Failures of nonviolent action?

Brian Martin

followed by

Learning from failures in the 2011 uprisings

Thomas Davies

Published in Global Change, Peace & Security, vol. 28, no. 1, 2016, pp. 117-122

pdf of published communications

Failures of nonviolent action?

Brian Martin

Go to

Responses to critiques of nonviolence

Brian Martin's publications on nonviolence

Brian Martin's publications

Brian Martin's website


Thomas Richard Davies in his 2014 article 'The failure of strategic nonviolent action in Bahrain, Egypt, Libya and Syria' in Global Change, Peace & Security addresses an important topic.[1] As he notes, there are many lessons to be learned from failures of nonviolent struggles, but far more attention is given to successes. Davies analyses the 2011 nonviolent campaigns in four countries, diagnoses four corresponding modes of failure, and argues that in each of them nonviolent action may contribute to its own failure. Specifically, he argues that in Bahrain nonviolent action made resisters vulnerable to repression, in Egypt the nonviolent campaigners' strategy of alignment with the military made subsequent military repression easier, in Libya the repression of nonviolent resisters generated international outrage, making external military intervention easier to justify, and in Syria nonviolent action contributed to splits in the military, encouraging a transition to armed struggle.

Davies' analysis is well informed in terms of nonviolent action theory; he looks in particular at the process of political jiu-jitsu, a process by which violent attacks on peaceful protesters can generate greater support for the protesters, by analogy with the martial art of jiu-jitsu.[2] Davies draws on numerous commentaries on the four cases, providing an especially useful overview of international factors. Despite these strengths, there are some questionable aspects to Davies' analysis and conclusions.

Consider the first mode of failure: nonviolent action can make resisters vulnerable to repression. This risk is greatest when using methods of concentration, such as rallies, when large numbers of protesters are in specific locations in public, making them prime targets for attack by police and troops. In such circumstances, nonviolent action strategists recommend switching to methods of dispersion, including strikes, boycotts and symbolic methods such as turning off lights at a particular time.[3] These sorts of methods reduce the physical risk to protesters while enabling a continuation of the struggle. Davies refers to 'widely dispersed nonviolent action, with protests in locations encompassing population centres from Al Dair to Dar Kulaib, Alexandria to Aswan, Zawiya to Tobruk and Daraa to Qamishli',[4] but this is better thought of as use of concentrated methods - furthermore the same method, rallies - in many different locations. A shortcoming in these nonviolent campaigns was that there was not a greater and earlier switch to different sorts of dispersed methods, such as strikes and boycotts, that do not make protesters easy targets for state violence. Davies is right to point to a weakness in the use of nonviolent methods, but it is one already addressed in the literature.

Next consider the mode of failure that Davies diagnoses in Egypt: that protesters' alignment with the military made subsequent military repression easier. This is indeed a worthwhile point: nonviolent activists need to be careful about who they recruit and support as allies, especially when these allies, domestic military forces or foreign governments, have the capacity for repression. In the 1986 Philippines people-power revolution, breakaway troops were defended from attack by massive rallies of peaceful protesters. Was this a bad idea?

After-the-fact critique of nonviolent strategies is far easier than making astute decisions at the peak of struggles. A nonviolent strategist would have to be far-sighted to be able to anticipate the post-Mubarak trajectory of the Egyptian struggle. Is incorrectly forecasting the future a failure of nonviolent action?

Davies claims that the toppling of Mubarak in 2011 was not a success of nonviolent action, and that 'military dictatorship persisted'[5] after Mubarak's departure. However, there was a reasonably free election in 2012, hardly compatible with military dictatorship. Using Chenoweth and Stephan's criteria,[6] the Egyptian campaign was both nonviolent and successful, and most nonviolence scholars would concur.

Even with these reservations, Davies' analysis offers useful lessons for nonviolent campaigners. It is well recognised that fraternisation with troops is a key method of undermining their loyalty to the regime and is crucial to the success of nonviolent revolution.[7] As well as targeting troops' loyalty to the regime, fraternisation efforts desirably should also aim at fostering troops' commitment to nonviolent struggle. This can take many forms, as shown by the variety of methods used by troops refusing to join military coups, for example in Algeria in 1961 remaining in barracks and flying planes out of the country.[8] It is possible to imagine other methods of military resistance, including disabling weapons systems. Effective methods of fraternisation deserve further investigation, testing and evaluation.

For the case of Libya, Davies argues that government repression of peaceful protesters generated international outrage, providing a rationale for foreign military intervention. The resistance to the regime quickly morphed from nonviolent to violent, including armed opposition within Libya as well as Western military involvement. Davies treats this as a failure of nonviolent action. More specifically, he analyses this dynamic in terms of political jiu-jitsu, a concept deriving from Richard Gregg's idea of moral jiu-jitsu and being broadened and renamed political jiu-jitsu by Gene Sharp. The basic idea, in Sharp's formulation, is that if protesters remain nonviolent in the face of attack, it is possible that violent attacks on them will generate greater support for the protesters, with three groups specifically affected. More members of what Sharp calls the grievance group will become active, some third parties will become sympathetic to the protesters, and even some members of the regime may change their commitments.

In the case of Libya, the grievance group was composed of all members of the population oppressed by the regime. Some of these people became more active in the nonviolent struggle, but others initiated armed resistance. This violated one of the key principles of nonviolent struggle, namely to maintain nonviolent discipline.

Is it sensible to say, as Davies does, that political jiu-jitsu in this case operated in reverse, serving to undermine the nonviolent struggle and making it fail? A different interpretation, following Sharp's 'dynamics of nonviolent action',[9] is that the groundwork for nonviolent struggle had not been sufficiently prepared, and that prior to challenging the government there had been insufficient training in nonviolent techniques, including the necessity to avoid using violence. These shortcomings arguably were fatal to the nonviolent struggle.

Davies refers to 'strategic nonviolent action', which has the implication that the anti-regime campaigners are unified behind a strategy. Revolts triggered by external exemplars may lack sufficient preparation. Following a nonviolent strategy implies a degree of planning, training, coordination and agreement about methods and goals, along the lines of what Sharp calls 'laying the groundwork'.[10] This sort of preparation occurs in some struggles, but certainly was not sufficient in Libya in 2011.

That Western governments justified their intervention in part on the basis of attacks on peaceful protesters seems a secondary issue. Furthermore, it seems strange to interpret a pretext for Western military involvement as a failure mode of nonviolent action. This would imply that any action used as a rationale for military action potentially represents a failure of nonviolent action. Following Davies' approach, nonviolent action fails when violent attacks on peaceful protesters are ignored internationally - a common occurrence - and also fails when such attacks are used to justify armed intervention. This implies that part of the task of proponents of nonviolent action is to maintain nonviolent discipline not only among protesters but among external actors. This is incredibly difficult. Given that external actors usually serve their own interests, is it reasonable to treat their actions - supplying arms as in Syria or intervening militarily as in Libya - as failures of nonviolent action? This seems to be expecting far too much of campaigners.

Davies sees political jiu-jitsu as central to strategic nonviolent action. However, Sharp stated that political jiu-jitsu was neither necessary nor sufficient for the success of a nonviolent campaign. It is not necessary because success often occurs without a counterproductive attack on protesters. It is not sufficient because in many cases there is a jiu-jitsu effect but it is not enough to make the campaign successful.

Davies repeatedly cites a paper by myself, Wendy Varney and Adrian Vickers on political jiu-jitsu.[11] Subsequent to that paper, I further investigated why many attacks on peaceful protesters generate little or no public outrage, and studied five types of methods used by perpetrators to reduce outrage: covering up the action; devaluing the target; reinterpreting the events through lying, minimising consequences, blaming others and framing; using official channels to give the appearance of justice; and intimidating and rewarding people involved.[12] I called this the backfire model to distinguish it from Sharp's political jiu-jitsu: the model includes both the potential of a counterproductive outcome (backfire) and processes of outrage management, and can be applied to struggles outside the usual scenarios of violence-versus-nonviolence, for example sexual harassment and genocide.

One of the key points arising from the backfire model is that the existence and level of public outrage over perceived injustices is the outcome of a struggle to manage this outrage, with perpetrators and their allies typically seeking to reduce outrage and activists seeking to increase it, though either group may misjudge the consequences of their efforts in this regard. From this perspective, it is entirely predictable that governments will try to reduce outrage that does not serve their interests and to stimulate it when it does. In this context, public reactions do not autonomously trigger government reactions; rather there is an interplay between publics, activists and governments. The use of attacks on peaceful protesters as a pretext for action is simply one particular technique in a wider package of outrage management methods. It is something that activists should take into account, but hardly merits being called a mode of failure.

Davies advises strategists of nonviolent campaigns to take into account the international context, which is a reasonable suggestion, and indeed something that experienced and knowledgeable activists do routinely. More questionable is Davies' conclusion that, so far as the audience for attacks on peaceful protesters is concerned, 'where that sympathetic audience consists only of segments of national armed forces or of international actors prepared to use violence, the logic of political ju-jitsu plays a part in the transformation from nonviolent into violent action'.[13] This is indeed an undesirable outcome, and one not anticipated by Sharp or others who have studied political jiu-jitsu. However, it does not seem to be an outcome inherent in political jiu-jitsu or in nonviolent action more generally. If success in nonviolent campaigns depends crucially on persisting in struggle while maintaining nonviolent discipline, then in the cases of Libya and Syria the failure to remain nonviolent is far more significant than offering alleged justifications for armed resistance or intervention.

In summary, Davies usefully points out that there is much to be learned from unsuccessful nonviolent struggles. His emphasis on understanding the international context is worthwhile, as is his noting of pretexts for armed struggle. However, rather than accepting that his four modes of failure are inherent in nonviolent action and more specifically in political jiu-jitsu, it is likely to be more productive to start with Sharp's dynamics of nonviolent action - especially laying the groundwork and maintaining nonviolent discipline - supplemented by findings from more recent studies, and to analyse failures from that perspective.


I thank Tom Davies, Maria Stephan and Tom Weber for valuable comments on drafts.

Learning from failures in the 2011 uprisings

Thomas Davies

Go to

Tom Davies' web profile


Until recently the study of strategic nonviolent action has, as Erica Chenoweth and Maria Stephan argue, been considered to be 'something of a pariah' in strategic studies.[14] It is in this context that literature on the subject has tended to concentrate on apparently successful cases of nonviolent resistance, which has helped to secure its position alongside traditional military strategic analysis in a growing array of literature.[15]

In his response my article on strategic nonviolent action in the 2011 uprisings in Bahrain, Egypt, Libya and Syria, Brian Martin observes that 'many lessons' may be learned from instances of failure. His analysis is particularly helpful for pointing towards a number of aspects of nonviolent strategic thought that may be reconsidered and improved in the light of the experience of the 2011 uprisings.

Among the key contributions in Martin's analysis is its elaboration on the need for more sophisticated consideration of the relationship between nonviolent actors and armed forces. A central claim in literature on strategic nonviolent action has been that persuading loyalty shifts among the armed forces can be crucial to success of a nonviolent campaign.[16] As my article indicated, the experience of the 2011 uprisings indicates key potential problems in this relationship, such as legitimation of military pre-eminence in Egypt, and helping to facilitate splits in the armed forces that contributed towards civil war in Libya and Syria. Martin's analysis helpfully indicates some of the potential methods of more effective fraternisation with armed forces that deserve far greater attention than they have received to date.

Martin's analysis is also helpful for its clarification of the roles of dispersion and concentration in nonviolent action. A key distinction that is sometimes overlooked is between methods of concentration that are widely dispersed, such as protests in multiple cities, and methods of dispersion such as strikes and symbolic actions. As my article notes, the 2011 uprisings involved both methods of concentration that were widely dispersed and methods of dispersion such as the strike actions in Bahrain, Egypt and Syria. The distinction Martin highlights with respect to forms of dispersion is a significant one, but adoption of both strategies was insufficient to prevent the failures in the uprisings in these countries.

There is a significant contrast between understandings of the definitions of success and failure for strategic nonviolent action in Martin's analysis, and those put forward in my article. Strategists of nonviolent action need to consider carefully whether use of force legitimated with reference to attacks on peaceful protesters is merely 'one particular technique in a wider package of outrage management methods' or a mode of failure that nonviolent strategists should seek to avert.

According to Martin, 'most nonviolence scholars would concur' that the 2011 uprising in Egypt was a success for strategic nonviolent action, with a 'reasonably free election' taking place in 2012. This appears to be in sharp contrast to analyses from beyond their discipline, where it has been emphasised that the aftermath of the uprising was not regime change, but rather 'appears to have amounted to an intra-regime coup, with the military faction prevailing over a rival business faction.'[17] If the criteria Martin uses to deem the Egyptian uprising to be successful were to be applied to violent strategy, it could be argued that the military invasion of Iraq in 2003 was a 'success', given the 'reasonably free' elections that took place subsequent to the invasion, a conclusion with which many would not concur.

If it is the aim of strategists of nonviolent action to be considered on the same terms as strategists of violent action, the failures of strategic nonviolence need to be subjected to the same critical scrutiny as the failures of violent strategy. In his conclusion, Martin suggests that the reader should return to Gene Sharp's original formulation of the dynamics of nonviolent action, especially laying the groundwork and ensuring nonviolent discipline, rather than the dynamics of failure set out in my article. However, the role of nonviolence in legitimating the use of force by other actors, is not, as Martin argues, 'a secondary issue', but instead should be a key concern in nonviolent strategic analysis. In elaborating on effective methods of fraternisation with armed forces, Martin makes his most helpful contribution to advancement of non-violent strategic thought. We should also take into account the interplay of geopolitical and societal factors with the dynamics of failure, elaborated in my article but in need of further testing. As Martin argues, the expectations set in my article with respect to the diverse array of circumstances and dynamics that protesters need to take into account demands much more of nonviolent strategists than traditionally put forward. However, the invigoration of dictatorial rule in Bahrain and Egypt and the escalating violence in Libya and Syria since 2011 reveal how disastrous the consequences of failing to do so can be.


1. Thomas Richard Davies, 'The failure of strategic nonviolent action in Bahrain, Egypt, Libya and Syria: "political ju-jitsu" in reverse', Global Change, Peace & Security 26, no. 3 (2014): 299-313.

2.   Gene Sharp, The Politics of Nonviolent Action (Boston, MA: Porter Sargent, 1973), 657-703.

3. Kurt Schock, Unarmed Insurrections: People Power Movements in Nondemocracies (Minneapolis, MN: University of Minnesota Press, 2005).

4. Davies, 'The failure of strategic nonviolent action in Bahrain, Egypt, Libya and Syria', 308-9.

5. Ibid., 310.

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

7. Sharon Erickson Nepstad, Nonviolent Revolutions: Civil Resistance in the Late 20th Century (Oxford: Oxford University Press, 2011).

8. Adam Roberts, 'Civil resistance to military coups', Journal of Peace Research 12 (1975): 19-36.

9. Sharp, Politics of Nonviolent Action, 449-814.

10. Ibid., 451-520.

11. Brian Martin, Wendy Varney and Adrian Vickers, 'Political jiu-jitsu against Indonesian repression: studying lower-profile nonviolent resistance', Pacifica Review 13, no. 2 (2001): 143-156.

12. Brian Martin, Justice Ignited: The Dynamics of Backfire (Lanham, MD: Rowman & Littlefield, 2007).

13. Davies, 'The failure of strategic nonviolent action in Bahrain, Egypt, Libya and Syria', 313.

14. Erica Chenoweth and Maria Stephan, Why Civil Resistance Works: Why Civil Resistance Works: The Strategic Logic of Nonviolent Conflict (New York: Columbia University Press, 2011), p. 17.

15. See, for instance, Lawrence Freedman, Strategy: A History (New York: Oxford University Press, 2013), chapter 23.

16. Maria J. Stephan and Erica Chenoweth, 'Why Civil Resistance Works: The Strategic Logic of Nonviolent Conflict", International Security 33:1 (2008): 42.

17. Ewan Stein, 'Revolution or Coup? Egypt's Fraught Transition', Survival: Global Politics and Strategy, 54:4 (2012): 45-66.