Shoot the Boer
shoot shoot
Shoot the Boer
shoot shoot
Shoot the Boer
shoot shoot
Shoot the Boer
shoot shoot These dogs are raping
shoot shoot
ay iyeah
shoot shoot
(Traditional version of “Dubul' ibhunu")
Introduction
South Africa has clearly, and quite correctly, accepted that in addition to “sticks and stones”, words can also harm people. Our legislation, case law and general societal living reflects a robust defence of people’s dignity, or efforts to try and protect respectful engagement. Our popular media reinforces these achievements, we have well-developed defamation, employment and general dignity laws, and we have more than one organization appointed to protect the harm that can arise from the words we use.
Given our history, we fully understand, maybe more than most other modern countries, the scars and damage that can flow from words. Outside of our own experience, and fully in accordance with those lessons, we find that conflict studies and a range of interlinked disciplines such as psychology, social sciences, law, and so on are all, along their own routes, arriving at the same conclusion: words matter.
Scope of our inquiry, and premise of the article
My article flows from current consulting work that I am doing for a European environment, but I want to focus my arguments of the harm done by certain types of song and rhetoric around the “Kill the boer” song. The song, and its recent public litigious history, is well-known in South Africa, and it contains most of the relevant components of the topic as a conflict study.
As such it is the perfect lens through which to study the larger question of the role of these slogans and songs in a conflict environment. Hopefully, the article then works either as a discussion on the “Kill the boer” song, and / or the wider conflict implications of certain songs and slogans on existing conflicts. The litigation thus far has, in each instance, had a limited scope, as is expected from litigation.
This article approaches these events, and the ongoing dispute, from a specialized conflict management perspective, and then leaves the reader to conclude as to whether allowing such conduct is responsible and ethical leadership, based on evidence, research and studies.
South African case law
Seeing that this is not a legal argument, I will very briefly record the litigation thus far, simply as a comparative background and logical point of departure. As full disclosure, and for the reasons we will examine herein, I believe the SCA decision to be unfortunate, and unintentionally harmful of our country’s best interest as seen from a conflict resolution perspective.
The case law and related public debates are all available, and I will not spend much time on repeating that process, other than a brief reminder of the events up to here.
The litigation starts effectively in 2011, with Afriforum suing Mr. Julius Malema (then the ANC Youth League leader) and the ANC in the Equality Court. This court finds the singing of the song to constitute hate speech, and interdicts Malema and the ANC from singing it publicly or privately, emphasizing the song’s potential to harm the dignity of the target group, and to incite harm in a post-apartheid South Africa.
In 2020 Afriforum brought a fresh complaint against Malema, the EFF and MP Dr. Mbuyiseni Ndlozi in the Gauteng Equality Court, for multiple instances of singing the song (between 2016 and 2019), now sung often with variations such as “burn the boers”. In 2022 Judge Molahlehi dismissed the application, finding that the song was not hate speech when viewed in the political context of the EFF's land reform rhetoric; it was not meant literally, lacked evidence of inciting actual harm/farm murders, and was protected speech. We will return to these findings below.
In May of 2024 the Supreme Court of Appeal via Saldulker AJ upheld the Equality Court ruling, emphasizing the context again, and finding that a “reasonably well-informed person” would see it as a political expression, and not a literal call to violence. The evidence will force us back to those words later on in the article.
In March 2025 the Constitutional Court dismissed Afriforum’s application for leave to appeal, on the basis that it had no reasonable prospect of success. The heart of this legal debate, the SCA decision, can be found at AfriForum v Economic Freedom Fighters and Others (1105/2022) [2024] ZASCA 82; [2024] 3 All SA 319 (SCA); 2024 (10) BCLR 1275 (SCA); 2024 (6) SA 1 (SCA) (28 May 2024)
This, for many, concluded the argument in favour of its proponents, as being protected political speech in the context of which it is used. We will examine that conclusion below.
Comparative studies
For ease of reading I will simply make the summarized point(s) of a particular study, and provide the links to the reference / study below a particular point or finding. These are representative examples, and in some instances meta-studies, and I am of course prepared to provide further references and links where reasonably required.
It would appear as if the litigation we briefly discussed above were poorly served by its access to various conflict studies dealing very directly with the topic at hand.
Let’s look at a few of the most important of these items of evidence.
(i) One such series of experiments shows that even mildly violent political metaphors (eg narratives referencing slogans such as “fighting for our future”) increased support for political violence among participants with aggressive personalities, with little effect on low-aggression individuals. I would think that we have a few “aggressive personalities” around.
- study by political communications Prof Nathan Kalmoe of Louisiana State University https://www.politico.com/magazine/story/2018/10/30/yes-political-rhetoric-can-incite-violence-222019/ )
(ii) A 2025 study (showing a 184 effect sizes) clearly showing situations where partisan elites threatening harm or violence, or violent rhetoric from co-partisans, increases support for political violence. This is exactly the norm for most of the settings in which the song is sung, and then carried out over social media.
- Taegyoon Kim article Springer Nature article, at https://link.springer.com/article/10.1007/s11109-025-10025-7
(iii) In high conflict settings (these experiments were conducted in Israel and India), “fighting” words combined with outgroup references boosted support for harmful policies towards that outgroup, even among less prejudiced or less aggressive people. Again, this is exactly the standard settings where these speeches and songs are sung.
- https://law-journals-books.vlex.com/vid/violent-rhetoric-in-protracted-874213342
(iv) The use of dehumanizing language (referring to the other as vermin, cockroaches or invaders) has a moderate association with aggression. A 2025 meta-analysis of 184 effect sizes dehumanization linked to aggression (r= 0.34 overall, stronger for blatant forms at r = 0.39)
Dehumanization and aggression: A meta-analysis - ScienceDirect
(v) The use of violent song lyrics show similar patterns, with the 2025 meta-analysis of 35 linked studies linking exposure to violent lyrics to higher aggressive behaviour, cognition and affect, including an increase in aggressive actions. As the study shows, prosocial lyrics can reduce aggression. https://www.hsaj.org/articles/22935
(vi) Music and songs in political contexts and environments can emotionally prime and normalize hostility, trigger narratives involving harm and violence, and aid radicalization. (The same article as referenced in (v) above)
(vii) Observational data also links this type of rhetoric to downstream harms, where exposure to a political leader’s tweets during protests predicted surges in violence severity, duration and weapon use. https://news.northwestern.edu/stories/2024/11/new-northwestern-research-links-social-media-and-growing-levels-of-political-violence
(viii) Broader media exposure of events such as we are discussing show positive association with violent radicalization in youth, with the multilevel meta-analysis showing r = 0.12. https://www.sciencedirect.com/science/article/pii/S1359178926000169
(ix) Violent rhetoric targeting officials or groups has risen sharply in some periods, often in feedback loops with real events – this is exactly what can be observed with this song and other rhetoric, sung at big public events, and then amplified by social media and other media attention.
(x) The causal effect between alleged cause and effect is sometimes disputed by sceptics, despite very clear evidence supporting the conclusions of such causal links. One of the clearest examples can be found in the 1994 Rwandan genocide, where village level data on radio reception for the RTLM, broadcasting dehumanizing calls to violence against Tutsis, mixing news, talk and songs in a sustained narrative, showed significant cause and effect links, with one major study attributed ~51 000 perpetrators, or about 10% of the total violence, to the broadcasts. Tellingly, the study also shows how the education levels of listeners played a meaningful role in the receptivity and results of these narratives in being linked to subsequent violence. https://www.hks.harvard.edu/publications/propaganda-and-conflict-evidence-rwandan-genocide
(xi) This type of study, and the causal linking of conflict cause and effect, goes back even to Nazi methods, where the denial of humanity, and narratives enabling violence which included the use of music, songs and slogans all have a very clear influence on such violent propensities and behaviour. https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0274957
(Some of these studies overlap, with the various different findings emphasized with different sources above)
How does this evidence fit in with the legal approach taken until now?
What distinguishes this evidence from the legal arguments? Critics and sceptics of these studies point to a range of caveats and cautions that we should take into account when assessing this debate. Let’s do so.
These studies, and simple human experience, show that most people so exposed to these songs and narratives do not commit violence as a result thereof. Good to hear. How many do we need to be so influenced before it becomes a valid argument? “Most” people is a low bar in conflict with high stakes such as this, especially when the solution is so easily achieved.
The studies and actual conflicts observed show that the education levels of participating audiences play a role. Again, easily conceded but how does that help as a defence against the charge that these songs should not be used at all? Are the songs only sung at gatherings of intellectuals? As our case law shows, context is an important perceived defence. What exactly is that context? As the case studies show, context can be an amplifier of these violent tendencies and events. Context can be as much of a conflict accelerant as it is a dampener of moods and receptivity.
It is also quite correct, and can be marked as common cause, that receptivity to these messages, and the resultant violence, is mostly exacerbated, even enabled, by underlying tensions and conflict causes such as unemployment, poverty, inequality and cynicism with democratic processes. How exactly is that a defence for the singing of this song in a South African environment and context? It all but concedes the tremendously volatile environment into which these messages are launched.
The court case(s) made much of the free-speech trade-off, the balance which democratic societies are obliged to seek between free speech and harm. Again, no argument there. That balancing exercise, in the text of the court’s own decision, seems to however not have had the benefit of the weight of the evidence we looked at. I believe that, even on the evidence that the courts did have access to at the time, the balance of interests in a free speech debate did not favour the singing a song that is clear on the face of it. It is demeaning to insist that it is a culturally fair reflection of the parties’ past, present or future. That context, that culture, can be wonderfully expressed, preserved and protected without the use of such rhetoric.
When we add the weight of the evidence we examined, the argument in favour of the use and protection of the song simply collapses on its own terms. The arguments in favour of retaining the public use of the song become even more unconvincing when we note that causality between cause and effect is not required in this type of litigation (cf par. 62 SCA). The requirements of the “reasonable listener” (par. 61 SCA) reads like a remarkably naïve approach when we have the abovementioned studies at our disposal.
The evidence that the court did assess (paras. 68 and 69 SCA) was, with all due respect, completely inadequate and out of line with the aims and findings that should have been placed before the court. The evidence that was found acceptable (for instance paras 96 and 97 SCA) are remarkably lenient, and again, seemingly unaware of the conflict studies. The correct answers are not arrived at because the correct questions are not asked. This is the case even when we remember that the cases themselves, in every instance, took a narrowly defined view to pronounce on, as court cases should. The clear, evidenced harm being done by allowing the song clearly cannot be defended, even in the narrow confines of a free speech debate.
Additional conflict management arguments
Where the impact and words of a song or narrative do not lead to the increased risks and harm dealt with in the evidence we considered earlier, the free speech arguments are well-protected in our Constitution, our case law and our lived everyday realities. When Jacob Zuma sings his signature song, Umshini Wami, no serious concerns are raised when we hear:
Please bring me my machine (gun)
Please bring me my machine (gun)
You are holding me back / pulling me back
My machine (gun), please bring me my machine (gun)
The song does not explicitly call for harm to a specific group, and intent must be inferred. It attracts much less attention and concern, and even many of its critics regard it as a legitimate struggle song. It is not as if there is an oversensitivity aimed at struggle songs, or that legitimate and harmless historic memories are being prejudiced. It also shows up the remarkable hypocrisy in allowing the singing of this song, evidence aside.
But the evidence, seen from our conflict studies perch, raises another big, waving red flag. For those of us who are not content with dismissing the court cases as matters of settled history, we see a repeat pattern in current events. Our current migration conflicts, so harmful, so deserving of our attention and focus, again show a troubling repeat pattern of the topics that we discussed up to now, if only we know where to look, and are prepared to do so. Again we have songs, slogans and rhetoric directly supportive of and potentially enticing violent behaviour, the cries and songs of "Mabahambe" / "Abahambe" ("They Must Go!") again influence and direct behaviour. We again see songs and slogans with its roots in struggle history, sung in charged crowds with weapons or gestures of force present, again people claim that “of course” these songs and slogans are not calls for literal violence, and again we see a difference between the deflection and the reality on our streets.
Again, the underlying causes play an important role in creating legitimate anger and despair, again by definition setting the stage for the exact foundations for violence that we find in the studies, again these causes are offered as justifications for irresponsible conflict behaviour.
Conclusion
I anticipate that our comparison of legal versus conflict management perspectives, evidence and conclusions will be met by some with a shrug and the pointing out that the debate is a legal dead-end after the history of the argument thus far. I could not agree with such a dismissal. We saw the narrow focus of the court cases. It is clear that not all relevant evidence was placed before the court. Some of the studies that we have looked at were also concluded after the closing of pleadings, especially in the SCA matter. The interconnectedness with other national conflicts, such as the spiralling migration conflict, may not have been fully understood or examined.
In any event, once the evidence and causal links are understood, responsible political leadership should not hide behind, or wait for, a perceived lack of legal developments. Being aware of the potential for such violence and harm, and nevertheless allowing the song to be sung as freely as it is done at this moment in our history, and having access to this evidence, makes government liable on a spectrum of events, at least based on a dolus eventualis basis.
Where there is evidence of real and potential harm, the government has an active duty to act reasonably in preventing or minimizing such actual or potential harm. Running the free speech argument again, knowing what we know, it cannot be argued that the right of free speech involving such a poorly defended instance of rhetoric, so clear to even a basic understanding of the words used, can outweigh the rights of those potentially affected, especially where the harm suffered is linked to violence inducing methods. No one has a right to use violent, or potentially violent, narratives, no one has a right to play dice with a song of such limited cultural and historic value, as balanced against its potential harm.
As a conflict management study the debate and its present impasse is an absurd one, not warranted by any meaningful form of evidence or logic. It is, at best, a tool manufactured and elevated to this damaging level by the clash between identity conflict dynamics, and it is exactly there where it is allowed to do its harm, and to hold the potential for violence, harm and atrocities that it does.
As we have seen from our President’s visit to the White House in May 2025, perceptions and words form weapons against our best interests, and if the arguments for the retention of the public use of the song are so flimsy on the domestic stage, they simply disappear when tested on the global stage.
The song, in this form, should not be allowed to be sung in public and as part of political events. It is, in its own way, a test for how we approach conflicts and seemingly intractable, unresolvable differences and interests. It is a small but important test, and we are failing it.
Summary of main sources, references and suggested reading
1. The research for this article is selected articles from my consulting work. While that work is done in a European context, the data easily and in all instances, carry across effectively to a South African context and application. Where my work deals with clearly European applications, I have left such evidence out of our discussion for this article.
2. For articles dealing with conflict in general, see our blog index at Conflict Conversations
(Andre Vlok can be contacted at andre@conflict1.co.za for any further information.)
(c) Andre Vlok July 2026
* Author’s note on the use of artificial intelligence in writing this article
I learned to draft, argue and write in the hard school of litigation. I enjoy and value the very human process of creating ideas, of testing my own knowledge and thoughts. It is a process that I need, for answering some of my professional and even personal questions, it is cathartic and inspiring. Other than the most basic research assistance I do not use any AI in the creation of my written work, this article included. It is a matter of pride, of preference, and of mental health. Whether that is a wise choice or not, I will leave to the reader to decide.