Towards strategic rioting?

Theory in Action, vol. 15, no. 2, April 2022, pp. 77-89
pdf of published article

Brian Martin

Go to

Responses to critiques of nonviolence

Brian Martin's publications on peace, war and nonviolence

Brian Martin's publications

Brian Martin's website


Benjamin Case argues that the framework of strategic nonviolence is limited by its assumption that violent protest necessarily demobilises movements, and that rioting can be empowering for participants. However, Case’s statistical analysis of US riots and peaceful demonstrations may not be a comparison of rioting and nonviolent action because it is questionable whether, in the US, peaceful demonstrations should be classified as methods of nonviolent action. Rioting can be empowering, but there is also considerable evidence that participation in nonviolent action can be empowering. Much research remains to be done to determine whether rioting can be a leading or major part of strategic action for social change.

KEYWORDS: Violence, Nonviolence, Social Movements, Rioting


Benjamin Case (2021) in “Molotov cocktails to mass marches” argues that, in the US, riots are beneficial for social movements in two ways. First, they mobilise greater support, including promoting more nonviolent activism. Second, they are emotionally invigorating for participants.

My aim here is to examine Case’s arguments with an eye for insights about nonviolent action. Case’s analysis raises questions about what counts as nonviolent action and about the emotional impact of participation in different forms of social action.

Case presents his argument as a counter to the approach of strategic nonviolent action, which in recent decades has become more influential in US social movements. This approach draws on the ideas of Mohandas Gandhi (Bondurant, 1958), who pioneered what has been called principled nonviolence, based on an ethical imperative not to use violence, and of Gene Sharp (1973), the preeminent advocate of what has been called pragmatic nonviolence, based on the view that nonviolent action can be more effective than armed struggle. Gandhi and Sharp were vital figures but many others have contributed to the development of nonviolence theory and practice.

It is useful to remember that the approach of strategic nonviolent action, drawing especially on Sharp, is relatively new as a way of thinking about social movement strategy. For decades, many on the left were inspired by armed struggle in countries such as China and Cuba. Guerrilla warfare was seen as the most promising way to challenge repressive governments, especially client states of the United States. The left ignored Gandhi and Sharp, adopting the prevailing assumption that violence, if sufficiently ruthless, would always be victorious over nonviolent opponents. With this assumption, liberation necessarily depended on the use of arms.

Gandhi and Sharp challenged the assumption about the necessary superiority of violence but their views had relatively limited impact outside the nonviolence community until research by Erica Chenoweth and Maria Stephan (2011). They examined 323 anti-regime, secessionist and anti-occupation campaigns from 1900 to 2006, classifying them as primarily either violent or nonviolent. They concluded that nonviolent campaigns were twice as likely to be effective as violent campaigns. The exact figure may be disputed, but undoubtedly Chenoweth and Stephan’s work changed the debate by undermining the assumption that only armed struggle has the capacity to bring down a ruthless government.

Chenoweth and Stephan examined campaigns with maximal goals, most of them to overthrow governments. Their findings can be applied to other sorts of campaigns, within political systems. In this, the US civil rights movement is the most common example, but there are many others. In this context, a number of writers and activists have argued for the value and effectiveness of rioting (Anisin, 2020; Meckfessel, 2016). Case (2018, 2021) is one of them. He presents evidence that rioting in the US is more effective than nonviolent action and that rioting can be empowering to the participants. In the following sections, I examine each of these issues.

Rioting and nonviolent action

Case uses data from the Banks Cross-National Time-Series Data Archive and concludes that, “… years with increased riots also see increases in nonviolent protest, and vice versa, indicating that violent and nonviolent protest are mutually constitutive of moments of uprising” (Case, 2021: 26). It is possible to raise questions about Case’s data and the robustness of his conclusions.[1] However, here I want to raise a more general question about what counts as nonviolent action.

The concept of nonviolent action has been debated by many scholars and activists. For the purposes here, it is useful to point to boundaries with other forms of social and political action, acknowledging that the locations of the boundaries may be imprecise and contested.

The boundary that has been most discussed is between violent action and nonviolent action. When Palestinian youth throw stones at Israeli soldiers, should this be considered violent or nonviolent action or perhaps a different category, unarmed struggle (Rigby, 1991) or unarmed violent action (Pressman, 2017)? When workers in factories (for example in Nazi Germany) use sabotage to halt production, should this be considered violent or nonviolent action? Should self-immolation be considered violent or nonviolent action? Is cutting a fence at a military base violence against property? These sorts of questions have been debated by activists for decades. There are no right or wrong answers, although the answers chosen may affect the way people think about particular actions.

Another boundary, much less discussed, is between nonviolent action and conventional social and political action. This boundary is especially important when looking at Case’s analysis. To consider this boundary, it is useful to turn to the pioneering work of Gene Sharp (1973), who famously catalogued 198 different methods of nonviolent action, giving historical examples of each one. Among the methods are numerous types of protest and persuasion such as signed public statements, petitions, deputations, religious processions, protest meetings, and marches. What is not often noticed is that Sharp says that such methods are only sometimes in the category of nonviolent action.

Although Sharp does not make a clear statement about the boundary between nonviolent and conventional social and political action, his intent can be inferred from various comments. His basic idea is that nonviolent action goes beyond what is normally acceptable to authorities, and this depends strongly on the circumstances. In a dictatorship, a political petition may be treated as subversion and put signers in jeopardy. In a political system where civil liberties are respected, a petition is nothing special: it is part of conventional political activity.

In writing about characteristics of nonviolent action, Sharp distinguishes between action and inaction, emphasising that nonviolent action is not passivity. He says that within the category of action, “a dichotomy into violent and nonviolent is too simple” (65), and then distinguishes nonviolent action from verbal persuasion, violence against people, and material destruction. Importantly, he says that nonviolent action is different from “peaceful institutional procedures backed by threat or use of sanctions” (65). In distinguishing nonviolent action from other forms of action, he writes,

Nonviolent action is so different from these milder peaceful responses to conflicts that several writers have pointed to the general similarities of nonviolent action to military war. Nonviolent action is a means of combat, as is war. It involves the matching of forces and the waging of “battles,” requires wise strategy and tactics, and demands of its “soldiers” courage, discipline, and sacrifice. (67)

He goes on to say,

Another characteristic of nonviolent action which needs emphasis is that it is usually extraconstitutional; that is to say, it does not rely upon established institutional procedures of the State, whether parliamentary or nonparliamentary. (67)

The implication is that what counts as nonviolent action, in Sharp’s conception, depends on the circumstances, in particular whether the form of action is inside or outside what are routine or accepted modes of action within the society. This can be illustrated by Sharp’s comments about some methods of protest and persuasion. In writing about formal statements, Sharp says,

Normally, written or oral statements, whether by an individual, group or institution, are simply verbal expressions of opinion, dissent or intention, and are not nonviolent protest and persuasion as defined above. However, certain circumstances may give such statements a greater than usual impact and such an act may then fall within this class. (119)

For example, in writing about method 2, letters of opposition or support, Sharp writes, “Letters usually gain sufficient significance to be classed as a method of nonviolent protest because of the status of the signer or signatories, because of the number of persons signing the letter or sending identical or similar letters, or because the political situation has heightened the significance of such an act” (120). He then gives examples from Nazi-occupied Europe and from Czechoslovakia in 1968 after being invaded by Warsaw Pact troops. The implication is that letters of opposition or support constitute a method of nonviolent action when they challenge a repressive government, but not otherwise. It makes a difference whether writing a letter might lead to severe consequences or is just an everyday occurrence.

It is useful to note that Sharp’s main concerns were war, dictatorship, genocide and systems of social oppression. His most popular book, translated into numerous languages and widely used by activists in many countries, especially those ruled by repressive governments, was From Dictatorship to Democracy (1993). Although Sharp cited the US civil rights movement and a number of other campaigns within systems of representative government, it is reasonable to understand his selection of 198 methods of nonviolent action in the context of dictatorship.

As noted, in some countries, methods listed by Sharp would not be counted as nonviolent action. In the US, this might apply not just to statements and petitions but also to rallies and marches, depending on the circumstances. A march by African Americans in the US South in the 1950s was a serious challenge to oppression, and could be met by police violence. On the other hand, in parts of the US, rallies, marches and other forms of protest and persuasion have become normalised. They are either legal or not met with major police violence. They are often treated as expressions of opinion and sentiment rather than serious threats to systems of power. Therefore, following Sharp, it might be that many such actions should not be considered forms of nonviolent action.

This assessment puts Case’s analysis in a different light. In analysing patterns of peaceful protest and riots, he is not so much comparing nonviolent action and unarmed violence as comparing nondisruptive and disruptive protest, neither of which would readily fit within the category of nonviolent action. With this perspective, Case’s findings are compatible with the views of Piven and Cloward (1979), who argued that for poor people’s movements in the US, disruptive protest was far more effective in forcing concessions than building formally structured organisations with mass memberships. They said that movement leaders in the 1930s and 1960s were so preoccupied with building formal organisations that they actually curbed popular defiance, for example large-scale strikes and mass civil disobedience. Interestingly, Piven and Cloward’s work is taken as a touchstone by Engler and Engler (2016), who argue for disruptive nonviolent action linked with community organising.

To compare nonviolent action with riots in the US, examination is needed of some stronger forms of disruptive action, for example workers occupying a factory or protesters blocking construction of a pipeline. Such a comparison might or might not confirm the same patterns that Case found with peaceful protests.


Case, in the second main part of his article, gives evidence that rioting can be empowering for rioters. In particular, black bloc participants[2] gain a feeling of liberation by smashing shop windows and taking other actions against property which, in the US, has almost a sacred status. This is evidence of a significant psychological impact on rioters.

To better appreciate the role of empowerment through participation in rioting, here I look at empowerment in social action more generally, specifically through participation in nonviolent action and in warfare as well as in rioting. As well, I point to a number of other factors that should be considered in conjunction with empowerment of participants.

The empowerment of rioters, as recounted by Case, can be compared with the psychological impact of participating in nonviolent action. For this, it is useful to again turn to Sharp’s The Politics of Nonviolent Action, where he devotes twenty pages to a survey of the effects of nonviolent action on the actionists (Sharp, 1973: 777–799). He identifies several different effects, at a group level and/or individual level, including a reduction in submissiveness, an understanding of one’s own power, greater fearlessness, greater self-esteem, greater enthusiasm and hope, greater group unity and more cooperation. For each of these effects, Sharp gives examples from major campaigns, including the American revolutionary struggle in the 1700s, Gandhi’s campaigns in India, the US civil rights movement, strikes in Russia and Britain, the 1905 revolution in Russia, and anti-apartheid struggles in South Africa.

Of special interest is another effect: a challenge to the usual association of masculinity and violence. Sharp (1973: 790–791), citing work by Jerome Frank (1967), notes that participation in nonviolent action can reverse this association, replacing it with a sense of masculinity connected to being willing to suffer.

Sharp notes that most research on nonviolent action focuses on the effects on opponents and that much more attention should be placed on the effects on the activists. However, few have followed this recommendation. Martin and Varney (2003), in a study of nonviolence and communication, identified collective and individual empowerment as two facets of communication in nonviolent action, conceived as communication within groups and as inner dialogue. They pointed to evidence from two episodes in the Soviet Union, prison camp strikes at Norilsk and Vorkuta in 1953 and resistance to the 1991 coup.

It should also be pointed out that soldiers can be energised by their participation in violence (Bourke, 1999). The evidence is overwhelming that soldiers are powerfully affected emotionally by their experiences in wartime, so much so that this remains the time of their lives when they felt most alive. For example, Gray (1970, p. 44), in his classic reflections on the psychology of soldiering, writes, “Many veterans who are honest with themselves will admit, I believe, that the experience of communal effort in battle, even under the altered conditions of modern war, has been a high point in their lives.”

Evidence from Case, Sharp and warfare suggests that taking action is often a powerful emotional experience. In rioting, nonviolent action, and combat, the action goes against or beyond the accepted norms for conventional social and political action. In both rioting and nonviolent action, the effects can be empowering for those involved. However, participant experiences are only one factor to consider when examining forms of action. Other factors include harm, skills, participation, masculinity and prefiguration, each of which is addressed in the following paragraphs.

Harm here refers to physical, emotional and other damage to opponents, bystanders, participants and the natural and built environment. Harm always needs to be weighed against the benefits of successful action in reducing the harm from discrimination, exploitation and repression. These are not easy comparisons, because the prospects for success also need to be taken into account.

Militaries cause the most harm, including to their own soldiers. Nonviolent action can cause harm to opponents; for example, strikes and boycotts can cause economic damage. Much of the harm caused by rioting is to property, though people can be hurt too. An important example is the riot in south central Los Angeles in 1992 triggered by the not-guilty verdict by a jury in the trial of four police officers involved in the beating of Rodney King a year before. This riot led to the deaths of more than 50 people.

Participants in action can be harmed physically and also emotionally. It is well known that soldiers can suffer post traumatic stress disorder as a consequence of combat. Might there be some adverse psychological effects of participating in rioting? Or, on the other hand, could property destruction become so habitual that rioters seek greater thrills by causing ever more damage? There seems to be little investigation of these possibilities.

Skills are important for carrying out actions, and actions in turn can help develop skills. An action that fails to achieve its immediate goal may still be worthwhile if it increases the likelihood of future success. Soldiers are tested in battle and those who survive can become better fighters. The same applies to rioting, nonviolent action and conventional social and political action. Development of skills is often undertaken prior to action, as in military training. Arguably, the effectiveness of nonviolent campaigns can be improved by systematic training (Martin and Coy, 2017). Little has been written about training to become more effective rioters.

Participation in some forms of action is easier than in others. Participation in some forms of nonviolent action, such as strikes, boycotts and sit-ins, is possible for men, women, children, elderly and people with disabilities. On the other hand, some methods, such as perching in tripods or taking part in blockades of ships, require physical skills and endurance, restricting participation. Some of the methods commonly used in rioting, such as throwing objects, require some degree of physical fitness, as does escaping from police if that is part of the plan. In frontline combat, physical fitness is important, and young fit men play the biggest role. However, in other military roles there are tasks — such as building bombs and piloting drones — can readily be carried out by women, the elderly, children and some people with disabilities. What needs further investigation is the connection between the type of participation and empowerment. In particular, does putting one’s body at risk provide a special sort of emotional experience? How empowering is it to be involved in less direct ways, for example preparing a Molotov cocktail, joining a strike or working in an arms factory?

Masculinity plays a role in different forms of action. As noted, Sharp offered evidence that participating in nonviolent actions can counter conventional feelings of masculinity. In relation to warfare, on the other hand, there is a strong connection between use of violence and masculinity (Brock-Utne, 1985; Enloe, 1983, 1993; Ferguson, 2021). There seem not to have been studies on this topic concerning rioting. Women can certainly join riots but whether this is common and whether male rioters have their masculinity reinforced or mitigated remains to be studied.

Prefiguration refers to actions that embody the sort of society being sought. Warfare is obviously not prefigurative, at least if the goal is victory and then peace: it is seen as a means to an end, not an end in itself, and the exceptions to this, in which warmakers seek to perpetuate war (Keen, 2012) are undesirable.

Rioting is not prefigurative either; it is hard to imagine a desirable society in which rioting is a routine occurrence. Another important point is that rioting is not necessarily progressive. The worst race riot in US history was in 1921 in Tulsa, Oklahoma. It was a riot against the city’s black population, leading to numerous deaths (Hirsch, 2002).

Nonviolent action, on the other hand, is potentially prefigurative. It might be that in a society without organised violence, methods such as strikes and boycotts would be regularly used as part of conflicts. More obviously prefigurative is what is called constructive resistance, for example cultivating crops on unused private land, in which the action embodies the goal. Gandhi is most well known for campaigns using nonviolent action but he devoted just as much attention to promotion of alternatives in what was called the constructive programme, which was based on what we now call prefiguration (Ostergaard and Currell, 1971). Vinthagen (2015), in his four-dimensional reconceptualisation of nonviolence, highlights what he calls “normative regulation,” which involves prefiguring social relations without violence.

In summary, it is important to consider the mobilisation and empowerment effects of different types of action, but also important to consider other effects, including harm, skills, ease of participation, connection with masculinity, and prefigurative potential. In comparing armed struggle, rioting, nonviolent action and conventional social and political action, much remains to be learned. Most attention by researchers and practitioners has been on armed struggle and conventional action, about which there are vast bodies of research and commentary. Nonviolent action has received relatively little attention, and rioting even less.


It is valuable that Benjamin Case (2021) has set out to question assumptions underlying strategic nonviolent action, which has become an influential approach among US social movements. By presenting evidence about rioting as a method of action, in the US context, he usefully draws attention to areas that need further investigation.

Case uses data drawn from several sources to look for correlations between rioting and peaceful demonstrations. I have argued that this may not be a comparison of violent and nonviolent action, because in the US, peaceful demonstrations do not necessarily satisfy the criteria for nonviolent action, being closer to conventional political action, or at least on the boundary between nonviolent and conventional political action. Case’s comparison may be more like contrasting disruptive with nondisruptive or less disruptive action, showing the power of disruption in US social movement history.

Activists who try to follow the principles of nonviolent action often assume that every method listed by Gene Sharp (1973) falls in the category of nonviolent action. However, reading Sharp carefully shows this is not correct: according to Sharp, whether an action fits in one of his categories depends on the context, in particular whether it goes beyond conventional social and political action. Sharp in his writings mainly focused on highly repressive contexts, for example dictatorships, not on the use of methods tolerated by authorities. Following this line of thinking, US activists need to recognise that, in many cases, rallies and marches are closer to being conventional than nonviolent action. To make a fair comparison between riots and nonviolent action, stronger methods, such as those called civil disobedience, would be better comparators.

Case also provides evidence for the empowering effect of participating in rioting. There is also evidence of empowerment by participating in nonviolent action. Confrontations with opponents can generate powerful emotions. However, a full analysis of the emotional and mobilising effects of different forms of action — violence against opponents, property damage, and nonviolent action — remains to be carried out. Case has thus pointed to an important area for further research. Yet the impact of action on participants is only one factor to consider. Also important are harm, skill development, ease of participation, mobilisation of masculinity, and prefiguration.

Case has presented the beginning of an understanding of what might be called strategic rioting. Much more is required to fill out the picture, including recruitment, training and strategy. Indeed, it is not obvious that any movement has ever used rioting as its main method, nor how rioting as an occasional tool in a movement might relate to processes of legislative change, protection against attacks, negotiation with opponents, leadership dynamics and many other issues. Can rioting be the basis for a strategic approach to social change, or is it inevitably a supplement to nonviolent or conventional social and political action?


I thank Kristian Skrede Gleditsch, Andrew Rigby and Tom Weber for valuable comments.


Anisin, Alexei. 2020. Debunking the Myths behind Nonviolent Civil Resistance. Critical Sociology. 46(7/8): 1121–1139.

Biggs, Michael. 2018. Size Matters: Quantifying Protest by Counting Participants. Sociological Methods & Research. 47(3): 351–383.

Bondurant, Joan V. 1958. Conquest of Violence: The Gandhian Philosophy of Conflict. Princeton: Princeton University Press.

Bourke, Joanna. 1999. An Intimate History of Killing: Face-to-Face Killing in Twentieth-Century Warfare. London: Granta.

Brock-Utne, Birgit. 1985. Educating for Peace: A Feminist Perspective. New York: Pergamon.

Case, Benjamin S. 2018. Riots as Civil Resistance: Rethinking the Dynamics of “Nonviolent” Struggle. Journal of Resistance Studies. 4(1): 9–44.

Case, Benjamin S. 2021. Molotov Cocktails to Mass Marches: Strategic Nonviolence, Symbolic Violence, and the Mobilizing Effect of Riots. Theory in Action. 14(1): 18–38.

Chenoweth, Erica, and Stephan, Maria J. 2011. Why Civil Resistance Works: The Strategic Logic of Nonviolent Conflict. New York: Columbia University Press.

Dupuis-Déri, Francis. 2013. Who’s Afraid of the Black Blocs? Anarchy in Action around the World. Toronto: Between the Lines.

Engler, Mark, and Engler, Paul. 2016. This Is an Uprising: How Nonviolent Revolt Is Shaping the Twenty-First Century. New York: Nation Books.

Enloe, Cynthia. 1983. Does Khaki Become You? The Militarisation of Women’s Lives. London: Pluto.

Enloe, Cynthia. 1993. The Morning After: Sexual Politics at the End of the Cold War. Berkeley, CA: University of California Press.

Ferguson, R. Brian. 2021. Masculinity and War. Current Anthropology. 62(supplement 23): S112-S124.

Frank, Jerome D. 1967. Sanity and Survival: Psychological Aspects of War and Peace. New York: Random House.

Gray, J. Glenn. 1970. The Warriors: Reflections on Men in Battle. Lincoln, NE: University of Nebraska Press.

Hirsch, James S. 2002. Riot and Remembrance: America’s Worst Race Riot and its Legacy. Boston: Houghton Mifflin.

Keen, David. 2012. Useful Enemies: When Waging Wars Is More Important than Winning Them. New Haven, CT: Yale University Press.

Martin, Brian, and Coy, Patrick G. 2017. Skills, Training and Activism. Reflective Practice. 18(4): 515–25.

Martin, Brian, and Varney, Wendy. 2003. Nonviolence and communication. Journal of Peace Research. 40(2): 213–232.

Meckfessel, Shon. 2016. Nonviolence Ain’t What It Used to Be: Unarmed Insurrection and the Rhetoric of Resistance. Chico, CA: AK Press.

Ostergaard, Geoffrey, and Currell, Melville. 1971. The Gentle Anarchists: A Study of the Leaders of the Sarvodaya Movement for Non-violent Revolution in India. Oxford: Clarendon Press.

Piven, Frances Fox, and Cloward, Richard. 1979. Poor People’s Movements: Why They Succeed, How They Fail. New York: Vintage.

Pressman, Jeremy. 2017. Throwing Stones in Social Science: Non-violence, Unarmed Violence, and the First Intifada. Cooperation and Conflict 52(4): 519–536.

Rigby, Andrew. 1991. Living the Intifada. London: Zed.

Sharp, Gene. 1973. The Politics of Nonviolent Action. Boston: Porter Sargent.

Sharp, Gene. 1993/2010. From Dictatorship to Democracy: A Conceptual Framework for Liberation, 4th edition. East Boston, MA: Albert Einstein Institution.

Vinthagen, Stellan. 2015. A Theory of Nonviolent Action: How Civil Resistance Works. London: Zed.


1. For example, are the events in the Banks archive counted in consistent ways? Is it better to count the number of participants rather than the number of events? (Biggs, 2018). Is the ARIMA model used by Case appropriate if there are systematic changes in the data over time? How many of the events recorded in the data are routine political behaviour? I thank Kristian Skrede Gleditsch for suggesting these queries.

2. “Black bloc” refers to groups of activists who join demonstrations wearing black clothes and covering their faces to maintain anonymity. They often engage in violence against property or police. Many see themselves as anarchists (Dupuis-Déri, 2013).