From Sétif to Moramanga: Identifying Insurgents and Ascribing Guilt in the French Colonial Post-war

I: Anti-colonial Insurgency and the Violent ‘Post-war’

Reflecting on the upsurge in anti-colonial insurgency between 1945 and 1947, this article explores critical transitions in colonial state violence in two of the three French dependencies identified above: Algeria and Madagascar. The selection of these two cases juxtaposes one territory: Algeria, the epicentre of France’s empire and the clearest exemplar of ‘deep settler colonialism’ in the francophone imperial world, with another: Madagascar, something of a colonial outlier, little understood and often overlooked in French official rhetoric and public-sphere discussions of the empire’s possible futures. ³ So, too, concentration on the years 1945 to 1947 is intended to shed light on the most intense period of French domestic and imperial reconstruction, one that coincided with a pronounced ideological shift in the orientation of French government from the years of leftist tripartite government to the resurgent conservatism of the Third Force centre-right. ⁴ Finally, the designation of ‘post-war’ as a noun is also deliberate. Historians of the disorderly aftermath of the First World War, its legacies of forcible population transfers, paramilitarism, and continuing intra-state violence, discern a particular quality to the political uncertainties and social upheaval provoked by preceding global conflict. ⁵ So, too, historians of the occupations, displacements, and reconstructions, which followed the Second World War, detect similar particularities in the months and years after 1945, a transitional period that historian Peter Gatrell tellingly describes as ‘violent peacetime’. ⁶ These peculiarities reverberated through European empires with equal, if not greater force. Focusing here on the French Empire, the suggestion is that official and local responses to colonial disorder in these post-war years defined new, more violent parameters of colonial counter-insurgency. Its forms and presumptive rationale would remain substantially intact over the coming decades of decolonization. Indeed, the idea that anti-colonial subversion and repressive violence were now intrinsic to an intermediate space between war and peace would become integral to the articulation of French military concepts of revolutionary war refined during the conflicts of the 1940s and 1950s in Vietnam and Algeria. ⁷

Martin Conway and Robert Gerwarth get to the essence of the analytical problem here:

Partly because of the falsely unifying concept of ‘the Second World War’, with its concomitant emphasis on a narrative of military conflict, it is all too easy to overlook the other dynamics of violence that emerged in the maelstrom of ethnic, partisan and economic conflicts which overwhelmed Europe in this period, and which applied well beyond the conventional terminus dates of 1939 and 1945. ⁸

Whether it was revolutionary or counter-revolutionary, violence, they contend, was by the mid-1940s deeply embedded as a form of political action in Europe. My point is that such was even more the case within Europe’s overseas empires where the shift to dirty war practices was already well under way. Repressive action was typically intelligence-led, its organizers conflated civil and military authority, and its enactment was predicated on sweeping legal restrictions that served to conceal much of the violence enacted. ⁹ In the same way that Conway and Gerwarth identify a reciprocal relationship between social breakdown, political violence, and revolution in Europe at the close of the Second World War, so this reciprocity operated with even greater intensity in much of the colonial world of the late 1940s. ¹⁰

My argument is thus one about context, about a liminal space between war and peace that valorized a political rhetoric of resistance at home while sustaining the occlusions necessary to enable heightened violence in the colonies. It connects four factors, all of which proved pivotal to fundamental shifts in the nature and scope of French colonial repression after 1945. First is the ascendancy of a new French political elite at the Liberation, one whose ideas of political inclusion and legitimate political violence were shaped by particular views of republican integration and experiences of wartime resistance to Nazi occupation. The insurgents of Algeria, Madagascar, and Vietnam mimicked the associational forms of popular resistance and appropriated its claims to legitimacy. This presented unique dilemmas to prominent figures in government, colonial administration, and the security forces whose post-war primacy derived from personal or institutional connections with French resistance movements, whether at home or within the empire.

Linked to this is the second factor: the cognitive dissonance experienced by the French Empire’s senior administrative cadres as they sought to make sense of anti-colonial disorder. In Leon Festinger’s classic formulation of cognitive dissonance individuals seek recognizable patterns of behaviour in the actions of others. Dissonance results when those being observed do not act in the ways anticipated. ¹¹ The responses of French colonial analysts in Algeria and Madagascar may properly be called dissonant because, on the one hand, their expectations were confounded by the actions of insurgents. On the other hand, the tendency persisted to apply a cognitive framework that depicted rational grievance as violently irrational and thus non-negotiable. The insurgents’ claims were depoliticized, typecast as criminally illegitimate, and psychologized as infantile. At the same time, anti-colonial dissent was culturally constructed as viscerally menacing. The dissonance characteristic of administrative responses to rebellion echoed the destabilizing effects described earlier between the opening of possibilities for dialogue in some parts of the empire and the violent closing of political space in others. Locally, these opposing tendencies were not so much contradictory as temporally contingent: the counterpart to reformism and the prospect of greater political inclusion was exclusion, even repression, for those who rejected what was on offer. Failure of the protracted Franco–Vietnamese negotiations of 1945–46 is substantially explicable in terms of the ways in which opponents of compromise exploited the societal disruption in North Vietnam provoked by war, famine, and successive Japanese and Chinese occupations to spark confrontation between French and Democratic Republic of Vietnam (DRV) forces. ¹² But the outbreak of the Indochina War in December 1946 also represented something else: the Hanoi regime’s rejection of French schema designed to perpetuate imperial influence. If the embrace of reformism promised entry to the French republican cité, its rebuttal placed the individuals and organizations responsible outside it. In this way, talk of reform, although intrinsically welcome, also sharpened the distinction between partner and opponent.

The attendant reconceptualization of French imperial threat perceptions went furthest in those territories where political intelligence analysis and security policing became integral to day-to-day governance at provincial, prefectural or cercle (district) level. This, the third factor to be considered, explains the primary focus on Algeria and Madagascar. My suggestion is that the social unrest and ethno-cultural discrimination central to the outbreak of collective violence in these territories were all exacerbated by the manner in which these phenomena were described, processed, and misunderstood. Fourth, and finally, the article connects the preceding issues of resistance and political legitimacy, intelligence culture, and administrative reconstruction to the ways in which insurgent violence was countered. The interaction between these four elements defined the parameters of French counter-insurgency practice in the final decades of empire.

II: Violence, Resistance, and the New Rulers of Empire

Frantz Fanon, in his 1952 work Black Skin, White Masks, famously excoriated Octave Mannoni, a French psychiatrist, whose account of the Malagasy rebellion – Prospero and Caliban: The psychology of Colonization – reduced rebel motivation to a caricature of psychological dependency. According to Mannoni, the leaders of Madagascar’s Mouvement Démocratique de la Rénovation Malgache (MDRM) were sufficiently educated to recognize their reliance on French tutelage but were insufficiently mature to achieve genuine autonomy, psychologically or politically. The majority of their followers, in this reading, cleaved to the MDRM because they felt betrayed by a weak wartime administration, whose guiding hand they craved. Fanon rightly exposed the racial stereotyping at the core of Mannoni’s interpretation. He ridiculed the psychiatrist’s unwillingness to concede that Malagasy were reacting rationally against decades of economic exploitation and cultural denigration. ¹³ Mannoni’s egregious viewpoint offers a route into the themes explored below because his identification of the Malagasy people as authors of their own downfall in the crackdown that followed rebellion in 1947 contrasts starkly with the four arguments I wish to put. My concern is not with the collective psychology of infantilized colonial subjects but, rather, with attitudinal shifts among the new rulers of empire to emerge from the ashes of the Second World War.

Many of those at the apex of government – metropolitan and colonial – were new rulers indeed. Superficially at least, the French political figures so determined to revitalize the Empire after 1945 made unlikely imperialists. Most were ideologically left-of-centre. Several leading government figures had been imprisoned for resistance activities. Some, like the Corsican jurist Paul Giacobbi and Marseilles mayor Gaston Defferre, built powerful regional political networks on the foundations of their earlier resistance work. ¹⁴ Other discrete groups – Jean Monnet’s dirigiste planners and Pierre Mendès France’s political economists – were technocratic modernizers. Their reformist sympathies translated into commitment to mobilize state resources to develop colonial economies, improve welfare provision, and raise living standards. Even France’s influential community of colonial anthropologists, bitterly divided in their wartime responses to Vichy’s racial discrimination, shifted after the war. In 1950, echoing the newly founded United Nations Educational, Scientific, and Cultural Organization (UNESCO), they formally condemned scientific racism, dismissing received wisdom about hierarchies of civilization as a pernicious myth. ¹⁵

The problems facing these eclectic French reformers as the Second World War ended were daunting: not just the physical reconstruction of a home country ravaged by war but the cultural reconciliation and psychological healing of a society similarly torn apart. ¹⁶ Yet the post-war years also presented tremendous political opportunity at home and in the empire. Restoring republican democracy to France, and drafting a new constitution to enshrine it, implied nothing less than reorganizing the state, redefining the role and purpose of government, and reconsidering the reciprocal relationship – the social contract – between citizens and their leaders. Nor was this process confined to the hexagon of metropolitan France. ¹⁷ Rebuilding infrastructure, the reordering of politics inherent in a new constitution, and the reconceptualization of the rights and duties of governors and the governed were tasks equally applicable to France’s overseas empire. ¹⁸

The essential point here is that the ascendency of these politicians, their working coalitions, and the advancement of the colonial administrators serving them, all derived substantially from wartime resistance activities. Hang on a minute, one might say. For one thing, claiming resistance connections was quite distinct from translating resistance ideals or practices into post-war government. For another thing, Olivier Wieviorka is surely right to suggest that the significance of resistance credentials has been overblown, and that any revolutionary shift in post-war politics was illusory. ¹⁹ Among the early Fourth Republic’s parties of government only the UDSR emerged from a fusion of various non-Communist resistance groups. ²⁰ Admittedly, the MRP built on the foundations of Catholic resistance networks but its collective leadership also stressed their intellectual debt to the social Catholicism of the interwar years. ²¹ Notwithstanding their organizational rebirth, enough old faces remained within the local and national executives of the Socialist and Communist Parties to indicate continuity rather than rupture with the pre-war years. ²² Perhaps most important, neither in terms of purging France of all institutional and political remnants of the Vichy past, nor in terms of their continued preference for coalition-building did the administrations of the tripartism period signify transformative change. ²³ The épuration, although profound, was soon scaled down. Its colonial equivalent was superficial. ²⁴ Numerous Vichy sympathizers retained administrative or legal posts during French North Africa’s transitions of power from Generals Giraud to de Gaulle in 1943–44, and the later judicial purge was patchy. ²⁵

In two other respects, however, the resistance ticket mattered. Rhetorically, the language of politics was now inflected with concepts, even ‘traditions’, of popular resistance. And this, in turn, shaped political thinking – and constitutional guarantees – regarding citizens’ rights and, more specifically, the minimum protections that they should expect from the state. ²⁶ This returns us to the constitutional architecture of the Fourth Republic and its overseas dependencies if only because France’s rulers of empire were constrained by newly elected institutions, additional colonial administrative agencies, and the promise of more inclusive political cultures they implied. Put differently, the forms of government enshrined by the Fourth Republic and French Union began from the proposition that political engagement crystalizes in the juridical relationship between the individual and authority. As a result, the constitution-makers were acutely sensitive to colonial echoes of resistance, this time pursued by soon-to-be imperial citizens against a one-and-indivisible Republic. The resultant juxtapositions were stark. Anti-fascist republicans determined to restore democracy to France dismissed violent anti-colonial activism – ‘resistance’ of a sort – as illegitimate and criminal.

Conversely, French official justifications for colonial repression were framed in terms of the obligations of colonial subjects to imperial authority; obligations that, it was argued, were distinct from the rights and duties of French citizens to resist persecution. For many resisters, especially the non-communists, taking up arms against unjust rule was conceptualized in terms of republican defence, the origins of which were traceable to popular mobilization against counter-revolutionary forces in the early years of the French Revolution. Resistance signified the fulfilment of a citizen’s obligation to fight for republican ideals more than simply for family, community, or place. Despite their past associations with popular struggle against foreign occupation, the Fourth Republic’s republican elites endorsed harsh colonial clampdowns, not as a departure from their normative standards, but as their affirmation.

Security force repression was justified as socially disciplinary and politically preparatory. Restoring order was pathologized as a curative psychological process; the forcible correction of the misperceptions of colonial opponents led astray. ²⁷ Once this shock therapy took effect pathways would be cleared, mentally as much as administratively, for the elevation of dependent societies. ²⁸ Swift, decisive action was more than ever essential after France’s Liberation because, as we’ve seen, the mechanics of imperial government were being re-engineered. Forced labour was finally to be abolished. Citizenship, voting rights, and enhanced work and welfare entitlements were a realistic prospect, albeit hamstrung by pronounced gender and religious bias. ²⁹ Rocking the colonial boat at such a sensitive moment was read as further proof of political immaturity, naïve impatience rather than the release of pent-up wartime frustrations. Underlying these patriarchal presumptions were less readily acknowledged fears: that anti-colonial militants in Algeria, Vietnam, and elsewhere might risk the opportunity costs of violent social action because they discerned a France laid low by years of war as vulnerable to pressure for change. ³⁰

In practice, these militants only turned to violence when other avenues of dialogue were closed. Instead of release from internment and potential re-engagement, at war’s end Algeria’s PPA leadership remained in jail or under house arrest. DRV leaders in Hanoi, while certainly preparing for revolutionary overthrow, began definitive preparations for war only in late 1946 after France’s tripartite coalition split decisively over the continuation of high-level talks. Parliamentarians from colonial constituencies also made their influence felt during the Constituent Assembly debates in early 1946 over the projected French Union. Malagasy Deputies Joseph Ravoahangy and Joseph Raseta made the case for ‘statehood’ within the French Union only to find their arguments dismissed out of hand. ³¹

III: Resistance, Repression, and the Sétif Uprising

Discussion of the Malagasy and Algerian Deputies brings us to the regional examples of post-war repression that convulsed the empire as plans for the French Union took shape. The journey begins in the Sétif region of north-eastern Algeria in May 1945. Viewed through the prism of decolonization, the Sétif uprising and its attendant counter-violence exemplified in compacted form the struggles that erupted across Africa and Asia over the next generation: an asymmetric contest between a colonial occupier apprehensive of the loss of politico-economic control and an indigenous nationalism radicalized by legal proscription. ³² Placed within its local context, Sétif marked the intensification of another type of struggle. This was the battle between what I’ve termed elsewhere an ‘intelligence state’ against the upsurge in mass opposition that its information-gathering processes were supposed to predict, prevent, or contain. The analytic readings of the Sétif experience by a civil-military elite trained to apply particular filters of cognition and interpretation to violent dissent catalysed an institutional transition in French-ruled Algeria towards a repressive regime that might more properly be termed a police state. ³³ Where formerly the colonial authorities depended on the pre-emptive collection of sensitive information to sustain political control by intervention against presumptive opponents, after Sétif, targeted surveillance gave way to systemic repression. New-style methods included tighter restrictions on rights of assembly, additional police powers to arrest those accused of threatening state security, and the creation of dedicated detention centres to house those suspected (but not necessarily convicted) of sedition. ³⁴

As these juridical changes imply, the Algeria that emerged from the Second World War, although nominally a part of France’s republican renovation, was excluded from it in practice. This is not to dismiss official interest in lending greater substance to Algeria’s theoretical integration with France through economic modernization and wider access to citizenship. Nor is it to underestimate the symbolic importance attached by Algerian political actors to appropriating the ceremonial practices and discursive forms of French republicanism, its professed egalitarianism especially. ³⁵ Each gathered impetus as schemes for an Algerian Statute – essentially a constitutional redraft – emerged between 1945 and 1947. But whether we regard such measures as subtler forms of social control or as a real attempt to collapse Algeria’s communal divides, the outcome satisfied no one. ³⁶ The new statute, finally voted through in September 1947, did little to reverse the differential treatment of the subject population, whether on the basis of economic stratification, ethnicity, language, religion, or an amalgam of cultural differences. Prior to its adoption, acute inequalities in a post-war environment marked by economic deprivation, supply shortages, and political uncertainty created a climate of fearful anticipation among security force personnel and civilian intelligence analysts, notably the Arabic-reading specialists tied to prefectural administrations. ³⁷ Their anxiety impelled closer surveillance of those colonial subjects and groups judged most prejudicial to order. ³⁸

But how did the colonial authorities monitoring local opinion discern such shifts? More crucially, how did they read them? As Ann Stoler has argued by reviewing the colonial archive of the Dutch East Indies, the evidential narrative, judgmental but highly subjective, speaks to administrators’ anxieties about surrounding populations. Decisions about whom to monitor, what to report, and which to proscribe were reached on the basis of such material. The fear that drove these analyses was stoked by a volatile mixture of presumptive expectations about the behaviour of local communities and incomplete local knowledge about them. As Stoler puts it, ‘Partial understandings, epistemic confusion, and undigested bits of cultural “information” made up the modus vivendi of high and low civil servants across the Indies.’ ³⁹ The replication of such colonial knowledge construction in colonial Algeria – and the acute dangers it posed to the colony’s Muslim population – is abundantly clear in the surveillance reports compiled by the Algerian gendarmerie before and after the 1945 uprising. The policing agency closest to the colony’s rural poor, the gendarmerie could claim to be colonial administration’s primary gatherer of evidential data about Algeria’s peasant majority. The intelligence picture available to gendarmerie commanders seems, in hindsight, to have been richer than that of the colonial government’s native affairs section, the administrative branch supposedly best apprised of social conditions. In the four months before the May uprising, gendarmerie political intelligence surveys mapped connections between a sharp rise in minor criminal misdemeanours in north-eastern Algeria, including assaults, thefts, and vandalism, and the deepening inter-communal antagonisms that erupted into violence on 8 May.

These accounts may, to a degree, be explicable in light of the impending abolition of the hated native penal code, the indigénat. It had for decades enabled police officers and local colonial officials to issue punishments, usually fines, internment, or brief spells in jail, without recourse to the courts. ⁴⁰ As arbitrary as they were racially configured, the indigénat codes, which varied colony by colony, invited abuse. Their principal targets were young males, always judged the most threatening of colonial subjects. Abolition, finally enacted in 1946, was long overdue. ⁴¹ And it opened the door to a fundamental reordering of colonial law with new codifications of juridical powers and individual rights within the federal structures of the French Union. ⁴² Ending the indigénat had another, more mundane consequence: petty crime now required bureaucratization. Increased reports of assaults and malicious damage simply represented the recording on paper of infractions previously handled on the spot. Here, the epistemic anxieties identified by Stoler as typical of colonial officialdom were intensified by the greater recording of otherwise insignificant, low-level crime. Security assessments began to claim that popular deference to colonial authority was collapsing. ⁴³ Algeria’s senior gendarmerie commanders, General Taillardat and Lieutenant-Colonel Roubaud, concluded that the colony was gripped by a social ‘malaise’ marked by intense popular resentment at any symbol of French privilege. ⁴⁴

Gendarmerie accounts were borne out by French magistrates in their periodic summaries of Algerian public order offences sent to the Ministry of Justice from February to May 1945. These noted an explosion in the number of police and gendarmerie arrests for the distribution of banned nationalist tracts and posters, graffiti writing, and disruption of otherwise peaceable meetings. Several caïds and other local officials complained of being assaulted or intimidated, often by crowds in which young members of a Muslim scouting association played the leading role. ⁴⁵ Reports of this type from local authority figures confirm the observations of Stathis Kalyvas and Matthew Kocher regarding the relative dangers of participation versus non-participation in social movement protest and insurgent violence. ⁴⁶ Young adult males and day labourers on settler farms confronted difficult choices. Those reluctant to take part in anti-French actions faced condemnation by their communities and workmates. But participating in violence risked security force punishment. Crucially, it appears that in early May 1945 the dangers of the former outweighed the latter. Collective action, in other words, was not just a matter of political preference; it was the more rational alternative to passivity and ostracism.

The Constantine Information Studies Centre (Centre d’information et d’études, CIE), the regional clearing house for incoming political intelligence in Algeria’s eastern-most département, dwelt on the sociopolitical implications of this persistent harassment. Frightened for their safety, village headmen and other Algerian public servants were distancing themselves from a loathed colonial administration. Leading notable families throughout north-eastern Algeria were also conspicuously silent, reluctant to make themselves targets, either of PPA militants or the colonial security forces. ⁴⁷ CIE analysis, which spoke of collapsing social hierarchies, meshed with the shocking accounts of over a hundred settler deaths between 8 and 11 May 1945 to provide official justification for massive increases in security force deployments in Sétif’s aftermath. On 5 and 6 June Generals Henry Martin and Raymond Duval, the army commanders charged with directing the regional clampdown, toured several outlying settlements where French families and forestry workers had been killed. They found settlers terrorized by thoughts of Algerian neighbours, commercial employees, and farmhands joining forces as violence workers determined to kill them. ⁴⁸ Finding themselves the targets of Algerian anti-colonial violence, some Europeans adopted the language and associational forms of resistance in response. In the Guelma sub-district, another epicentre of the May outbreak, most adult male settlers signed up to paramilitary ‘Vigilance committees’ established on the initiative of their local mayor by reactivating the local Free French resistance network. ⁴⁹

IV: Ascribing Guilt and Depoliticizing Insurgency

It was always unlikely that French Algeria’s bureaucrats might concede that acts of anti-colonial violence signified any kind of resistance against unjust rule. Instead, they condemned the original attacks on settlers as beyond rational justification. Anticipating further race killing, regional officials, prefecture security analysts, and medical practitioners explained political dissent among the Muslim majority in pathological terms of individual and collective psychological neuroses. The Constantine CIE had surmised in April 1945 that Constantine’s rural population was suffering acute hardship owing to the combined effects of draught, harvest failure, and wartime disruption to foodstuff markets and transport systems. The region’s poor were ‘persuaded that Algeria’s European population were indifferent to their fate’. Blaming their woes on governmental ‘incompetence’, local Muslims ‘believed they sensed a need for change in a sort of collective psychosis that made them receptive to separatist theories and the most fantastic rumours’. ⁵⁰ CIE writers saw no need to explain why Algerians’ logical identification of food shortages and price inflation with ineffective administration signified psychological disorder. What we might read as indicators of collective politicization and social radicalization were, instead, depicted as affirmation that the Muslim population’s latent millenarian fervour had festered in the volatile socio-economic conditions preceding the uprising. The crisis, in other words, originated in the collision between the administrative hiatus of the war’s final months and Algerian proclivity to emotional over-reaction. PPA activists had exploited both to the full – an act, not just of sedition, but of criminal irresponsibility. ⁵¹

Not surprisingly, fine-grained analysis of the rebellion’s violence does not support security officials’ readings of criminal incitement of a fanaticism hitherto latent among Eastern Algeria’s rural poor. Nor do the initial attacks on settlers, gendarmerie and police stations, travellers, and farmsteads appear to have been driven by material gain in line with the greed thesis proposed by Paul Collier and Anke Hoeffler as a core motivation for violence actors in civil war. ⁵² Rather, the end-of-war combination of supply disruptions, harvest failure, foodstuff shortages, and spiralling prices intensified grievances felt by sharecroppers and day labourers against farm owners and wealthier European townsfolk. Thefts of grain, livestock, and other essentials certainly took place from farmsteads and homes after their settler occupants were murdered. But if these actions suggest some conformity with Collier and Hoeffler’s greed model, they do not explain the ability of PPA organizers to coordinate attacks. Contrary evidence of the politicization of popular grievances against a settler over-class insulated by preferential rationing and greater wealth from the worst of Constantine’s food crisis points instead to a combination of local antagonisms and external coordination. ⁵³ More typical of the civil war violence analysed by Stathis Kalyvas, it was, at the time, jarring to the French security analysts searching for affirmation of fanatical killings. ⁵⁴

Their normative presumptions challenged by the actions of the living, colonial security officials turned instead to the bodies of the dead. Police analyses of autopsy reports on those killed in the initial outbreak were cited as further evidence of a pattern of collective psychosis in which cultural prohibitions against interpersonal violence collapsed in an atmosphere of mob vengeance. Most of the 28 fatalities in Sétif died from multiple stab wounds or head trauma consistent with crushing blows from blunt objects. Twenty-six European victims were male; two were female. The manner of their deaths indicated numerous assailants purportedly acting in a ‘frenzy’ of violence. Subsequent mutilation of 23 of the dead, mainly by intestinal evisceration, suggested that victims were doubly violated in acts of ritual slaughter. Their bodies treated like animal carcasses, victims were dehumanized in death; evidence, it was said, of their attackers’ primal aggression when unencumbered by the social constraints of societal norms. ⁵⁵ (The total absence from the archive of comparably detailed reports for the far larger numbers of Algerian victims of settler and security force retribution says much about the relative importance attached by the colonial authorities to each.)

Police interview reports, collated by the CIE, followed a similar line, conflating racial and Islamophobic stereotypes with a pseudo-medicalization of insurgent actions. Male detainees ‘under the age of thirty’, it was alleged, withstood the pressures of interrogation because their religious ‘fanaticism’ inured them to pain – and to reason. According to the CIE analysts, such intense devotion indicated how successful Algeria’s leading Muslim cultural network, the Association of reformist ulama, had been in exploiting madrasas, boys’ sports clubs and scouting groups to ‘indoctrinate’ (façonner leurs cerveaux) these young men against France. Religious schooling cultivated the ‘virus’ of separatism, which youth organizations then disseminated. ⁵⁶ Scouting associations were singled out. Despite, or perhaps because of the internment of leading ulama clerics and the closure of their madrasas in Guelma, Akbou, Tazmalt, Djidjelli, Mila, and Tébessa, their young Muslim Scouting Association devotees were described as ‘uncommunicative and unwilling to be disarmed’. ⁵⁷

Another assumption intrinsic to such analyses was that the endemic criminality of young Muslim males made demonstrative security force retribution inevitable. ⁵⁸ Alienated from the social order of colonialism and unable to transcend their primitive instincts, these colonial subjects were inherently dangerous, it was averred, because they had no stake in French governance and were unreceptive to rational dialogue. What looked like a mixed message – young Algerian men were intrinsically prone to violence but their violence was compounded by alienation from authority – was boiled down to a simpler policy prescription. Atavistic, regressive violence was always liable to break out among an unstable local community susceptible to the influence of malevolent extremists. ⁵⁹ Enacting reform was pointless when disorder was embedded in the minds of Algeria’s majority population. ⁶⁰

The gendarmerie command, for instance, applied this logic to explain the recrudescence in interpersonal violence between supporters of rival Algerian nationalist groups during late summer 1945. PPA activists who escaped the post-Sétif crackdown sought to enforce boycotts against Algerian participation in municipal elections and the French general election scheduled for late October. Eager to cement PPA primacy in its working coalition with the Association of reformist ulama and Ferhat Abbas’s Amis du Manifeste et de la Liberté (AML), they insisted that ‘collaboration’ with colonial authority was impermissible. ⁶¹ As the boycott gathered momentum in August, several pro-administration candidates received death threats, some of which were carried out. Gendarmerie General Taillardat commented that this campaign of violence proved that generous colonial reform – in this case the enfranchisement of thousands of Algerian men by the French Provisional Government in 1944 – was wasted on a population too immature to engage in political dialogue. ⁶² To be sure, the secret police analysts of the Algiers Renseignements généraux (RG) were more sensitive to the transnational currents of Arab opinion that shaped Algerian nationalist thinking. Their confreres in Morocco and Tunisia were at the time building wider global support through appeals to the newly founded United Nations and the Cairo-based Arab League. The latter organization, in particular, repeatedly stressed that internal nationalist unity was a prerequisite to its endorsement of Algerian independence. Even so, the RG concluded that the rebellion’s original organizers had lost control of their ‘shock troops’, the ‘young extremist fanatics’ who rejected anything short of independence won through violence. ⁶³

These readings of local activism reached such distorted conclusions because sophisticated anti-colonial activism could not be credited by those accustomed to viewing colonial tutelage as the natural condition of social relations between Algerians and French. ⁶⁴ Equally unconscionable was to interpret such dissent either as resistance or as clandestine multi-organizational networking; in other words, as a colonial variant of what so many of post-Liberation France’s leading political actors had recently practised. To understand this cognitive dissonance we need to return to the rebellion’s outbreak. In the first week of May Algeria’s prefectural administrations accumulated intelligence concerning covert PPA meetings and mass demonstrations. These were reported from Mostaganem in the west to Souk Ahras near the Tunisian frontier in the east. Still, the inference that a nationalist uprising was imminent remained hard for officials to process precisely because the mounting evidence of meticulous preparation by committed nationalists jarred with administrators’ abiding presumptions about Algerian political immaturity. ⁶⁵ This dissonance endured beyond the uprising. The prosecuting magistrate presiding over the trials of alleged insurrectionists combined a patriarchal disdain for the political aspirations of the Algerians in the dock before him with undiminished certainty in the merits of colonial guidance. Even the rebels, he said, recognized ‘the benevolent character of our colonial policy’. They were ‘not animated by hatred of the French’ but by dreams of Arab nationhood that stirred them to frenzy. ⁶⁶

Once again, it was gendarmerie commanders, still unable to admit the implications of their own findings, who possessed the most detailed information before and after the uprising. The discovery of widespread arms caches and information that Algerian army veterans had been organized into PPA ‘resistance bands’ (the inverted commas were always preserved when the term ‘resistance’ was invoked) indicated that wartime subversion techniques were being copied, sometimes by former colonial soldiers who had encountered ‘real’ resisters first-hand during their military service in France. Gendarmerie interrogation reports also confirmed the existence of a cellular PPA structure devised to limit security force infiltration; again, a practice familiar to French and other European resistance networks. ⁶⁷

Why depoliticize these actions? A strategy of delegitimization perhaps, but also a logical mental leap for security analysts and prosecutors whose cognition of Algerians’ collective behaviour was circumscribed by colonial stereotypes and the expectations they fuelled. To designate such violence ‘political’ not only rendered Algerian grievances rational adult responses to discrimination but risked according the violent social action of the Sétif rebels equivalence with French resistance to foreign occupation. Especially revealing in this respect was the identification of PPA-affiliated Berber rebel groups in the Tizi Ouzou region of Kabylia as criminal gangs of ‘outlaws’. Accused of ‘banditry’ by the police and gendarmerie units that tracked them between 1944 and 1948, these fighters, many of them ex-servicemen and army deserters, called themselves maquisards in evocation of their French resistance counterparts and the rural scrub-land through which they moved. ⁶⁸ Both sides clearly appreciated the rhetorical – and material – stakes involved in these appellations.

French security forces meanwhile moved quickly from selective coercion to mass killing of civilians. ⁶⁹ Redolent of patterns of collective punishment in occupied Europe, such retribution would become more characteristic of colonial counter-insurgency in the years ahead. The town of Guelma was the scene of the most extreme counter-violence in what became a perverted settler adaptation of resistance styling. Guelma’s sub-prefect, André Achiary, a former Algérois police commander, organized resistance-style ‘self-defence groups’ into a milice civique of almost 300 paramilitaries. Settlers’ choice of the term milice would have been intolerable in post-Liberation France where Vichy’s milice paramilitaries personified the worst of collaborationism: working alongside German occupiers in hunting down resisters. But the members of Achiary’s new force were apparently untroubled by the Vichy echo, and embraced the milice’s basic organizational principles of coercive policing and collective punishment. ⁷⁰ Guelma’s milice units worked alongside the police and gendarmerie in rounding up the town’s Algerian menfolk. At least 320 Guelmois were killed, of whom 256 were français musulmans or Algerian colonial subjects. ⁷¹ Historian Jean-Pierre Peyroulou calculates that over a quarter of the town’s males aged between 25 and 45 were executed. Most were found guilty of involvement in the uprising by a milice tribunal. Hastily convened but brutally efficient, it issued death sentences before transporting the condemned to mass burial grounds outside the town. This pseudo-legal process, a dreadful parody of the resistance courts in post-Liberation France, had purged Guelma’s poorest Muslim quartiers of working-age males by the end of May. ⁷²

Elsewhere, some 3,000 Algerians were awaiting trial for alleged involvement in the May uprising by the end of August 1945. The Military Courts put in place immediately after the rebellion at Sétif, Guelma, and Constantine were supposed to impose rapid, salutary judgements without the need for jury trial. But a shortage of qualified prosecutors undid the plan. It fell to three Army-appointed judges to hear the entire caseload, slowing down the process. Faced with a mounting backlog, the head of the Algiers appeal court advised the Ministry of Justice to reinstate criminal trials in which a civilian prosecuting magistrate would make the case before mixed juries of French and Muslim jurors. Doing so would undermine nationalist accusations of unproven, ethnically biased judgements. The fact that the defendants were accused of criminal wrongdoing, rather than political activity, would be made clear. ⁷³ Yet this presumption that detainees were criminal – and not political – prisoners was harder to sustain in light of the original allegations made against them. The Constantine Military Tribunal caseload, for instance, included 13 prisoners accused of ‘rebellion’, 66 held for ‘provoking an anti-French demonstration’, 102 for ‘organizing a protest march’, and 38 for ‘reconstituting a banned organization’. ⁷⁴ Admittedly, these were all actions that were to varying extents proscribed under colonial criminal codes, which limited rights of assembly and expression. Their political motivation was abundantly clear even so.

The underlying judicial problem was therefore to find effective measures under Algeria’s penal code to dispense criminal punishments, thereby disguising the fact that the actions involved were politically motivated. Ultimately, the Algiers government would turn to the so-called Regnier Decree of 20 March 1935. Technically a legislative instrument introduced to outlaw ‘demonstrations against French sovereignty in Algeria’, it offered the broad powers needed to issue sentences of up to two years in jail through the colony’s penal code. ⁷⁵ As the Guelma example, albeit a grotesque disfigurement of resistance practice indicates, numerous Algerian detainees faced extra-legal punishment despite the official impulse to criminalize political dissent the better to discredit it. But there are striking parallels here with another post-war effusion of colonial bloodletting: the deaths of several thousand Malagasy at the hands of colonial security forces in 1947–48. ⁷⁶

V: Interpreting Violence in the Madagascar Uprising

The French psychoanalyst Octave Mannoni, the long-serving official in Madagascar’s colonial administration whom we encountered earlier, depicted the notoriously brutal repression of the Malagasy revolt as a form of ‘theatrical violence’. Village burnings and novel forms of murder such as dropping victims from aircraft were demonstrative acts staged to restore order to the minds of an indigenous population whom Mannoni considered psychologically dependent on the unflinching discipline of external authority. ⁷⁷ Fanon was surely right to expose the racism intrinsic to Mannoni’s thinking and, by extension, the deeper forces that drove the Malagasy to violence. ⁷⁸ Mannoni, though, was in step with the colonial officials who ordered the clampdown and the security forces who enacted it in the revolt’s aftermath.

The methods adopted in Madagascar reverberated beyond the island as well. The revolt had rural origins in the island’s central and eastern agricultural belt. ⁷⁹ Disturbances, while localized, spread quickly from the eastern garrison town of Moramanga to Madagascar’s capital, Tananarive. The colonial authorities fixated on this issue of transmission, convinced that MDRP activists had orchestrated the rebellion. What Fred Cooper dubs ‘the second Madagascar thesis’ – namely, the viewpoint amongst colonial governors and officials in Paul Coste-Floret’s Ministry of Overseas France that French West Africa’s rural interior might succumb to the same seditionist wave – helps explain the decisions made to crackdown against Félix Houphouët-Boigny’s Rassemblement Démocratique Africain (RDA) during 1948 and 1949. The underlying fear was less that West Africa’s rural populations might take up arms against the French. It was, rather, that the RDA, like its MDRM cousin, sought to establish a state within a state by means of local party networks. ⁸⁰ (Again, the resonance with resistance-style organizational methods went unacknowledged in official correspondence.)

In eastern Madagascar, as in eastern Algeria before May 1945, political violence took hold among rural populations hard hit by a wartime supply crisis. The agricultural economy was severely disrupted by the requisitioning of export crops, chronic price inflation, and recurrent foodstuff shortages. Many blamed rampant black market profiteering on the colonial government’s rice office, which failed to maintain affordable prices for this essential staple. ⁸¹ Allegations of administrative incompetence became integral to MDRM campaigning. In the two years that separated the end of war in Europe from the start of Madagascar’s revolt, party organizers, hitherto concentrated in Tananarive, radiated outward to other cities and market towns. Highlighting the absence of effective governance also helped the MDRM transcend its ideological origins. The governing executive and their local party cells were modelled on the Madagascan Communist Party. But this was a relatively common organizational practice among early nationalist movements in the French Empire, many of whose leaders, like Joseph Ravoahangy and Joseph Raseta, served their political apprenticeships as young Communist activists between the wars. ⁸²

If anything, the MDRM became more transparent, and less sectarian, during 1945–46. Supporters of the Parti nationaliste malgache (PA.NA.MA.), a powerful force in the south of the Island, as well as the ‘JINA’, first among the secret societies that remained influential in Madagascar’s political life, were drawn into cooperation with local MDRM branches. This was a breakthrough for a movement whose elite-status urban leadership was socially unrepresentative of Madagascar’s peasant society. ⁸³ Reflecting its surge in local popularity, on 21 March 1946 the MDRM executive demanded French recognition of Madagascar as a free state within the French Union with its own government, legislature, and armed forces. ⁸⁴

The response from Paris was to change governors, not government. Marcel de Coppet, a former chief administrator in French West Africa, took over as Madagascar’s High Commissioner in May 1946. MDRM hecklers disrupted his maiden speech on 19 May, demanding ‘an end to annexation’. Hours later the protesters laid siege to the Governor’s Palace, the prelude to a night of rioting in the capital during which scores were arrested. Demonstrators targeted police stations in outlying provincial towns as campaigning for local elections got under way later that summer. Officers responded by rounding up known MDRM supporters, a portent of the police methods that would account for a significant number of the fatalities in the following year. ⁸⁵ The crackdown backfired. Those Party executive members still at liberty redoubled their efforts to win over urban voters, a tactic that paid off when the MDRM triumphed in provincial elections that July. ⁸⁶

Further elections to local assemblies were scheduled for early 1947. Police monitoring of MDRM meetings, interception of correspondence, and detention of party activists intensified in response. The MDRM became more radical as this repression ramped up. ⁸⁷ Party documents and editorials in the pro-MDRM newspaper Fahaleovantena (Independence) indicated a shift towards a more integral nationalist platform. ⁸⁸ At this stage, though, neither the MDRM leaders’ written communications nor their pronouncements suggested an imminent turn to violence. Admittedly, activists sang the banned MDRM anthem during electoral marches, mimicking the actions of the PPA in its pre-Sétif demonstrations. In another Algerian parallel, MDRM organizers in Moramanga approached Malagasy ex-servicemen to help train paramilitaries. ⁸⁹ If this smacked of insurgency, most MDRM activities did not. Party members throughout the island were, for example, instructed to send passport-style photographs to their local branch. Supporters’ pictures were to be distributed in France to prove how ‘civilized’ (for which might be read non-black) they were. ⁹⁰

MDRM leaders in Tananarive were also slow to harness outrage among the poorest sections of Madagascar’s population over continuing rises in the price of staple foods, rice and bread in particular. ⁹¹ Only after the revolt began on 29 March did the MDRM make sustained efforts to build mass support. This may be explained in the language of civil war analysis as a joint action model of micro-violence. The MDRM’s politicization of local grievances over rising prices, unfair rationing, and uneven availability of essential foodstuffs connected supra-local actors – in this case, MDRM activists – with local populations economically marginalized within the colonial agricultural market. ⁹² Notable in this respect was widespread incendiarism, much of it targeted at settler-dominated export agriculture. Coffee estates bore the brunt of these attacks. Plantations were ransacked and burned. Their stores of coffee were looted and later resold, both to secure essential cash and to drive down prices. Estate workers were urged to join the rebels. ⁹³

Jacquerie-style violence of this type sent shockwaves through the colonial administration because its police intelligence suppliers had ignored Madagascar’s peasant workforce, focusing instead on signs of urban politicization in the months before the rebellion. ⁹⁴ Just as MDRM activism pointed to a raucous, but elite-focused electoral strategy, so, too, civil society activism, notably among public sector workers and city consumers, offered little indication of an uprising. In January 1947 the administration’s Malagasy clerical workers formed a Comité d’Union des Fonctionnaires Malgaches in Tananarive to lobby for the same wages and working conditions as their French colleagues. A ‘Consumers’ League’ (Ligue des consommateurs) emerged weeks later to protest against rising food prices. ⁹⁵ Put differently, Malagasy urban politics in early 1947 seemed more akin to the vibrant – and non-violent – ‘claims-making culture’ of French West Africa’s coastal cities than pre-rebellion Algeria. ⁹⁶

De Coppet’s High Commission was thus caught off guard when, just before 1 a.m. on the night of 30 March 1947, several hundred attackers descended on the colonial army garrison at Moramanga on the island’s eastern highland rim. Twenty-two soldiers were killed during an assault whose main aim was to plunder the garrison arsenal. That same night, a smaller, less lethal raid on the Lazaret military base outside the northern port of Diego Suarez netted an assortment of small arms and ammunition. Almost simultaneous attacks in nearby settlements and the east-coast towns of Manakara, Farafangana, Vohipeno, and Vatomandry signalled the outbreak of a concerted rebellion. ⁹⁷ At its maximum in June–July 1947 the rebel-held area stretched to 100,000 square km, or one-sixth of Madagascar’s landmass, mostly in the south, east, and central uplands. A separate Mandritsara uprising was confined to the far north. ⁹⁸

The colonial administration was humiliated, its extensive political intelligence network undone by misguided threat assessments and a fatal underestimation of rural grievances. While loyal colonial troops were dispatched to secure key arterial rail links, de Coppet requisitioned Air France civil aircraft to enable him to visit Moramanga and other sites of rebel violence. ⁹⁹ Still uncertain about the scale of the unrest, the High Commission showed no hesitation in ascribing blame to MDRM ‘plotters’. Minds made up, de Coppet’s security personnel extracted the necessary evidence to prove the High Commission version of events. Another spate of arrests, police torture, and forced confessions confirmed the official view that MDRM activists worked with former soldiers and serving Malagasy troops to coordinate the initial attacks. Aside from the intrinsic limitations of the colonial archive, the administrative inclination to a priori judgement and manipulation of evidence – particularly from confessions extracted under duress – makes such material especially problematic. The point bears emphasis because of the onus placed on two confession transcripts to vindicate the clampdown. In the capital a serving Malagasy officer, Lieutenant Albert Randriamaromanana, reportedly met with MDRM leaders on 25 March to plan a citywide insurrection. Europeans were to be disarmed, the national radio station seized, and independence proclaimed. ¹⁰⁰ Randriamaromanana’s confession implicated the MDRM executive and was used to justify the first mass arrests. But it was soon overshadowed by a much fuller account.

The critical evidential breakthrough came on 12 April. After a night of police ‘interrogation’, Martin Rakotovao, editor of the MDRM newspaper La Liberté and secretary-general of the Party’s Tananarive office, identified 19 co-conspirators. They had met at an address on the rue Gallieni on 27 March to agree fine details of an insurrection in the capital. Administrative buildings were to be overrun. Stolen weapons were to be distributed, and, it was alleged, war unleashed by a self-styled MDRM ‘general staff’. Among those present were two of the MDRM’s French parliamentarians: Joseph Ravoahangy and Jacques Rabemananjara, plus the Party Treasurer, and several provincial councillors. ¹⁰¹ Similar plans were agreed for takeovers in other cities. ¹⁰² Extracted in circumstances that invalidated its citation as legal evidence, the evidential and ethical basis of Rakotovao’s account was not questioned in Paris. Instead, the French government deemed Rakotovao’s admissions serious enough to revoke the named MDRM deputies’ immunity from prosecution. ¹⁰³

Joseph Raseta, the remaining MDRM Deputy, who was in Paris at the time of the uprising, worked hard to highlight the abuses of judicial process and detainees’ rights before he too was stripped of his parliamentary privileges and arrested. ¹⁰⁴ It would be several months before the French Communist Party, Félix Houphouët-Boigny’s RDA, and France’s pre-eminent human rights organization, the League for the Rights of Man, took up the MDRM Deputies case. ¹⁰⁵ Security force repression meanwhile entered a more violent phase.

Despite its geographical diffusion, Madagascar’s civil and military authorities grew increasingly confident that the insurgency would be contained once reinforcements of colonial and Foreign Legion troops arrived in July. Areas of intense rebel activity were closed off (literally ‘caged’ or encager) and designated free-fire zones as mobile units swept through them. ¹⁰⁶ This was when substantial deaths occurred, often in circumstances that, to this day, remain unrecorded. Historian Jean Fremigacci stresses that the rebels’ lack of modern weaponry prohibited major firefights and ensured that army units were able to impose their supremacy swiftly. Both may have limited heavier casualties than would otherwise have been the case. But while the number of fatalities attributable to fighting with army units has probably been overestimated, police beatings and the unsanitary conditions inside overcrowded detention camps accounted for a larger proportion of Malagasy deaths than is widely assumed. ¹⁰⁷

Reflecting in 1949 on the speed with which organized resistance had collapsed in the latter half of 1947, de Coppet’s successor as High Commissioner, Pierre de Chevigné, argued that the rebellion was never deeply rooted. The militancy of Madagascar’s notorious secret societies was marginal, the integral nationalism of the MDRM intelligentsia even more so. According to de Chevigné’s advisers, ‘ordinary’ Malagasy preferred attentisme to political engagement. ¹⁰⁸ This interpretation was unconvincing. Wait-and-see attitudes surely reflected people’s awareness of the acute dangers of open attachment. The High Commission’s analysis was doubly reactionary insofar as its outcome realigned colonial administration with trusted, conservative clan elders and chiefs, particularly among non-Merina communities, whose living standards had steadily eroded since the depression years of the 1930s. ¹⁰⁹ For de Chevigné, as for the French governments he served in 1948–49, the answer to the Madagascar problem was to belittle popular grievances and turn back the clock.

Conclusion

Estimates vary wildly, but it seems credible to deduce that in the 27 months between May 1945 and August 1947 French security forces, their paramilitary auxiliaries, and settler vigilantes killed between 40,000 and 70,000 subjects of the French Empire. However configured, these numbers raise discomfiting questions when placed alongside other, better-known episodes of mass loss during the Second World War era. Direct comparison with French fatalities in the battle of France in May–June 1940 (approximately 85,000) or French Jews murdered in the Death Camps (just under 77,500) would be crass. In one sense, though, such comparison has a point, underlining that what sets security force violence in the French Empire between 1945 and 1947 apart is the statistical fog enveloping it. These colonial fatalities remain little known to most and, at the same time, hotly contested by others. Algeria’s leading pro-regime newspaper El Watan, for instance, insisted at various points in the early 2000s that 60,000 died in the post-Sétif repression in Algeria between May and August 1945 – a stark doubling of the previous FLN claim of 30,000 killings. By contrast, the official death toll of 1,500 conceded by the French colonial authorities has never been formally revised or challenged. Imprecision is everywhere. We lack exact figures for the numbers killed in the May 1945 bombardment of the Damascus parliament because French colonial troops were ordered to remove bodies to mass graves outside the city. Similarly, the numbers killed by artillery shelling, naval bombardment, and aerial strafing at Philippeville and Bougie in May 1945 and at Haiphong in November 1946 will never be known, particularly as the population of these port cities was first inflated by an influx of refugees fleeing violence elsewhere, then deflated by a larger refugee exodus after the bombardments began. The largest number of fatalities was in Madagascar: even rejecting as untenable the original French Communist claims of up to 100,000 killings, upwards of 20,000 died, whether in the three months of army sweeps through the island’s centre and south-east or during periods of confinement in detention camps and police custody. Much of this violence was openly retributive. Often it was performative with calculations of audience and popular response central to the choices and methods of violence applied. Always it was highly imbalanced. Targeting was, at best, arbitrary, at worse completely unscrupulous. The asymmetry of archival evidence about colonial deaths next to European ones reflects the asymmetry of the violence thereby concealed.

Extreme repression marked by collective punishments, free-fire zones, and widespread rights abuses are all familiar to scholars of colonial counter-insurgency and of asymmetric warfare more generally. ¹¹⁰ The tendency to indiscriminate violence is also heightened in conflict situations in which insurgents are typecast as rebels pursuing criminal ends, not recognized combatants fighting for a putative state. ¹¹¹ Limited protections of international law are nullified; repression proceeds unencumbered. Should we, then, regard French official justifications for their counter-insurgency methods as merely a rhetorical sleight of hand to conceal what remained quintessentially colonial violence? Three things suggest not. One is the symbolic and material importance of resistance precedents, evident in the forms of paramilitarism adopted by vigilante groups in eastern Algeria over the summer of 1945. Another is the pressing importance of the citizen–state relationship within French imperial affairs in the immediate post-war, a transitional period in which the social contract between governors and governed at home and in the colonies was fundamentally redefined. Finally there is the prevalence of government and security force narratives, which defended counter-insurgency as a necessary prelude to modernization. ¹¹²

Insurgencies and counter-insurgencies are shaped by the estimations of one side for the other. These readings, in turn, combine cultural presumptions and political predilections with more analytical calculation, itself the product of specialist administrative and intelligence services. The examples from the French Empire considered here indicate the extent to which such information analysis drew on the particular perceptions and worldviews of a governing elite whose prior experiences of resistance against foreign occupation affected their cognition of anti-colonial insurgency. Ironically, the interpretations attached to intelligence of impending inter-communal unrest in eastern Algeria made it harder to read the uprising, when it came, as a violent social movement against colonial inequalities – a variant of resistance, not its antithesis. The conviction that anti-colonial violence was unrepresentative and conspiratorial hardened in Madagascar, producing even more excessive retribution.

Here we get to the heart of things, to an epistemic violence in which particular forms of language and consequent interpretive frameworks devalued colonial grievances. This repressive turn was doubly tragic insofar as its authors were politicians, officials, and security force personnel, many of whom suffered similar processes of exclusion and silencing during their wartime years. ¹¹³ So it was that the architects of the Fourth Republic took the decisive steps between 1945 and 1947, setting France on a road to decades of extreme colonial violence more usually associated with colonialist diehards than erstwhile resisters.