• 1 . Refugee numbers come from “Note sur les problèmes d’emploi et de logement dans les villes d’Algér (...)

1I approach this collection’s discussion of war and imprisonment through the example of the forced displacement and confinement of rural Algerians in the war and revolution of 1954-62. During this time at least 2. 4 million peasants were forcibly relocated or “regrouped” in internment camps and another 1. 1 million rural people fled to shantytowns in cities throughout Algeria, France, and neighboring countries (Cornaton, 1998, 122-123, Kateb, 2001, 315)1. These Algerians did not suffer conventional forms of imprisonment in the sense that people in the camps had not been convicted of a crime, nor were they formally suspected of criminal activity; and in the case of the shantytown, no walls guarded people inside. But like the incarcerated criminal, the Algerians living in camps and shantytowns represented a threat to the French state, and their confinement promised to neutralize it. This threat took shape through the crisis of sovereignty France faced in its Algerian departments, sovereignty that the National Liberation Front-Army of National Liberation (FLN-ALN) contested with its secessionist agenda.

• 2 . Pierre Mendes-France cited in Todd Shepard, The Invention of Decolonization: The Algerian War and (...)
• 3 . By 1949, schools were available but enrollments were often refused because families could not pro (...)

3The question of confinement and sovereignty reveals itself between the camps and shantytowns in separate but overlapping modalities. Within the regroupment camps, France sought to inscribe its sovereignty, targeting rural people whom the French state had long held weakly, and which the rebellion brought into focus as the greatest crisis point for France’s claims in Algeria. How could the government’s slogan that “Algeria is France, not a foreign country,” ring true given the poverty and political disenfranchisement of most of its people?2 The shantytown represented a different problem than the camp, but one that was also linked to the rural crisis in as much as the people living here had fled to the city to escape the poverty, hunger, and violence they faced in the countryside. Treating shantytowns and regroupment camps together reveals certain questions that might otherwise remain elusive. As a scholar has recently remarked, when treated together, camps and shantytowns reveal the “spacial operations of counterrevolution” (Henni, 2017, p. 13). I too, place them in the same analytic frame in order to think about war and debates on discipline and sovereignty, biopower and security (Foucault, 2007, Foucault, 2008, Agamben, 1998, Mbembe, 2003, Khalili, 2013). The Algerian shantytown, an improvised housing agglomeration within or adjacent to the city, housed people who lived, literally and figuratively, off the grid. Shantytowns, by definition, had none of the amenities of modern urban life. Moreover, their inhabitants generally existed outside of the civic-administrative apparatus of city and the state. There were no schools or places of public assembly, and many people had no identity papers, escaped censuses, and did not have an état-civil, the foundation of one’s juridical and political subjectivity.3 For this reason, critics at the time called them zones of “non-droit absolu” (“absolute lawlessness”) (François Maspero in Hervo, 2001, p. 17), which confined a group of city-dwellers to whom the droit de cité (rights of citizenship) was refused (Berque, 1958). However, the state was hardly powerless in the shantytowns. It exercised, or attempted to, a form of statecraft rooted in political economy and the eighteenth-century’s “economic turn” (Kaplan and Reinert, 2019). In this respect, state power realized itself not by imposing a juridical conception of law, nor even the modern state’s Gewaltmonopol (monopoly of violence) linked to its right to kill, but by allowing demographic and economic “truths” to run their course towards maximum efficiency (Foucault, 2008).

• 4 . Jean-Philippe Talbo-Bernigaud, “Zones interdites,” Temps modernes 16, no. 177 (January 1961), 708 (...)
• 5 . Parlange, “Des regroupements aux milles villages,” n.d., ANOM 81f444.

4The regroupment camp was called a “centre de regroupement” in French, a designation that conformed to the French government’s careful avoidance of the word “camp” in official texts concerning the many different types of detention centers in wartime Algeria because of the word’s association with totalitarian regimes (Kuby, 2019, p. 197). The displacement of civilians began in the early weeks of the war, and the first regroupment camp opened in the Aurès Mountains in 1955 (Tengour, 2010, p. 199-213). This camp was an improvised affair, but it fit into the military’s effort to establish a security grid across the Aurès where the armed struggle began. This grid proved hard to transfer from the map to the ground. It relied upon indiscriminate forms of violence like airpower, as used commonly in the twentieth century (Satia, 2008, Headrick, 2010). The areas where it could be concentrated were called the Zones Interdites (Forbidden Zones). Here people faced evacuation and anyone remaining was considered an enemy: unprotected by the normal apparatus of the law, they could be shot on sight.4 After 1957, the forbidden zones expanded and control of the civilians fell upon the French security services, strictly military until 1959 (Cornaton, 1998, Sutton, 1999). At this time, it was recognized that the camps had won a military victory but suffered a political defeat, and their mission changed under shared military and civilian direction.5 Rather than an emergency response to military necessity, the camps were tasked with social reform.

• 6 . Yusuf to Ministry of War, 10 November 1837, SHD 1h53.
• 7 . Ferdinand Lapasset, “Mémoire sur la colonisation indigène,” 1848. ANOM 8h10.

5Social reform or modernization has become a hallmark of civilian confinement in 20^th century counterinsurgency warfare, even serving as “constructive war” (Feichtinger and Malinowski, 2011), but the novelty claimed by the military theorists of today is belied by a longer history of using social transformation to consolidate colonial conquest (Khalili, 2013, p. 200-212). For example, in 1837, at the beginning of France’s occupation of Algeria, General Yusuf warned his French superiors that the conquest was not complete with the city of Constantine’s fall. The crux of the problem lay in the countryside. “[T]he rural population must be brought to bend (assouplir) to our domination”6. The language that Yusuf uses here, assouplir, reveals his belief that he could not impose order by force alone. France must find the “means to transform this population,” he wrote. Yusuf’s vision realized itself in plans for “indigenous colonization” which sought to bring Algerians into French-style agricultural settlements (Bugeaud, 1847). As one officer wrote in 1846 the “essential is to group together these people who are everywhere and nowhere; the essential is to make them perceptible to us and put them within our grasp (saisissable)” (Richard, 1846, p. 191). The most successful colony was the Smala d’Aumale constructed near Ténès by Ferdinand Lapasset, an officer of the Indigenous Affairs Office. By 1848 it had 33 homes, laid out barracks-like, where 205 Algerians lived, 58 of whom served in the local militia with their families confined to the village (Yver, 1955). Lapasset’s self-presentation reveals someone who might be thought of as an early “lesser evil” humanitarian (Weitzman, 2011), a person who could resolve the question of what France would do with Algerians without mass killing. Rather than “destroying more than three million inhabitants” as many claimed necessary to complete Algeria’s conquest and colonization, he wrote, his village would “resolve the problem of dominating the country [by] using Muslim society towards our own ends”7. He imposed a strict regime of self-sufficiency, authority, and collective labor. In return, people received subsistence from collective provisions, which were stored in 19 silos, tightly controlled by the commander himself who distributed rations based on need and rank, a non-egalitarian notion of communalism typical to colonial-era Algeria (Marçot, 2012). Lapasset’s projects of docility extended well beyond the garrison space itself and the individualized, drilled body of his militiamen. In January 1851, after several years of harvest failures, locals suffered a deadly subsistence crisis and resorted to eating arum bulbs for survival. Lapasset wrote about the crisis favorably:

• 9 . Délégation générale et Commandement en chef des forces en Algérie, 20 July 1959, no. 4426, ANOM 8 (...)

6This early camp project expressed the sort of political relations that existed between the French state and Algerians in the mid nineteenth century. Apart from an anemic “convention” from 1830, France did not negotiate a social contract with Algerians; instead it arrogated power for itself and claimed sovereignty by the right of conquest, a pattern that repeated itself across Africa (Mbembe, 2001, p. 34). This left political relations between Algerians and the state as primarily within the realm of victors and vanquished; and as Hamdan Khodja (Ḥamdān b. ‘Uthmān Khawāja) famously wrote in 1833, vae victis (woe to the vanquished)! (Khodja, 1985, p. 184). By the 1950s, these relations had changed dramatically: Algerian territory had undergone a process of assimilation to the metropole, and after the Second World War Algerian Muslims themselves had become, legally-speaking, French citizens and equal before the law, even as they did not enjoy political equality and suffered from many discriminatory measures. But when the armed struggle for independence began in 1954, old reflexes showed themselves in institutions like the Zones interdites where French security forces could kill at will. This expressed not only the military’s sense of expediency but also a form of sovereignty informed by the original logic of conquest and the state’s right to take life. However, as the war progressed the state’s right to kill increasingly uncoupled itself from the law. This might have expressed a pattern typical in colonies, where the colonized were subjects rather than citizens, but Algeria was not a typical colony and this lawlessness compromised French claims that Algeria was an extraterritorial extension of the metropole. Although a variety of legal and extra-legal measures ensured that security forces continued to exercise a largely unrestricted freedom to kill typical of dirty wars (Thénault, 2007), in 1959 the territorial dimensions of legal exception represented by the Zones Interdites had to be dismantled. The order read: “The essential mission of the Army is therefore to give concrete expression to this sovereignty everywhere on this territory, by its presence and especially by the presence of the administration… The existence of the Zones interdites is contrary to this basic notion”9. The same order made provisions for continued free fire in some areas such as the Sahara and the border regions, but “the expression, ‘Zones interdites’ must disappear from the current vocabulary.” In other words, officials came to realize that the notion of an internally “forbidden” or lawless space threatened basic principles of Algeria’s legal status as part of single field of sovereignty linking Algeria and France.

• 10 . Papon, “Directive sur la lutte contre l’appareil politico-administrative rebelle” 19 June 1957, A (...)
• 11 . Délégation générale du Gouvernement en Algérie, Cabinet, “Note d’information sur les regroupement (...)
• 12 . “Des regroupements aux milles villages,” op. cit., ANOM 81f444.
• 13 . Colonel Goussault in L’Atlantique nord (June 1958), revue de presse, Sept-Oct 1958, no. 28, ANOM (...)

7This upended the original dynamic of the military’s efforts to constitute sovereignty through killing people it declared criminals, which in turn increased the stakes of the camps. Originally, interned civilians were not invested with any specific value beyond that which came with cutting “umbilical cord that linked the rebel organization to the population”10 and thereby denying the FLN/ALN this point of logistical support. As the war progressed, it was no longer possible to leave the people interned in the regroupment camps in a perpetual state of political indeterminacy, with camps serving primarily as holding pens. A robust project of social and political reeducation emerged in the camps, which became “homes for human evolution and social advancement”¹¹. Education, women’s emancipation, agricultural science, hygiene, and “participation in political life” became central terms.¹² The goal then was to produce a modern political subject, with the outward form of a loyal French citizen. In 1958 an officer expressed this task as “making the population adhere to the great thoughts upon which Western Civilization sits, to make them enter on their own will into the French Community”13. This figured within a larger effort to use economic development towards social and psychological regeneration of Algerians based on the inculcation of materialist and consumer values as shown in major development projects such as the Constantine Plan announced in 1958 (Davis, 2017). Thus, French claims depended upon producing a semblance of a social contract by which French sovereignty seemed to derive from Algerian consent. This proved an improbable scenario given the violence of the
situation: politics could not easily disentangle itself from war.

• 14 . Fatima Larit, personal communication, September 2010. Spelling from Marcelin Beaussier, Dictionna (...)
• 15 . Delouvrier administrative instructions to Prefects, 13 June 1960, no. 707/AP. SHD 1h4063.
• 16 . Carton “Toufik” with notation “je trouve cela idiot!” ANOM 93/4429.
• 17 . “Les colonialistes jouent les ruses,” Armée et Front de Libération Nationale Algérienne, May 1957 (...)
• 18 . Inspection report of Aïn Sidi Ali, 5 June 1961, SHD 1h4063.

8The effort fell apart not only under the weight of its own contradictions but because of the refusal of interned people themselves. Algerians knew the regroupment camp as a “qīṭūn” (pronounced gīṭūn), a word designating a canvas tent¹⁴. And they experienced firsthand the material failures of the camps, where they suffered extremely harsh living conditions, marked by high mortality, illness, and poor diet (Sacriste, 2016). They also witnessed the failure of the camps’ political and socialization projects, which based themselves upon making a case for the validity of the French cause in Algeria, a case that the French administration’s own ignorance, racism, and hostility compromised.¹⁵ Even French officers found that the camp propaganda designed to instill pro-French political dispositions was condescending and “idiotic”.¹⁶ For their part, the FLN-ALN correctly identified that “the confidence of a people, its personality and its dignity cannot be bought with candy, barley, or a crocodile smile”17. Moreover, Algerians protested against the conditions in the camps, staging strikes and demonstrations. In one 1961 incident at the Aïn Sidi Ali camp near Laghouat, 300 people assembled at the commander’s office to protest a wave of deaths caused by hunger and disease (50 people had died of illness in 12 days), threatening to walk out of the camp and complain directly to the prefecture’s office¹⁸. They could be killed as rebels for traveling without permission, but they argued that this was preferable to the slow death they faced in the camp. 

• 19 . Ouelhadj, Note de service, Wilaya III, 15 March 1960, SHD 1h1619.
• 20 . État Major de Guerre, Wilaya III, Rapport moral, June 1959, SHD 1h1616.
• 21 . Front et Armée de la Libération Nationale, État-Major de Guerre, Wilaya III, Rapport Moral, Decem (...)
• 22 . “Traduction de documents recoupées le 18-8-60 en PX 95 SC,” September 1960. SHD 1h1638.

9While the camps failed to win over the political loyalties of most interned people, French efforts did enjoy some successes. These reveal themselves not in the carefully staged French propaganda that came out of the camps, a Potemkin villages record (Feichtinger, 2017, p. 60), but in the FLN-ALN’s archive which clearly reveals problems they faced in the camps. The FLN-ALN’s efforts to combat the camps included forcing civilians slated for regroupment to evacuate to Tunisia, as well as reserving exemplary punishments for “traitors” and the especially feared French officers of the Sections administratives spécialisées (SAS) who ran many camps (Mathias, 1998)¹⁹. The fact that the Algerians interned in camps were overwhelmingly women increased their fears. The FLN-ALN recognized that sexual violence had served France as a weapon to divide Algerian society (raping women compromised the family structure) as it later did in other contexts (Shalhoub-Kevorkian, 2009), and they worried that the same tactic would be used in the camps²⁰. These fears manifest themselves in ALN reports about girls recruited as French spies in the schools and women who had been drugged and raped by the French, forcing them into prostitution and alcoholism21. Such women were considered so dangerous, the documents suggest, that they were subject to execution²². In short, FLN-ALN documents show that they placed a priority on women as an important site for their own claims to sovereignty. Apart from the logistical challenges posed by the camps, which cut off a significant supply line, they were on the front lines of larger battle concerned with women’s sexuality and the family (Vince, 2015; MacMaster, 2009).

• 23 . The average per decade increase in the period 1891-1921 was 22%. René Lespès, “Projet d’enquête s (...)
• 24 . R. P. Letellier, “Le Prolétariat indigène musulman,” 3 June 1946, ANOM 10APOM815; Letellier, “Les (...)
• 25 . “Note pour M. le Préfet,” 10 Febuary 1951, no. 153.1/PR, ANOM 10i9; United Nations Human Settleme (...)
• 26 . “Note pour M. le Préfet,” 10 February 1951, no. 153.1/PR, ANOM 10i9; Çelik, 1997, p. 202.

11The rural crisis put people in motion. Poverty forced hundreds of thousands of people to the coastal cities, inland towns, and metropolitan France. The Algerian population of Algerian cities over 10,000 inhabitants grew 54.5% in the decade between 1921-31, doubling the pre-war growth rate23. In Algiers, migrants first found housing in the Casbah, the old Ottoman section of the city. Although a few families had lived here continuously since 1830, by the mid-twentieth century, the old quarter was home to about 60,000 new arrivals24. The Casbah had all of the features that define modern slums: poor access to water, no sewage or sanitary services, high rents and insecure contracts, and, of course, poverty, precarious employment, disease (tuberculosis and syphilis), and high mortality25. It was also extremely crowded. Algiers had been a densely populated city back in Ottoman times: when the French first arrived, the people of Algiers lived about 500 to the hectare (Berque, 1958, p. 21). The colonial era saw the Casbah’s population skyrocket with average densities reaching
3 333/ha by 1951, and some neighborhoods in the upper Casbah numbering
3 800/ha26. At this time, people used every part of the Casbah as sleeping and living quarters: roofs, workshops, courtyards, stairways and halls, balconies, cellars, and the street. With the rural insecurity caused by the Algerian war this density further increased to 4 885/ha by 1966 (Benatia, 1978, p. 68). This made it one of the most densely populated slums of the twentieth century.

• 27 . “Recensement des bidonvilles, Alger Centre,” March 1941. ANOM 10i10.
• 28 . “Réunion en vue d’examiner certaines questions de nature à pallier la crise du logement,” 11 Sept (...)
• 29 . “Cité Mahieddine, recensement du 7 mars 1941”. ANOM 10i10.
• 30 . The 1954 population figures come from Çelik, 2003, p. 188; the 1960 numbers from “Monographie de (...)

12As the Casbah reached saturation, new arrivals congregated in shantytowns, where only the occasional European ever lived (most often as spouses of Algerian men; in one case a working-class Spanish family)27. The first modern shantytown of Algiers, Mahieddine, emerged just prior to the First World War on the slopes rising above Belcourt and Mustapha, the working-class outer districts of Algiers–near the port, warehouses, and slaughterhouse (House, 2014, p. 47-54). Other shantytowns followed, such as El-Kettar, Frais Vallon, Clos Salembier, Bel Abri28. Not all shantytowns were equal. Mahieddine offered good proximity to jobs and streetcars, and it became a preferred neighborhood. It housed the comparatively fortunate families who had lived the longest in the city (in 1941 40% of people in Mahieddine had lived in Algiers for more than a decade) and worked at more regular jobs29. At the beginning of the Algerian war in 1954, 7 400 people lived in Mahieddine, and six years later (1960) its population had risen to about 10-12 000 people30.

• 31 . Minutes of meeting led by Perillier, n.d. ANOM 10i10; Lespès, “Projet d’enquête,” op. cit.,
  431-3 (...)
• 32 . Girardiere, “La lutte contre les bidonvilles,” June 1939, ANOM CHEAM 1005.
• 33 . “Recensement des bidonvilles, compte-rendu,”19 March 1941. ANOM 10i9.
• 34 . ANOM 10i10.
• 35 . Office publique d’HBM de la ville d’Alger to Préfet du département d’Alger, 17 March 1941, no. 54 (...)
• 36 . Roland Simounet: d’une architecture juste (Paris: Moniteur, 1997).

17There is also the question of the state’s position. The state’s indifference to the harsh living conditions did not mean they had forsaken the shantytowns. The abandonment suffered by the shantytowns is best understood as a form of relinquishment, with the state letting them to their fate, or, rather, letting things happen that are understood to be governed by inevitable processes. This was the first response to the crisis of the shantytowns prior to 1954 when France still enjoyed an uncontested sovereignty in Algeria. To this end, municipal authorities steadily collected information about the urban poor. Although they were hidden from daily life, as reformers complained, people living in shantytowns were far from invisible to government agents. This work began during the 1930s and was motived by a series of fears: fear that the shantytowns represented a health risk, the fear that shantytowns might give rise to revolt, the fear of international condemnation on humanitarian grounds (during the Vichy era this included fears that Mussolini would exploit the deplorable poverty to embarrass the French and put forward Italian claims in North Africa)³¹. These yielded a call to suppress the shantytowns, the so-called “struggle against the shantytowns”32. But rather than undertaking a frontal assault—pulling them down, burning out their inhabitants—city planners responded with an effort to know their populations. Government agents completed a detailed census of all the shantytowns in Algiers in March 1941³³. Undertaken by a team of 18 (including 6 Arabophone and Berberophone translators), this census shone a clear light into the shantytowns, listing names, family composition, places of birth, time in the city, as well as people’s work, their pay, and the rent they paid. Rather than the “epistemological fog” (Davis, 2006, p. 45) that surrounds global slums today, Algiers’s shantytowns revealed themselves in neatly bound dossiers, full of laboriously typed figures, fully ordered and tabulated³⁴. With this information, government agents made decisions about rehousing, and what elements of the population, deemed “socially interesting,” they could move to new homes³⁵. But in as much as they recognized that shantytowns were useful and cheap sources of labor (House, 2018, p. 140), and facing the challenges of destroying them, authorities opted to manage the shantytown, conducting, arranging, and directing their inhabitants. Even after authorities had turned towards intervention and reform in 1953 (House, 2018, p. 147-55), architects studied the shantytowns for the “discovery of a primordial organic tissue” of vernacular building forms that they could reproduce in their designs³⁶. Zeynep Çelik has suggested that the idea was to “learn from the bidonvilles” (Çelik, 2003a). In this sense, these architects did not seek to reform the deracinated and normless “new man” who so worried reformers but to find “man himself,” the bearer of authentic and uncorrupted deep structures which could be known, rationalized, and then freed to reproduce themselves to the benefit of colonial society.

• 37 . Action menée dans Alger par l’autorité Militaire responsable du Maintien de l’Ordre, 10e div para (...)

18This approach to the shantytowns ended under the force of war. In February 1957, French soldiers invaded Mahieddine, attempting to break up support of the FLN-organized general strike during the Battle of Algiers (House, 2014, p. 49-50). It was only at this time that authorities turned towards investing themselves in the shantytowns, seeking for the first time to ensure that people became consolidated under state control. Thus, in February 1957 a new census focusing on Algiers’s Muslim population sought to “establish files on and the control over all the population”37 and this project seconded itself with army units housed in the shantytown. This effort at direct control, reminiscent of the modality of the regroupment camps, moved forward with mixed results capped with the staged burning of veils by Algerian women in 1958 and other projects to mobilize Algerian Muslims in Algiers to oppose independence (MacMaster, 2009).