SciELO - Scientific Electronic Library Online

 número46FE EN EL PAPEL: LA INSCRIPCIÓN DEL DOMINIO DE LAS TIERRAS DE COMUNIDAD EN EL ALTIPLANO CHILENOArchivo Fotográfico. Manuel Yanulaque Scorda (1850-1934). Historia e imágenes ariqueñas Rodrigo Ruz Zagal y Ariela González Yanulaque índice de autoresíndice de materiabúsqueda de artículos
Home Pagelista alfabética de revistas  

Servicios Personalizados




Links relacionados

  • En proceso de indezaciónCitado por Google
  • No hay articulos similaresSimilares en SciELO
  • En proceso de indezaciónSimilares en Google


Diálogo andino

versión On-line ISSN 0719-2681

Diálogo Andino  no.46 Arica mar. 2015 





Ann  Stoler*

* University of Michigan, Department of Anthropology. Michigan, Estados Unidos. Correo electronico:


It is an enormous pleasure and privilege to give this year's lecture in honor of Eric Ketelaar, such an inspired and creative figure for those of us whose work is shaped by the practices and perceptions of archives and by the challenge of how best to respect and register the provenance of documents and the rich possibilities for the future they serve and store.

Professor Ketelaar has long been an archivist ahead of his time, one committed to the complex sedimentations of historical process, a scholar willing to question openly how archivists might entreat historians to treat documents as part of a 'transactional process', only as important as 'the fabric of relationships and contexts' in which they are embedded and we as historians can bring to the fore (Ketelaar 1997).

Archiving as process

As a student of colonial histories, and Dutch colonial archives of the Netherlands Indies in particular, it is an argument that has defined how I see my task. Archival documents are neither originary fonts nor sources of fact to be 'extracted' and showcased as coveted jewels, apart from where they came. My work on Dutch, French and U.S. imperial formations has sought to demonstrate why our task as historians is not to engage in archival labor as an 'extractive' enterprise, but rather as an 'ethnographic' one. This latter enterprise attends to the densities and distribution in which these documents were produced, to their tone and temper, to how they moved, to the logic of appended documents and accumulations, to what kinds of authorizations and what claims about causation such weighty dossiers served. In this venture to read 'along the archival grain' of colonial governance rather than against it, details and slips of the pen matter a lot (Stoler 2009). Acerbic asides, crossed-out words and marginalia are storied notations in the making of truth-claims, evidence of the labor that went into crafting archival narratives and the deceptively smooth coherence of their authority. It is not that historians have not attended to such asides. It is rather a question how we analytically mobilize them, and what subjacent sensibilities we use them to trace.

I see analytic traction both in the rote repetition and deadening weight of officialese and in what prescriptions they harbor. But more of my attention has been on the quixotic shifts of what constituted 'context' for colonial agents themselves. Contexts were never givens. The choice of one over another could give credence to one set of truth-claims, discount the validity of another, or radically reframe the very temporal and spatial parameters of what circumscribed an 'event.' I take choices of context to explore epistemic commitments of colonial agents and actors - what people thought they needed to know and how they imagined they could best know it. In Sumatra's plantation belt of Deli, it is no surprise that the 1876 murder of a white planter's wife was 'explained' by different actors and witnesses in very different ways.1 Each rendition of the story, by the inept civil servant Frans Carl Valck (whose dismal record I have tracked for years), by Demmemi, the military commander more versed and concerned over the Aceh war, by the European planters bent on protecting their autonomy from Dutch authorities, produced different circuits of knowledge and different social topographies. Each traces an archival map not only with different coordinates, 'contexts' and different plots. Each depends on distinct notions of reliable evidence, whose evidence might garner authority, whose could be disqualified and dismissed - or not. What I have called these 'hierarchies of credibility' were not fixed. They are spliced through the folds of documents, rarely in explicit words, and underscored by what warranted repeating to a superior and what could not. Such concerns over credibility could invert rumor, turn

it from a derided form of evidence to one of value, as Frans Carl Valck did when he insisted on making the unpopular case that the murder of the planter Luhmann's wife and the mutilation of her children was 'caused' by and retribution for the brutalities of the planters. In Valck's account the 'barbarisms' were perpetuated by the planters 'in cold blood'; it was their hands that were sullied, not those of the Javanese who worked for them.

The new colonial histories that can emerge are not dependent on archives as sources alone, but on an appreciation that the very production of documents and the genres in which they appear (commissions, semi-official correspondence, district reports, classified inquiries) are not just traces of colonial bureaucrats securing their jobs. These are powerful technologies of rule where racial categories were altered, where new classifications were tried out, and where selective knowledge and disregard rather than more knowledge could confer more power. If an historical event may be defined as Michel Foucault once so succinctly did, as 'the breach of self-evidence', that which defies the rules and prescriptions of what came before, that which disrupts the unquestioned and received prescriptions that made up 'colonial common-sense,' it is precisely identification of such disquieting moments that has directed and drawn my attention (Foucault 1994).

Along the Archival Grain makes the argument that we historians have much to learn about the nature of colonial rule and the dispositions it engendered from the writerly forms through which it was managed, how attentions were trained and selectively cast. As Michel de Certeau some thirty five years ago wrote, 'transformation of archival activity is the point of departure and the condition for a new history' (de Certeau 1988). Archivists seem to know this already. Historians should, and for the most part do not. What De Certeau presciently had in mind were the effects of computers and digitization on what archives are imagined to be. But I think one could take his observation to create another future for writing richer colonial histories. Transformations of archival activity have been brought about not only by technological innovations, but as importantly by the changing circuits of knowledge production that governing institutions demanded and that colonial governance's paper trails bear witness to.

On this premise, I have sought to track the colonial order of things as seen through the record of archival productions, to ask what insights into the social imaginaries of colonial rule might be gained from attending not only to colonialism's archival content, but the principles and practices of governance lodged in particular archival forms. By 'archival form' I allude to multiple registers: commissions as a genre of their own, but also to how issues of confidentiality, criteria of credibility, categories of classification were framed, and not least to what sorts of documents moved rapidly through the bureaucratic and administrative corridors - or were abruptly arrested in their tracks.

Treating archiving as process rather than archives as things has a fan-like effect. It becomes flatfooted to treat colonial archives as only skewed and biased sources (as those of us in critical colonial studies have treated them for too long). Rather they emerge as entry points for understanding the regimes of truth that underwrote them, the logics and sensibilities that joined what historians have long taken to be unrelated histories and unrelated things. I have written and cared so much about the kindergartens and nurseries for European or 'Indo' children, domestic service in Dutch colonial homes, the orphanages for Inlandsche kinderen and the persons who attended to them only in part out of a contemporary feminist interest of my own. What drew me to these concerns was their presence and placement, often dead center in administrative archives, and most pointedly in the making of race. It was Ministers of Colonies and Governors-General for whom these sites and subjects mattered so much. The very location of such documents in the commissions on impoverished whites and 'needy Europeans' disturbs what counts as a 'security' concern: it throws into new relief how capacious the category of security was seen to be. It makes sense of the breadth of activities the watchmen of empire thought necessary to broach. In short, it allows us to track what domains of the intimate were both private and public, within the purview of governance because they were thought to be sites of potential danger to a racialized order in need of control.

Or let us take another set of unintelligibilities that render problematic a transparent narrative of a particular event. In Chapter 3 of Along the Archival Grain, 'Habits of a Colonial Heart,' I argue that the demonstration in May 1848 of Batavia's Dutch and creole city fathers against the schooling of their young sons in the Netherlands and the financial and emotional burden it placed on them, opens to a different history when family sentiments are not dismissed as the exaggerated pathos riled up by journalists whose papers needed to sell, but as part of the very substance of politics. The stress of family estrange-ments are just some of the details of the everyday that seem to spill over the edges of official format as authorities scrambled to learn what they could not really know. And they asked themselves and others again and again: whether the demonstration was confined to local matters, about family attachments, racial equality, or the contagious effects of something else - reverberation of Europe's spring 1848 revolutions, news of which came in newspapers and private letters that alluded to 'communist thinking' spreading throughout Europe, among the middle classes and working poor.

It is not that this story of the Harmoniehof 'disturbance' (gisting) in Batavia has not been told. But it has been told, even by such a renown historian as Cees Fasseur, as a straightforward story, rather than as one whose ambiguities and significance lay not in 'what happened' at the 'disturbance' turned demonstration, but in the uncertainty about what authorities thought it meant and where it might lead. Their mounting disquiet was evident in circulation of documents, in where and to whom reports were sent, and most palpably in the impatient tone of authorities, ill at ease with certitudes they sought and what they did not know. If archival science was produced out of the rationality of a bureaucracy as Max Weber would have it, there is as much in the circulation of documents that suggests that it was not rationality alone that put Dutch troops readied with arms on the outskirts of Batavia in March 1848, but fear that the familial attachments and sentiments of the Dutch colonial administration's very own agents, partisans and supporters could not be controlled.

In short, it is work with the archives, not in them, not abstract theory of Derrida, De Certeau or Foucault, that may generate new conceptual formations. It is in refashioning and reworking what constitutes the boundaries of the colonial archive and in how we conceive of 'content' that is key to making new histories of our own. It is by attending to the paper trails of epistemological hesitation (when officials ask, 'how should we designate the Inlandsche Kinderen and how can we know who is really one?'), the traces of affective strain (when Valck cries out in an otherwise staid missive 'please do not think that I write in a moment of agitation, please do not think that I exaggerate... believe me, I do not see the situation blacker than it is') and of political anxiety as in May 1848, that an otherwise 'minor' event can prove to be a diagnostic of uncertainty and the fear of unreason within the ranks of colonial agents themselves. A minor event allows something different: to craft a colonial history in what I have called 'a minor key' - neither biographic nor personal alone.2 Rather, it is one that binds official decrees to a child's estrangement, to a father's heartbreak, to a mother's madness, to nerves that expose the body and sear the soul.

Lessons of apprenticeship

For Aristotle, the historian's task was to 'tell about things that have been' and pxoets 'about things as they might be', but those of us who work on colonial histories know better. Colonial historiographies may be written in the past tense, but colonial documents are not. They cannot be broached in the passe compose. For these are documents that were visions of what a colony could be, rehearsals of what was thought to be a dangerous precedent, fears of what it might become, made up of strategies of defense against enemies in the flesh and as yet unrealized dangers in formation. The colonial archives that I know best of French authorities in Indochine and Algeria, and the Dutch in Sumatra and Java, are documents often in the subjunctive, expectant, conditional tense. They address more than unrealistic and unrealized visions. They allow us to 'develop historical negatives'. Infused with speculations about possible futures, they are diagnostics of the fears about a present that its authors hurry to relegate to the past.3

This conditional tense, this 'if only x, then y is possible' conveys more than prevailing fears. It taps into the political logics of governance, the forms of political reason that cut across metropole and colony. It situates the politics of comparison as central to making racial grammars commensurable, colonial congresses feasible (how else could colonial 'experts' from diverse colonies and diverse fields construe such a subject as 'the mixed-blood problem' about which they all presumed to share and know?), and taxonomic categories sufficient to designate some people as threats, others as potential partisans, and still others as dispensable.

In this current project, I take those insights and lessons from archival labor outside of the designated official colonial archive proper in a different direction, one that seeks to think differently about what a colony is, one that asks what unrealized histories might be gleaned by exploring the portability of documents, and the trail of their citations that transgress imperial borders and colonial sites. The problematic is straightforward: what might a genealogy of imperial practices look like that neither succumbs to a teleology of the present nor is one restricted to the policed categories of colonial archives themselves? Such a venture opens to the political logics that joined penal and social reform in the mid-19th century and ultimately the conceptual and political matrix that joins colonies and camps.

Archival labor of the sort briefly described here, provides lessons of apprenticeship in reading along the archival grain for a much wider set of histories than those derived from catalogued archives alone. Archival labor makes one attend differently to reference within documents and cross-reference between them. It trains our analytic sensibilities on placement and the arts of culling and cribbing. It trains one to be attentive to verbal tense, to track how histories are made, fixed, and rewritten in the very process of 'merely' reporting on an event.

This paper looks to the political lives of documents, to the logics that underwrote their citations, and not least to the powerful work of comparison (what I call the 'politics of comparison') and what it tells us about how commensurabilities were drawn across seemingly disparate domains. Tracking what colonies were and intended to be, it attends carefully to the future conditional tense as it charts a geopolitical map that spans both the colonial archives proper as well as those that hovered outside its constricted edges to think differently about the scale and scope of imperial formations and the unstable political logics that underwrote their expansive and varied projects.

It wrestles with the challenge to think about what methodological renovations we might need to write histories that yield neither to smooth continuities nor to abrupt epochal breaks, but that rather capture the uneven, recursive qualities of the visions and practices that imperial formations have animated, both what they have succeeded and failed to put in place. I see such new colonial histories not as a rehearsal or condemnation of the past, but rather as tracing the jagged lineaments, political scissions, and some of the deep fault lines of our world today.

The political life of a document

There are many ways to imagine the political life of a document.4 We might explore the relations it forged, the places it inhabited, the sentiments it provoked, or the hands it passed through. Alternately, that life might manifest in the singular boldness of the intervention made at the time of its writing, or the affective relations mobilized or bore witness to. Or it might be found in the contrast between an unremarked life when it came into being and its capacity upon later retrieval to breathe sustenance into political truths long denied or demands unfulfilled.

But in thinking about the political life of a document here, I want to engage in another sort of exercise, with another set of questions and provocations in mind. Can a public document have a political life if it is rarely referenced, or elevated to the noteworthy status that warrants quotation or censure? Can its political life be lodged in its unexceptional quality, in the density of similar documents that surround it, in its rehearsal of common refrains?

The answer depends on what we imagine a 'life' to be. Here I want to think of the life of a document as the kinships it recognizes, the genealogies it may ignore, and the alliances it calls up and commands. Here I think of a 'life' as a lively node of connectivities, concerns and preoccupations shared and exchanged, commensurabilities sought and assumed, proximate and distant affiliations affirmed. All these give contour to futures aborted and tenacious attachments that might no longer be. What interests me are documents that strain our conceptual categories and chronologies, disturb our contemporary common sense about what words belong with what things, disrupt the confident lineages that students of European and colonial history so assuredly trace, call into question the notion of 'context,' confounding what we imagine should be deemed 'in' or 'out' of place.

The particular document I look at here provokes such an unsettling precisely because its citations and comparisons seem unfamiliar, counter-intuitive and out of place. Most strikingly, it locates what were called 'agricultural colonies' ('les colonies agricoles') for impoverished, orphaned, abandoned, and/or delinquent children and youth (hailed as one of the most successful social reformist projects of mid-19th century Europe) in an archival field and political web of linkages that prompt historians of European child welfare and those who work on Algerian military camps to reconsider both in a broader arc of governance - and as I will argue - as part of a wider imperial domain. It is a document whose content breaches institutional forms that seem incommensurable in our present. It is in these couplings and affiliations that I broach the multiple arts of imperial governance and the shifting geographies of imperial interventions.

Thus my intent is three-fold: one, to trace the political life of this document through its self-defined subject, these colonies agricoles; second, to identify the unlikely kinships made through this minor node; and third, to use these kinships to plot a geo-political topography carved out of the distribution of containments of peoples, strategies of displacement, definitions of security, and tactics of defense.

Together they make up the genealogies of what I call 'the imperial modern.' I use the term to think through some of the principles and practices by which such institutional arrangements were blatantly and oblique joined: in relations of force and philanthropy, of care and coercion, of curative and punitive instruments of constraint and enclosure that were forged, fumbled, dismissed, reversed, and reworked again.

Colonies agricoles

Regimes of security and containment share a deep temporal and spatial imperial history of their own. They are predicated on making common sense of the categories of persons against which society must be defended and reshuffling this membership as situations change. They produce an archival map made up of failed experiments and exuberant visions that sustain urgent reassessments of those dangers to be defended against.

So let me introduce the document. It is a public document, available in libraries. Two, it is an essay of some five hundred pages, in four volumes, written by a certain Count A. de Tourdonnet, printed in Paris in 1863 under the title An essay on the education of poor children: The agricultural colonies (de Tourdonnet 1863). It was neither an exceptional essay nor the first of its kind, rather part of a proliferating if minor genre of reformist writing from the corridors of law, penal reform and philanthropy, most marked between the 1830s and 1880s. Intensifying after the revolutions of 1848 throughout Europe in ostensible response to the political threats such 'disorder' was seen to pose, literally tens and thousands of pages of detailed commissions, reports, and multi-sited surveys were produced to assess, project, and give substance to the viability and widely shared conviction that a return to agriculture and an attachment to land and labor were economic, political and moral solutions to alleviate urban pauperism, quell the revolutionary potential of the Europe's urban masses and avert a new generation of wayward and abandoned children and youth who would otherwise turn to revolt and crime.

The children's agricultural colonies thus described were understood as reformist enterprises in and for France, designed to remove children from adult prisons, and rescue them from moral harm.

It is in this general context that they long have been studied.5 Some historians have traced their rapid spread to progressive Catholic philanthropic initiatives, while others have compared them to utopian agrarian settlements in their communal vision. Still others have understood them as modernized variations on the penitentiary system, elaborations on the houses of corrections that were turned in the 1830s into colonies penitentiaires. Others still situate the colonies agricoles within a long duree of confinement of the poor, and especially of children, dating back to the seventeenth century.

If some historians have emphasized pedagogic goals, others have described them as 'depots of mendacity', or more critically as 'prisons in the fields'.6 Colonies agricoles came in many forms, with proliferating distinctions among them, about which social reformers rarely agreed. This taxonomic work took up much of the time. A pedagogic model of relief circulated, crisscrossed and was continually 'contaminated' by programs and practices of incarceration: both passed through 'dressage' of the body, and severe punishment if that training was not fully adhered to or achieved. Even a sketchy graph of their changing taxonomic grids yields a striking observation: how many features of what De Tourdonnet would distinguish as religious colonies, military colonies, settler colonies (colonies de defrichement) and what he called, 'charitable agricultural colonies for children', were sustained by principles and procedures shared among them. Distinctions among the children's colonies were especially hard to draw.

De Tourdonnet devised a new typology to differentiate those devoted to poor relief from those designed for repression and to distinguish both from those focused on education. But harsh discipline and poor relief marked them all. The overlap was not a shortcoming of De Tourdonnet's sorting methods. As we shall see, this morphing and meshing of principles and practices were key features of how this knowledge was assembled and disseminated across the globe. If anything, De Tourdonnet's essay might rather serve as a diagnostic of the multiple logics that underwrote them and thus of what constituted a 'colony' and who belonged in them. Orphans, children who had stolen a loaf of bread, and those who slept under the bridges of Paris came under one rubric and one system of internment in part because they were subsumed under the same constellation of political logics: to deter a new generation through rehabilitation from becoming enemies of the state; to defend society from a potential threat, and to secure France's colonies with productive and loyal settlers.


Among the most famous and allegedly most successful of these colonies was the colonie agricole et penitentiaire of Mettray whose very descriptive label (combining penitentiary and agricultural colony) confounds its curative project. Mettray was lauded as exemplary from its founding in 1840. Among all the children's colonies created throughout Europe, none was more often visited, depicted, and discussed by dignitaries and officials from, among others, Britain, Belgium, Germany, the Netherlands and the U.S.7 Mettray's founder, D.A. Demetz, a former magistrate, believed his motivation provided a moral mission to assure that poor, abandoned and delinquent children should not be just punished, but saved and redeemed as citizens through strict discipline, agricultural labor and supervision based on a familial model whose guardians would be assigned as 'fathers' to train and care for children's bodies and souls.

Mettray quickly became the paragon of private initiatives in social reform. Its fame was not accidental. Demetz was a philanthropic entrepreneur who encouraged high placed visitors who he housed in his 'Hotel de Colonie,' built just outside the colony's grounds. Copying Mettray became virtually obligatory to procure funding for the sixty-odd establishments created over the next decades, all claiming to follow Mettray's moral and organizational model. Some even borrowed its name, like the Mettray colonies in the Netherlands or the shortlived 'Petit Mettray' in Northern France.8 Such colonies agricoles were to be 'seedbeds' to raise honest citizens and hardworking laborers with limited aspirations, by removing them from the unhealthy immoral cityscapes of indigent adults and by investing their time in soil and soul. For De Toudonnet as well, Mettray was a benchmark against which to measure the costs and benefits for reform of the new colonies he proposed.

If these children's agricultural colonies had iconic status in the mid nineteenth century, they have since been endowed with more. This is not due to their reformist success (about which historians disagree), but because Michel Foucault singled out Mettray in probably his most widely read book, Discipline and Punish, to mark what he saw as 'the completion of the carceral system,' and the condensation of 'the art of punishing that is still more or less our own.' As he put it:

Were I to fix the date of completion of the carceral system, I would choose not 1810 and the penal code, nor even 1844 when ... the principle of cellular internment was passed ... I would choose ... 22 January 1840, the date of the official opening of Mettray. In the arrangement of a power-knowledge over individuals, Mettray and its school marked a new era (Foucault 1974).

He saw it concentrating all the coercive technologies of comportment and behavior through meticulous and multiple forms of training - for its staff (its 'technicians of behavior' as he called them) and for its 'colons' (colonists), the children themselves. In his hyperbolic style, he dubbed it 'the first training college in pure discipline' and in the 'art of power relations'.

But Mettray, and the colonies agricoles that so riveted De Tourdonnet, was also something else, central to Foucault's larger conceptual project: an opportunity to identify 'a whole series of institutions' which constituted, what he famously called, 'the carceral archipelago', a 'gradated', 'far-reaching' network of 'compact and diffuse institutions and methods' that were to be both curative and punitive. The carceral archipelago for Foucault was part of a new social apparatus, a dispositive - 'a subtle, graduated, carceral net' - a form of modern discipline that turned away from the spectacle of punishment to the isolation of the cell. As importantly, it was an emergent vehicle in the power of normalization that collapsed into a single entity the 'multiple dangers of social disorder, deviance, madness, and crime'.

De Tourdonnet's children's agricultural colonies may have represented the 'soft' end of the carceral spectrum (as Foucault himself quipped, the choice of Mettray was 'somewhat unjust'). But that was precisely where his favored site for analysis would rest, on 'the frontiers' of techniques that erased distinctions between the undisciplined and the dangerous. The agricultural colonies that superimposed five models - part prison, cloister, family, school, and military regiment - embodied a quintessential carceral form.

This is a compelling and now famous story shaped by Foucault's dramatic prose. But it also a partial one, schematic, and skewed in time and place. Mettray was a node in a network, many of whose linkages Foucault chose to bypass or at least had not yet sought to name. What was 'new' for him about Mettray was not only how effectively it combined multiple models outside of and within institutions, but their dispersal across a broad societal domain. French historians were quick to critique his generalizations, noting that Mettray was unique, with practices that were almost never followed. Unperturbed, Foucault merely insisted that they missed the point.

His response to questions as to why he chose the prison as the center of new penal system and not deportation or the penal colonies, however, is curious. He claimed to prefer not 'to dwell in universals but in work that takes place between unfinished abutments and anticipatory strings of dots'. But if these were his chosen sites, the oddly selective genealogy deserves attention. Despite his emphasis on gradations of punitive and curative arrangements within the carceral archipelago, those beyond northern Europe had little interest for him: settler colonies made up of French soldiers, labor camps of Europeans that served colonial public works projects, and most notably penal colonies, like those of New Caledonia, and French Guiana, scattered across the topographies of Europe's empires. As anthropologist Peter Redfield rightly notes, penal colonies were not 'marginal spaces on the edge of the nation' as Foucault located them, but central technologies of it (Redfield 2000).

Still, for Foucault, these kinds of colonies remained outside his territorial, political, and analytic frame. In brief mention of the possibility of a broader carceral net in Discipline and Punish, he writes that 'the example of colonization comes to mind', but then dismisses it as 'not the most convincing example'. But this begs the question: 'not the most convincing example' of what? Of discipline? Surely not. Of normalization? Perhaps. In the one reference to colonialism, he dismisses all the projects that were drawn up '... for the delinquent, undisciplined soldiers, prostitutes and orphans to take part in the colonization of Algeria, that colony was formally excluded, by the law of 1854, from becoming one of the overseas penal colonies: in fact deportation to Guiana or later to New Caledonia had no real economic importance'.

But this too is strange. Economic efficacy was never a sufficient benchmark to account for the political imaginaries he sought to name. Failed projects were the fabric of experimentation and crucial sites of 'political dreams', indices of the anxieties and fears that produced the very categories of person to be contained. One methodological insight of genealogy is precisely a refusal to distinguish among the implemented, the failed, and imagined. Differential histories are the product and productive possibility of genealogies that track dispersed and disparate beginnings.

Genealogy traces histories in a piecemeal fashion from alien forms. And this is exactly what De Tourdonnet did. Although Algeria was formally excluded as an overseas penal colony in 1854, it was not, as he shows us, excluded as a site of children's colonies agricoles in the same years. In fact, he and others promoted Algeria as a renewed and excellent site for them. The children's colonies agricoles dotted throughout rural France, serving to raise strapping youths with a love of the 'sweat of labor' and a ready love of the land were imagined to become the fodder for colonizing Algeria for France's expanding empire. The ties between children's agricultural colonies and the settlement of Algeria are threaded through De Tourdonnet's essay, through the logics of colonization, punishment, and rehabilitation of youth, and as we shall see, drawn as Foucault conceived it, in 'a piecemeal fashion' from 'alien forms' (Foucault 1984:76-100,78).

De Tourdonnet places Mettray in another frame: he notes, the 'transplantation of welfare children in Algeria was nothing new' in the 1860s. Earlier commissions from 1832, 1852 and 1856 had too sought 'to bind' destitute French children to Algeria. The St. Ilan orphanage in Brittany proposed such a project to the Emperor in 1830. In the 1850s, more than 200 children were sent from Paris to agricultural colonies outside Alger to learn how to cultivate the soil, under the supervision of a well-known Catholic Father. Eleven more regional departments were granted subsidies to do the same. Two hundred more children were sent to Father Abram with the support from the war ministry - a failed project because the children were too young to work and ill-prepared for the climate. It is not noted whether they died, were sent back to France or endured some other fate.


Warning against taking small children from their nursemaids (and arguing In any case, that French authorities in Algeria would be unable to absorb them), Tourdonnet proposed a three-stage process to better prepare children for a hot climate and strenuous labor. They would be sent first to 'preparatory colonies' from ages one to twelve (colonies with nursemaids), then to 'colonies of transition' for those aged 12 to 14, where adolescent bodies might be first 'bronzed by the sun of Provence' and finally to 'colonies of application,' for those aged 14 to 21, primed to be cultivators and equipped with a disciplined cultivation of the self. As a commission published two years put it, 'The goal is not to produce flabby townsmen but to form robust workers, hardened to fatigue who have the courage for the job, for whom labor is the life of their organs' (de Lurieu et Romand 1854:192). This was not the only way in which Algeria was so vitally caught up in the carceral net. The agricultural colonies established there morphed between permanent agricultural colonies, detention centers, and temporary military settlements. That historians offer inconsistent and confused accounts of the numbers of colon who left for Algeria between 1848 and 1851 - and the conditions under which they did so - may partly reflect the multiple streams of influx and the settlements' rapid change of function. By some counts, 14,000 Parisian workers were sent to Algeria in September 1848, only some of whom were political deportees. Three thousand republicans were sent after Louis Napoleon's 1851 coup. Others estimate that nearly 15,000 people were arrested in 1851, of which 6,000 were deported Republicans. Disorder in Paris led to 450 political deportations to the Algerian penal colony of Lambesc. Significantly, there were also deportations to at least six former colonies agricoles that had been newly converted into colonies penitentiaires (penal colonies). Colonists from Malta, Italy, and Spain as well as other Parisian workers encouraged . . . to emigrate voluntarily were added to the mix of French soldiers established in villages-militaires in 1840. Thousands more were recruited under an intensive colonization program to make colons out of a toxic mix of ill-prepared urban poor, alternately referred to as les sans-travail, les revoltes, and les deracine. De Tourdonnet's account traverses some of these morphings, noting that the Director of Algerian Affairs in 1856 was commissioned to establish a large farm, with 'concentrated buildings' populated with 500 colons under military protection and surveillance (de Tourdonnet 1863;III:66). Specific enterprises failed again and again, butspecific enterprises again and again, but what is more striking is the resilient commitment and development of a dispositif of displacements and dislocations, of banishments that marked imperial expansion and ever new reinvestments in its potential success.

Transpositions of function

This is a truncated account but the point should be clear: this shift of function produced an experimental space in which an agricultural colony could serve as a military zone, and a penal institution could be transformed into a 'normal,' civilian colonial settlement. Transpositions of function brought penality, philanthropy, social welfare, and imperial conquest and intervention into proximity and adjacency, mixing strategies that have lasted long past the mid nineteenth century. Olivier Cour de la Grandmaison argues that the colonies agricoles were part of double system to disgorge France's carceral institutions and to protect the social and political order through colonial deportation. Marc Bernardot documents a sedimented history of internment that used the legalties of psychiatric internment in 1838 and 1912 vagabond laws as a model for future decrees of political internment within France. Fusing traditions of control and surveillance, successive waves of populations were treated in the same sites, most notably for colonial subjects, at least until the 1960s.9

Such observations make more curious Foucault's insistence that his project was to write a history of the present 'of the arts of punishment that are still very much our own' for his history abruptly stopped short of these exemplary instruments to cordon off specific populations under modern governance. The political philosopher Giorgio Agamben has famously called the paradigmatic 'political space of modernity' the camp, not the prison, the hidden matrix of our time, borne out of a state of exception that became a norm. But Agamben's genealogy is selective, his archive invisible, his history thin as well. Like Foucault's, Agamben ignores the deeper archives of empire (leaving historians to quibble over the 'trivial' point of whether the first camps were those of the Spanish in Cuba in 1896 or those in which the British herded Boer families in southern Africa at the turn of the century). Nor were these states of exceptions. Rehabilitation camps, re-concentration camps, forced labor colonies and agricultural colonies for children were only as provisional and exceptional as imperial formations themselves. These were forms of enclosure and containment that were part of the very architecture of empire, not vestiges of earlier systems of order and control. They are predicated on moving categories and populations that 'create new subjects that must be relocated to be productive, dispossessed to be modern, disciplined to be independent, converted to be human, stripped of old cultural bearings to be citizens, coerced to be free' (Stoler & McGranahan 2007:8). Nor were these confined to nation-state projects as Agamben claims but geographic arrangements fundamental to defining and maintaining the racial distributions of the Imperial Modern. That Foucault never touches on the camp as a singular carceral entity is all the more surprising given that his acknowledged inspiration for his use of the term, 'archipelago' was a direct reference to Solzhenitsyn's Gulag Archipelago, published just the year before Discipline and Punish. When questioned in 1976 about his geographic metaphors, he had only this to say: '.only one notion here...' (Foucault 1980:68) is truly geographic, that of an archipelago. I have used it only once . to designate via, the title of Solzhenitsyn's work, the carceral archipelago (how) ... a form of punitive system is physically dispersed yet at the same time covers the entirety of a society.

But why does his carceral archipelago, directly inspired by penal colonies, exclude them? Perhaps these 'epochal' distinctions, as Paul Rabinow argues, actually get in the way. In subsequent years, Foucault will abandon these sharp breaks between the classical and modern ages for a more subtle, recursive historical formulation.

One could imagine that the carceral archipelago would look different in this new analytic space, that it would perhaps not have excluded the Gulag as part of the modern, nor the military run agricultural colonies of Algeria, nor the re-establishment of penal colonies in French Guiana and New Caledonia, nor the isolated forest hamlets where Harkis, Algerians who served as auxiliary military for the French during the Algerian war, were dumped for decades throughout rural France. Nor might it have looked past the successive waves of populations housed in the internment camps of Larzac and Thol in France that grew up out of 19th century traditions of internment.

We would do well not to follow the ease with which Agamben generalizes the camp. Genealogy and historical specifity tell another tale, one animated by multiple logics at play. As Peter Redfield notes these logics of the penal colony '(ed) between the need to colonize, the need to punish and the need to reform.' In the logics of governance, nothing has been more normalized than the varied forms of colonies and camps whose genealogies have long linked defense of society against its dangerous elements to the varied security tactics of imperial formations.

The 'breach of the self-evident'

Again, taking an historical event to be the 'breach of the self-evidence', the relevant event is not Mettray's opening or the advent of the colonies agricoles, or the strategic relocations that accompanied France's conquest of Algeria alone. It is the elaboration of a carceral network that was inclusive of all three. The history of les colonies agricoles has been written through the prism of Europe's metaphoric carceral archipelagos, but more evidence favors an account of its complex connectivities through a broader geopolitical space, what I would call a 'carceral archipelago of empire'. Tracing this network through other sites and other lines of force does more than fill in peripheral colonial gaps. It reconfigures the carceral paradigm, its constitutive elements and not least alters our temporal frames and archival maps. Changes in the penal system were tethered to revised colonizing projects in nearly every decade of the 19th and 20th centuries. Such a perspective reveals new entanglements between institutions and visions whose histories have been severed and pulled apart. A new cartography entails a new conceptual map, dense with shared knowledge and newly fashioned recombinations of experiments in discipline, defense and biosecurity across imperial maps.

De Tourdonnet's extended essay on les colonies agricoles alerts us to just this, as it benignly endeavors to situate les colonies agricoles as potential sites for securing imperial frontiers, as instrumental in the imperial dispossession of land, and in the military strategies designed to call on children in the agricultural colonies and on freed penal colony inmates as a frontline of settler colonialism, with the adult 'colon' as a reserve army for defense. Foucault's development of the concept of a dispositif after writing Discipline and Punish captures some of this movement in uncanny ways. A dispositif is neither a device nor a technology as it is so often translated but a 'network,' (a reseau) a 'heterogeneous ensemble of institutions, discourses, architectural forms, scientific statements, moral and philanthropic propositions.' The apparatus is the system of connections between them. It is not a 'thing' but an active 'response' toward an 'urgent need.' Its commanding feature is an 'interplay of shifts of position and modifications of function that can vary widely.' What is key to the connections that link these coercive and curative projects, that which muddles the space between colony and camp is a political matrix of managed mobilities. This I would argue is how we might rethink the carceral archipelago of empire. It is precisely how the colonies agricoles trans-mutated with multiple functions, something De Tourdonnet seemed so readily to grasp.

Social and political horizons

Let us look at the social and political horizons of his vision for the children's colonies and their spatial coordinates. De Tourdonnet's essay begins not with the children's colonie agricole in Northern France, but with a prelude to comparisons across the imperial world: to the initiatives of Catherine II and her Russian imperial successor to create a fortified empire within its interior frontiers and on its external borders. Her projects were to establish colonies on Russia's vulnerable borders as well as on the outskirts of Moscow; to house abandoned children in rural colonies on the fringes of St. Petersburg as well as the 'remarkable Russian colonies of Saratoff in the Volga basin', made up of German immigrants recruited to settle and colonize Russia's steppes. He takes note of the penal colonies in Guiana, New Caledonia and both the colonies in Siberian and the colonie agricole in Algeria at Lambesc, reserved for political dissidents (de Tourdonnet 1863;I:32). Connections that seem to have escaped Foucault's carceral archipelago are placed full front on De Tourdonnet's cognitive, visionary, and political maps.

Secondly, De Tourdonnet's use of the terms colon and colonies are telling in their referents. His purview includes religious colonies, military colonies, failed 17th century pauper colonies and those active in mid 19th century. There are children's agricultural colonies as well as retirement colonies for the infirm. A colonist could be a new French settler in Algeria, an adult inmate in a private establishment for paupers outside Paris, a penal colony inmate in French Guiana, an orphan in a colonie agricole in Provence. His cartography spills across institutions of care and correction, dotted across a vast imperial space. A historian of 18th century empire might find it familiar, evocative of the well-trafficked networks within and among empires themselves. Relocations of populations, dislocation of dangerous elements, and deportation of dissidents were fundamental features of imperial governance of a longue duree. Historian Kerry Ward describes 'an imperial network of forced migration' in the 1700s with multiple circuits that joined Java to Ceylon and the Cape of Good Hope, networks that were critical 'to maintain civic and political order in the posts, settlements, and colonies that comprised the East India Company's territorial nodes' (Ward 2008).

De Tourdonnet's reference to these disparate forms captures something that contemporary studies of empire have since lost: how different notions of a colonie should be organized and who should rightly be in them were shared and compared. Imperial expansion and modes of confinement, resettlement of delinquents, pauper programs, and the recruitment of empires' pioneers were not separately conceived and executed projects with different architects. Nor did De Tourdonnet conceive them as wholly interchangeable or all the same. These were political rationalities not settled, but in formation. If carceral concerns and care for poor were the emphasized goals, they were equally under-stood as connected and sometimes contiguous sites to experiment with and implement the convergent tactics deployed for 'security' and territorial control.

De Tourdonnet teaches us something else. French blueprints for agricultural and pauper colonies drew on imperial rearrangements of people as much as the other way around. But imagining a direct exchange between metropole and colony misrepresents the circuits of movement, the comparative work of global reach, and the seemingly disparate forms of containment and displacement that were conjoined. What appears now as a seemingly indiscriminate range of comparisons references and revives what were once politically tethered terms, both shallow and deep historical connections. Ambiguous nomenclatures, substitutable populations, partial appropriations were part of the precocious movements of people and things that secured the viability of imperial formations. That De Tourdonnet could look at once from the Crimean colonies in the Russian south to those established in the Amur basin on the Chinese frontier, underscores not only the sweep of his comparative vision but the poaching of practices that so marked cross-imperial knowledge acquisition and application.

The date of 1840 may mark the opening of an emblematic art of punitive and curative power in Europe, but it is the surreal drawings of Mettray with a replica of a sailing ship appearing next to its church in the central courtyard that speaks to promised opportunity and forebodes the violence of something else: on the one hand, the forced and free settlement in Algeria of political exiles, convicts, military, and children who would eventually become free and reliable colons; on the other, the racial logic of military and settler occupation of Algeria that dispossessed Algerian families, by seizing their land, murdering thousands, and burning their fields and homes. If 'knowledge is made for cutting', for severing received categories and common sense, De Tourdonnet's essay slices with sharpened shears (Foucault 1984:88). It offers the historian the opportunity of making those cuts herself: his hesitant attraction to the central logic of 'displacement' (deplacement) of children joined his temperate case for how the colonies agricoles could one day make Algeria a 'second continental France' (de Tourdonnet 1863:162-4).

Carceral networks

I came to Tourdonnet and Mettray by a more circuitous colonial route, at another node in the imperial carceral net, at another site of 'urgency.' It was in the context of colonial documents with which I had long worked, a charged debate in the 1850s about Java. Dutch colonial authorities spent decades worrying and rehearsing the dangers posed by what they saw as a rapidly increasing population of 'wavering classes,' imagined as a threat to the security of the colony and to the clarity of its racial divide.10 At the center of their anxiety was a category that defied their social categories, the 'so-called' Inlandsche Kinderen, the destitute mixed-blood descendants of native women and Dutchmen who fathered the children but refused to legally acknowledge them as their own. Their numbers eluded officials because the population was amorphous, neither easily identified or named. Sometimes it included pauper whites born in the colonies, but always those who lived on the economic, moral, and social fringes of colonial society proper and who were deemed fabricated - or 'quasi-European'.

Through much of the mid nineteenth to early twentieth century, colonial officials designed an extraordinary set of minutely conceived plans to confine their movements, disrupt and reconfigure their ties to the native population, and to nurture docile dispositions toward the racial impositions of the colonial state. Some of the visions turned on making them into artisans and craftsmen, others turned on making them into an industrial working class set apart from native workers. But the most popular if infeasible plan envisioned a return to the land for those who had never lived off it, reclaiming a mode of existence that would forge a new settler colonialism and loyal colonial citizens. And it was here that Mettray emerges in the archives as a preferred model, unadorned with its history, both in and out of place. In this vision, Mettray-like colonies would dot the Indies archipelago to instill bodily discipline, limited aspirations, love of the soil through supervised training and care for the soul.

The Dutch colonial appeal to Mettray was but a small node in a broader carceral net. Mettray's founder, Demetz, as Tourdonnet and others noted, drew his inspiration in turn from Van den Bosch, who established the first private agricultural colonies for the criminalized poor in the Netherlands in 1817 and a forced cultivation system for Javanese farmers lasting some thirty years, making what one historian has called a 'para-penal colony' out of Java. His forced cultivation system was oppressive and his benevolent colonies in the Dutch countryside were an equivocal success. Still De Tourdonnet notes his efforts with admiration.

Demetz also looked elsewhere. In 1836 he visited Philadelphia's penitentiary to learn more about its acclaimed, innovative system of cellular isolation, as De Tocqueville had done several years earlier.

De Toqueville also praised Heurne de la Pommeuse's 1832 treaties on the children's colonies agricoles that De Tourdonnet often quoted. De Tocqueville and Huerne de la Pommese articulated a political logic that joined the children's agricultural colonies to colonization, and both to the domestication of the poor, the dissident, and new subjects of empire. It was also of course De Tocqueville who aided in casting the carceral net back through the conquest of Algeria11

The carceral archipelago of empire could be conceived not as a prelude to the modern but the gradated sites of its instantiation. New concepts provide, as the philosophers Deleuze and Guattari write, 'the conditions under which not only subject and object are redistributed but also figure and ground, margins and center, moving object and reference point, ... length and depth.' The political life of De Tourdonnet's essay takes shape in the comments and critiques he makes about a range of containments and enclosures, about their principles and practices - in the kinships he turned from and those kinships he sought to claim.

These carceral networks were literal archipelagos but also figurative in more legal senses than Foucault might have envisioned. These were zones of semi-extra territorial status, part of but removed from certain legal restrictions, inside and outside the nation proper. Like the detention and refugee camps in Europe today removed from the polis proper, they were clusterings of types of persons, both 'of' and 'at' biopolitical risk, relegated to the edges of legality, to the outskirts of the nation and on the edges of empire, but not outside security, surveillance, and intelligence networks or their visionary bounds.

The crafting of new imperial histories

The projects designed to house children in France's agricultural colonies for conquest and protection hardly got off the ground, but the political logics that underwrote them have left indelible watermarks of empire that trace the lineaments of how broadly security regimes were conceived.

There are no straight historical lines that lead from these colonies agricoles to Guantanamo. Nor do the late nineteenth century reconcentration centers in Cuba or the German camps for the Herero provide a direct link to Nazi concentration camps or humanitarian refugee camps, as some German historians now claim. This is not the point. Reorienting our archival and conceptual maps has to do more than reverse directionality, revise chronologies and the spatial scope of European history. Rather it raises new questions about the shared spaces, forms of knowledge, types of instruments, and persons dislocated, dispossessed, or driven in and out of capitivity and into circulation. An imperial genealogy of security regimes is not an inexorable teleology of the present. Rather it might chart how political common sense is forged, how relational histories are severed, how specific places are retooled with substitutable populations in them. Such a genealogy would ask what durabilities of empire shape the distribution of displacements and the logics that underwrite containments today. Minor documents like De Tourdonnet, so broad in its scope and innocuous in its claims, offer an oblique if chilling perspective on them.

Attention to the entailments of archival labor can make possible the crafting of new imperial histories, providing us with the discernments that allow us to better trace the connectivities between documents and what we imagine those connectivities to look like and in what circuits of knowledge production we expect them to be. As De Certeau so well understood, copying, binding, classifying are all practices not of a preordained history, but 'of a history to be made'(de Certeau 1988). The challenge in writing new colonial histories is just that: not histories that follow the well honed paths that historians have travelled, but the possibility of imagining other ways of making what De Certeau saw as a critical task: to make, as he put, 'an entirely different history possible'.



1 See Chapter Six, 'Hierarchies of Credibility' in: Stoler, Along the Archival Grain.

2 See Stoler, Along the Archival Grain, Chapter Seven, 'Imperial dispositions of dis-regard.

3 See Stoler, Along the Archival Grain, Chapter Four, 'Developing historical negatives.'

4 Parts of the essay which follow were first delivered as an opening address for the conference 'The Political Life of Documents' at The University of Cambridge in 2010 and at The Max Planck Institute for the History of Science in Berlin in 2011.

5 See, for example, Ceri Crossley, 'Using and trans-forming the French countryside: The 'Colonies Agri-coles' (1820-1850),' French Studies 44 (1) (1991), 36-54.

6 See, for example, Christian Carlier, La prison aux champs: les colonies d'enfants delinquants du nord de la France aux XIXe siecle (Paris, les editions de l'atelier, 1994).

7 See Sophie Chassat, Luc Forlivesi et George-Francois Pottier, Eduquer et punir: La colonie agricole et penitentiaire de Mettray, 1939-1937. (Rennes, Presses Universitaire deRennes, 2005).

8 On the Dutch Mettray, see Jeroen Dekker, Straffen, redden en opvoeden: Het ontstaan en de ontwikkeling van de residentiele heropvoeding in West Europa, 1814-1914, met bijzondere aandacht voor 'Nederlandsche Mettray'. (Assen, Van Gorcum, 1985).

9 See Marc Bernardot, Loger les immigres. La Sonacotra 19562006 (Editions du Croquant, 2008) et Camps d'etrangers (Editions du Croquant, 2008).

10 See Stoler, Along the Archival Grain, Chapter IV 'Developing Historical Negatives.'

11 See, for example, Jennifer Pitts, ed. and transl., Alexis de Tocqueville: Writings on empire and slavery (Baltimore, Johns Hopkins Press, 2001) and Seloua Luste Boulbina, ed. Alexis de Tocqueville, Sur l'Algerie (Paris, Falmmarion, 2003).


References Cited

Bernardot, M. 2008 Loger les immigres. La Sonacotra 1956-2006. Editions du Croquant.         [ Links ]

Carlier, C. 1994 La prison aux champs: les colonies d'enfants delinquants du nord de la France aux XIXe siecle. Paris, les editions de l'atelier.         [ Links ]

Crossley, C. 1991 'Using and trans-forming the French countryside:The 'Colonies Agri-coles' (1820-1850)', French Studies 44(1), 36-54.         [ Links ]

Chassat, S.; Forlivesi, L. et Pottier, G. 2005 Eduquer et punir: La colonie agricole et penitentiaire de Mettray, 1939-1937. Rennes, Presses Universitaire de Rennes.         [ Links ]

de Certeau, M. 1988 The Writing of History, transl. Tom Conley. New York, Columbia.         [ Links ]

de Lurieu, G.; Romand, H. 1854 Sur les colonies agricoles de mendicants, jeunes detenus, orphelins et enfants trouves: Hollande-Suisse-Belgique-France. Paris, Librarie agricole de la maison rustique.         [ Links ]

de Tourdonnet, A. 1863 Essai sur l'education des enfants pauvres. Des colonies agricoles d'education. Paris, Brunei         [ Links ]

Dekker, J. 1985 Straffen, redden en opvoeden: Het ontstaan en de ontwik-keling van de residentiele heropvoeding in West Europa, 1814-1914, met bijzondere aandacht voor 'Nederlandsche Mettray'. Assen, Van Gorcum.         [ Links ]

Foucault, M. 1974 Discipline and Punish. New York, Vintage.         [ Links ]

Foucault, M. 1980 Questions of Geography' in Power / Knowledge, ed. Colin Gordon. New York, Pantheon.         [ Links ]

Foucault, M. 1984 'Nietzsche, Genealogy, History', in: The Foucault Reader, ed. Paul Rabinow. New York, Pantheon.         [ Links ]

Foucault, M. 1994 'Table Ronde du 20 mai 1978,' in: Dits et ecrits. Tome IV Paris, Gallimard.         [ Links ]

Ketelaar, E. 1997 The Archival Image: Collected Essays, Hilversum, Verloren, 1.         [ Links ]

Luste Boulbina, S. 2003 Alexis de Tocqueville, Sur l'Algerie. Paris, Falmmarion.         [ Links ]

Pitts, J. 2001 Alexis de Tocqueville: Writings on empire and slavery. Baltimore, Johns Hopkins Press.         [ Links ]

Redfield, P. 2000 Space in the Tropics: From convicts to rockets in French Guiana. Berkeley, University of California Press.         [ Links ]

Stoler, A.; McGranahan, C. 2007 "Refiguring Imperial Terrains," in Imperial Formations, eds. Ann Laura Stoler, Carole McGranahan and Peter Perdue. Santa Fe: School of Advanced Research.         [ Links ]

Stoler, A. 2009 Along the Archival Grain: Epistemic Anxieties and Colonial Common Sense. Princeton, N.J., Princeton University Press.         [ Links ]

Stoler, A. 2010 Conference The Political Life of Documents in The University of Cambridge and        [ Links ]

Stoler, A. 2011 Conference The Political Life of Documents in The Max Planck Institute for the History of Science. Berlin.         [ Links ]

Ward, K. 2008 Networks of Empire: Forced Migration in the Dutch East India Company. Cambridge: Cambridge University Press.         [ Links ]

Creative Commons License Todo el contenido de esta revista, excepto dónde está identificado, está bajo una Licencia Creative Commons