1The election of 23 April 1848, held under direct universal suffrage, has been represented in the history books as the symbolic end of the barricades. [1] The election was immediately celebrated as the birth of “modern democracy”, of “popular sovereignty”, and of the “Republic of voters”. [2] Proclaimed from all quarters, these claims assumed the status of prophecy. No, in this revolutionary spring, electoral democracy would not turn its back on the Republic: a conviction shouted from the rooftops by the vast majority of the Parisian press, such that our reading of the works of chroniclers and memorialists has been deeply influenced by this interpretation of events. [3] According to Garnier-Pagès, “In the midst of a revolutionary explosion, France managed to establish the most honest, the most sincere, the most devoted, the most liberal of all the Assemblies she had ever constituted”. [4] A similar opinion was expressed by Georges Weill, who focused on how exemplary it was that the election managed to fight off the threat of disorder: “The calm with which we voted, almost everywhere, is a striking refutation of the fears and jeers of those who have for so long claimed that universal suffrage was unrealizable.” [5] Daniel Stern, meanwhile, was not afraid to exaggerate: “Everywhere order and calm reigned over this immense moral and physical movement of a whole people. Not a single accident, not a single incident of disorder came to disturb the operation that had been deemed by experts to be physically impossible”, [6] a view which was echoed even by supporters of the socialist movement. [7] As a result, the event has been locked into an interpretation that is simultaneously ahistorical and heroic, inaugural and teleological.

2But were these first elections under universal suffrage as “calm” as described? To answer this simple question, I launched a research project investigating law enforcement and judicial archives, as well as departmental and local press archives (to read more about the research protocol, see the methodological annex). Over time, about one hundred incidents of physical confrontation were selected, then subjected to close examination. The selection consists of cases involving open conflict which, caught up in this “election day”, provoked the intervention of the authorities, leading to arrests, injuries, and even deaths – enough to discredit the idealistic vision of a peaceful election. It is also enough to re-open an ongoing theoretical debate.

3Arguments about the pacifying role of electoral democracy are often normative, and risk organizing in advance what is supposedly under investigation. Two theories about the relationship between voting and violence thus confront each other. The first relies on the existence of a natural history of democracy. Its driving force is a mysterious “law” that allows for the gradual eradication of the barricades to be replaced by the ballot box: in short, from bullets to ballots. [8] This well-known argument claims that the appropriate means of pacifying conflicts can curb the “need” for, and therefore the probability of, violence. This theory dovetails with the accepted view of “democratic development”, or, in rational choice literature, with the idea that there exist mechanisms that naturally occur within electoral competition which are altered by violence. [9] This theory claims the moral crusades which are periodically fought in the name of the “defense of democracy” as supporting evidence. The second theory is the complete opposite of the first: it suggests that democracy encourages violence. Following the pioneering work of Douglas A. Hibbs, [10] the theory goes as follows: democracies have “a special proclivity” for stimulating political violence. This characteristic of democracy has often been obscured by cold war preoccupations and ignorance about historical experiences of democratization. Viewing electoral democracy as a means of pacifying conflict is a mistake; the electoral bargain does not at all lead to compromise or moderation. Jack L. Snyder even describes this idea as “dogma” intended to make the vote a kind of political cure-all unconnected to conditions or circumstances. Where the preconditions for democracy (a state based on the rule of law; a free press; flexible governing elites) do not exist, ballots may in fact encourage the use of bullets. In other words, from ballots to bullets. This conclusion proves especially true in ethnically divided authoritarian states. [11]

4Be that as it may. Although completely opposed, these two theories have generated a shared blindness when it comes to what exactly is at stake in the practice of violence, due to their fundamental premise. Partly because of the way the issue has been set out, both understand electoral violence as the break between “norms” and “deviance”. This distinction, which is brought to play in cases of electoral violence, works equally on an analytical level, separating two completely independent universes. But we must push this point. This article maintains that the study of how electoral violence becomes discredited requires the refutation of such idealism, whether it is based in law or economics. Electoral rule wasn’t drawn up against violence, nor outside it, but at the heart of it, using its physical elements as a basis and as a justification. This means that there is not an impermeable boundary between, on the one hand, the electoral norm that supposedly assures the authentic expression of the will of the people, and on the other, destructive acts that democracy should monitor and punish. Without a “theory capable of exploring the complicated and paradoxical relationship between bullets and ballots”, [12] understanding the way in which the rules of elections can pacify struggles for power means focusing on the practices that shape the recourse to violence, and, perhaps even more so, on the social relationships that are created or modified through these practices. [13] In short, we must acknowledge that electoral violence has a history, one that is as inseparable from the state as it is from the nature of social ties or relationships between competitors.

5Studying the violence that arose during and because of the April 23 election tests this theoretical framework. This is especially true because this first experience of French universal suffrage brought two different relationships to politics head to head: electoral citizenship, and insurrectional citizenship. [14] Since the French Revolution, insurrection and the barricade had been supposedly “democratic” means of acting politically. [15] Yet during this election, these methods were deemed disreputable by the notion of voter sovereignty, believed to be the source of “political truth”, superior to all “social truths”. [16] Yet while the stigmatization of violence may have been due to the particular context of this election, the pacifying action of elections is not an illusion. Over time, tallying votes has become a means of domesticating violence. To understand the process by which this happened, we need to look into the very heart of electoral “collusion”. Even better: we need to clarify the way in which an opposition was constructed between democratic deliberation and the taking up of arms, civilized behavior and deviance. To be clear, it was not legal scruples that reduced violent confrontation, nor the ideal of fraternity that had been promoted by the provisional government. It was rather the identification of the will of the People with electoral interdependence. In identifying the whole body of male voters (henceforth without conditions regarding wealth or education) with “a People” who were considered sovereign, the March 5 decree laid the foundation for the legitimacy of an abstraction. [17] This was a symbolic act of force that this time celebrated majority rule for its apparent arithmetical and moral objectivity and, and therefore consecrated it as a metonym for democracy. With “universal suffrage”, politics was no longer a battle, it was a competition: a vote, cast directly and made available to more of the population, whose verdict effectively doubled as aspirations for peace and national unity, by delegitimizing the motives of those who made their demands through violence, if not the actors themselves.

Electoral civility

6The pacifying dimension of electoral activity has often been graphically represented. Consider the 1869 engraving by printmaker Honoré Daumier of a republican worker brandishing his ballot and exclaiming aloud, “Here’s my cartridge”. Or, staying with 1848, consider the print by Marie Louis Bosredon depicting a rebel who abandons his rifle for the ballot – a way of saying that the vote is the only legitimate way to devolve political power. Moreover, it shows that violence is a repertoire of action that is fundamentally opposed to the exercise of the right to vote. The 12 April 1849 issue of La Gazette des tribunaux reads: “With the all-powerful weapon of universal suffrage, it is foolish and shameful to appeal to force and violence.” This is what, for Tocqueville, ensured that suffrage was part of the ideal of justice, for it substituted “the idea of law for violence”. [18] This principle was elevated to the level of mythology by Victor Hugo. During his speech before the National Legislative Assembly of the Second Republic on 31 May 1850, he stated: “By giving the ballot to those who suffer, universal suffrage takes their rifles from them… Now, what does that mean, sirs? It is the end of violence, the end of brute force, the end of rioting… the right to insurrection is abolished by the right to vote (applause).” That said, electoral violence has rarely been explained, or it has been explained etiologically by generic “causes” of violence: for example, in Marxist theory, one response given to a community experiencing persistent social frustration is that the frustration is tied to a state of economic dispossession. The schema dispossession/frustration/aggression ignores the purely political conditions of these acts of violence. In the same way, the existing literature on the Second Republic does not dwell on this first example of universal suffrage, preferring instead to focus on the legislative vote of June 1849 or the presidential vote of December 1848. As for historians of violence, they have ignored electoral confrontations, focusing instead on the republican demonstration against the Constituent Assembly on 15 May 1848, or the June Days uprising, or the Bonapartist coup of December 1851, whose human toll (and political resonance) was truly striking. [19]

7And yet, violence was clearly evident from 23 April 1848. My study of 104 cases constitutes strong evidence of this. These data are even more fascinating because, among them, only a half-dozen enjoyed any mention in the history books (mainly the confrontations in Rouen, Limoges or Nîmes). The sample is divided up among 50 departments of France, with an over-representation of rural cantons. We will not spend too long on total numbers (237 wounded, 49 deaths and 981 recorded arrests): they are indicative amounts only, due to the incomplete state of the archives, the vagueness of judicial proceedings, or the politicization of some of the numbers. [20] Recall that the verification of credentials carried out by the members of parliament at the beginning of May had “cleaned up” a number of ballots by covering up a number of idiosyncrasies. One such example: in the canton of Aigrefeuille, the elections included “regrettable violence and manipulation”, notably because roughly a dozen individuals with guns had prevented citizens from exercising their right to vote. And yet according to the public prosecutor of Rennes, “As the National Assembly has already validated the elections for the Lower Loire department, taking this back to court would be inappropriate.” [21] The electoral inquiry, although publicized by posters throughout the department, was abandoned. Electoral violence, whether witnessed by judicial authorities or the local and national press, did not happen by itself. It was constructed by creating a narrative which resulted in an administrative categorization, in the form of a ‘report’. This goes to show how much data remains inaccessible, especially those incidents that public prosecutors or the police dismissed as “unimportant” or “apolitical” demonstrations. [22]

8One hundred and four cases of electoral violence: a lot or a little? Either way, it is an opportunity to overcome the skepticism that would prefer this statistic to remain impossible, or at least unrealistic. [23] What these violent incidents which sprung up around polling stations show us, in the first place, is the continuation of the recourse to force in various forms – in other words, the vote of 23 April was part of the revolutionary dynamic of that February, which had seen the end of the Orleans monarchy and the beginning of the Second Republic. [24] It was the product of its revolutionary context, not an “electoral interlude” nor an “outbreak of democracy”. [25] While the National Guard went on patrol, with a drummer at the front of its lines, altercations arose in the administrative centers of the cantons: between electoral officials, in the polling stations, and in public squares. Such violence, however, should be carefully characterized. As historians of peasant revolts and of the police (la gendarmerie) have emphasized, the spring of 1848 was punctuated by anti-tax and forestry rallies. [26] These were “outbursts of emotion from the common people”, which the election echoed, but itself remained clearly distinct, framed as a state ritual. [27] These “electoral disturbances” draw our attention to the act of voting, in particular to the conditions of its social and territorial appropriation. Certain pieces of received wisdom may get a rough ride in the process, such as the idea that the election was immediately ‘democratic’, or that the ballot unreservedly assumed national importance. [28]

Figure 1

The distribution of types of electoral violence and their outcomes

The distribution of types of electoral violence and their outcomes

9The kinds of conflict inspired by the vote can be categorized into five different types of violence on a scale of their relationship to politics. Category one consists of physical attacks against poll workers and, more generally, against figures of public order (mayors, justices of the peace, the National Guard). These acts of rebellion were most often rooted in the repertoire of communitarian politics. [29] Directly challenging the state, these acts resulted in high injury and death counts. The second category was material violence against electoral equipment: the smashing or theft of ballot boxes as well as the destruction of minute books and bundles of counted votes. These acts, targeting symbols of electoral democracy, had varying motives but all lead to a rejection of national interdependence, which was understood to oppose regional or class solidarity.

10The third category consists of brawls between groups of voters or electoral officials. It highlights the difficulty of passing from struggle to competition, that is to say from thinking of one’s opponent as an enemy to considering him as an adversary, a rival who knew both how to behave and how to control himself. If the orderliness of the vote was threatened in both of these first two categories, what is at stake in the third category is the competition between groups and candidates, with all its excesses and escalations of violence; an attitude that electoral politics tried to domesticate by stigmatizing extreme conduct ( “we do not fight, we count”). The fourth category includes a different form of violence: altercations tied to coercive voting, especially intimidation used by officials and notables to extort ballots. This category of electoral violence is truly political: at stake was the recognition of the vote as an opinion that belongs to an individual and is kept secret. Finally, the fifth and last category consists of clashes tied to the presence of “seditious” signs, chants, or symbols. These examples of “voice” (Hirschman’s term) amount to symbolic violence to the staging of the vote (recall that, etymologically speaking, violence comes from the Latin vis, meaning the transgression of a proscription or a boundary). Hostile to all delegation of authority, this category of violence strongly broke with the idea of the electoral bargain. In distancing themselves from physical violence, these acts, isolated and anonymous, were structurally closer to political opinions than was violence against ballot boxes or justices of the peace. They were public statements that targeted the standardized rituals of the polls.

Figure 2

Typology of forms of politicized violence – Insurrectional citizenship/electoral citizenship

Typology of forms of politicized violence – Insurrectional citizenship/electoral citizenship

11During the vote, many feared the unruliness of uneducated populations. Others feared the opposite – as dependents, would these voters be putting their vote into the hands of the “powerful”? Beyond concerns over how people would vote, however, one question was asked by everyone: would the most radical groups respect the will of the majority without firing a shot? There were other repertoires for political action, repertoires which were still available to these radical groups regardless of the outcome of censitary suffrage. [30], [31] The idea of an “archaic resistance” or of a “conspiracy of clubs” that is often mentioned in authority accounts (especially in their reports on violence) must not distract us from the essential. A good number of these conflicts were not “instigated”; an examination of the data demonstrates that they were a result of the same logic that is behind universal suffrage, a logic that has become a kind of state ritual in France. [32]

12One example: in Amiens, more than 4,000 municipal workers demanded to be paid for the Monday, which they had spent at the election. This claim ran into a problem: Easter Monday had never been a working day in the town, and therefore had never been salaried. The refusal of the mayor quickly provoked rallies and the raising of several barricades. It led to eighteen arrests, but also the death of a young worker, struck by a bullet fired from the rows of troops. [33] For the democratic-socialists, the integrity of the notion of “the people” required that they themselves administer the elections. In the event, it was the state authorities, however, who took charge. In 1848, there was no longer an electoral college nor assembly of electors, but simply a “polling station”; a room for the elections dominated by the ritualistic figures of a state which, in a context of social revolt due to the crisis of the manufacturing industry and the abrupt slowdown of agricultural sales, became a target for local resistance.

A state ritual

13At the national level, a third of the violence reported fell within the category of insurrectional citizenship. This is the most important category of this sample: more than 20 departments are involved, and it resulted in nearly half of all reported injuries, nearly all of the deaths, and an overwhelming majority of the arrests. These rebellions primarily struck against the polling stations, symbol of the state. Forgotten was the law of 19 April 1831 that established a return to the principle so dear to the French Revolution: the election of poll workers. At the beginning of 1848, the principle was abolished. The instructions of the provisional government given on 8 and 10 March 1848 were clear: the polling station was henceforth to be presided over by the justice of the peace in each canton, or if he were not available, one of his deputies; a significant position when one realizes it meant that the returning officer, in charge of the police, could request the assistance of armed forces. [34] This power also marked him as a broker of the state.

Figure 3

Analytical structure of electoral violence

Analytical structure of electoral violence

14Each case of electoral violence roughly followed the same pattern: the conflict responds to an injustice that is attributed directly or indirectly to constitutional power, which sparks a gesture of protest. Then comes the mediation of local authorities, who are called upon to deliver not the kind of verdict that occurs with a vote, that is to say a public, deferred abstract, legal decision, but to deliver personal reparation, the kind that is collective, immediate, and symbolic: the kind of reparation that is capable of restoring honor to each one of the protagonists. Since the logic of elections was unable to satisfy this desire, rioting was reignited: the call to the barricades, to take up shovels, pickaxes, and clamps to dig up paving stones from the street or brutalize others, all accompanied by the ringing of church bells in alarm or the redoubled cries of “To arms!” Shotguns, sharpened files, pikes, sticks, and bayonets evidenced the refusal of the people to negotiate; or, at any rate, their distance from the principle of a legally articulated national will.

15While the disqualification of violence did not happen in a linear fashion, universal suffrage still acted as a catalyst for that process. If we think of it in terms of framing theory, we could say with Snow, Rochford, Worden and Benford that the March 5 decree was an example of “frame bridging”: it linked together two or more interpretive frames that were “ideologically congruent but structurally unconnected” on the question of legitimate repertoires of political action. [35] Electoral interdependence signaled a growing delegitimization of violence among all the sectors of the population involved in the ritual of voting. [36] For example, it was against this interdependence that the barricades in Rouen were raised. Led by textile workers and shopkeepers, the insurrection was a way of “forcefully asking for what suffrage had refused them”. [37] This head-on conflict was particularly intense. From figures from the public prosecutor, we can count 39 deaths, 53 injured, and 521 arrests, including more than 315 charges brought. In Elbeuf, where the insurrection had spread, the figures are similar: 3 dead, 20 wounded, 140 arrests. The refusal to accept electoral defeat was not just an episode of social struggle; it was also the rejection of the state’s stranglehold on the ideals of the revolution. [38] Described by Blanqui as another “Saint-Bartholomew against the workers”, the repression was denounced as having been “plotted in secret by the royalist bourgeoisie of Rouen”. This, however, forgets the infantry troops and cavalry of General Gérard: a public force requested by the government of the February revolution.

16Ideological reversal? Far from it. The intervention of the forces of order had a more structural explanation. The state was directly involved with the staging of elections. It was a systematic and premeditated presence – National Guard or police units were stationed near most polls, serving as a guarantor of “orderliness” during the civic drama of the vote. In France, elections are a state ritual. This is why the National Guard, the symbol of the unity of the country and a rampart against “barbarous customs” (Guizot’s “mœurs barbares”), has a role in them. [39] This had a downside, however: it dramatized the smallest hostility, and provoked multiple confrontations. Out of the 108 incidents mapped by Rémi Gossez during the summer of 1848 in response to the “45 centimes” tax increase, only one quarter of them became rebellious confrontations. The reason: the distancing of the police from hotbeds of protest. [40] Another example: out of the 120 cases of troubles forestiers (forestry unrest) documented by Nadine Vivier for the same period, consisting of disturbances centered on the Pyrenees and the mountains on the northeastern border, only twelve ended in open hostilities against the police. Elections, on the other hand, put populations into direct contact with the bureaucracy of the nation-state. It is a drama that is played out under the vigilant eyes of the forces of order. The smallest brawl brought about an intervention from the National Guard, if not the police or even troops, depending on the intensity of the fighting. The general recognition of the right to vote, in this respect, should be seen alongside other processes which were brought under state control, like direct taxation, military conscription, or mass education. All are, after all, part of the universalization of obligations to the state. [41]

Figure 4

Socio-professional characteristics of 79 protesters convicted of having taken up arms in Rouen, April 1848 [42]

Socio-professional characteristics of 79 protesters convicted of having taken up arms in Rouen, April 1848 [42]

17The rejection of the normative order of elections was expressed most clearly by the theft of ballots, or the election minute-books. Regarded as sacred by the electoral code, these pieces of voting equipment were, for exactly that reason, deliberately profaned during voting days. In Villeneuve d’Aveyron, a group of voters sent the National Guard fleeing before taking the ballot box and burning it. They then headed for the house of the deputy judge who presided over the polling station. They broke furniture, cut down trees, and threatened to start fires – in the end, there were three arrests. [46] In Carhaix, the polling station was occupied by “troublemakers”: the ballot box was overturned and several men were wounded. [47] In Oust, in the Ariège, “the state representative” was overseeing voting from the commune of Seix, when several voters, after having protested against the process, burst into the polling station and destroyed it entirely, then took the ballot box back home with them. Although manhandled, the returning officer was nonetheless eventually saved, but this was not the case for one of the political leaders of the riot who, that very night, was gunned down. [48]

18In the large towns, the potential for insurrection was controlled in the days preceding the election, as seen in the aborted attempts at vandalism in polling stations in Paris, Marseille and Lyon. [49] The success of the election’s organizers was primarily due to having dispersed the alreadydivided “democ-soc” movement in large urban centers. The only notable exception, along with Rouen, was Limoges, whose departmental census was interrupted by several hundred workers and “ponticauds”; the town was controlled by the rioters for fourteen days. [50] The violation of a ballot box was described by the prosecutor in this radical paradox: “The will of the people”, he wrote, “finds itself handed over, defenseless, to the whims and caprice of three to four thousand men who are more or less fanatical, and for the most part from the popular classes”. After shouting at two men about “the gravity of the crime they were committing”, he was told in response, “We want our five candidates from the working class, and we will have them. After all, the worker is everything – you, you are nothing. We want to conduct our own affairs ourselves. We have no need for you, go away.” The judge concluded: “I had to protest this violation of the rights, liberty, and sovereignty of the people by my silence.” [51]

Figure 5

Socio-professional characteristics of the 45 protestors convicted of electoral violence in April 1848 in Limoges [52]

Socio-professional characteristics of the 45 protestors convicted of electoral violence in April 1848 in Limoges [52]

19Max Weber observed this in the chapter he wrote about the idea of the market: the expansion of exchange goes hand in hand with relative pacification. [53] We could add that electoral change obeys the same logic. Once available to all, it implies a constitutional order that “disarms” violence. This constitutional order is perhaps the basic precondition for its establishment. Electoral competition implies the renunciation of taking power by force. It disqualifies a certain kind of political engineering: that which long endured among the working-class milieus of shop workers, industrial factory workers and artisans, and which used the street as a stage and the crowd as actors.

Challenging interdependence

20«Everyone finds a voice in the majority”: this majoritarian logic speaks to the power of accreditation through voting, and thereby delegitimizes “direct action”. For the working classes, 1848 sanctioned this new state of affairs: it was necessary to submit to the law of the greatest number, and to leave decisions to a majority rule extended to the whole of the People. A division of political work should result from it that allowed individual expression of consent, while opening up electoral competition to representatives who were more socially heterogeneous. Three corollaries were associated with this style of democracy, which was synonymous with a more open-ended bargain (political promises in exchange for electoral support): the idea of a (political) opinion as the rational choice of an individual; the identification of aggregated votes as the will of the people; and the notion that the numerical equality of the vote ( “one man, one vote”) forms the very foundation of the dignity of citizenship. It is precisely this electoralization of politics that, with the arrival of new groups of voters and the canvassing of votes, consecrated the new authority of this turn to suffrage.

21Of course, not everyone submitted to this authority, but everyone was obliged to recognize it. The universalized order of the vote came at a price: it required people to take part, but within the limits of a delegation of power which was put forward as the opposite of violence; within the limits, also, of a reciprocal engagement to respect the results of the vote. This was a difficult challenge for those supporters of democracy who came to settle conflicts rather than institutionalize them. When the electoral outcome is no longer recognized, the struggle is no longer contained and spills out into the public arena. In Rodez, on April 28, the result was greeted by a rather vocal annoyance expressed toward the clergy, to whose debts they wholeheartedly agreed to attribute it. This caused an “electoral uproar”: the windows of the elected deputy were broken by stones, the door of the bishop’s palace was broken down and the minor seminary was lain waste. [54] In Langon, supporters of one candidate simply refused to accept defeat. They took it out on the houses of “enemy voters”, a violent reaction kindled by the conviction that “the bourgeoisie and the landowners wanted to re-establish tithes”. [55] Opponents of the results of the vote found it even harder to come to terms with because the winners, losing all sense of moderation, provoked their adversaries. Clamors and rallies thus turned into “disturbances” or “assault and battery”.

22In Nîmes and elsewhere, the geography of violence marked the civic limits of the election ritual: the public square, the school yard, the area immediately surrounding bars, or the entrance to the town. Generally, the street was the site of a rebellious citizenry, a citizenry opposed to the abstraction of the vote. So it was by running all over town after the results were announced that most people showed their commitment to this militant geography. The electoral party, with its acts of bravado and songs, sometimes created the opportunity for release to make up for the frustrations created by the regimented procedure of counting votes. Why did people continue to question the vote? The vote was not seen as an expression of opinion right away. In 1848, it was still largely a technique of ratification. [60] This is why emphasis was placed on the approval dimension of the vote: voting enabled, locally, the magnification of a verdict that became the social seal of approval for an authority or a group.

23Is this nostalgia for non-electoral democratic practices? We must not forget the failures of the vote itself. In 1848, the codification of the electoral system was embryonic: repeated offenses and collusion were not punished; as for the statute of limitations, it was limited to three months. But there was more. A tradition since 1789, the electoral oath was abolished, and with it, a registration technique that placed at the very least a moral limit on the expression of “electoral passions”; a limit that, in 1848, was the only form of civic socialization. [65] This was similar, in practice, to a limit on public discourse. If discursive violence was the target in particular, it is because it became more appreciable: henceforth, electoral rhetoric was aimed at voters in the popular classes. One single example of this policing of opinions: the condemnation by the criminal court of Vienne, on 10 August 1848, of Louis Charles de Saint Savin, a 32-year-old landowner, who, with the help of his personal fortune (an income of more than 80,000 francs), had spread throughout the neighborhood of Civray the call to “Trim the frock coats to lengthen the vests”. His discourse “against the rich” was judged too fanatical: “The rich are like tall trees that, through the shade they cast, prevent the little ones from prospering. Prune the tall, cut them down by their feet and by the head, and the little ones will prosper.” And what about majority rule? “I will not listen to the majority of the Chamber if it does not share my opinion; we will revolt and we will throw the members of this majority into the Seine.” It was discourse, but also action, because the man encouraged his friends to “cut the throat of a landowning rival” and to “hang another from the tree of liberty”. [66] The role of civic socialization must also be understood in the context of the administration of the vote: with the elections held in each department, across a large number of cantons, and sometimes with polling subdivisions, control over the vote was dispersed. No longer could there be a focus on a small number of voters brought together in neighborhood electoral colleges. This was a bureaucratic extension of the vote as outlined in several prefectoral declarations; [67] it reached even the most far-flung hamlets, stifled the hubs of self-legitimation and, with them, one of the foundational elements of communitarian revolt. One thing is clear: through this normative interdependence, local order could externalize itself, and new deterrents came into effect.

24The forced substitution of ballots also conveyed a rejection of the mathematical sovereignty of suffrage. In Beaune, when workers presented printed ballots “to their brothers, the voters of Volnay”, the priest had them ripped up, “considering that the democrats who brought them there seemed too progressive to him”. Blows with fists and sticks were exchanged, until a picket from the artillery intervened. [68] In Levroux (Indre), when voters from Verneuil brought “ballots that had been written in advance”, electoral agents “either through intimidation or violence ripped up their ballots and substituted new ones”. These were common occurrences. In the countryside, and even in certain industrial centers, it was less of a deliberative election process than it was the mere submission of ballots, because it relied on the mobilization of prominent individuals more than on a sense of civic duty. [69] What of “electoral choice”? It resulted less from partisan splits or from opinion than from prevailing conformity – that is to say, a complicity of place or circumstance, or a shared religious solidarity or material relationship. We might allude to the temptation to ‘influence’ voters using the social domination they experienced on an everyday basis as a lever, but once again we would lose sight of the political reasoning at work here.

25With the rejection of the assembly structure, of which the end of the electoral colleges of the July Monarchy was a part, the vote was separated into two distinct sequences: on the one hand, the material expression of suffrage, henceforth synonymous with casting a vote; on the other hand, the deliberation of the voter that happened outside the polling station, and was subject to “influences” and “pressure”. [70] Strictly speaking, the notion of the electoral campaign was born from this fundamental split… except it was in no way regulated during the spring of 1848. If bureaucratic organization maintained a tight hold on the polling stations, the electoral battle itself was allowed to run unchecked, and notions of respect for adversaries and the autonomy of the voter were largely ignored. Neither the principle of declaring one’s candidacy, nor the financing of canvassing operations, for example, forged the kind of electoral competition that could be separate from relationships of domination. The destruction of statements of faith, aggression towards those distributing pamphlets, slander and libel in all forms: these attacks on to the fairness of the competition were rarely punished, or were punished by common law. [71] The parliamentary laws concerning the “verification of credentials” did not offer any more security: “It is not necessary to focus on protests that make accusations of violence or fraud, without presenting any facts or explanation… if a large number of votes separates the winning candidate from his closest rival, where the latter is the one making the complaint.” [72] In other words, disorder and violence became a motivation for annulling a vote only if the difference in the number of votes was so narrow as to suggest that such disorder might have had a direct impact on the results. [73] This is the logic of the decision the court returned in a case that had been brought to it the day after the vote in Privas.

26

“The fact that the polling station was surrounded by people who engaged in acts of violence against voters, forcing them to show their ballot papers and tearing up those they didn’t like; or even that threats were issued, blows exchanged and an arbitrary arrest made, does not in itself nullify the election if, on counting all the votes from the college where these events took place, the winning candidate in the election in question is found still to be in receipt of several thousand more votes than the candidates who came in behind him.” [74]

27The weakness of this codification was obvious: the equality of citizens was not yet the same as the equality of candidates. This pattern is also quite clear concerning ballots. The administration did not yet have a monopoly on the physical production of ballot papers, and pushed for standardized, legal ballots. [75] But such standardized ballots were in competition with private ballots that, printed with candidates’ money, were denounced by the commissioners of the Republic as “inundating citizens”. According to the representative of the Lower Loire, “priests and noblemen” were behind the “150,000 ballots printed by press or lithograph” and distributed throughout the department. Hence his plea: “If it were possible to forbid the use of these ballots, it would perhaps make victory for our enemies less easy.” [76] A whole range of practices was developed to combat these rival distributions: from the technique of colored, numbered, or stamped copies, to the requirement that voters wear pants without pockets, to the use of “double-sided ballots” (billets à double face) that served to foil the stratagems of the unscrupulous. A reflection of unregulated competition, these techniques occasionally gave rise to brawls and litigation. Since they were incapable of pursuing those who acted in this way, the public authorities opted for clemency. There was no law addressing this level of practices which remained therefore outside the circle of electoral violence.

A restrained struggle

28Comparing scenes of voting to those of violence should not cause us to forget how much tension is created by the split of the vote itself. Tension is part of the very nature of the electoral operation. As the prosecutor of Bordeaux stated, an election is “unrest, but justified, reasonable, legal, unrest that leads to the peaceful gathering of citizens who seek to be heard so that their various candidates may win”. [77] A candidate in Creuse, Émile de Girardin, was more radical: according to him, an election “is civil war, with this one difference that instead of using cartridges, we use ballots”; [78] which amounts to describing elections as a restrained battle.

29Depending on which way you looked at it, emphasis was placed on one version of elections or the other. In La Châtre, according to the prosecutor, near the polling station of the first voting sub-section, there was an office intended for ripping up “democratic ballots”: they were “to replace what had been taken away from the voters”. There was also no shortage of intimidation. The household of government commissioner Fleury was besieged by 200 people; George Sand’s residence was “threatened with fire and death”; the father-in-law of general secretary of the prefecture – whom Charles Delaveau, mayor and deputy of La Châtre, “had publicly denounced as his enemy” – was also threatened and told that he should “withdraw to the countryside and abstain from voting in order to save his life”. In the canton of Sainte-Sévère, the violence had been, in the opinion of the prosecutor, “widespread”. There were insults against “democratic voters” (meaning those whose distrust inspired them to “make their own ballots”), and blows were exchanged with the “recalcitrant ones” (one of them would be left for “nearly dead”). As the magistrate noted, “The ardor with which each one wants to make his political opinion count in these solemn circumstances in which we find ourselves naturally excites the spirit.” [79] We find the same view of the situation in the canton of Saint-Pons, where the election was held “under an atmosphere of terror organized by the local authorities”. Acts of violence were so numerous that the police commissioner worried, “civil war has essentially become the true nature of the election”. [80] And, in fact, a number of voters did not dare to go to the chef-lieu where the Mayor of Vellieux was nearly assassinated during the elections of April 23.

30In 1848, the observations of commissioners and prosecutors generally tended to downplay the public expression of violence. A strategic reaction? Undoubtedly. But that does not mean we should exclude empathy with the rioters as a possible explanation, given the social conditions of these people as well as their political orientation, which was often on the extreme left. What the Orleanist powers hastened to denounce as archaic behavior – the image of a barbarous multitude rising up in the middle of civilization [81] – the republicans described in far more banal terms. Hurled rocks and beatings with sticks? “Nothing worse”, for this commissioner, “than what ordinarily happens during the festivals and markets in Finistère”. [82] In his report, the prosecutor of Riom remarked, laconically: “Some disorder in the neighborhood of Thiers in St. Rémy; and in the neighborhood of Clermont in Olby, peasants from different communes who were returning from voting encountered one another, and a fistfight with sticks and rocks ensued.” Yet twenty people were wounded, three seriously. [83] While some people could see in these incidents a preparation for armed struggle, few really looked into the true significance of these acts, and even fewer have considered the connection between these revolts and the social context that inspired them.

31And yet the reality of electoral violence surpasses the language of disapproval and forgiveness, terms in which it is almost always framed. For example: between the villages of Guchen and Ancizan, at the foothills of the peak of Arbizon, the dispute went back to 1830 when the government decided to give a flag to the Aure valley and Ancizan promptly seized it. On April 24, the day of the vote, the residents of this village proudly displayed this symbol of their superiority, to the great displeasure of the voters of Guchen, who were scandalized by the parading of a “royalist” flag. After casting their ballots, the voters of the two villages encountered one another on the main road. There was an attempt to chalk the republican motto onto one of the flags, and this was provocation enough for a fight to break out. [84] The conflict was also informed by social structure: Guchen, along with the villages of Ancizan, Cadéac, and Grézian was part of a pastoral association called the Quatre Véziaux [Four Neighbors]. The shared land, possessed in joint ownership, consisted of nearly 3,000 hectares of mountain pastures, upriver from the valley of Campan, on the village lands of Ancizan. This supremacy was contested by Guchen, one of the larger villages of the valley (nearly 700 residents compared to a thousand in Ancizan) and almost a small market town, especially since the development of its textile industry. Historians, focused on the campaigns of political clubs or the administrative supervision of the vote, have paid little attention to this dimension, at the risk of leaving unexplored the way in which expressions of disagreement struggled to become differences in opinion in the spring of 1848.

32On the day of the vote, plenty of opportunities for fighting presented themselves: during discussions in bars, on the road home, in waiting lines, or in front of the polling station. Who were the people fighting? For the most part, they were young men. What caused the fighting? Often an attack on someone’s honor that brought tension between villages to the surface, where a group would want to “settle its own scores”. The government commissioners, not well informed about life in the countryside, attributed this kind of behavior to “long hours waiting in bars” or to the importance of “tradition”. [85]

33Each election is a battle, but a battle whose political nature transforms social energy. The point merits review. The regularity of elections does not signify that they are fair, simply that they bear the signs of conformity required by the law. “Regular” and “fair” are not the same things. This requires us to accept that elections do not create a separation between the political and the social, because in the process of the election, one reinforces the other. Struggle is only possible when there is contention, but fairness does not prevent the thirst for victory. The evolution of electoral politics bears the traces of this, providing candidates and voters alike the opportunity to do things right: to act, and to claim one’s actions as being, within acceptable boundaries. But this does not mean that non-political motivations are absent from the ways people behave on election day.

34Explaining electoral turmoil often means relating it back to the ordinary violence of social life. Of course, there are rarely deaths involved, or even injuries. Usually it’s nothing more than intimidation, or last-minute plans. But there could also be “overt force”, as the police reports mentioned: this was notably the case when inter-village brawls or popular rebellions re-emerged. However, electoral politics were supposed to diminish this form of expression. Didn’t this type of politics introduce structured and visible opposition parties? Relayed through the authority of prominent bourgeois citizens, codified by electoral law, electoral politics channeled disagreements (over property boundaries, rights of passage, questions of irrigation, even rivalries between candidates) by subjecting them to the rhythm of regular elections. Hence the progressive pacification of customs and conflicts.

35Let us turn to the final category of electoral violence: “seditious” chants and symbols. They are a separate category that, it should be noted, explicitly breaks with the notion of electoral change. These rebellious forms of expression did not belong within republican institutions, within the idea of popular universal suffrage as ordained by the March 5 decree; but were directed at these institutions. The dissenting voices were raised away from the polling stations, and refused moreover to become ‘electoral voices’. They were anonymous, public, and pithy, and diametrically opposed to the idea of making one’s voice heard through the ballot box. In Clouay, on the evening of April 24, at the moment when the voters of the village of Buby in the neighborhood of Lorient were leaving the polling station, a cry of “Long live Henry V!” [86] rang out several times. [87] On the morning of April 23 in the village of Coex, in the neighborhood of Sables d’Olonnes, a white flag floated at the top of a poplar tree, “as if to intimidate those people who left to vote”, noted the prosecutor. [88] The white flag was seen again, but this time displayed by the voters of the village of Saffré who went to Nozay for the election. [89] Such were the signs that were perceived as “disturbing the public peace”.

Figure 6

Substantiated cases of electoral violence during the elections of 23 April 1848

Substantiated cases of electoral violence during the elections of 23 April 1848

36The map of electoral violence does not really with the traditional areas where tax revolts took place. Generally, this geography of “electoral collusion” suggests the opposite of what we typically understand about regional temperaments. Of course, there is a great temptation to attribute these incidents of electoral violence to a local “propensity” or “penchant”. But, when looking at the data, what is striking is how the localization of electoral violence is tied to something else: the interaction of militant mobilization with state intervention; even more so because the replacement of mayors during the spring had led to a number of frustrations. Although the areas where the Montagnards were strong stand out, [90] as do the traditional bastions of revolt against the state, the areas of “electoral violence” do not map onto them, nor onto the areas where the forestry protests took place in the spring, or where the rebellion against the 45-centimes tax happened, or the riots against the banning of hunting, except in a small number of cantons in the Dordogne, Alsace, Bretagne, or the Pyrenean valleys.

37On voting days, seditious printed material found an echo in whispers and rumors: “news” that, while false, aroused an atmosphere of anxiety. On April 23, two posters were found in public squares and on the walls of Noirmoutier’s town hall. One stated “Good citizens of Noirmoutier, the worker, with his face to the ground and no idea of what is to come, demands work or to die straightaway!” The other stated, “Messieurs les Bourgeois of Noirmoutier, we ask you to pay close attention to the poverty of the worker: work and bread or death!” [91] In Bordeaux, when the voting had just begun, on all the walls of town could be found a poster lashing out at the government and its representative in the department, who was running for parliamentary election: “Voters… Do you want to send to the national assembly another nonentity who will cost France 10,000 francs a year for saying nothing? Do you want to sanction the savage dictators who were wicked enough to send into several departments escaped convicts and assassins with unlimited powers for insurrection?” [92] Fears about church bells being melted, statues of saints being burnt, “communist” intrigues, looters ready to invade the commune: these images of “threat” express specifically rural anxieties about the integrity of the community being strained. The predominant fear in the towns was of election side-effects: volatile political movements were hard to distinguish from the outbursts of a crowd. Weren’t they likely to forget their early resolutions of good behavior? This is what the commissioner in Toulouse feared, where a poster for the Democratic Club sowed terror by announcing that all citizens 21 years and older could be members of the National Guard. [93] In Riom, a piece of news that was “spread by the mail coach from Paris” provoked anxiety: “a group of 50,000 communists had come to blows with members of the Parisian National Guard the day before”. [94] It is not by accident, then, that following the June Days Uprising there was a significant intensification of the fight against “those who chant, sing and affix posters”. Peddlers of “false information”, and “the display or distribution of signs or symbols intended to spread the spirit of rebellion or to disturb the public peace” [95] were the particular targets of law enforcement in a society where suffrage was henceforth universalized.

Electoral sovereignty

38What could be legitimate about the recourse to violence in 1848, at the very moment when the people declared themselves sovereign through their ballots? The culturalist approach answers this question with a stock response: the weakness of pluralist electoral culture. It is not surprising that historical sociology prefers to explore other explanations, as well as to highlight how each election created a unique new situation, thanks to the orchestration of numerous activities and specialized services by the state. While bureaucratization, at least within consolidated democracies, made elections routine procedures, the result of those procedures remained a source of tension and uncertainty. Tocqueville estimated that each presidential election in America in the 1830s generated “a period of national crisis”, where the “entire nation falls into a feverish state”. What is certain is that elections suspend the objective order of public institutions. We have seen it here: the election appeared in 1848 to be a mathematical division, a competition prescribed by a certain number of conventional abstractions – the counting of votes, rule of the majority, putting political aspirations to the electoral test. To understand how it was accomplished still means, even today, recognizing these innovations, which were sources of specific tension, and linking them to what characterizes this form of electoral competition: the battle to increase one’s share of the vote.

39While the vote of April 23 effectively closed the door on the practice of voting by assembly, [96] the process of individualization was just beginning. No rules were established for supervising electoral campaigns or a system of candidacy. This amplified the risks surrounding the ballot box. Some might claim that the standardization of voting procedures was very advanced, [97] or that transferring the creation of ballot papers away from the polling stations shortened the time it took to vote. But such procedures came with a drawback, which the ranking of the communes or the alphabetized list of voters only served to exacerbate: that they increased the opportunities for intimidation right outside the polling stations themselves. It has to be admitted that the individualization of suffrage (which contemporaries referred to as the “isolated vote”) was not then perceived as the affirmation of autonomy. In the eyes of many, individual suffrage was a result of “backwardness”, or a “lack of mobilization”, just like those “latecomers” who were admitted to vote at the end of each election day. To enter directly into a relationship with each citizen was, however, the goal of April 23’s election. Individualizing opinions – through the use of handwritten ballots or placing the polls in the administrative center of cantons – was to strive to neutralize community hierarchies. In a nutshell, it undermined traditional organic representation.

40Pagès, a winning candidate in the Haute-Garonne, points this out in his proclamation published in Le Journal de Toulouse on the day of his departure for the capital: “[The workers] have today made their entry into electoral life, but have not been able, despite their numbers and despite their associations and their alliances, to enter into political life. At fault is electoral law, written under their eyes and practically for them. This law admits all citizens but only as individuals, and the election is an affair of figures and numbers. If ever an electoral law were decided by profession, workers would make up a necessary part of national representation.” In 1848, electoral change collided with the mobilization of community and business leaders, who rekindled their influence over professions and communities. For them, the vote did not so much present the possibility to decide between competitors as the chance to see their “natural representatives” ratified. Hence the often unanimous character of votes from voters who arrived at the polling stations in processions, singing parish hymns or workers from the same guild bearing their shared flag.

41All the same, even in this hierarchical universe, each voter was declared to be equal to the next. More importantly, into the confrontations between village communities or professions, the possibility of a personal expression of political volition had been introduced; a focus on individual conscience which was magnified after 1848 as the source of the “wisdom of voters”, if not their “electoral dignity”. [98] This claim was formulated as much by some officials of the state as by those who upheld the old order who were worried about the pressure applied by the government. [99] It seems as if voting was above all else the display of an interest in and an aptitude for political classification. It is a principle that implies that the value of the individual is independent from all other allegiances. Thus the democratic fiction that one citizen is indistinguishable from another. [100]

42If violence was was what spoke loudest to men in 1848, then this was as a result of these normative factors; not just to the specific context (the social rifts unique to a revolutionary period), or the arrival of a new cohort of voters (the expansion of suffrage to nine million voters). The use of weapons, whether overtly or covertly, the crowds and the threatening demonstrations, the distortion of votes and the seditious symbols: all these kinds of violence were delegitimized by a civility promoted as ultima ratio, or the last resort, of democracy. [101] This brought to light the divisions that existed within each community, and at the same time both aroused them and transformed them into a purely mathematical rivalry. In addition to promising to “institutionalize uncertainty”, [102] this electoral interdependence signified that the ballot box could be a legitimate measure of popular consent. Even better: that each voice could become, with expanded suffrage, an anonymous unity, abstract and perfectly interchangeable. Of course, the voters did not come to choose. They came to vote. The events of 1848 – whether it was the elections of April 23, the municipal elections of July, the cantonal elections of August, or the presidential elections of December – restored a largely communitarian spirit. But the tallying of votes obeyed a mathematical objectivity, that of electoral sovereignty that, in distancing itself from the procedures of deliberation that were dear to the systems of vote by assembly and censitary suffrage, truly paved the way for a vote by citizens.

43The history of the relationship between voting and violence cannot, then, be written in the simple, if not simplistic, terms of prohibition and legality, even less so in terms of the “choice” between the ballot box and the shotgun. [103] Not that it is not a question here of the slow but orderly creation of a normative division. But it is very clear that this history is, first and foremost, that of the creation of new social rules, the most important being that which relates to the norms of socialization in electoral competition. [104] If there exists a legitimate exercise of political competition, it is because social norms disqualify transgressive practices as contrary to its autonomous functioning. That said, electoral democracy does not settle for punishing deviations from normative constructs; it absorbs their influence, such that deviance in turn plays a part in the constraining and even the codification of the vote. [105] Of course, it needs time for the ballot to constitute true opinion, that is to say for it to become part of the ongoing relationship of exchange within this space of ordered and pacified competition that we call electoral democracy. In the meantime, universal suffrage is a first step, which did not only declare a right: it began the process of learning about how interdependent politics – previously untested – might work, while naturalizing the act of making reference to it.