial judge has the duty of explaining to the jury, by reference to the facts, how they may reason to their verdict on that charge. These are powerful reasons for reducing the disharmony in the separate modes of reasoning which are occasioned by the present law on extended common purpose liability. The applicants specifically requested this Court to do so. On this occasion, there is no procedural or technical impediment to the Court's responding to the request.
JUSTIFICATION FOR JUDICIAL RE-EXPRESSION
I accept that reasons may be collected for washing our hands of responsibility for the present state of the law, as revealed by these applications and earlier cases, and leaving its repair, if at all, to the several Parliaments of the States and Territories concerned. Deciding when that course is appropriate and when judicial re-expression of the common law is proper often presents a difficult question on which judicial minds may differ - Brodie v Singleton Shire Council (2001) 206 CLR 512 at 591-597 -. Nowadays, the general responsibility for the reform and restatement of the criminal law lies with the legislature. Especially where there is any suggestion of the expression of a new or additional form of criminal liability, the courts, in recent times, have disclaimed a creative role - Knuller (Publishing, Printing and Promotions) Ltd v Director of Public Prosecutions  AC 435 at 466; Director of Public Prosecutions v Withers  AC 842 at 857-859, 863, 877; cf Lipohar v The Queen (1999) 200 CLR 485 at 563 .
In matters of criminal law, I have sometimes concluded that this Court should not re-express the law, despite the demonstration of deficiencies - as in Lipohar at 563-565 , - and Lavender at 110 -. However, in the present proceedings, I have reached a different conclusion. In summary, my reasons are as follows:
What is proposed is not the enlargement of criminal liability or the imposition of liability where it did not previously exist. On the contrary, it is the rationalisation and simplification of the present rules of secondary criminal liability so as to remove current disparities and disharmonies that result in the overreach of criminal liability;
The particular law challenged in these proceedings is not, as the prosecution suggested, a law of long standing. Effectively, it dates in Australia from the early 1980s. Its deficiencies have been noted ever since. They have been complained about and criticised. In these applications, a specific request has been made to the court having the power and responsibility to do something about the defects and to remove the potential for injustice that they occasion;
The law in question was originally expressed by judges. It can therefore be re-expressed by them. The advantage of this Court's doing so is that the injustice and the anomalies, asymmetries, complexities and unconceptual defects of the law can be repaired by this Court for all jurisdictions of Australia where the common law of criminal responsibility applies. This Court has often referred to the desirability of maintaining a general uniformity in the expression of the basic rules of criminal responsibility in Australia - R v Barlow (1997) 188 CLR 1 at 32; Darkan (2006) 80 ALJR 1250 at 1274 ; 228 ALR 334 at 363. Complete uniformity in this aspect of the criminal law is impossible, given the different approach taken in the Code jurisdictions of Australia. However, what this Court stated in Johns, McAuliffe and Gillard it can now restate. It can do so in the light of the criticisms addressed by the applicants to the present law as applied in their trial;
With all respect to the recent conclusion that the problem should be left to the legislature, as stated by the House of Lords in Powell, that conclusion does not govern the response of this Court to the applicants' submissions. We have a constitutional and judicial function in Australia, in deciding matters brought to this Court, to state the common law for every part of this nation - Lange at 562-567. A different conclusion by courts in other nations on like questions is no longer determinative of this Court's decisions;
Also, with respect to the conclusion on the point in issue in these proceedings, expressed in the recent report of the Law Commission of England and Wales (Inchoate Liability for Assisting and Encouraging Crime, Law Com No 300, (2006) Cm 6878 at 13 [2.3], 19 [2.24]-[2.25]), I am not convinced by its reasoning and I note that it has proved controversial in its own jurisdiction. More convincing on this point are the criticisms of the present state of the law on extended common purpose liability expressed by many distinguished scholars in the United Kingdom, not least the late Professor Sir John Smith in "Criminal Liability of Accessories: Law and Law Reform", (1997) 113 Law Quarterly Review 453 at 464, upon whose writings and criticisms the applicants heavily relied. The flaw in the reasoning of those who have defended the current law is that the legal fiction they endorse - e.g. Simester, "The Mental Element in Complicity", (2006) 122 Law Quarterly Review 578 at 600 - can operate with a serious lack of proportionality. Culpability is too easily trumped by the desire for criminalisation. That may not concern those who live in a world of theory. It should, however, concern judges of the common law whose orders often result in multiple cases of very prolonged imprisonment for wrongdoing where individual culpability does not warrant that course - even allowing for the collusion in a joint enterprise of some kind;
Suggestions for the parliamentary rationalisation and modernisation of the English law were made in R v Cunningham  AC 566 at 583. There, Lord Edmund-Davies expressed the hope that the legislature would undertake the necessary and urgent task of reform. That hope was repeated in 1997 in Powell by Lords Mustill at 12 and Steyn at 15. The latter pointed to the availability of a "precise and sensible solution, namely, that a killing should be classified as murder if there is an intention to kill or an intention to cause really serious bodily harm coupled with awareness of the risk of death". A similar solution was supported by a House of Lords Committee (United Kingdom, House of Lords, Report of the Select Committee on Murder and Life Imprisonment, HL Paper No 78-I, (1989) at 25 ). However, these solutions had not been implemented in England. The House of Lords' suggestion would have the merit of requiring a finding on the specific intention of the individual accused in each of the three bases upon which the prosecution presented its case against the applicants, suggesting that they were guilty of the offence of murder. The more recent attempt of the Law Commission to find a solution more palatable to the legislature can only be explained by the failure of the United Kingdom Parliament to adopt this earlier principled approach to reform. In the present age, waiting for a modern Parliament to grapple with issues of law reform of this kind is like waiting for the Greek Kalends. It will not happen. Eventually courts must accept this reality and shoulder their own responsibility for the state of the common law; and
Within Australia, piecemeal reforms by way of limited statutory enactments have been achieved - see, e.g., Criminal Code (Cth), s 11.2. However, for the common law jurisdictions of Australia, the applicable law remains unchanged. The result is unconceptual and unduly complex. It is also unjust. It casts the net of liability for murder too widely. It is insufficiently discriminating in respect of individual responsibility. It catches potentially weak and vulnerable secondary offenders and, by a legal rule, attributes to them liability that may properly belong only to the principal offender. Effectively at the option of prosecutors, it fixes people with very serious criminal liability because they were in the wrong place at the wrong time in the wrong company. It is prone to misuse by public authorities. It deflects prosecutors and juries from the difficult but ordinarily necessary task of assigning criminal liability appropriately by reference to proved moral culpability, particularly in the circumstances of homicide which attract the serious punishments properly imposed in respect of conviction for such offences.
CONCLUSION: RE-EXPRESSION REQUIRED
It follows from these considerations that I would accede to the applicants' submissions. This Court should re-express the principle of the common law concerning extended common purpose liability in Australia for criminal acts done by others. The present applications afford a proper opportunity to do so.
THE RE-EXPRESSION OF EXTENDED COMMON PURPOSE LIABILITY
ALTERNATIVE RESTATEMENTS OF THE RULE
Once the foregoing conclusion is reached, two possibilities exist for the re-expression of the common law, at least in the case of extended common purpose in homicide cases. The first, and simpler approach would be for this Court to replace the foresight required of the secondary offender under the present law by foresight of an outcome that was regarded as probable. Thus, the present test, addressed to what is viewed as possible, would be altered to address attention to what the secondary offender viewed to be probable. Such a change would be modest, confined but consistent with the earlier recognition of liability for a form of recklessness as explained by this Court in R v Crabbe (1985) 156 CLR 464 at 468-470; cf Lavender (2005) 222 CLR 67 at 107-108 .
The second possibility, which the applicants submitted was preferable, because it was more principled, would be to follow the suggestion of Professor Smith who said in "Criminal Liability of Accessories: Law and Law Reform", (1997) 113 Law Quarterly Review 453 at 465 (original emphasis):
It may be that the law is too harsh and, if so, it could be modified so as to require intention (or even purpose) on the part of the accessory that, in the event which has occurred, the principal should act as he did .... If it were to be decided that intention should be required, the jury would be told that they should not find D guilty of murder unless they were sure that D either wanted P to act as, and with the intention which, he did, or knew that it was not merely a 'real possibility' but virtually certain that he would do so.
THE PREFERABLE REFORMULATION
The substitution of "probability" for "possibility" would have the advantage of simplicity. It would adopt a more stringent criterion for liability than that contained in the existing formulation. It would provide a test that would leave an enlarged scope for the operation of manslaughter. It would produce a result similar in some ways to that adopted in the Criminal Code (Cth) (s 11.2) and somewhat similar to that adopted in the Criminal Codes of Queensland, Western Australia and Tasmania - Darkan ALJR at 1257-1260 -, 1273-1274 ; ALR at 341-345, 363. [Note that the Code provisions in Canada and New Zealand require a more stringent test. See Darkan (2006) 80 ALJR 1250 at 1259-1260 -, 1266-1267 , 1274 ; 228 ALR 334 at 343-344, 353, 363.]
However, the disadvantages of the first solution are obvious. The omission of any reference to the accused's "intention" would remain. Asymmetry between the different paths for permissible jury reasoning would persist. There would still be a serious lack of harmony between legal responsibility and moral culpability. The outcome would not be sound in principle. To that extent, it would defeat the essential purpose of re-expressing the governing law. The compass that will allow us to find our way again in this field of law is the now usual requirement of mens rea - cf Simester, "The Mental Element in Complicity", (2006) 122 Law Quarterly Review 578 at 582, 590.
It is for this reason that I prefer the formulation proposed by Professor Smith. In the place of telling the jury, relevantly, that they might convict a secondary offender for a crime actually committed by another in the course of a common enterprise if it was proved that that offender participated or continued to participate in the enterprise aware that it was possible that another participant might commit murder, the judge would explain the need for the jury to be sure that the secondary offender either wanted the principal offender to act as he or she did, with the intention which he or she had, or knew that it was virtually certain that the principal offender would do so.
This formula for extended common purpose liability would result in a proper correlation between legal responsibility and moral culpability in the crime of murder, proved by reference to the intention of the accused. It would restore symmetry with the existing principles of complicity at common law. In the words used in McAuliffe at 118, it would be "in accordance with the general principle of the criminal law that a person who intentionally assists in the commission of a crime or encourages its commission may be convicted as a party to it". It would be simpler for judges to explain to juries. It would be easier for jurors to understand. It would be more consonant with the requirement of the intention needed to establish each of the other modes of reasoning relied on by the prosecution (acting in concert and aiding and abetting) to sustain an accused's guilt of murder. And it would leave an enlarged and more appropriate scope, in the proper case, for the operation of the law of manslaughter. It would encourage juries to focus on the proved responsibility of each accused where the present law too readily sweeps them all up into a quasi-fictional responsibility for acts done by someone else, which acts the secondary offender may not have desired, anticipated, expected or intended.
THE CORRECT DISPOSITION OF THE PROCEEDINGS
ALTERNATIVE DISPOSITIVE ORDERS
Having reached the foregoing conclusion, it is my view that this Court should have granted special leave to the applicants and upheld their appeals. A difficult question then arising is what this Court should have ordered.
The applicants asked that their convictions be set aside and a new trial ordered. In favour of that course is the fact that there was nothing that the applicants could have done, at trial or in the Court of Appeal, to secure acceptance of the proposition which they urged on this Court. At both levels in the Supreme Court of Victoria, the judges were bound to conform with the holdings of this Court on extended common purpose liability as expressed in McAuliffe and Gillard. Furthermore, because the mode of reasoning by the jury to the verdict of guilty of murder returned in the case of each applicant is unexplained, it is possible that such verdicts rested on the jury's conclusion based on the prosecution case expressed in terms of extended common purpose liability alone. Because that mode of reasoning was the one most favourable to the prosecution, and relieved the jury of the necessity to evaluate the specific intentions of each applicant, it would not be difficult to infer that it might have been the course that the jury took in reaching their verdicts in the trial of the present applicants.
THE PROPER ORDERS
Nevertheless, since this Court's decision in Weiss ALJR at 453-456 -; ALR at 672-675 re-expressed to some extent the way in which courts of criminal appeal in Australia should consider the disposal of appeals where legal error has been shown, anchored in the statutory necessity to demonstrate a miscarriage of justice in the case, it is appropriate that the prosecution should have an opportunity, if it should so wish, to persuade the Court of Appeal of Victoria that this would be a case for the application of the "proviso" [cf Darkan ALJR at 1269 -, 1277-1278 -; ALR at 356-357, 368-369].
In a matter so important as the instruction provided to the jury on the proper way of reasoning to their verdicts, it would be consistent with Weiss for the Court of Appeal to conclude that the demonstrated defect in the judge's instruction, shown by the development and re-expression of the law as I would favour, amounted of itself to a miscarriage of justice, was a fundamental defect in the conduct of the trial and required a retrial. However, in the view that I take, that would be a conclusion that I would leave to the Court of Appeal to decide.
Although the majority of this Court was of the view announced at the conclusion of oral argument that the applicants' applications for special leave should be dismissed, I would grant each of the applications and allow each appeal. I would set aside so much of the orders of the Court of Appeal of the Supreme Court of Victoria as dismissed the applicants' appeals against their convictions of murder and return the proceedings to the Court of Appeal for disposition consistently with these reasons.
 This was the situation in the two cases considered in Powell  1 AC 1 and in the case presented in Gillard (2003) 219 CLR 1. See also R v Gamble  NI 268 described in Powell  1 AC 1 at 28-30.
 The test propounded by the late Professor Sir John Smith, "Criminal Liability of Accessories: Law and Law Reform", (1997) 113 Law Quarterly Review 453 at 465. See below at .
 This test bears some similarity to the test applied in the Australian Code States. See Darkan v The Queen (2006) 80 ALJR 1250 at 1257-1260 -, 1266-1267 , 1273-1274 -; 228 ALR 334 at 341-345, 353, 363. See below at .
 Their Lordships heard the appeals in February 1997. They pronounced their orders in July 1997. They published their reasons on 30 October 1997; cf Kitto, "Why Write Judgments?", (1992) 66 Australian Law Journal 787 at 792; NAIs v Minister for Immigration and Multicultural and Indigenous Affairs (2005) 80 ALJR 367 at 380 ; 223 ALR 171 at 184-185; Hadid v Redpath (2001) 35 MVR 152 at 163-164  per Heydon JA.
 The prosecution relied on Johns (1980) 143 CLR 108. In Johns, at 122, 130-131, this Court approved a statement as to accessorial liability expressed by Street CJ in the Court of Criminal Appeal of New South Wales in that case in terms of "an act contemplated as a possible incident of the originally planned particular venture" (emphasis added). Of course, forms of secondary liability long pre-dated Johns and have been traced to Bracton and the Statute of Westminster of 1275: see Simester, "The Mental Element in Complicity", (2006) 122 Law Quarterly Review 578 at 578.
 See e.g. Victoria, Law Reform Commissioner, Law of Murder, Report No 1, (1974); Law Reform Commission of Victoria, The Law of Homicide in Victoria: The Sentence for Murder, Report No 1, (1985); Law Reform Committee of South Australia, Suggested Amendments to the Law Regarding Attempted Suicide, Report No 14, (1970); Victorian Law Reform Commission, Defences to Homicide: Final Report, (2004).
 Queensland Law Reform Commission, Abolition of the Distinction Between Wilful Murder and Murder, Report No 2, (1970). This report resulted in the amendment of the Criminal Code (Q). The crime of "wilful murder" remains in the Criminal Code (WA), s 278; cf Ugle v The Queen (2002) 211 CLR 171 at 180-181 ; Law Reform Commission of Papua New Guinea, Report on Punishment for Wilful Murder, Report No 3, (1975), proposing incorporation of differential punishment for cases of wilful and other murders under s 309 of the Criminal Code (PNG).
 See e.g. Criminal Law Consolidation Act 1935 (SA), s 11; Criminal Code (Q), s 305(1); Criminal Code (WA), s 282(a); Criminal Code (NT), s 164. In New South Wales, Victoria, Tasmania and the Australian Capital Territory, whilst the maximum sentence for murder is life imprisonment, imposition of that sentence is discretionary. See Crimes Act 1900 (NSW), ss 19A(1), 19A(3); Crimes Act 1958 (Vic), s 3; Criminal Code (Tas), s 158; Crimes Act 1900 (ACT), s 12(2); Crimes (Sentencing) Act 2005 (ACT), s 32(1). See also discussion in Lee Vanit v The Queen (1997) 190 CLR 378.
 Criminal Code (Q), s 8. See also Criminal Code (WA), s 8; Criminal Code (Tas), s 4. But note Criminal Code (NT), s 8(1).
 Clarkson, "Complicity, Powell and Manslaughter", (1998) Criminal Law Review 556 at 557-558. See also Mueller, "The Mens Rea of Accomplice Liability", (1988) 61 Southern California Law Review 2169; Cato, "Foresight of Murder and Complicity in Unlawful Joint Enterprises Where Death Results", (1990) 2 Bond Law Review 182; Bronitt, "Defending Giorgianni - Part One: The Fault Required For Complicity", (1993) 17 Criminal Law Journal 242; Odgers, "Criminal Cases in the High Court of Australia", (1996) 20 Criminal Law Journal 43; Smith, "Criminal Liability of Accessories: Law and Law Reform", (1997) 113 Law Quarterly Review 453; Kadish, "Reckless Complicity", (1997) 87 Journal of Criminal Law and Criminology 369; Gray, "'I Didn't Know, I Wasn't There': Common Purpose and the Liability of Accessories to Crime", (1999) 23 Criminal Law Journal 201; "McAuliffe Revisited", (2004) 28 Criminal Law Journal 5; Bronitt and McSherry, Principles of Criminal Law, 2nd ed (2005) at 386-387.
 Somewhat analogous and disparate outcomes arose in the case of Mr Bentley. See Bentley (Deceased)  1 Cr App R 307.
 The tension between Giorgianni and McAuliffe has been noted. See Simester, "The Mental Element in Complicity", (2006) 122 Law Quarterly Review 578 at 596.
 (1985) 156 CLR 473 at 506 per Wilson, Deane and Dawson JJ. See also at 481-482, 487-488 per Gibbs CJ, 500, 504-505 per Wilson, Deane and Dawson JJ.
 See e.g. Davis v The Queen (1991) 66 ALJR 22 at 23-24 per McHugh J; 103 ALR 417 at 420-421; Edwards v The Queen (1992) 173 CLR 653 at 657-658; Stokes and Difford (1990) 51 A Crim R 25 at 37-39; R v Le Broc (2000) 2 VR 43 at 60-65 -.
 Referring to Criminal Law Revision Committee, Offences against the Person, Report No 14, (1980) Cmnd 7844 at 14  adopted in Law Commission, A Criminal Code for England and Wales, Law Com No 177, (1989), vol 1, cl 54(1) of the draft Criminal Code Bill.
McAuliffe v The Queen (1995) 183 CLR 108
Gillard v The Queen (2003) 219 CLR 1
Chan Wing-Siu v The Queen  AC 168
Hui Chi-ming v The Queen  1 AC 34; R v Powell  1 AC 1
R v Crabbe (1985) 156 CLR 464
Alford v Magee (1952) 85 CLR 437
Bowers v Hardwick 478 US 186
Lawrence v Texas 539 US 558 (2003)
R v Hartwick (2005) 159 A Crim R 1
Johns v The Queen (1980) 143 CLR 108
R v Powell  1 AC 1
Chamberlains v Lai  NZSC 70
Lange v Australian Broadcasting Corporation (1997) 189 CLR 520
R v Hartwick, Hartwick and Clayton  VSC 63
Domican v The Queen (1992) 173 CLR 555
Adelaide Company of Jehovah's Witnesses Inc v The Commonwealth (1943) 67 CLR 116
Weiss v The Queen (2005) 80 ALJR 444; ALR at 671-675
R v Tomkins  2 NZLR 253
R v Lavender (2005) 222 CLR 67
Charlie v The Queen (1999) 199 CLR 387
Wilson v The Queen (1992) 174 CLR 313
Darkan v The Queen (2006) 80 ALJR 1250; 228 ALR 334
R v O'Connor (1980) 146 CLR 64
He Kaw Teh v The Queen (1985) 157 CLR 523
Osland v The Queen (1998) 197 CLR 316
Giorgianni v The Queen (1985) 156 CLR 473
Brodie v Singleton Shire Council (2001) 206 CLR 512
Knuller (Publishing, Printing and Promotions) Ltd v Director of Public Prosecutions  AC 435
Director of Public Prosecutions v Withers  AC 842
Lipohar v The Queen (1999) 200 CLR 485
R v Barlow (1997) 188 CLR 1
R v Cunningham  AC 566
NAIs v Minister for Immigration and Multicultural and Indigenous Affairs (2005) 80 ALJR 367; 223 ALR 171
Hadid v Redpath (2001) 35 MVR 152
Ugle v The Queen (2002) 211 CLR 171
Lee Vanit v The Queen (1997) 190 CLR 378
Bentley (Deceased)  1 Cr App R 307
Edwards v The Queen (1992) 173 CLR 653
R v Le Broc (2000) 2 VR 43
Crimes Act 1958 (Vic): s. 568
Authors and other references
UK Law Commission, Inchoate Liability for Assisting and Encouraging Crime, Law Com No 300, (July 2006) Cm 6878
Prof Simester in "The Mental Element in Complicity", (2006) 122 Law Quarterly Review 578
A P Simester & G R Sullivan, Criminal Law Theory and Doctrine, 2nd ed (2003)
New South Wales Law Reform Commission, Partial Defences to Murder: Diminished Responsibility, Report No 82, (1997)
Criminal Law Officers Committee of the Standing Committee of Attorneys-General in Australia, Discussion Draft: Model Criminal Code, Chapter Two, General Principles of Criminal Responsibility, (1992)
Mueller, "The Mens Rea of Accomplice Liability", (1988) 61 Southern California Law Review 2169
Prof Sir John Smith, "Criminal Liability of Accessories: Law and Law Reform", (1997) 113 Law Quarterly Review 453
House of Lords Committee, Report of the Select Committee on Murder and Life Imprisonment, HL Paper No 78-I, (1989)
Kitto, "Why Write Judgments?", (1992) 66 Australian Law Journal 787
Law Reform Commissioner of Victoria, Law of Murder, Report No 1, (1974)
Law Reform Commission of Victoria, The Law of Homicide in Victoria: The Sentence for Murder, Report No 1, (1985)
Law Reform Committee of South Australia, Suggested Amendments to the Law Regarding Attempted Suicide, Report No 14, (1970)
Law Reform Commission of Victoria, Defences to Homicide: Final Report, (2004)
Law Reform Commission of Queensland, Abolition of the Distinction Between Wilful Murder and Murder, Report No 2, (1970)
Law Reform Commission of Papua New Guinea, Report on Punishment for Wilful Murder, Report No 3, (1975)
Clarkson, "Complicity, Powell and Manslaughter", (1998) Criminal Law Review 556
Cato, "Foresight of Murder and Complicity in Unlawful Joint Enterprises Where Death Results", (1990) 2 Bond Law Review 182
Bronitt, "Defending Giorgianni - Part One: The Fault Required For Complicity", (1993) 17 Criminal Law Journal 242
Odgers, "Criminal Cases in the High Court of Australia", (1996) 20 Criminal Law Journal 43
Kadish, "Reckless Complicity", (1997) 87 Journal of Criminal Law and Criminology 369
Gray, "'I Didn't Know, I Wasn't There': Common Purpose and the Liability of Accessories to Crime", (1999) 23 Criminal Law Journal 201
"McAuliffe Revisited", (2004) 28 Criminal Law Journal 5
Bronitt & McSherry, Principles of Criminal Law, 2nd ed (2005)
C B Boyce with J P Wheelahan for the applicant in Matter No M156/2006 (instructed by Patrick W Dwyer)
L C Carter with M J Gumbleton for the applicant in Matter No M157/2006 (instructed by McNamaras)
M J Croucher for the applicant in Matter No M158/2006 (instructed by Robert Stary and Associates)
P A Coghlan QC with C M Quin for the respondent in all matters (instructed by Director of Public Prosecutions (Victoria))
all rights reserved