Corps de l’article

Introduction

Aminor league hockey spectator shouts racist insults to an Asian child on the opposing team and encourages others to chant the insult in unison; a court acquits him of the crime of causing a disturbance because “obscene language and racially insulting comments in a public place does not by itself attract penal sanctions.”[1] A municipal policy prohibits a bus advertisement by an anti-Muslim organization displaying images of women who were “honour killed by their families”; a court holds that the policy justifiably limits freedom of expression under subsection 2(b) of the Canadian Charter of Rights and Freedoms (Charter) to promote a safe and welcoming public transit system, since the ad could be perceived as “offensive, discriminatory and demeaning.”[2]

Cases such as these prompt a series of questions, which we explore in this paper. When is expression discriminatory? What reasons do we have to legally prohibit discriminatory expression? Does the offensiveness of such expression alone give such a reason or does the only reason derive from the need to prevent a specific kind of harm it may cause? How should reasons for prohibition figure into our deliberations when we weigh them against reasons to protect freedom of expression? Our inquiry into these questions takes inspiration from how myriad areas of law deal with discriminatory expression, although we focus on prohibitions of discriminatory expression under antidiscrimination statutes or human rights codes. These statutes prohibit conduct that (directly or indirectly) disadvantages a victim within a specified sphere of private life (such as employment or housing) because of a specified trait of a victim (such as her race, sex, or religion).

We develop theoretical perspectives on the above questions by means of a case study of Ward v. Quebec (Commission des droits de la personne et des droits de la jeunesse). In that case, the Supreme Court of Canada considered when discriminatory expression should be prohibited by Quebec’s Charter of Human Rights and Freedoms (Quebec Charter).[3] A disabled boy who achieved public notoriety as a singer in Quebec argued that he was discriminated against after a comedian made disparaging comments about his appearance in live shows and online videos. According to the Court, the jokes were not discriminatory because of the legally prohibited ground of disability. Even if they were, they would not be prohibited by the Quebec Charter; they did not cause the specific kind of harm whose prevention gives a reason to prohibit them that could justify limiting freedom of expression, namely the distinctive harms of hate speech.[4]

We do not take a position on the factual result of Ward. Rather, our contribution is pitched at a higher level of generality, with the legal framework constructed in the Ward majority’s reasons serving as our interlocutor. First, we argue the Court’s view of how to define expression as discriminatory is at odds with standard views in law and the scholarly literature. Second, while the Court held that only preventing the harms caused by discriminatory expression, as opposed to mere offence, could give a justifying reason to limit freedom of expression, it relied on an impoverished conception of harm that was restricted to the societal harm of hate speech. In our view, there are reasons to prohibit discriminatory expression to prevent other types of harms—reasons that can conceivably be sufficiently weighty to justify limiting expressive freedom.

In Part 2, we analyze when a person is disadvantaged by expression because of a prohibited ground of discrimination and show how the Ward majority’s view on this issue conflicts with standard views on how to define something as discriminatory. In Part 3, we explain how the tension between protecting free expression and promoting egalitarian ideals played out in Ward before exploring how it arises in other legal contexts. In Part 4, we explain how the majority resolves this tension by rejecting the offensiveness of discriminatory expression as a reason for prohibition and insisting on the need for the harm of hate speech. In Part 5, we identify harms other than the harms of hate speech whose prevention can give a reason to limit free expression without impugning the claim from Ward that offence alone gives no such reason. Part 6 responds to potential objections to our view on how to conceive of these other harms.

I. When is Expression Discriminatory?

We begin by considering when expression is properly classified as “discriminatory.” This inquiry must preface any analysis of reasons to legally prohibit such expression, since it identifies the target of the inquiry into reasons for prohibition. However, classifying expression, including Ward’s jokes, as “discriminatory” does not automatically mean that it is morally wrong or that there is a reason to prohibit it that can defeat expressive freedom.[5]

We proceed by canvassing perspectives from the analytic philosophy literature on discrimination before turning to the Canadian legal approach. Our aim is to articulate a generic, non-legal definition of discriminatory expression and evaluate the extent to which the majority’s legal analysis in Ward assists in this task. As alluded to above, we claim that Ward is of little assistance in this regard given the majority’s unorthodox approach to defining discrimination. However, we also aim to establish the broader point, which this critique helps elucidate; to determine whether expression is discriminatory, we can apply a definition of “discrimination” that parallels the standard philosophical definition of discriminatory actions or mental states.

A. The Philosophical Concept of Discriminatory Expression

In analytic legal philosophy, an act is defined as “discriminatory” against a victim if, roughly, it (i) disadvantages the victim (ii) because (iii) the victim possesses some trait.[6] If Jack dislikes deaf people and fires Jill, who is deaf, this act is “discriminatory” since it disadvantages Jill in employment, Jack is motivated by dislike for deaf people, and Jill is deaf.[7] This is an example of “direct discrimination.”[8]

Things other than discrete actions, like a rule or policy, can be “discriminatory.” If fewer women than men can meet a specified aerobic standard, Jack’s policy requiring all employees to meet this standard as a condition for promotion discriminates against Jill. It disadvantages her in employment by imposing disproportionate burdens on women compared to men, and Jill is a woman. This is a case of “indirect discrimination.”[9]

Mental states that are not manifested in external conduct can also be “discriminatory.” Consider stereotypical beliefs that ascribe negative qualities to a person simply based on the person’s membership in some social group. When asked by an elderly person for assistance, an IT technician might form a discriminatory belief that the person lacks technological proficiency based on a stereotype about the client’s age.[10] The technician makes the victim the object of a negative assessment by mentally associating older age with technological ineptitude, and the victim is elderly.

Now, expression is the outward conveyance of a meaning through sounds or visible signs or symbols. We propose that expression is “discriminatory” against a victim if it meets the same definition given above. For example, a “Whites Only” sign above the entrance to a store is discriminatory against a Black person. It denies access to the store using a criterion that explicitly mentions Black people and the person is Black. The anti-Muslim bus ad declaring that Muslim people “honour kill” their daughters is discriminatory, too. It conveys a negative appraisal of Muslim people in the form of the endorsement of a belief that Muslim people engage in barbaric practices as supposed tenets of their faith.

Discriminatory expression thus sometimes resembles an outward discriminatory act or policy. Like the “Whites Only” sign, it may constitute a “speech act” that communicates a directly discriminatory policy.[11] Or it can resemble a discriminatory thought that, like the anti-Muslim ad, externalizes a negative mental evaluation directed at a victim.

The fact that expression is uttered by a person with the intent of eliciting laughter in an audience does not preclude it from being discriminatory. For example, expression may convey a negative appraisal of members of the victim’s social group, and the speaker may view this appraisal as integral to its comedic value. Also, whether discriminatory expression has a comedic purpose may be relevant to assessing the strength of reasons against legally prohibiting it, even if it causes harm that the state has a reason to prevent. For example, its comedic purpose may qualify it as artistic expression, which bolsters existing reasons to refrain from prohibiting it.[12]

B. The Legal Concept of Discrimination in Canada

Our proposal to apply the same generic definition of discriminatory acts or thoughts to expression gains support from Ward, where the Court applied the same generic legal test for determining whether actions or policies are discriminatory to expression.[13] But when determining whether the comedian’s expression was discriminatory, it encountered a difficult question of when disadvantage imposed on a victim is “because” of a trait of the victim. In what follows, we discuss two approaches to this issue that can be gathered from the philosophical literature. We then look for guidance in determining which approach to favour by considering which the Court favoured in Ward and which better comports with Canadian antidiscrimination law more generally.

The two approaches can be understood as divergent interpretations of the idea that, for disadvantage to be “because” of a trait, the trait must be a reason for the disadvantage. The first approach is “victim-focused.”[14] On this approach, a victim’s trait must be an explanatory reason for the disadvantage.[15] For example, when Jill is fired because of her disability, her disability causally explains why Jack fired her. This indicates that a trait can explain the disadvantage a victim suffers if it figures into the perpetrator’s subjective reasons or motivations for causing the disadvantage. But the point of examining the perpetrator’s mind on this first approach is to explain how the victim’s trait caused her disadvantage by tracing the causal chain through the perpetrator’s mind. It is not to pinpoint some moral flaw in the perpetrator’s deliberative process. Mental attributes need not be part of the causal narrative at all. In a case of indirect discrimination in employment, a victim’s trait may causally explain why she suffers disadvantage under a promotions policy if, regardless of the employer’s own reasons for instituting the policy, it has a disparate impact on employees sharing that trait.

The second approach is “perpetrator-focused.”[16] A victim’s trait may be the motivating reason the perpetrator took within his internal deliberations as justifying the disadvantage. This approach scrutinizes the perpetrator’s cognitive activity to identify the true basis or ground of his choice to disadvantage the victim. That his reasoning was on the causal pathway between the victim’s trait and her being disadvantaged is beside the point. It is possible for Jack to subjectively take Jill’s disability as a good reason to disadvantage her without ever actually firing her or unsuccessfully attempting to fire her.

To determine whether to prefer the victim—or perpetrator-focused approach—we can look for guidance to the Canadian legal test for discrimination.[17] To obtain a statutory remedy, an alleged victim of discrimination must show that (i) she was disadvantaged by the perpetrator by being deprived of a right conferred under the statute (such as a right against discrimination in employment or housing) (ii) because (iii) the victim has a trait that is protected as a prohibited ground of discrimination.[18] The legal definition of discrimination is narrower than the philosophical definition, as it applies only to deprivations of statutory rights (not any disadvantage whatsoever) because of certain prescribed traits (not any trait whatsoever). But since they share the crucial second element, the legal treatment of this element can inform the broader non-legal definition.

Ward addressed whether the comedian’s expression was discriminatory “because” of the prohibited ground of disability. The Commission des droits de la personne et des droits de la jeunesse brought a discrimination claim against Mike Ward on behalf of a minor, Jérémy Gabriel. Gabriel was born with Treacher Collins syndrome, which caused malformations of his head and deafness. When he was six, he received a bone-anchored hearing aid that enabled him to learn to speak and sing. He became a professional singer, performing at large events that received significant media coverage. He released an album and autobiography and participated in international initiatives to raise awareness of Treacher Collins. Ward was a professional comedian in the style of dark comedy, which features shocking or taboo subjects. The main theme of one of his public shows was tolerance for differences between people. He performed a routine during which he mocked figures in Quebec’s artistic community who he described as “sacred cows” that could not be made fun of due to their wealth, influence, or perceived vulnerability. One figure was Gabriel. Ward also made disparaging comments about Gabriel’s physical appearance in a video posted online. Afterwards, students attending Gabriel’s school teased him, drawing inspiration from some of the comments in that video.[19]

Section 10 of the Quebec Charter provides that every person has the right to the full and equal recognition and exercise of other rights conferred by the Quebec Charter “without distinction, exclusion or preference based on ... handicap.”[20] The Commission argued that Ward’s jokes deprived Gabriel of his right under section 4 of the Quebec Charter to “the safeguard of his dignity, honour, and reputation” because of Gabriel’s disability. The Quebec Human Rights Tribunal, adjudicating the Commission’s claim at first instance, distinguished between Ward’s decision to comment on Gabriel’s appearance and the comments themselves. The decision was not discriminatory because of disability: “[Gabriel] was chosen as a target because he was a public personality who attracted public sympathy and seemed to be ‘untouchable’ ... [Ward] did not choose [Gabriel] because of his handicap.”[21] But the comments themselves were discriminatory. They referred to Gabriel as the “ugly kid who sings” whose hearing aid resembled a “speaker” on his head.[22]

However, a majority of the Supreme Court, in reasons written by Chief Justice Wagner and Justice Côté, held that after maintaining that the decision to make fun of Gabriel was not because of Gabriel’s disability, the Tribunal should not have gone farther and considered the message Ward chose to express in isolation from Ward’s decision itself. Even if Ward’s comments mentioned Gabriel’s disability, it did not follow that the disability was a factor contributing to Gabriel’s being targeted for ridicule.[23]

The majority’s approach to defining discrimination was perpetrator-focused. By accepting that Ward’s decision to mock Gabriel was not because of disability but public notoriety, the Court accepted that its task was to identify the true basis of Ward’s choice to comment on Gabriel, i.e., to assess Ward’s internal deliberations and discern whether disability was the ultimate subjective motivating reason for expressing a negative appraisal of Gabriel’s appearance. A victim-focused approach may well have led to the opposite conclusion. One can tell a plausible story according to which Gabriel was disadvantaged by being made an object of derision, which was explained by the content of Ward’s comments and how they emphasized Gabriel’s appearance. Since Gabriel’s appearance was explained by his disability, his disability explained why he suffered derision. The connection between derision and disability travelled through the meaning of the words used in the jokes that Ward expressed. No doubt, before being expressed, the jokes were formulated in Ward’s mind, and Ward may have had several factors in mind when deciding who to make fun of (including celebrity status and “untouchability”). But on a victim-focused view, it did not matter precisely where disability figured into Ward’s deliberations or whether disability was the most fundamental factor in Ward’s choice to express his jokes. All that mattered was whether the link between disability and disadvantage could somehow be traced through Ward’s reasoning, even if disability was not the dominant or sole reason.

The problem is that under Canadian antidiscrimination law, the victim-focused approach enjoys ascendency over the perpetrator-focused approach. The dissent in Ward, led by Justices Abella and Kasirer, pressed this critique against the majority, writing that it was not necessary that Ward “intended to single out Mr. Gabriel based on his disability, or that the distinction was based exclusively, or even primarily, on his disability. There must only be a connection between the distinction and the ground.”[24] For something to be discriminatory “because” of a prohibited ground, there need only be some link, nexus, or connection between the disadvantage it causes for a victim and the victim’s trait. The trait must be a factor in the disadvantage. But it need not be the only reason for it. Nor need it be the closest, most direct, significant, material, or dominant reason. It need only contribute to the disadvantage in some way. It is irrelevant whether the disadvantage had its true basis in, for example, stereotypical or arbitrary decision-making of the perpetrator. What is relevant is not the perpetrator’s attitude, but the role the victim’s trait played in the negative impact she suffered.[25] In an incisive dissenting opinion in Stewart v. Elk Valley Coal Corp., Justice Gascon explained that the issue of whether disadvantage is “because” of a prohibited ground is concerned with effects on victims, not perpetrators’ intentions:

The difference between effect and intent, analytically, is best understood from the standpoint of the relationship under consideration. If discriminatory intent were dispositive of contribution, the relevant relationship would be that between an employee’s protected ground and the corporation’s intent to harm that employee. But contribution emphasizes discriminatory effect. Indeed, for human rights legislation to protect against “indirect discrimination”—i.e. neutral rules with adverse consequences for certain groups—intent cannot be a requirement for prima facie discrimination. Therefore, the relevant relationship addressed by contribution is that between an employee’s ground and harm.[26]

What Justice Gascon says is that, when analyzing whether something is discriminatory because of a specified trait, courts must consider whether the trait contributes to disadvantage suffered by the victim by focusing on the relationship between trait and disadvantage (in a victim-focused sense), not the trait and the perpetrator’s reasons for causing the disadvantage (in a perpetrator-focused sense).[27] Ward turned this approach on its head.[28]

This point is not of mere technical or methodological significance. To suppress the victim-focused approach is to tacitly endorse a specific substantive view of the purpose of antidiscrimination law. Given how the perpetrator-focused approach involves scrutiny of a perpetrator’s true motivating reason for disadvantaging a victim, it aims to identify some inherent moral defect in the perpetrator’s mental deliberations. The sought-after defect would reside in how, most of the time, the traits covered by the legally prohibited grounds—race, sex, disability, etc.—do not give any objectively real or genuine reason to disadvantage someone in areas like employment or housing.[29] If something constitutes illegal discrimination just when this defect is present, the point of outlawing discrimination must be to prohibit morally defective deliberations and eradicate a type of morally blameworthy process of individual decision-making from society.

A victim-focused approach to defining discrimination presupposes a distinct conception of antidiscrimination law’s purpose. By focusing on the relationship between a victim’s trait and a disadvantage experienced at a perpetrator’s hands, rather than any moral fault in the perpetrator’s decision-making processes, this approach aims to identify scenarios where that trait has become a locus of disadvantage, setting aside the blameworthiness of the perpetrator for causing disadvantage and trait to align. But the traits usually included in the legally prohibited grounds characterize social groups that have been historically marginalized. So, if something is illegal discrimination whenever trait and disadvantage coincide, the goal of prohibiting discrimination must be to disrupt patterns of social inequality faced by historically marginalized groups and prevent the harms such patterns cause for these groups.[30] The moral blameworthiness of individual agents who contribute to structural inequality is immaterial for achieving this goal.[31]

The way that the perpetrator-focused account conceives of antidiscrimination law’s purpose is unacceptable because the law’s point is not to condemn morally blameworthy conduct like the criminal law does.[32] For example, according to the Supreme Court, an employee who has been sexually harassed by a coworker may bring a human rights code complaint against her employer, and it is irrelevant whether the employer was at fault for intentionally encouraging the harassment or negligently failing to prevent it. Concepts of vicarious liability from tort law, or strict liability from criminal law, are inapplicable because a human rights code “is not aimed at determining fault or punishing conduct. It is remedial. Its aim is to identify and eliminate discrimination.”[33] The suppression of the victim-focused approach to defining discrimination in Ward therefore contradicts settled understandings of the overall purpose of antidiscrimination law.

In sum, although Ward supports the application of the standard philosophical concept of “discrimination” to expression, despite initial impressions, it cannot support the use of the perpetrator-focused approach to ascertaining whether disadvantage suffered by a victim is “because” of a specified trait. This is because Ward neglects the prevailing legal concept of discrimination, which informs the philosophical concept by indicating that expression is discriminatory when it makes the victim the object of a negative assessment “because” of the victim’s trait in a victim-focused sense.

II. Clashes Between Freedom of Expression and Antidiscrimination

We now wish to address what reasons exist for legally prohibiting expression that is properly classified as discriminatory. Here we confront a conflict between antidiscrimination statutes’ goal of reducing inequality and reasons to protect expressive freedom. As a prelude to examining how the Court resolved the conflict in Ward, we will survey how it arises across various legal contexts—human rights codes, constitutional law, criminal and quasi-criminal law, and private law. This survey will yield analytical resources for when we eventually turn to critique the reasoning in Ward.

A. Human Rights Codes

In Ward, the Commission asserted Gabriel’s right under section 10 of the Quebec Charter against disability discrimination that deprives him of the safeguard of dignity under section 4. Section 4 is subject to section 9.1, which provides that in exercising fundamental rights and freedoms, “a person shall maintain a proper regard for democratic values, public order and the general well-being of the citizens of Quebec.”[34] Thus, the scope of the section 4 right is internally limited by democratic or public values. The Commission had to prove that Ward’s jokes deprived Gabriel of a right to the safeguard of dignity whose protection was justified despite the need for it to be exercised consistently with Ward’s freedom of expression.[35]

Human rights statutes outside Quebec contain provisions that prohibit a person from publishing or displaying a sign, symbol, or representation that indicates an intention to discriminate or incites others to discriminate. These provisions sometimes contain a qualifier that they must not be interpreted as interfering with freedom of expression.[36] Similarly, antidiscrimination statutes typically prohibit sexual harassment or harassment on other grounds.[37] “Harassment” is often defined as including a course of “vexatious comment” against a victim that ought reasonably to be known to be unwelcome.[38] Finally, among the traits listed as prohibited grounds of discrimination in many provinces are a person’s gender identity and gender expression. It has been suggested that deliberately misgendering a person could be a legally prohibited form of expression.[39] In all these ways, the human rights codes’ objectives are pursued by restricting free expression.

B. Constitutional Litigation

Some human rights codes contain prohibitions of discriminatory expression that, mirroring the crime of hate speech, prohibit expression that is likely to expose a person or group to hatred or contempt because of a prohibited ground.[40] These provisions have been subject to challenges on the basis that they infringe subsection 2(b) of the Charter in a manner that cannot be “demonstrably justified in a free and democratic society” under section 1.[41]

The prohibition of hate speech under the now-repealed section 13 of the Canadian Human Rights Act was challenged in Canada (Human Rights Commission) v. Taylor as infringing subsection 2(b) of the Charter.[42] John Ross Taylor distributed telephone messages asserting that Jewish people were part of a conspiracy to take over Canadian society. The Supreme Court, guided by its approach to the crime of hate speech, held that section 13 limited freedom of expression in a manner that was justified under section 1 of the Charter.[43] It rejected the argument that Taylor’s discriminatory messages were not constitutionally protected at all by being antithetical to the values underlying subsection 2(b) (the search for truth through the free public exchange of ideas, democratic political participation, and self-actualization on the part of the speaker). Subsection 2(b) must protect any expression that conveys a meaning irrespective of its content. While violent forms of expression do not receive protection because they are inimical to expressive freedom’s underlying values, Taylor’s messages did not take the form of direct application of physical violence, and section 13 of the Canadian Human Rights Act targeted their content, not their form.[44] Taylor thus makes clear that the protection of freedom of expression requires inclusive protection of even expression that is viewed as unpopular or contrary to mainstream normative beliefs and practices. Subsection 2(b) of the Charter cannot achieve its purposes unless it embraces messages that dissent from the views of the social or political majority.[45]

The prohibition of hate speech in Saskatchewan’s antidiscrimination statute was more recently challenged in Saskatchewan (Human Rights Commission) v. Whatcott.[46] William Whatcott distributed flyers on behalf of a Christian group that urged elementary schools to not hire homosexual persons as teachers or teach students about homosexual practices. These flyers allegedly contravened paragraph 14(1)(b) of the Saskatchewan Human Rights Code by exposing people to hatred, or ridiculing, belittling, or otherwise affronting their dignity because of their sexual orientation.[47]

The Supreme Court interpreted the term “hatred” in paragraph 14(1)(b) as referring to expression that vilifies, rejects, delegitimizes, inspires enmity, or galvanizes ill-will against a group characterized by a prohibited ground, rendering them abhorrent, dangerous, or unacceptable in the eyes of the audience.[48] The reason to prohibit hatred is to prevent certain societal effects, particularly its potential to lead listeners to engage in discriminatory acts or otherwise harm members of the group.[49] It is not that listeners may find it offensive or repugnant to mainstream values and beliefs.[50] The objective of paragraph 14(1)(b) was thus to reduce the harm of discriminatory conduct by tackling one of its root causes. Expression that delegitimizes vulnerable groups and rejects them as undeserving of social standing and acceptance makes it easier to justify discriminatory conduct toward them. The Court described hate propaganda as in need of proscription, not because of how it causes emotional distress for individual members of targeted groups, but because of its societal impact:

If a group of people are considered inferior, subhuman, or lawless, it is easier to justify denying the group and its members equal rights or status ... As the majority becomes desensitized by the effects of hate speech, the concern is that some members of society will demonstrate their rejection of the vulnerable group through conduct. Hate speech lays the groundwork for later, broad attacks on vulnerable groups. These attacks can range from discrimination, to ostracism, segregation, deportation, violence and, in the most extreme cases, to genocide.[51]

Furthermore, hateful expression undermines the ability of the targeted group to find self-fulfillment through their own expressive activity and participate in democratic deliberation. It inhibits their ability to respond to the ideas being expressed by forcing them to argue for their basic social standing as a precondition of participating in public debate.[52]

However, not all the words in paragraph 14(1)(b) advanced the objective of eliminating the harms of hate speech. Expression that “ridicules,” “belittles,” or “affronts the dignity” of marginalized groups may be considered repugnant, hurtful, disdainful, or offensive. But it typically does not have the potential to recruit others to engage in discriminatory conduct toward victims in the way that hateful, delegitimizing, or vilifying expression does. Therefore, the limit on subsection 2(b) of the Charter these words caused could not be justified under section 1, and they were severed from paragraph 14(1)(b).[53]

Whatcott involved a challenge against both a statutory provision and an administrative tribunal’s decision that Whatcott’s flyers were prohibited by that provision.[54] This indicates that another context where the legal regulation of discriminatory expression raises a conflict between freedom of expression and the pursuit of equality is where an executive government agent, to promote equality, makes a decision that is said to limit freedom of expression. The agent must appropriately weigh the Charter value of free expression.[55] For example, if a municipality refuses to allow an anti-abortion group to place a bus ad that uses graphic imagery and language to describe abortions, the refusal may be challenged on judicial review for whether it complied with principles of proportionality. The municipality must properly balance between the subsection 2(b) Charter value of freedom of expression and the need to prevent psychological harm for women, fear, and confusion for children, and inaccurate or misleading advertising.[56] Similarly, a human rights tribunal must satisfy proportionality principles and account for freedom of expression when deciding whether the statutory prohibition of discriminatory expression in employment extends to sexist comments made by a worker about a manager in online blog posts outside work hours.[57]

C. Criminal and Quasi-Criminal Law

Taylor and Whatcott sought guidance from the criminal offence of hate speech. But other offences also limit freedom of expression. These include causing a public disturbance,[58] distributing obscene material[59] or child pornography,[60] criminal harassment,[61] uttering threats,[62] publishing defamatory libel,[63] or causing mischief.[64] Each of these offences could be committed by means of discriminatory expression, in which case they criminalize such expression.

For example, a racial epithet used by a white supremacist to threaten to harm a Black person could be criminalized under the uttering threats offence.[65] If a person shouts Islamophobic comments in a public space, such as a subway train, causing the train to be shut down and delaying the commute of other riders, he may be convicted of causing a public disturbance.[66] That said, shouting “insulting or obscene language” is alone insufficient to constitute a disturbance; the expression must cause “... disorder calculated to interfere with the public’s normal activities.”[67] Thus, the fact that discriminatory expression is insulting or obscene is insufficient for criminalizing it. But its discriminatory nature is not irrelevant. For example, deliberately shouting inflammatory, racist remarks on a subway platform is more likely to rouse public disorder than polite conversation about the weather. Similarly, if expression interferes with another’s use or enjoyment of property, its criminalization under the offence of mischief limits expressive freedom. For instance, a person could conceivably interfere with a Jewish person’s use or enjoyment of their home, and thereby commit mischief, if he stood outside the house waving Nazi flags.[68]

Discriminatory expression can also be limited by regulatory offences. For example, a provincial statute may impose a fine for the use of language that interferes with others’ use and enjoyment of a public space. It is not difficult to conceive of a case where a person is prosecuted under such a statute for, say, displaying a sign denigrating recent Hispanic immigrants that reaches the threshold of interfering with a Hispanic citizen’s use and enjoyment of a park.[69]

D. Private Law

Finally, certain private wrongs can be committed through discriminatory expression, giving rise to a cause of action for victims to obtain a personal remedy, such as monetary compensation from perpetrators through a civil proceeding. The Commission’s claim on Gabriel’s behalf in Ward resembled a civil action for defamation founded on a breach of the safeguard of dignity under section 4 of the Quebec Charter. The majority highlighted distinctions between discrimination and defamation.[70] However, at least at common law, compensation for the tort of defamation is available to victims of defamatory discriminatory expression whose reputations are harmed.

To elaborate, an action in defamation lies where a defendant publicizes a statement about the plaintiff that tends to lower the plaintiff’s reputation in the eyes of a reasonable person.[71] The defendant may establish a defence, such as truth or fair comment.[72] The common law must be developed in accordance with the Charter value of freedom of expression. But given the importance of the individual interest in good reputation and the tenuous connection between defamatory statements and the core values underlying subsection 2(b) of the Charter, the common law of defamation is consistent with the value of freedom of expression.[73] And a defendant may lower a plaintiff’s reputation via discriminatory expression. For example, defamation can take the form of referring to a Muslim person as a terrorist who is plotting to commit acts of violence and war crimes in Canada.[74]

Another private wrong that may be occasioned by expression is intentional infliction of mental suffering (IIMS). This tort requires a plaintiff to prove that the defendant engaged in flagrant and outrageous conduct calculated to produce harm that results in visible and provable illness.[75] The impugned conduct of an IIMS defendant often takes the form of expression, although the high thresholds for the expression to be flagrant, outrageous, and intended to cause harm allow space for people to freely make critical, unflattering, or distasteful statements about others without attracting civil liability.[76] Tort law scholars have advocated for the use of IIMS to address discriminatory expression.[77] For example, a woman may be entitled to damages for IIMS if she is subjected to sexual harassment in the workplace through sustained and vexatious comments by male coworkers who intended to cause her mental anguish.[78] Similarly, police officers may be liable for IIMS if they utter racist insults toward an Indigenous man in the course of an arrest.[79] Further, consider a civil cause of action created by Nova Scotia legislation for electronic communications denigrating another person because of a prohibited ground of discrimination, where those communications are maliciously intended to harm the other person’s health.[80] This supplies a model for understanding how the IIMS tort could be committed through discriminatory expression that is calculated to cause harm.[81]

This completes our survey of various legal contexts that feature a tension between reasons to protect freedom of expression and the egalitarian objectives of antidiscrimination law. Next, we consider how the Ward majority resolved this tension before drawing on our survey to critique the majority’s resolution.

III. The Rejection of the Offence Principle in Ward

In Ward, the Commission had to prove that Gabriel’s dignity warranted protection from Ward’s comments despite Ward’s freedom of expression. The Court defined the right to dignity under section 4 of the Quebec Charter as protecting a person from being “stripped of their humanity by being subjected to treatment that debases, subjugates, objectifies, humiliates or degrades them ... [and] does no less than outrage the conscience of society.”[82] It also affirmed Whatcott’s definition of hate speech as expression that lowers the social standing of the victim’s group and has the harmful downstream effect of heightening the group’s vulnerability to further discriminatory conduct.[83] By aligning the scope of the protection afforded by the section 4 dignity right with the definition of hate speech, the majority held that the only reason to protect a victim’s dignity that could justify prohibiting discriminatory expression is to prevent the harm of hate speech, not mere hurt feelings, humiliation, or offence.

The Human Rights Tribunal had accepted that discriminatory expression can be legally prohibited if a reasonable person, in the circumstances of its victim, would not temper their reaction to it given the need, in a pluralistic society, to tolerate even immoderate expression.[84] But the Supreme Court maintained that invoking the perceptions of a reasonable person allows expression to be limited to prevent mere offence or repugnance and emotional harm to an individual rather than harmful societal effects.[85] It formulated a new test for justifiably limiting discriminatory expression that approximated the principles from Whatcott. The safeguard of dignity is violated by such expression if, first, a reasonable person, aware of the relevant context and circumstances, would view it as inciting others to vilify the targeted individual or group or detest their humanity. Second, the reasonable person must view the expression as likely to jeopardize the targeted group’s social acceptance and lead to harm against them in the form of discriminatory treatment by third parties.[86]

The Court stated that discriminatory humour is unlikely to lower the targeted group’s social standing and encourage harmful conduct against its members, as the audience in a comedic context ordinarily accepts that ridicule involves deliberate exaggerations and distortions of reality that are not to be taken at face value.[87] Indeed, the Court concluded that Ward’s jokes did not infringe Gabriel’s right to the safeguard of dignity. While Gabriel testified about the deep pain the comments caused him, the applicable test focuses not on the emotional harm suffered by the target of discriminatory expression, but on the likely effects of the expression in the form of harmful conduct by third parties against the target.[88] Ward’s jokes were offensive, but they did not incite others to vilify Gabriel.[89] Moreover, although the jokes exaggerated Gabriel’s physical appearance to deride him, they were made by a career comedian known for dark humour who exploited feelings of discomfort for entertainment. Considered objectively, they “were not likely to have a spillover effect that could lead to discriminatory treatment of Mr. Gabriel.”[90]

Ward thus clearly rejects what has been called the “offence principle”—that is, the idea that the offensiveness of discriminatory expression gives a reason that is strong enough to potentially justify limiting liberty under a human rights code. The Court instead endorses a version of the “harm principle,” according to which only the prevention of harm to others caused by discriminatory expression gives such a reason.[91] It is useful here to grasp the distinction between harm and offence often drawn by legal theorists. A harm is a change in a person’s condition that makes her worse off by impairing her physical and psychological capacities to flourish or diminishing her opportunities to pursue valuable, self-chosen projects. Offence is usually understood as a relatively more trivial affront to a person’s sensibilities; an unpleasant or aversive psychological experience that nevertheless does not impair well-being. Oftentimes, the taking of offence is mediated by a judgment that the conduct deemed offensive has violated some social or conventional norm.[92]

Discriminatory expression is offensive in this mediated way. It affronts sensibility because of a judgment that it fails to abide by norms of appropriate interpersonal interaction. Although Ward’s jokes caused offence and emotional harm to Gabriel, every content or meaning conveyed by expression must be protected by freedom of expression, even if it violates those norms. In a liberal society committed to free expression, expression that only offends others must be tolerated.[93]

According to Ward, only a very specific type of harm gives a sufficiently weighty reason that can justify limiting discriminatory expression: the conception of the harm of hate speech articulated in Whatcott. This is the broader societal harm of lowering the victim’s social standing and laying the groundwork for them to be discriminated against by others. Whatcott and Ward emphasize the harmfulness of discriminatory expression that is addressed to “recruits”[94] and aims to inspire third parties to accept the illegitimacy of the target group. On this “recruitment” view, the harm whose prevention justifies limiting discriminatory expression is a diffuse environmental harm. It consists in rendering the background social climate in which the target group lives inhospitable by undermining the public good of assurance that they are safe from further spillover harm by third parties.[95] When an audience is encouraged to embrace the attitude that the target group has inferior social status, they become less reluctant and emboldened to manifest this attitude by means of discriminatory conduct toward the target group. Ward’s jokes did not cause this general atmospheric harm, so safeguarding Gabriel’s dignity was unwarranted.

However, in what follows, we argue that the Ward framework is incomplete. We do not dispute that preventing offence is alone incapable of justifying limits on free expression. But the Court’s view of the harm whose prevention justifies such limits is myopic. Preventing harms other than recruitment harms, which are not identical to the more trivial taking of offence, also gives a reason to prohibit discriminatory expression under antidiscrimination statutes, and this reason can sometimes outweigh the reasons to protect free expression.[96] We do not claim that Ward’s jokes caused proscribed harm within this unacknowledged legal category. We also do not deny that, even if they did, the fact that Ward expressed them with a humorous or artistic intent adds to the reasons for protecting Ward’s expressive freedom and raises the bar for justifying their legal prohibition. We simply claim that the Court failed to recognize the additional category of harm in its account of the relationship between antidiscrimination and free expression.

IV. Beyond the Recruitment View

We start with a query: Can there be instances of discriminatory expression that, intuitively, should be prohibited by a human rights code to prevent harm, even if they do not cause the kind of environmental harm contemplated by the recruitment view? We think so.

Suppose that a same-sex couple asks a baker for a cake celebrating their wedding, but the baker refuses on the ground that this would require her to endorse sexual practices condemned by her religious beliefs. Instead of politely declining, the baker utters a homophobic insult toward the couple as she expels them from her bakery. Imagine that when news of this exchange reaches the public, several other bakers offer to bake for the couple—free of charge—a cake that is more beautiful than the one the initial baker would have made.[97] In this hypothetical, far from encouraging third parties to accept the lower social standing of gay people, the baker’s expression inspires support for same-sex marriage. We suggest that there is still an intuitive reason to prohibit the expression to prevent harm to the victimized couple, not just offence. If the couple were to bring a claim before a human rights tribunal, they may not be able to demonstrate compensable losses in the form of a lost wedding cake, but they may still obtain a remedy of nominal damages in recognition that they had some legal right violated.[98]

To explain the intuition in question, we must investigate whether discriminatory expression causes some type of harm other than recruitment harm that there is good reason to prevent. The Supreme Court’s decision in R. v. Keegstra offers guidance here. Chief Justice Dickson enumerated two kinds of harms caused by discriminatory expression whose prevention gave the criminalization of hate speech a valid objective. Only the second concerned effects akin to recruitment—that is, hate speech’s “influence upon society at large,” or the potential to “attract individuals to its cause, and in the process create serious discord between various cultural groups in society.”[99] The first was immediate and direct harm to the individual victim of discrimination deriving from disrespect shown toward the social group to which the victim belongs. Criminalizing hate speech prevents “emotional damage” to the victim, “grave psychological” consequences, “humiliation and degradation” and “a severely negative impact on the individual’s sense of self-worth and acceptance.”[100]

The Chief Justice referred to the writings of American theorists who explore the justifications for limiting free speech under United States constitutional law. These theorists argue that race-based discriminatory expression causes serious and tangible damage to the psychological integrity and emotional well-being of its victims.[101] Their arguments have received support from more recent empirical research. American psychologists have articulated a psychological condition, similar to post-traumatic stress disorder, called “race-based trauma” to capture the emotional injury of racial slurs.[102] Similar conditions have been documented in victims of anti-Semitic and homophobic expression.[103] In a study of how Black and Indigenous victims of racial insults self-report their experience of the impact of racial insults, the authors conclude that discriminatory expression causes emotional injury to individual victims in addition to inciting third parties to engage in disorderly conduct. They contend that the law must aim to prevent this individual harm or else it risks denying victims’ experience with racism.[104]

Accordingly, it is possible to vindicate the intuition that there can be a reason capable of justifying prohibitions on discriminatory expression to prevent harms other than recruitment harms by invoking a reason to prevent immediate and direct harm to the psychological integrity of an individual victim. Earlier, we defined discriminatory expression as the external manifestation, through words or symbols, of a mental state, belief, or attitude that makes the victim the object of a negative appraisal because of the victim’s group membership. In our view, people who are members of the historically marginalized groups within patterns of social inequality that the legally prohibited grounds of discrimination characterize have an interest in not being victimized in this fashion. When they are, they can suffer emotional or psychological damage that amounts to a non-trivial setback to a person’s well-being and an impairment of the person’s capacity to flourish and fully achieve their aspirations in life.[105] Preventing that harm provides a reason that is potentially strong enough to limit freedom of expression and is entirely independent of the need to safeguard people’s assurance that they live in a safe and accommodating social environment. As stated by Justices Abella and Kasirer in their Ward dissent, “speech can cause individual harm without being hateful, and there is no constitutional bar to legal recourse in such circumstances where the limit on freedom of expression is aimed at different objectives.”[106]

Yet we do not rest exclusively on moral intuitions about the hypothetical case we described above, for the prevention of psychological harm is not unheard of in Canadian law. We can see this by revisiting the different contexts where the law deals with conflicts between free expression and antidiscrimination.

Consider first the crime of distributing child pornography. This offence is subject to a defence if the material the accused distributed depicting a person under eighteen as engaging in sexual activity “has a legitimate purpose related to the administration of justice or to science, medicine, education or art” and “does not pose an undue risk of harm to persons under the age of eighteen years.”[107] The second aspect of this defence has been interpreted in a manner that ensures that the criminalization of child pornography only punishes not mere offensiveness against prevailing moral views of the community, but a type of harm whose prevention can justify limiting freedom of academic or artistic expression.[108] Now, among the many evils of child pornography is that it amounts to discriminatory expression against children; it disadvantages its victims because of age, similar to how adult pornography has been said to involve sex-based discriminatory expression against women.[109] It follows that the offence of child pornography, insofar as it is interpreted as preventing harm of the proper kind (not just offence), justifiably limits discriminatory expression. Although child pornography can plausibly cause the kind of recruitment harm contemplated in Whatcott by rendering children more susceptible to mistreatment throughout society, it also causes serious psychological harm to the individual children portrayed in it. Preventing that harm gives just as good a reason to criminalize child pornography as does preventing recruitment harm.[110]

Similar remarks apply to uttering threats.[111] Threats of violence are not constitutionally protected forms of expression.[112] But threats can be conveyed through discriminatory expression, as when a white supremacist threatens bodily harm against a Black person using racial slurs[113] or a man writes letters threatening sexual assault to women.[114] Although its commission can cause recruitment harm by, for example, lowering the social standing of Black people or women, it also plausibly causes psychological harm to its individual victims by instilling fear or intimidation, which also gives a justificatory reason for the offence. Indeed, the prohibition of threats of bodily harm encompasses threats of “psychological harm [that] substantially interferes with the health or well-being” of the victim, and because of the emotional trauma experienced by victims of sexual assault, the prohibition must capture threats of sexual assault.[115]

Recall as well that certain torts can be committed through discriminatory expression. A tortfeasor may utter xenophobic remarks to publicly defame a Muslim business owner.[116] The reason why a cause of action for defamation justifiably limits freedom of expression is that forestalling attacks on a victim’s reputation prevents harm to “the innate worthiness and dignity of the individual,” their “sense of worth and value,” and their “self-image and sense of self-worth.”[117] Publishing defamatory comments also constitutes “an invasion of the individual’s personal privacy and is an affront to that person’s dignity.”[118] But if defamation sets back victims’ interests in their sense of self-worth, dignity, and privacy, it can cause psychological and emotional injury whose prevention warrants limiting free expression. Defamation partially overlaps with other “dignitary torts,” including IIMS and invasion of privacy, in that it protects victims from the emotional distress, embarrassment, humiliation, or disrespect that can come from being the target of diminished regard by others.[119] Its commission can cause psychological harms when it occurs via discriminatory expression, i.e., when the tortfeasor’s attack on a specific victim’s reputation proceeds by denigrating the social group with which the victim is associated. If so, the prevention of these harms justifies the limit on free expression imposed by a private cause of action.[120]

That justification is freestanding from how discriminatory expression must be limited to prevent the cultivation of a social climate within which third parties are emboldened to mistreat members of the victim’s social group. Certain private law actions make this distinction explicit. Recently, a tort of “internet harassment” has been recognized, drawing an analogy with Nova Scotia’s Intimate Images and Cyber-protection Act.[121] This statute creates a private action for “cyber-bullying,” which includes electronic communications that either “denigrate another person because of any prohibited ground of discrimination” or “incite or encourage another person to,” inter alia, engage in discriminatory electronic communications.[122] So, the statute protects against the immediate harm of being denigrated by online discriminatory expression, which would encompass psychological harms suffered by the individual victim of the expression, such as damage to self-esteem, depression, or anxiety,[123] but these harms are treated as independent from the harms of inciting or encouraging third parties to commit online discriminatory expression against the victim. Thus, Nova Scotia’s statutory cause of action for cyber-bullying is available for online discriminatory expression that is “maliciously intended to cause harm to another individual’s health or well-being” if the relevant harm is not recruitment harm but harm to mental health.[124]

That human rights codes may prohibit discriminatory expression to prevent psychological harm explains the nature of the compensatory remedies typically available for breaching the codes. For example, the British Columbia and Ontario codes provide that a human rights tribunal may order compensation “for injury to dignity, feelings and self-respect.”[125] Among the factors that tribunals consider when assessing the quantum of damages to award for unlawful discrimination are the victim’s “vulnerability,” the “psychological impact on the victim” or the “immediate impact” on the victim’s emotional health, and the “degree of anxiety” the discrimination caused.[126] The same-sex couple who are subjected to homophobic remarks by the baker but who receive a substitute wedding cake from third parties may therefore be entitled to more than nominal damages after all. The extent of their injury might not be quantifiable by reference to the value of the service they were denied, but it may be quantifiable by reference to the nature of the immediate harm to their mental health, if any, they might have experienced. This is consistent with the prohibition of purely verbal harassment under the human rights codes. Given that harassment is a “demeaning practice” that “attacks the dignity and self-respect of the victim,”[127] it can cause psychological harm that is compensable under antidiscrimination statutes’ remedial provisions.

The difference between this kind of harm and recruitment harm has already been recognized, albeit inchoately, in human rights law. Subsection 7(1) of the British Columbia Human Rights Code prohibits the publication of any discriminatory representation that a) “indicates discrimination or an intention to discriminate against a person or a group or class of persons” or (b) “is likely to expose a person or a group or class of persons to hatred or contempt.” In Oger v. Whatcott (No 7), the British Columbia Human Rights Tribunal held that flyers distributed by William Whatcott, which disparaged a transgender political candidate as living contrary to God’s will, were prohibited by paragraphs 7(1)(a) and (b) of the Code.[128] In rendering its decision, the Tribunal had to proportionately balance between the Charter value of freedom of expression for Whatcott and the Code’s statutory objectives. A full proportionality analysis was not necessary for paragraph 7(1)(b), which prohibits hate speech and is identical to the Saskatchewan hate speech prohibition that the Supreme Court held justifiably limits subsection 2(b) of the Charter in Whatcott.[129] By contrast, paragraph 7(1)(a) “targets a different kind of speech than was at issue in Whatcott,” and a de novo proportionality analysis was needed for it.[130] Member Cousineau stated that whereas paragraph 7(1)(b) targets “speech that exposes or is likely to expose a protected group to ‘detestation and vilification,’” paragraph 7(1)(a) targets discriminatory expression that does “not meet the threshold for hate but clearly indicates discrimination.”[131] It captures expression that causes “some real-world, discriminatory effect, beyond merely being offensive,”[132] or “specific adverse consequences” for the victim, which need not be limited to adverse effects within the areas of employment, housing, and services, although the prohibited expression must be “more than a mere statement of opinion.”[133]

While Member Cousineau recognized that the species of harm that paragraph 7(1)(a) prohibits is distinct from offence and from the category of harm that paragraph 7(1)(b) prevents, she unfortunately left the parameters of the middle category ambiguous. The “specific adverse [consequence]” she fixated on was how, by casting transgender persons as inherently immoral, Whatcott sought to dissuade others from voting for the candidate based purely on her gender identity and to “block the doors of government with a message that the political realm is for ‘cisgender people only.’”[134] The goal of forestalling this outcome was proportionate to limiting Whatcott’s freedom of expression. But it is difficult to see how it differs from preventing recruitment harm.

However, even if Member Cousineau did not draw the distinction between the harms that paragraphs 7(1)(a) and 7(1)(b) respectively prevent quite so neatly, it is easily grasped by invoking the distinction between the psychological harm to the individual victim and the diffuse environmental harm of recruiting third parties to accept the reduced political standing of transgender persons more broadly. Member Cousineau highlighted the contrast more clearly in her reasons for awarding monetary compensation to the victim for injury to dignity, feelings, and self-respect. When considering the impact on the victim of Whatcott’s flyers on the one hand, she described the victim’s generalized fear for her safety and concern that the flyers would foster an atmosphere of rejection of transgender persons that would encourage people to attack her because of her gender identity.[135] On the other hand, she also accepted that the victim’s plight was “degrading” and “humiliating” and caused her to “worry,” feel “hurt and angry” and “emotional,” and doubt her own worthiness to hold a public office—“this type of effect is a serious impact on a person’s dignity.”[136] The recruitment harm the flyers caused was therefore considered in tandem with the emotional damage to the victim in the justification for the Tribunal’s monetary award.[137]

We have argued that although the framework developed in Ward rejects the offence principle and accepts the harm principle, the Court’s account of harm is incomplete. We do not deny that preventing recruitment harm is a reason that can justify limiting freedom of expression. But other kinds of harm can do so, too. To raise awareness of this conceptual space, we have sketched an intermediate category of harm situated between recruitment harm and the more trivial taking of offence—roughly, non-trivial injury to a person’s psychological integrity—whose prevention can also supply the requisite justification. In various ways, Canadian law already recognizes this category. The Court ought to have done so in Ward by considering whether the discriminatory jokes directed at Gabriel fell within this category.

V. Objections and Replies

One possible objection to our argument is that if preventing harm other than recruitment harm is a potential justifying reason for limiting discriminatory expression, the jurisdiction of administrative tribunals would be unduly expanded. The Ward majority resisted an interpretation of the Quebec Charter that, it thought, would inappropriately expand the Human Rights Tribunal’s jurisdiction. While most of the statute’s prohibitions of discrimination pertain to deprivations a victim might experience within employment, housing, and the provision of services, section 10 applies to a deprivation of any statutory right, even if its protection is not restricted to those spheres. The section 4 safeguard of dignity usually grounds defamation claims before civil courts, which can be perpetrated through expression occurring outside employment, landlord-tenant, or service provider-consumer relationships. It can thus ground section 10 discriminatory expression claims before the Tribunal that are similarly unbounded. But the majority thought that this could allow the Tribunal to encroach too intrusively on freedom of expression. To contain that encroachment, such claims can be permitted only if they aim at preventing the recruitment harm of hate speech.[138]

Yet it is unclear why, unless a claimant establishes recruitment harm, a tribunal cannot have jurisdiction over an allegation of discriminatory expression occurring outside employment, housing, or services. In many contexts, the law sees fit to limit freedom of expression outside these spheres. Crimes like uttering threats need not occur within an employment relationship, nor do torts such as IIMS, but they limit freedom of expression to prohibit, inter alia, harms to victims’ psychological integrity. Holding that human rights tribunals cannot have jurisdiction over discriminatory expression outside the usual statutorily regulated spheres because this would overlap with the jurisdiction of civil courts over defamation presupposes exactly what we have argued is false—namely, that the harm caused by discriminatory expression is distinct from the harm caused by defamation. Because the harms are similar, any jurisdictional overlap is unobjectionable.

Moreover, the Supreme Court has recognized that, to be prohibited under antidiscrimination statutes, purely verbal harassment of an employee in a workplace does not only reflect an exercise of economic coercion by a person in a position of economic authority. It can also reflect the exercise of “gendered power” by men over women, or “racialized power” by white people over people of colour, and “[t]he exploitation of identity hierarchies to perpetrate discrimination against marginalized groups can be just as harmful to an employee as economic subordination.”[139] Preventing non-economic harms justifies extending harassment prohibitions beyond the employer-employee relationship to capture, for example, coworker harassment.[140] And like the harms of defamation, these harms are “not limited to the employment context” even if they are “exacerbated in the employment context where a complainant is particularly vulnerable” because, for example, employees often constitute a “captive audience” who are unable to avoid an employer’s expression.[141] The Court has accordingly rejected any strict dichotomy between the contexts of employment, housing, and services under human rights codes and discriminatory expression occurring outside these contexts—the codes capture all “contexts of vulnerability.”[142] In our view, given the non-economic harms that such expression can cause, the relationship between its speaker and victim essentially constitutes a context of vulnerability unto itself within the jurisdiction of human rights tribunals regardless of whether it is instantiated within a sphere of formal power imbalance such as employment.

A second objection might hold that there can be no viable legal mechanism for determining whether expression, once properly classified as discriminatory, has caused the type of psychological harm whose prevention we say gives a reason for limiting expressive freedom. The purported difficulty lies in distinguishing this harm from offence—a comparatively trivial affront to a person’s sensibilities by virtue of the impugned expression having a content that flouts conventional standards of appropriate interpersonal behaviour. If this difficulty is insoluble, so the objection goes, the prevention of psychological harm is nothing but a proxy for—or a slippery slope toward—preventing offence, which gives no reason for limiting freedom of expression.[143]

Until Ward, the Court of Appeal of Quebec had maintained that discriminatory expression contravenes the Quebec Charter when a reasonable person in the circumstances of the victim would not temper their reaction to it given their awareness of the need, in a pluralistic society, to tolerate forms of expression that contravene mainstream norms.[144] The Supreme Court majority stated that “the perception of a reasonable person targeted by the same words must be excluded”; a test that focuses on “the repugnant or offensive nature of the expression or on the emotional harm caused to the person” would undermine expressive freedom by censoring expression due to its merely offensive content.[145] The dissenting judges in Ward, by contrast, accepted the Court of Appeal’s reasonable person test.[146]

This test strikes us as potentially able to identify the type of (non-recruitment) psychological harm to which we draw attention. However, it must be refined to make more explicit the point that mere offence is incapable of justifying limits on freedom of expression. What is more, the Court of Appeal judgment from which the test derives involved a discrimination claim against an employer who called all his Asian employees into a meeting and berated them for being dirty because they were Chinese, and the court rested its conclusion that the employer’s comments were justifiably prohibited despite not causing the recruitment harm on factors unique to the employment context. For example, the employees were a captive audience, their employment was precarious, and there were already generally accepted limits on employers’ freedom of expression.[147] Although the category of psychic harm that we say justifies limiting free expression may be exacerbated by the unique economic vulnerabilities of an employee, it is not restricted to the employment context. Given these considerations, it seems more accurate to maintain that discriminatory expression may be justifiably prohibited if, as a matter separate from the economic hardship it visits on the victim in the employment context, it causes serious mental injury that rises above what a reasonable person in the circumstances of the victim would be ordinarily expected to tolerate in a pluralistic society, assuming the person is aware of the importance of protecting expression with offensive content.

If, in rejecting the Court of Appeal’s test, the Supreme Court majority in Ward thought that no kind of damage to psychological integrity ever gives a sufficient reason to justify limits on freedom of expression, again, this position is question-begging. As we have seen, Canadian law does regard such harm as sufficient in some circumstances, even if it does not occur in a context such as employment. The qualification, in our view, is that the damage must amount to a genuine, non-trivial, legally cognizable harm that rises above mere offence. What we endorse in the Court of Appeal’s judgment is the approach of identifying that harm and distinguishing it from mere offence by considering whether the reasonable person would place the damage on the offence or harm side of the divide.

This approach is modelled on the recent loosening of strictures surrounding compensation awards for mental injuries in negligence law. Historically, compensation for mental injuries for negligence victims was extremely limited out of skepticism that people may feign or exaggerate objectively trivial psychological disturbances. Tangible physical damage was seen as the paradigmatic form of compensable harm, so any compensable intangible mental injury had to be consequential on physical loss. The standard of proof of mental injury was also set very high. Plaintiffs had to show that they sustained a “recognizable psychiatric injury” according to medical diagnostic classification criteria. But recently that standard has been relaxed.[148] A plaintiff must show a form of objective psychological harm, as distinguished from mere subjective “psychological upset,” “disgust, anxiety, agitation or other mental states that fall short of injury.”[149] The harm must be “serious and prolonged and rise above the ordinary annoyances, anxieties and fears that people living in society routinely, if sometimes reluctantly, accept.”[150] In other words, the goal is to distinguish objective harm and subjective offence: “[M]ental injury is not proven by the existence of mere psychological upset.”[151]

The reasonable person test for identifying the type of psychological harm whose prevention can give weighty reasons for limiting discriminatory expression has the same goal. The dividing line between harm and offence may be vague and difficult to discern, but it does not follow that there is no difference between these concepts. Contemporary negligence law puts faith in triers of fact to distinguish between harm and offence based on the evidence before them and to make credibility determinations of those who claim to have suffered a harm that transcends offence.[152] Human rights adjudicators have the same capabilities. Ultimately, the rejection of the reasonable person test by the Ward majority underappreciates the capacity for fact finders under antidiscrimination statutes to manage psychological harm. By relegating all claims of emotional injury to the category of mere offence and equating the category of harm to recruitment harm, it undermines how the law attempts to avoid perpetuating the stigmatizing prejudice that mental injuries are trivial or that claims for their compensation are inherently unmeritorious.[153]

The final objection we shall consider concerns evidentiary challenges associated with proving that discriminatory expression causes psychological harm to individual victims whose prevention warrants limiting freedom of expression. Emmett Macfarlane, for example, has recently sought to expose methodological flaws in some of the empirical studies about the self-reported injurious emotional effects of discriminatory expression on religious or racial minority groups and members of the LGBTQ2S+ community.[154]

But even if these flaws are present, there are just as acute problems with proving the recruitment harm of hate speech, whose prevention is a good reason to limit free expression. When a person circulates a piece of propaganda that aims to instill in its audience a belief in the lower social standing of members of certain groups, it is exceedingly difficult to tell with scientific accuracy whether it is causally responsible for harming these groups by diminishing their assurance or confidence in their security against spillover harm throughout society. If it ever brings this outcome about, it likely does only by dispersing throughout the relevant society’s culture and combining, in multifarious ways, with other diffuse systemic and structural attitudes or beliefs held by third parties that already pollute the culture. Its causal path is neither linear nor easy to trace. Often, it may simply dissipate into the social atmosphere without achieving any recruitment at all. Macfarlane points out that for proponents of hate speech prohibitions, “[m]ore often than not, the harms of hate speech are asserted more than they are demonstrated, with even book-length analyses on the topic presenting virtually no empirical evidence to support their conclusions.”[155]

The challenges in proving recruitment harm have not deterred the Supreme Court from holding that prohibitions of hate speech are justified under section 1 of the Charter by the objective of preventing this harm. Recruitment harm is said to be so important to prevent that limits on free speech are justified if “common sense and experience” suggest that hateful expression will cause that harm without the precise causal link having to be scientifically established.[156] In our view, any problems in proving that discriminatory expression causes psychological harm to its individual victims should similarly not create an obstacle to justifying the prohibition of that expression under human rights codes. If anything, the empirical support for this harm seems to be superior to evidence demonstrating how hate speech causes recruitment harm. Assessing the credibility and reliability of individual claimants and witnesses is the bailiwick of triers of fact. Fact finders are better equipped to assess the nature of a harm in a particular case as opposed to arriving at an opinion rooted in competing social science evidence about broad and intangible harms pervading society. We have also claimed that there is a powerful intuition that, for instance, homophobic insults, racial slurs, or sexist threats can cause harm to victims.

Any suggestion that preventing psychological harm is somehow less important than preventing recruitment harm rests on the devaluation and denial of the reality of damage to mental health that modern negligence law, for example, attempts to correct. Instead of this skepticism, the law trusts triers of fact to use their common sense and experience to find whether negligent conduct has caused compensable mental injury without demanding scientific or medical certitude. Antidiscrimination law should be no different.[157]

Conclusion

In this paper, we have defended two conclusions. First, the majority in Ward endorsed a perpetrator-focused conception of how to define expression as discriminatory that is inconsistent with the standard victim-focused approach to defining discrimination in Canada. Second, although Ward held that the only reason to justifiably limit discriminatory expression is to prevent harm—not offence—its conception of harm was restricted to the recruitment harm of hate speech; we have argued that discriminatory expression may be legally prohibited if it causes other types of harms, particularly serious psychological harms to individual victims.