Tuesday, 10 May 2022

Post-Pell – Intermediate appellate courts in evidence evaluation in the digital age

Post-Pell - Intermediate appellate courts in evidence evaluation in the digital age

The Honourable Helen Murrell SC
former Chief Justice, ACT Supreme Court

Sir Richard Blackburn Lecture

Tuesday 10 May 2022

 

 

Introduction

I acknowledge the traditional custodians of this land and I pay my respects to their Elders, past, present and emerging. I acknowledge that sovereignty over this land has never been ceded.

The first Blackburn Lecture—titled “The Courts and the Community”—was delivered by Sir Richard himself. He spoke of the connection between the community and the administration of justice. As to the role of the jury, he concluded that:

the jury—a collaboration of twelve heads—is less likely to be wrong than a single judge in answering the question “is this person guilty beyond reasonable doubt?^[1]

Like Sir Richard, Lord Devlin considered that juries were superior fact finders. With a characteristically dramatic flourish, in Trial by Jury he described jury trials as “the lamp that shows that freedom lives”.^[2]

But is it correct that, when considering whether an accused person is guilty beyond reasonable doubt, twelve heads are less likely to be wrong than one head, regardless of whether the head is screwed onto the shoulders of an appellate judge?

If not, should an appellate court ever intervene on the ground that the unanimous factual decisions of the twelve were necessarily wrong?

The decision in Pell v The Queen^[3] may prompt us to reflect on this issue.

Today, I will explore the questions:

1. Are juries the better fact-finders? Do judges and juries decide facts in the same way and, if not, which is the better way?
2. Is an intermediate court of appeal well placed to decide whether the factual findings of the jury—or the trial judge where the trial has been by judge alone—have resulted in an “unreasonable” verdict?

Traditional arguments supporting juries as superior fact finders

In 1846, Vice-Chancellor Knight Bruce proclaimed that “[t]he discovery and vindication and establishment of truth are main purposes certainly of the existence of Courts of Justice.” But, he also acknowledged that “[t]ruth, like all other good things, may be loved unwisely—may be pursued too keenly—may cost too much.”^[4]

The law understands that rarely, if ever, can we be sure of the truth. Human decision-making is “inherently disputable”.^[5]

Consequently, in the adversarial trial process, we do not ask whether the relevant facts exist. We ask whether, having regard to the facts that we think exist, the case has been proved to the requisite standard.

At a dinner party, one of Australia’s greatest jurists, Sir Owen Dixon, was seated next to a guest who remarked that it must be splendid to dispense justice. Dixon replied:

I do not have anything to do with justice, madam. I sit on a court of appeal, where none of the facts are known. One third of the facts are excluded by normal frailty and memory; one third by the negligence of the profession; and the remaining third by the archaic laws of evidence.^[6]

Members of the profession may disagree with at least one of these categories, but his Honour’s point about human frailty is unassailable. We humans are not adept at fact-finding, particularly when, as is often the case, we are presented with apparently conflicting information.

There are many cases in which reliable evidence—such as DNA evidence—has later come to light and shown that the jury did wrongly convict.^[7]

Which brings us to the question of whether juries are better than judges at the difficult task of fact-finding.

Lord Devlin thought so. He proffered two explanations that are relevant for present purposes.

First, he opined that juries were better at appreciating human fallibility:

the hope of the defence very often lies in impalpabilities—the willingness to make allowances for muddle-headedness, illogicalities and unreasonableness—impalpabilities that are less appealing to the legal mind than to the lay.^[8]

Second, Lord Devlin considered that juries were better at assessing whether a witness was reliable, especially in the sense of whether the witness was lying:

Whether a person is telling the truth, when it has to be judged, as so often it has, simply from the demeanour of the witness and his manner of telling it, is a matter about which it is easy for a single mind to be fallible.^[9]

I do not accept Lord Devlin’s first point that juries have a better understanding of human fallibility and the propensity to behave irrationally than do coolly logical judges. As I observed in R v Ali (No 3) [2020] ACTSC 103, these days judges who practise in crime and common law gain a deep understanding of the range of unusual and irrational human behaviour and the impact on behaviour of social deprivation, intergenerational trauma, drugs and alcohol, mental illness and cognitive impairment. Among other contexts, almost every sentencing exercise requires judges to engage with such issues. Trial judges are better qualified than most lay decision makers to appreciate the range of normal and abnormal human behaviour.

Lord Devlin’s second point retains significant contemporary relevance. “Demeanour” is a blunt tool for the task of assessing witness reliability. Fortunately, these days, relatively few cases hinge on assessing witness credit by considering demeanour. But where such assessment is important, unlike many lay people, trial judges understand the dangers of relying on witness demeanour. The assessment of credibility is their daily chore and they have been educated about the assessment of credibility. In this respect, judges may have an advantage over juries.

But Lord Devlin was not making the point that, as individuals, jurors were better at assessing credibility by reference to demeanour. Rather, he was speaking of the capacity of a group to negate the bias of an individual—a critical point, to which I will return.

The framers of the Australian Constitution accepted that trial by jury was not universally appropriate. Andrew Inglis Clark, a great admirer of American democracy, composed the first draft of what was to become section 80 of the Constitution. Drawing heavily on Article 3 of Section 2 of the Constitution of the United States, Clark circulated a draft that provided that the trial of ‘all crimes cognisable by any Court’ would be by jury.^[10]

Clark’s draft was diluted in two substantial respects. The words “all crimes cognisable” were replaced by “trial on indictment”. Further, consistent with the federalist compact that was being negotiated, section 80 of the Constitution did not purport to cover trial by jury for offences against State or Territory laws. Today, most states and territories permit trial by judge alone in some circumstances.^[11]

Historically, the right to trial by jury was not a given.

And removed from its historical pedestal, the jury may seem a peculiar concept. As Baldwin and McConville put it:

Twelve individuals, often with no prior contact with the courts, are chosen at random to listen to evidence (sometimes of a highly technical nature) and to decide upon matters affecting the reputation and liberty of those charged with criminal offences. They are given no training for this task, they deliberate in secret, they return a verdict without giving reasons, and they are responsible to their own conscience but to no one else.^[12]

When I reflect on many years’ experience as a trial judge, I conclude that the jury verdicts returned in the trials in which I have sat were usually the verdicts that I would have returned had I been sitting as a judge alone. There are a few cases in which, unlike the jury, I would have returned a guilty verdict. In some of those cases, the jury acquittal may have been a “merciful verdict”, or a compassionate verdict, as explained in Gammage v The Queen.^[13]

A recent review by the ACT Supreme Court of the verdicts returned in judge alone versus jury trials suggested that judges were just as likely as juries to convict or acquit. Historically, this has not been the case; in July 2011 the range of matters in which an accused person could elect trial by judge alone was narrowed because it was considered that judges were acquitting too readily.^[14] Time does not permit the exploration of that interesting anomaly.

Although judges and juries tend to reach the same verdicts, this does not necessarily mean that they apply the same decision-making process.

Same decision, different process?

Vicki Waye looked at the method by which judges and jurors evaluated evidence and concluded that there was little to distinguish the general approaches of each.^[15] Both followed the “story model” of evidence evaluation.^[16] In other words, both tested the evidence by asking questions such as:

1. How does the evidence fit with the narrative (the “story”) that the party (most relevantly, the prosecution) is advancing?
2. Does the evidence provide internally consistent, plausible and complete support for that story?
3. How believable is the story?

Conversely, both judges and juries eschewed a Bayesian approach.

However, Waye opined that, as between judges and juries, different safeguards operated to reduce irrational decision making. In the case of juries, group decision making reduced cognitive illusion and moderated individual biases. In the case of judges, the need for written reasons directed the judge to articulate and “reality test” the inferences underlying conclusions of fact.^[17]

Waye’s point about the safeguard of group decision-making by juries is Lord Devlin’s second point about the superiority of the jury as fact-finder:

The impression that a witness makes depends upon reception as well as transmission and may be affected by the idiosyncrasies of the receiving mind; the impression made upon a mind of twelve is more reliable.^[18]

Ideally, a jury comprises of decision-makers who reflect the diversity of experience that reflects that within the community from which they are drawn. Consequently, the jury’s combined experience—or “collective wisdom”—is greater than that of any one or even two or three of its members. To apply an overused expression—the jurors bring their combined “lived experience” to deciding whether the prosecution case “stacks up” beyond reasonable doubt. If diversity and the sharing of experience is the core strength of the jury system, then we should constantly review our jury selection process to ensure that our juries do reflect the diversity within the community

The work of Kahneman and others, popularised through the 2011 publication of “Thinking Fast and Slow” has transformed our understanding of how decisions are made, and the fragility of decision-making.^[19] Incidentally, this research may also prompt us to question the efficacy of the standard “beyond reasonable doubt” (is it effectively unattainable?), or at least to rethink (and possibly even try to articulate) what we mean by “reasonable doubt”.

Put simply, the work explores the fallibility of human decision-making to both predictable and random error. Predictable error is strongly associated with “fast thinking”, which depends on mental shortcuts. Without recourse to fast thinking, the task of making the many decisions of daily life would be intolerable. But complex and important decisions demand “slow thinking”.

The safeguards of group decision-making by a diverse group and considered written reasoning by judges support slow thinking.

Justice Gageler has observed that a reliable judicial decision requires the judge to consciously put aside emotion and prejudice, keep an open mind throughout, and focus intently on the task at hand.^[20] By doing so, the judge minimises the influence of confirmation bias (looking for information to support a view that has already been formed), anchoring bias (overvaluing initial impression) and substitution (the tendency to supply an answer to a simple question rather than working out the answer to the relevant but complex question). Like other commentators, his Honour considered that the requirement to produce reasons was a critical safeguard that promoted slow thinking and the associated avoidance of predictable error.

Bennett and Broe have explained the neurobiology of decision-making in the civil context, but their insights are also helpful when considering decision-making in the criminal context.^[21] They examined the suggestion made by three of the judges in Briginshaw v Briginshaw^[22] that a decision-maker must “feel” satisfied or be “comfortable” that the standard of proof had been reached; in other words, a good decision evokes emotion and is not just a cold, rational conclusion. The authors observed:

Where decisions involve personal, social or moral factors, it is accepted that particular parts of the brain are recruited to participate in the decision-making process, which in turn have links to bodily sensations and feelings and emotions.^[23]

Most criminal trials do involve personal, social and moral factors.

According to Bennett and Broe, the prefrontal area of the frontal lobes is critical in decision-making. In early sorting of information, the orbitobasal region (which includes the ventromedial cortex) draws on the previous experience of the decision-maker. The ventromedial cortex assimilates information with emotional content into the reasoning process, enlivening visceral and other somatic responses.

It is at this level that the past experience of the judicial officer acquired both professionally through the law and from their private life in the world at large, and the relative salience of options and outcomes based on this experience, will be processed. Decisions based on this level of processing alone will appear to the judicial officer to have been “intuitive”.^[24]

The second prefrontal brain region, the dorsolateral cortex, then holds “working memory” for a short period, enabling complex information from various sources to be carefully and consciously compared and weighed.

As this stage of processing is thought to occur at a conscious level, it will also allow inappropriate emotional bias to be detected and excluded. It will be on the basis of all this information, both cognitive and somatic, that any ultimate decision will be made.^[25]

Referencing the Briginshaw test, of the second and fully conscious stage of processing, the authors stated:

If this critical, analytical, and deliberative process is able to achieve a “comfort”, or reconciliation, between the two opposing “decisional trends”, then the judicial officer will be in a position to have achieved a comfortable satisfaction that a fact has been proved.^[26]

The authors concluded that, the references by three of the Briginshaw judges to “feeling” satisfied was consistent with neurobiological science, in that:

Any ultimate decision may be a product of a reconciliation of varying and competing mental and somatic inclinations.^[27]

What this may mean is that immediate, ex tempore decisions rely more on intuition and are more prone to predictable error or bias. However, very slow decision-making may also be problematic—and not only because it impacts on the relevant court’s list of reserved judgments. Intuition and the wisdom derived from the decision-maker’s “lived experience” may be sacrificed. Appellate courts recognise that the advantage of the trial judge in assessing oral testimony decreases with delay in judgment delivery. Generally, they speak of the trial judge’s loss of recollection about the demeanour of witnesses. But the real loss may be the loss of intuitive response to the evidence.

Taken together, these findings suggest that:

1. Both judges and juries find facts applying the “story model” of evidence evaluation, rather than a quantitative method of analysis.
2. Like all decision-makers who address complex questions, when evaluating evidence and making decisions about the facts, judges and jurors may start with “fast thinking”, drawing on their personal and professional experience in a seemingly intuitive way.
3. At the second and fully conscious stage of decision-making, the decision maker engages in “slow thinking”, scrutinising the information that has been sorted and prioritised in the first stage. Bias can be consciously addressed. Nevertheless, the “intuitive” aspect of decision-making (which draws on the decision-maker’s professional and life experience) continues to play a role. The ultimate decision draws on both intuition and conscious reflection; the decision-maker will want to “feel” that the decision is the right one.
4. Jurors are likely to be significantly less proficient at slow thinking in the legal context, and judges are likely to be more proficient, by reason of their training and experience, and because of the discipline of articulating their reasons.
5. Although judges bring only one set of personal experience to the task of decision-making, they bring many years of professional experience (as either counsel or a judge who has heard evidence from and about people in all walks of life), and this experience is relevant at both stages of decision-making.
6. In the case of a jury, the diversity of the jury and the dynamics of group decision-making is likely to enrich the product of the first stage of decision-making and tease out biases in the second.
7. In most cases, judges and juries are likely to arrive at the same verdict. There is nothing inherently better about the fact-finding or decision-making of judges or juries, but the process does differ.

This begs the question: can we utilise the strengths of each approach?

Arguably, this is what occurs in tribunal decision making. For example, in relation to disciplinary proceedings brought against health care professionals, the tribunal may be comprised of a judge, two members of the relevant health care profession, and a community representative. Having sat on such tribunals, I can confirm that the input of the health care professionals is often most valuable at the beginning of the decision-making process (when their experience is drawn upon) and that of the judge most useful at the judgment writing stage. Almost invariably, the result of the “combined wisdom” of tribunal members is that each “feels” a very high level of comfort in what is almost invariably a joint decision.

The conclusion that neither judges nor juries are inherently better at fact-finding or decision-making does not resolve the issue of whether more judge-alone trials would dilute the quality of the justice delivered in serious criminal matters. There are other considerations that favour trial by jury.

First, the public may more readily accept jury verdicts. Second, the jury system nurtures public trust in the justice system because it embeds members of the public in its administration, consistent with democratic process.^[28]

Wayne Martin QC has put the case that “reasoned justice is preferable to inscrutable justice” and is more likely to inspire public confidence.^[29]

Speaking for myself, I doubt that the public places more faith in judge- alone decisions because of the judge’s experience as a professional decision-maker and because reasoned decisions are available. The public is rightly—but not unreservedly—confident in both processes.

What then is—and should be—the role of the appellate court when it is alleged that the tribunal of fact, be it judge or jury, has decided the facts wrongly?

The appellate court’s role and the competing temptation to intervene

In the 19th century in England and Australia there was no right to appeal on the ground that the jury had made an incorrect factual determination.^[30]

In 1912, New South Wales was the first state to adopt a general right of appeal, including an appeal against a verdict that was “unreasonable, or cannot be supported having regard to the evidence”.^[31] Relatively quickly, similar provisions were adopted in other jurisdictions.^[32]

It is well established that, on an “unreasonableness” appeal, the appellate court must not substitute trial by jury with trial by appellate court.^[33]

It is also well settled that, in considering whether a verdict is “unreasonable” the question is whether, on the whole of the evidence, it was open to the jury to be satisfied beyond reasonable doubt that the accused was guilty.^[34] Or, as propounded by Hayne J in Libke v The Queen, whether the jury “must,” as distinct from “might,” have entertained a doubt.^[35]

In M at [9] the majority set out the now well-established two stage test to be applied by an appellate court. McHugh J dissented. Unlike the High Court in Pell, he considered that the tests of “open” to convict and “must” have had a reasonable doubt differed; the “must” test was narrower, and it was the correct test. However, this view did not prevail, and was later abandoned by his Honour. Applying the “must have entertained a doubt” test, in M, McHugh J would have upheld the jury’s verdicts of guilty, although they depended on the uncorroborated evidence of a complainant that exhibited discrepancies.

It is interesting to consider whether the verdicts of guilty that the majority in M considered “unreasonable” would be regarded as “unreasonable” today. An understanding of what is “reasonable” is not immutable; it is culturally loaded.

In theory, in an “unreasonableness” appeal against a judge-alone verdict, the same rules apply as apply to an appeal against a jury verdict. But, as Heydon J observed in AK v Western Australia, because juries do not give reasons, “the jury has very great independence in relation to the facts.”^[36] In contrast, as Brereton JA said in Ford v The Queen:

in the context of a judge alone trial, the availability of reasons for a judge’s decision will inform consideration of whether [the guilty verdict] is unreasonable, because, unlike in a jury trial, the process of reasoning is exposed.^[37]

Consequently, in practice, an appellate court may be more inclined to overturn a verdict given by a judge alone.

I note that, while it may appear that the whole of the reasoning process is exposed in a judge’s reasons, science tells us that it is not. Just like a jury decision, the decision of a judicial fact-finder is necessarily informed by the judge’s “lived experience”, both personal and professional.

These days, for both jury and judge alone trials, the appellate court can access all the evidence given at first instance, often in the same form as it was given at the trial. Knowing that it has good access to all the evidence, subconsciously the appellate court may be sorely tempted to push the boundaries of appellate restraint and resolve the appeal on the court’s own assessment of the case.

The temptation may be even greater when the appellate court is not comprised of trial judges; such appellate courts may feel less constrained by respect for the trial judge’s findings, both because of a lack of collegial loyalty (not a bad thing) and because of a lack of experience as a trial judge and an associated lack of appreciation of how trial judges reach factual decisions and resulting verdicts (a fundamental difficulty).

Justice Gageler has argued that Condorcet’s Jury Theorem explains the inherent logic or “ecological rationality” of the typical appellate structure in which a judgment of a single judge is appealed to an intermediate appellate court of three or five judges, with a capacity to appeal to an ultimate appellate court of five, seven or nine judges; the further the matter proceeds up the appellate hierarchy, the more likely it is that the outcome will be correct, if only because the number of decision makers is greater.^[38]

Assuming the Theorem to be correct, it provides little reassurance where an appeal from the decision of a jury or a judge alone is allowed by a two/one majority in the intermediate appellate court; we then have the unedifying result that two decision-makers consider that the accused should be or was rightly convicted and two consider that the accused must be acquitted—highlighting that, when it comes to decisions based on witness reliability, there is no verifiably correct outcome.

Is an intermediate court of appeal well placed to review the factual findings of juries or trial judges sitting alone?

In the article “Where does the truth lie?”, Michael Kirby proposed that, at least in the civil context, the decision in Fox v Percy^[39] was a watershed in the reasoning approach of trial and intermediate appellate courts, introducing a shift from extreme deference to the supposed “advantages” of the trial judge in assessing witness credibility to much greater reliance on other means of fact-finding. He stated:

Now deference to the judicial impression of witnesses is a last consideration, after exhausting any relevant contemporaneous evidence and analysis of the inherent logic of the proved facts. Technology is coming to the aid of the law and the courts. The endless stream of emails and text messages, all phone location records and other objective testimony makes it much less usual for judges and decision-makers now to rest their conclusions on the fragile foundation of human assessment of truthfulness, based on witness appearances.^[40]

Similar observations could be made about criminal trials. Persuasive contemporaneous records are often available and may corroborate or cast doubt on a witness’ “story” about what happened. In such cases, an assessment of witness credibility does not depend on an assessment of the witness’ demeanour.

The availability of evidence such as contemporaneous records or DNA evidence diminishes the importance of human recollection and means that an appellate court may be well placed to decide whether the fact-finder must, as opposed to might, have had a reasonable doubt about the prosecution “story”.

However, there are still many criminal prosecutions that are essentially “word on word” cases. Often such cases are emotionally heavy as they involve allegations of sexual assault.

Pell was such a case.

Pell v The Queen

At one level, the High Court’s unanimous decision in Pell changed nothing. There was no dispute that, in deciding whether the verdict was unreasonable, the applicable law was that set out in M: whether, on the whole of the evidence, it was “open” to the jury to be satisfied of guilt beyond reasonable doubt. Nor was it controversial that this question should be resolved in two steps. First, the intermediate appellate court was to make its own independent assessment of the record to determine whether it had a reasonable doubt about the guilt of the accused. If it did, the court was to consider whether “the jury’s advantage” could explain the doubt.

In MFA v The Queen, the majority had spoken of the need to accord “special respect and legitimacy to jury verdicts deciding contested factual questions” because of the jury’s “advantage”.^[41] As to what “the jury’s advantage” meant, the Court in Pell said:

generally speaking, the appeal court should not seek to duplicate the function of the jury in its assessment of the credibility of the witnesses where that assessment is dependent upon the evaluation of the witnesses in the witness box. The jury performs its function on the basis that its decisions are made unanimously, and after sharing the benefit of the jurors’ subjective assessments of the witnesses. Judges of courts of criminal appeal do not perform the same function in the same way as the jury, or with the same advantages that the jury brings to the discharge of that function.^[42]

The Court emphasised that the jury’s “advantage in seeing and hearing the witnesses” was not just the advantage of first-hand observation (which may now be available to an appellate court through digital recording) but the “functional or ‘constitutional’ demarcation between the province of the jury and the province of the appellate court”, i.e. that jury decisions are unanimous decisions that reflect the jurors’ shared subjective assessment of witnesses.^[43]

Given the deference that must be accorded to the collective decision-making process—a deference that science would endorse—on any appeal it is a “big call” to overturn a jury’s finding that the account of the critical prosecution witness or witnesses should be accepted beyond reasonable doubt.

Nevertheless, the law requires that the appellate court examine the record:

…to see whether, notwithstanding that assessment – either by reason of inconsistencies, discrepancies, or other inadequacy; or in light of other evidence – the court is satisfied that the jury, acting rationally, ought nonetheless to have entertained a reasonable doubt as to proof of guilt.^[44]

In Pell, the Court assumed that the jury had found the complainant to be a credible and reliable witness.^[45] The central issue was whether, given the evidence of credible witnesses concerning “usual” or even “invariable” practices that obviated any opportunity for the alleged offences to have occurred, the prosecution had excluded the reasonable possibility that the accused did not commit the offences. Determination of this issue was critically affected by the considerations that important “lack of opportunity” evidence was unchallenged, and there was to be allowance for the forensic disadvantage associated with historical allegations.^[46] Undoubtedly, it was a case in which a jury, acting reasonably, might have acquitted.

The Court was critical of the fact that the Court of Appeal had viewed the video-recorded evidence of some witnesses, stating that such a course should be “exceptional”.^[47] The High Court considered that the majority’s erroneous engagement with the complainant’s credibility and their conclusion that he was a "compelling” witness had driven the majority decision.^[48]

The dissenting judge in the Court of Appeal, Weinberg J, had accepted a probabilistic analysis; he agreed that the “compounding improbabilities” presented by the evidence of usual practice meant that the prosecution could not exclude a reasonable doubt as to the accused’s guilt.^[49]

Adopting a similar approach, the Court found that, given three Church practices, each of which affected opportunity, as a whole the evidence was not capable of excluding a reasonable doubt about guilt.^[50]

Significantly, the High Court (and Weinberg J in the Court of Appeal) assumed that:

Evidence of a person’s habit or practice of acting in a particular way to establish that the person acted in that way on a specific occasion may have considerable probative value.^[51]

This proposition has been disputed by psychologists. Goodman-Delahunty, Martschuk and Nolan have observed:

Memory for repeated events is especially vulnerable in comparison with single event memory because experienced details become linked to the memory script rather than to the episodic memory traces for the instance in which they were experienced.^[52]

The Court’s decision about “unreasonableness” was not founded on inconsistencies or discrepancies within the complainant’s evidence, which was accepted as credible. Rather, it was based on a body of “other evidence” that, in the Court’s opinion, necessarily created a doubt.

Is it logically possible for direct evidence from a complainant to be both accepted and doubted?

Other broader issues raised by Pell follow.

These days, what is the “jury’s advantage”? Or that of the trial judge sitting alone?

If it is just the advantage of seeing the demeanour of witnesses, then in many—probably most—cases, that is no longer much of an advantage. Often, witnesses can be assessed by reference to external, relatively objective evidence. And in cases where demeanour is important, why not let the appellate court view the video recorded evidence? What is to lose? After all, in many cases the jury or trial judge will have seen the critical (pre-recorded) evidence on the screen, rather than “live” in the courtroom.

But maybe there is more to it. The fact-finder—be it judge or jury—who is immersed in the evidence as it unfolds sequentially will “intuitively” respond to the evidence, based on their “lived experience”. In the case of a jury, the experiences of the twelve will be brought to bear. In the case of a trial judge sitting alone, the wisdom gained from their personal experiences is supplemented by that derived from their professional experiences. Unless it views all the evidence sequentially, no appellate court will fully engage its life knowledge and common sense.

Are we content to retain the “story model” of decision-making?

If we choose to decide by reference to “compounding improbabilities” or, for that matter, “compounding probabilities”, then as judges we will need an education in statistics, and we will need to convey our learning to juries.

More fundamentally, as Justice Gageler has explained,^[53] courts do not search for the truth (or what is probably true) but consider what facts have been proved.

Is the two-step process of evaluation by an intermediate appellate court scientifically valid and helpful in practice?

If the only advantage of a jury is the observation of demeanour, then there are very few cases in which the jury’s view of the facts will prevail over that of an appellate court that harbours a doubt.

But if the real advantage is that the jury harnesses the life experiences of diverse community representatives who, having been immersed in the unfolding evidence, make unanimous group decisions, the position is otherwise. Appellate courts cannot replicate this advantage as they rely primarily on conscious analysis. It is not necessarily correct that a doubt experienced by the appellate court will be a doubt that the jury should have experienced.

Similarly, although arguably to a lesser extent, trial judges sitting alone enjoy this “advantage”. Most trial judges sitting alone “feel” how the trial is going, a “feeling” that ebbs and flows as the trial progresses, and is subject to rigorous, conscious review at the judgment-writing stage.

It may be time to reconsider our understanding of what constitutes an “unreasonable” verdict, appreciating that a decision about “unreasonableness” is not a decision that reflects an immutable truth but one that incorporates the lived experience or experiences of the decision-maker.

The Honourable Helen Murrell SC

I would like to thank my former associates, Ross Mackey and Bridie Adams, for their extensive assistance in the preparation of this address.