The Harm In Qwelane

This is the second brief dealing with the Constitutional Court’s decision in Qwelane v South African Human Rights Commission. This brief considers the issues that arise from the Court’s interpretation of the word “harm” in the Promotion of Equality and Prevention of Unfair Discrimination Act.
The Harm In Qwelane

The Constitutional Court’s judgment in Qwelane v South African Human Rights Commission[i] (“Qwelane”) cleared up many interpretive issues that arose from the wording of section 10 of the Promotion of Equality and Prevention of Unfair Discrimination Act[ii] (PEPUDA). This brief provides a critique of the judgment by exploring the issues raised by the Court’s interpretation of the word “harm” in PEPUDA.

The Link Between Hateful And Harmful

The wording of s 10 of PEPUDA does not contain identical wording to the prohibition of hate speech in s 16(2) of the Constitution. The Constitution defines hate speech as words which advocate hatred (based on specific grounds) and which 
constitute incitement to cause harm. In distinction, after the Court’s judgment, s10 of PEPUDA prohibits “...words... that could reasonably be construed to demonstrate a clear intention to be harmful or to incite harm and to promote or propagate hatred”.

Both sections consider that the promotion or advocacy of hatred is prohibited. In considering what qualifies as words which promote hatred, the Constitutional Court quoted approvingly from Canadian case law. In the Canadian law cited by the Court, promoting hatred means to “instil detestation, enmity, ill-will and malevolence in another”[iii]. The Court accepted that hate speech does not include merely offensive speech but that the speech must demonstrate “abhorrence, delegitimisation and rejection”[iv] against someone based on their membership in a group.[v] Such speech should “abuse, denigrate or delegitimise [targeted persons and] render them lawless, dangerous, unworthy or unacceptable in the eyes of the audience.”[vi]

Next, the Court considered what “harmful” means. One should note that where the Constitution prohibits words which incite harm, PEPUDA also prohibits words which are harmful. Various meanings of the word harm were advanced, but the Court ultimately accepted that “harmful” can mean “deep emotional and psychological harm that severely undermines the dignity of the targeted group”.[vii]

With this definition of harmful, it is unclear how speech which promotes hatred against members of a group (that is, speech which seeks to instil extreme detestation and delegitimisation of persons because they are a part of a group) will NOT be harmful. It seems that such speech will always cause emotional and psychological harm and violate the dignity of a member of the group and the group itself. In other words, it appears that speech which promotes or propagates hatred will always be harmful. The court removed the word hurtful from s10 because of its superfluity since it “may be so that harmful communication is always hurtful”.[viii] Including emotional and psychological harm may result in the word “harmful” being superfluous too, since speech which promotes hatred will be (psychologically and emotionally) harmful.

It is sensible that “harm” should be interpreted as more than physical harm and extend to social, economic, political or legal harm. However, psychological and emotional harm which impairs dignity seems a step too far, since all one would actually need to prove is that there has been expression that promotes or propagates hatred, since psychological and/or emotional harm which impairs dignity would follow as a natural consequence.

Do We Need To Prohibit Harmful Speech?

One can appreciate why the court interpreted “harm” in the way it did. It noted emphatically that prohibitions of hate speech seek to prevent “undermin[ing] the dignity and humanity of the target group”. [ix] However, as the Supreme Court of Appeal noted:

We can all agree that it is important to protect the dignity of all our citizens. Equally, we must agree, given our history, that freedom of expression must also be prized. That does not mean that hate speech beyond the provisions of s 16 cannot be proscribed. It must, however, be tailored so as to comply with constitutional prescripts and it must survive a justification analysis.[x]

Even though the inclusion of the words “be harmful” in PEPUDA is clearly a widening of the limitation of freedom of expression beyond what the Constitution contemplates, no justification analysis was carried out by the Constitutional Court.

Moreover, international law does not demand that States prohibit speech that is “harmful”. Article 20 of the International Covenant on Civil and Political Rights[xi] provides that ‘[a]ny advocacy of national, racial or religious hatred that constitutes incitement to discrimination, hostility or violence shall be prohibited by law’. States may go further and prohibit hateful speech that does not constitute incitement, but as the Special Rapporteur noted in 2019, this would constitute a limitation of freedom of expression and must meet the conditions of legality, necessity and proportionality, and legitimacy.[xii] It is unfortunate then that the Court did not tackle whether prohibiting speech that does not constitute incitement is justifiable either in terms of the Constitution or international law.

If speech which does not incite harm must be restricted under hate speech prohibitions, it is arguable that there should be a different remedy than that imposed for speech which does incite harm. As the Court noted, the jurisprudence of the European Court of Human Rights (ECtHR) distinguishes between the “gravest forms of hate speech” which includes speech which incites harm, and “less grave forms of hate speech” which covers speech which does not incite harm. A distinction would not be necessary if both types of speech were equally heinous, and it seems sensible that the remedy should be commensurate to the graveness of the hate speech. 

Offensive Speech

The Court was at pains to clarify that offensive speech does not qualify as hate speech. It quoted from its earlier judgement in Islamic Unity Convention v Independent Broadcasting Authority that “[f]reedom of expression is applicable... also to those [expressions] that offend, shock or disturb the state or any sector of the population”.[xiii]

Hate speech, said the Court, was expression which shows “extreme detestation and vilification which risks provoking discriminatory activities against that group”.[xiv] The distinction between hate speech and offensive speech is murky and may be lead to conflicting decisions. For example, is referring to black persons as monkeys, as Penny Sparrow did in 2016, offensive and disgusting (but not hate speech), or is it extreme detestation which risks causing discrimination against black persons (which is hate speech)?[xv]

It is arguable that speech which may offend any sector of the population and should be protected (per Islamic Unity) will now be categorised as hate speech. This is suggested from the Court citing ECtHR cases where it had been said that “insulting, holding up to ridicule or slandering specific groups of the population can be sufficient [to qualify as hate speech]”. [xvi]

In 2010, constitutional law Professor Pierre De Vos argued that the element of hurtfulness in the definition of hate speech would lead to much offensive speech being unlawful such that:

[I]t may potentially be unlawful to refer to homosexuals as “moffies” ... or to tell a person that he or she is a pervert and will burn in hell because he or she is gay. Calling someone a “kaffir” or a “Boer” could constitute hate speech as well. When a religious person tells an atheist that he or she has been condemned by God to burn in hell this could also potentially constitute hate speech.

...Telling a polygamist that polygamy was a backward and oppressive cultural practice and that he was an oppressor of woman for indulging in polygamy could constitute hate speech. Telling someone that members from his or her ethnic group are lazy, cunning, dishonest or greedy could constitute hate speech.

Though the Court removed the word hurtful, it only did so because it was redundant in that both “hurtful” and “harmful” could mean emotional and psychological harm.[xvii] What the court has done is to incorporate a problematic aspect (hurtful) into another (harmful), and as a result, the hypotheticals considered by Professor De Vos may reasonably occur.


This brief does not contend that nothing should be done against speech which is hateful but which does not incite harm. Rather, it merely highlights that the Constitutional Court’s treatment of this type of speech leads to legal absurdities, will cause problems in future hate speech adjudication, and that the Court did not justify its restriction in line with the Constitution. It further contends that if such speech must be included in hate speech prohibitions, a lighter remedy may be more appropriate. Apart from prohibiting such speech, states should also, in line with Human Rights Council resolution 16/18,[xviii] publicly condemn expressions of prejudice, offer education and training to counter intolerance, and create community projects to combat the causes of discrimination. Such efforts may be more valuable in eradicating prejudicial hateful expression than legal solutions.

Media Enquiries
Zeenat Emmamally

[i] (CCT 13/20) [2021] ZACC 22 (31 July 2021). 

[ii] Act 4 of 2000.

[iii] Note i, fn 100.

[iv] Ibid para 103

[v] Ibid para 129

[vi] Ibid para 80.

[vii] Ibid para 154.

[viii] Ibid para 155

[ix] Ibid para 130.

[x]Qwelane v South African Human Rights Commission 2020 (2) SA 124 (SCA) para 85.

[xi] International Covenant on Civil and Political Rights, 16 December 1966 (ICCPR). The ICCPR was signed and ratified by South Africa in 1994 and 1998, respectively.

[xii] Thematic report by Special Rapporteur on the promotion and protection of the right to freedom of opinion and expression to the General Assembly, on online hate speech (A/74/486) (9 October 2019), para 20.

[xiii] 2002 (4) SA 294 (CC) at para 26.

[xiv] Note i, para 81.

[xv]ANC v Sparrow (01/16) [2016] ZAEQC 1 (10 June 2016).

[xvi] Note i, para 146 and fn 203.

[xvii] Ibid para 155.

[xviii] Resolution adopted by the Human Rights Council 16/18 “Combating intolerance, negative stereotyping and stigmatization of, and discrimination, incitement to violence and violence against, persons based on religion or belief” (12 April 2011) A/HRC/RES/16/18.