‍ ‍The Implied Freedom of Political Communication and the Present Crisis

‍ ‍Henry Ergas, AO

‍ ‍Rule of Law Institute Inaugural Twilight Colloquium on Civic Order and Democracy

‍ ‍Sydney, 10 August 2026

‍ ‍Since October 7, 2023, Australia has experienced an outbreak of antisemitism of a scale, persistence and virulence without precedent in this country’s history, culminating, just a few months ago, in the massacre of Jews peacefully attending a Hanukkah celebration at Bondi Beach. That outbreak has strained this country’s long tradition of toleration to, and at times well beyond, the breaking point. Its intensity, and the tragedy it has produced, raise urgent questions about the rule of law and about how far a liberal democracy is required to permit its own freedoms to be used—and abused—by those intent on destroying them. Those questions lie at the heart of the issues I wish to examine this evening.

‍ ‍

My argument has six parts. I begin with speech, and with the harm that organised and repeated demonisation inflicts on the autonomy, dignity and equality of a targeted minority, as well as on public discourse itself. From there I move to assembly, examining both the damage caused by demonstrations that incite hatred and what our own history, and Britain’s, show timely regulation of those protests can achieve.

‍ ‍

That history bears directly on the present, when the demonstrations we have witnessed since October 2023 have been defended by reference to a longstanding “right to protest” that, on closer examination, proves to be a recent invention. But thanks to the expansive interpretation of the implied freedom of political communication, that supposed right has become a widely held presumption that threatens legislation aimed at curbing potentially dangerous forms of militancy. Having critically examined that interpretation, I conclude by locating the paper’s central arguments within an older liberal tradition—a tradition forged by the tragic experience of democracies that lost the will, or the permission, to defend themselves.

‍ ‍

I

‍ ‍

That speech can cause harm is obvious; it is the basis for the restrictions on the freedom of expression imposed by laws that target defamation, breach of copyright, fraud and misleading and deceptive conduct. It is equally apparent that speech can cause grave personal offence, which is addressed, however poorly, by section 18C of the Racial Discrimination Act.

‍ ‍

But the harms with which I am concerned are not associated with these familiar concerns. They arise when expression erodes the conditions in which citizens can participate freely, fully and safely in public decision-making, damaging the fabric of public life and undermining the values central to liberal democracy.

‍ ‍

Thus, a virulently hostile environment—one in which attending a synagogue, wearing a Jewish symbol or speaking publicly as a Jew carries a real risk of harassment—does not merely wound feelings; it contracts the range of options genuinely available to those targeted, as people withdraw from places they would otherwise frequent, refrain from speaking under their own names, or calculate the risk of intimidation before every public act.

‍ ‍

Formal liberty of expression and of participation remains; effective liberty shrinks. As a result, the members of a minority are deprived of their autonomy, that is, the practical capacity to form, revise and act upon their own choices — a capacity which, since at least Kant, has been considered foundational to the functioning and benefits of liberal democracy.

‍ ‍

Because the harm is social rather than merely psychological, it extends beyond autonomy to dignity and equality. As John Rawls, perhaps the most influential political philosopher of the postwar period, has stressed, a well-ordered society depends upon a background assurance that each citizen will be treated as a full member of the political community, entitled to the same basic security as everyone else. Public, durable and repeated denunciation can corrode that assurance, not through any single insult but by accumulation: prejudices that begin as lurid slogans become familiar assumptions, and familiar assumptions shape whom institutions trust, whose claims they credit and whose political participation they discount. The result is not simply offence; it is the progressive diminution of civic standing and hence of the ability to influence democratic decision-making.

‍ ‍

The same process can impoverish public discourse itself. If members of a targeted group speak less because they fear the consequences, or continue speaking only to find their words discounted in advance because of who voices them, the range of voices effectively heard contracts. Speech can therefore diminish speech; and when it does so systematically, it undermines the very rationale on which the protection of political communication depends.

‍ ‍

All of those dangers are magnified, and new ones added, when hostile expression ceases to be an isolated utterance and instead acquires organised, repetitive and coercively visible form in the streets—when demonstrations shouting the same slogans, hurling the same accusations and parading the same symbols return week after week, occupying public space and altering the environment in which others must live.

‍ ‍

The resulting repetition does not only multiply the disruption, it changes its social meaning. Slogans and symbols that might otherwise be episodic expressions become, through constant reiteration, entrenched features of the public environment itself, altering the atmosphere in which social and political life unfold.[1]

‍ ‍

The language deployed in those demonstrations accentuates the harm. Words such as “genocide”, “apartheid”, “baby killers” and, more recently, “Zio” function less as descriptions open to rebuttal than, in Reinhart Koselleck’s phrase, as Kampfbegriffe—battle-concepts whose force lies not in any pretence to precision but in the images and associations they invoke; their purpose is not to describe but to condemn, denigrate and exclude.

‍ ‍

“Zio”, offered as a mere contraction of “Zionist”, functions much as “Yid” once functioned as a contraction of “Yiddisher”. A term that needs no elaboration, it carries its hostility within itself, in a form perfectly legible both to those who hurl it around and those at whom it is directed. Yet its lexical ambiguity—and the ability to claim it is a mere contraction—allows its users to deny, when challenged, any pejorative or antisemitic intent.

‍ ‍

The same reliance on ambiguity—which is the hallmark of the expressions commonly referred to as dog-whistles—is palpably true of “globalise the intifada”. The fact that intifada has more than one lexical meaning in classical Arabic no more settles the meaning of that slogan when it is chanted by keffiyeh-sporting crowds than the fact that Sieg Heil literally means “Hail Victory” settles its meaning when chanted in unison by Nazis draped in Hitler-era iconography. These terms’ political meaning lies not in the dictionary but in use: what matters is the meaning conveyed in its actual social and historical setting. Repetition—of the language and of the emotion-charged context in which it is deployed—cements that meaning and gives the slogan its mobilising power and its implied menace.

‍ ‍

The meticulous diaries of language use under the Third Reich kept by the philologist Victor Klemperer show that the sustained public repetition of ordinary terms, slogans and coded expressions entrenched contempt without ever explicitly arguing for it. It was instead built into what he called the lingua Tertii Imperii, which normalised a rhetoric that transformed its Jewish targets from objects of ridicule into monstrous adversaries.

‍ ‍

Recent research on the behaviour of hateful rhetoric once it enters circulation—particularly online, and within a fierce, identity-charged conflict of the kind this country has experienced since October 2023—bears out Klemperer’s intuitions. The term “affective polarisation” describes the first stage: disagreement over ideas or interests gives way to active dislike and distrust of the opposing group. Demonisation, the central feature of the second stage, goes further, presenting the adversary not merely as mistaken or hostile but as the embodiment of absolute evil, often through images of vermin, disease or pollution. At that point the moral restraints that normally protect an opponent begin to weaken, because the opponent is increasingly imagined as having no rightful place within the moral community.

‍ ‍

But repeated demonstrations do not only normalise dangerous expression; they increase its visibility. By returning to the same streets, courting the same controversy and repeatedly disrupting ordinary civic life, demonstrations force their cause upon the attention of those who have chosen not to engage with it. Their frequency and scale magnify the movement’s apparent strength, dominate the news and social media, and make its hostility inescapable, altering the pictures in the public mind. The rising incidence of “cancellation”, including by disrupting, or threatening to disrupt, speakers the protesters object to, does much more than demonstrate a complete disdain for freedom of expression; it transfers the repeated hostility into a form of public elimination.

‍ ‍

The sociologist Lewis Coser, himself a refugee from Nazism, and the scholars who built on his work, described how those processes reinforce one another—and how they lead to the outcomes Coser had so closely observed.[2] First, a sustained and disproportionate focus on a target community draws a symbolic moral boundary between the transgressor and ordinary citizens. Next, activists deploy scripts that magnify the transgressor’s guilt and the immorality of its conduct. Finally, repeated participation in public displays of those scripts draws in the weakly committed, incites the already committed to escalate, and increasingly places the targeted group beyond the sympathy owed one’s own kind.

‍ ‍

Writing at a time of intensifying social conflict, Shakespeare supplied, centuries ago, the dramatic paradigm for what happens next. Antony calls Brutus “an honourable man” again and again over Caesar’s corpse, never instructing the crowd to riot, yet making the words perform the opposite of their lexical meaning. The effect is immediate: exactly as Antony intended, the crowd riots and shortly afterwards murders Cinna the Poet, an innocent man, essentially because he bears the wrong name.

‍ ‍

That is the point J. L. Austin later formalised: speech does not merely say things — it does things. In a long past world, great orators, such as Lincoln, Churchill, Roosevelt and Menzies, used it to forge unified nations out of fractured publics. Now, in an overcrowded information environment, where torrents of words degenerate into bite-sized pieces that circulate within algorithmic mobs and metastasise into mere emotive triggers, it stirs hatred and division far more readily than it informs. There may be no direct appeal to violence; but study after study shows that combining the glorification of violence with the demonisation of an identifiable minority supplies the mood music to which violent actors dance.

‍ ‍

II

‍ ‍

A state need not and should not await the descent into all-out civil conflict before recognising the risks those phenomena pose; indeed, to await the explosion is to ensure its causes will resist cure. Repeated demonstrations that entrench hatred against an identifiable minority, train participants to regard it as polluting the community’s moral universe, and normalise language and conduct that would be intolerable if directed at another group, present precisely the kind of danger against which a well-ordered polity must act if it takes seriously its duty to protect minorities, guarantee public safety and ensure the equal protections of the law.

‍ ‍

Our own history shows what timely legislation can achieve. The worst sectarian violence in Australia’s history before the murderous attack at Bondi occurred in Melbourne in 1846, when Catholic and Protestant mobs fired upon each other on the anniversary of the Battle of the Boyne. The colonial legislature promptly prohibited processions commemorating festivals, anniversaries or political events related to religious or political divisions between Her Majesty’s subjects. Banners and music calculated to provoke animosity were forbidden. The result was immediate. As John Hirst put it, “the poisonous cycle of demonstration and counter demonstration by which Northern Ireland keeps alive its troubled past was nipped in the bud”.

‍ ‍

Melbourne demonstrated the principle Leonard Hobhouse would later formulate in his influential Liberalism, published in 1911: “Social freedom … rests on restraint.” Or, in his still more pointed formulation, “the restraint of the aggressor is the freedom of the sufferer”. Restraint did not extinguish civic freedom; it preserved the conditions in which civic deliberation could endure.

‍ ‍

What happens when that restraint is either absent or inconsistently enforced was demonstrated in Interwar Europe, with tragic consequence. As escalating street violence fed upon political division and deepened it, Weimar Germany and France enacted sweeping repressive powers, but enforced them only fitfully. The Weimar Republic oscillated between impotence and overreaction: constitutional scruples paralysed it, while emergency powers hollowed out its legitimacy. Sweeping powers without the demonstrated willingness to put them to use secured neither order nor legitimacy—and much the same was true of France.

‍ ‍

Britain chose differently. It did not seek to suppress fascism as a belief, but to regulate the intimidatory and violence-prone forms through which it was practised—forms epitomised by Oswald Mosley’s strategy of staging parades in Jewish neighbourhoods and holding inflammatory rallies outside synagogues. The Public Order Act 1936 strengthened police powers to restrict or prohibit marches, banned political uniforms, and authorised the suppression of flags and banners likely to incite disorder. Introducing the Bill, with Labour leader Clement Attlee’s strong endorsement, the National Government’s Home Secretary, Sir John Simon, stated the distinction plainly: “The Bill … does not touch belief; it deals only with intimidation.” Without martyring the fascists, the Act deprived Mosley’s movement of the performative tactics central to its appeal and hastened its marginalisation.

‍ ‍

Watching those developments closely, Robert Menzies, then federal Attorney-General, asked his department whether equivalent legislation was needed here. He was advised that it was not: the States already possessed ample powers to prevent disorderly processions and were plainly willing to put them to use. A policy of “restrained firmness” achieved the same result as in Britain—the effective disappearance of street militancy by the mid-1930s.

‍ ‍

Discontent was not silenced but channelled: away from intimidation and harassment, and into the ballot and parliamentary rivalry—the ordinary, productive forms of political competition that representative government exists to sustain.

‍ ‍

III

‍ ‍

The relevance to the present moment of that history—which is so poorly known and understood in this country, including on the bench—scarcely needs to be laboured. What Mosley pioneered, others have revived: marches directed at Jewish neighbourhoods and communal institutions; rallies timed and routed to intimidate rather than persuade; and the predictable descent into disorder when authorities decline to act. Thus, in November last year, demonstrators attempting to storm Darling Harbour’s international maritime exposition left police officers badly injured. Weeks earlier in Melbourne, counter-protesters attacked a peaceful “March for Australia”, burning the national flag and severely injuring police. Meanwhile, university campuses have become no-go zones for many Jewish students, as protesters have occupied public spaces, disrupted lectures, broken into the offices of Jewish faculty members, and forced the cancellation of visits by Jewish scholars, all with little or no disciplinary action being taken. And slogans glorifying violence have been repeatedly chanted, ensuring that violence and the threat of violence become an integral part of the movement’s moral vocabulary and permeate the climate around it, even when they do not become an immediate part of its conduct.

‍ ‍

The challenge, then, is to deny intimidation and disorder the licence conferred by misplaced deference to an inadequately qualified right of assembly. But there is, today, a widespread belief in the existence of a settled, historically rooted “right to protest”, allowing any legislative restraint upon its exercise to be cast as an infringement on an ancient liberty.

‍ ‍

The notion of such a time-honoured “right to protest” lacks historical foundations. That there was, in English law, no positive constitutional, statutory or common-law right of public meeting was entirely clear by the turbulent 1880s. Liberty of this kind, the Queen’s Bench Divisional Court held, “always consists in doing something a man is not forbidden to do”. Assembly in public places was a residual freedom, not an affirmative right.

‍ ‍

Australian courts took the same view. In Melbourne Corporation v Barry (1922) 31 CLR 174, Higgins J held that a claim to assemble upon a highway “to the detriment of others having equal rights” was “irreconcilable with the right of free passage”, there being “no authority whatever in favour of it”. Fourteen years later, Lord Hewart restated the English position in Duncan v Jones [1936] 1 KB 218: English law “does not recognize any special right of public meeting for political or other purposes”.

‍ ‍

As late as 2016—the year before Brown v Tasmania (2017) 261 CLR 328 marked a significant departure from the traditional understanding—the Australian Law Reform Commission described freedom of assembly not as an independent right but, at most, as a derivative corollary of the implied freedom of political communication, subject to significant limitations. Reporting the effect of Wainohu v New South Wales (2011) 243 CLR 181, in which the Court held that any freedom of association implied by the Constitution would exist only as a corollary of the implied freedom of political communication and would be tested in the same way, the Commission adopted the assessment of Professors Williams and Hume that the result was to leave that freedom with only “limited constitutional vitality”.

‍ ‍

In short, what existed, for most of two centuries, was a discretionary tolerance—in other words, a privilege—extended or withdrawn by police and magistrates according to the circumstances of the times. To invoke a right to protest as though it named a longstanding inheritance now under assault is not to defend a tradition; it is to invent one. But the language of recent judicial decisions shows how deeply that invention has taken hold.

‍ ‍

Thus, assembly is described as “the right, jealously guarded, of the citizen to exercise freedom of speech and assembly”. Disruption is treated as inherent in an “entitlement to peaceful protest”. And in Commissioner of Police (NSW Police Force) v Lees [2025] NSWSC 858, which permitted last year’s Harbour Bridge march to proceed, the public interest in that form of expression was held to be “very high”. That language borrows the vocabulary—and the presumptions—of settled entitlement for a doctrine that, on examination, has never been anything of the kind.

‍ ‍

The pivotal word is “entitlement”. Simpson J introduced it in Commissioner of Police v Rintoul [2003] NSWSC 662; twenty-two years later, without close examination, it helped underwrite the decision allowing the Harbour Bridge march to proceed. Underpinning each of those decisions, and many others in between, is the judicially defined implied freedom of political communication — a freedom which can, and repeatedly has, been invoked to challenge and disallow legislative measures affecting public assembly.

‍ ‍

IV

‍ ‍

The threat of legislation protective of the civil peace being disallowed by the implied freedom is now an omnipresent feature of the legislative environment. In the space of thirty months it has been realised three times, and on the third occasion against legislation passed in the immediate aftermath of the antisemitic murder, by Islamic terrorists, of fifteen people at a Chanukah celebration at Bondi Beach.

‍ ‍

In Kvelde v State of New South Wales [2023] NSWSC 1560, decided in 2023, parts of the New South Wales anti-protest legislation of the previous year were held invalid on the grounds that they were inconsistent with the implied freedom. Equally, in October 2025, in Lees v State of New South Wales [2025] NSWSC 1209, Mitchelmore JA declared invalid a provision inserted seven months earlier by the Crimes Amendment (Places of Worship) Act 2025, which had restored to police the power to direct participants in an unauthorised protest occurring in or near a place of worship. Her Honour held the provision broader than necessary, since a rule confined to protest affecting those entering or leaving would have imposed a lesser burden while serving the same end.

‍ ‍

Then came Bondi. The Terrorism and Other Legislation Amendment Bill was introduced on 22 December 2025, passed both Houses within three days, and received assent on Christmas Eve. It permitted the Commissioner of Police, with the concurrence of the Police Minister, to declare an area within which public assemblies could not be authorised — for fourteen days at a time, to a maximum of ninety, and only in the fourteen days following an authorisation under the terrorism legislation, a power exercised once before in its history, on the afternoon of the Lindt Café siege. In April this year, in Jarrett v State of New South Wales [2026] NSWCA 62, the Court of Appeal declared the whole of that scheme invalid.

‍ ‍

It would, unfortunately, take far too long to present a forensic analysis of Jarrett. But its salient holdings are telling of where the doctrine can lead.

‍ ‍

The first involves what Parliament was trying to do. Representative government, the Court held, “entails acceptance of the potential for disharmony, incivility and disruption that is part and parcel of democratic intercourse”. The State’s contrary argument — that a community traumatised by terrorism may temporarily require a measure of civic calm — was dismissed as involving “a particular conception of civic life” at odds with the Constitution.

‍ ‍

But what makes the Court’s conception any less particular? There is nothing in historical experience, in democratic theory, or in ss 7, 24, 64 or 128 of the Australian Constitution which says that representative government requires the degree of disruption the Court constitutionalised.

‍ ‍

Having thus drawn the picture of representative government it claimed the Constitution embodies, the Court then held that the circumstances in which the legislation was adopted and intended to operate were irrelevant to its assessment. It accepted that the legislature “perceived that community tensions were running high”, and that assemblies in that condition might place the community at risk — but held that a terrorist attack “does not make it any more constitutionally permissible” to act on the perception.

‍ ‍

Yet democracies, if they are to survive, need the power to meet emergencies by law, provided so far as possible in advance and limited in time. Clinton Rossiter, writing in 1948 with the previous decade fresh in mind, concluded that such powers are tolerable, indeed essential, in a democracy so long as they are temporary and self-terminating: their only reason for existing is the crisis, their purpose is to end it, and their life ends when it ends. Measured against Rossiter’s widely accepted standard the New South Wales scheme was close to a model. It was made by a Parliament sitting in public rather than claimed as executive prerogative; it required a Minister’s concurrence; it came up for renewal every fortnight; and it expired after ninety days. None of that mattered.

‍ ‍

That it was rejected isn’t because our courts cannot accommodate an emergency. They can and have: the preventative detention and questioning powers enacted after 2001 have survived challenge more than once. Equally, other legal systems have built a vocabulary for the problem: Article 15 of the European Convention, to take but one example, allows derogation in time of public emergency subject to conditions, notification and review. Comparative material on precisely that point was put before the Jarrett Court by the United Nations Special Rapporteur on counter-terrorism and ultimately considered not significant.

‍ ‍

But the Court was not content to rest with dismissing the importance of context. Quelling political communication “in the interests of protecting another part of the community from a sense of unease or threat (not associated with any proximate physical threat)” is, it held, not a legitimate purpose.

‍ ‍

Violence, however, does not emerge out of the ether. As experience and analysis have shown time and again, it breeds wherever a small community is repeatedly demonised and placed outside the bounds of the moral restraints that normally prevent even sharp disagreement from descending into attempted destruction. That harm is, by its nature, neither proximate nor physical. Going by Jarrett, it is therefore not something an Australian parliament may legislate against at all — until there is blood, or its imminent likelihood, on the streets.

‍ ‍

Finally, because the case was decided on the objective rather than the drafting, the decision left no scope for the legislation to be redrafted. The vice, the Court held, “lies in the purpose for which the power has been conferred”, so that any exercise of the power “will necessarily be in pursuit of the illegitimate purpose”. So long as it pursues the impugned end — the protection of a minority at a time of crisis — the New South Wales parliament cannot narrow this law and hope it survives; it must stand by while the civic peace is poisoned.

‍ ‍

There is no sound justification for the risk this poses to parliament’s capacity to protect the civil peace. Legislation, like all rules, is never perfect, over-inclusive in some respects and under-inclusive in others. But preventing harassment, intimidation and the demonisation of a visible minority is not a retreat from representative government — it strengthens and safeguards it.

‍ ‍

The High Court has already accepted that form of reasoning in Coleman v Power (2004) 220 CLR 1, with McHugh J stating the principle in terms of general application:

‍ ‍

“Communications on political and governmental matters are part of the system of representative and responsible government, and they may be regulated in ways that enhance or protect the communication of those matters. Regulations that have that effect do not detract from the freedom. On the contrary, they enhance it.”

‍ ‍

If regulation directed at the corrupting effect of money on political communication can be understood as enhancing rather than burdening representative government, there is no principled reason why regulation directed at the corrupting effects of demonisation, harassment and intimidation should be understood differently.

‍ ‍

It is true that a few paragraphs from the passage that has just been quoted McHugh J held insults to be a legitimate part of the political discussion the Constitution protects, and that Gummow and Hayne JJ held that a law aimed at ensuring the civility of discourse was incompatible with the constitutionally prescribed system. The same point was pressed further in Monis v The Queen (2013) 249 CLR 92. Whether those contentions are persuasive is debatable. But even if they were, the civility of discourse and the organised, repeated and targeted demonisation of an identifiable minority are different things with different effects. A doctrine that cannot tell them apart is a doctrine that is flawed.

‍ ‍

The deeper difficulty lies in what Jarrett said the freedom is for. The implied freedom, the Court said, guards against the systemic risk, in a majoritarian system, that political communications unhelpful or inconvenient to a current majority might be unduly impeded. That is a real risk. But it assumes the danger runs in one direction only — the many against the few, acting through legislation — so that the remedy against the “tyranny of the majority” must be a restraint on Parliament’s power.

‍ ‍

However, the harm at issue here runs the other way. It comes not from a majority and not through a statute, but from a mobilised faction in the street, and it falls on a minority’s practical capacity to take part fully and equally in civic life. On the Court’s own account of why the freedom exists — to preserve the conditions in which unpopular views can be heard, and so to secure the free and true choice that ss 7 and 24 contemplate — organised demonisation is not a competing interest to be weighed against the freedom. It is an instance of the very harm the freedom exists to prevent, inflicted by private rather than legislative means: speech diminishing speech.

‍ ‍

The most famous footnote in American constitutional jurisprudence recognised two ways in which democratic self-correction can break down. In United States v Carolene Products Co, 304 US 144 (1938), Justice, later Chief Justice, Harlan Fiske Stone identified the first as legislation that limits the political processes ordinarily relied on to repeal bad laws — restrictions on voting, on the spread of information, on political organisation, and on peaceable assembly. In the very next paragraph he identified a second: prejudice against “discrete and insular minorities”, which “tends seriously to curtail the operation of those political processes ordinarily to be relied upon to protect minorities”. The first is a blockage the legislature creates; the second is one it does not create, and may be the only body able to remove.

‍ ‍

A doctrine that sees only the first, and forbids parliaments to address the second, leaves the minority exposed to all the dangers discussed above. John Locke observed that “that ill deserves the name of confinement which hedges us in only from bogs and precipices”. The restraints needed to preserve freedom are not restrictions on it but the means of its survival. Jarrett stands Locke’s admonition on its head: it is now the hedge, and not the precipice, from which the Constitution is said to protect us.

‍ ‍

V

‍ ‍

Those errors, tensions and inconsistencies reflect the far broader afflictions from which the implied freedom suffers. Some fifteen years after the High Court first read the freedom into the Constitution, Professor James Stellios—now a Justice of the Federal Court—described the doctrine as “unsettled”. He meant unsettled in the ordinary, curable sense: the Court was still “incrementally feeling its way forward with a fledgling constitutional doctrine”, and clarity would prevail as the doctrine matured.

‍ ‍

But Justice Stellios’ expectation has not just been disappointed; the divisions on the Bench have worsened and the confusion in which the doctrine is enmeshed has worsened with them. The court’s opinions are pervaded by stark differences of valuation; more recently even the doctrine itself has been doubted. Thus, having observed in LibertyWorks Inc v Commonwealth (2021) 274 CLR 1 that the freedom identified in Lange v Australian Broadcasting Corporation (1997) 189 CLR 520 “is not yet settled law” and might even “justify a reconsideration of the implication itself”, Steward J was blunter still in Ravbar v Commonwealth (2025) 99 ALJR 1000, describing the doctrine as “very wrong” and “just not sustainable”.

‍ ‍

Nor is the disorder confined to the implication itself. The structured proportionality test adopted in McCloy v New South Wales (2015) 257 CLR 178 and applied in Brown has since been demoted: in Babet v Commonwealth (2025) 99 ALJR 883 and again in Ravbar it is described as a tool of analysis whose invocation is by no means necessary in every case, and the Court of Appeal in Jarrett resolved the matter before it with only limited reference to it. A decade of litigation has produced not a settled method but a menu, with no clear principles guiding the choice between the options it offers — and none of the clarity that lower courts, legislatures and affected parties are legitimately entitled to demand.

‍ ‍

That is far from being an accident; it reflects, at least in part, an ambiguity that was present in the doctrine from the outset. The unanimous formula adopted in Lange confines the freedom to communication “necessary for the effective operation” of representative government. What it, and subsequent decisions, never clarified was where the emphasis in that phrase falls.

‍ ‍

If the emphasis in “necessary for the effective operation” is placed on the word “necessary” the formula establishes a floor: the minimum political communication without which representative government could no longer function, with a law being invalid—and then plainly invalid—if it takes the system of government below that threshold. But if the emphasis is instead placed on “effective operation”—which means that the system operates well rather than merely operating—then the formula points towards an ideal, against which almost any restriction may appear a burden, no matter how clearly representative government remains intact. The Court has never chosen between those readings; in fact, it has frequently applied both in the same judgment.

‍ ‍

The practical import of the difference is readily illustrated. Until recently, the states had public order laws that could and at times of acute social conflict did severely restrict public assembly. Now, no one could seriously deny that Australia was a representative democracy before the 1980s; but an advocate of the second reading of the Lange clause might argue (in my view wrongly) that the system of representative government operated less “effectively” before the transition to those laws’ relaxation.

‍ ‍

Such an ideal-regarding reading is, I believe, fatally flawed, for two reasons. First, an inquiry that focuses on whether legislation moves us towards or away from some notion of fully effective representative government inevitably requires a conception of exactly what such a fully effective form of representative government looks like. But “representative government” is not a term behind which some Platonic eidos awaits to be discovered or a Hegelian idea awaits to unfold. The entire history of political philosophy shows that there is no perfect concept of representative government, of which the various conceptions are more or less adequate instantiations, for any institution, judicial or legislative, to identify and proclaim. Like “democracy”, representative government is, at best, what W. B. Gallie called an “essentially contested concept”, whose underlying exemplar and proper application is perpetually disputed. As a result, the search for an ideal type that could serve as a benchmark is not merely difficult; it is doomed to fail.

‍ ‍

Secondly, even if such a disembodied ideal existed—and it does not—there would be no clear, much less uniquely correct, way of weighing movement towards it against the benefits of some other social objective. Adapting Justice Antonin Scalia’s memorable aphorism, asking whether a governmental interest outweighs a purported freedom is like asking whether a rock is heavier than a line is long.

‍ ‍

To say that is not to suggest that judgments involving incommensurable values are impossible. Rather, the point is that they call for what the Greeks termed phronesis: practical judgement. Aristotle was surely right to regard phronesis as the characteristic excellence of the statesman, rather than the philosopher, precisely because it lays no claim to a uniquely correct answer. Its exercise is therefore inherently political: informed, as Jeremy Waldron argues in his magisterial Law and Disagreement, by the changing circumstances of politics and perpetually revisable as those circumstances change.

‍ ‍

The case for leaving that judgment to legislatures does not rest on their being wiser. It rests on three things a court cannot supply. The first is that where the disagreement is genuine the question of who should settle it cannot be answered by identifying who is right, since there is, by the nature of the subject, no single “right” answer. The second is that a legislature produces a settlement openly proportioned to circumstances and openly revisable as they change. Third and last, while legislatures can undoubtedly err in their assessment of the competing values, as well as in determining the best means of giving them practical expression, they are directly accountable to the electorate, and hence can be disciplined by the electoral process.

‍ ‍

In contrast, when courts assume that task, they risk becoming entangled in disagreements that erode the clarity, stability and predictability central to the rule of law—and, ultimately, their own legitimacy.

‍ ‍

Lon Fuller, one of the leading American legal philosophers of the twentieth century and for many years Professor of General Jurisprudence at Harvard, analysed the underlying problem in his famous essay The Forms and Limits of Adjudication (1978). Some questions, he argued, are “polycentric”: there is no uniquely correct answer waiting to be discovered through adjudicative reasoning, because choosing among alternative social arrangements alters a web of interdependent interests. In discussing what can happen when adjudication nevertheless assumes such functions, Fuller warned against the “formless exercise of a kind of wisdom-directed personal power”: the adjudicator can become a “czar”, drift towards “charismatic leadership” and ultimately risk “playing god”.

‍ ‍

Applied to the implied freedom, the objection is therefore not merely that judges disagree about how competing values should be balanced. Once a constitutional minimum is transformed into an ideal of effective representative government, the Court risks assigning itself precisely such a polycentric task: choosing among rival forms of democratic order where no uniquely correct constitutional answer exists. Constitutional interpretation then shades into constitutional construction—the judicial selection of a preferred democratic settlement. That is the kind of transformation Fuller’s analysis warns against: adjudication ceases merely to determine rights and begins to assume the open-ended responsibilities of political choice, altering the role and damaging the standing of the legal system.

‍ ‍

None of this means that the implied freedom is irremediably flawed. No matter how improbable it might seem, a legislature could enact a law that negated the constitutionally prescribed system of representative government, and the doctrine can legitimately guard against that possibility. But that is the work of a constitutional minimum, not a constitutional ideal. The Constitution protects democratic self-government from destruction; it does not commission courts to perfect it by testing every measure against some version of political nirvana.

‍ ‍

The High Court should therefore return to a minimalist reading or at least adopt an approach that gives the legislative function the deference, or margin of appreciation, it is owed. As Joseph Schumpeter argued long ago, democracy, which is the essence of representative government, simply stipulates the basic rules needed to allow conflicting beliefs to compete for elected office, along with periodic voting procedures that allow voters to sanction elected governments’ performance. Accepting such a bare bones benchmark would leave legislatures free to perform the continuing work of calibrating competing ends—autonomy against expression, order against assembly.

‍ ‍

To argue that legislatures undertake that task perfectly would be ludicrous; but if the task of calibrating conflicting values properly belongs to ordinary politics, it is not because it has shining virtues. It is because, in a world sadly lacking in omniscient philosopher kings, ordinary politics is the only means a pluralistic society has of resolving questions to which there is no single right answer — and to do so without the grandiose pretence, which is the hallmark of the courts, that what has been determined is a constitutional truth, rather than a temporary accommodation. However inadequate, legislatures are the best we have — and their answers are better than no clear answers at all.

‍ ‍

VI

‍ ‍

It remains, in closing, to draw these several threads together. There is a frame available for doing so that is older, and more thoroughly tested, than the doctrine framed by the High Court that has been the subject of this paper. Writing in the long shadow cast by Nazi and Stalinist totalitarianism, the great political philosopher Judith Shklar distilled the moral temper those times demanded into a single phrase: the liberalism of fear. That phrase named a liberalism disabused by experience rather than derived from first principles: anti-utopian, attentive to political evil, suspicious of certainty, and alert above all to the gap between the world as moral programmes imagine it and the world as human beings actually inhabit it. Its overriding concern was not to realise a chimerical summum bonum, some perfected condition of political life, but to avoid a summum malum: the concrete, recurring catastrophes through which free political communities destroy themselves.

‍ ‍

Today, the liberalism of fear speaks with an authority that doctrine alone cannot supply, because it was forged in proximity to political evils whose underlying dynamic has unfolded before our eyes: the onward march of demonisation; the drawing of a stark moral boundary between a designated enemy and the rest of the moral community; the scripts that convert condemnation into outrage; and the repeated, ritual participation that trains a movement to regard an entire community as beyond its universe of sympathy.

‍ ‍

A liberalism that has genuinely confronted the outcomes to which history shows that path leads does not mistake indulgence for tolerance. It certainly recognises that governments may overreach, and that any measures must therefore be adopted with prudence, designed with care and accompanied by strong and credible safeguards. But it also knows that a democracy unwilling vigorously to defend itself against those working, whether patiently or violently, for its destruction has not remained liberal; it has simply forgotten what its own liberalism is for.

‍ ‍

These convictions were not abstractions then, and they are not abstractions now. As Leonard Hobhouse argued more than a century ago, and as the liberalism of fear so clearly understood, far from being the negation of liberalism, the organisation of restraint in the service of civility, tolerance and freedom, is one of liberalism’s oldest and greatest achievements. I can only hope that achievement—which is, in the end, what we cherish when we cherish the rule of law—will be preserved for all Australians, and for generations to come.

‍ ‍
Bibliography

Books, Articles and Reports

Aristotle. Nicomachean Ethics. Translated by Terence Irwin. 2nd ed. Indianapolis: Hackett, 1999.

Austin, J. L. How to Do Things with Words. Edited by J. O. Urmson and Marina Sbisà. 2nd ed. Cambridge, MA: Harvard University Press, 1975.

Australian Law Reform Commission. Traditional Rights and Freedoms — Encroachments by Commonwealth Laws. ALRC Report 129. Sydney: Australian Law Reform Commission, 2015. Tabled in Parliament 2016.

Beaver, David, and Jason Stanley. The Politics of Language. Princeton, NJ: Princeton University Press, 2023.

Cepollaro, Bianca. “Slurs as the Shortcut of Discrimination.” Rivista di estetica 64 (2017): 53–65. https://doi.org/10.4000/estetica.2063.

Coser, Lewis A. The Functions of Social Conflict. Glencoe, IL: Free Press, 1956.

———. “The Visibility of Evil.” Journal of Social Issues 25, no. 1 (1969): 101–9.

Ergas, Henry. “Anti-Zionism and Antisemitism: Identifying and Analysing Antisemitic Speech.” Background paper submitted to the Royal Commission on Antisemitism and Social Cohesion, May 2026. Draft; available on request

Fein, Helen. Accounting for Genocide: National Responses and Jewish Victimization during the Holocaust. New York: Free Press, 1979.

Fogal, Daniel, Daniel W. Harris, and Matt Moss, eds. New Work on Speech Acts. Oxford: Oxford University Press, 2018.

Fuller, Lon L. “The Forms and Limits of Adjudication.” Harvard Law Review 92, no. 2 (1978): 353–409.

Gallie, W. B. “Essentially Contested Concepts.” Proceedings of the Aristotelian Society 56 (1955–56): 167–98.

Hirst, John. Sense and Nonsense in Australian History. Melbourne: Black Inc. Agenda, 2005.

Hobhouse, L. T. Liberalism. London: Williams and Norgate, 1911.

Javed, Jeffrey A. Righteous Revolutionaries: Morality, Mobilization, and Violence in the Making of the Chinese State. Ann Arbor: University of Michigan Press, 2022.

Klemperer, Victor. I Shall Bear Witness: The Diaries of Victor Klemperer, 1933–1941. Translated by Martin Chalmers. London: Weidenfeld & Nicolson, 1998.

———. The Language of the Third Reich: LTI — Lingua Tertii Imperii: A Philologist’s Notebook. Translated by Martin Brady. London: Athlone Press, 2000.

Koselleck, Reinhart. Futures Past: On the Semantics of Historical Time. Translated by Keith Tribe. Cambridge, MA: MIT Press, 1985.

Lamont, Michèle, and Virág Molnár. “The Study of Boundaries in the Social Sciences.” Annual Review of Sociology 28, no. 1 (2002): 167–95.

Locke, John. Two Treatises of Government. Edited by Peter Laslett. Cambridge: Cambridge University Press, 1988.

McGowan, Mary Kate. Just Words: On Speech and Hidden Harm. Oxford: Oxford University Press, 2019.

Rawls, John. Political Liberalism. Expanded ed. New York: Columbia University Press, 2005.

———. A Theory of Justice. Rev. ed. Cambridge, MA: Belknap Press of Harvard University Press, 1999.

Rossiter, Clinton L. Constitutional Dictatorship: Crisis Government in the Modern Democracies. Princeton, NJ: Princeton University Press, 1948.

Saul, Jennifer Mather. Dogwhistles and Figleaves: How Manipulative Language Spreads Racism and Falsehood. Oxford: Oxford University Press, 2024.

Schumpeter, Joseph A. Capitalism, Socialism and Democracy. New York: Harper & Brothers, 1942.

Shklar, Judith N. “The Liberalism of Fear.” In Liberalism and the Moral Life, edited by Nancy L. Rosenblum, 21–38. Cambridge, MA: Harvard University Press, 1989.

Smith, David Livingstone. Making Monsters: The Uncanny Power of Dehumanization. Cambridge, MA: Harvard University Press, 2021.

Sosa, David, ed. Bad Words: Philosophical Perspectives on Slurs. Oxford: Oxford University Press, 2018.

Staub, Ervin. The Roots of Evil: The Origins of Genocide and Other Group Violence. Cambridge: Cambridge University Press, 1989.

Tilly, Charles. The Politics of Collective Violence. New York: Cambridge University Press, 2003.

Trachtenberg, Joshua. The Devil and the Jews: The Medieval Conception of the Jew and Its Relation to Modern Antisemitism. New Haven, CT: Yale University Press, 1943.

Waldron, Jeremy. Law and Disagreement. Oxford: Clarendon Press, 1999.

Williams, George, and David Hume. Human Rights under the Australian Constitution. 2nd ed. South Melbourne: Oxford University Press, 2013.

Wimmer, Andreas. Ethnic Boundary Making: Institutions, Power, Networks. New York: Oxford University Press, 2013.

Cases

Babet v Commonwealth (2025) 99 ALJR 883.

Brown v Tasmania (2017) 261 CLR 328.

Coleman v Power (2004) 220 CLR 1.

Commissioner of Police (NSW Police Force) v Lees [2025] NSWSC 858.

Commissioner of Police v Rintoul [2003] NSWSC 662.

Duncan v Jones [1936] 1 KB 218.

Jarrett v State of New South Wales [2026] NSWCA 62.

Kvelde v State of New South Wales [2023] NSWSC 1560.

Lange v Australian Broadcasting Corporation (1997) 189 CLR 520.

Lees v State of New South Wales [2025] NSWSC 1209.

LibertyWorks Inc v Commonwealth (2021) 274 CLR 1.

McCloy v New South Wales (2015) 257 CLR 178.

Melbourne Corporation v Barry (1922) 31 CLR 174.

Monis v The Queen (2013) 249 CLR 92.

Ravbar v Commonwealth (2025) 99 ALJR 1000

United States v Carolene Products Co, 304 US 144 (1938).

Wainohu v New South Wales (2011) 243 CLR 181.

Legislation and International Instruments

Commonwealth of Australia Constitution Act 1900 (Imp) 63 & 64 Vict, c 12.

Convention for the Protection of Human Rights and Fundamental Freedoms, opened for signature 4 November 1950, 213 UNTS 221 (entered into force 3 September 1953).

Crimes Amendment (Places of Worship) Act 2025 (NSW).

Public Order Act 1936 (UK) 1 Edw 8 & 1 Geo 6, c 6.

Racial Discrimination Act 1975 (Cth).

Terrorism and Other Legislation Amendment Act 2025 (NSW).

‍ ‍[1] On the general account of political speech as the imposition of a perspective rather than the assertion of propositions open to rebuttal, see David Beaver and Jason Stanley, The Politics of Language (Princeton, NJ: Princeton University Press, 2023). On the particular mechanism by which derogatory terms impose their content as presupposed rather than asserted, and so make rejection costly, see Bianca Cepollaro, “Slurs as the Shortcut of Discrimination,” Rivista di estetica 64 (2017): 53–65; David Sosa, ed., Bad Words: Philosophical Perspectives on Slurs (Oxford: Oxford University Press, 2018); and Geoffrey Nunberg, “The Social Life of Slurs,” in New Work on Speech Acts, ed. Daniel Fogal, Daniel W. Harris, and Matt Moss (Oxford: Oxford University Press, 2018). On the deniability that gives dog-whistles their utility, Jennifer Mather Saul, Dogwhistles and Figleaves: How Manipulative Language Spreads Racism and Falsehood (Oxford: Oxford University Press, 2024). On speech that harms by altering the norms of a situation rather than by persuading, Mary Kate McGowan, Just Words: On Speech and Hidden Harm (Oxford: Oxford University Press, 2019). I have applied these tools to the slogans discussed here at greater length in Henry Ergas, “Anti-Zionism and Antisemitism: Identifying and Analysing Antisemitic Speech,” a background paper submitted to the Royal Commission on Antisemitism and Social Cohesion in May 2026, a draft of which is available on request.

[2] Lewis A. Coser, The Functions of Social Conflict (Glencoe, IL: Free Press, 1956), and “The Visibility of Evil,” Journal of Social Issues 25, no. 1 (1969): 101–9. On the cumulative effect of such language once it enters circulation, building on the account of presupposition and perspective in Beaver and Stanley, The Politics of Language, cited above: the sequence described here follows Jeffrey A. Javed, Righteous Revolutionaries: Morality, Mobilization, and Violence in the Making of the Chinese State (Ann Arbor: University of Michigan Press, 2022), which traces the drawing of a moral boundary between a virtuous ingroup and a demonised outgroup, the mobilisation of that boundary into violence, and its reinforcement through the complicity of those who participate. On the making and policing of symbolic boundaries more generally, see Michèle Lamont and Virág Molnár, “The Study of Boundaries in the Social Sciences,” Annual Review of Sociology 28, no. 1 (2002): 167–95, and Andreas Wimmer, Ethnic Boundary Making: Institutions, Power, Networks (New York: Oxford University Press, 2013); on the mechanisms connecting boundary activation to collective violence, Charles Tilly, The Politics of Collective Violence (New York: Cambridge University Press, 2003). On the exclusion of a group from the universe of moral obligation as a precondition of its destruction, Helen Fein, Accounting for Genocide: National Responses and Jewish Victimization during the Holocaust (New York: Free Press, 1979); on the incremental steps by which perpetrators come to accept what would earlier have been unthinkable, Ervin Staub, The Roots of Evil: The Origins of Genocide and Other Group Violence (Cambridge: Cambridge University Press, 1989); and on the psychological mechanisms of dehumanisation, David Livingstone Smith, Making Monsters: The Uncanny Power of Dehumanization (Cambridge, MA: Harvard University Press, 2021). On the specific content of anti-Jewish demonisation and the durability of its medieval inheritance, Joshua Trachtenberg, The Devil and the Jews: The Medieval Conception of the Jew and Its Relation to Modern Antisemitism (New Haven, CT: Yale University Press, 1943).

‍ ‍