January 10, 2020

Some Observations on Hate Speech in Armed Conflicts

By: James Hendry

In proceedings instituted before the International Court of Justice, The Gambia alleged that Myanmar’s treatment of the Rohingya people breached the Genocide Convention, relying in part on the expressions of hate by Myanmar as evidence of the special intent required. Professor Rikhof recently wrote an article about hate speech in international criminal law in this Journal The-ICC-and-Hate-Speech. In this article I will offer some thoughts on the hate speech in The Gambia’s pleadings, the arguments before the International Court of Justice (ICJ) and the only conviction this year at the International Criminal Court (ICC) in the Ntaganda case from a Canadian legal perspective.

Hate speech at the Supreme Court of Canada

In Canada, the prohibition of hate speech in the Criminal Code and in human rights legislation at both the federal and provincial levels has been challenged on a number of occasions as an unconstitutional infringement of freedom of expression. The Canadian approach to the interpretation of these provisions has been influenced by both domestic and international concerns. 

In Taylor, an individual found civilly liable for a breach of a provision of the Canadian Human Rights Act that prohibited telephonic communication of hatred and contempt for members of a protected group, challenged s. 13, now repealed, for breaching his right to freedom of expression for his publishing a telephone number that would allow the caller to hear a message denigrating the Jewish race and religion among other groups. A bare majority of the Supreme Court held the provision breached freedom of expression but was a reasonable limit on that right. The constitutional justification required the government to show a pressing and substantial objective for an infringement of the right and that it was proportional to the objective. A bare majority held that Parliament’s objective was to prohibit a serious harm to equality, which was recognized in the Canadian Charter of Rights and Freedoms and Canada’s international commitments. It was a proportionate limit because it prohibited speech that undermined equality and it was limited in its infringement on expression by being specifically directed to hatred and contempt targeted at a protected group that conveyed strong feelings of detestation, calumny and vilification. The Court had previously held that intent was not a necessary element to prove discrimination. The requirement in s. 13 of proof that the messages be repeated further limited the provision’s scope to public dissemination of hate and so would not interfere with private conversation. The majority thought that hate expression has little or no value anyway. The prohibition did not exceed the value of the harm prevented because human rights legislation made this a remedial civil matter and not a crime. The dissenting justices emphasized the breadth of the prohibition, its subjectivity, the absence of defences, the lack of a requirement of actual harm and that it might create a ‘chilling effect’ for other expression that might be part of democratic debate of some sort.

The accused in R v. Keegstra challenged s. 319(2) of the Criminal Code prohibiting statements that wilfully promote hatred against any identifiable group, alleging it breached the protection of freedom of expression in s. 2(b) of the Canadian Charter of Rights and Freedoms. Once again, a bare majority of the Supreme Court of Canada held that, though s. 319(2) infringed this right, it was justified under s. 1 of the Charter as a reasonable limit prescribed by law consistent with the values of a free and democratic society. The majority held that preventing harm caused by hate propaganda was a pressing and substantial purpose drawing on domestic and international sources. Domestically, studies showed there had been a growth of hate propaganda in Canada that degraded vulnerable groups in society deterring them from participating in Canadian society at large and embedding hatred for the target group in minds that might even rationally reject it giving it a credence that could lead to social discord and violence. The majority also referred to the values in international law mirrored in the values underlying the Charter, the most significant being that of discrimination. The majority referred to the International Convention on the Elimination of All Forms of Racial Discrimination (CERD) which requires at article 4(a) that States Party “Shall declare an offence punishable by law all dissemination of ideas based on racial superiority or hatred, incitement to racial discrimination, as well as all acts of violence or incitement to such acts against any race or group of persons of another colour or ethnic origin…” The majority also referred to the International Covenant on Civil and Political Rights (ICCPR) which provides at article 20(2) “Any advocacy of national, racial or religious hatred that constitutes incitement to discrimination, hostility or violence shall be prohibited by law.” The majority noted that the freedom of expression protection in the ICCPR and CERD did not extend to prohibiting hate advocacy. First, it was agreed among States Party to CERD that this was the case (citing the Study on the Implementation of Article 4 of the International Convention on the Elimination of All Forms of Racial Discrimination, prepared by Special Rapporteur Mr. José D. Inglés, A/CONF. 119/10, May 18, 1983, para. 108). Second, in a communication under the Optional Protocol to the International Covenant on Civil and Political Rights John Ross Taylor, who had been found to breach s. 13(1) of the Canadian Human Rights Act, and whose challenge to the prohibition of the communication of hate messages against identifiable groups, noted above, violated article 19.  The Committee rejected this argument holding that it was incompatible with the provisions of the ICCPR and in particular with article 20: “…the opinions which Mr. [Taylor] seeks to disseminate through the telephone system clearly constitute the advocacy of racial or religious hatred which Canada has an obligation under article 20(2) of the Covenant to prohibit” [quoted by the majority at pp. 752-3].  The majority also referred to other provisions of the Charter that constitutionalized the right to equality and the interpretive value of Canada’s multi-cultural heritage. The impugned provision of the Criminal Code was a proportionate response to the objective in that it protected vulnerable groups, fostered social cohesion and expressed societal reprobation of hate messages. It minimally impaired rights through the definitional limit to hate and contempt, it required proof of the mens rea of wilfulness and excluded private communications. The challenged provision did not advance the truth and aspirations of Canadians through expression or promote democracy and so did not outweigh the value of the prohibited expression. The dissenting justices held that the provision might hamper the creative and beneficial expression of ideas. It was too subjective and did not require proof of actual harm. Finally, it was outweighed by the value of the marketplace of ideas, open democratic debate and self-actualization and freedom of members of society.

By bare majority, the Supreme Court of Canada then struck down a now-repealed provision in the Criminal Code that prohibited the wilful publication of a false statement likely to cause public mischief arising from Ernst Zundel’s publication of a Holocaust denial pamphlet alleging that it was part of a worldwide Jewish conspiracy. The majority this time held that there was an unjustified breach of the freedom of expression. It held that there was no social problem here creating a pressing and substantial objective. It was not a proportional response to the objective because the determination of ‘false news’ was highly subjective, the provision could extend to any controversial statements thereby chilling a great deal of expression.

In Mugesera, the Minister of Citizenship and Immigration started proceedings for deportation under the Immigration Act against a prominent member of a Hutu political party who had addressed a large crowd in Rwanda at a time of escalating ethnic violence and massacres of Tutsi, and who fled to Canada and became a permanent resident, on the basis that his speech incited hatred, murder and genocide and constituted a crime against humanity. Speaking of Tutsis, Mugasera said in his speech, quoted at para. 76 from the speech annexed as Appendix III to the Judgment:

[TRANSLATION] So in order to conclude, I would remind you of all the important things I have just spoken to you about: the most essential is that we should not allow ourselves to be invaded, lest the very persons who are collapsing take away some of you. Do not be afraid, know that anyone whose neck you do not cut is the one who will cut your neck. Let me tell you, these people should begin leaving while there is still time and go and live with their people, or even go to the “Inyenzis”[cockroaches], instead of living among us and keeping their guns, so that when we are asleep they can shoot us. Let them pack their bags, let them get going, so that no one will return here to talk and no one will bring scraps claiming to be flags!

The Court unanimously upheld the decision that Mugasera had incited murder and incited genocide contrary to s. 318 of the Criminal Code. His speech conveyed to the audience in violent terms that they should either exterminate the Tutsi or they would be exterminated by them, referring to them as ‘cockroaches’, intending to incite them to murder along with the specific intent to incite genocide. His speech was delivered in the context of the ethnic massacres already underway. Mugasera had also breached s. 319(2) of the Criminal Code because he intended to incite hatred and violence toward the Tutsi. The Court also held that hate speech may constitute the crime against humanity of persecution within the definition of s. 7(3.76) of the Criminal Code (now the Crimes Against Humanity and War Crimes Act) where it incites hatred and violence and thereby furthers a widespread, systemic attack on the target group of civilians with discriminatory intent.  The Court agreed with the findings that there were reasonable grounds to believe Mugasera had incited hatred and violence furthering the systemic attacks against the Tutsi civilian population with discriminatory intent based on ethnicity, knowing that the speech would further the already ongoing attack. 

This author’s view is that Mugasera’s unanimous statement “The intention of Parliament was to prevent the risk of serious harm and not merely to target actual harm caused.  The risk of hatred caused by hate propaganda is very real.  This is the harm that justifies prosecuting individuals under this section of the Criminal Code…” shows the point at which the balance in the Court changed on hate propaganda.

Finally, in Whatcott, the Court unanimously agreed that provincial human rights legislation prohibiting hate was constitutional. The issue was whether statements in public flyers such as those equating homosexuals with carriers of disease, sex addicts, pedophiles and predators who would proselytize vulnerable children and cause their premature death breached a provincial anti-hate prohibition. Though the impugned provision breached the freedom of expression, the Court unanimously held that it was a reasonable limit on that right. The legislature had a pressing and substantial objective in the reduction of the harms of discrimination. The Court wrote at paras. 74 and 75:

Hate speech, therefore, rises beyond causing emotional distress to individual group members.  It can have a 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 observed by this Court in Mugesera v. Canada (Minister of Citizenship and Immigration), [2005] 2 S.C.R. 100, at para. 147, the findings in Keegstra suggest “that hate speech always denies fundamental rights”.  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: see Taylor and Keegstra.

Hate speech is not only used to justify restrictions or attacks on the rights of protected groups on prohibited grounds. As noted by Dickson C.J., at p. 763 of Keegstra, hate propaganda opposes the targeted group’s ability to find self-fulfillment by articulating their thoughts and ideas. It impacts on that group’s ability to respond to the substantive ideas under debate, thereby placing a serious barrier to their full participation in our democracy. Indeed, a particularly insidious aspect of hate speech is that it acts to cut off any path of reply by the group under attack.  It does this not only by attempting to marginalize the group so that their reply will be ignored: it also forces the group to argue for their basic humanity or social standing, as a precondition to participating in the deliberative aspects of our democracy.

The prohibition was proportional to this objective. It was rationally connected to it by requiring of the decision-maker to objectively assess whether the expression sought to marginalize the target group in the eyes of the majority through extreme feelings of hatred. It minimally impaired the right of expression. It was far from the values that freedom of expression protects. Hate speech was not adequately dealt with in the marketplace of political ideas because it seeks to exclude the targeted group from the market. 

Canadian hate propaganda law evolved from a concerns that narrowly balanced discrimination through the denigration of the target group with the marketplace of ideas to a post-Mugasera concern for not only the damage to equality in society, but the potential for social unrest and even violence, and the fact that it could effectively silence the voice of the target group in a democracy. 

The Gambia v. Myanmar 

In its Application Instituting Proceedings and Request for Provisional Measures, The Gambia alleged Myanmar engaged in genocide by killing, causing serious bodily and mental harm, inflicting conditions calculated to bring about physical destruction, imposing measures to prevent births and forcible transfers all done with the special intent to destroy the Rohingya ethnic, racial and religious group in whole or part (para. 2). The Gambia argued that the UN Fact-Finding Mission Reports documented atrocities ‘similar in nature, gravity and scope to those that have allowed genocidal intent to be established in other contexts’ (para. 21). It also argued that Myanmar’s deprivation of the legal rights of the Rohingya provided some evidence of genocidal intent as well (para. 32). The Gambia also alleged that some of the evidence of genocidal intent was to be found in Myanmar’s hate propaganda that portrayed the Rohingya as a threat to the Buddhist character of the country and dehumanized them (paras. 37, 40). The President of Myanmar referred to Rohingya terrorists alleging that ‘our ethnic people are in constant fear in their own land’ on Facebook (para. 44). The Gambia also noted that as the ‘clearance operations’ started against the Rohingya, the Tatmadaw (Myanmar military) members called their victims: ‘people from Bangladesh’, ‘Bengalis’, using the racial slur ‘Kalars’ (para. 48). The argument for genocidal intent proceeded from deductions from the actions of the Tatmadaw to words of hate which required no interpretation to convey the dehumanization and alleged fear of the Rohingya from the highest authority in Myanmar to those members of the military and others who carried out the atrocities.  

The Agent for The Gambia commenced his address to the Court on the first day by saying that genocide is ‘preceded by a history of suspicion, mistrust and hateful propaganda that dehumanizes the other…’ (p. 17). Counsel for Gambia argued that the genocidal acts incubated in hate speech on Facebook and Twitter dehumanizing the Rohingya using every type of epithet from illegal immigrant to their being animals and creating an existential threat to Burmese racial purity (p. 23) drawn from the UN Fact-Finding Missions led to an inference of Myanmar’s genocidal intent (pp. 23, 26-7). Its second counsel placed even greater emphasis on Myanmar’s dehumanizing and fear-mongering hate propaganda against the Rohingya (pp. 28-30). Another counsel pointed out that the UN reports concluded that State-mandated segregation fostered an environment for dehumanizing hate campaigns (p. 39). Though the last counsel speaking for the Rohingya noted that the Foreign Ministry of Myanmar explicitly refused to recognize the Rohingya as a group and has ‘overtly sought to dehumanize the group and its members’, he sought no specific provisional measures to stem the tide of hate (p. 68).  Myanmar's response on the second day was that The Gambia had failed to establish the necessary standard of special intent for genocide to indicate provisional measures because that special intent was not the only explanation for the pattern of atrocities committed as required by the cases: they were also consistent with forced displacement (p. 38-40). There was no submission on the issue of the expression of hatred. During the first half of the final day of hearings, counsel for The Gambia noted that Myanmar had not denied the hate campaign but that its Agent had emphasized it by saying “Hate narratives are not simply confined to hate speech – language that contributes to extreme polarization also amounts to hate narratives” (p. 13).   During the second half of the final day counsel for Myanmar dealt with the issue of hate expression, focusing on the whether the expression ‘clearance operations’ alleged in the 2018 short UN report to have been contained in a Facebook post of the Tatmadaw commander and which formed the basis of the Fact-Finding Mission’s ‘damning evidence of genocidal intent’ was not accurate (pp. 24-5).  Further, he argued the Facebook post had many meanings and was certainly not hate speech in the context. He noted that the many hateful Twitter messages circulating in Myanmar were the acts of individuals and that similar hate messages of various kinds against various groups would be circulating everyday in The Hague that day and were as such not evidence of the State’s intent (p. 26).

Hate speech triggers for killing

Psychologist David Grossman identifies factors that enable a soldier to overcome a natural moral reluctance to kill a fellow human being. In addition to institutional factors involving authority of the order to kill, membership in an organization supporting the killing, the predisposition of the soldier to kill by training and temperament, the ‘relevance’ of the victim, Grossman notes that ‘distance’ from the victims is a vital factor. Distance is created by social beliefs that the target group of people is less than human, cultural belief in racial and ethnic differences that allow the soldier to dehumanize the victim, moral beliefs involving a sense of moral superiority and revenge and mechanical distance allowing the soldier to kill from a physical distance without fully appreciating the humanity of the victim (On Killing, 2009, Sec. IV, Ch. 6). Much the same distance is required to compel soldiers to commit atrocities (On Killing, p. 212).

Clearly hate propaganda aimed at dehumanizing the members of a group and alleging that they are a danger to the soldier’s own group serve to create the distance needed to overcome a soldier’s moral reservations against killing and committing atrocities.

Hate speech in the Ntaganda conviction

Hate propaganda can be found in the Ntaganda conviction decision where he was convicted as a direct perpetrator of murder as a crime against humanity and war crime and persecution as a crime against humanity and also as an indirect co-perpetrator on 18 charges including murder and persecution. While the stated ambition of the UPC/FPLC was to defend the population of Ituri, the Chamber found that its aims were contradicted by its operations against the Lendu, that recruits were taught that the Lendu were the enemy and a song inciting recruits to attack the Lendu was recited during training (para. 687). Ntaganda spoke at a rally where he called on parents to send their children for military training by the UPC/FPLC to defend themselves against the Lendu (para. 357). The expression ‘kupiga na kuchaji’ was commonly used within the UPC/FPLC and was understood by its soldiers to mean kill all the Lendu which they did as instructed according to a policy that was ‘actively promoted’ (paras. 687-689). In relation to the charge of persecution, the Chamber said that it would evaluate the alleged status of a persecuted group in the political, social and cultural context and take into account not only the objective factors proving discrimination but the subjective perception of belonging of both perpetrator and victim (para. 1010). The Chamber recalled testimony that UPC/FLPC soldiers said in the course of their operations “Lendu are useless wild animals and we can do with them anything we want. They are not humans; and that they would ‘exterminate’ the Lendu within three days.” The same witness was also told by a commander that she and others were not human beings, but beasts, animals and hostages making her feel as if she were dead (paras. 599, 628, 1021). Hate expression was sufficiently important to the UPC/FLPC that a soldier killed a non-Lendu for singing an anti-Hema song (the ethnic group of the perpetrators) (para. 998).

The United Nations has promulgated a strategy and action plan on hate speech in recognition of the growing problem posed by this kind of expression.  


Hate speech directed at dehumanizing the ‘enemy’ group and alleging their malicious intentions towards the speaker’s group are common in conflicts. They may well be the most dangerous forms of hate propaganda as some experts see them as creating the kind of distance between antagonists that enables soldiers and others affiliated with them to overcome the natural propensity of people not to kill one another. Perhaps hate propaganda varies in the amount of distance it creates between perpetrator and victim, enabling stronger inferences of intent from specific forms of hate propaganda.

Please cite this article as James Hendry, “Some Observations on Hate Speech in Armed Conflicts” (2020), 4 PKI Global Justice Journal 1.

About the author

James HendryJames Hendry is a Co-Editor-in-Chief of the PKI Global Justice Journal. He served as counsel to the Canadian Human Rights Commission before joining the Department of Justice in 1989. He was General Counsel in the Human Rights Law Section at the DOJ until retirement in 2011, working in civil Charter social policy review, specializing in equality rights, human rights legislation, and human rights act design. He has also published extensively on Canadian and comparative constitutional issues and has lectured in Canada, Spain, South Africa, the United States and Hong Kong.

Image: Flickr, ‘No violence, no hate speech’, John S. Quarterman, Nov 9/11, Creative Commons License Attribution 2.0 Generic, (CC BY 2.0).