Hunger Strikes by Suffragettes and Irish Republicans, 1909-1923

Suffragette posterPeople sometimes invite suffering—or actually inflict it on themselves—as an act of political protest. They march long distances, go willingly to jail, welcome or provoke the blows of police, refuse to eat, and even kill themselves. Such actions, which I term 'communicative suffering', have been neglected by social scientists. Systematic analyses of protest events have focused on tactics like strikes and demonstrations. Communicative suffering may be less common, but it raises an important theoretical puzzle: why would protesters treat their own suffering as benefiting their cause? (Suicide terrorism is less puzzling because the perpetrators, by accepting death, can inflict greater harm on the enemy.) My research on self-immolation—where someone kills him or herself for a cause, without harming others—reveals the various ways in which suffering can become a source of power. It can win the sympathy of 'bystander publics' by signalling injustice, and it can exhort sympathizers to act by evoking anger or guilt. To extend and deepen my analysis of communicative suffering, I will turn to the phenomenon of hunger striking. It is closely akin to self-immolation, though the possibility of threats and bluffing on either side introduces another layer of complexity—and theoretical interest.

Hunger strikes were central in two movements that convulsed and reshaped the British polity in the early twentieth century, the campaigns for women's suffrage and for an Irish Republic. Although the first is conventionally treated as a "social movement" and the second as a "nationalist insurgency", it is natural to investigate them together. Republicans borrowed the tactic of hunger striking from suffragettes, and the Westminster government used the same range of measures to deal with both movements. These two movements together provide an exceptionally dense concentration of hunger strikes: the tactic was used by hundreds of suffragettes from 1909 to 1914, and many thousands of Republicans from 1916 to 1923 (including prisoners of the Irish Free State from 1922). These numbers dwarf the more famous hunger strikes in Northern Ireland in 1981. A further advantage is that government records from the early twentieth century are open.

Research objectives

My research will contribute to the literature on social movements in two ways. A growing body of scholarship focuses on the role of emotions (Goodwin, Jasper, & Polletta 2001; Flam & King 2005). Hunger strikes obviously play on emotions, in two ways. First, graduated starvation requires almost superhuman self-control, and emotions are central to this process; for example, the anticipated shame of surrender could be used to offset the pangs of hunger. Second, the hunger strike is potent because it can evoke powerful emotions among adherents and bystanders. Perhaps less obviously, the protagonists themselves made careful calculations about the emotional consequences of their actions; for example, officials envisaged what would happen were they to allow a prisoner to die. This combination of strategy and passion provides a rich field for sociological analysis.

The second contribution will be analysing the dynamics of interaction between movement and government. Social scientists have generally focused on the movement, treating its antagonist to a static configuration of "political opportunities" (a deficiency emphasized by McAdam, Tarrow, & Tilly 2001). Hunger strikes allow a dynamic analysis of interaction on two levels. Each hunger strike is a sequence of interaction, in which either the prisoner or the authorities can decide to back down—as indeed happened in the vast majority of cases. I have analysed the logic of these interdependent decisions using game theory. At a higher level, both sides learned new strategies over time. The "Cat and Mouse" Act of 1913, for instance, enabled the government to release starving prisoners only to re-arrest them when they had recovered.

Although gender is not the specific focus of investigation, the juxtaposition of these two movements will provide fruitful comparisons, especially regarding the differential treatment of women and men by the authorities.


There are extensive literatures on the campaign for woman's suffrage and the war for Irish independence (e.g. Joannou & Purvis 2001; Hopkinson 1998, 2002; Purvis & Holton 2000). Building on the work of historians, this project will employ a methodology pioneered by sociologists investigating social movements: the systematic compilation of data on discrete protest events over a period of time. Conventionally these events are "contentious gatherings" such as demonstrations or strikes (Olzak 1989). This project will collect standardized data on each individual's hunger strike, beginning with the prisoner's decision to go on hunger strike and ending when the government offered concessions (including release), the prisoner backed down, or the prisoner died. Usually several prisoners went on hunger strike together—the largest involved 7,800 prisoners—but it is vital to collect data on individuals, in order to trace multiple hunger strikes by the same person.

This systematic compilation of data is important because historians have naturally concentrated on the most committed prisoners, like Emmeline Pankhurst and Terence MacSwiney. In the case of Irish Republicans, the six hunger strikers who died have gained far more attention than all the others put together (I estimate at least 10,000). Systematic data on every hunger strike will reveal the contours of the overall phenomenon for the first time. These data will flesh out the game-theoretic models. They will also enable me to trace longer-term strategic shifts on each side. To take an example, I hypothesize that hunger strikes make less sense when the government is executing prisoners, for two reasons: self-inflicted suffering is less necessary with state-created martyrs; the hunger strike becomes unattractive as a bluff, because the government has shown that it is not averse to the death of prisoners. This hypothesis can be tested by comparing time series of hunger strikes and executions.

An analysis of hunger strikes must also consider two kinds of control groups. First, some suffragettes and Republicans chose not to go on hunger strike. Fortunately there is a clearly defined population of prisoners, and their incarceration generated records. I will gather summary information on the total number of the movement's prisoners at monthly intervals, and then compile comprehensive data on non-participants at selected prisons at times of mass hunger strikes. Movements that eschewed hunger strikes constitute the second kind of control group. I will explore this using secondary sources on the labour movement and on conscientious objectors in the United Kingdom during this period.

(Funded by the British Academy.)

Michael Biggs, Department of Sociology, University of Oxford