On Sunday 22 January 2017, a new and evocative term entered mainstream English usage. How that occurred was as follows. On Friday 20 January 2017, the 45th President of the US was inaugurated in a public ceremony on the West Front of the Capitol Building at the end of the National Mall in Washington, DC. The US National Park Service, which controls the National Mall, does not publish statistics on the sizes of the crowds that gather there. The Washington Metropolitan Area Transit Authority does publish statistics on the numbers of people who ride the Washington Metro so as to be able to get to the National Mall.

On Saturday 21 January 2017, the White House Press Secretary, Sean Spicer, held his first press conference. Mr Spicer criticised mainstream media reporting of the estimated number of people who had gathered in the National Mall for the President’s inauguration the previous day. Mr Spicer then said, “we know that 420,000 people used the DC Metro public transit yesterday, which actually compares to 317,000 that used it for President Obama’s last inaugural”. The Washington Metropolitan Area Transit Authority’s published statistics on the number of people who rode the Washington Metro on the morning of President Obama’s inauguration on 21 January 2013 was indeed 317,000. The Authority’s published statistics on the number of people who rode the Washington Metro on the morning of Friday 20 January 2017 was not 420,000, but 193,000.

That formed the background to an interview which aired live on NBC’s “Meet the Press” program on the morning of Sunday 22 January 2017. Moderator Chuck Todd interviewed publicist Kellyanne Conway, who by then held the new position of Counsellor to the President. Mr Todd asked Ms Conway why the President, when putting his press secretary in front of the podium for the first time, had chosen to cause him to utter a “provable falsehood”. Ms Conway’s answer, when it eventually came, was as follows: “Don’t be so overly dramatic about it, Chuck. … You’re saying it’s a falsehood. … Sean Spicer, our press secretary, gave alternative facts to that.” She went on to say: “I don’t think you can prove those numbers one way or the other. There’s no way to really quantify crowds.”

Those watching the interview on that Sunday morning had just witnessed the creation of a meme. The term “alternative facts” went viral and global. Many, especially in the mainstream media, derided Ms Conway’s use of the term. Some, particularly in the blogosphere, defended it.

Ms Conway subsequently explained what she meant. She said: “Two plus two is four. Three plus one is four. Partly cloudy, partly sunny. Glass half full, glass half empty. Those are alternative facts.” What Ms Conway seemed to be saying, in context, was that the ultimate question of whether the crowd in the National Mall at the time of the inauguration was big or small was an inherently evaluative question. Because the ultimate question was evaluative, she seemed to be saying, it was acceptable to marshal tendentious material to support one answer over another.

One of Ms Conway’s defenders was journalist Joel Pollak. The day after the “Meet the Press” interview, on Monday 23 January 2017, he published an article in Breitbart, in which he explained “alternative facts” to be a “harmless, and accurate, term in a legal setting, where each side of a dispute will lay out its own version of the facts for the court to decide”.⁴ Some short time later, on 2 February 2017, blogger David Allison published an article in American Thinker entitled “‘Alternative facts’: a common legal term”.⁵ The term “alternative facts”, Mr Allison wrote, “is known to most lawyers”. Mr Allison wrote that the category of “most lawyers” presumably included Ms Conway, who had a degree from George Washington University Law School and who could therefore be presumed to have known exactly what she was saying. Mr Allison went on to write that when the non-legally-trained Mr Todd had described an alternative fact as a “falsehood” in his interview with Ms Conway, Mr Todd was “not only wrong, but propagating an ignorance born out of lazy and shallow thinking”.

The online version of Mr Allison’s article provided a link to a Wikipedia entry entitled “Alternative facts (law)”.⁶ The “history” of that Wikipedia entry recorded that the entry was created on 25 January 2017. The entry defined “alternative facts” as “a term in law to describe inconsistent sets of facts put forth by the same party in a court given that there is plausible evidence to support both alternatives” and as a term “also used to describe competing facts for the two sides of the case”. The entry went on to refer to a number of English and American cases on pleadings in which, lo-and-behold, the term “alternative facts” has been used.

Despite being legally trained in Australia and the US, I confess to having been one of those lazy and shallow thinkers who had previously been ignorant of the term “alternative facts”. Through Wikipedia, I became better informed. But I have not come to shed newfound enlightenment by delivering a lecture on pleading.

What I want to do is to reflect, in a post-truth era, on how our legal system deals with the phenomenon of the assertion of alternative versions of a fact. My topic is legal epistemology from the perspective of a lawyer rather than a philosopher. What do we, as lawyers, mean by truth? How does our conception of truth relate to our conception of justice?

A fundamental similarity between the US and Australia is that both are inheritors of the common law system of justice. The central feature of the common law system is that there is committed to a distinct judicial branch of government the unique and essential role of conclusively determining disputed questions of law and of fact in the context of an adversarial trial. For present purposes, I do not draw any distinction between civil and criminal trials and I do not draw any distinction between the US legal system and the Australian legal system as each has developed over the past two centuries. To the extent that there is a relevant difference, it is that Australia has abandoned trial by jury in almost all civil cases.

As someone who has studied and practised law mainly in Australia, however, I naturally approach the topic from the perspective of an Australian. And as an Australian lawyer I cannot help but to be influenced in my thinking by Sir Owen Dixon, who was a dominant figure within the Australian legal hierarchy for a substantial part of the 20th century and whose intellectual influence is still felt within it. Much of what I have to say will be drawn from his insight into what it means for a tribunal of fact to make a finding on a disputed question of fact. To the extent that I add anything, it will be limited to some contemporary observations about probability and heuristics.

First, I should tell you who Sir Owen Dixon was. He was a Justice of the High Court of Australia from 1929 to 1952 and Chief Justice from 1952 to 1964. Between 1942 and 1944, he took time off the Bench to become Australian Ambassador to the US. In Washington, he became a friend of Justice Felix Frankfurter.⁷ In the library of the High Court in Canberra, where I now sit, there is a first edition of Judge Learned Hand’s book on the Bill of Rights published by Harvard University Press in 1958. Inside the front cover is a handwritten inscription which gives an indication of the relationship between Dixon and Frankfurter. It reads:

Dixon was the principal exponent and exemplar of a judicial method which he famously described at the time he was sworn in as Chief Justice as “strict and complete legalism”.⁸ In a later address at Yale Law School, he explained that judicial method in the words of Professor Maitland as cleaving to that strand of the common law tradition which emphasised “strict logic and high technique”.⁹ Legalism has often been associated with formalism, including by me,¹⁰ but to brand it as mere formalism would be wrong to the extent that the description would suggest that it was mechanical or Austinian. Dixonian legalism, as practiced by Dixon himself as distinct from some of his lesser imitators, was sophisticated, intellectually inquiring and subtly innovative. Dixon’s legalism was not markedly different from the judicial method of Frankfurter. It might fairly be described as an understated, scholarly and sceptical version of the “grand style” admired by Professor Llewellyn.¹¹

Not long before he retired, Dixon delivered what was in retrospect his capstone speech. He delivered it not to a group of lawyers, but to a gathering of surgeons. The speech was about different forms of decision-making under conditions of uncertainty. Speaking of the role of a judge in contrast to that of a surgeon or a general, he said this:¹²

Whether truth was an attribute which could properly be ascribed to a legal conclusion, was the question which Dixon deliberately left hanging.

Dixon chose to entitle that speech “Jesting Pilate”.¹³ He took that title from a line in Francis Bacon’s essay entitled “Of truth”.¹⁴ Bacon’s inspiration came from the biblical encounter between Jesus and Pontius Pilate as told in the Gospel of John. Bacon penned as the first line of his philosophical dissertation: “What is truth? said jesting Pilate; and would not stay for an answer.” Dixon quoted those words as the penultimate line of the last significant speech he was to give in his long career as a judge. Dixon’s last line was, “I have not forgotten that when Pilate said this he was about to leave the judgment hall”.¹⁵

The same weary scepticism is evident in a story, recounted by his biographer, about Dixon seated at a dinner party next to a woman who was enthusing about how splendid it must be to dispense justice. Dixon’s reply: “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.”¹⁶

What I find interesting about the story is that a man who has been a senior appellate judge for a very long time gets asked about justice and responds by talking about facts. He responds by expressing in the negative what Justice Brandeis expressed affirmatively in the quotation with which I opened. He says, in effect, that there can be no justice without knowledge of the facts.

Of the three impediments to an appellate court having knowledge of facts which Dixon identified, the first two might together be described in terms more general and more generous than those used by Dixon as “human imperfection”. The third — what Dixon described as “archaic laws of evidence” — is, I think, inextricably linked with the other two. I will try to explain why.

Historically, rules of procedure and rules of substantive law were very much more blurred than they are now. By the time of Chief Justice Dixon’s retirement in 1964, many rules that a century before would have been considered rules of evidence had been transmogrified into rules of substantive law. Conventional estoppel and the parol evidence rule are examples. Other rules like client legal privilege have since gone the same way.

That still leaves many truly procedural rules of evidence derived from the tradition of the common law that persist within our contemporary legal system. They include rules which regulate the form and method by which evidence is adduced in a court. They include rules which regulate when evidence adduced is admissible to prove or disprove a fact in issue.

The basic rule of admissibility, to use the language of its modern statutory restatement in Australia, is that evidence is admissible if, but only if, it is relevant and that evidence that is relevant is evidence that, if it were accepted, could rationally affect (directly or indirectly) the assessment of the probability of the existence of a fact in issue.¹⁷ From that basic rule of admissibility, there are exceptions. Except under strict conditions, by way of example, we exclude hearsay, we exclude opinion, and we exclude evidence of a person’s predisposition.

Were the question of the size of the crowd at the Presidential inauguration ever to arise for determination in a court, the question would be seen as a question of fact to be decided by reference to inferences drawn from evidence. Evidence of the number of people who rode the Washington Metro on the morning of the inauguration would be relevant. But evidence of that number, although relevant, would be excluded as hearsay or as opinion except under strict conditions.

Worth pondering is how exclusionary rules of that nature should have come to exist and persist. How is it that our system of justice should have come to accept that less relevant evidence should be preferable to more relevant evidence? Apart from evidence of such low probative value that its admission would be a waste of time or lead to undue expense, why is evidence that could rationally affect the assessment of the probability of the existence of a fact excluded from the assessment of the probability of the existence of that fact? Why have we not opted for what Jeremy Bentham suggested at the beginning of the 19th century was the “natural system” of adjudication: “Be the dispute what it may, — see everything that is to be seen: hear everybody who is likely to know anything about the matter”?¹⁸

Scholars who began to ponder those sorts of questions in the 19th century tended to see the answer as lying in the traditional distinction in common law courts between the role of the legally trained judge to orchestrate the proceeding and to state the law and the role of the randomly chosen jury of laymen to find the facts. The exclusionary rules had evolved, as they saw it, to limit the scope for the jury to be swayed by prejudice.¹⁹

The problem with that explanation is that not many exclusionary rules in their overt formulation differentiate between fact-finding undertaken by a jury and fact-finding undertaken by a judge. More recent historical research has tended to link the emergence of exclusionary rules of evidence less to the emergence of the jury as the finder of facts than to the somewhat later emergence of the adversarial system under which a trial of fact, whether in a civil proceeding or a criminal proceeding, came to be treated as a contest between parties.²⁰

The essential feature of fact-finding within an adversarial system is that the tribunal of fact — whether it be a jury or a judge — is tasked not with the independent pursuit of some ultimate truth but with arbitration of a contest between parties who assert different versions of the truth. Within an adversarial system, the party who asserts the existence of a fact which another party disputes ordinarily bears the burden of its proof. The question for the tribunal of fact is not the abstract question of whether the fact exists but the more concrete question of whether the tribunal is satisfied at the conclusion of the contest that the fact has been proved to the requisite standard. The requisite standard of proof in a civil proceeding is traditionally expressed as proof “by the preponderance of the evidence” or more commonly, at least in Australia, as proof “on the balance of probabilities”.

Expression of the standard of proof in a civil proceeding as satisfaction on the balance of probabilities is an acknowledgment that the judgment to be made by the tribunal of fact is inevitably to be made under conditions of uncertainty. Unless we were present, and perhaps even if we were, we can never have absolute certainty that an historical event occurred. Our memories are at best impressions of fragments of the past.

Satisfaction on the balance of probabilities, however, has not been equated with mere satisfaction as to the balance of probabilities. There are many examples in the decided cases, but an example from the academic literature perhaps illustrates that point best. There are many variations of the example. The earliest, I think, was given by Professor Tribe.²¹ The example is of a town in which there are just two bus companies. The Blue Bus Co has 95% of the buses. The Red Bus Co has the other 5%. The evidence shows only that the plaintiff was knocked down by a speeding bus in the dead of the night and nobody saw the colour. The question is whether the plaintiff recovers damages in negligence from the Blue Bus Co on the basis that there is a 95% chance that the bus was blue. The answer our legal system gives is “no”.

Theoretically, our legal system could have taken a different approach to fact-finding and a correspondingly different approach to the imposition of liability. We could have accepted an entirely probabilistic approach to finding facts in issue and we could have adjusted liability to reflect the probabilities. On the basis that the split reflects the probability of each company being responsible for the speeding bus, we could have held the Blue Bus Co liable for 95% of the damage sustained by the plaintiff knocked down by a bus and held the Red Bus Co liable for the other 5%.²²

The reason we do not impose liability in that probabilistic way, as Professor Nesson to my mind convincingly demonstrated, is deeply rooted in our conception of justice according to the rule of law.²³ The rule that a bus shall not be driven at more than 60 miles an hour is a rule that the driver of a bus is expected to obey, and either obeys or does not. Only its breach gives rise to civil or criminal liability. If liability was to be imposed in proportion to the probabilities without need to prove an actual breach of the rule, the Blue Bus Co could stay within the speed limit and still be liable for 95% of the damage sustained by the plaintiff being knocked down by a red bus, and the Red Bus Co could exceed the speed limit and still only be liable for 5% of the damage sustained by the plaintiff knocked down by one of its own buses. Neither bus company would have an incentive to obey the rule. The normative force of the rule itself would be destroyed.

To preserve the integrity of the rules we enforce, the approach we have therefore taken to fact-finding is all or nothing. We treat past events the occurrence of which is uncertain as either proved to have happened or not proved to have happened. We force a tribunal of fact to decide one way or the other, and we impose or decline to impose liability according to the outcome of that binary decision. Whatever its underlying probability, a disputed fact once found is a fact which is taken to exist for the purpose of resolving the legal rights or liabilities that are in contest. The fact once found is treated for that purpose as certain even though the fact might be found only on the balance of probabilities and even though the existence of the fact remains uncertain outside the scope of the controversy that is concluded by the judgment of the court. Sir Owen Dixon, to whom I have already referred, and to whom I will refer again, wrote of that phenomenon that: “[C]ourts have an advantage over other seekers after truth. For by their judgment they can reduce to legal certainty questions to which no other conclusive answer can be given.”²⁴

The process of fact-finding, however, is not entirely linear. The certainty attributed to a found fact loops back to affect in practice the way evidence is evaluated to find that fact.

Quite what is involved in the notion of satisfaction on the balance of probabilities was spelt out by Dixon J in 1938 in a case called Briginshaw v Briginshaw.²⁵ According to a recent survey, Briginshaw comes in at number seven of the 200 most frequently cited cases in Australia.²⁶ Like many frequently cited cases in many jurisdictions, it is one of the most persistently misunderstood. It is often treated as standing for the exact opposite of what it held.

What Dixon J was immediately concerned to do in Briginshaw was to reject the notion, since taken up in the Supreme Court of the US,²⁷ that some categories of facts in civil cases — of which fraud is the prime example — demand a higher standard of proof than proof on the balance of probabilities: a standard which might be expressed in contradistinction to proof “by the preponderance of the evidence” as proof “by clear and convincing evidence”. His explanation provided an analysis of the nature of fact-finding more generally. What he said was as follows:²⁸

The main thing Justice Dixon was saying, consistently with mainstream judicial²⁹ and academic³⁰ opinion in the US, is that satisfaction on the balance of probabilities involves the formation under conditions of acknowledged uncertainty of a subjective belief. The requisite belief is an “actual persuasion” that the fact in issue actually exists — that a past event the occurrence of which is uncertain and is disputed did indeed occur. What he was emphasising is that belief, as Bentham put it, “is susceptible of different degrees of strength, or intensity”.³¹ The belief involved in having a state of satisfaction “beyond reasonable doubt”, the universally accepted expression of the requisite standard of proof for a fact asserted by the prosecution in a criminal proceeding, is similar to the belief involved in having a state of satisfaction “on the balance of probabilities” in that it is subjective belief and different only in that it is belief that must be held with a greater degree of intensity.³² We have refused in Australia to define what we mean by “beyond reasonable doubt”,³³ but in England and NZ, where a judge is permitted to translate the meaning of “beyond reasonable doubt”, the standard instruction to a criminal jury is that it means “you must be sure”.³⁴ In the US, in some States, juries are instructed that proof beyond reasonable doubt is proof “that leaves you with an abiding conviction that the charge is true”.³⁵

Advances in biological and social sciences mean that recognition of the inherent subjectivity of belief in or satisfaction of a fact has the potential in a contemporary context to give rise to a number of avenues of inquiry. The suggestion has been made that what Dixon J referred to as “actual persuasion” of the existence of a fact in issue has a somatic component in that the feeling of persuasion can be related to activity in the region of the pre-frontal cortex associated with emotion and not simply with activity in the region that is more classically associated with reason, deliberation and judgment.³⁶ Expansion of the frontiers of neuroscience might well produce a more profound neurobiological understanding of the process of fact-finding.

In the meantime, our understanding of the cognitive processes involved in evaluating evidence and subjectively finding facts has been increasing through advances in behavioural science, particularly those building on the work on judgment under uncertainty pioneered by Daniel Kahneman and Amos Tversky in the 1970s.³⁷ The take-out point for present purposes is that humans, whether judges or jurors, are not particularly good at estimating probabilities. Tasked with forming a judgment as to the likelihood of the existence of a fact on the basis of incomplete information of uncertain veracity, most of us have an innate tendency to adopt rules of thumb or “heuristics” which work well enough in most situations, but which can lead in other situations to systematic and predictable errors or “biases”.

The subjective explanation of the process of fact-finding provides a bridge between those developing fields of biological and behavioural science and orthodox legal theory. When acknowledgment of the inherent uncertainty of the existence of a fact in issue is combined with acknowledgment of the inherent subjectivity of the process of finding that fact, scope emerges within the confines of mainstream legal analysis for conceiving of at least some exclusionary rules of evidence as measures serving a function of compensating for or mitigating difficulties faced by a tribunal of fact attempting to weigh some types of logically probative evidence. What emerges is the potential for conceiving of the existence and application of at least some of those rules as methods for correcting and improving the making of judgments of fact under conditions of uncertainty within the context of an adversarial system — for conceiving of rules which Dixon sardonically described as archaic rules impeding a court’s knowledge of the facts as measures which, to the contrary, might serve to mitigate unconscious biases so as to align the fact-finding tribunal’s perception of what is likely to have occurred more closely with a statistically objective assessment of what probably occurred.³⁸

But the subjectivity of fact-finding also allows us to understand why a statistically objective assessment of what probably occurred based on the evidence cannot be the measure of the veracity of curial fact-finding and should not be its goal. Recognising that the process of forming a state of mind cannot be divorced from the consequences that flow from such a state of mind being formed assists in understanding why we do not aspire to what Professor Tribe famously derided as “Trial by Mathematics”.³⁹ Improbable things, by definition, sometimes happen. Experience teaches that it is the happening or asserted happening of an improbable thing which in the majority of civil cases gives rise to the underlying dispute about liability. The risk of error inherent in finding an improbable thing to have happened on the balance of probabilities, like the risk of error inherent in finding a probable thing to have happened beyond reasonable doubt, has a human cost and a social cost.⁴⁰ The necessity for the tribunal of fact to feel actual persuasion of the existence of a fact in issue accommodates those realities by requiring the tribunal of fact, in effect, to factor in the cost of error.

There is a final aspect of Dixon J’s Briginshaw explanation that is of overriding importance. Frank acknowledgment of the inherent subjectivity of the fact-finding process highlights the ultimate dependence of our legal system’s discernment of the existence of a fact on the honesty and integrity of the person or persons who constitute the tribunal of fact. The notion of a judge or a jury needing to feel an actual persuasion of the occurrence or existence of a fact before that fact can be found is meaningful and workable only if the judge or each member of the jury brings to the fact-finding function a mind genuinely open to persuasion.

I referred earlier to Francis Bacon, who among his many other achievements held the office of Lord Chancellor in the early 17th century. I will conclude with reference to Matthew Hale, who among his many other achievements held the office of Lord Chief Justice around the middle of the same century, during a particularly turbulent period of English history in comparison with which Brexit might fairly be portrayed as a minor blip. In 1660, Hale produced a set of rules for judges. Two centuries later, the then Lord Chief Justice, Lord Campbell, would say that Hale’s rules “ought to be inscribed in letters of gold on the walls of Westminster Hall, as a lesson to those entrusted with the administration of justice”. In his popular yet profound little book entitled “The rule of law”, the Senior English Law Lord, Thomas Bingham, said as recently as 2010 that Hale’s rules “would still today be regarded as sound rules for the conduct of judicial office”.⁴¹

Of the 18 rules Hale laid down, four of them amount to different ways of saying the same thing. The first is “That in the execution of justice, I carefully lay aside my own passions, and not give way to them however provoked”. The second is “That I be wholly intent upon the business I am about, remitting all other cares and thoughts as unseasonable and interruptions”. The third is “That I suffer not myself to be prepossessed with any judgment at all, till the whole business and both parties be heard”. The fourth is “That I never engage myself in the beginning of any cause, but reserve myself unprejudiced till the whole be heard”. The four rules are important, because together they capture the personal intellectual and moral discipline of decision-making on which the integrity of the system depends.

We have an adversarial system of justice. Our system has never adopted the illusion that it is involved in an ultimate quest for truth. But nor has it descended to the relativism of countenancing the dispensation of justice as nothing more than the making of an unconstrained choice between so-called alternative facts. Our best answer to Pilate’s eternal question lies in our system’s deeply rooted commitment to the ideal of an honest and impartial tribunal, mindful of the gravity of the decision to be made, finding disputed facts by assessing evidence that has been filtered to correct for cognitive bias, so as to arrive at an actual state of satisfaction or lack of satisfaction as to the existence of those facts. Polarisation is the antithesis of impartiality; and impartiality is the hallmark of justice.